Cover Image
close this bookHandbook for Agrohydrology (NRI)
View the document(introduction...)
View the documentAcknowledgments
View the documentSummary
Open this folder and view contentsChapter 1: Introduction
Open this folder and view contentsChapter 2: Measurement of runoff
Open this folder and view contentsChapter 3: Erosion and sedimentation data
Open this folder and view contentsChapter 4: Rainfall and other meteorological data
Open this folder and view contentsChapter 5: Soils and soil moisture data
Open this folder and view contentsChapter 6: Catchment characteristics
Open this folder and view contentsChapter 7: Water harvesting and field structures
Open this folder and view contentsChapter 8: Data analysis

(introduction...)


S. Miller

NRI
NATURAL RESOURCES INSTITUTE
Overseas Development Administration

Acknowledgments

The role of the Overseas Development Administration (UK), in funding the production of this Handbook through the Natural Resources Institute is gratefully acknowledged. Particular thanks are owed to David Jackson of the NRI Land and Water Development Section's Agronomy and Cropping Systems Research Programme and Stephen Walker of NRI. The ODA-NRI / SADCC Land and Water Management Development Project based in Botswana (1987-92), provided much of the field data and practical experience on water harvesting research that are presented herein. ELK International Ltd gave permission for the reproduction of figures 2.42 4.1, 4.11 and 4.13. The United States Department of Agriculture gave permission for reproduction of its material.

© Crown copyright
1994

The Natural Resources Institute (NRI) is an internationally recognized centre of expertise on the natural resources sector in developing countries. It forms an integral part of the British Government's overseas aid programme. Its principal aim is to alleviate poverty and hardship in developing countries by increasing the productivity of their renewable natural resources. NRI's main fields of expertise are resource assessment and farming systems, integrated pest management, food sciences and crop utilization.

NRI carries out research and surveys; develops pilot-scale plant, machinery and processes; identifies, prepares, manages and executes projects; provides advice and training; and publishes scientific and development material.

Short extracts of material from this publication may be reproduced in any non-advertising, non-profit-making context provided that the source is acknowledged as follows:

Miller, J. (1994) Handbool for Agrohydrology. Chatham, UK: Natural Resources Institute.
Permission for commercial reproduction should, however, be sought from the Head, Publishing and Publicity Services, Natural Resources Institute, Central Avenue, Chatham Maritime, Kent ME4 4TB, United Kingdom.

Printed by Hobbs the Printers of Southampton

Price £30.00

No charge is made for single copies of this publication sent to governmental and educational establishments, research institutions and non-profit-making organizations working in countries eligible for British Government aid. Free copies cannot normally be addressed to individuals by name, but only under their official titles. When ordering please quote WR3.

Natural Resources Institute
ISBN: 0-85954-389-7

Summary

This handbook provides detailed information on the practical aspects of hydrological research in agriculture. It has been written not only for hydrologists, but also field scientists for whom hydrology lies outside their particular specialisation.

Theoretical methods for the estimation of flow peaks and volumes are evaluated. Techniques for the measurement of runoff and its associated meteorological variables are presented with details on equipment and circumstances of suitability, selection, manufacture and operation. Soil erosion and sedimentary processes are discussed in terms of both field data collection and the use of empirical formulae. Alternative field and laboratory methods of measuring soil moisture are described.

Water harvesting techniques are in discussed in the context of increasing catchment size, peak flows and runoff volumes, and the field data from research trials are given for each main category of water harvesting technique. The planning, design and construction of the field structures that are essential in water harvesting research and practice: bunds, ridges and waterways, are also discussed.

Methods of the analysis of hydro-meteorological data are described, illustrating both statistical and non-statistical techniques.

1.1 The role of hydrology in agriculture

Agrohydrology can be regarded as the study of hydrological processes and the collection of hydrological data, aimed at increasing the efficiency of crop production, largely by providing beneficial soil moisture conditions. However, the influences on the production of runoff and the ways that runoff affects the environment within which crops grow are very diverse and agrohydrological study, of necessity, also includes the collection of information on climate, soils, vegetation and topography. Rainfall amount and its spatial and temporal distributions determine the quantity of water that reaches the land's surface. Temperature and humidity, the type, amount and distribution of vegetation cover determine what proportion of this water re-evaporates. Vegetation, soil conditions and topography determine how much water infiltrates into the soil, how much runs off the land's surface and where it goes. It is the interaction of these complex processes and the volumes of runoff that these processes produce that form the core research of agrohydrology.

Hydrological practice has been developed to its greatest extent and sophistication in the provision of water resources from large catchments, usually for industrial and domestic consumption. Historically, hydrological practice has had a limited role to play in agriculture even where large-scale irrigation schemes have been undertaken, because civil or agricultural engineering expertise has usually taken a dominant place in such circumstances. But with the increasing interest in improving poorly developed and more marginal regions of agricultural activity, where large capital investment is uneconomic, a thorough understanding of the hydrological conditions that prevail has become essential. This understanding is particularly important where agriculture is a subsistence activity, or where water harvesting is proposed to improve agricultural production. Knowledge of the hydrological environment is necessary to determine whether or not opportunities to create optimal soil moisture conditions exist, and how these opportunities can be exploited.

Also, the understanding of the hydrological environment links together with other important environmental issues: the removal of natural vegetation, soil erosion, flooding, drought. The influence of hydrological conditions on farming practice and farming systems is substantial, and in the case of water harvesting, the availability, timing and volume of surface runoff will be critical to success or failure.

The actual techniques and methods that are used to collect information for agrohydrological study are, in general, very similar to those used in more orthodox hydrological field practice and the transfer of technology is not a problem, but there are some differences. These differences may be summarised as follows:

- the catchments from which runoff is measured are usually smaller.

- runoff peaks and volumes are also usually small. This often necessitates the modification of otherwise standard equipment.

- often, studies concentrate upon the particular conditions under which runoff is produced and particular conditions may even be imposed upon a catchment. The use of "natural" catchments is not common.

- it is often necessary to study many replicates of catchment types.

- particular conditions of climate (especially rainfall) and catchment may have only a very localised extent.

- a close connection with farming practice will be desired.

- historical hydrological (and sometimes other) data will be limited.

- methods of analysis of data are identical, but the shortness of records often imposes severe constraints.

The aim of this guide is to show which hydrological factors are important and how they can be measured, so that any opportunity for improving the water supply to crops can be taken. The core interest of the guide is runoff, the surface flow of water and the rainfall and catchment conditions that cause it.

1.2 Summary

This book is directed at agricultural projects whose staff do not have the specialised skills of the hydrologist and at hydrologists whose experience in the agricultural field is limited. It is aimed especially at those working in developing countries, where resources will be limited and where it is essential to put the right equipment in the right place, in the right way. Projects usually have a very short life-span compared to the time it takes to collect a comprehensive set of data and a season that does not yield information that can improve the quality of future decisions is, effectively, a season lost.

The Contents of this Handbook

A deliberate attempt has been made to bridge the gap that is so commonly found between textbooks and guides to field practice. It is often the case that textbooks discuss in detail the theory of hydrology, but give little or no explanation of how this theory is applied to hydrological work. The material may make sense in isolation, but it often cannot be translated into real field activity. For obvious reasons the authors of texts do not select limited data of dubious quality for illustrative examples, but this is exactly the kind of data that are found frequently when work is undertaken at the project level. Similarly, practical guides often leave the reader with no accurate background to the theoretical basis upon which their research is founded. When research moves away from its theoretical basis, to accommodate the realities of everyday life, it is important to know exactly where it stands and whether or not the links with sound theory have been stretched too far.

There are three main components to this handbook. These are chapters 2, 7 and 8 which cover Runoff and its measurement, Water Harvesting and Field Structures, and Data Analysis, respectively. The other chapters are important, but these three cover the fundamental aspects of agrohydrology. Equipment used for the collection of data is essential to the successful acquisition of agrohydrological knowledge. Its manufacture, installation and maintenance are covered in depth. A breakdown of the main topics of each chapter is given below.

Breakdown of Chapters 2 - 8

Chapter 2: Measurement of Runoff
- theoretical estimates to help in the selection of appropriate equipment
- hydrometrics/ runoff controls, both natural and artificial
- measurement of hydrological variables
- equipment descriptions
- equipment manufacture? installation and maintenance

Chapter 3: Sedimentation Data Collection
- soil erosion and methods of estimation
- total sediment and suspended sediment measurement
- equipment
- laboratory analysis of water and soil samples

Chapter 4: Rainfall and Meteorological Data Collection
- equipment descriptions for all major meteorological variables
- installation and maintenance
- siting and operation
- raingauge networks

Chapter 5: Soils and Soil Moisture
- soil classification
- soil textures
- methods of determining soil moisture
- infiltration
- equipment

Chapter 6: Catchment Characteristics
- natural vegetation
- catchment size, land slope, topography
- field orientation
- geology and other influences

Chapter 7: Water Harvesting and Structures
- types of on and off field systems
- results from research examples of these systems
- design criteria. channels and waterways
- practical aspects of laying out fields/ agricultural engineering

Chapter 8: Analysis of Data
- runoff data, non-statistical analysis
- statistical analysis, theoretical distributions of data
- rainfall and other meteorological data
- rainfall intensity
- rainfall/runoff relations
- evaporation and evapotranspiration

Bibliography

The list of books and papers given below has been limited to those which bear directly on the text of this handbook and most of them should not be difficult to obtain. The range of reference material on hydrology and agrohydrology is extremely comprehensive but it is recognised that some field workers may find difficulty in obtaining such material. Some of the references below are orthodox textbooks that deal largely with theory, while others are field manuals or research papers that report experience at first hand. It is hoped that this mixture of theory and practice will provide a good basis from which research can be undertaken? while at the same time allowing researchers to follow their own preference toward particular reference material.

A generalised computer program for the solution of the Penman equation for evapotranspiration. Chidley, T R E and Pike, J G, 1970.
Journal of Hydrology 10 (1970): 75-89

A rapid method of computing areal rainfall. Chidley, T R E and Keys, K M, 1970.
Journal of Hydrology 12 (1970): 15-24

Agrometeorological crop monitoring and forecasting' FAO Plant Production and Protection paper 17, 1979
FAO, Rome

Applications of remote sensing to hydrology, S T Miller, 1986
PhD Thesis, University of Aston, Birmingham, UK

Annual floods and the partial duration series. Langbein W, 1949
Transactions of the American Geophysical Union, 30: 878-881

Arid zone hydrology. FAO Irrigation and Drainage Paper 37, 1981
FAO Rome

Climate of Botswana, part II Elements of climate. Bhalotra, Y P R, 1984
Department of Meteorological Services, Gaborone, Botswana

Effect of slope and plant cover on runoff, soil loss and water use efficiency of natural veldt. Snyman, H A and Van Rensburg, H A, 1986. Journal Grassland Society of South Africa 3,4: 153-158

Field manual for research in agricultural hydrology, 1979
USDA Agriculture Handbook 224. Washington DC, USA

FAO Soils Bulletin no. 1(), Physical and chemical methods of soil and water analysis 1970.
FAO Rome

Flood studies report, 1975 Natural Environment Research Council, UK

Field directors handbook, Oxfam 1985
Oxford University Press, Oxford. UK

Handbook of applied hydrology, 1964. Ven Te Chow (Editor in chief) McGraw-Hill, New York

Hydraulics of runoff from developed surfaces. Izzard, C F, 1946 Proceedings of the High Resolution Board, 26: 129-150

Hydrology for engineers, 1985. R K Linsley, M A Kohler and J L H Paulhus McGraw-Hill, New York

Hydrology for soil and water conservation in the coastal regions of North Africa United States Department of Agriculture Soil Conservation Service, 1974. Washington DC, USA

Instructions and tables for computing potential evapotranspiration and the water balance. Thornthwaite, C W and Mather, J R, 1957. Drexel Institute of Technology, Publications in Climatology Vol X, No 3.

Irrigation principles and practices, 1962. O W Israelson and V E Hansen John Wiley and Sons, New York

Land husbandry manual, 1977 Ministry of Agriculture & Natural Resources, Lilongwe, Malawi

Land and water management project ODA/SACCAR. Annual Reports, 1989 - 1992 Natural Resources Institute, Chatham, UK

Land and water management project. Hydrology final report, Miller S T, 1994 Natural Resources Institute, Chatham, UK

Measurement and prediction of actual evaporation from sparse dryland crops, Wallace, J S, Gash, J H C, McNeil, D D, Sivakumar, M V K. 1986. Report on project 149. Institute of Hydrology, Wallingford, Oxfordshire, UK

Meteorological observers handbook 805, 1969
HMSO, UK

Microtopography and agriculture in semi-arid Botswana 1. Soil variability, Miller S T, Brinn, P J, Fry G J and Harris D, 1994. Agricultural Water Management (in press)

Prediction of variation in grassland growth in semi-arid induced grassland. Dye, P J, 1983. PhD Thesis, University of Witwatersrand, Republic of South Africa

Predicting rainfall erosion Losses, 1978
United States Department of Agriculture report 537

Principles of hydrology, Ward, R C. McGraw-Hill Ltd, 1975
London, UK

Probability and statistics in hydrology, V Yevjevich
Water Resources Publication, 1972
Fort Collins Colorado

Probability and statistics for engineers and scientists. Walpole, R E and Myers, R D Collier Macmillan, 1985, New York, USA

Rainfall-induced runoff computed for fallow fields, Hauser,V L, Hiler, E A, 1974. Soil and Water Division Paper 73:2520, ASAE.

Remote sensing and image interpretation, 1979. TM Lillesand and R W Kiefer Wiley and Sons, New York

Soil conservation, Elwell, H A
College Press, 1986. Harare, Zimbabwe

Soil water balance in the Sudano-Sahelian Zone. IAHS Publications, 1991 Institute Of Hydrology, Wallingford, Oxfordshire UK

Soil and Water Conservation Engineering. G O Schwarb, R K Frevert, T W Edminster, K K Barnes Wiley and Sons 1981, New York

Stochastic considerations in optimal design of a microcatchment layout of runoff water harvesting, Oron. G and Enthoven, G. 1987. Water Resources Research 23:7:1131-1138

The quantification of runoff and factors influencing its production, Miller, S T and Veenendaal, E M,1990. Proceedings of the Land and Water Management Research Programme Scientific Workshop, Gaborone, Botswana

Three years experience with an on-farm macro-catchment water harvesting system in Botswana Carter. D C and Miller ST 1991. Agricultural Water Management 19 (1991) 191 -203

Time domain reflectometry in soil science: theory, operation and use. Robinson, D, 1993. Institute of Hydrology, Wallingford, Oxfordshire, UK

User's handbook for the Institute of Hydrology's neutron probe system, 1981. Institute of Hydrology, Wallingford, Oxfordshire, UK

Vegetation management guidelines for increasing yields in a semi-arid region: an Arizona case study, Fogel M M, Report of the School of Renewable Natural Resources, University of Arizona, Tucson, USA

Water harvesting for plant production. Reij, C, Mulder, P, Begmann, L, 1988. World Bank Technical Paper 91. Washington DC, USA

Equipment Cost Lists

At the end of each chapter, there is a list of basic equipment and 1993 prices in US $. The cost of scientific equipment is often surprisingly high and although these prices may soon be out of date they provide a useful basis for early planning until current prices can be obtained, a process which can take weeks or even months. Below are the names and addresses of a number of UK manufacturers and suppliers of scientific equipment. This should be of use when researching equipment prices and availability.

Note that inclusion in this list does not confer an' recommendation by NRI

Company

Address

Equipment

Casella London Ltd

Regent House, Wolsley Road,

Meteorology


Kempston, Bedford MK42 7JY



England. (Fax: 0234 841490)


ELK International Ltd

Eastman Way, Hemel Hampstead

Hydrology


Herts. HP2 7HB, UK.

Soils


(Fax: UK + 0442 252474)

Agronomy



Laboratory


Mon Plaisir 25, Postbus 373 4879 AK

Meteorology


Etten-Leur, Nederland.



(Fax: 01608 33181)



Schonbergstrasse 47? D-7302



Ostfildern 4, Deutschland



(Fax: 0711 457 09 51)


Smail Sons & Co Ltd

Unit 1, St. Andrews road,

Hydrology


Glasgow G41 lPP. U

Groundwater


(Telex: 041 429 4429)


Vector Instruments

115 Marsh road, Rhyl, Clwyd

Wind monitoring


LL18 2AB. UK.



(Fax: UK + 0745 344206)





Delta T Devices Ltd

128 Low road, Burwell, Cambridge

Agronomy


CB5 OEJ. UK

Meteorology


(Fax: UK + 638 743155)

Loggers & Software

Valeport Ltd

Unit 7, Townstal Industrial Estate,

Hydrology


Dartmouth, Devon TQ6 9LX. UK



(Fax: UK + 0803 834320


Soil Instruments Ltd

Bell lane, Uckfield, East Sussex,

Soils and Geotechnical


TN22 1QL. UK



(Fax: UK + 0825 761740)


Didcot Instrument Co Ltd

Unit 14, Thames view Industrial Park,

Meteorology


Abingdon, Oxon, OX 3UJ. UK

Soils & Neutron Probe


(Fax: UK + 0235 522345)


Appendices to Chapters

For ease of reference and where appendices are appropriate, the material is placed at the end of each relevant chapter and the sequential page numbering of the chapter is continued. The page numbers of the appendices are listed in the Contents section.

Computer Models

A very short list of "off the shelf" models and database systems, that can be used for catchment and agricultural flow simulations is given below. There are great difficulties in providing such a list: very many research institutes, university departments and private companies have developed or are developing models based on some or all of the physical processes involved in rainfall/ surface flow/ soil moisture/ crop/ natural vegetation/ groundwater recharge, etc., to determine various water balances. Some of these models will be relevant only to specific research purposes, while others will be intended for general release, to be used in many different circumstances. To keep track of all of these developments is an impossible task. It is recommended therefore, that enquiries for details of models for field projects should be directed to the funding organisations of such projects. These organisations will be in a good position to contact research organisations and national institutes for details and may indeed be funding the development of water balance models themselves. Local organisations (Water Authorities, Ministries of Agriculture, etc.) may also be able to provide useful information on any models that have been developed or modified for use under local conditions. The following models and databases are presented because they are (mostly) in widespread use, and they have been used and tested for some years. The programs are not suitable for use with small runoff plot data, but are designed to be used with a p.c. Enquiries need only be directed to one organisation, the Institute of Hydrology, Wallingford, OX10 8BB, UK.

"HYDATA" is a hydrological database and integrated analysis system, best suited for catchment purposes. Currently it is used in 15 overseas countries. It stores station data (location, name etc.), stage and rating equation data, flows, rainfall and meteorological information. It has been developed to be compatible with the WHO's meteorological database, CLICOM via a transfer utility "HYCOM". Analysis gives comparison plots (eg double mass curves) or time series (eg hydrographs), flow duration and low flow information. Current price £5,000.

"HYFAP" is a frequency analysis and modelling package developed for use with extreme event information (eg annual maximum flows) for the prediction of magnitudes and return periods. Various distributions and fitting methods (see chapter 8) are available. Data can be transferred directly from HYDATA using an optional utility package "HYDOUT". Current price £495.

"HYRROM" is a relatively simple conceptual, deterministic rainfall/runoff model. Rainfall is routed through an interception and soil store, with evaporation and transpiration deducted. Runoff is given after losses to groundwater are accounted for, though groundwater contributions are added after a time delay. The model can be manually calibrated. It is compatible with HYDATA. Current price £825.

"Micro-LOW FLOWS" is a modelling program that incorporates the findings of the Natural Environment Research Council's Low Flow Study. Catchment characteristics are used to provide mean, 95 percentile and mean annual minimum flows; low flow frequency and flow duration curves. Current price £1100.

"Micro-FSR" is a software package for estimating design flood flows and probable maximum precipitation, using statistical modelling techniques (see chapter 8). Current price £995.

"SWIPS" is a soil moisture quality control, processing database for data from neutron probes, capacitance probes and tensiometers, and is run under Windows 3.1 operating software. Current price not available.

* For updates of packages and prices contact IOH. Bona fide research organisations/projects can purchase at large discounts.

1.3 Project planning and practical problems

This handbook assumes that work is being undertaken in developing countries and usually, though perhaps not always, will be implemented through a project with a finite lifetime. It is important to consider briefly, the manner in which projects are formulated and evaluated. Project staff should be aware of how and why projects have been devised and funded, and understand the work that has gone into developing the project proposals. Their own experience may be invaluable for future proposals for project development.

Projects have a finite life, but should seek to attain their goals and leave behind continuing benefits that come from the successful integration of new developments; technical, economic and perhaps social. Technical staff have important contributions to make in these areas, to both current activity and future planning, but it would be naive to believe that technical improvement and social benefits are the only aims of funding agencies. Policy and administrative considerations are often paramount and it is essential that technical assessments should be thorough and realistic.

Proposals and Planning for Projects

Project proposals are the first tangible evidence of possible future activity. They collect together ideas generated by the previous work and experience of individuals and organisations and will pass through many different stages of development before final acceptance or rejection. Because of this, project proposals may develop over a long period of time and it is important that their relevance is continually assessed. It is also advantageous, and in the case of most projects essential, that in addition to technical and logistical enquiry, the specialist skills of sociologists and economists be applied at the earliest stage of any proposals, to define the possible consequences of implementation. It is also important to remember that each funding agency will have its own individual character and particular spheres of interest and experience.

A list of basic conditions that project proposals should fulfil is given below. These address the structural and material content of proposals and do not consider any of the important social issues that can undermine the success of any technically feasible project. Project proposals should take into account:

Clearly defined aims and objectives

These should explain precisely the long and short term goals that the project seeks. They should be agreed upon and documented prior to any implementation. In some cases they may be limited to the stages of technical implementation and their results, in other cases it may be necessary to include the socio-economic effects that are expected and the development of the activities of the project, in the light of these effects.

Institutional framework

This will identify the interested parties and clarify the position that the project occupies between them. The responsibilities of the organisations involved and the financial, staffing and logistical support that a project receives, should be explained in detail.

Lines of communication between Project, Donor, Recipient Organisation and Participants

These are often complex, but vital to the success of a project. Misunderstandings may lead to a lack of amicable cooperation. They provide essential conduits for reporting, review procedures and information, and keep everyone involved aware of the progress of work and need for revision.

Reporting Procedures

Strong lines of communication are useless without defined reporting procedures. If these are well established within project proposals, it is not easy to neglect them, even when a project is running and day to day tasks have a more immediate attraction.

Evaluations and Reviews, both internal and external

Project proposals (and activities) are not immutable and a flexible approach is essential. Conditions change continually and it is not possible to design, implement and complete a project without considering improvement as experience is gained. Review procedures are important to ensure that sensible assessments of progress are made and to initiate discussions on alternative courses that may be followed. Both internal and external reviews should be timetabled, with at least one major review allocated to a project at the most suitable stage of development. The difficulty here is to balance the timing; it should be neither too late to be of use, nor be too premature to review sufficient material. Short term reviews, perhaps annual internal reviews, can be made available and help in selecting appropriate timing, but much will depend on the nature of the project and its duration. The nature of the review bodies should be stated, as should to whom they will report and their constituent members.

Funding

Funding may sometimes be contentious, sometimes easily agreed upon. In terms of project success, it is often the distribution of funding that causes problems rather than (within reason) the amount. It is usually most convenient for funding agencies to disburse an evenly spread level of funding. This allows consistent administrative procedures and easier future budget planning, but it is rarely appropriate for the efficient running of projects. Project capital investment expenditure is relatively high at first, gradually decreasing over the project life. Conversely, funds for employing local staff, equipment repair, vehicle maintenance and fieldwork will become greater as the project grows. Funds for training tend to peak during the mid and final term when staff have sufficient experience to require enhanced expertise and suitable courses have been identified. Problems can arise not only from the varied levels of funding needed in each case, but also because each may be obtained from different areas of fiscal responsibility within the funding organisation. It is clearly to everyone's advantage to assess realistic levels of funding in detail, relate them to the stages of the project's life and identify remedial procedures, should these be necessary. The responsibilities of all organisations concerned with supporting the project, should be clearly stated.

Long-term Obligations

The project proposals should place the role of the project clearly within the framework of past activity and where long term obligations are planned, agreements on these should not be postponed nor over looked. The manner in which the results of the project fit into the social and institutional framework of the host country and whether or not they can be maintained, should be assessed. Projects usually become self funding over a timetabled period, but it is easy to overestimate local sources of support where attention is distracted from problems of future funding by the overall appeal of the project.

Format

Project proposals should be clear and concise, but comprehensive. Different funding agencies use different formats of presentation and it is sensible to adopt these formats as early as possible. The process of preparation is sufficiently time consuming without additional unnecessary delays, especially where projects have a defined season of implementation.

Basic Questions

There are several basic questions that have to be asked when a project proposal is being developed. The list below is not necessarily exhaustive.

a. What are the genuine needs to be served ?
b. Can they and have they been identified ?
c. What objectives can the project actually achieve ?
d. Are the technologies appropriate and economically feasible?
e. What constraints, technical, social and economic are to be overcome ?
f. What are the long term implications ?

When these questions are answered it is important to present the basis of each answer and provide a summary of the research material. For example the answer to questions 'a' and 'b' may be based on extensive questionnaires; government economic or agricultural statistics; discussions with research organisations or workers already in the field. Questions 'c', 'd' and 'e' demand recourse to previous experience from other projects, noting advances made and failures due to identifiable causes. Sensible answers to question 'f' show that the long term development of a project has been well thought through. They indicate a familiarity with the host country and recipients and an understanding of what can and cannot reasonably be expected. Negative answers do not necessarily prove that a project is totally unsuitable, the long term implications may be simply too optimistic. Modification of the proposals may overcome any long term difficulties that come to light.

Background Information

The type of information needed will obviously be determined by the kind of project that is proposed. Consideration should be given the following:

a. National, regional and socio-economic information.

In general, bi-partite projects will have fewer problems in obtaining this information than ones involving various groups. However, such information is not always easy to get; it may even not be available, it may be seriously outdated and governments are sometimes reluctant to give it. Previous reviews and surveys provide a good indication of the width and availability of researchable material.

b. Previous, current and future projects.

These can often be a very valuable source of information, in addition to background research, technical data and research conclusions may be available. Learning from previous mistakes is an opportunity to be taken. Areas of cooperation can be explored, sometimes to the general benefit of all, but these areas should be clearly decided upon and defined.

c. Host government policies.

It is essential that projects be concordant to the policies of the host government. Any that are not are bound to fail and any long terms benefits will be lost. The opportunity to gain familiarity with organisations and individuals within government should be taken to the full.

Reporting and Evaluation

Reporting on project achievements should be undertaken in a systematic manner. Agreed arrangements should be made, which specify the details of reporting methods:

- From whom / to whom
- How often and at what length
- Whether technical or financial or both

Evaluations should include:

- Against what objectives any achievements should be assessed
- Details of the use of funds
- What dissemination of information has been undertaken
- Technical evaluations

Project Support

Project support can take many forms and should come from both inside and outside government. It may be through the cooperation with complementary projects which saves costs and provides a wider range of inputs. It may take the form of organised seminars and discussions which give a wider audience to the aims and achievements of project work. It may include the dissemination of information through government departments. The involvement of local institutions and experts is beneficial to both project and participants. Where a considerable financial burden may fall on participants (for example travel fares and housing of conference delegates), it is important that the responsibility for these costs to be identified and budgeted for.

Training

One of the greatest benefits a project can leave behind is well trained staff There are however several conditions to this statement that must be given serious thought:

-The correct type of staff must selected and the level of suitable training must be clearly identified. This may be vocational or institutional. If the latter, the correct course must be sought and places obtained. A careful timetable is necessary.

- Host government obligations. Funding may be the responsibility of the project or government, or both. It may not be to the advantage of host governments to pay for training for local project staff who later return expecting an enhanced position and better career prospects. The details of such matters should be settled well before training is arranged.

- The training must be appropriate for re-deployment when the project is finished. Any training must be in the long term interests of the host government, especially where counterpart staff are concerned.

- Project obligations. It should be ascertained that the project budget provides funding for training, as some may not.

- International Centres. Some international centres will sponsor candidates for training, but early enquiries should be made, because such sponsorship is eagerly sought.

Practical Problems

The collection of good quality field data is frequently very difficult and many problems will be peculiar to individual projects. However, the two most general but almost universal problems that must be faced by research and field staff are:

- Collecting adequate information during the limited lifetime of a project
- Balancing resources between the amount and quality of data that can be collected.

These are problems that face almost every project as a whole and a cooperative effort is needed from all staff to overcome them, but below is a discussion of common difficulties that will probably be encountered by individual field workers in the attempt to obtain good data.

Limited Project resources: Projects in developing countries often succumb to the temptation of over-stretching their resources. In areas where little data is available and there is much to collect, the relative merits of few sites with intensive data collection and many sites collecting fewer data, must be carefully weighed.

Difficult access to sites: This is especially true of wet seasons when bad roads often become impassable. In such situations sites cannot be visited, equipment repaired, conditions observed nor site staff consulted. Sites should be carefully selected so as not to impose an intolerable burden on the data-collection routine.

Restrictions on transport to cover sufficient sites: The availability of transport and the means by which it is kept in good repair often pose some of the most serious logistic/resource problems that projects face in developing countries. Roads are usually bad, vehicle maintenance standards low and shortages of spare parts common. There is an understandable temptation for vehicles to be put to non-project uses in countries with rudimentary public transport facilities. This situation is not helped by the status which is conferred upon drivers, where the private ownership of vehicles is a great luxury.

Hostile physical environment: Equipment has a hard life. Rough and inexperienced handling, transportation in difficult conditions often leads to early breakdown. High humidity and large ranges of both seasonal and diurnal temperatures frequently take their toll. Given the difficulties of repair and replacement, equipment should be treated carefully and be well maintained. An adequate provision of spare parts should be made.

Inappropriate equipment: Very often equipment must be ordered from overseas. It is essential that the correct equipment be selected in the first instance. Replacement may be impossible or may take many months.

Inexperienced staff: It must be recognised that educational and training levels in developing countries are commonly lower than those in developed countries. This puts a great responsibility on professional project members to give as wide a range of relevant training as possible to technicians and field staff. Ideally, initial planning should place training as a core component of project activities, but this is by no means always so and in some instances research budgets actually preclude the use of project funds for training purposes. Initial project proposals should seriously consider the role of training in the future of any project. There is little more dispiriting to all concerned than a project which leaves no continuing activity behind at the end of its life.

Under-motivated staff: It may seem to project professionals working abroad, that local staff do not always give what they can toward the success of a project. When this is true, it is usually for very good reasons. Local staff almost always have very poor pay and conditions compared to expatriates. Often they are seconded from other areas of activity, sometimes on a temporary basis, with little hope of an enhanced career or improved personal prospects. They are often not trained nor aware of the opportunities that a project may offer. In developing countries, as in developed countries, technical personnel are undervalued in general; an administrative post in a ministry is far more likely to lead to promotion than supervising a field team. It is essential that local technical and field staff are shown that the success of any project lies very much in their hands. Without reliable, accurate information collected at the correct time, projects in the area of agrohydrology are little more than an exercise in redistributing a given amount of money.

These difficulties can never be totally overcome, but the careful selection, siting, installation and maintenance of equipment allied with good staff training can keep damage and disappointment to a minimum.

(introduction...)

Background

"Runoff" is the term usually employed to distinguish the flow of water running off the land's surface during and shortly after rainfall, from the longer term flow of groundwater to rivers. This distinction is achieved by the analysis of flow data from perennial streams and rivers in humid climates, but in many agrohydrological and water harvesting situations, groundwater contributions are not present and all flow is runoff. This will almost certainly be the case in arid and semi-arid climates.

The collection of runoff data is very site and purpose-specific, both in terms of the kind of information that will be required and the manner in which it is best obtained. Runoff events are less frequent than rainfall events, in all climates. In areas of low rainfall, where agrohydrological and water harvesting projects are usually located, the number of runoff events may be fewer than ten per season. This compounds the unhelpful fact that problems with equipment and installations are not encountered until runoff occurs. If these problems are not rectified quickly, a large proportion of a season's data can be easily lost. Moreover, both equipment and experimental designs have to cope with a large range of runoff volumes and peak flows, so careful planning and a quick response to unexpected situations are very important in the success of collecting comprehensive, accurate information.

In planning runoff experiments it is important to have a clear idea of the object of research. For example:

1. In some regions, there may be no existing hydrological data. In this case it may be most suitable to spread project resources thinly and to collect information on as many different hydrological factors, at as many sites as possible. The replication of experiments will be limited and the development of hydrological models using these data will probably not be possible. A careful selection of off-station sites will be needed; adequate field staff and vehicles should be available and routine visiting schedules should be drawn. Site observers may be necessary and a greater proportion of automatic equipment will be needed. But a great number of varied circumstances will be documented and the data should be suitable for input into existing models. The information will be especially useful for projects that are following on and the research may provide insight into areas of hydrological behaviour not previously observed.

2. In other instances a much narrower focus may be desired. The development of hydrological models may be a priority and a large number of replicated experiments will be needed so that the data are amenable to statistical verification. It may be important to work cooperatively with existing projects, to extend the range of project activities and it may not be possible to undertake research in many environments.

Some projects will have a more practical applications bias and research may be combined with farmer participation and the implementation of farming systems. The overall objectives of any project will determine the financial commitment that is placed on the measurement of runoff, but it is assumed for the purpose of this guide that in the field of agricultural hydrology, runoff measurement is of primary importance.

2.1 Estimates of runoff

Estimates of runoff are made for two reasons:

1. They are essential guides in the decision of which system of runoff measurement to use, either volumetric or continuous. After this decision has been made, these estimates must be used to determine the size and capacity (peak flow and flow volume) of the equipment.

2. If the measurement of runoff is not to be undertaken, then calculations must be made to estimate the design specifications of bunds, channels etc. that are to be used in the mechanics of water harvesting and field layouts.

2.1.1 Estimates Based on Previous Data

A project may be fortunate in succeeding previous work that has already undertaken runoff measurement. These data can provide a good starting point for the selection of appropriate equipment. For example "Will a 100 litre collection tank be large enough to collect runoff from a 50 m² plot when previous data show that a 50 litre tank was large enough for a 20 m² plot?". The answer in this case would probably be "Yes", but caution must be exercised. Comparisons of catchment size, land use, slope and rainfall must be made to ensure that these data can be used for the purpose of installation design. Generally, if a project is succeeding another it is likely that most of the physical conditions under which the new project will operate will be similar and catchment size may be the primary concern in estimating flow peaks and volumes. Catchment size is by no means the main influence on runoff amount, however.

Consider Table 2.1 below, data from a strip tillage experiment undertaken in semi-arid SE Botswana. All physical conditions for the plots, including rainfall for each season, were identical. In both seasons the smallest plots produced most runoff (both seasonal and event by event) and the conclusion may be drawn that catchment size exerts an influence on runoff, though this influence appears to be relatively small and is unlikely to cause serious problems in estimating probable runoff proportions, volumes or peak flows, by a simple ratio of previous to proposed catchments areas. The complete set of rainfal/runoff data relating to these plots is given in chapter 7, section 7.1.

Other research has shown that for small plots, runoff proportions (percentages, sometimes called "runoff efficiencies") tend to be larger for small catchments until the distance of flow is about 80 m. Thereafter it is assumed that runoff proportions remain the same for catchments with the same conditions.


Table 2.1: Runoff Proportion (%) versus Catchment Size (average of 2 replicates)

Other factors can be much more influential than catchment size. In the example above, the variation in runoff from catchments of the same size due to differences in rainfall between the seasons was as great as that due to plot size. It must be remembered that data from previous work may have been collected during historical periods with greater or lesser rainfall amounts and/or intensities than those of a new project, even though the location of the data collection site is the same. Variations in other catchment physical conditions, such as vegetation and crop cover, the level of which are determined by yet other environmental circumstances (especially rainfall and human activity), may also be important. The data in Table 2.2 shows runoff from the same-sized plots at the same site for three different rainfall seasons, each plot has a different kind and density of vegetation cover. The data are from a rangeland area of SE Botswana.


Table 2.2: Variation of Seasonal Runoff (%) with Vegetation Cover

Note that not only does vegetation cover type and extent determine runoff, but once again the difference in seasonal rainfall exerts a strong influence. In this case the season with most rainfall produced least runoff from all plots. This kind of comparison shows the influence of different, variable conditions on runoff.

The difficulties in estimating future runoff from past data are best overcome by a statistical analysis of the information: changes in catchment condition or rainfall pattern which affect runoff amount will be present in the data and will be accounted for.

The main design criterion of hydrological equipment is whether or not this equipment will cope with the largest individual runoff event in any given number of seasons. In experimental and equipment design, it is the balance struck between the cost of the over-design of equipment and the possibility of equipment failure during large storms, that is particularly important. A probability must be assigned to the occurrence of the particular design flow or peak that is selected. This can be done by several statistical methods the simplest of which is the annual maximum series, outlined below:

a. Theoretical Distributions

Runoff data are often matched to statistical distributions with known forms. Extrapolation can be made relatively simple where a good adherence to a statistical distribution can be found, but hydrological data may not conform, or different distributions may be more suitable in different geographical regions.

b. Partial Duration Series Methods

These methods do not evaluate the bulk of the data, but use a number of flows for each season that are greater than a selected runoff threshold. The pattern of these values is determined and linked to a statistical distribution from which flows of a specified return period can be extrapolated. The use of these methods is suitable when the number of seasons for which data have been collected is small, perhaps only 10 or less.

Annual Maximum Series

This method, which is a particular kind of partial duration series, selects the largest event of each year or season, tabulates them in order of magnitude and from this list derives a flow peak with a probability of occurrence and return period. To extrapolate for large events, the data may be plotted as in Figure 2.1 below on log-normal probability paper. This is a very simple and straightforward method, its main limitation is that it requires a large number of years' data to be useful, as it selects only the greatest flow from any season.


Figure 2.1: Annual Maximum Series, Runoff versus Probability

These methods of estimation of design flows are discussed in detail in Chapter 8, Data Analysis.

2.1.2 Theoretical Estimates of Flow

A great deal of research has been undertaken to develop hydrological models that can predict runoff peak flows and volumes. The majority, however, are not suited to general use. Sometimes they are too complex but most frequently they are limited by the geographical localities and hydrological conditions within which the data were collected. Many models are regression models and their value is difficult to assess outside their own particular circumstances.

Presented here are five models that can be used to predict peak flows and three models that are suitable to estimate runoff volumes. They are suitable for use with a wide range of catchment sizes and conditions. These methods of estimation have certain drawbacks: they can be relatively inaccurate because they make simplifying assumptions. They demand the availability of some primary data such as catchment physical characteristics and rainfall. However, they have been used for some time in a variety of environments with success and are based on measurements from a great number of catchments, with a wide range of physical characteristics.

2.1.2.1 Peak Flows

Peak flows determine the design specifications of structures such as bunds, channels, bridges and dams. Peak flows also determine the capacity of the control sections of flow-through measurement systems and the collection pipes and transfer conduits of volumetric collection vessels. Some estimate of peak flows must be made before the design of these systems can be completed.

Design peak flows are linked to particular return periods, such as the maximum flow in 5, 10, 25, etc. years and design specifications are a balance between economic cost and the prevention of failure of the structure. Where no serious damage will result, for example on field bunds, a low return period ( say 5 or 10 years ) can be used. The 10 year return period is commonly used for agricultural purposes . Where serious damage or the loss of life is involved, then designs for large return periods, perhaps 50 or 100 years, are necessary. The return period most appropriate to the objectives of the project should be decided upon.

a. Rational Method

The Rational Method which estimates peak flows, is a simplified representation of the complicated process whereby rainfall amount and intensity, catchment conditions and size as well as human activity, determine runoff amount, but it is suitable where the consequences of the failure of structures are limited. The method is usually restricted to small watersheds of less than 800 ha and is based on the rainfall/runoff assumptions of the hydrograph below.


Figure 2.2: Hydrographic Basis of the Rational Method

The equation to calculate peak flows is:

q = 0.0028 CirA where (2.1 )

q = peak flow in m³ s-1
C = the runoff coefficient
ir = maximum rainfall intensity in mm h-1 for the desired return period and the "time of concentration" of the catchment, Tc.
A = area of the watershed in hectares (1 ha = 10,000 m²)

The rainfall intensity is assumed to be uniform for the period and over the whole catchment for a time at least as great as the time of concentration of runoff, (Tc).

Values of Coefficient C

The value of C is the ratio of the peak runoff rate to the rainfall intensity and is dimensionless. It represents the proportion of rainfall that becomes runoff and is determined, to a large extent. by catchment conditions. Work by the US Soil Conservation Service has enabled the influence of many of these conditions to be expressed in various values of C. Examples of these C values are given in Tables 2. 3 for the USA (temperate region' 700-1000 mm average annual rainfall) and 2. 4 below for Malawi in central southern Africa (sub-tropical region' average annual rainfall variable. from < 400 mm to > 1000 mm). Different hydrological conditions according to soil groups are accounted for.


Table 2.3: Coefficient C values for USA

Values of coefficient 'C' for Malawi are given in Table 2.4:


Table 2.4: Coefficient C values (Malawi)

Rainfall intensity, ir

The rainfall intensity value used in the Rational Method is selected according to the desired return period for the design of the structure under study. The duration of the rainfall intensity is, for the purpose of the Method, said to be equal to the time of concentration of the runoff, Tc.

A graph or set of graphs can be drawn, to determine the maximum rainfall intensity for a particular return period and a particular rainfall duration (equal to Tc for the purposes of the Method). Such graphs demand the availability of many years of data, as they represent the line of best fit through a group of data points drawn from a wide range of rainfalls and their intensity measurements. Extensive records are especially necessary for long duration-intensity periods, which are not experienced frequently. Obviously, the climates of geographical regions will vary and even local differences can be great where a country shows a marked variety of topographic form. Areas of uniform rainfall characteristics should be provided with unique sets of rainfall intensity graphs. Figure 2.3 shows the manner in which these graphs are drawn.

A simple alternative way to calculate the return period of the maximum rainfall intensity for a specific duration, where data are too sparse to plot graphical relations, is shown below in Table 2.5 using 10 years' hypothetical example data. Where extrapolation is concerned, it should be remembered that the accuracy of estimation is related to the quantity of available data and the length of record. Note that in Figure 2.3, the lines defining Tc/Intensity relations are lines of best fit obtained from many storm data.


Figure 2.3: Example Graphs of Return Period, Intensity and Duration (which = Tc)

Example return periods used widely for different structures are: Field structures, 5-10 years; Gully control and Small farm dams, 20 years; Large farm dams, 50 years.

List the data as follows (duration in mins, intensity in mm h-1, m is order number of the item in the array). The rainfall intensity is the maximum intensity recorded that season or year, for the particular duration. The return period, in descending order of magnitude, of the rainfall intensity in years - (n+1)/m., where n is the number of years of record. Note that in the example Table 2.5 below, the exact values for the 5 and 10 year returns must be interpolated from the table and the values are given in bold type. Although the relation between intensity and duration is in fact curvilinear, linear interpolation does not lead to important inaccuracies. Making the time steps between the durations smaller, increases accuracy.


Table 2.5: Annual Maximum Series (Hypothetical Example Data.)

Intensities for the same return period increase with shorter duration (and Tc). It is also clear from the example above, that long records of data are necessary to obtain rainfall intensity values associated with long return periods as well as long durations and this may be a limiting factor with work in developing countries where records are frequently short. Great care must be exercised in using data that are imported from other regions, if local data are not available.

Time of concentration, Tc

The time of concentration (Tc) is the time by which water from most distant parts of the catchment has reached the outlet. The following formula has been developed to estimate Tc., with example values given in Table 2.6.

Tc = 0.0195 L0.77 S-0.385 where (2.2)
Tc is in minutes
L the maximum length of the catchment in m, and S = slope of the catchment in m m-1 over the total length L


Table 2.6: Values of Tc using Formula 2.2

The time of concentration, when calculated from equation 2.2 or obtained from Table 2.6, can be used to obtain the desired maximum rainfall intensity, depending on return period.


Figure 2.4: Time of Concentration, Tc, for Catchment Areas 0 - 36 Hectares

Equation 2.2 is not universally accepted and alternatively, the time of concentration can be found by dividing the measured length of flow by the estimated flow velocity. Manning's formula can be used to estimate flow velocities, although the estimation of flow velocity using Manning's formula can be a complex matter for large catchments where changes in channel form, size, slope and roughness can vary greatly and where the evaluation of these characteristics may be difficult. Figures 2.4 (above) and 2.5 (below) give values of Tc for a range of catchment areas, slope categories and qualities of protection. In all cases, it is important to calculate runoff peaks for the catchment conditions most likely to produce them, so that maximum peak flows are estimated: for example before cultivated land has been ploughed and before dense natural vegetation has regrown on non cultivated areas.


Figure 2.5: Time of Concentration, Tc, for Catchment Areas 40 - 200 Hectares
Source: Land Husbandry Manual, Ministry of Agriculture and Natural Resources, Malawi

Worked Example

A catchment of 15 ha is composed of 5 ha of permanent pasture (Soil Group B) and 10 ha of row crop in poor condition (Soil Group C). What peak flow is to be expected from a 1 in 5 year storm? The maximum flow length is 610 m, with a gradient of 2%.

From Table 2.6 or equation 2.2, Tc = 12 minutes

From Table 2.5 (hypothetical illustration), Rainfall intensity 73.0 mm h-1

Runoff coefficient C for permanent pasture (Group B, 5 ha ) = 0.14

Runoff coefficient C for poor row crop (Group C, 10 ha) = 0.71 therefore weighted value of C for whole water shed - 0.52 substituting in equation 2.1:

q = 0.0028 × 0.52 × 73.0 × 15 = 1.6 m3 s-1

b. Cook's Method

Developed by the USCS, this method essentially provides a simpler and more generalised, but similar approach to the estimation of peak flows to the Rational Method. Catchment size and conditions are accounted for. Table 2.7 gives catchment condition details.


Table 2.7: Values () for Catchment Conditions Cook's method

Catchment conditions are assessed and the numerical values assigned to each are added together. For example, if conditions are those in the right column of Table 2.7, a total value of 25 would be found and peak flows could be expected to be low, the exact size depending on catchment area. The conditions of a particular catchment will probably be found to be listed in different columns, but the relief condition is most heavily weighted and, in general, the four columns list conditions that describe "type" catchments. It was found generally that for African conditions, surface storage had little effect and a different set of values for catchment conditions were determined, as presented in Table 2.8. Soil type and drainage conditions were found to be especially important.


Table 2.8: Catchment Condition Values for African Conditions

When a total of catchment condition values is made, the peak flow is estimated using Table 2.9, below.


Table 2.9: Peaks Flows (m³ s-1) According to Catchment Condition Total Values and Area Using 10 Year Probability High Intensities for Tropical Storms

c. TRRL (UK Transport and Road Research Laboratory) Model

Work in East Africa, by the UK Transport and Road Research Laboratory has led to a model designed to overcome two serious problems associated with data in many developing countries: that rainfall/runoff correlations can only be developed using large amounts of data and that extremes in the data are rare. The US SCS method was found not to give acceptable results for East African conditions.

The concept of a "contributing area" (CA) is used to avoid the use of a uniform coefficient throughout the catchment. Early rain fills the initial retention (Y) and runoff et this stage is zero. A lag time (K) was incorporated to account for routing on larger catchments. Total Runoff Volume was found to be defined by:

Q= (P-Y) CA · A · 103 (m³ s-1) where (2.3)

P = storm rainfall (mm) during time period equal to base time of the hydrograph.
Y = initial retention (mm)
CA = contributing area coefficient
A = catchment area (km²)

The average flow QM is given by:

QM = 0.93 · Q/ 3600 · TB where (2.4)

TB is the hydrograph base time (hours)

Initial Retention (Y)
A value of 5 mm for Y was found to be appropriate for arid and semi-arid conditions.
A value of 0 mm for Y was found to be appropriate for wet zone areas.

Contributing Area (CA)

Soil type, slope, land use and catchment wetness were found to be the most influential factors in determining catchment contributing area. The design value is of the form:

CA = CS CW CL where (2.5)

CS = a standard value of contributing area coefficient for grassed catchment at field capacity.
CW = catchment wetness factor
CL = land use factor

Lag Time (K)

Lag time was found only to have a relation with vegetation cover.

Base time of the hydrograph (TB)

Simulation studies showed that TB could be found from the equation:

TB = Tp + 2.3 K + TA where (2.6)
TA = 0.028L / QM0.25 S0.5 where (2.7)

L = main stream length (km)
QM = average flow during base time (m³ s-1)
S = average mainstream slope
K = lag time
Tp = rainfall time

The value of QM can be estimated, through a trial and error iteration of equation 2.6, with TA initially being zero. Below are the tables necessary to estimate the various runoff factors.


Table 2.10: Standard Contributing Area Coefficients, CS (wet zone areas, short grass cover)


Table 2.11: Catchment Wetness Factor, CW

The table (2.13) gives rainfall time (Tp) for East African 10 year storms as a guide. The values for the localities under study can be obtained from local data if available.


Table 2.12 Land use factors, CL Catchment Lag times, K


Table 2.13 Rainfall time, Tp, for East African 10 year storms

Procedural steps for calculation

- Measure catchment area, land slope and channel slope.

- Establish catchment type and from Table 2.12 Lag time K.

- Establish soil type and using land slope, estimate standard contributing area coefficient, Cs. from Table 2.10.

- Establish antecedent rainfall zone and catchment wetness factor, CW, from Table 2.11.

- Use Table 2.12 to estimate land use factor, CL.

- Calculate contributing area coefficient by: CA = CS · CW · CL

- Find initial retention Y (0 or 5 mm).

- Using Table 2.13 or local data find rainfall time, Tp.

- Calculate design storm rainfall to be allowed for during time interval TB hours (P mm).

- Runoff volume is given by:
Q = CA · (P - Y) · A · 103 (m³).

- Average flow is given by:
QM=0.93 Q/3600 · TB

- Recalculate base time using
TB = Tp + 2.3 K + TA where TA = 0.028 L/ QM0.25 S0.5

- Repeat steps 6 to 9, until QM is within 5% of previous estimate.

- Design peak flow, QP is given by:
QP = F · QM where the peak flood factor, F is 2.8 when K < 0.5 hour and 2.3 when K is > 1.0 hour.

Worked Example
What is the 10 year peak flow of a catchment with the following details? Area 5 km²; Land slope 3%; Channel slope 1%; Channel length 2 km; Soils with slightly impeded drainage; Good pasture. 10 year daily point rainfall of 80 mm.

From Table 2.12, Lag time K

= 1.5 hrs

From Table 2.10 Standard contributing area coefficient CS

= 0.38

From Table 2.11, Catchment wetness factor CW

= 0.5 (dry zone ephemeral)

From Table 2.12, Land use factor CL

= 1.0

Therefore. design CA = 0.38 · 0.5 · 1.0

= 0.19

Initial retention Y

= 0 mm

From Table 2.13, TP

= 0.75 hrs

Using equation 2.6 with TA = 0


TB = 0 75 +2.3 .1.5

= 2.59 hrs

Rainfall during base time is given by

RTB = TB/24 (2.33/ TB + 0.33 )n · R10/24

where R10/24 = 10 year daily rainfall
and n = 0.96 (Table 2.13)
Therefore R2.59 = (2.59/24 · 24.33/ 2.92)0.96 · 80 = 72.2 mm

An areal reduction factor is used to take account of the fact that rainfall depths are smaller over catchment areas than they are at spot measurement points.

The Areal Reduction Factor (ARF) was found to be = 1 - (0.04 · TB-0.33·A0.55)
which for the value of A and TB = 0.93, thus:

Average rainfall P

= 72.2 · 0.93



= 67.5 mm.


The Runoff volume Q = CA (P - Y ) · A · 103

= 0.19 (67.5 - 0) · 5.103

= 64.13 × 10.3 m³.

QM = 0.93 · Q/3600 · TB

= 6.40 m³ s-1


First iteration of TA = 0.028 L/ QM0.25 S0.5

= 0.04 hrs


The value of TA (very small) indicates that no re-calculations of TB, the Rainfall time and QM are necessary.

Therefore the design flood is:

QP = F · QP where the flood factor F is 2.3 (as K is > 1 hour).
Therefore the Peak Bow QP - 2.3 · 6.4 = 14.72 m³ s-1

d. US Soil Conservation Service Method

This method is founded on the rainfa/runoff relation for the triangular hydrograph illustrated below in Figure 2.6. It is important to note that the method is used to calculate the peak flow of a known runoff event volume or to calculate the peak flow for an expected or desired runoff event volume. A specific discharge must be designed for. Knowledge of rainfall intensity is not needed. Peak flow is defined by:

q = 0.0021QA/Tp where (2.8)

q = runoff rate (peak flow) in m³ s-1
Q = runoff volume in mm depth (the area under the hydrograph)
A = area of water shed in ha
Tp = time to peak in hours, defined by:

Tp = D/2 + TL where (2.9)

D = duration of excess rainfall
TL = tune of lag, which is an approximation of the mean travel time and can be obtained from the nomograph below, Figure 2.7. Alternatively, Time of lag = 0.6 × Time of concentration which is the longest travel time of the runoff (not the time to peak as in the Rational Method).


Figure 2.6: The US SCS Triangular Hydrograph


Figure 2.7: Lag Time and Time of Concentration, US SCS Method

Source: US SCS Hydrology, National Engineering Handbook, 1972

Worked Example Determine the peak flow from a 10 ha catchment with a 0.5 hour storm that produced a runoff volume of 7 mm . Time of lag is 0.1 hours.

Substituting in equation 2.9,
Time to peak, Tp = 0.5/2 + 0.2 · 0.35 hours
Peak Bow, q = 0.0021 × 7 × 10 / 0.35 = 0.42 m³ s-1 or 420 l s-1

e. Izzard's Method

The previous techniques have been developed to estimate runoff rates from catchments ranging in size from a few hectares to several hundred hectares. However, agrohydrological experiments frequently make use of very small runoff plots only tens of square metres in area. This is for two reasons. First, they are easy to replicate and many such plots can be placed in a small area to study a range of catchment conditions. Second, they can be used conveniently to look at interventions that work on a small scale and which are intended to be installed within the boundaries of individual fields. In these circumstances, a method developed to estimate runoff from sheet flow and limited channel flow may be more appropriate. Izzard made extensive experiments with flows from various surfaces, over relatively small areas. He found that the overland flow hydrograph could be drawn as a composite of the two dimensionless curves illustrated by Figure 2.8 and peak flow is found when q/qe = 0.97 and t/te = 1.0. and is given by

qe, the flow at equilibrium = iL/ 3.6 × 106 whereh (2.10)

i = rainfall rate in mm hr-1 and L= length of flow surface (in m) for a portion of the flow surface that is 1 metre wide.

For the purposes of calculating only the peak flow, it is not necessary to enter into the relations between the other runoff parameters which allow the construction of the overland flow hydrograph and the calculation of total flow volume. This is discussed below in the section on runoff volumes.

Peak flow, qp in m³ s-1 = 0.97 (i L /3.6 × 106)

Worked example
What is the peak flow of runoff from a 25 m wide strip catchment length 10 m, as the result of a rainfall with a maximum intensity of 60 mm hr-1?

Using formula 2.10, Peak flow 0.97 × 60 × 10 × 25 × 1000 / 3.6 × 106 = 4.0 l s-1.

2.1.2.2 Runoff Volumes

It is necessary to estimate the size of likely runoff volumes as accurately as possible, as they will determine whether volumetric or continuous data collection methods must be used. If the former is selected, these estimates will ensure that the design of collection tank size is suitable. Tanks that are too small will be over-filled, data will be lost during large runoff events. This will be a great set-back because obtaining information about large runoff volumes and the probabilities associated with them is critical for agricultural planning purposes. The over-design of collection tanks incurs unnecessary expense and can lead to difficulties of installation.

For water harvesting schemes, it is necessary to estimate the size of runoff volumes that catchments are likely to shed. Over-estimation of runoff volumes can lead to serious under-supplies of supplementary water, whereas volumes much larger than those expected can result in flooding and the physical destruction of crops and structures. It is important to stress once more, however, that the methods for calculating runoff volumes shown below, can only provide estimates.

a. US Soil Conservation Service Method

This method is applied to small agricultural watersheds and was developed from many years of data obtained in the United States, though it has been used successfully in other regions. The Method is based on the relations between rainfall amount and direct runoff. These relations are defined by a series of curvilinear graphs which are called "Curves". Each curve represents the relation between rainfall and runoff for a set of hydrological conditions and each is given a "Curve Number", from 0 to 100. The equation governing the relations between rainfall and runoff is:

Q = (P-0.2S)2 /P+ 0.85S where (2.11)

Q = direct surface runoff depth in mm
P = storm rainfall in mm
S = the maximum potential difference between rainfall and runoff in mm, starting at the time the storm begins.

The parameter S is essentially composed of losses from runoff to interception, infiltration, etc.

The US SCS calculates S by:

S = (25,400 / N) - 254 where (2.12)

N is the "Curve Number", from 0 to 100. Curve Number 100 assumes total runoff from the rainfall and therefore S = 0 and P = Q.

Values of curve numbers for different hydrological and agricultural conditions are given in Tables 2.14 and 2.15. Note that the values for these tables are separated on the basis of antecedent soil moisture condition, that is the state of "wetness" of soils prior to rainfall. The basic assumption for this separation is that wet soils shed a higher proportion of rainfall as runoff than dry soils and therefore the same soil will have a higher curve number when wet, than when dry.


Table 2.14: Curve Numbers for Soils and Catchment Condition, Antecedent Soil Moisture Condition II


Table 2.15: Curve Numbers for Soils and Catchment Condition.

Local conditions, especially rates of evapotranspiration, should be considered to assess whether the categorisation of antecedent soil moisture conditions I to II should be modified. For example, soils in a region with summer rain and a summer growing season may fall within category 1, despite a previous 5-day rainfall of 40 mm. Similar soils with a winter growing season and the same antecedent rainfall could fall into category III.

One difficulty in using the US SCS method is the derivation of the value the rainfall parameter P. This parameter is usually defined as a specific return period storm of a known duration, for example "the 12 hour rainfall with a return period of 50 years", expressed in mm. Rainfall P is calculated from a relatively complex linear relation with several rainfall duration and return period factors. In the US, these data are easily available and can be obtained from published maps and although regional variations in the relations defining the rainfall intensity parameter exist to cover climatic variation, there are serious difficulties in transferring this kind of information to other geographical areas. In developing countries it is unlikely that such comprehensive data will be available. Even if they were, the work involved in converting raw data into a series of useful tables, graphs or maps would be beyond the scope of most projects where all that is sought is an estimate of runoff.

As an alternative, long-term daily rainfall is usually available even in countries with only basic meteorological information. The 24 hour rainfall is a frequently-used value and in these circumstances, it is best to use a simple estimate of rainfall that can be obtained from a listing of annual maxima, such as shown in Table 2.5. For example the 10year return daily (assume it to be the 24 hour) rain could be used in equation 2.11 to calculate runoff. Relations between daily and other period rainfall can be established by regression analysis where records exist. When available, local records should be used even if they are less amenable to sophisticated treatment


Table 2.16: Antecedent Soil Conditions

Worked example

Given that the 25 year return rainfall is 85 mm, calculate the total runoff volume from a catchment of 46 ha, of which 13 ha are poor pasture (soil group A), 25 ha are contoured under small grain crops with poor treatment (soil group C) and the remaining 8 ha are fallow (soil group B). Antecedent soil moisture condition 1.

Subarea A (ha)

Soil Group

Land Use

Curve No. N

N×A

25

A

Pasture, poor condition

68

1700

13

C

Small grain, poor condition

63

819

8

B

Fallow

86

688




Total N×A =

3207

Therefore the weighted curve number = 3207 /46 = 69.7

From S = (25,400 / 69.7) - 254 = 110 mm

From equation 2.11 Q = (85 - 0.2x 110)2 / 85 + (0.8 × 110) = 22.9 mm over the catchment (1mm on 1 ha = 10 m³)

The total runoff, Q = 22.9 × 10 × 46 = 10,534 m³

b. TRRL Model

Reference to this model is made for an alternative method of calculating runoff volume later in this chapter.

c. Izzard's method

The use of work by Izzard for the calculation of peak flows was discussed earlier, in the relevant section. The results can also be used to calculate flow hydrographs and volumes and is especially useful for small runoff plot calculations.

Izzard found that the time to equilibrium,
te =2 Ve /60 qe where (2.13)

te is the time to which flow is 97% of the supply rate and Ve is the volume of water in detention at equilibrium. The volume Ve in cubic metres was found to be:

Ve =kL1.33 i0.33/ 288, where (2.14)

i is in mm hr-1 and L, the length of the strip is in m. k was found experimentally to be given by:

k = 2.76 × 10-5 i + c / s0.33 where (2.15)

s is the slope of the surface and c is given in Table 2.17.

The average depth over the strip is = Ve/L = kqe0.33 (2.16)


Figure 2.8: The Dimensionless Hydrograph according to Izzard


Table 2.17: give values of the surface retardance coefficient c for various surfaces. Table 2.17 Surface Retardance 'c'

Note that for low slopes and small rainfall intensities, the value of c is relatively important.

Procedure to calculate runoff volume

- Values of te and qe can be calculated from equations 2.13 and 2.10, respectively.
- With te and qe known, the plot of the rising limb of the overland flow hydrograph, plotted as q (volume) against t (minutes) can be found from Figure 2.8.
- The recession curve of the hydrograph can be plotted using the factor B which is:

B = 60qeta/VO where (2.17)

VO is the detention volume given by equations 2.15 and 2.16, taking i = 0 and ta is any time after the end of rain.

- When the hydrograph is drawn, the runoff volume is the area under the hydrograph.

2.2 Collecting runoff data

2.2.1 Volumetric data

Simple Tanks

Complex Tank Systems

Multislot Dividers
Multipipe Dividers
Rotary Dividers

2.2.2 Continuous Systems

2.2.2.1 Natural Controls for Runoff Measurement

Rating Curves

Methods of Flow Measurement

Velocity Area Method
Float Gauging
Chemical Gauging

Stream Flow Networks

2.2.2.2. Artificial Controls for Runoff Measurement

Flumes

HS, HL and H Flumes
Parshall Flumes

Weirs

V-notch Weirs
Triangular Weirs

Culverts and Existing Structures

Methods to calculate the runoff that is likely from various rain storms on catchments of various sizes and with a range of conditions have been discussed. The following chapter describes the equipment that is needed to measure runoff using these systems. There are two main types:

Volumetric equipment

Continuous or Through-flow equipment

A theoretical estimation of runoff peaks and volumes will enable the choice between these two methods to be made, as choice is determined essentially by the size of runoff volumes. Other considerations such as the risk of sedimentation, debris in the flow, site and cost also play a part in the decision, but the amount of water to be measured is by far the most important criterion. In general, the two methods of data collection are used in the following circumstances and have the concomitant advantages and disadvantages that are outlined below. After considering these advantages and disadvantages and after estimating the size of flows that are likely, the most suitable method of measurement should be clear. If doubt remains between the suitability of using simple or complex volumetric measurement systems, further discussion is presented in section 2.2.1.

Volumetric Data Collection Methods

Advantages:

- Can be used easily to measure small volumes of runoff.

- The most basic equipment is a simple tank, although more complex systems will be described later, whereby a small, known proportion of the flow is collected and the total is found multiplying the collected flow according to this proportion.

- This equipment is relatively cheap.
- It can be manufactured locally and is relatively simple to use.
- Lends itself to the easy replication of experiments.

Disadvantages:

- The main limit on the equipment is the physical size, in particular the depth of the collection vessel, the top of which must be installed at a lower elevation than the runoff area.

- It also has the disadvantage of only collecting "lump sum" runoff volumes and gives no other hydrological information. This limits the usefulness of the data.
- No idea of the varying contributions within complex storms is available.
- No information on runoff duration can be found, nor how much rain fell before runoff started.

- The equipment must be well-serviced and be emptied (of sediment as well as water) after every runoff event. It is therefore not a good system for field station runoff measurements, because runoff may not be suspected and an arduous and often fruitless visiting schedule would be necessary to cover all eventualities on a routine basis.

- There is the risk of over-filling of the tanks which can result in the loss of accurate data, though a limited number of experiments with very small vessels can be used successfully, if a reliable field observer is engaged at the site.


Continuous Data Collection Methods

These can be undertaken in many different ways, depending on the physical properties of the flow and characteristics of the site, but the basis of measuring the runoff remains the same for all. Runoff is channeled to flow through a "control" section. This control section may be artificially constructed and as such, will have pre-determined hydraulic properties. Alternatively, a suitable section of a natural channel may be used, though the hydraulic properties of a natural channel must be determined by measurement. In either case, the volume of water passing at any time is found by measuring only the height of water in the channel (the "stage" of the flow). The measurement of stage is effected by the use a water level recorder (WLR) which records changing flow stages over a desired period. Integration of the various stage heights over the period of record gives the total flow.

Advantages:

- These methods also give information on flow durations, peak flows and on when runoff started in relation to rainfall.

- WLRs can hold the data relating to many runoff events.

- The equipment can be left unattended for months if necessary and is well-suited for use at remote field sites.

- The disruption of routine visiting schedules is not a serious problem.

- There is no limit to the flow volumes that can be measured, if the control section is large enough to pass the runoff.

Disadvantages:

- The greatest restriction on the use of this method is the cost of the water level recorders (similar to that of recording rain gauges) and, in common with any complex machinery, the possibility of malfunction. Artificial control sections can be designed to be built locally, for an outlay similar to that of volumetric collection tanks.

2.2.1 Collection of Volumetric Data

Simple Tanks

Simple tanks are used to collect runoff from the very smallest catchments. It is important to put into perspective the size of plot for which simple tanks are suitable:

Example

For every 1 mm of rainfall that is shed per 1 m² of catchment, the collection tank will receive 1 litre of runoff. A 1 in 10 year storm is an appropriate return period for which to design.

- A 100 mm storm (assumed to be a 1 in 10 year rainfall) over a 1 m² plot with a runoff efficiency of 50 % ( for example a sandy loam soil, with a slope of 2%, relatively bare of vegetation ) would give 50 litres of runoff.

- Over a 10 m² (1 m × 10 m) plot runoff would be 500 litres

- Over a 100 m² (10 m × 10m) plot runoff would be 5000 litres.

A tank built to contain 5000 litres would need to have dimensions greater than 2.5m × 2.0m × 1.0 m, (an adequate freeboard is always essential), about equal to the capacity of 25 large oil drums. This is too big for easy manufacture, installation and replication. Even to contain the runoff from the 10 m² plot, the tank would have to be greater than 1.0 m × 1.0 m × 0.5 m. Alternative complex tank systems can be designed to collect runoff amounts of this order of magnitude, and these systems are discussed later. It is reasonably obvious, therefore, that even though most runoff events will be much smaller than the example above, simple tanks should not be used on plots larger than a few square metres. The importance of measuring extreme event runoff volumes, to ensure the widest range of data for analysis, cannot be over emphasized.

Design

On the whole, the design of simple tanks is not a difficult task. The dimensions should be appropriate to the size of the estimated maximum runoff volume. Remember that an adequate freeboard is necessary. Generally it is best to have the tanks made specially, as the modification of containers used previously for other purposes may be as expensive and can lead to compromises in design.

An example simple tank and plot layout is illustrated below in Figure 2.9.

Galvanized steel plot boundaries, about 15 -20 cm above ground


Figure 2.9: Typical Design of Simple Runoff Tank and Plot Layout

Features to note are:

- Light gauge galvanized steel boundaries, about 15 cm high to avoid rain shadow. These can be easily bent into shape and be knocked into the ground or, if soils are very hard, dug in. Any seam or joint that is not sealed should have its outer edge pointing downslope to prevent the entry of outside runoff.

- "Funnel" neck to direct flow into tank. This should be large enough to allow unrestricted peak flow into the tank. It should not be liable to blockage by debris; a wire mesh may be fitted.

- Tank cover to prevent rainfall entering tank directly and prevent animals interfering with collected runoff.

- The tank should be made of inexpensive mild sheet steel, painted against rust.
- Welding destroys zinc protection at seams, if galvanised steel is used.
- Plastic containers usually have a short life, due to rotting by UV light.

One suitable alternative to the specially manufactured tank is the ubiquitous large oil drum (usually about 40 imperial gallons or a little more than 200 litres). These can be cut to size along a horizontal axis to give 2 × 100 litre containers with some freeboard. These tanks are generally too deep to allow easy installation as a whole unit (deep soils would be needed and emptying a deep container is awkward) and their horizontal installation is very inconvenient.

A section must be cut to allow a funnel arrangement to enter the tank and lids need to be manufactured as separate items. Oil drums are often prized objects in developing countries, put to many varied uses and their cost or scarcity may make them unsuitable. A typical installation is shown below in Figure 2.10.


Figure 2.10: Modified oil drum as 100 litre collection tank

Installation

The tank, of whatever type, must be sunk into the ground, therefore ensure that depth is the smallest dimension where possible and completely fill in the whole of the excavated area to avoid water collecting around the tank and disturbing it by flotation or subsidence. Settlement of the soil may take some time. Ensure that animals cannot remove the lid to drink from the water. Regard theft (and possibly vandalism) as a serious threat in field locations.

Cementing the tank into the ground should be avoided as it is likely that the tank will have to be removed on occasions for a thorough cleaning and possibly repair. Easy removal is especially important if the installation is in a farmer's field where pre- and post-ploughing runoff data may be required and removal will be necessary. The tank's presence could prevent correct ploughing. The plot boundaries should not cast a serious rain shadow and should be tapped into the ground or eased in with a spade. The galvanized sheeting recommended can be re-used over many seasons and is cheap and easy to work with. In all respects metal sheeting is superior to earth-dug bunds for small plots. During and after installation, ensure that the minimum disturbance is made to the catchment. In particular ensure that digging etc. does not impede or encourage runoff flow into the tank. Remember that surface flow on small catchments is very shallow, perhaps less than 1 cm deep and even very small surface features, inadvertently produced, can affect runoff. If installed at a field site, a permanent reader will probably be necessary.

Measurement

Carelessness while measuring the amount of runoff that has been collected can lead to as inaccurate results as careless installation of the equipment. A standard data sheet should be drawn up, such as the one below. A separate sheet per site is recommended for each visit. Each individual tank should be clearly numbered where several tanks are installed together. A record of the capacity of the tanks is essential to check against spurious measurement. The tanks can be emptied by scooping out the water using a small measuring vessel. This is preferable to having to remove the tank when measurement is made. The quickest and most reliable way to read the tanks is to have 2 × 20 litre (marked with 1 or 2 litre graduations) and 2 × 5 litre containers (marked in 0.5 litre graduations) available. The runoff is scooped out in a marked vessel (2- 3 litres capacity is the most manageable) and poured into one of the 20 litre containers. When this is full, it is noted on the sheet. This tank is emptied while the other is filled. When most of the runoff is emptied, the 5 litre vessels are used. Unless runoff amounts are very small, results to the nearest 0.2 litre are adequate. Care is needed during the process. Sediment is likely to be present.

Dip-stick measurement of runoff is difficult since accurate leveling of the tank is necessary during installation and the tank may later re-settle or be disturbed. Graduations marked on the tank are an unsatisfactory method of measurement, they become erased or covered with mud, etc. Emptying the tanks carefully is surprisingly time-consuming and an adequate period should allowed, especially where groups of many tanks are involved. It is best that at least two people should undertake the task together and three is to be preferred, one concerned only with the recording of data.


Figure 2.11: Collection tank data sheet example

Complex Tank Systems

The examples of possible runoff amounts, given in the previous section on simple tanks, indicate that small tanks can only be used on runoff plots of about 10 m or even less. For larger catchments, that for one reason or another are not fitted with flow-through measuring devices, a more complex system of collection tanks is needed.

Design

The basic arrangement of these systems is to install a first tank which collects initial runoff. When this tank is full, further runoff entering the tank causes overflow into a second tank. However, only a small, known proportion of this overflow runoff is passed into the second tank, most of the runoff is allowed to run to waste. Such systems can deal with runoff from catchment areas with an upper limit (depending on storm size and catchment conditions) of about 100 to 200 m², though this will be determined by the character of the individual site. There are two main types of complex tank systems:

Multi-slot dividers and Rotating slot dividers.

The major limitations are:

- Tank size

The manufacture and installation of tanks with capacities of several hundred litres can be difficult. Generally, they are made of heavy-gauge galvanized or mild steel sheet to retain some rigidity and are therefore quite expensive to manufacture and unwieldy to install. Because of their size and the fact that they lie one behind the other, downslope, considerable earth-moving may be necessary, especially where land slopes are low. Shed-like constructions may be needed to prevent rain from entering the tanks and flow channels. Dug earth channels behind the installation are necessary to remove the waste water during rainfall/runoff and to prevent the ponding of water which may otherwise enter the tanks and confuse measurement. These channels are also essential to remove the water as the tanks are emptied, which requires the use of a small pump. Pumps of about 0.1 horse power combine sufficient power with precision of control. The tanks should be protected from accidental runon from areas around the installation. Installations for rotary dividers are considerably smaller than those of multi-slot dividers.

The tanks cannot be considered portable in any sense of the word and a permanent location is required for their use. It is important therefore that very careful consideration is given to site suitability from the viewpoints of installation and experimental objectives. In particular, problems can occur in areas of low slope (< 2%) where back-up ponding and drainage difficulties can be severe.

- Number of tanks

If small tanks are preferred, a greater number must be used for each installation and this number is limited by accuracy, as well as the space that they occupy. The first tank in line measures all runoff, conventionally the second tank measures 1/10 th of the overflow and a third tank would usually do the same (though these proportions can be altered to suit individual needs). Thus by the third tank, only 1% of the real runoff amount is actually being collected. If the tanks are not well-manufactured and properly installed, every 1 litre inaccurately measured will affect the calculated volume by 100 litres. If yet another tank were added to increase the catchment area from which runoff was collected, then each litre measured would represent 1000 litres (1 m³) of runoff. Small inaccuracies of collection or measurement would lead to seriously flawed data. Although the difficulties of tank size can be overcome by making the tanks smaller, three tanks in line is probably the sensible limit. The problems of accuracy can be overcome by calibration (see below), but some inaccuracies in measurement and accidental spillages must be accepted as a fact of life.

Rotating slot dividers do not suffer from exactly the same limitations as multislot dividers, since the division of runoff into various proportions is only undertaken once. However, high quality design and manufacture is essential and any flaw in the mechanism may prevent operation. Peak flow estimates by calculation should be made to ensure that all inlets/outlets can cope with the flow.

a. Multislot Divider

Runoff draining from a collection gutter (which should be covered with a wire trash screen ) on the downslope side of the plot flows into a conveyance channel or pipe to the first tank. Heavy sediment will settle here. When this tank is full, 10% (for example) of the excess is passed through a vertical slot and drops down into the second tank. Various sizes and numbers of slots can be used, according to need, examples of approximate slot sizes and capacities are given below in Table 2.18.

The remaining portion of the runoff continues along the channel to be discharged as waste, or to further tanks where a similar proportion is retained for measurement. The tanks and slot plate should be made of suitable galvanized steel sheet with welded, water-tight seams. Slots should be made in the plate as accurately as possible and all angles should be 90°. However, sophisticated workshop facilities may not be available and equipment may have to be transported under difficult conditions to a field site. Thus, while it is important to construct the dividing system with care, it is necessary to calibrate the equipment after installation, to correct for any inaccuracies of manufacture.


Table 2.18: Approximate Number, Dimensions and Capacities of Divider Slots

Figure 2.12: shows a typical arrangement for a multislot dividing system. Construction details are given in Appendix


Figure 2.12: Multislot Divider System a) Plan b) Side Elevation


Figure 2.12 continued

Where the materials and workshop facilities are not available to manufacture a multislot system, a cheap alternative "multi-pipe" divider can easily be made, so long as basic welding equipment is obtainable. The system uses 200 litre oil drums, sheet steel and neoprene pipes. Calibration, as described later in this section, is essential for this system.

A metal box, with a handled lid, is welded from cut pieces of steel plate. On one side, a rectangular section, approximately 15 cm square, is cut out to accept an intake gutter. Ten 5 cm. stub pipes are welded over 5 cm. holes cut into the opposite side of the box to the intake, to form spouts. A 200 1 drum (usually 90 - 100 cm high), is cut into two portions, the first 60 cm and the second 30 cm high or thereabouts. One 5 cm hole is cut into the side of the larger piece of drum, leaving a freeboard of 5 - 10 cm. A short metal stub pipe is welded on to the hole to a form an intake. On the opposite side of the drum ten 2.5 cm spouts are fitted radially in the same way, at a slightly lower level than the intake. A 2.5 cm intake pipe is fitted on the smaller piece of drum. The box, large section of drum and small section of drum are fitted together with the neoprene pipes. All containers should have lids to prevent direct rainfall from entering them. The nine spouts that are not connected act as waste drains. The containers should be leveled when installed.

Figure 2.13 below shows how the containers are fitted together. All dimensions of containers and pipes are approximate and can be altered to suit locally available materials. The number of pipes can be selected according to need.

Installation

The installation of a system requires care more than expertise.

Large tank sets (Figures 2.13, 2.16 and 2.17) must have the site completely prepared by earth-moving if necessary, beforehand. All the tanks in the series should be carefully placed into their excavations and the conduits fitted. They should be leveled and completely assembled together before being set in concrete, where appropriate. All conduits should be leveled.

Where lowering of the tanks may be necessary later, for instance if the surface of the plot is expected to erode as part of the experimental intention, tanks should not be concreted in but placed on durable stands, the elevation of which can be reduced. Multi-pipe systems are best in such circumstances, because they are smaller. Great care should be taken not to distort the dividing system on installation. The system should be protected from rain and runon as soon as possible, where open tanks are used.

The free flow of water, with no ponding, from all tanks in the series should be ensured. Calibration should not be attempted until the concrete is set, or the tanks are secured to their stands. Procedures for the small tank system are essentially the same, but less labour will be required and alterations to the site can be made during installation.

The detailed procedures of calibration are given below.


Figure 2.13 Multipipe Divider System

Calibration

Calibration is straightforward, though the large tanks will need several oil drums of water.

- Ensure the tanks are firmly set in the ground and all the components are secure.

- Ensure the outlets are as level as possible

- Ensure the whole system is clear of debris.

- Ensure all seals are water tight

- Fill the first tank with water until it overflows to a small degree into the second tank.

- Ensure there is no ponding between the tanks.

- Remove any water from the second tank

- Using an accurately graduated vessel ( for example 10 litres), pour water carefully into the (primed) first tank.

- Record the amount poured.

- This should be repeated until an easily recoverable quantity of water can be removed from the second tank. The waste water from the first tank can be ignored, but ensure that it flows away from the site easily.

- Measure the water recovered from the second tank.

- Repeat the process but start by filling tank 2 and measuring from tank 3

The calibration factor of the tanks, or more correctly the dividing system is then the ratio: Water Poured / Water Recovered.

For example:

Tank 2
275.0 litres / 25.6 litres,
Calibration Factor (C.F.) = 10.74

Tank 3
109.0 litres / 11.2
Initial Calibration Factor = 9.73
Actual Calibration Factor- CF Tank 2 × CF Tank 3, therefore = 10.74 × 9.73 = 104.50

All quantities runoff measured from the second tank should be multiplied by the calibration factor to calculate the true runoff volume. A mean value should be obtained from a number of calibrations undertaken for each tank. For a series of three tanks, the same procedure is followed. The total runoff from rainfall represented by the portions actually collected in all the tanks would be (for the examples given above):

Runoff from tank 1 × Calibration Factor ( = 1.0) + Runoff from tank 2 × Calibration Factor (= 10.74) + Runoff from tank 3 × Calibration Factor (=104.5).

Routine data collection sheets for a multiple tank should be prepared in a manner similar to Figure 2.14:


Figure 2.14: Collection tank data sheet example

b. Rotating Slot (Coshocton) Divider
The rotating slot divider is a much smaller device than the equipment that has been described above, but involves a high degree of precise manufacture. Welding must be accurate and discrete, bumps and distortions of the metal wheel must be avoided. High quality bearings are needed. Detailed drawings for the construction of this runoff sampler are given in Appendix A3. Figure 2.15 shows a sketch of the mechanism fitted below an H flume.

Runoff is directed from a collection conduit and pours over a horizontal sampling wheel, the slot divider. The action of the water forces the wheel to rotate and the sample slot cut into the wheel continually passes under the water then away as the rotation continues. Runoff that goes through the slot enters a sump and then is conducted away to a collection tank. That which does not pass through the slot, runs to waste.

Installation

Plot design and collectors are as described above.
Installation design often depends on the land form of the runoff area.
Figure 2.17 shows an example.

The complex tank systems described above are frequently used for sampling sediment load and the combination of runoff and sediment measurement is obviously a cost-effective manner of organising activities, the value of which should be carefully considered at the planning stage of any project.


Figure 2.16


Figure 2.17

Figures 2.16 and 2.17 Alternative Installations of the Rotating Slot Divider Source: USDA Handbook 224

2.2.2 Collection of Data from Continuous or Flow-through Systems

Flow-through systems are used where catchments provide too much runoff to be collected in tanks. They are also necessary where knowledge of the start, the duration, peak and end of flow of runoff (the flow hydrograph), is needed. These systems are particularly useful at remote sites, where visits cannot be made after every rain storm. In humid climates, they can be used to make continuous readings of permanent streams. Measurements are made at a control section with known hydraulic properties, where the flow volume can be measured simply by recording the depth (stage) of water passing through the control section at any time.

Controls fall into two main categories:

Natural controls

exploit the physical features of the stream channel or other waterway, (be it ephemeral or permanent) to provide a location for the measurement of flow. This is not as simple as it may appear.

Artificial controls are made to pre-designed specifications, according to their use and are placed where required. There are many designs of artificial control.

In general, the most useful controls for work in agrohydrology and water harvesting are artificial. However, it is possible that natural controls may be used of necessity, especially where natural channel flow in large amounts is present. These volumes of water may be too large for the use of a pre-calibrated device or conditions may be unfavourable for its operation. The cost of building artificial controls on any but the smallest river channels will probably be prohibitive.

2.2.2.1 Natural Controls for Flow Measurement and Stream Gauging

The first problem to overcome is the identification of a suitable site. Bends in channels should be avoided. Eddying and spiral flow occur and cause changes in the river bed and the undercutting of banks, making such locations unsuitable. Straight sections of channel are desirable. Sites should be located where the bed is as uniform as possible, away from tributaries and other flow disturbances. Changes in vegetation, human activity etc. can all affect the control at a site.

The effects of controls (the influence of the channel having a particular form that restricts flow) can be present at low water, high water or may change as the depth of the river alters in flood. Usually low water controls become ineffective as stage increases. Contracted sections such as bridge openings may operate at highwater which may be a disadvantage, though bridges are frequently convenient access locations for measuring stream discharge.

At gauging sites, locations where frequent flow measurements are made, gauges that facilitate the recording of stream depth are placed. If possible, an automatic water level recorder and manually-read posts are used, though this is expensive. Where a continuous record of stage is not required, manual posts alone are sufficient. The hydraulic properties of a natural channel, which must be determined to allow the use of stage-only recordings to measure flow, are defined by rating curves, sometimes by rating tables.

Rating Curves

A rating curve is a calibration curve, a graph of the relation between stream depth (stage) and flow (discharge). Obviously, as stream depth increases, so does discharge. However, this relation is unique at each location on the channel and is rarely if ever a straightforward linear relation. When sufficient depth/discharge data are collected (see below) to define this relation, unknown discharge can be found by simply reading the known depth from the rating curve. Further discharge measurements are taken on a routine basis to up-date the rating curve. This will be necessary if the stream channel changes, for example after a severe flood. It is a serious problem for short-lived projects that it may take many years to compile a rating curve which, of necessity, should include a wide range of discharge, from low to high.

A typical rating curve will be similar to Figure 2.18. The relation can also be defined in a rating table, which allows more convenient use in computer programs. Note that the curve will plot as a number of straight lines when logarithmic axes are used, indicating a change of control at the inflection point(s) as stage/discharge relations change.


Figure 2.18 Example Rating Curve

Data points collected after the compilation of the rating curve, should lie within 10% of the curve. Values that do not, indicate a change in control of the river section or large measurement error. Values not within +/- 2% of rated discharge can indicate that re-drawing of the curve may be necessary in some circumstances, but it is assumed here that moderately accurate values of discharge are satisfactory, and that agricultural and water harvesting projects will not wish to invest the time nor resources to delve deeply into the theory and practice of hydrometry, nor undertake the rigorous field schedule of data collection that would be necessary to achieve wholesale rating curve revision. A simple stage/discharge plot should be adequate to define the relation.

From a graph such as 2.18, the flow hydrograph can be obtained from a continuous reading of stream height.

There is no easy way to extrapolate extreme discharges from a rating curve and although the general equation of any curve is assumed to be q =k (g - a)b, where a, b and k are site constants, this formula cannot account for changes of channel geometry at higher stages and an abrupt discontinuity of the relation will be seen when bank-full conditions are experienced; that is when the river floods and is no longer confined to the channel.

Stream gauging procedures measure the stage and discharge of a stream and provide the basic data for rating a channel location. Rough estimates of discharge for different stages can be made by taking the cross-sectional area of the channel and multiplying the channel area by stream velocity. Values of velocities can be estimated by using Manning's equation (see later in this chapter).

Methods of Flow Measurement

a. Velocity-Area method

This most accurate and usual method uses a current or flow meter and associated equipment. The meter consists of a set of horizontally mounted cups (vertical axis) that move a contact breaker as they rotate. This breaker, wired to a battery and either an automatic counter or head phones, registers each rotation of the cups. Each meter is individually calibrated and provided with a table (usually) fixed to the carrying case. This table is used to convert the number of rotations per time period into flow velocity in feet or metres per second. For very small streams or highly vegetated conditions a horizontal axis propeller-type meter is used, but the method of measurement is the same for both instruments. Figures 2.19 and 2.20 show typical cup-type and propeller meters.


Figure 2.19: Typical Cup-type Flow Meter


Figure 2.20: Propeller-type Flow Meter

A marked cable or tape is stretched across the stream channel at the gauging site, at right angles to the flow. The tape is used to divide the stream width into convenient sections. In Figure 2.21 below, 14 × 2.0 m sections are used, starting from the initial point (0) on the left bank. This leaves one section of 1.2 m at the right bank, to complete the full stream width of 29.2 m. To ensure a sufficient number of velocity readings across the stream width, no section should be greater than 10% of the total width, where possible 5% gives greater accuracy. The subdivision of the channel width allows the measurement of the different velocities and discharge within each section, due to friction and eddying. On the other hand, the time taken to complete the procedure should not be too lengthy, because minimum change in stage of the stream, during the time of gauging, is desirable. Any large change in stage can cause an inaccurate measurement of discharge, unless corrections are made. These are quite complicated and best avoided. Stage readings should be taken from the staff gauges located at the site, before starting and after completion of discharge measurements to check for excessive change of stage.


Figure 2.21: Velocity-Area Method

Figure 2.21, shows the arrangement of channel sections used in the velocity-area method. The flow meter is positioned at 0.2 and 0.8 times the vertical depth of the water, pointing against the flow. At each position the number of rotations in a given time (or given number of rotations in a measured time) is counted and the velocity calculated from the calibration table of the flow meter. The average velocity of the two readings is taken as the overall average velocity of the whole vertical section. The distribution of velocities from bed to surface in a stream is parabolic and the average of the two measurements gives an accurate measure of true mean velocity.


Figure 2.22: Sample Discharge Measurement Form

The depth of the stream is noted from the graduated bar that holds the current meter. The procedure is followed for all the sections of the stream. Where the stream is too shallow (< 0.50 m) to allow two velocity readings to be taken, the 0.6 of stream depth position, below the stream surface, alone is used. Calculations of depth, velocity and thereby, discharge are made as illustrated in Figure 2.22.

In some cases, usually when the river is at a high stage, wading the stream to effect measurement will be impractical. In such circumstances a bridge or other convenient structure must be used. Readings are taken with the meter suspended from a cable and held down by a large, streamlined weight. The current will carry the instrument downstream and it will not hang vertically, but a small correction can be made to overcome this. With the cable 12° from the vertical the inaccuracy of measured depth is about +2%, but corrections will depend on exactly how much cable is paid out and how much is in and out of the water. A table can be drawn up and used, according to Figure 2.23 below.


Figure 2.23: Geometry of a Cable-suspended Flow Meter

Operation and maintenance

On the whole, the equipment is easy to operate and maintain. Regular oiling of the cup bearings with light lubricant is essential (sewing machine oil is a good substitute if manufacturers' oil is unavailable, but do not foul the contact points) and spare bearings and a spare set of cups should be purchased. The cups in particular should be treated with care as any damage will alter the rating of the instrument. Equipment should be cleaned and dried after use. The electrical contacts should be kept clean as they tend to burn out at the tips with use. Propeller meters usually have bearings of synthetic plastic material and usually should not be lubricated.

It is desirable for current meters to be re-calibrated by the manufacturer or a hydraulics laboratory every few years, therefore careful use of the equipment is essential if this costly inconvenience is to be kept to a minimum.

Equipment suspended from a bridge or similar structure will necessitate the use of a winch. This is purchased with meter weights, cable and fittings and will include an integrated depth-counter. A simple board can be made to which the winch can be fixed for manual operation. Figure 2.24, below shows a vertical view of the winch board.

It is suitable for use by one person with all but the largest of weights used for very large rivers. Some help may be needed when the weight and meter is lowered over the side of the bridge. Purchased stands for winches tend to be expensive and large.

It assumed that suspended cable ways, which are sometimes used in operational hydrology, will be far beyond the resources of an agrohydrological or water harvesting project and that less accurate but cheaper methods of flow volume estimation, such as float-gauging would be more appropriate. Inflatable dinghies are not too expensive, but the need for an outboard motor adds to the cost.


Figure 2.24: Simple Hand-held Winch Board

b. Alternative methods of discharge estimation

Float gauging

This method probably provides the most suitable alternative way to measure stream velocity and discharge. Surface floats travel at about 1.2 times the mean stream velocity. See Figure 2. 25 below which shows the distribution of stream velocity with depth.


Figure 2.25: Depth / Velocity Relations of Stream flow (Velocity Profile)

Floats should be clearly visible and of uniform size and material. A straight stretch of channel should be used to avoid velocity changes and eddy currents. Where possible, changes in velocity should be accounted for by placing floats across the width of the stream. Several floats should be used and average reading taken.

Remember that a cross-sectional profile of the stream with depth measurements to calculate area will be essential to allow the computation of discharge. This will require a survey of the channel at a later date. Permanent staff gauges can be emplaced and a rating curve determined. This method may be useful when a river is at a very high stage and impossible to gauge by the velocity-area method. In general the cost of an automatic water level recorder to measure river stage would not be warranted if float gauging is used to the exclusion of more accurate methods.

Chemical gauging, whereby salts, dyes or radioactive materials are introduced into river flow is sometimes used. With this " dilution" or "tracer" method, a concentration ct of the tracer is injected into the flow at a rate qt. Downstream, samples are taken when equilibrium concentration cc has been achieved and the discharge qt is

qt = (ct/ce - 1) qt (2.18)

Methods of concentration determination however, involve expensive detection equipment that would not be appropriate to most projects.

Stream Flow Networks

It is likely that a project that measures stream flow will require a number of stations, but at the same time will need to keep down costs. A basic network should consider:

- Determination of the minimum catchment area to be monitored, perhaps as large as 250 km² per station in developing regions .

- A station should always be located at the catchment outlet.

- Gauging of major tributaries should be undertaken.

- Locations/streams of particular significance such as those in areas of future development should be targeted.

- Regard should be given to the kind of use the information will be put to: flood forecasting; irrigation development etc.

- A good sample of hydrological, topographical and geological types could be monitored.

- Probability forecasts are usually an important factor in collecting stream flows and the longer the station records, the better.

- However, if budgets are severely restricted, then stations can be moved after 5 or 10 years and synthetic data derived thereafter.

- Where possible sites should be located near bridges etc. for ease of measurement and instrumentation and should have good, all year access.

- Gauge height readings from manually-read, graduated posts will require a site reader.

2.2.2. Artificial Controls for Runoff Measurement

Natural controls are limited by the occurrence of natural channels, whereas artificial controls can be placed wherever there is need for them. This can be in natural channels if desired, but bunds and channels can also be installed to bring dispersed surface flow to a point suitable for measurement. Furthermore, artificial controls are pre-calibrated with known rating curves which do not have to be compiled using flow-discharge information. These advantages make artificial control structures the most suitable for agrohydrological applications. There are many designs of artificial controls, each developed to be suitable in different circumstances and it is important to select the correct design of structure for the job in hand. Figure 2.26 presents a diagram to aid selection.

The basic assumption is made here that agrohydrology and water harvesting projects will measure runoff from relatively small catchments, fields and experimental plots, though equipment suitable to measure runoff from areas in the order of square kilometres is considered. Peak flows will be relatively small, probably no more than 1- 2 cubic metres per second, in many cases peaks will be only a few litres per second. Therefore, from the wide range of artificial controls available, those that are most appropriate to small peak flow measurement have been selected and are described below in detail, with examples of installation, problems of operation, etc.


Figure 2.26: Selection of Artificial Control Structures

Situations may be encountered where small structures are inadequate, and examples of large artificial controls are given, but in less detail. These are minor works of civil engineering and their construction is usually undertaken only by River Authorities and similar organizations. They are costly and permanent, but in some situations may be essential if stream flow data are to be collected where natural controls are unsuitable. Any project proposing to enter into the construction of such controls is urged to approach the relevant Authority and seek advice as to those which have proved most suitable for local conditions, their likely cost and problems of installation.

2.2.2.1 Flumes

Flumes are essentially long, box-like structures that allow the flow of water to retain or increase its kinetic energy as it passes through them. They have the advantages of being able to measure small flows accurately while allowing debris and sediment to pass. They can be made light and portable and can be located in most situations. They may be fitted to small experimental plots which do not have natural channels, or be placed in steam beds. They are probably the most suitable of all artificial controls for agrohydrological applications.

a. H, HS and HL flumes

The H flume is the basic instrument of this group. HS flumes are designed to measure very small flows accurately (flows < 28 1 s-1), while HL flumes are capable of measuring much greater flows (up to 3.3 m³ s-1). The materials from which they are made, their installation and operations are similar to the H flume, though their dimensions are somewhat different. H flumes fill a wider niche of runoff measurement than HS and HL flumes, therefore this section will discuss them in detail. The design criteria of HS and HL flumes are given below in Figures 2.27 and 2 28. Rating tables for the conversion of recorded stage to discharge are given in Appendix A1. The main point to note is that the dimensions and quality of manufacture strictly determine the rated capacities of the flume, therefore they should be made as accurately as possible to retain the correct rating characteristics.

HS and HL Flumes


Figure 2.27: Design Specifications of HS Flume


Figure 2.28: Design Specifications of HL Flume


Table 2.19: Capacities of HS and HL Flumes

Rating tables for these HS and HL designs are given in Appendix A1.

H flumes

For the range of flow measurement met in agrohydrology, especially runoff plot and farmer's field studies, H flumes are very useful measurement devices, but it is important that they are manufactured and installed with precision. Figure 2.29 gives the design dimensions.

Construction

The flume is the component of the whole structure to which the water level recorder is attached to measure stage.

- To manufacture the flume, prepare detailed drawings of the design with the maximum capacity needed.

- Make a paper and then a thin sheet metal template of the flume that can be used many times.

- Use either heavy gauge galvanised or mild steel sheet that can be rust-resist painted to make the flume.

- The thickness of metal should be appropriate to the overall size of the flume.

- Support all edges with angle iron or structural steel to prevent warping.

- Welded joints should be water tight, strong and ground smooth.

- Vertical sides should be exactly vertical and made from one piece, the bottom plate should contain no more than one joint and it should not be closer than 30 cm to the outlet.

- Avoid all distortions, dents and warps when cutting and fixing plates.

- Before installation, the flume should be checked for adherence to the proportional dimensions in Figure 2.29.

Approach section and stilling well

The flume head or measuring section acts as an artificial control to allow stage measurement, in addition all flumes need an approach section, attached to and upslope of the flume. They also need a stilling well upon which to site the water level recorder (WLR). Remember that the top of this well must be of a suitable design to take the specific manufacture of WLR. It is usually essential therefore, to obtain the WLR before the well is designed and fitted. Stilling wells are best made of the same metal as the flume and welded to the head measuring section. Openings allow the passage of water between the flume and stilling well.

H Flume Specifications

For small flumes (D = 20 to 60 cm), it is good procedure is to construct the approach section out of flume metal according to the specified dimensions and weld it to the flume in the workshop, with the required slope of 2%. In the field, the flume measuring section is installed with its floor horizontal (use a spirit level) and the approach section will then be set at the correct slope without further action. Such an installation has the advantage of being portable and can easily be removed for ploughing or relocation. Large flumes may require under-floor support of the approach section. Handles on the walls allow ease of portability (see Figure 2.30 below).


Figure 2.29: Design of H flume

Note: for flumes with D < than 30 cm, length of flume is made greater than 1.35 D to allow for float and stilling well.

Source: USDA Handbook 224

For large flumes (depth 1.2 m+), construction of the approach section can be completed in the field using cement block walls to the appropriate dimensions in Figure 2.32. All block work and cement floors should be rendered smooth and the join between flume and approach section should be well sealed. Alternatively, treated wood (tongue and groove with water tight joints) can be used for large or small flumes. This can be sheet metal-covered if preferred and makes a good, cheap temporary structure, but consider the problems of termite damage and rot.


Figure 2.30: H Flume, Approach Section and Stilling Well

A concrete approach floor with a 2% slope can be used, with the (metal) flume discretely bolted to it and the join sealed. However, concrete floors do not allow the same flexibility of removal and may be problematic where installation is dependent on seasonal ploughing. In all cases, angle iron should form the sloping edges of the flume, to prevent any distortion.

Installation
Before installation, it is well worth considering the following points:

- Has the flume been checked for correct manufacture?
- Has a test fit of WLR, float and counter weight been made?
- Has permission to install been given where required?
- Will the installation be permanent / be there for many seasons ? or
- Will it have to be removed for ploughing and then replaced?
- Is the design too big for easy transport and installation?
- Is it located in the correct position?
- Would many cheaper but shorter-lived flumes serve the purpose better?

Installation should take place with the approach section just below ground level and at the lowest elevation of the plot or catchment. This is convenient if in a natural stream channel, but for agrohydrological measurements, this may entail the construction of bunds (typically earth or galvanised steel) to concentrate the flow. If gutters are used instead, they should be covered or the runoff from rain falling directly into them must be taken into account in runoff calculations. To avoid scouring and undermining of the approach section, a small clay or cement apron can be positioned where water runs into it from the plot. A hard surface (tiles, cement) is placed below the outlet point to prevent erosion.

Remember that the measurement of runoff from ploughed fields in particular involves sedimentation as a problem. If so, then the flume should have a 1 in 8 sloping floor fined as shown in Figure 2.31. A sloping floor makes no significant difference to runoff measurement.


Figure 2.31: Front Elevation of H Type Flumes Showing Sloping Floor

Setting the water level recorder
Flumes require water level recorders to be fined at the measuring section to record water height. The simplest way to fit the WLR is as follows.

- Test that the WLR float and counterweight move freely up and down within the stilling well, with the WLR sitting on top of the well, but not fixed.

- The stilling well should be designed to have its base at a lower level than the flume floor by about 10 cm, thus forming a sump.

- The sump is filled with water until it flows out into the flume.

- The WLR is fined with tape, float and counterweight.

- When the float is lowered, water is displaced until the float rests at the zero position.

- The WLR pen or electronic level is set and the WLR stand can be bolted into fixed position.

- The WLR is then set in relation to the flume floor and any accidental subsidence will not affect readings.

- As the sump water evaporates the WLR will register negative readings, but in the event of runoff, the sump will fill and the float will rise rapidly.

- The amount of runoff needed to fill the sump is negligible.

- Alternatively, a wire frame in the sump set to a level whereby the float rests on it at zero level with the flume floor can be used to prevent negative values. - Check the sump for sediment and clear as necessary.

Rating Tables and Equations

Strictly speaking, current metering checks should be made on the operation of flumes, to ensure that design specifications have been followed precisely and rating is accurate. However, in practice the facilities and time to do this will rarely be available for project staff working in the field, especially in developing countries where suitable facilities may not exist. It is essential therefore to construct the flumes accurately and avoid accidental damage to them. Damage is most likely during transit to the installation location, or during seasonal removal.

The equations that govern the rating (stage/discharge) of H flumes are complex and it is advised that the rating tables provided in Appendix A1 are used.

The rating equations for H, HS and HL flume stages are given below for discharge in cubic feet:

Low Flow: Transition:

Q = A0 (280 + B1H) H( H - 0.01)A1
Q = (K0B0 + K1B1H) (2g)0.5 H1.5 (2 19), (2 20)

Medium and High Flows:

Q = {(E0 + K1D) B0 + (F0 + F1D) B1 (H +v2/2g )}(2g)0.5 (H + v/2g)1.5 (2.21)

with D = 1 and v = average velocity at the head measuring section


Table 2.20: Coefficients of Rating Equations 2.19, 2.20 and 2.21

Siting and Plot Construction

For small catchment and plot runoff measurement, the simplest location for the flume is in the lowest-lying corner. Use of a simple levelling instrument from one base will identify this point, without the need to undertake a comprehensive plot survey. Few plots, even those that have been land-levelled are absolutely square to the land contours and it is likely that the lowest point will lie in a corner of the plot, which is conveniently the focus of the defining bunds or walls. However, the use of a level is necessary as an assessment of elevation by eye can be misleading. Galvanised sheeting cut to 30 cm wide strips and dug into the ground will form a durable perimeter for small plots. The edges must overlap well, with upslope ends of the metal sheets on the inside of the overlap. Where they meet the flume approach section, they can be bolted to it and provided with a water-proof seal. Earth banked against the inside of the sheets at this point can help prevent scouring.

Rarely, scouring around the flume approach mouth can be a problem. The metal approach in Figure 2.30 has a step dug under the ground to help prevent this. Compacted soil, clay or cement aprons can also be used, but a solution will depend on the particular circumstances of the site and workers should be prepared to use their imagination in overcoming small problems such as these. Earth bunds will be cheaper for larger catchments. Experience shows that for a 0.4 ha plot (4,000 m², 100m × 40m) a perimeter bund built to 50 cm takes about 4-8 days to complete by 4 people on a hard, compact sandy loam soil. Picks, shovels and mattocks will be needed. Settling of the soil reduces the height of the bund to approximately 35 cm after a few weeks, with no further reduction. Weed growth soon aids stability. As a guide to perimeter bund construction, pegs with string at the desired height are adequate. Soil is dug and thrown in, with the trench on the outside of the plot. Obviously 35 cm bunds enclosing a plot with a flume of greater depth would not be adequate, or if they were, they would indicate that the flume had been over-designed. For any but the smallest plot, it is not necessary to cover the flume or account for rain falling directly onto it. In the example plot above, the H flume and approach section represent only about 0.01% of the runoff area.


Figure 2.32: Alternative Installations for H Flumes

Submergence

Wherever possible, flumes should not be located where submergence, that is the ponding of discharged water around the outlet, will occur. Drainage channels (where necessary) should be adequate to deal with the removal of discharge. This is not a problem at locations with any reasonable slope, but in low-slope areas (1% or less) it can cause difficulties. H flumes are well designed to cope with the submergence problem, 30 and 50 % submergence cause less than 1 and 3% inaccuracies in the measured flow, respectively.


Figure 2.33: Head and Submergence, H flume

Figure 2.33 gives the relation between the increase of flume water head due to submergence and depth of submergence. It is defined by the equation:

H = d1/ 1 + 0.00175 (ed2/d1)5.44 where (2.22)

H = free flow head;
d1 = actual head with submergence;
d2 = tail water depth above flume zero head
e = base of natural logarithms (2.71828)

However, before seeing the head / submergence relation as a way out of this problem it must be remembered that a second WLR, or some other method, is needed to measure the depth of submergence. Given the high cost of WLRs, (as well as the extra time needed to analyze the data) it is best to choose a less problematic site for installation, if at all possible.

b. Parshall flumes

Parshall flumes are a particular type of Venturi flume, their chief advantage being that they cause only a low loss of head during operation. Their design is based on a long constricting section or throat, the floor of which is flat. They are more difficult to construct than H-type flumes, having a more complex shape, but in general they have no significant advantages for measuring runoff in most circumstances, except that they can be constructed on site to measure very much larger flows. Field calibrations, with velocity recordings and large flows of water, which are difficult to arrange, could be necessary. Small Parshall flumes can be bought relatively cheaply, but they are too small to be fitted with WLRs and therefore are only suitable for regular, predictable flow, such as that in irrigation channels. In these circumstances, stage can be measured manually on a regular basis. Small flumes can be manufactured from welded sheet metal, following the careful practice outlined above in this section, though the design is complex and the tolerances of dimensions are very small. They are installed with the flume floor level and care must be taken that they are stable and undermining by erosion cannot take place in front of the converging section. The stilling wells are located in the adjacent banking (see Figure 2.34) and hydraulic connection to the water level recorder is provided by a connecting pipe at flume floor level. WLR installation procedure is the same as for the H flume.


Figure 2.34: Parshall flume

Source: USDA Handbook 221

Rating Equations
The general rating equation for small Parshall flumes is:

Q=4Wha1.522 W0.026 where (2.23)

Q = discharge in cubic feet
W= throat width or length of crest in feet (the size of the flume)
Ha = gauged head, 2/ 3 {(W/2)+4)}feet back from the crest in feet

In metric form, with dimensions in metres, the rating equation is:

Q= 4 ( 0.3048)2-1.57(W)0.026 WHa1.57(W)0.026 where (2.24)

Q = discharge in m s-1
W = throat width in m
Ha = upper gauge head in m at a point 2/3 (W/2 + 1.219) metres back from the crest.

The general formula for large flumes (> 3 m) is given by Parshall as:

Q= (2.29265 W + 0.47376) Ha1.6, with all values in metres (2.25)

Large flumes are constructed of reinforced concrete in the field and their manufacture is a difficult task and accordingly expensive. They are generally used where flows are large (they can measure flows much greater than H-type flumes) and where backing up of water and submergence can be a problem. They do need more than one water level recorder where this latter condition is met, however. Where flows are regulated and orderly (for example during irrigation procedures) they can be used with manual gauges, which should be read at short, regular intervals. This saves greatly on the cost of WLRs, as is also the case for H flumes. Capacities and dimensions for Parshall flumes are given in Appendix A 5.

Submergence

Submerged conditions occur when water in the diverging section impedes flow in the converging section and they demand a more complex formula than for H flumes. It is accurate for values of Hb / Ha up to 0.96:

Q = C1 (H2 - Hb)n1 / {-(log Hb/Ha + M2)}n2 (2.26)

Values for submerged flow coefficients and exponents, C1, Ha, Hb, n1, C2 and n2, are given below in Table 2.21. St is transition submergence, where free flow changes to submerged flow.


Table 2.21: Submerged low Coefficients and exponents for Parshall Flumes (m)

2.2.2.2. Weirs

A weir is a low dam or wall built across an open channel and has a specific shape and size. Water flows over in a free-falling sheet (nappe), but if the nappe is partially under the water downstream of the dam, it is said to be submerged. In this condition the accuracy of measurement is reduced. There are many types of weir, but none are suitable for locations other than those with light concentrations of sediment. Some common designs are described below.

a. V-notch Weirs

These are often used to measure low flows, as they do so accurately. They are therefore relatively useful in agrohydrological situations. The common V-notch is a 90° opening (usually cut from a metal plate) with the sides at 45° to the vertical. The approach velocity of flow can be ignored if the distance from weir to bank is twice the head and the height from channel bottom to the crest is twice the head. To fulfil these criteria, modification to the approach section is not usually difficult. V-notch weirs are also useful in the agrohydrological context, because not only can they be used to measure flow from plots and small catchments, they are relatively easy to make and install. Their rating equation for various flows is simple. Their biggest disadvantage is that they are unsuitable for locations with any other than low concentrations of sediments. The V-notch should be kept clean and sharp at all times.

Manufacture and Installation

The 90° V-notch is cut from rigid 10 mm sheet mild steel, which is galvanised or carefully painted to resist corrosion. This is bolted to the cement block approach section (with a rubber gasket sealed joint), which also acts as a sediment sump.

Figure 2.35 below shows a typical installation of a small V-notch weir.


Figure 2.35: Installation of V-notch Weir

The V-notch is bevelled to a sharp edge and must be maintained in this condition. The stilling well is located away from the weir, at a convenient point and hydraulic connection is made to the sump (at the level of the V apex) by a 5 cm diameter pipe. The stilling well can be any convenient dimensions. A small oil drum fixed into the ground, with a suitable outlet for the connecting pipe, makes a good form for the stilling well. It should be treated to resist corrosion. The top of the drum has welded or bolted onto it, fixings appropriate to the type of water level recorder to be used. It should also be fitted with a lid with small holes adequate for the passage of the float / counterweight tape or wire. The connecting pipe should be fixed to the drum prior to installation, laid horizontal by levelling and sealed at the sump end. Bunds are raised to direct flow to the weir and should be solid enough to resist erosion. The simplest way to set the level of the WLR and to check that the levels of the V-notch and pipe are the same, is to fill the sump and stilling well with water (though this may take several hundred litres, depending on the size of stilling well and sump). The water level should be allowed to settle until it is just at the apex of the V-notch and at the bottom of the pipe in the sump and stilling well. The float can then be lowered, a small amount of displaced water will drain and the pen or electronic counter on the WLR set to zero. The WLR will register negative values due to evaporation from the stilling well during long periods without runoff. Some account must be made of runoff collected in the sump and refilling of the stilling well after rainfall, especially if the runoff event is small. The capacity of the sump may be several hundred litres. This procedure will depend upon the exact circumstances of the installation, the size of runoff event and the degree of sedimentation of the sump, etc.

Rating Equation

The rating equation for a 90° V-notch weir is relatively simple and from it a rating curve or table can be derived. It is:

Q = 2.49 H2.48 where (2.27)

Q = discharge in cubic feet ( 35.3 ft3 = 1 m³ = 1,000 litres)
H = head above lowest part of the V-notch in feet ( 1 foot = 0.305 m)

For V-notch weirs with angles not equal to 90°, the rating equations are complex and discussed below for large structures. For small weirs, however, there is little or no advantage in diverging from this orthodox design.

b. Large Weirs

Examples of large artificial controls are given below, but in less detail than those used for the measurement of small runoff flows. These structures are minor works of civil engineering usually undertaken only by River Authorities and similar organisations with the necessary equipment and skills. They are costly and permanent, though in some situations they may be essential if stream flow data are to be collected and natural controls are unsuitable. Any project proposing to enter into the construction of such controls is urged to approach the relevant Authority and seek advice as to those which have proved most suitable for local conditions and their likely cost.

Broad Crested V-notch Weirs (Triangular Weirs)

Large versions of the 90° V-notch weir can be used to measure large volume of runoff, but as runoff amounts increase so does (usually) the presence of debris which may block the outlet. Triangular weirs pass floating debris easily.


Figure 2.36: Broad Crested Weir

They are large, permanent concrete structures, capable of measuring flows greater than 30 m³ s-1 and involve a considerable input of finance and labour. Backwater ponding is not permitted. However, they are relatively simple in design and construction' compared to alternative weirs and so are discussed here in some detail. Figure 2.36 shows the dimensions of such a weir with a 3:1 sloping section.

A straight section of channel is needed for 20 m upstream and a concrete apron 4m long is needed downstream. A large end cutoff wall is necessary to prevent the structure being undermined. The calibration of these weirs is affected by the approach velocity, the cross-sectional area of the approach 3 m upstream from the weir being a measure of this. Rather than providing a series of rating equations, which are very complex for these weirs, rating tables are given in Appendix A 4.

c. Culverts and Similar Existing Structures

In some instances it is possible to use existing structures such as road culverts to measure runoff. The advantages of such structures are that they may be fairly common and will be already in place. Sometimes they may have to be built out of necessity for other project activities and so impose no extra cost on the hydrology budget. It is important to remember however, that in most cases existing structures will not have been made with runoff measurement in mind and modification may be necessary.

This can be costly and time-consuming. They may not be conveniently located and serious errors of estimation can occur when such structures are used without knowledge of suitability. The basic aspects of flow in culverts are discussed here because culverts and their runoff capacities can be an important aspect of water harvesting schemes, farm layouts and irrigation projects. In particular, square concrete and circular corrugated metal culverts are frequently encountered.

Flow in culverts

Culvert capacity can be controlled by the inlet section or the conduit. In either case the head water elevation may be above or below the top of the inlet and the solution to calculating culvert flows depends on the head and tail water conditions. Square or circular sectioned culverts may be used, but neither are accurate meters of low Bow when compared to pre-calibrated artificial control sections. The three main types of flow in culverts are:

1. Where the slope of the conduit is less than the neutral slope. The conduit is full and therefore controls the flow. Inlet submerged, outlet submerged or not.

Use Pipe Flow Equation (2.29).

2. Where conduit slope is greater than neutral slope. Inlet submerged, outlet is not submerged. Entrance controls exist (inlet submerged) .

Use Orifice Flow Equation (2.30) or Figure 2.38

3. Inlet not submerged, outlet not submerged, culvert slope less than neutral slope.
Conduit controls exist and Entrance controls do not.

Use Manning's Open Channel Formula (2.31)

Neutral slope is defined for small angles of the conduit to the horizontal by:

Neutral slope=tan x = sin x = Hf/L = Kc (v2/2g) where (2.28)

x is the slope of the conduit,
Hf is friction loss in conduit of length L (m)
Kc is the friction loss coefficient v is the velocity of flow in m s-1 g is the gravitational constant in m s-2

Figures 2.37 (a) to 2. 37 (c) illustrate these conditions, respectively.


Figure 2.37 (a)


Figure 2.37 (b)


Figure 2.37 (c)

1. Pipe flow (that is when the conduit controls the capacity of flow) usually occurs when the slope of the conduit is less than the neutral slope. The pipe flow equation is:


where (2.29)
Q = flow capacity in units of L3 T-1
a = cross-sectional area of conduit in units L2
H = head causing flow in units of L
Ke = entrance loss coefficient
Kb = loss coefficient for bends in the culvert and can therefore often be ignored.
Kc =head loss coefficient (which = (1,244,522 n2)/ d1.33 where d =diameter in SI units and n = Manning's n)

Values of a, L and H are measured.

To see if conduit slope ( x ) is less than the neutral slope, the latter is found by equation 2.28:

Sn = tan x = sin x = Hf / L = Kc (v2 / 2g) where

x = slope angle of the conduit
Hf = friction loss in conduit length L in m
L = conduit length
Kc = friction coefficient
v = velocity of flow in m s-1
g = gravitational constant in m s-2

Worked example

What is the capacity of a 600 mm diameter culvert, 15.0 m long with a square edged entrance? Survey shows the inlet elevation to be 456.35 m, the outlet elevation = 456.20 m, the head water elevation is 457.95 m and tail water elevation is 455.25 m ?

The first step is to assume that pipe flow prevails and use equation 2.29

Ke = 0.5 (square inlet)
Kc= 0.319 (with Manning's n estimated at 0.036)
Kb is only used for conduits with bends and therefore is not included
H=0.94 m
a =3.14 m²

Therefore:

m² s-1

To determine if the pipe flow assumption is correct, the neutral slope is calculated from equation 2.28 and substituting discharge / area for velocity in SI units,

Sn = 0.319 × (0.283)2/ 2 × 9.8 × (0.28)2 = 0.0166,or 1.7%

As the actual slope of the culvert S = 456.35 - 456.20) / 15.0 = 0.01, or 1.0 %, then as culvert slope < than neutral slope, pipe flow conditions prevail. To check whether orifice conditions (as opposed to pipe flow) prevail, the orifice equation 2.30 below, is used. The values of h and C in this equation are 0.85 and 0.6 respectively. The discharge is:

m³ s-1

This discharge is greater than the full pipe capacity and therefore pipe flow must prevail. Using the procedures above and in cases where the culvert slope is found to be greater than the neutral slope, pipe flow cannot prevail, the orifice equation should be used. Outlet not submerged.

2. Orifice flow is the second type of flow and is found when the conduit slope is greater than the neutral slope, the inlet is submerged but the outlet is not (i.e. inlet controls exist). The equation for orifice flow is:

where (2.30)

a = cross sectional area
h = head to the centre of the orifice
C = 0.6 for sharp-edged orifices
g = the gravitational constant in m s-2

Alternatively, Figure 2.38 can be used to determine flow from an orifice with inlet (entrance) controls where H= head and D= diameter of a circular sectioned culvert in m.


Figure 2.38: Stage Discharge Relation for Control by Square Inlet to Circular Pipe

3. Channel Flow (Manning's Formula).
In some cases the headwater elevation is lower than the top of the inlet and control is by the channel or conduit itself. This occurs where the conduit slope is too shallow to allow the maximum possible flow that could be provided. Manning's formula is used to calculate flow. In other cases, despite the low head, the inlet section still provides the control and restricts the flow. This occurs when the slope of the conduit is greater than that required to move the possible flow through the inlet. Figure 2.38 can be used in these cases. To determine whether the results as given by either channel flow or inlet restriction calculations are correct, follow the steps below.

Assume restriction at the entrance does not exist.

Calculate channel flow using Manning's formula:

v = R0.667 S0.5/ n where (2.31)

v = average velocity of flow in m s-1
n = roughness coefficient of the channel
R = the cross-sectional area divided by the wetted perimeter (a/p) in m
S = hydraulic gradient (channel slope)

The use of this formula depends on making the dimensions of the channel such that v = Q/a and where Q the flow rate of the channel.

Then use Figure 2.38 to check the flow through the inlet. If the Manning's solution gives a flow greater than that from Figure 2.38 then obviously the latter is correct since this represents the upslope control of all flow through the conduit. Where the situation is reversed (channel flow < inlet flow), channel flow will prevail.

Manning's Formula Worked Example

Determine the capacity of a pipe of 1.20 m diameter, 20 m long with a square-edged entrance. Elevation of the inlet is 224.0 m, outlet is 223.95 m . Head water elevation is 225.0 m and tail water elevation is 220.0 m. In this case, the conduit has a very shallow slope and channel conditions may prevail. In the first instance assume a flow depth in the conduit of 0.6 m. Then

a = 0.57 m²
n = 0.015
R = 0.185 m
S = 0.0025

Substituting in Manning's formula (2.31), v= 1.08m s-1 and Q=0.62m³ s-1

If it is assumed that the approach velocity is negligible, then the loss of static head due to acceleration is = v2/2g, = 1.082 / 19.6 = 0.06 m. The depth of water at the entrance (headwater elevation minus inlet elevation) is 1.00 m and a loss of 0.06 m would give 0.94 m, which does not correspond with the assumed depth of 0.6 m. The process of iteration can be continued. If the flow depth is now assumed to be 0.90 m, then:

a = 0.94 m²
n = 0.015
R = 0.269
S = 0.0025

Substituting into Manning's formula: v = 1.39 m s-1 and the loss of head, v2/ 2g, now = 0.10 m, which when subtracted from 1.00 m = 0.90 m, the assumed depth. Thus flow is limited by the conduit and the discharge of the flow is: Q =1.39 × 0.94 = 1.31 m 3 s-1

Very many different designs of culverts are constructed and it is recommended that specialist manuals be consulted if work in the area of culvert structures and their hydraulic properties in to be studied in great detail.

2.3 Water level recording instruments

Crest Gauges
Manual Gauges
Automatic Water Level Recorders
Electronic Logging Recorders
Chart Recorders
Bubble (Servo-manometer) Recorders

The measurement of flow volumes that use control sections, be they natural or artificial, necessitates the collection of water level data for the passing through the section, either automatically or by observer. The collection of these records is made by the use of one of three types of instrument. In order of providing least data first they are:

Crest gauges:

These gauges record only the highest level of


flow, but do so automatically.

Manual gauges:

These are simple gauges and provide records


whenever an observer is present to read them.

Water Level Recorders:

These are relatively sophisticated gauges that


provide a constant record of water levels.

The choice of gauge will depend upon the importance of the data: the first two types of instrument are cheap to manufacture from local materials, whereas the latter must be purchased at considerable expense. However, to help in correct choice, here is a list of the advantages and disadvantages of each, with a description of the circumstances for which they are suitable.

Crest gauges

Advantages:

- Cheap to manufacture out of local materials
- Easy to transport, place and maintain
- Very little instruction is needed for correct reading

Disadvantages:

- Provide very little data, only maximum peak flows
- Must be visited and read after a high flow

Suitable For:

- Situations where peak flow and maximum discharge only are required, for example estimating maximum flood levels, survey work, maximum flow probabilities. Some manner of converting flow stage to discharge must be available, if discharge values are needed.

Manual Gauges

Advantages:

- Cheap to manufacture from local materials
- Can be easy to install
- Need little maintenance
- Can provide good data from streams that flow regularly
- Essential backup and check on natural controls that have WLRs
- Data is a permanent, written record

Disadvantages:

- Do not give continuous records
- Need to be visited regularly or retain a gauge reader
- Data is as good as the reliability of the reader
- Can be washed away in flood
- Installation can be difficult
- Analysis demands manual input into computer storage

Suitable For:

- Commonly used on permanent streams
- Irrigation schemes
- Can be used on flumes etc. instead of WLRs where flow is regular and easily monitored
- Not suitable where a continuous record is needed
- Distant sites visited from base on a regular but not frequent basis can often be provided cheaply, but a local reader will be necessary.

Water Level Recorders

Advantages:

- Automatic, need infrequent visits

- Give complete runoff record: duration, peak flows, flow recession, volumes

- Data can be linked well to rainfall data (from intensity gauges)

- Data can often be downloaded directly into the computer for analysis with a great saving of time.

- Most suitable for remote sites, need no reader at site

Disadvantages:

- Expensive
- Need regular checks and maintenance
- May be difficult to repair.
- Solid state electronic instruments will have to be returned to manufacturers for repair.
- Need a higher level of training for correct usage.

Suitable For:

- Used where good quality data is essential
- Often used at remote sites where data collection would otherwise be impossible
- Especially useful at base stations where core research is being conducted

a. Crest Gauges

A typical crest gauge is illustrated below in Figure 2.39. Crest gauges provide useful information on peak flows when no observer is present. They can be fixed to bridges, stable stream banks or spillways. One of their most important advantages is the ease at which they can be constructed easily from cheap, locally-available materials.

Construction and Installation

Galvanised steel water pipe or plastic water pipe, with a 5 - 8 cm diameter is suitable. The former is less prone to damage from flood water, debris and rough handling, but plastic is easier to work with, lighter and cheaper.

Suitable caps for plastic pipe are sometimes more easily obtained and do not need to be screw-threaded to fit. Any suitable length from 1.0 - 1.5 m can be used conveniently. A series of 0.5 cm holes are drilled into the lower pipe cap and act as intakes to provide hydraulic continuity with the flow. A vent hole must be provided in the top cap or upper portion of the pipe. Inside the gauge a wooden measuring stick is placed, graduated with clear markings. One centimetre marks give adequate accuracy. Screwed to the measuring stick is a small, perforated container of plastic or non-corroding metal. For this purpose it is better if the stick has a square section. Within this container is placed powdered cork or fine polystyrene granules. When peak flow occurs, this material floats out of the perforated container and deposits itself on the measuring stick, from which the peak flow reading is taken.


Figure 2.39: Typical Crest Gauge

Note that the lower cap has a support for the measuring stick, which is securely screwed to the top cap. The gauge should be installed precisely in a vertical position, using rust-resistant, bolted brackets. The openings should face the direction of flow. The gauge should be levelled to a permanent bench-mark, so that in the event of removal, replacement can be effected from the same base level. Where other gauges are used (for example manually-read staff gauges), it should be levelled in sequence with these if possible, or at least to the same bench mark, so that a relative reference point is available. Care should be taken when replacing the measuring stick if the water level is higher than the bottom of the gauge, as a temporary displacement of water in the gauge could lead to a false reading.

As well as measuring peak flows, crest gauges can be used to measure volumes at unattended sites, or at times when the observation of maximum runon volumes is impractical. For example, some water harvesting systems direct runoff into small basins that provide supplementary water for fruit trees. Knowledge of the basin symmetry and water level can provide an estimate of total received runon, though infiltration losses need to be accounted for. Evaporation is unlikely to cause serious inaccuracy.

b. Manual Staff Gauges

Staff gauges are made of metal (often ceramic-covered) or plastic strips, about 12- 15 cm wide in 1 - 2 m long sections. They have numbered, graduated markings at 1 cm intervals. They are placed in a low-velocity location on structures (bridges, stilling wells, etc.) or on posts in the river bank. The water level is read from their graduated markings. They may be installed singly or in sequence, as illustrated below. In the case of ponds and reservoirs, the posts are set inclined for increased accuracy, as a small increase in water level can represent a large increase in volume. In this case the graduation and setting must be done very carefully.


Manual staff gauges installed in sequence

Staff gauges can be purchased and these are of high quality, but very expensive compared to those manufactured from local materials, especially when shipping costs are taken into account. On the whole, plastic gauges are best avoided as they eventually become brittle and breakage can result when they are placed in rivers that carry much debris. Resetting (which in flashy streams may be necessary each season) of these gauges can also lead to damage and they make popular targets for shooting practice.

Construction and Installation

The facility to manufacture locally is a great advantage. As well as providing large savings in cost, it is important to have replacement gauges available immediately. Perfectly serviceable gauge plates can be made in the following way, for less than one tenth of the commercial cost.

Cheap gauges can be made from flat wooden boards which have been treated to prevent rotting though these will last a few seasons. Aluminium sheet and galvanised steel sheet provide better alternative materials and can be cut into strips for use.

A suitable stencil with which to paint the graduations can be cut (0.5 m is a practical length) from acetate or thin, stiff card,. Thin metal makes a durable stencil, but tends to bend and work less well. The metal strip is painted black as a background. The stencil is placed on the sheet and the markings spray painted in white. Numerals 5 cm high are sprayed at 10 cm intervals in white on the black background Numerals 7.5 cm high are sprayed on at every metre interval. These can be sprayed on in the field according to particular need, another advantage over pre-marked, purchased plates. Two suitable designs are shown in Figure 2.40. The finished plates are then screwed (brass or stainless steel screws) to a treated back-board to maintain a suitable rigidity.

Staff gauges are emplaced at the gauging station during the dry season when permanent streams are at their lowest, or at any convenient time for ephemeral flows. Where possible, they should be fixed to bridges etc. to reduce the risk of loss in floods.


Figure 2.40: Example Graduated Markings for Gauge Plates

The type of fixing will depend on the structure available, but all plates should be vertical. Set the lowest one first and then in sequence. If they are to be placed on a stream bank, 5 cm galvanised water pipe sections make good posts. Alternatively rot-resistant local timbers can be used, but these are difficult to hammer into the stream bed or banks and may necessitate the use of a manual post hole digger. River Authorities and similar bodies have access to heavy installation equipment that is unlikely to be available to most projects. In all cases, the gauges must be levelled to ensure the sequence is accurately placed and the bottom of the gauge must be levelled to a permanent bench mark. If man-made structures are not available, nailed and painted marks on several large trees, well away from the river, will suffice. At many sites it will be too difficult to level into a national survey, but a site plan including all levelling details, should be made. Checks on the level of the gauges should be carried out at least once each year. Staff gauges can be used with artificial controls where flow is regular and reading can be arranged.

c. Automatic Water Level Recorders (WLRs)

There are many different manufacturers of WLRs. There are, however two main types:

Float and Counterweight Recorders
Pressure Sensing (Bubble gauge ) Recorders

Advantages and disadvantages:

In general, the former are the cheaper and more commonly encountered. They are the most suitable for agrohydrological applications because of their small size and ease of siting on artificial controls, especially H flumes which have integral stilling wells. They are easier to install. They do not need special housing, unlike bubble gauge recorders and are therefore more easily re-located. Either type will measure large differences in water level.

Float and Counterweight Recorders

There are two main types of these instruments, according to the manner in which data is recorded and stored: those with electronic data loggers and those which record with pens and paper charts set on a clockwork drum. The relative advantages and disadvantages of each type are listed below. The costs of both types are similar.

Electronic

Chart

Advantages


- Compact and robust

- Widely known

- Wide range of easily set recording times

- Sometimes possible to repair locally

- Good precision

- Do not need computer facilities

- Wide range of level differences


- Long periods between visits if necessary


- Download direct to computer


Disadvantages


- Batteries can fail

- Sensitive to rough handling

- Cannot be repaired locally

- Time/level adjustments limited by

- Need computer facilities

charts and clock which can be


difficult


- More limited recording time


- Manual data entry into computer

Operation

1. WLRs with Electronic Data Loggers

In keeping with the general trend towards solid state electronic instrumentation, this type of WLR is becoming increasingly common, but the float and counterweight, mechanical aspect of these recorders is still very much the same as orthodox chart recorders. The operation of these WLRs is discussed prior to installation procedures, as it is assumed that familiarity with the equipment will be desired before selection or installation in the field.

Changes in water level are detected by a float which sits on the water level in the stilling well, connected to a stainless steel tape or wire that ascends to and over a pulley connected to the recorder. At the other end of the tape or wire which descends from the pulley a counterweight is fitted to balance the mass of the float. As the float rises and/or falls, this movement is registered via the pulley axle. The rotational movement of the axle is converted into electrical signals by an electronic integrator. These signals are passed on to the main processing unit and then to the data logger, at pre-set time intervals specified by the operator. Figure 2.41 below shows a typical electronic WLR set to an H flume stilling well.


Figure 2.41: Electronic Water Level Recorder set in H flume Stilling Well

These recorders are compact and good designs are very robust. The main processing unit is powered by dry cell (preferably alkaline) batteries which should last for a year and which can easily be replaced. In many cases the loggers are powered by integral lithium batteries which last for up to ten years, but which can only be replaced by the manufacturer. These batteries enable the loggers to be removed from the recorder without the loss of data. The electronic components of the loggers are usually resin-sealed to prevent damage. Their operation is relatively simple. Once installed, facilities are available to label the recorder number, date and time. (Typically, these are recorded as a heading prior to the water level data and can be viewed when the data are down-loaded). This information is usually displayed on an LCD screen, which is located on the processing unit and is easily revised by using various switches.

The time interval that is desired for the data to be recorded is adjusted and displayed in a similar way and provides a very flexible facility. Time periods usually range from 1 minute to 24 hours, in 1 minute steps. Changes of date are normally recorded. Readings are precise to 1 mm, but the accuracy of this sensitivity depends on the correct installation and operation of the recorder and measuring section of the control. When a replacement logger is installed, the heading information is usually written on to it automatically. The recorders are capable of recording level differences (zero to maximum) of 100 m, but tapes and wires can be purchased or cut to any desired length. Ensure before purchase, that the correct type of power batteries are easily available.

Avoid the temptation to leave recorders untended for very long periods, just because the loggers allow this. It increases the probability of undetected faults, damage by flood, vandalism and theft. Data that are lost can never be replaced. In addition to the annoyance and loss of data, the misuse of such expensive equipment will greatly reduce its cost-benefit to the project. The more frequent the visits, even to automatically recording equipment, the better, though of course each project must decide upon the priority that this activity can take..

For small catchments and plots, which will provide short periods of runoff, it is important to make the time interval between the logging of water levels short, perhaps no longer than 5 to 10 minutes. A 32 kb logger should not need to be replaced more frequently than once each week or ten days with a 10 minute record interval. For much larger catchments with longer durations of flow, half or one hour periods may be adequate. For seasonal or perennial streams, records once, twice or four times each day may be suitable. Loggers in these circumstances can remain unchanged for many months. In all cases, the most suitable time interval is a balance between these factors:

- logger memory size;
- frequency of site visits;
- duration of runoff

Logger memories vary in size, but 100 kb+ or so is typical. An example data set is illustrated below, from a recorder on a 30 cm H flume: note that no-flow data are also recorded.

Level recorder Nr. 0045

Level in mm

Repeat Period 5 (min)

D T

"90/02/07

23:05 "

P O



P O



P 3



P 7



P 11



P 11



P 55



P 110



P 114



P 107



P 86

"90/02/08

00:00"

P 65



P 33



P 27



etc.



Spare batteries and the tools to replace them should always be carried on site visits, a note of the visit and logger change should be kept. It is difficult to check batteries with a voltmeter and experience is the best indication as to how long they will last. Recorders set on flumes etc. are unlikely to require replacement floats and counterweights if treated properly, though sometimes spares are useful.

Data are usually down-loaded into computer storage by a program provided by the manufacturer. When this is done, the data in the logger is marked for erasure by new data.

2. Chart Water Level Recorders

These recorders have a relatively complicated mechanical action, though this will vary to some extent according to manufacturer, whose instructions must be closely adhered to. The float and counterweight system is similar to that describe above. Typically, the action of the pulley, as the float rises, rotates a horizontal bar along which is a sunk spiral thread. Along this thread a pen and ink carriage is moved to the right by the rotation until it reaches the end of the bar, if the rotation of the bar continues the pen action reverses and it moves to the left along a counter-spiral. In this way the pen traces zig-zags along the chart. When the float falls, the action of the pen is reversed. This allows a wide range in levels to be recorded.

Particular care is needed on two points. First, the chart must be accurately placed on the drum according to its marked, correct position. Second, the pen must be accurately placed at the zero position after the chart is replaced. The adjustments for the speed of the drum can be altered to allow longer or shorter times between chart replacement. This is effected by changing part of the gearing mechanism (provided by the manufacture according to request) or engaging different cogs, often by a lever or sliding rod. Figure 2.42 below shows an example of this type of equipment.

It is important that the ink supply is adequate and that the pen functions properly, drying can be a problem in hot climates. The timing of the clock should be monitored and corrected if necessary. Before and after field installation, the pen should be checked that it turns at the correct place on the chart. Details of recorder number, date etc. can be written onto the chart. Analysis of the data is according to the level/time pen trace and can be undertaken (for flow volumes) by digitiser or by hand.

Charts should be clearly identified with station, recorder number date and time of removal, checks with manual gauges and checks on pen reversal. Change-overs from rising to falling flow, time corrections because of fast or slow clock running marked to the nearest minute and pen relocations should be recorded in pencil on the chart at the correct point. Make sure that the pen moves freely by rotating the pulley to raise the float tape. A free pen will make a perpendicular mark which should be noted as a check. The chart should be replaced on immediate arrival so that time can be spent to check that the equipment is working correctly. It is important to instil a regular routine for each inspection.

In the case of both electronic and chart recorders, the diameter of the float and the length of the counterweight should be appropriate to the size of the stilling well. No contact with the well sides should be allowed.

Installation of WLRs on Small Artificial Controls, Flumes and small V-notch weirs

The installation of water level recorders on to such equipment as H flumes and V-notch weirs was covered earlier. This procedure is straightforward and the same for both kinds of recorder. The main points of installation are the same for all control types: the tape and counterweight should move freely after the recorder has been set horizontally, as indicated by the spirit level provided on the recorder. The details of setting the recorder heading and time period or placing the chart and pen will depend on the manufacturer's specifications, but will be broadly similar to the above.

Installation on Natural Controls and Large Artificial Controls

The installation of the stilling wells for WLRs using large artificial and natural controls, whether electronic or chart, is a costly and difficult process. On the whole, agrohydrological and water harvesting projects will be concerned with small plot or catchment runoff, but the need to install WLRs to measure larger runoff amounts may be an important adjunct to these activities. The basic requirements of installation are described here.

The need in these circumstances, is to provide a large stilling well upon which the WLR can sit. The structure should be:

- Robust enough to withstand peak flows.

- Sited upstream of the control.

- It is of great advantage if the WLR can be secured to a solid structure such as a bridge, or attached to the control by supports.

- It should be placed in a relatively protected location.

- Access should be available at all stages of the river (for example by providing steps which ascend the structure or a walkway from the bank).

- Installation is best done in the driest season.

- Perforated steel water pipe or cemented pre-cast concrete sections can be used.

- To place the well in the lowest part of the channel, (perennial) streams must be diverted to allow access to bedrock. In sand rivers, air or water jetting can be used to sink the pipe, but the danger of it being washed away remains.

- If absolute minimum flows are not required, a diversion is not necessary.

- Heavy lifting gear will be necessary.

- Artificial controls may have to be dug clear of sediment during the dry season, especially in sand rivers.

- A series of staff gauges must be emplaced as a check on WLR operation.

Advice from local organisations familiar with the installation of such structures should be sought at the earliest planning stage. Construction may be beyond the time and resources of the project and less precise estimates of flows may release valuable resources for other work.

3. Bubble gauge (servo-manometer)

Generally, most advantages lie with the WLRs described above and so this description of bubble gauges, is brief, though this type of gauge is quite commonly used in the USA.

Where rivers are subject to violent peak flows with the likely loss of WLRs, these gauges have the distinct advantage of being sited away from flood water.

The manufacturer's manual should be consulted for detailed testing, installation and operation. Bubble gauges work on the principle of depth of water exerting an opposing pressure, registered by a mercury manometer, on that exerted by a regulated gas (nitrogen) supply from a cylinder. An orifice, with a vented cap to minimise sediment entry, is fixed to the river bed from which a plastic pipe leads to the equipment. The orifice should be located below minimum expected river stage. The changes of river level are recorded by a pen, on to a chart fixed to a clockwork drum.

Important points to note are:

- The equipment must be housed in a water and vandal-proof hut away from maximum floods.

- Sediment entry into the orifice vent must be prevented.

- The orifice installation can be jetted or driven into the channel bed, but changes in bed topography may arise.

- A spare nitrogen cylinder is necessary, though the rate of gas discharge may be regulated.

- One cylinder can last many months.

- Care should be taken to ensure the gas regulator operates correctly. This must be
tested before installation.

- The chart recorders are designed specifically for this kind of manometer instrumentation

- Especial care should be taken to avoid damage to the mercury manometer

Figure 2.43 shows how the components of a bubble gauge are assembled


Figure 2.43: Bubble Gauge with Servo-Manometer

Equipment costs

All costs of locally made equipment are very approximate. The costs of raw materials and especially labour are highly variable from country to country, but a good idea of cost magnitude can be gained from the figures quoted below. The costs of manufactured equipment are based on mid-1993 prices, and where possible have been obtained from a range of manufacturers.
Shipping costs, agents fees and fluctuations in exchange rates cannot be taken into account.


Figure

Appendix A: Measurement of runoff

Appendix A1: Rating tables for H flumes, HS flumes and HL flumes

Rating Tables given in the USDA Agriculture Handbook 224 are in feet and inches. If metric measurements are required, conversions can be made using the following conversion factors. 1 inch 2.54 centimetres 1 foot 0.3048 metres 1 cubic foot 0.02832 cubic metres 28.32 litres

Because of the size of increments used in the rating tables (0.1 and 0.01 ft.), interpolation may be necessary when conversions to SI units are undertaken. Linear interpolation is permissible and does not lead to serious inaccuracies. Below are presented two conversions of rating tables to SI units, for a 30 cm deep H flume and for a 90 cm deep H flume. Together, these two sizes of flume and their rating table conversions will cover the range of discharge measurements encountered by most agrohydrological projects.


Rating Table 30 cm H Flume (litres second-1)


Rating Table 90 cm H Flume (litres second-1)


90 H Flume Rating Table continued

Rating Tables for Various Depths of H Flume (feet3 second-2)


Flume 0.5 foot deep


Flume 0.75 foot deep


Flume 1.0 foot deep


Flume 1.5 feet deep


Flume 2.5 feet deep


Flume 2.5 feet deep - Continued


Flume 4.5 feet deep


Rating Tables for Various Depths of HS Flumes (feet3 second-1)


Rating Tables for HL Flume 4 Feet Deep (feet3 second-1)

Appendix A2: Construction details of multislot dividers


Figure


Figure

Appendix A3: Construction details rotary slot dividers


Figure


Figure


Alternate Design for N-1 Coschocton-Type Runoff Sampler


Figure

Appendix A4: Rating table for broad crested (triangular weirs)

For conversions into SI units:

1 foot = 0.3048 metre,
1 square foot = 0.09290 square metre

Rating Tables for Various Cross-sectional Areas of Channel 10 Feet (3m) Upstream of Centre of Crest (feet3 second-1)


2:1 Triangular weirs


3:1 Triangular weirs


3:1 Triangular weirs continued


5:1 Triangular weirs

Appendix A5: Capacities and dimensions of parshall flumes

For conversion to SI units

1 inch = 25.4 mm
1 foot = 0.3048 metre
1 cubic foot per second = 1 second-foot = 0.02832 cubic metres per second = 28.32 litres per second


Figure

(introduction...)

This chapter covers five main topics:

- The estimation of soil loss by the application of empirically derived equations.

- Methods of measuring soil loss by the collection of the total amount of eroded material from runoff/soil erosion plots.

- Methods of sampling runoff that are carried in suspension to determine overall soil losses.

- Methods of laboratory analysis that determine the quantities of suspended soil material.

- Methods of laboratory analysis that determine soil particle size.

For practical purposes, soil erosion is regarded as the detachment of soil materials from their previous location. Sedimentation is the transport and deposition of eroded soil and although erosion due to wind occurs, the main transporting agent in almost all environments is water. Thus the theoretical calculation of soil losses are covered by estimates of erosion, whereas the actual measurement of material lost from a catchment is regarded as the measurement of sedimentation.

The erosion that leads to a wide range of environmental problems is usually the result of human activity; such as deforestation, cultivation and over-grazing, and is intimately linked with the runoff process. Agriculture is often a powerfu1 agent in promoting soil erosion and water harvesting frequently exploits the conditions that promote runoff and which can intensify rates of soil removal. The processes of sedimentation are frequently attendant.

Soil loss data are expressed in terms of weight per unit area, for example tonnes per hectare, per season. The measurement of sedimentation, the quantification of eroded soil materials, is likely to be of importance to projects that engage in the activities of agrohydrology and methods of reducing soil erosion are an important aspect of this book. A background is given to the main methods of erosion control, below, but the mechanical and constructional aspects of these controls are discussed in detail in chapter 7, Water Harvesting.

3.1 Soil erosion

The climatic factors that influence erosion and sedimentation processes are chiefly rainfall amount and intensity, which largely determine rainfall energy and runoff quantity, though the relation between them and soil loss is very complex.. Vegetation cover reduces rainfall energy and retards surface water flow, encourages infiltration through the physical perforation of soils and the reduction of the soil moisture reserve. Topography determines land slope and the length of flow of surface runoff, while the character of soils themselves, in terms of texture, structure, density, etc., influences the rate at which soil loss and sedimentation will take place.

Erosion can be summarised as taking five characteristic forms:

Rain splash erosion is the local movement of soil particles under the influence of raindrop impact. Soil particles are detached from the soil surface, elevated by the action of splash and return to the surface somewhat lower down the land slope. Areas with high slopes suffer from such erosion to a much greater extent than flat land. Large quantities of soil are removed from their original ground location by the action of raindrops. The energy equation that relates rainfall energy to intensity, developed by Wischmeier and Smith is of the form:

(kinetic) Energy, E 12.1+8.9 i where (3.1)

E is in m-Mg/ha-mm (metre-metric tonnes per hectare-millimetre) and rainfall intensity "i" is in mm h-1

Clearly such factors as wind direction, drop size, velocity and the nature of the soil will also affect the quantity of erosion that occurs and although small, clayey particles are more easily transported than larger sandy ones, they are not so easily detached from the soil surface.

Sheet erosion is a simplified term for the formation of extremely small channels or rills. These are created under the influence of rain splash and microscopic topographical variability and their wandering causes the eventual erosion of soil in the manner of a sheet. For a given soil and vegetation cover, the erosive power of the overland flow will be related to its velocity and depth.

When runoff concentrates into small streamlets Rill erosion takes place, forming small channels. These channels are clearly visible, but their size is such that they can be obliterated by ploughing. The concentration of flow into rills is particularly important because it leads to the concentration of runoff, increasing its velocity and erosive power. On high slopes and shallow soils such erosion can be destructive.

Gully erosion occurs when channel flow is sufficient to overcome tillage practices and is yet another stage of the increasing concentration of runoff, even though it may take place on an ephemeral basis. Gullies extend by the cutting back of their head and by progressive channel erosion, but stabilisation may occur naturally as the channel becomes harmonious with its slope and vegetation growth is established

Stream channel erosion is dissimilar to gully erosion in that it represents a process that occurs in channels with lower gradients, in which streams often flow continuously. Suspended material is carried along without contact with the bed, while material moved by the process of saltation bounces or skips along. The bed load is rolled or pushed along the channel bottom.

Soil erosion is a natural recycling process that has continued throughout geological time and has been responsible for the formation of sedimentary deposits that are now the major components of the continental land masses and sea floor. Tolerable rates of erosion have been suggested as being between 5 and 10 tonnes per hectare per year, but extremely different local circumstances of soil depth, formation and productivity make such values rather meaningless.

Influences on Sedimentation

Catchment Size. Generally, an increase in catchment size reduces the proportion of material removed (the "sediment delivery ratio"). This is due the increased opportunities for the entrapment of sediment within the catchment area. The presence of flood plain areas adjacent to large streams and rivers provides an environment for deposition that does not exist with smaller, steeper channels.

Topography and Channel Density. Important topographic factors that influence sediment removal are channel slope and the channel density of a catchment. The sediment delivery ratio is highest for steep channels with well defined courses, rather than low slope streams with ill defined channels. The use of channel density factors (see chapter 6) is common in the regression analysis of sedimentation data.

Precipitation and Runoff Regimes are critical factors in determining the removal of sedimentary material. Streams of a flashy nature are effective at its removal. High intensity storms not only give large runoff amounts, but also increase soil erosion by rain splash from high energy drops. Catchments that suffer from such storms display high delivery ratios.

Loss of Vegetation Cover and Agricultural Activity play a major part in creasing soil erosion.

These influences are built in to the theoretical models of soil erosion that are discussed below.

3.1.1 Theoretical Estimates of Soil Erosion

a. Universal Soil Loss Equation (USLE)

The Universal Soil Loss Equation (USLE) is widely known and was developed in many locations of the US by Wischmeier and Smith. In the 1978 publication (USDA Handbook 537), site data from US locations are given. For example, the full set of data of which Table 3.3 is a sample, covers 160 crop to fallow conversion ratios of soil loss (see Appendix B). The USLE is used to determine the value of conservation measures in farm planning and predicts non-point sediment losses. As its name suggests, it is the most widely accepted method of estimating soil loss and has generated variations that are adapted to various local conditions. Special note is made of the difficulty of applying sod-based rotations in semi-arid areas as are soil and moisture conservation opportunities of residue/mulch management (Table 3.3). However, it is important to point out that despite the simplification of the variables involved in the erosion process by the USLE, its use is often limited because the evaluation of these variables has not been achieved in many regions of the world. Those wishing to apply the USLE are recommended to obtain Handbook 537, but an outline of procedures is presented below. The average annual soil loss is given by:

A = 2.24 RKLSCP where (3.2)

A = average annual loss of soil in Mg ha-1 (tonnes ha-1)
R = rainfall and runoff erosivity index by geographical location
K = soil erodibility factor
LS = topographic factor
C = crop management factor
P = conservation practice factor

The factor R was found (under fallow conditions) to be related to the maximum 30 minute rainfall intensity and the kinetic energy of storms. The factor K in t ha-1 was assessed by measurements of actual soil loss for a series of soils with a range of physical and chemical properties. The factor LS converts soil losses from the experimental plot length and slope (22m and 9%, respectively). The conversion formulae for different slopes and values of L are:

L =(1/22)x and (3.3)
S=(0.43 + 0.30s + 0.043s2)/ 6.574 where (3.4)

x - a constant, 0.5 for slopes > 4%, 0.4 for slopes 4% and 0.3 for slopes < 3%
I = slope length in m
s = field slope in %.

C, the cropping management factor includes the effects of cover, crop sequence, length of season, tillage and storm time distribution. Conversions may be made from cropping to continuous fallow. P, the conservation practice value, discriminates between contouring, strips and terraces. Terraces alter the value of the slope length L, which becomes the terrace interval for losses from the terrace, whereas if losses from the terrace channel are required, the contour factor is applied.


Table 3.1: Values of R for Different Return Periods, Annual and Single Storms, USA

Estimates of soil losses can be made and if found to be unacceptable, the manipulation of farm management and conservation practices can be planned for its reduction: for example contour and terrace size and intervals. Although soil loss in terms of t ha-1 (with account taken of economic productivity) is the usual criterion by which the acceptability of erosion is judged, the effects of sedimentation may demand that smaller soil losses should be aimed for.

Tables 3.1 to 3.3 and Figure 3.1 give typical values for the variables in equation 3.1.


Table 3.2: Soil Erodibility Factor, K, by Soil Texture in Mg ha-1 (t ha-1) *


Table 3.3: Ratio of Soil Losses from Crops to Corresponding Loss from Continuous Fallow

Crop stages:

0 Turnploughing to seabed preparation: 1 Seedbed to first month after seeding: 2 Establishment to second month: 3 Growing cover from 2 months: 4 stubble or residue to new seedbed.

RdL = Crop residues left and incorporated by ploughing: RdR = residues removed


Figure 3.1: Factor LS soil loss for Length and Slope

Some success was obtained using the USLE at Patancheru, India by ICRISAT, though generally the slope factor was not seen to influence soil loss strongly. In South Africa, research in the Orange Veldt region, using rainfall simulator tests to validate USLE information derived largely from the USA, has shown that the use of these inputs was acceptable.
Table 3.4 below gives values for recommended conservation practices. These are greatly simplified from the complex indices given by Wischmeier and Smith (1978) which include values for rangeland, pasture, crops, woodland mulches, etc. Slope length limits are derived from work by the US SCS.


Table 3.4: Recommended Conservation Practices

Worked Example

Calculate the annual soil loss for a location with R= 310, loamy sand soil with K= 0.10 (Table 3.2), C = 0.32 (Table 3.3) length L= 60 m and slope S =12%.

From Figure 3.1, LS=2.5 and the field is to be contoured, therefore from Table 3.4, PC=0.5.

Annualsoilloss = 2.24 × 310 ×x 0.1 × 2.5 × 0.32 × 0.6 = 33.3 Mg ha-1

A reduction of these levels of soil loss would appear desirable. This must be undertaken by changing the cropping factor C and/or the conservation practice P and thereby reducing the value of these indices.

b. Modified USLE

The replacement in the USLE, of the rainfall energy factor with a runoff energy factor, has been undertaken in the USA. The equation of prediction is:

Y = 11.8 (Qqp) 0.56 KCPSL (3.5)

All components of the equation are as for USLE, except the energy factor, 11.8 (Qqp) 0.56 where Q = runoff volume in m³ and qp is the peak flow in m³ s-1.

A combination of measurements from a variety of catchments were tested against this equation and the results encouraging, however the limitations in this model must be regarded as being similar to the original USLE model.

c. Soil Loss Estimation for Model for Southern Africa (SLEMSA)

This model was developed in Zimbabwe, following disappointing results using the USLE. It is based on defined agroecological zones, their physical environments and soils. In particular, the concentration of the USLE on cropped areas and cropping systems was regarded as unsuitable in a region where rangeland conditions are very important. The structure of the model is shown in Figure 3.2 which also provides details of the model components. However, values of many of these components have not yet been determined outside Zimbabwe. Some extrapolation of the model to other countries in the region has been made, but this work has largely depended upon the direct translocation of experimentally-derived values from Zimbabwe, in particular the regression coefficients for the bare soil sub model K, and the soil erodibility factor F. Other components can be measured or determined at a location. Compared to USLE, this model may become more widely used because of the relative simplicity in obtaining empirical values.


Figure 3.2: Structure and Components of the SLEMSA Model

Empirical studies have defined the values of the parameters K, C, and X in the following terms.

E = Seasonal rainfall energy

Joules m²

Energy of mean annual rainfall

F = Soil erodibility

-

Index of soil characteristics related to known erodibilities

i - Rainfall energy intercepted

%

Vegetation cover

S = Slope steepness

%

From contours

L = Slope length

metres


Sub-models



K = Bare soil condition

tonnes ha-1 linking E and F

C = Canopy cover

-

Soil loss ratio related to i

X = Topography



Main Model



Z = Predicted soil loss

tonnes ha-1 = K·C·X

The value of index K, the bare soil submodel, is obtained by estimating rainfall energy. For example in Zimbabwe and Botswana, research has indicated that with an average annual rainfall of 550 mm, the energy received = 10, 000 joules m-² . Luvisol and Regosols were given an erosion index of 4 (the F value). The equation combining rainfall energy and erosivity was found to be:

K= exp (0.461 + 0.7663 F) ln E + 2.884 - (8.1209)F (3.6)

The canopy cover submodel converts the bare soil submodel K soil loss prediction, to a prediction for an area with vegetation, in the form:

C = exp(-0.06) i where (3.7)

i is the intercepted energy and = mean cover.

The topographic submodel, X, takes account of slope with the equation:

X= L0.5 ( 0.76 + 0.53 S + 0.076 S2) / 25.65 where (3.8)

L = slope length in m
S = slope in %

Calculations of soil loss were undertaken in grid cells of 1 km².

d. Other Models

The inability of general models to give good results under different geographical conditions has led many researchers to develop their own for particular localities. The majority of these have been regression models, whereby the dependent variable, soil loss, is regressed against a combination of independent variables. In most cases, erosivity equations giving the highest statistical correlations have been developed from the use of a rainfall /energy factor.

Some examples of rainfall/energy relations that have been examined and found to be significantly related to soil loss are given below, but the problems of site specificity are equally as important with these examples as with more general types. Variations in soils, slope and vegetation cover, as well as the characteristics of rainfall, render such equations subject to misuse. However, they do illustrate the main direction of research into soil erosion studies and the main factors in soil loss processes.

Some examples are given below:

Index

Localities

1. KE > 25 index

Nigeria, Zimbabwe

2. EI30

index USA, Kenya

3. EI5

index Zimbabwe, Kenya

4. E15

index Zimbabwe, Kenya

5. AI30

index Nigeria

6. p2/P

index Kenya

7. SUM of pi2/P index

West Africa (related to EI30)

In case I, the energy of all rainfall intensities greater than 25 mm hr-1 are summed for example:

(a)

(b)

(c)

(b)

x ( c)




Intensity (mm hr-1)

Amount (mm)

Energy (J m-2 mm-1)

Total

0-25

24

-

-

25-50

25

25.2

630

50-75

20

27.8

417

> 75

10

28.3

290




1337

Energy values for column (c) are obtained from E = 30 - 125/I

In the cases of 2, 3 and 4 (EI indices) the subscript refers to the intensity period in minutes. The energy values are obtained from E = 11.9 + 8.8 log I, then total energy is multiplied by the intensity to give an erosivity value.

In example 5, the index is obtained from the total storm rainfall (A) and peak intensity, I30. In 6 and 7, p is mean rainfall for the wettest month and P is mean annual rainfall.

Examples of EI regression equations from Kenya are:

Soil loss in tonnes per hectare

= 0.026 (EI15) - 1.18 with

R2 = 0.71 and standard

error = 4.19



Soil loss

= 0.35 (EI30) - 1 11

R2 = 0.69 se = 4.26

Soil Loss

= 0.0054 (AI15) -1.35

R2 = 0.73 se = 4.07

3.1.2 Soil Erosion Control Practices

Soil erosion practices are widely known in agricultural practice and a brief description is given of those most widely implemented.

a. Rotations

Rotations assist in erosion control by ensuring that soils are not exposed to the same risks each season. Differing crop plant covers, growing period and rooting densities, especially when periods of grass cover are incorporated, help to reduce erosion and increase the binding organic matter content of soils.

b. Tillage

Tillage helps to increase infiltration and reduce runoff and soil loss, at least for a short time. However, excessive tillage destroys the natural structure of soils and exposes organic matter to oxidation. Thus a balance must be made between sufficient tillage to achieve a good growing environment for crops and too much tillage which can lead to soil crust development and enhanced runoff. Ploughing depths should be varied to reduce the risk of a hard plough pan forming beneath the soil surface.

c. Minimum and Mulch Tillage

Minimum tillage can give better erosion control and reduce costs compared to conventional methods. Mulch tillage involves covering the soil surface with suitable residues and can reduce runoff and soil losses considerably. It tends to even out temperature differences and helps to protect the soil when plants are small and cannot do so themselves.

d. Grazing Control

The control of stocking rates can be an extremely important factor in preventing soil erosion by maintaining sufficient vegetation cover, preventing soil compaction and thereby reducing runoff.

e. Water Conservation

Water control and conservation, which covers a wide range of practices: tillage, cropping systems, farm planning and physical conservation techniques, is the most effective way to reduce soil erosion. It amounts to an integrated approach to land management.

Contouring: any farm operation carried out on the contour, ploughing, planting, weeding, etc. may be regarded as contouring. Surface runoff is reduced by the physical barriers thus formed which restrict the movement of water to small distances and low velocities. Ridging increases its effectiveness, but field with microtopography, drainage and gullies may be unsuitable, especially on high slopes. Breakage of the features formed by contouring concentrates runoff and increases soil erosion.

Strip cropping: this practice may be regarded as a type of contouring, but different crops, grass or fallow are placed alternately on the contour to increase infiltration and reduce runoff.

The design and construction of physical conservation practices (ridges, bunds, terraces and flow channels) is covered in detail in chapter 7.

3.2 Field measurement of sediments (eroded material)

3.2.1 Total Sediment Collection

Equipment and Collection of Data

3.2.2 Suspended Sediment Samplers

Equal Transit Rate Method
Depth Integrated Sampling
Point Integrated Sampling

3.2.3 Pumping Samplers

Considering the site specific limitations of soil erosion models, it is perhaps not surprising that erosion losses are still widely quantified by direct measurement. The practical aspects of the measurement of eroded material are discussed below.

As with runoff studies, the selection of site locations will be largely determined by their suitability to project objectives. From the agricultural point of view, sedimentation and the loss of soil is the main focus of attention and in many ways it is sensible to measure runoff and sediment at the same site and, where possible, on the same experimental plots. Runoff plots are usually used to measure basic erosion rates, or total sediment load, under specified soil/cover/ slope conditions, and replication will probably be necessary. Total sediment loss measurements obtained from small plots may not reflect real catchment conditions, because the natural conditions of runoff loss and redistribution are imperfectly represented. Natural catchment stream-gauging location points are usually suitable for suspended sediment sampling, as are artificial controls emplaced on catchment outlets. Little extra investment is needed to obtain sediment samples manually, though laboratory facilities must be available for soil loss determination. Where the site is remote and the cost can be justified, pumping sediment samplers are sometimes used. A wide range of sampling devices are available and may be adapted to suit local conditions. A number are described below.

3.2.1 Total Sediment Collection

Total sediment collection is obtained from the kind of runoff plot described in the section on volumetric runoff data collection systems, in chapter 2. These systems are constructed and operated exactly as described in the Runoff chapter, and the design criteria of the tanks, peak flows and total volume are calculated in the same way. The limitations of very small catchment size and regular site visits are also the same. The data from these plots are important in agrohydrological research and such erosion plots are commonly found on agricultural research stations. Arrangements for ploughing, cropping will be made and bare or uncultivated plots may be necessary to provide a controlled comparison. Natural slopes are best suited to the collection of useful data, as reshaped areas will not contain normal soil nor slope profiles.

Unlike most plots used for runoff measurement, the surfaces of plots which suffer from high levels of erosion may become lower over several seasons and it is worth considering whether or not to build in facilities that allow the collector gutter at the downslope end of the plot to be lowered accordingly. The gradient of the pipes leading to the collector tanks must be retained to prevent sediment collecting in them (a velocity of 0.6 m s-1 is adequate). This may mean lowering the tanks themselves, and during installation they should not be set in concrete, but put on stands suitable to accommodate the lowering. Over-deep excavation of their position may be necessary and this is often a laborious process necessitating the use of heavy earth moving equipment. Chapter 2 gives details of construction, installation and maintenance.

Samples taken from multislot and rotary dividers are expected to be representative of the total runoff and the following equations may be used in the calculation of sediment discharge.

GS =(QxC)/(Ax 103) where (3.9)
GS = Sediment discharge in kilograms ha-1
Q = Discharge in m³
C = Storm weighted concentration in ppm
A =Area of plot in hectares

Instantaneous sediment discharge rates can only be found if records of runoff rates are available, that is if a control section and water level recorder are installed on the catchment. If these data are available, then the formula to calculate instantaneous sediment discharge 'g' is:

g = (q × C) / 103 where (3.10)
q = Instantaneous discharge rate in m³ s-1
C = Concentration in ppm by weight for runoff rate q

Sediments collected in the conduit and tanks of the runoff measuring equipment are weighed in the field. Samples are taken as the material is weighed and the percent of dry material is determined in the laboratory. Dry weight of the deposits is given by:

Wsd = Wsw × Pdm where (3.11)

Wsd = weight of dry sediment
Wsw = weight of wet sediment
Pdm = percent dry material

Equipment and Collection of data

a. Multislot Dividers

For details of manufacture, installation and use of this equipment, see chapter 2. It is likely that some runoff events will provide no soil material, perhaps because the runoff plot is heavily vegetated or the rainfall is small. In these instances, sedimentation data will not be available.

In cases where runoff provides sediment that can be suspended by stirring the following procedure is followed:

- Agitation should be done energetically using flat paddles. It will probably take two people.

- Fill three 1 litre sample containers, using a smaller container to fill them. It is best to use 2 or 3 samples to fill each litre container so that examples from the mix are selected.

- The containers should be pre-labelled to avoid accidental confusion with date, time, plot, tank and sample no.

- The total measurement of runoff can then be taken, adding the 3 × 1 litre samples to the total runoff amount.

- It is important to ensure agitation of the mix is continuous, so in total at least three people will be necessary.

- This should be done for all tanks.

- As samples will have to be sent for analysis, it is important to check with the laboratory that the required number of samples can be dealt with conveniently. If not, it is probably best to reduce the overall number but increase sampling dips, so as not to become involved with problems of storage and the possibility that results will be late, samples lost or accidentally destroyed. To a large extent this will depend on individual circumstances, but it is an important point to note.

In cases where sediment is too heavy to be stirred totally into suspension, the following procedure should be followed to collect samples:

- In the divider system described in chapter 2, the sludge will be trapped in the first container in line.

- The supernatant (water and suspended) material should be removed to within about 2 cm of the top of the sludge.

- This should be done carefully, with no disturbance of the deposited material. It is likely that a syphon will have to be used rather than a pump, which could suck up the deposited material.

- Allow more time for this procedure than would be needed for runoff measurement alone.

- The sample bottles can be filled as the supernatant is siphoned and measured for runoff volume

- Stir the sludge aggressively until it is liquid enough to find its own level.

- Measure its depth. It is necessary to ensure that the form of the tank/container is known. This can be achieved by calibrating the volume of the container for a given depth (using water), after installation. The precision of calibration depends on the size of the tank, the advantage being with smaller tanks, for which greater imprecision of depth measurement gives smaller inaccuracies.

- Mix the sludge once more and during the process take the samples.

- Tanks and site should be left clean and free of debris.

b. Rotary Samplers

For details of manufacture, installation and use of this equipment, see chapter 2. - Heavy sediment should be collected and weighed from the collection conduit and transfer pipes. Tanks samples are taken as outlined above for multislot dividers.

3.2.2 Suspended Sediment Samplers

Suspended sediment samples are taken from water bodies such as streams and reservoirs and do not involve the wholesale collection of runoff. However, knowledge of the location of the source of material (other than it being somewhere within the catchment), is difficult to obtain.

Suspended sediment samplers are designed to sample the sediment/water mix and should ideally have the following features:

- They should sample with the intake at the same velocity as that of the stream.
- They should sample away from the disturbance they cause.
- They should be able to sample close to the stream bed.
- They should be rugged and inexpensive.

In general it is probably best to purchase specially manufactured equipment, though one set of such equipment could be used to calibrate the results of any locally made equipment capable of fulfilling the criteria listed above.

There are two main types of samplers: depth and point integrating samplers.

The former instrument takes a continuous sample as it is lowered and raised from and back to the water surface. Point samplers are equipped with an electrically-operated valve which can be opened at desired depths and samples taken. If the valve is left open, it will operate as a depth integrating device. These samplers are streamlined bomb-shaped instruments, housing a glass bottle type sampling container. The samplers can be used by wading in streams, but are often used on rivers that necessitate the suspension of the instrument from bridges or cable ways. Several sizes of sampler may be necessary where variations in flow volume are present. The USDA gives recommendations on sampler types according to speed of flow and depth (USDA handbook 224), but these relate only to equipment obtained from within the US. Suspended sediment samplers are especially important where runoff from river catchment areas is under study. Figure 3.2 shows a typical suspended sediment sampler


Figure 3.2: Suspended Sediment Sampler

The concentration of sediment within any flow is not only a function of the physical characteristics listed earlier in the chapter, but also of time of flow. In general, the rising limb of a hydrograph is prone to greater changes in sedimentation as the runoff process gets underway and therefore requires more frequent sampling. Similarly, small streams undergo a more rapid change in flow stage than large streams and these too require more frequent sampling.

Table 3.5 gives guidelines for sampling frequencies, but it is important to determine precise rates of sampling according to experience:


Table 3.5: Approximate Frequency of Sediment Sampling

Figure 3.3 shows the theoretical distribution of suspended sediment concentration in a stream section, compared to the velocity distribution. Coarse, sand-sized particles count for most of the variation in concentration, fine particles are usually fairly evenly distributed throughout the stream section. As in many cases actual concentrations vary with stage and turbulence, samples must be collected systematically throughout the stream section.


Figure 3.3: Sediment concentration and Stream Velocity

a. Equal Transit Rate (ETR) Method

This method is most commonly used on small streams on agricultural land and gives a discharge-weighted sample for the stream cross-section. Six to twelve sampling verticals spaced equally across the stream usually give sufficient accuracy. Stream discharge measurement at the time of sampling is not necessary, total flow may be estimated from gauge height and a rating curve or rating table. The collection and computation of data are relatively simple using this method, though an integrating sampler is needed and the transit (lowering/highering ) rate of the instrument must be uniform. The composite sample that is collected represents the mean, discharge-weighted concentration. Suspended sediment discharge is calculated from the mean concentration and the total water discharge. Because it is commonly used, this method is covered in detail below.

Figure 3.4 shows the path of a sampler during a typical sampling procedure:


Figure 3.4

Path of Sampler During Equal Rate Sampling Procedure

- Select a straight section of stream, with as uniform a cross section as possible, near the gauging station. Avoid shallow sections.

- Lay out a tape or line across the stream, standing downstream of the line if wading.

- Determine the position of about 6 verticals (sufficient for a wadeable stream of 10 m width)

- Record the stage of the stream.

- Rinse the sampler bottle, check for a clear orifice and stand about 1 m down stream of the line at the first position to take the sample.

- Hold the sampler rod and move the sampler down at a constant speed, touching the stream bed at each vertical.

- Sampling from surface to bed is recommended to avoid sampling disturbed bed material at higher sampling points.

- The sampling velocity should not exceed 0.4 times the stream velocity.

- When a bottle becomes almost full, replace it and mark the sequence of verticals and replacement .

- Usually between 1 and 6 bottles are needed.

- Record water temperature.

- Record stage if stream has fallen or risen significantly.

- It is convenient to put as much information on the bottle cap as possible for ease of future sorting, but in any case ETR; the stream/catchment name; date and time; stage; bottle sequence numbers; temperature and signature should be written on the bottle label.

Where sandy bed streams are being sampled, a second sample run will reduce sampling errors. Deeper streams can be sampled in a similar way, but a sampler for use with line and winch equipment will be necessary. Different nozzles will probably be available, so that the rate of collection of samples can be adjusted to such prevailing stream conditions as depth and velocity. Charts are available that indicate the speed of sampler transit and whether a single or two way transit is recommended for particular nozzles, but the details will depend on manufacturer. Errors that may be caused because of different flow velocities close to the stream banks are not serious.

Where panicle-size analysis is to be undertaken, a second sample must be taken. For all other methods, discharge must be measured at time of sampling.

b. Depth Integrated Sampling

A relatively large number of depth integrated samples must be taken on verticals at the midpoint of equal sections of the width of the stream. Usually 6 to 12 are sufficient, each located within cross-sections of equal discharge. Mean sediment concentration is found by weighting the mean concentration in each sampling vertical with respect to the discharge in the vertical.

Total suspended sediment is found by mean cross section concentration and total water discharge. Discharge must be measured at the time of sampling. Variations in sediment concentration across the stream may be obtained. Alternatively, the collection of depth integrated samples at verticals that represent the middle of sections of equal discharge may be undertaken.

c. Point Integrated Sampling

Samplers used in this method are equipped with an electrically operated valve which takes samples on command. With the valve continuously open, they perform in a manner similar to depth integrating samplers. Point samples are taken in stream verticals which represent equal or known discharge. Mean values are weighted accordingly, but the number of points depend on the physical character of stream flow.

In general, the ETR method (a.) is most widely used.

3.2.3 Pumping Samplers

These samplers are complex pieces of equipment and relatively costly to obtain and install. It is probable that their cost can only be justified if soil erosion studies are a major activity of a project and if suspended sediment sampling from streams is an important component of this activity. Pumped samples do not represent discharge weighted samples as they are point measurements. Therefore, calibration curves that plot pumped samples against discharge weighted samples (taken simultaneously) must be compiled until a known relation is established. These must be revised should the relation change with time.

They are located in a shelter at the side of the stream and take samples of the flow, a portion of which is retained. They are useful for remote locations where site staff cannot be stationed. Samples are collected and stored in bottles. An intake is placed in the stream and typically a float activated, battery powered system comes into operation at a predetermined stream level. Samples are pumped into bottles at selected, predetermined time intervals until the stream level falls or the containers are full. Most of the pumped water goes to waste, to remove any debris drawn in by the previous sampling process, samples only being taken at the end of the pumping period. It is usual to site the equipment at a gauging station so that an indicating mark can be made on the water level recorder chart, when sampling takes place. Records of stage and sampling are thus linked. Some systems rely on a gravity feed sampling and are usually sited at such locations as reservoirs and weir installations.

Bed samples may be collected from perennial streams using special sampling dredgers, sampling cores and spuds but in most cases the costs of such equipment will not be justified by the information returned. Deposits of bed sediment may be sampled by soil sampling cores for the determination of bulk density and chemical analysis, during the dry season.

The mapping and surveying of channels, changes in gullies and the upslope movement of gully scarps can be important aspects of erosion and sedimentation studies. Mapping is required in great detail

3.3 Laboratory analysis

3.3.1 Sediment Concentration

Evaporation Method
Filtration Method
Separating Fines and Sands
Particle Size Analysis
Pipette Method
Hydrometer (Bouyoucos) Method
Wet Sieving
Dry Sieving

The analysis of samples will probably be undertaken at a specialist laboratory, but this is an expensive procedure and may even involve the dispatch of samples to another country. Therefore descriptions of common analysis techniques are given below. In some cases analyses can be undertaken with relatively simple equipment and is possible even where orthodox laboratory facilities are not available, if costs can be met. Evaporation and filtration are the two usual methods of determining sediment concentration. As it is more convenient to work with weights rather than volumes concentrations are usually determined as a ratio of dry sediment weight to sediment/water mixture. Conversion to units of milligrams per litre or parts per million is undertaken afterwards.

Packing and Transport of Soil and Water Samples

The most suitable container for a soil sample is a thick polythene bag which can be sealed with tape. This can then be put into a second paper or cloth bag for extra protection. For all analyses except bulk density, there is no problem if the sample is disturbed. Samples of about 1 kg are suitable, large stones having been removed. Where gravel content is of interest the sample may be 2-3 kg. Very wet samples may be dried, but any mixture of samples or contamination should be avoided. Samples should be placed in small wooden or cardboard boxes for transport as soon as possible to the laboratory; the addition of tuline to kill organisms may be necessary if samples are to be tested for nitrate and cannot be delivered promptly.

Water samples should be carried in screw-top polythene bottles and placed in boxes that are fitted with sections to separate each bottle from the next. Any empty space in packing cases or boxes should be packed with wadding to prevent movement of the samples during transport. Samples of 1 litre are adequate and soda glass containers should be used if an analysis for boron is to be undertaken.

Water and soil sample containers should be clearly marked, at least twice, with a water proof pen. So should any outer container. Details of samples should be recorded in a sample book and generally the less detail on the sample the better, to avoid confusion, however a detailed packing note should accompany any container of samples. It may be necessary to check with the laboratory in case it has a preferred system of labelling.

3.3.1 Sediment Concentration

a. Evaporation Method

Basic equipment is as follows:

- Graduated containers 0.5 -1.0 litre capacity.
- Large container, 5 litres or more
- Distilled water
- Vacuum source
- Evaporation dishes of several sizes
- Convection drying oven
- Pipette
- Desiccator
- Flocculating agent
- Balances accurate to 0.1 gram and 0.1 milligram

The procedure is as follows:

- Prepare a worksheet (see Figure 3.5)

- Transfer all sample details from the bottles to the work sheet.

- Weigh the total sample, less the container(s) weight(s) and record.

- If the colloidal material is in suspension, flocculate adding 0.40 millilitre of 0.2 molar solution of alum (90.7 g l-1).

- Allow the sample to settle overnight or for at least 12 hours.

- Note that flocculant is not usually used for concentrations < 1000 milligrams per litre. If this is done then the introduced error can be calculated for example: 50 milligram per litre solution added to a concentration of 1000 milligrams per litre will give an error of 5%, if all the flocculant is sorbed by the sediment.

- If samples have relatively little colloidal clay, they can be allowed to settle for a few hours and no flocculant is needed.

- Using the vacuum source and appropriate tubing, remove all water except 30 millilitres from the sample. This amount can be approximate, so long as it is the same amount for all the samples. All the effluent from the samples from one sampling site can be combined in one large container.

- Wash the remaining sample into a numbered evaporation dish with distilled water.

- Place in the oven and dry overnight at 105 - 110°C. It is best to avoid vigorous boiling and splashing of the sample. The sample container can be used where possible.

- Remove from the oven and place in the desiccator, weigh as soon as possible after removal.

- Enter gross and tare weights and obtain net weight of sample on the worksheet.

- Mix effluent thoroughly and withdraw 100 millilitres. Oven dry at 105 - 110°C, weigh sample enter on the worksheet.

- Compute the correction factor for the effluent by dividing the net amount of dissolved solids (the residue) by the aliquot volume (in this case 100 millilitres) and multiply the volume of water left in the large sample container × 1 million. (0.3888 g / 178.1 g) × 106 = 2183 ppm

To compute the concentration in milligrams per litre:

Concentration = B × ( weight of sediment × 106/ weight of water-sediment mixture) in mg l-1.

The value of the factor 'B' can be obtained from Table 3.6 which is based on specific weights for water and sediment of 1.000 and 2.65 g cm-3.


Figure 3.5: Example Worksheet for Evaporation Method of Sediment Concentration

Compute sediment concentration (in parts per million) as follows. Figure 3.5, the worksheet, provides the example:

Subtract correction factor from the net sediment weight. In Figure 3.5 for example.

0.3930 g - 0.0042 g = 0.3888 g

Divide oven dry weight of the sample by the net sample weight of the sediment plus water and multiply by one million.


Table 3.6: Values of Factor 'B' for the Computation of Sedimentation Concentration in mg l-1 When Used with Ratio (x 106) of Weight of Sediment to Weight of Water/Sediment Mixture (0-29°C)

b. Filtration Method

The filtration method works well with low sediment concentrations and obviates the need for the dissolved solids correction. Compile a work sheet as shown in Figure 3.6

Equipment is as follows:

- Got crucibles, at least 25 millilitre capacity with perforated bottom, suitable to be fined to a vacuum system.

- Filters. Commercial glass fibre or cellulose are satisfactory for most sediments.

- Distilled water

- Vacuum system

- Evaporation dishes of several sizes

- Convection drying oven

- Desiccator

- Flocculating agent

- Balances accurate to 0.1 gram and 0.1 milligram

Set up as follows:

- Determine the weight of the sediment/water mixture (the sample).

- Allow to settle until clear then decant the excess liquid into another beaker.

- Install suitable filter into crucible and determine tare weight.

- Connect crucible to vacuum system and transfer sample.

- When filtration is complete place crucible into oven and dry at 105 -110 °C.

- Remove and place in desiccator. Remove and weigh and compute concentration. Other methods of filtration can be used.

- A simple glass funnel fitted with a filter paper, the sample draining under gravity can be used when samples have a relatively high colloidal content and if care is taken.

Procedure:

- Follow the first three steps of the evaporation method.

- Wet sieve the material using sample water.

- Remove the material < 0.062 mm.

- Dry and weigh this material.

- Remove the material > 0.062 mm (sands) and put in a tared evaporation dish.

- Oven dry and weigh.

- Adjust to pH 3 - 5 with HCl, using pH paper.

- Add about 1 millilitre of 30 percent H2O2 (hydrogen peroxide) per gram of dry sample, in 40 millilitres of water.

- Allow to stand to oxidise the organic material (this will take a few hours) and remove any floating organic material.

- Destroy any remaining organic material and hydrogen peroxide by bringing the sample to a boil.

- Oven dry sand sample and weigh.

- Determine the organic content by subtracting the gross weight of sands from gross weight of sands before peroxide treatment and record.

- Record gross and tare sand weights on worksheet, compute net sand weight and record.

- Subtract weight of organic matter from weight of sand and record. No correction for dissolved solids is necessary as the effluent was washed into the fines portion during sieving.

- Compute sands and fine concentrations as described in the Evaporation method. Total concentration is equal to sum of fines and sands.


Figure 3.6 Example Worksheet Filtration Method

c. Separating Fines and Sands

The separation of fines and sands is frequently used in analysis and is normally made at 0.062 mm, though different preferences can be catered for. The equipment is the same as the Evaporation method, with the addition of a 0.062 mm (or 0.053 mm) meshed sieve. Figure 3.7 gives an example worksheet.

Procedures:

For the sands:

- Follow the first three steps of the Evaporation method
- Wet sieve on the 0.062 mm sieve using sample water.
- Remove material < 0.062 mm. Dry and weigh to the nearest 0.1 ma.
- Adjust to pH 3 -5 with HCl using pH paper.
- Add about 1 millilitre of 30 percent H2O2 per gram of dry sample in about 40 ml of water.

Allow to stand to oxidise the organic matter.

- Destroy hydrogen peroxide by boiling
- Oven dry sample and weigh to nearest 0.1 ma.
- Determine organic content by deducting weight of sample after from before hydrogen peroxide treatment.
- Record gross and tare sand weights, compute net sand weight.
- Subtract organic matter weight from sand weight and record.
- Compute sand and fines concentrations as for Evaporation method.
- Total concentration is found by totalling concentrations of sands and fines.


Figure 3.7: Example Worksheet Sands and Fines


For the fines:

- Continue with the steps of the Evaporation method

3.3.2 Particle Size Analysis

Particle size analysis is undertaken not only for sedimentation work, but also to determine soil textural type. Several methods may be employed because of the wide range of particle sizes frequently present in samples. Table 3.7 gives size range, analysis concentration quantity of sediment and methods of analysis, recommended by the USDA SCS. Table 3.8 gives a grade scale of sediment particle sizes.


Table 3.7 Recommended Particle Size Analyses


Table 3.8 Soil Textural Classes and Particle Sizes

The most commonly used methods of analysis are:

Fine sediments: Pipette, Hydrometer and Bottom Withdrawal (B W) Tube Methods. The former two methods are most commonly used and are detailed below.

Coarse sediments: Sieving and the Visual Accumulation Tube Methods. Details of these methods are given below.

In many cases, suspended sediment samples do not contain sufficient material for accurate analysis and thus procedures must necessarily be limited to the separation of fines and sands, though the possibility of combining samples from the same runoff event could be considered where practicable. Soil samples, by comparison, usually require splitting to the required sample amount before analysis. Dry samples can be split by a commercial sample splitter; moist samples by quartering or by extracting two or three samples from a thin tube inserted to the bottom of the material.

Fine Samples

a Pipette Method

This method is based on the principal that thoroughly dispersed particles of a given size will settle below a withdrawal point in a given time, according to Stokes law which is defined by:

2(D1-D2)GR2/9p where (3.12)

D1 = density of particle in g cm-3
D2 = density of the liquid in g cm-3
G = acceleration due to gravity in cm s-2
R = radius of the particle in cm
P = viscosity of the liquid in g cm-1 s-1

Equipment is as follows:

- 25 millilitre pipette apparatus (see Figure 3. 8 below)

- A vacuum source.

- Sedimentation cylinders of 1,000 millilitre capacity with rubber bungs.

- Stirring rods, brass of 6.4 mm diameter by 61 cm with a perforated plastic disc 5 cm in diameter, attached to end.

- Thermometers, Evaporating dishes.

- Desiccators, Stopwatch.

- Worksheets, 0.062 mm sieve.

Predetermined depths and times of withdrawal are given below in Table 3.9, based on the assumptions that particles have a spherical shape and a specific gravity of 2.65. The viscosity of the fluid is assumed to vary from 0.010087 cm²s-1 at 20 ° C and 0.008004 cm²s-1 at 30 °C. The gravitational constant is 9.80 m s-2.

Procedure:

If an organic matter and soluble salts content is required, the supernate is removed, the sample dried and weighed before these items are removed.

Otherwise the first step is to remove any organic material by oxidation.

- Use HC1 to adjust the sample pH to 3 - 5 for oxidation.

- Add about 1 millilitre of 30% H2O2 for each gram of dry sample. Stir and allow to stand for several hours.

- Any floating material can be removed.

- Usually samples need to be heated (less than 70°C) and more H2O2 should be added.
- When the reaction has stopped, the sample is boiled or washed with distilled water to remove the H2O2.

The second step is to remove soluble salt material.

- Effervescence, evident when a little dilute HCl is added to the sample indicates the presence of carbonates.

- To remove them, add 50 millilitres of a slightly acid sodium acetate (Na O Ac) buffer solution ( Na C2 H3 O2 3H2O,1N, 136 g l-1 adjusted to pH 5 with acetic acid) to each 5 grams of sample.

- Bring to suspension by stirring with a rubber tipped glass rod. Digestion is helped by heating the beaker in a water or sand both at near boiling temperature.

- After 30 minutes the suspension is washed by filtering with a filter candle (= to Pasteur-Chamberian or Selas type FP porcelain candle, 02 or 03 porosity). Some samples may need two or more treatments.

- After salts and organic matter have been removed, use the filter candle to remove excess liquid.

- When the candle is coated with soil, reverse the stopcock and add pressure with the rubber bulb.

- Touch the candle with inner surface of the beaker to remove any soil. Repeat the filtration and soil removal process.

- When free water has been removed, mix the sample with a jet of distilled water.

- Repeat the filtering and mixing process several times and when complete, add pressure as before to dislodge as much soil as possible and wash the soil back into the beaker. The rubber tipped glass rod can be used as an aid to this.


Table 3.9: Time of Pipette Withdrawal for Given Temperature, Depth of Withdrawal and Diameter of Particle


Figure 3.8: Schematic Diagram of Pipette Method Equipment

Thereafter analytical procedures are as follows:

- If both concentration and particle size are needed, weigh the water/sediment to the nearest 0.1 gram before proceeding.

- Remove excess water with a filter candle and place sample in a tared evaporation dish, dry at 100 -110° C and weigh to the nearest milligram.

- Deflocculate by adding a dispersing agent (40g sodium hexametaphosphate (Na P03)6 and 8 grams sodium carbonate in distilled water to 1 litre) for each 5 to 10 grams of sample.

- Transfer to a 250 millilitre shaker beaker adding distilled water to bring the volume to 180 millilitres, shake overnight. More convenient is the use of a mechanical analysis stirrer which will complete the mixing in 2 -5 minutes.


Figure 3.9: Worksheet for Pipette Method

In some cases, for example where the concentration of suspended sediments is very low, dispersion may not be necessary. Where it is, the dissolved solids correction must be determined for each new solution of the dispersing agent as follows:

- Add 10 millilitres of dispersing agent to a calibrated sedimentation cylinder, dilute to volume with distilled water, mix thoroughly and remove 25 millilitres, transfer to an evaporation dish and dry overnight at 105 - 110 ° C, weigh the dish and contents to the nearest 0.1 milligram.

- Perform in triplicate and use the average. The net weight of the dissolved solids is subtracted from the net weight of each pipette withdrawal.

Continue as follows:

- Weigh each sedimentation cylinder while empty then fill to between 500 and 1000 millilitres with distilled water. Weigh again several times and use the average. Do the same for the 25 millilitre pipettes. The ratio of mean weight of water in the cylinder: mean weight of water in the pipette is used as a volume ratio in comparing the results of the withdrawals.

- Select particle size determinations and using Table 3.9 set up a schedule for time and depth of withdrawals.

- Use distilled water to wet sieve (0.062mm) the dispersed sample, passing material into a sedimentation cylinder. Place the sands into a tared evaporation dish and dry, weigh to the nearest 0.1 milligram for the net weight.

- When ready to pipette, bung the cylinder and shake vigorously while turning end over end then plunge with the brass stirring rod.

- Immediately lower the pipette 10 cm into the sample and take a "zero time" withdrawal. Take the temperature and plunge again for 1 minute. After this make withdrawals according to the depth/time schedule, always measuring the sampling depth from the existing surface of the suspension.

- Each time, the pipette is flushed with distilled water and with the withdrawn sample, this is put into numbered and tared evaporating dishes. A rubber bulb may be used to blow out remaining droplets.

- Oven dry the withdrawals overnight (100 - 110 ° C) cool in a desiccator and weigh to the nearest milligram to determine net weight. Results may be tabulated as in Figure 3.9.

Calculations are carried out as follows:

- From the "zero time" withdrawal determine the net weight of fines in the sample and record. Make a dissolved solids correction if a dispersing agent was used. Compute the total weight of fines by multiplying the weight of fines in the suspension by the volume ratio.

- Determine the net dry weight of the sediment in subsequent withdrawals and multiply by the volume ratio. Note: this gives the weight of sediment in suspension finer than the size corresponding to the time and depth of withdrawal.

- To obtain the fraction of total sediment finer than the indicated size, divide the weight of sediment in the sample finer than the size corresponding to the time and depth of withdrawal by the dry weight of the total sediment in the sample.

- To obtain the concentration of the fines, sands and total sample in parts per million, divide the total net weight of each by the weight of the total sample (water/sediment mixture) by one million. Record on the form.

b. Hydrometer (Bouyoucos) Method

The density of the soil suspension is measured with a special hydrometer which is marked with percentage calibrations, calibrated at 20°C. The technique as proposed by Bonyoucos does not remove organic or calcium carbonate material and therefore gives approximate results where these are present in large quantities.

A second hydrometer (ASTM 152 H) was developed by Day with a more rigid adherence to Stokes Law.

The original hydrometer method was devised to provide a quick and easy method, the accuracy of which could be established by comparison with pipette analyses and thereafter be used with confidence. The modifications introduced by Day increase accuracy, but the analysis is no longer rapid (it takes about as long as the pipette method) and it is essential that organic matter and calcium be removed by pretreatment.

Bouyoucos Method

Equipment:

Apparatus and reagents as for pre-treatments and separation of sand
Mixing plunger
Thermometer including the 15 - 25 °C range
Accurate clock or stop watch
Bouyoucos or ASTM hydrometer
Hydrometer jars marked at 1 litre

Procedure:

- Estimate whether sample is sandy (silt and clay < 15%) or not sandy. In first case transfer 100 g oven dry sample to 250 ml beaker and add 100 ml 5 percent solution of hexametaphosphate-sodium carbonate. In second use 50 g.

- Rest overnight, transfer to mechanical stirrer, washing out the beaker and making up the volume to 500 ml with water.

- Stir for 2-3 minutes

(- If sample contains much carbonate or organic material, remove by treatments described previously. Dry and weigh.)

- Transfer suspension to hydrometer jar, wash out stirrer cup and adjust volume to 1 litre. Mix with glass rod.

- Take temperature, which should be 15 - 25 °C. Mix with plunger.

- At selected times lower Bouyoucos hydrometer carefully into centre of solution. Read the scale to nearest 0.5 unit.

The times advocated by Bouyoucos are:

Particles less than

50 microns

40 seconds


20 microns

4 minutes


5 microns

1 hour


2 microns

2 hours

In the case of particles < 2 microns, 6.5 hours may be more accurate for samples with organic matter removed. - Prepare a 0.5 percent solution of sodium hexaphosphate-sodium carbonate, transfer to a hydrometer jar and adjust to 20°C. Insert hydrometer and take reading as a blank. Discard solution. - If the temperature of hydrometer readings was not at 20 °C, the following corrections apply:

Temperature (°C)

Correction (g per litre)

Temperature (°C)

Correction (g per litre)

15

-2.0

21

+0.5

16

- 1.5

22, 23

+ 1.0

17, 18

- 1.0

24

+ 1.5

19

- 0.5

25

+ 2.0

- Subtract the blank reading obtained with the dispersing solution at 20 °C.

The values obtained are direct percentages of clay (< 5 or 2 micron according to time) or silt + clay (< 50 micron or 20 micron according to time) if 100 g were used. Values to be × 2 if 50 g were used.

Coarse Sediments

Both wet and dry sieving procedures may be followed for material greater than 0.062 mm. In each case it is essential not to overload the sieve, the USDA recommends 25 - 50 g for 8 inch diameter sieves for medium - fine sands (3 -7 g for 3 inch sieves) and 10 -20 g for fine sands ( 1.5 - 3.0 g for 3 inch).

Wet Sieving

- Immerse the sieve having the coarsest screen in a ceramic dish containing distilled water until the water surface is about 5mm above the sieve.

- Wash the sample on to the sieve and shake until all small particles have passed through.

- Pass the material and washing water onto the next smallest sieve and continue to repeat the process until the smallest sieve is reached.

- Transfer the material on to a tared container, dry and weigh each fraction. Any material passing through the 0.062 sieve is to be analysed by other methods

Dry sieving

- Set a nest of sieves on a mechanical shaker, coarsest on top, proceeding to the finest.
- Weights for each fraction are determined after about 10 minutes of shaking.

Equipment costs

All costs of locally made equipment are approximate. The costs of raw materials and especially labour are highly variable from country to country, but a good idea of cost magnitude can be gained from the figures quoted below. The costs of manufactured equipment are based on 1993 prices. Shipping costs, agents fees and fluctuations in exchange rate cannot be taken into account.


Figure

Appendix B: Erosion and sedimentation data

Appendix B1: Ratio of soil loss from crop land to corresponding loss from continuous fallow


Appendix B1 part I


Appendix B1 part II

4.1 Rainfall

The collection of snowfall data is not covered here. If detailed information is needed, consult a handbook such as USDA No. 224 or approach the local meteorological service. Ordinary, standard gauges with the collection/funnel component removed and a measured amount of antifreeze added can be used to measure snowfall, but errors due to wind effects can be very large.

Rainfall is the most important single factor in determining whether runoff will or will not occur for a given set of environmental conditions. It determines runoff amount and frequency. There are two measurements of rainfall amount that are commonly collected for hydrological purposes: Daily and (runoff) Event rainfall.

Daily Rainfall is probably the most ubiquitously measured meteorological variable. It is the rain that falls awing a 24 how period starting in the morning of one day (commonly 06:00, 07:00 or 08:00 furs) until measurement is made at the same time the following day. Event Rainfall by contrast is the rainfall occurring awing an unspecified time period usually, but not always less than 24 hours, that can be seen to be responsible for subsequent runoff. The collection and use of each has advantages and disadvantages.

In the case of daily rainfall, data are usually available from many stations, even in countries with only the most basic meteorological network. The equipment to measure daily rainfall is relatively cheap, simple to install, read and maintain. All projects should easily achieve adequate instrumentation. In most cases, many years of historical data will be available for analysis from a variety of sources, in addition to that obtained from meteorological offices: these sources include various government departments, water resource and construction projects, state and private farms, schools and interested individuals. Often basic analyses will have been performed on the data (average monthly and annual totals, spatial distribution, etc.). For the analysis of runoff relations, however, daily rainfall can have one serious drawback. It is the lump sum rainfall awing a 24 how period and in some climatic environments may greatly exaggerate the amount of rainfall thought to be responsible for runoff, but despite this drawback, it is the most commonly used climatic variable in runoff studies.

Event rainfall, obtained from the careful examination of the records of an automatically recording rain gauge, can provide a precise and accurate evaluation of the rainfall responsible for runoff and it is often to be preferred for rainfall/runoff analyses. However, recording rain gauges are not usually in widespread use except at important synoptic stations (especially in developing countries). They are expensive to buy, can be difficult to maintain and staff must have a higher level of expertise to operate them. The analysis of data is more complex and time-consuming.

It is possible to determine whether or not daily rainfall and runoff event rainfall are for all practical purposes, the same, though a number of historical data are necessary to do this. Values of daily rainfall are plotted against values of runoff event rainfall, as illustrated in Figure 4.1. A visible correlation exists for these data from sites around Gaborone in SE Botswana, confirmed by the statistical significance of the coefficient of determination of the relation (R2 = 0.99) The gradient of the line which is for all practical purposes 1:1, the regression equation being y= 1.025x + 0.068. In this case, daily rainfall values could be used instead of event rainfall, without fear of inaccuracy. The ratio of Daily to Event rainfall may not be 1:1, but any relation that is highly significant can be used. Where the relation is not significant, daily rainfall cannot be used to estimate the actual rainfall that caused runoff. The similarity between daily and event rainfall depends on climatic regime but it will be more evident in areas that have convective rainfall, giving heavy, short periods of rain and many thunderstorms. The analysis of rainfall data is discussed in greater detail in Chapter 8.

Rainfall Intensity defines the amount of rain falling during a specified time within the most intense period of the rain. This value is then converted into the amount of rain that would fall in one how at this intensity. For example if 5 mm of rain falls in a 2 minute period, the 2 minute duration rainfall intensity is 150 mm hr-1 and if 27 mm falls in 30 minutes, the 30 minute duration rainfall intensity is 54 mm hr-l. Rainfall intensity can have an important effect on runoff proportion, as it determines the rate at which rain arrives at the soil surface and, consequently, whether the infiltration rate of the soil is sufficient to allow absorption. Automatic recording gauges are needed to measure rainfall intensity The duration under study will be determined to a large extent by catchment size; small plot runoff is often closely related to the 2 or 5 minute intensity durations while large catchment relations are more evident with long durations. The use of daily rainfall becomes a more attractive proposition under these circumstances, especially as the use of long duration intensities can lead to a serious reduction in data, when rainstorms are short-lived.


Figure 4.1: The Relation Between Daily Rainfall and Event Rainfall

4.1.1 Non-Recording (Daily) Rain Gauges

All agrohydrological projects will collect daily rainfall from manually read rain gauges. Even when automatic recording gauges are installed it is important to cross-check the amounts that they measure and ensure that at least daily values are collected if they develop faults. Some automatic gauges have integral manual gauges sited under the recorder to facilitate this. Different models of daily rain gauges are used in different countries. In the case of some developing countries, which can ill-afford even basic meteorological equipment, national networks may use more than one type, having been supplied by different aid agencies or having inherited them from various projects. Usually this does not cause serious problems. It is essential however, to ensure that the correct measuring vessel into which the rain is poured from the gauge (and calibrated only for one particular model) is used to measure the rainfall amount. As a general rule, it is sensible to purchase equipment compatible to that already used by the national Meteorological Service. This has several advantages:

- Data will be strictly compatible.

- In the case of loss or breakage of measuring vessels (glass), a temporary loan may be possible.

- The Service will be familiar with procurement/replacement procedures.

- The Service will be familiar with the most suitable methods of installation.

- Service staff and other gauge readers who may be called upon to read the gauges, will be familiar with the equipment. In some cases it may even be possible to arrange the loan of gauges, if the Service has sufficient reserve.

Types of Gauge

Any open ended vessel can be used as a rain gauge, but uniformity of design to provide consistent splash characteristics necessitates the use of purchased equipment. Most gauges are made of corrosion-resistant metal such as brass.

Gauges with orifices of different sizes measure rainfall with about the same degree of accuracy. Results of tests show that readings are within 1% from gauge openings of between 5 to 50 cm. Differences in measurement usually result from installation and reader error rather than gauge design.

The standard US Weather Bureau non-recording gauge has an 8 inch (20.3 cm) orifice, the UK standard gauge (commonly adopted by former colonies) has a 5 inch (12.7 cm) opening. There is little to choose in design accuracy, but in general, smaller gauges tend to be less expensive. Figure 4.2 shows examples.


Figure 4.2 Non-recording Rain Gauges.

Non-recording gauges usually consist of a collector above a funnel which passes collected rain into a receiving vessel. Important requirements are that the collector walls should be vertical inside and steeply bevelled outside. It should prevent rain splashing in or out by having a sufficiently deep wall and a funnel with steep sides (at least 45 degrees). The area of orifices should be consistent. The receiving vessel should have a narrow neck (to prevent evaporation losses). It is usual to use a larger vessel in the same gauge, where it is impractical to visit the gauge on a daily basis.

The measuring (calibration) vessel should be of clear glass with engraved graduations (usually at 0.1 mm intervals). The type of gauge that it is to be used with, should be clearly marked. To achieve accurate readings for small rainfall amounts, the base will be tapered. Dip rods are sometimes used instead of measuring vessels, but this is unusual.

Plastic rain gauges, often inverted cones in shape and marked with mm gradations, are available in some countries and can provide a good, cheap alternative to expensive standard rain gauges. However, three important facts should be recognized.

- They eventually degrade due exposure to UV light, after one or two seasons.

- They cannot be recommended in areas of frequent hard frost.

- Not all such gauges are produced to accurate specifications and the accuracy of the gradations should be checked and if necessary, calibrated before use.

- They are difficult to install with the orifice exactly horizontal, using a single post.

Installation of Non-recording Gauges

The location of the gauge is the primary consideration in obtaining accurate rainfall measurements and the most serious problem is wind turbulence. Buildings, trees, fences produce eddies and reduce accuracy. Isolated obstructions should not be closer than twice their height to the gauge (further away if possible). However, openings in woods and orchards are suitable places (so long as the trees are no closer than specified); they act as windbreaks and reduce violent air currents. Sloping ground should be avoided and surrounding vegetation should be cut low. In general, smooth artificial surfaces are not suitable as they tend cause splashing and may attain high surface temperatures.

It is very important that the height above ground level of gauge orifices should be the same at all sites. It is preferable that the orifice be as close to the ground as possible. Wind velocity increases with height and the catch of the gauge is reduced thereby underestimating rainfall, however gauges positioned at ground level are prone to in-splashing, which is also undesirable. The UK Meteorological Office recommends that gauges' orifices be located at 305 mm above ground level as a compromise to minimise both of these effects. The influence of wind on rain catch can be expected to be greater in areas of convective rainfall where air turbulence is inherent.

In difficult field conditions however, the practical limitations of accidental disturbance and vegetation growth can result in problems with gauges set close to ground level. From experience, a suitable height is between 0.8-1.2 m. In this range of heights the gauge should be clearly visible and vegetation growth between visits from project staff should not affect readings. A consistent height for all gauges based on local field conditions is probably more important than following hard and fast recommendations. Ground-level gauges can be seen at meteorological stations, but here constant attention can be given to vegetation clearance, animals should not be a problem and checks can easily be made on the equipment. This may not be so at field sites.

The stand supports that hold the gauge should be inexpensive, strong, durable, rigid and easy to build from local materials. Wood is best avoided due to susceptibility to termites and rot. Materials should be easy to transport into the field. Although a concrete base in which stand legs can be set or bolted seems sensible and is often advocated, in areas where animal damage is possible, it is better to install the stands with metal pegs or by digging the legs into the ground to an depth of 50 cm, which is usually adequate. This may lead to the toppling of a gauge if pushed by cattle, but it can be reset immediately and will not be out of operation until the stand is replaced and re-cemented. In all cases, the gauge should be set horizontally by testing with a spirit level.

Gauges may often be placed on private land and it is essential to obtain permission from the owner before doing so. Usually a discussion as to the aims of the project, the (realistic) long-term benefits that are hoped for, and the purpose of the gauge are sufficient to obtain permission. Sometimes and if suitable, land owners can be recruited as gauge readers. It is important to provide an enclosure for the rain gauge to reduce the risk of damage from animals and to prevent interference with normal land use. Barbed wire is most suitable, with strands placed closer together (10-15 cm apart) at the bottom. Not only does this prevent access to animals and deter vandals, but also has no discernible effect on rainfall catch. Cut thorntree branches (e.g. Acacia species) woven into the lowest metre of the fence are a good deterrent against goats and other agile animals, although in areas of termite activity this will need to be replaced each season. For the same reason, enclosure posts are best made of metal, though they are more expensive than wood. If they are too costly or unavailable, wooden posts will last several seasons if soaked in creosote or other preservative. In some areas, termite-resistant tree species can be found for use.

Even though wind turbulence is important in reducing rain gauge accuracy, the use of windshields can generally be discounted. Their main aim is to reduce wind speed, turbulence and splash over the gauge and allow siting close to the ground, but in practice they are not easy to make locally and increase problems of interference by people and animals. Research has shown that differences in shielded and unshielded gauges are relatively small (2-8%).

When locating rain gauges in the field it is important to keep in mind the problems of theft and vandalism. Prevention can be aided by the provision of fencing and enclosures, but the safest action is to enlist the help of local people. Sites placed by the homes of gauge readers are usually convenient and secure. When positioned in more public places, such as common village land, it is advisable to discuss the purpose of the equipment at an open meeting, where the use of rainfall data can be explained. An agreement for safe custody of the equipment can (usually) be made without difficulty.

The UK Met Office and US Weather Bureau recommend that rain gauges in exposed positions be located within a circular turf wall to negate the effects of wind turbulence. Figure 4.3 illustrates the construction of the protective wall which should be kept in a good, clean condition. Note that the rain gauge orifice is level with the top of the wall. It should be stated that in many circumstances, recourse to such a structure will be very difficult.


Figure 4.3 Construction of Turf Wall to Remove the Effect of Wind

Non-Recording Rain Gauges: Routine Data Collection and Site Staff

At field sites that do not have automatic recording gauges, locally-hired gauge readers will almost certainly be required and it is essential that a good working relationship be established and maintained with such gauge readers, observers or site guards. More difficulties in the collection of field data arise from people than from instruments and reliable field staff are an invaluable asset to any project. The selection of suitable field staff can be difficult and common sense and judgement of personality are important criteria in arriving at a suitable choice. The opportunities to recruit will vary from country to country and will depend on social as well as educational factors. The following points can help.

- The person should be reliable and suited for the job. They should be literate and numerate at least to a basic level. It is unlikely that they will need skills sufficient to understand instrument manuals (this can actually be a disadvantage if they are "zealous" in the performance of their duties), but they will be using printed record sheets and measuring and recording numerical values.

- Village school teachers are often used as recorders of rainfall, by national meteorological services. However, they have other responsibilities and often take holidays away from their village. Rainfall records often suffer as a result.

- If possible, it is useful to have someone who can speak your own language, or an interpreter will always be necessary.

- A permanent resident is preferable, even if less educated. Their new responsibilities should not be in conflict with their normal day-to-day business.

- Be aware that the recruitment can give rise to discontent within the family/community by imparting status and/or financial gain to the individual.

- It is important to resist the recruitment of individuals who may be using their status within the community to acquire more and/or who expect financial gain, without being prepared to carry out the work properly.

- The gauge reader should be able to reach the gauge easily, therefore they should live close by and have permission for access if it is sited on someone else's land.

- After selection, the reader should be given a thorough explanation of why the data are being collected and the importance to the village and project.

- They should be tested for competence in all their responsibilities.

A contract should be drawn up covering all duties and rewards in a simple but comprehensive list. For example:

- Reading the gauge at specified times (check that this can be done, or at least that the time of reading can be noted).

- Keeping the record book and measuring vessel at a specified place so that they can be checked on field visits.

- Cleaning the gauge when necessary.

- Noting damage to the gauge, enclosure and keeping down vegetation etc.

- Repairing any such damage when possible.

- Keeping the site clear of vegetation on a routine basis.

- Informing in advance the use of the deputy and expected departures from site.

- Basic pay, holidays or pay in lieu.

- Deductions or action in the event of neglect of duties.

- Best method of contact in the case of unforeseen events.

Any seasonal changes in these duties should be stated, as should a clause covering unsatisfactory performance. A suitable deputy should be appointed as it is inevitable that at some time the gauge reader will have to leave the site. The contract should be translated into the local language if necessary. It is best if payment is made for the services of the gauge reader. This puts the arrangement on a business footing and no favours are asked by either side.

Payment should reflect local employment conditions, bearing in mind that while the work is not arduous, it does restrict the mobility of the gauge reader. All payments should be recorded, copied and signed for without failure. Most of these points will be influenced by local conditions, but a friendly involvement by the project field staff in the reader's work coupled with a direct and business-like approach seem to get best results.

The maintenance of non-recording gauges is straight forward and should be carried out by the gauge reader. The gauge should be checked for any blockage (cobwebs, insects, leaves etc.) in the funnel at each reading by holding the orifice component to the light. Even partial blockages should be cleared. The collection vessel should seen to be watertight and clean and the felt washer on the funnel of the gauge which fits over the vessel mouth (and helps to prevent evaporation) intact. The measuring (calibration) vessel should be clean and in good repair. It is useful to have a spare at site as these are made of glass. It should be ensured that the gauge is correctly seated on or in its stand, often gauges are disturbed when the tightly-fitting collection component is removed to measure the rainfall


Figure 4.4: Routine Data Collection Sheet, Non-Recording Rain Gauges

The stand should be checked for stability and deterioration and repairs effected if necessary. The area around the gauge should be kept clear of vegetation and the enclosure fence kept in good condition. Any repairs to or re-instatement of the gauge, including the date of original installation should be noted in the record book. The field team should carry a book identical to the site record book, with the same headings and columns, so that information can be copied from one to the other with minimum error. A hardback notebook, protected from the weather with stitched pages (rather than a file with loose, detachable pages) is best. Columns to be set out in a form similar to those in Figure 4.4, in addition to the name and address (and telephone number if appropriate) of the project, contact name etc. being clearly printed, in more than one language if necessary, on the book.

Field teams should routinely carry certain items such as a tool kit for impromptu repairs to gauges and stands (depending on the types used), spare record book, spare measuring glass, pay sheets, etc.

It is important to remember that meteorological offices "throw back" daily rainfall records, i.e. to attribute rainfall recorded on a morning to the previous day. This may not be appropriate for some studies.

4.1.2 Recording (Intensity) Rain Gauges

There are three main types of recording rain gauge systems: Tipping bucket, Syphon and Weighing. Recent advances have led to the use of electronic loggers which now frequently replace the usual chart and pen clockwork systems.

Alternatively, very compact intensity gauges measuring to a precision of 1 mm of rain and reputedly accurate to 2% over 2 years are available, though the cost of this equipment is not low. A liquid crystal display readout is given via a connecting cable and daily, weekly and monthly accumulated totals may be collected.

a. Tipping Bucket Gauge

These are commonly seen as in Figure 4.5, below. There are many different types and the manufacturers manual must be followed carefully, as is the case for all recording rain gauges. A dual tipping bucket pivots on a horizontal axis which lies beneath the funnel of the orifice, such that only one bucket receives rainfall at a time. When filled to a preset, calibrated amount (for example 0.2 mm) the bucket tips and is emptied, leaving the second bucket to receive rain. Tips are recorded electronically and individually.

Data are downloaded and analysed by computer software, though hard copies used for manual analysis are sometimes available via a portable printer carried to the field. Alternatively, the data may be recorded on a mechanically driven chart. In general the tipping mechanism works well, but sometimes does not register very light rain in hot climates. It may also under-register during very intense storms, because of the finite time taken for the buckets to exchange positions.

Electronic logger type

Advantages:

- Simple download directly to a computer.

- With computer programs, the analysis of intensity data is quick and easy.

- No problems with mechanical clock, ink, pen etc.

- Very large amounts of data can be stored (32 - 120 kb)

- Options for storing data, time, etc. only when rain occurs, thereby saving memory.

- Can operate for very long periods.

Disadvantages:

- Some designs are new and may not be well field tested.

- If the logger or battery fails, then all the data can be lost.

- Needs computer facilities at base (good electricity supply, etc.).

- Needs spare loggers or portable computer, both of which are expensive.

Maintenance

Buckets must move freely and be oiled on a regular basis in some cases. Mechanism must be checked frequently to ensure that it is horizontal.
Battery and connections must be tested at each visit and replacements made when necessary. Gauges come precalibrated, but they must be re-calibrated at the end of each season in the laboratory. This is a simple process whereby known volumes are emptied into the gauge from a pipette and checked against the record. Check with the manufacturer's instructions.

The tipping motion closes an electrical contact (usually current is provided by a 6 volt dry battery) which registers a pulse on an electrical counter or logger, each pulse representing the bucket contents (the example 0.2 mm of rain). The data are usually in the form of each tip represented by a time (Month: Day: Minute: Second). The opportunity to record the gauge name/number may or may not be offered, so it is wise to keep a careful note of from where the logger came. A 'record only with rain' facility is usual.

Various methods of down-loading can be used. In some instances the logger must be removed and taken to base to be down-loaded via an interface and computer. A replacement must be provided. In some cases the data can be downloaded in to a portable computer and the logger can remain at site. In all cases some sort of set-up procedure is necessary to re-activate the logger once the data are extracted.


Figure 4.5: Tipping Bucket Raingauge, Electronic Logger Type

Mechanical (Clockwork drum and chart)

Some gauges provide a chart record of the tipping instead of an electronic recorder. A permanent pen record is kept of each tip on a clock driven drum chart. Clocks usually work for about 30 days without attention, but this can be altered in most cases, by replacement of parts of the gearing mechanism.

Advantages:

- Permanent record is kept on chart, therefore cannot be lost.

- Less likely to be affected by adverse conditions

Disadvantages:

- Analysis of data is lengthy and must be entered manually into computer storage.

- Needs more frequent visits

- Not as flexible in terms of alteration of the instrument settings

Some recorders offer both electronic and mechanical records. This gives a good back-up facility and some models even record river levels simultaneously. In some cases the tipped water runs to waste, but some gauges have the provision for the rainfall to be collected in a vessel below the gauge via a funnel and so the total rainfall for the period between readings is known. This can be very useful if the gauge or logger develop problems.

Maintenance

This is a little more complicated than for the electronic gauges, though the bucket check is the same. Clock accuracy must be tested on a regular basis, even though charts last 30 days without attention. Chart replacement should be done with an accurate pen reset and any malfunction noted. Chart drum motor may be clockwork or electrical.

b. Tilting Syphon Type

Versions of this type suitable for use in tropical countries are available. Figure 4.6 illustrates a typical syphon instrument.

Rain is collected and falls into chamber A, and raises float B. In response, the pen moves upward and its trace is recorded on the chart fixed to the drum, H. The chamber is on a pivot (C), over-balances when full and empties through the syphon tube (D). The pen is then reset to the zero position while lifted clear of the chart by the rod G. The over- balancing is controlled by the trip, E and the chamber is restored to its original position by F, the counterweight. The siphoning takes approximately 15 seconds.


Figure 4.6: Tilting Syphon Rain Gauge

It is important to test the mechanism regularly by pouring in water through the inlet and to keep the syphon tube and its gauze filter clear of blockage at all times.

c. Weighing Type

This type is less common than the former kinds, but is especially useful where snow is frequently experienced. Precipitation is collected from a funnel into a bucket which, as the frame upon which it stands falls with increasing collection, stretches an isoelastic spring. The movement of the frame is proportional to precipitation and linked to a pen by a series of levers. This records on a clock-driven drum chart.

The maintenance of recording gauges will be the responsibility of visiting field teams who must be trained to a higher level than the field gauge reader responsible for daily gauges. Tools, loggers, charts, pens and ink will be carried routinely. Although gauges are of a type (syphon, tipping bucket, mechanical, electronic etc.), each will vary according to the manufactures' particular specifications and it is impossible here to list specific instructions for all gauges. In addition to the points of particular care noted in the sections describing the types of recording raingauge, the manufacturers instruction manuals should be carried to the field and studied carefully.

Factors Affecting Accuracy of Rain gauges

Many factors affect the accuracy of rain gauges. These include evaporation, adhesion, inclination of the gauge, condensation and splash. However, these are unlikely to cause differences of more than about +/- 1%, whereas wind turbulence at a poorly sited and maintained station can account for much larger errors. The precise effects of wind speed is still contentious despite many years of research. Some authors predict large deficits (for example 17% with winds of 16 km hr-1 and 60% at 48 km hr-1) while others (see Figure 4.7 below) expect the effect of wind speed on rain gauge catch to be much less. Agreement that wind effects cause an underestimate of rainfall (and more especially snowfall) is universal, however.

Damage to the instrument during carriage should be avoided, especially denting about the orifice, which can cause discrepancies in readings. Splits and cracks in the receiving vessel can cause serious losses. Care should be taken to ensure that all the water is emptied into the measuring vessel and that the measurement is made accurately at the bottom of the meniscus.


Figure 4.7: Catch versus Wind Speed

A list of errors and causes inherent in measuring rainfall from standard, non-recording gauges are given below. Other factors, notably poor maintenance, faulty resetting of loggers, pens and charts can affect the accuracy of rainfall records obtained from intensity gauges.


Figure 4.8: Errors in the Observation of Point Rainfall

Rain Gauge Networks

Many projects will be adequately served by the installation of rain gauges at each experimental site, when the interpolation of data between sites is unnecessary or can be easily achieved, perhaps using data from the national network. In some circumstances it may even be best to install two daily gauges, to allow for occasions when one gauge may be accidentally inoperative, though of course this must be balanced against cost. If the site area is large, these should be separated, with one gauge in the centre of the site and one at its boundary. When placed on a line at right angles to the prevailing wind, it is sometimes possible to collect information on the distribution of rainfall. If resources allow, the installation of a recording rain gauge at each site, in addition to at least one non-recording gauge, is to be preferred for agrohydrological purposes. Rainfall intensity is an important influence on runoff and its study will undoubtedly play an important part in the research agenda. However, recording rain gauges are very expensive (even "low cost" data-logger versions are currently several hundred pounds sterling, each) and project resources must be considered carefully.

If resources are inadequate for total coverage by recording gauges, then partial cover must be budgeted for. The success of partial cover will depend on a sound instrumentation strategy, which can only decided upon by the staff of each individual project. This will be dependent on the number of sites, their proximity to the base station and each other, the spatial variation of rainfall characteristics, the time between visits and field staff reliability. The two extreme options are:

1. to place all recording instruments at the most distant stations. This will provide event, daily and intensity data from widespread, infrequently visited sites which would otherwise give only rainfall totals from several days using non-recording gauges. However, it is wise to recognise that infrequently visited sites are always the most troublesome. Faults in and damage to the equipment will not be seen for some time, gauges can be interfered with or even stolen. Data and equipment could be totally lost. Access may be impossible at times during the wet season.

2. The second extreme option is to place all such valuable equipment at or near the base station. There is little danger of problems with the equipment not being rectified quickly, but the opportunity to collect data from diverse areas is lost. The best solution, clearly, lies between these two examples and to some extent trial and error (especially becoming familiar with the reliability of the gauges under field conditions) will be needed to determine which outlying stations are most suitable. It is essential to monitor one gauge carefully at the base station and check its readings and reliability of operation.

In many areas of convective rainfall, a statistical randomness means that over time, the average number of storms of a given intensity will be experienced at all locations within the study area. Thus it is reasonable to presume that rainfall intensity can be extrapolated from one site to another, if certain characteristics of a rainstorm (for example the amount and duration of rain) are known. It is convenient if a statistically significant relation exists between rainfall amount and rainfall intensity and allows the substitution of one type of data for another. Figure 4.9 below shows 30 minute duration rainfall intensity against daily rainfall. The significance exceeds the 99.9% level. In some instances, a clustering into groups of data points may be seen with a strong correlation among them, probably indicating that different types of rainfall (for example low intensity frontal and high intensity local convectional) have been experienced. The 2 minute duration intensity against daily rainfall showed no significant relation for the data.

If a comprehensive raingauge network is proposed for an agrohydrological project, the number and density of instruments will depend on several factors. Those relating to the physical environment are:

- Size of area covered

- Prevailing storm type

- Topography and Aspect

- Variation in seasonal rainfall

In general, more gauges will be needed for large areas and denser networks will be needed where storms are convective and localised with high intensities (as opposed to cyclonic areas where rainfall tends to be widespread and of more uniform, low intensity). Convective rainfall is characterised by the predominance of thunderstorms. Mountainous areas, which create orographic rainfall, are expected to have localised rainfall regimes and to need a more dense network than plateau areas (Table 4.1). However in practice, the rainfall in mountainous areas may be of a more regular distribution than extremes of elevation may suggest whereas flat plains dominated by very local convective rain storms often exhibit very large coefficients of variation of rainfall distribution.

It is important to plan with the hydrological characteristics of the area in mind. For instance it is more important to place a denser network of gauges in areas which contribute most to runoff, than in homogeneous areas which contribute little. It is possible to use correlation analysis in determining network densities. For instance if the correlation of daily rainfall between adjacent gauges is high (say, r = 0.90 or greater), a firm basis is provided to reduce the number of gauges in the network. Rainfall is spatially variable to a high degree and even the densest network of gauges can provide no more than an estimate of areal precipitation.

Networks are best planned in the preliminary stage of a project by the use of a map desk study to provide a picture of the overall pattern of gauge distribution. The distribution should not be random; random events are studied by a systematic arrangement of sampling points. Minor revisions of the network pattern can be made during installation, if unexpected problems of siting are found. It is useful to place some gauges outside the study area to ensure that extrapolation is possible to the boundaries of the study area. More gauges are needed if results are to be taken to other areas, rather than limited to the original study area.


Figure 4.9: Daily Rainfall versus Rainfall Intensity

As stated above, project objectives and resources will determine, to a great extent, the level of instrumentation. A watershed study that relates total precipitation and total annual runoff yield would need fewer gauges than a study of rainfall on a storm-by-storm basis. The USDA regards the following gauge densities as suitable (see Table 4.1), but states clearly that the size of the study area is the only criterion used to determine them. Factors such as climate and topography are not considered. These recommendations can be compared to those of WMO which take into account, in a limited way, climate and topography. However, it is likely that the needs of the project will be of paramount importance when compared to such general recommendations.


Table 4.1 Recommended Density of Daily Rain Gauges by Size of Study Area

4.2 Other meteorological data

Rainfall data collected for hydrological purposes will also be useful to other project members, such as agronomists and soil physicists. The same is true of other meteorological data which may be important, for example in assessing crop performance under varying climatic conditions. These other data will help categorise climate in general and are essential to estimate evapotranspiration and soil moisture conditions which can have an important effect on runoff production, soil moisture availability and crop growth. Indeed, the calculation of evapotranspiration is one the most important uses to which comprehensive climatic data are put during agrohydrological investigations. This chapter concentrates on instrumentation; the uses and analysis of data are covered in chapter 8.

Site of Meteorological Stations

The installation of climatic instruments requires a suitable site which should be representative of the macroclimate of the study area. Where climate varies greatly, perhaps due to topography, several stations may be necessary, though the spatial variability of some meteorological variables is greater than others. The site should not be in an exposed position on a steep slope, nor should it be within the distance of four times the height of any nearby trees or buildings. In semi-arid areas, sparsely vegetated open areas make good, representative sites. The site will probably represent the greatest concentration of instruments for a project and it is essential that a suitable, secure location be selected. Advice on instrumentation should be sought from the local Meteorological service and where possible instruments of the same manufacture should be acquired.

The site can be instrumented according to particular need, but special attention should be paid to such details as shading from elevated posts and other instruments. Doors to equipment should be away from direct sunlight, areas of artificial surfaces should be kept to a minimum. The area should be well fenced and gated, not higher than 1m, with wire mesh which is fine at the foot of the fence, to deter animals. Birds can be prevented from roosting on instruments by the provision of alternative high perches. Fence posts should be of metal to avoid termite damage. It is useful to retain extra space within the compound, for instruments that may be added at a later date. Where possible recording instruments should be used. Experience will show that records obtained manually, during holidays and weekends often appear suspiciously inconsistent when compared to weekday readings.

Records of equipment should be kept secure at the station. Loose-leafed books are more prone to damage and loss than those with permanent bindings. A site map of water pipes, cables etc. is useful if the station acquires permanent buildings. Longitude and latitude should be noted on any such map. Records of instruments added or removed are very useful as are schedules for routine repair, painting and grass mowing. Files should be kept which contain notes on the instruments; when bought, invoices, serial numbers, calibration tests, instruction manuals, repairs etc. Having this information easily at hand can save a great deal of time and frustration. Such details should be kept separate from the day-to-day records of measurements, and in a secure place.

4.2.1 Air Temperature

Air temperature is one of the most commonly measured meteorological variables. Maximum and minimum temperatures are used to calculate mean daily values for use in evapotranspiration estimations. Figure 4.10 illustrates the installation of maximum/minimum thermometers.


Figure 4.10: Maximum and Minimum Thermometers

They should be housed within a screened building or box.

The temperature readings are obtained from two separate thermometers. The maximum thermometer is of mercury in glass, secured so that the bulb end is 5° above horizontal. The minimum thermometer is alcohol-filled, with the bulb end about 5° below the horizontal.

As temperature drops, the alcohol retreats into the bulb, inducing an index (a small, dark dumbbell) to move within the bore of the thermometer, until the minimum temperature is reached. When the temperature rises, this index is left behind to give the minimum reading. This reading is taken at the end of the index furthest away from the alcohol-filled bulb. Both temperature readings are usually taken at about 08:00 each day and the instruments are then reset. In the case of the mercury-filled maximum thermometer, a rise in temperature forces the expanding mercury through a constriction above the reservior. The mercury cannot return when the temperature falls and the maximum temperature is shown. The mercury must be shaken gently back into place after the reading is taken.

4.2.2 Humidity

The moisture status of the air has a strong influence on rates of soil evapotranspiration (Et) and open water evaporation, both of which are greater when the humidity of the air is low. Relative humidity values are widely used in evapotranspiration equations and two main methods are used to measure humidity:

The first uses thermodynamic principles and measures temperature differences between wet and dry thermometers ("psychrometers"). They are set together, with the wet thermometer slightly lower than the dry and are usually housed as one unit in a secure metal frame. Around the bulb of the wet thermometer is placed a wick sheath, which trails into a container of clean, distilled water. The wick should fit tightly; dust, dirt and insects are sometimes a problem and the wick may need replacement or cleaning each week. The simplest and most common type of psychrometer (sometimes also called a "hygrometer") is housed in a screened box. The natural flow of air around the wet bulb thermometer results in it registering a lower temperature than the dry bulb thermometer. The use of psychometric tables converts the readings into dew point temperatures and relative humidity (RH). Hand-held versions are available for spot readings, these being whirled around on a handle to encourage ventilation; others use the assistance of electric fans to achieve this effect.

In the second case the instrument uses the hygroscopic properties of a material (usually human hair) to determine humidity, and is called a "hair hygrometer". A series of hairs expand and contract according to atmospheric moisture and oscillate a pen which marks a trace on a chart, moved by a clockwork drum. Adjustments to the instrument can be made by altering the arrangement of linking levers.

Hygrographs are usually placed on the floor to ensure stability. Shelves used for such instruments increase the possibility of readings being affected by vibrations and for this reason the housing fabric should be strong and rigid. The chart can be annotated with date of chart replacement, station, reader etc. Checks for correct readings should be made against psychrometer values when the humidity is high (early morning or during a rainy period) and low (mid afternoon). Hairs that become dirty should be cleaned with a soft brush, but eventual replacement will be necessary. Very often this instrument is linked to a temperature sensor that gives a continuous record on the chart and can be used as a check or back-up to the maximum and minimum thermometers. In this case the instrument is called a hygrothermograph. Less costly, non-recording hair hygrometers are also available.

4.2.3 Wind Speed and Direction

Wind also has an important effect on levels of evaporation and evapotranspiration. It removes humid surface air layers from above land and water and can physically damage crops. Anemometers are used to measure wind speed and duration and thereby windrun, in km day -1. A standard anemometer has three cups mounted at 120° to each other on a vertical axis. The movement of this rotor closes an electrical contact which measures and records a standard distance of wind movement. A continuous record of wind speed and direction is also provided. Wind direction is obtained by the operation of a single-panel vane. Pen and chart recorders are usual, but electronic recording of these data on data loggers is now common.

A site that is relatively level is to be preferred, with no obstructions within 100 x the height of the nearest obstacle. The World Meteorological Organisation's recommended height is 10m for general speeds, but this height imposes the need for larger and more expensive mast structures. For use in the calculation of Et values by the widely-used Penman method, 2m is recommended. Fortunately an estimate of wind speeds for levels other than of the instrument can be obtained by the formula:

u2 = (ln z2/ln z1) au1 where (4.1)

u2 is the estimate of wind speed

u1 is the known speed at instrument height

z1 and z2 are the heights in cm of the known and estimated wind speeds

a is an exponent between 1 and 0.6 according to ground surface roughness

Empirically this can be stated as Hellmann's formula:

Velocity at height 'h' / Velocity at 10 m = 0.233 + 0.656 log10 (h + 4.75) (4.2)

Hand-held anemometers with digital displays of wind speed are available, but these do not incorporate a wind direction sensor. Relatively compact, portable systems that can be quickly assembled at site, can be purchased. It is useful to note that accurate readings of wind speeds less than 5 km per hour are difficult to achieve.

4.2.4 Solar Radiation

Solar radiation provides energy for evaporation and plant development. Several methods of calculating Et use solar radiation as a key parameter, often converted to net radiation, which takes into account the portion of solar radiation that is reflected back into the atmosphere. Total incoming shortwave radiation is measured by solarimeters, sometimes called pyranometers, which sense the intensity of radiation from the sun and sky, that falls in a horizontal plane.

The portion of all radiation that is transformed into other forms of energy is called "net radiation". Net radiation is measured as the difference between incoming (downward) and outgoing (upward) radiation of all wavelengths by the net radiometer. As radiation varies between night and day, counters can be linked to the radiation measuring device to record these values separately for easy reading. Electronic pulses may be recorded on a strip chart, but increasingly (especially with small, automatic meteorological stations) the data are recorded on an electronic logger and can be downloaded directly in digital form on to a computer, for viewing and analysis.

Sites should be clear of obstructions, with a view to the horizon that is not affected by nearby trees or buildings. Under no circumstances should shading occur and artificial surfaces that can direct radiation to the instrument should be avoided. The site should be typical of conditions under study, but compromise is inevitable where conditions of vegetation type and cover, soil reflection, etc. vary from place to place within the local area. Placing the instrument high, perhaps at 3m, increases the field of reception which is useful in areas that are heterogeneous. Cultivated field situations are more likely to give representative values that rangeland areas which tend toward heterogeneity. As a guide, an instrument set at X metres above the ground will receive 90 and 99% of its upward flux from ground areas with radii 3X and 10X respectively. Figure 4.12(b) shows a typical net radiometer that would be one component of a small, automatic weather station.

It is very important to keep the glass or plastic domes clean and undamaged. The presence of dust is a common problem. A photographer's air brush is very useful for cleaning, but if not available, soft tissues can be used. Care should be exercised as the domes are prone to scratching. The instrument should be kept horizontal at all times. Calibration is important because of deterioration of the domes and black reception faces and should be carried out every few months. However, this involves the use of a replacement radiometer to continue the record and it may only be possible to check the upper and lower sensors during an off-season period, when a break in the record may not be important. This is best done at a time of steady radiation, when the faces of the radiometer are inverted for 10-minute periods. Averages are taken and both faces should give readings within 5% of each other. Alternatively, a second instrument can be kept in good storage conditions and used only as a standard for the field instrument. As different ground conditions may be measured by the two adjacent instruments, it is as well to exchange their positions around to check the first results.

Solarimeters (pyranometers) measure only incoming short-wave radiation and are sometimes used at meteorological stations. These can be used to calibrate radiometers. They should be shaded from direct solar radiation by placing the shadow of a black matte disc (about 1m away and held by a thin support) over the instruments. The response by both sensors should be the same such that Dr/ Cr= Ds/Cs. As Dr and Ds, the change in response of the radiometer and solarimeter are measured (in mv) and Cs, the calibration constant of the solarimeter is known, Cr can be found. This should be done several times.

Sunshine:

Sunshine hours are commonly recorded where the cost and practicability of maintaining radiation meters are limiting. The widely-used Campbell-Stokes recorder consists of a glass sphere mounted on a pedestal which concentrates bright sunlight onto a chart and so burns a trace along it, thus sunlit periods are recorded. Instruments which are specified by ranges of operation latitude (for example 0° to 60° N or S) and the correct charts for the N or S hemisphere should be used.

Electronic meters that measure photosynthetically active radiation (400 - 700 nm) are also available.

4.2.5 Evaporation

Evaporation Pans:

The measurement of free surface water evaporation depends on air temperature, wind, humidity and solar radiation. It is a commonly measured index that integrates these meteorological factors and to some extent, illustrates the behaviour of evaporation from water bodies and evapotranspiration from wet soil, where the availability of water is not limiting.

The most commonly used instrument, which is an international reference instrument, is the US Weather Bureau 1.22 m (4 foot diameter) A-pan. This can be purchased complete, or made from local material such as a suitable gauge, galvanised steel sheet. Pressed seams should not be allowed to cause buckling. Welded seams should be treated to prevent rusting The pan should be mounted horizontally on a wooden platform with tamped soil below, to allow a 13 mm air space (half an inch). In humid climates the areas around the pan will be grass, whereas in arid areas, vegetation will come and go according to season. However, vegetation should not be allowed to grow above the level of the pan. Locations near areas with artificial surfaces, boggy areas or water surfaces should be avoided.

The water level in the pan can be measured by an inclined, graduated gauge staff, but accurate reading in this manner is more difficult than by using a stilling well and micrometer hook gauge. The level is measured with the hook tip, lowered below the water and then raised until the tip just pierces the surface. The mechanism is removed from the stilling well and the reading taken from the graduated vernier scale. Water levels should be kept at 5 cm below the rim of the pan (+/- 2.5 cm) and water should be added or removed to maintain this level. A reading should be taken before and after this has been done. The differences in daily levels give evaporation, with additions and removals of water and daily rainfall (measured nearby) being taken into account. Readings are taken at the same time each day, normally 08:00 hours, though alternatively a WLR could be used if the cost is not prohibitive.

Daily maximum and minimum temperatures of the pan water are often taken by floating thermometers, kept at least 30 cm from the side. Problems can occur with birds and animals drinking from the pan. Fences will keep out larger animals, but a wire screen fixed over the pan itself, may be necessary. This can effect readings by the suppression of evaporation and in semi-arid climates a correction of 16% is made to measurements. Algal growth can be prevented by a small addition of copper sulphate to the water and the pan should be kept clean of debris and insects. Figure 4.13 shows a hook gauge version of the Type A evaporation pan. Data may be lost during periods of heavy rainfall and over-topping.

Lysimeters:

Lysimeters most commonly measure evapotranspiration by changes in the weight of containers filled with soil, to which water is added. Losses by evapotranspiration are then calculated. Lysimeters can be very large, weighing several tons, while others used in field locations may only measure water losses from a few kilogrammes of soil. Crops and vegetation may or may not be grown in them. Large containers are weighed by permanent pressure transducers, though the construction of large lysimeters is normally beyond the scope of many agrohydrological projects and is not practicable under field conditions. Meteorological services may find it useful to install them.

Evapotranspiration from small lysimeters is measured by their removal and these instruments are more commonly used for field research. Suitable ones can be made from PVC water pipe with a diameter of 15 cm and a length of 20 30 cm. They are filled by attaching a steel cutting edge to the lower edge of the plastic, and a metal ring to the top. The latter prevents damage when they are hammered into the ground at the required location, when soils are at field capacity and drainage has ceased. Jacking lysimeters into the ground may be necessary, but the appropriate equipment must be available.

The lysimeter is removed and the soil is retained by a wire mesh (5 mm is suitable), which is screwed into the base. They are then replaced into their holes which have been fitted with tubes, a few millimetres larger in diameter and deeper than the lysimeters. Lysimeters are removed and weighed every hour or so throughout the day. Because the cores are isolated from root activity, new cores should be taken every 2 to 3 days, to account for plant extraction.

There are other problems of representing true conditions with lysimeters: filling with soil can disturb the profile, edge-effects are great especially with small models, isolation leads to hydraulic continuity being lost at the sides and affected at the base. Intermittent, unexpected rainstorms can affect readings.

4.2.6 Soil Temperatures

Soil temperatures have direct effects on the germination and root growth of crops and natural vegetation, the state of which can greatly affect runoff. Soil temperatures determine the micro climate of the overlying air and are important for assessing the growing environment of crops. Soil temperatures are not only dependent on incoming and outgoing radiation, but also on the thermal properties of the soil which can change greatly with the addition of water by rain and its removal by evapotranspiration. Temperatures reach maximum some time after local noon and minimum after midnight.

Continuous records of temperature can be taken with thermocouples (thermographs) linked to pen and chart or electronic loggers. Thermocouples should be calibrated with mercury thermometers at least twice a season, at the beginning and end, and any corrections must be noted. Bent-stem mercury thermometers for 10 and 20 cm depths and encased mercury types with bulbs in crystalline wax and suspended in steel tubes for 50 and 100 cm depths, are used where recording instruments are not available. These depths are recommended by WMO. Readings are normally taken at 08:00, 14:00 and 20:00 Local Time. Good contact with the soil is necessary and accidental trampling should be prevented.

4.2.7 Automatic Weather Stations

Research applications may demand the collection of climatic information at sites in addition to base stations. In such cases the use of automatic weather stations may be more suitable than an array of individual instruments. Electronic logging is used to keep records (usually on a multi-channel data logger) and to avoid the need for frequent visits. The period between visits is determined by the number of instruments used, the frequency of record and the memory size of the logger.

Different formats for the presentation of data will be used according to manufacturer.


Figure 4.14 Example Meteorological Data sheet

Stations should be enclosed by fences in a suitable position, as discussed in the sections on individual instruments. Considerable thought should be given to the possible problems of vandalism and theft because of the cost of automatic weather stations and the ease with which the array of instruments can be damaged.


Figure 4.15: Typical Weather Station Layout

Figure 4.14 shows an example data sheet for the meteorological variables discussed above.

Figure 4.15 shows a plan of suitable weather station site, equipped with a basic list of individually installed instruments. Note that sufficient space is left within the compound to accommodate instruments that may be installed at a later date.

It is important to select automatic weather stations according to particular project needs, for example some stations place anemometers below the recommended 10 m elevation, and if wind speed and direction are important factors in the research agenda, this may not be suitable. Automatic weather stations can be very cost effective when their prices are compared to those of collections of individual instruments. Generally, the seven following meteorological parameters are measured:

- rainfall and relative humidity
- air and soil temperature
- wind speed and direction
- solar radiation.

Equipment costs

All costs of locally made equipment are approximate. The costs of raw materials and especially labour are highly variable from country to country, but a good idea of cost magnitude can be gained from the figures quoted below. The costs of manufactured equipment are based on 1993 prices. Shipping, agents' fees and fluctuations in exchange rate cannot be taken into account.


Table

(introduction...)

Soil physical, chemical and moisture properties constitute a study in their own right and it is possible that any agrohydrological or water harvesting project may have available the services of a soil specialist, but this is not always the case. The effects of soil physical properties on hydrological behaviour are very important.

Four main aspects of soils and their influence on runoff and agriculture are considered. These are:

1. The physical and textural nature of soils which are influential in determining runoff.

2. The soil moisture status which can also influence runoff and control water availability for crops.

3. How to measure soil moisture.

4. The influence of these soil factors on the process of infiltration, the ability of soils to absorb water.

In many respects soil textures and soil moisture status are closely linked; the physical characteristics of soils may change with the addition or removal of water, while the physical characteristics of soils will determine their ability to absorb and retain rainfall. In terms of the study of soils for agrohydrological purposes and the quantification of soil characteristics, it is most convenient to study these aspects separately. Methods of determining infiltration, which is strongly influenced by texture and moisture status, are also discussed.

5.1. Soil classification and soil textures

5.1.1 Soil Horizons and Their Characteristics

The soil profile, as exposed by the side of a pit is usually divided into 3 horizons which are frequently further divided into sub-horizons:

A horizon constitutes the top soil, where any organic matter is found and within which cultivation is initiated. B horizon is the subsoil, without organic matter. C horizon which is composed of weathered rock, usually the parent material.

Soil pits, dug to give an exposure of the soil to the C horizon where possible, provide a great deal of information which is used in the classification of the soil types. From the agrohydrological viewpoint however, it is the practical effects on farming and hydrology of such factors as the effective depth of soil (that is the depth that can provide a medium for roots) that are important. In most cases the effective depth is limited by the nature of parent material and the manner in which it has weathered; climatic influences are often strong. In other cases, gravel bands may be present and if tightly bound, will restrict the development of crop roots. Such bands should be noted as the limit of the effective depth. Roots may be evident in partially weathered parent material but it is unlikely that they contribute much to the intake of crop water and nutrients. Information on parent material, erosion, formation history and climate, indicate past periods of waterlogging and other aspects of the nature of the soil moisture reserve.

Topsoils

Of particular importance is the character of the top 20 cm or so of soil. This is the soil layer that influences soil surface/rainfall relations by its texture and aeration, and represents the approximate depth of cultivation. The top soil layer also determines structural stability, fertility, and the tendency for a soil to cap or erode. There are obvious limitations to digging large numbers of pits in order to determine soil characteristics; the job is a long and arduous one and pits must usually be filled in after examination. The textural definition of surface soils is therefore more commonly assessed by working the soil by hand, when wet. Where an accurate textural analysis of soils is needed, samples are taken and analysed in the laboratory (see chapter 3). Table 5.1 below lists the characteristics of soil textural types when manipulated.

Sandy soils have high rates of infiltration and percent runoff is usually low. They tend to be infertile, relatively acid and prone to leaching. At the other extreme, clay textured soils give high percent runoff in general, though cracking vertisol soils may absorb water until the clay particles swell, the cracks close and runoff results from later rain. Fine textured soils normally have a higher water holding capacity than coarse, sandy soils and their chemical mix is more varied and nutritious for plants.


Table 5.1: Soil Textures According to Manipulation When Wet

Soil depths are also important; whatever the inherent water holding capacity of soils on a unit volume basis, the absolute volume of water available to crops will be small if soils are shallow. This is an important consideration when the viability of water harvesting opportunities is being assessed, as it will be a critical factor in determining how frequently water must be added to the soil moisture reserve.

Soil textures are determined precisely and classified most rigorously in the laboratory as described in chapter 3. The FAO has now adopted the USDA soil classification triangle which categorises soils into textural types according to the percentage of silt, sand and clay components, and is shown in Figure 5.1. The "International" classification (Figure 5.2) is now used in few countries. Relatively small differences exist between them. Several approaches can be taken to the selection and collection of soil samples. Spatially, soil textures can be highly variable, so that when top soil samples are collected to assess the general textural type of a large area (for instance a whole field), samples are taken at individual points and combined well before submission for analysis. If the spatial variation of soil textures is in itself a characteristic under investigation, samples should be taken systematically on a marked grid basis, with each sample given a point reference number accordingly. This avoids subjective sampling.

Where microtopographical features are under study, sampling should take place along defined transects at every one or two metres, according to transect length. Again, the samples are referenced to the sample points and also to the elevations above a base level (a levelling survey will be necessary). Wash-ins, ploughing, crops, vegetation and faunal activity may be recorded. Loose samples (not cores) are collected and sealed in polythene bags and the depths to which they are taken are noted. See chapter 3 for details regarding the dispatch of soil samples.


Figure 5.1: FAO/USDA Soil Classification Triangle


Figure 5.2: International Soil Classification Triangle

Subsoils

Subsoils affect soil water permeability and thereby runoff. In the field, permeability is usually assessed by the observation of soil physical characteristics rather than direct or laboratory measurement of hydraulic conductivity. Common terms used in descriptions are:

Compacted: Firm or hard consistency, close packing of particles resulting in a dense material with reduced pore space.

Cemented: Hard and brittle, soils which do not soften with prolonged moistening.

Deflocculated: Soils in which sodium has entered the exchange complex and dispersed the colloids. This leads to reductions in pore space, aeration and permeability. High levels of pH and electrical conductivity are found. Columnar horizons which are hard and dense may be found.

The colour of soils may give information on aeration and drainage and are described according to the standard Munsell notation. Colours may vary between and within horizons, for example:

Drainage:

Reds

Well drained


Yellows, Greys

Poorly drained

Organic Matter:

Browns & Blacks

High in organic matter

Leaching:

Paler

Moreleached


Darker

Less leached and higher mineral fertility

Wet soil colours are usually darker and there may be the presence or absence of mottles. The colours on Munsell charts that provide the standard reference (described in detail below), are arranged to give the three variables used to define all colours and are recorded in a standard order:

Hue: The dominant spectral colour (increase in redness or yellowness).

Value: The lightness of colour and total amount of light reflected.

Chroma: The purity or strength of colour (increases with a reduction of greyness)

On each card the colours are of a constant hue. The colours increase in lightness vertically and in equally visible steps. The colours increase in chrome to the right and become greyer to the left. In the field, a 1 cm fragment is selected from the sub soil, untainted by organic matter. After deciding whether it is predominantly yellow or red, a colour chart is selected and the sample compared through the most appropriate hole in the chart. Intermediate matches are not uncommon. Check that the hue is correct. Avoid sweat on the colour charts (not always easy).

Mottles, very pale and very dark colours indicate reduced permeability or groundwater near the surface. Rust-coloured mottles along root channels suggest periodic waterlogging, as does an abrupt change from reddish to greyish colouration. Grey mottles in an otherwise reddish weathered rock zone indicate a seasonal water table.

Bulk Density

Soil texture is largely responsible for the bulk density of soils, that is the weight per unit volume, most commonly expressed as g cm-3. Imperial units of lb ft-3 may be seen. Bulk densities are found by comparing the oven dry weight of samples and their volume. Samples are taken from soil pits using standard soil sampling cores, driven into the exposed face below the top soil when the soil is neither very wet nor completely dry. The sample must not be disturbed, so as to maintain its original volume. The sample should be oven dried at 105 °C to 110 °C and weighed to the nearest 0.1 gram.

The bulk density (sometimes called the "specific weight") dry weight of sample/ volume of sample The bulk density in g cm-3 can be converted to lb ft-3 by multiplying by the factor 62.4.

Soils with high bulk densities have a paucity of pore space, impede root penetration, make cultivation difficult and promote runoff.

5.1.2 Pedological Classification

The pedological classification of soils, although basically created with agriculture in mind, is described only briefly here. It is relatively complex and includes an extremely wide range of soil types. Many of the terms and names derive from the Russian language. Soil surveys and maps use the orthodoxy of pedological classification, but in developing countries soil mapping is usually at an early stage or restricted to localities of special interest. Map scales are commonly 1: 250,000 to 1:1,000,000 and cannot be expected to depict the variability of soil types with accuracy.

The pedological classification of soils is broken into two main groups: Higher and Lower categories. Of the higher categories, the nature of Zonal soils depends greatly on the prevailing climate at the time of formation. Intrazonal soils not only are influenced by climate, but also localised conditions, for instance poor drainage, and therefore cross the boundaries of zones. Azonal soils such as lithosols (rocky) and regosols (dry sandy) are not zonal.

In arid and semi-arid regions, regosols, lithosols and lateritic soils (which are red and have a high iron oxide and aluminium hydroxide content) are commonly found. Variation in soil types is wide and intrazonal soils may commonly occur due to changes in local conditions of geology and drainage. Calcareous bands may be common at depth.

Glei are indicative of impeded drainage and a rising and falling of the water table. These mottled colourations may be red, yellow or brown when the water table is low, or grey or blue when it is high, resulting from the oxidisation or non-oxidisation of iron and manganese.

Higher Categories

There are three main orders of soils (Zonal, Intrazonal and Azonal) which are sub-divided into Suborders and Great Soil Groups:

Zonal Soils Suborder Great Soil Groups

1. Cold zone

Tundra,

2. Light coloured arid zone

Desert, Red desert, Sierozem,


Brown/Reddish-brown soils

3. Dark coloured soils of semiarid sub humid and humid grasslands

Chestnut, Reddish-chestnut, Chernozem,


Prairie, Reddish prairie soils

4. Forest grassland transition

Degraded Chernozem and Noncalcic brown soils

5. Light coloured timbered regions

Grey wooded or grey podzolic soils

6. Laterite soils of forested warm temperate and tropical regions

Reddish brown and Yellow brown lateritic and Laterite soils

Intrazonal soils

1. Halomorphic (saline and alkali)

Saline, Soloth, Solonetz soils soils of imperfectly drained arid regions


2. Hydromorphic soils of marshes

Humic-glei, Low humic glei, Bog, and swamps

Groundwater packed soils

3. Calcimorphic soils

Brown forest soils


Azonal soils


Lithosols, Regosols, Alluvial soils

Further classification is beyond the scope of this book. The Unified System, developed in the USA, is concerned with the engineering aspects of soil classification, rather than agriculture.

Lower Orders

The Great soil groups are subdivided into Soil Series and then Types. Series are soils developed from the same parent material and soils within a series have the same profile characteristics except for the texture of the surface layer. Types are determined by the texture of the A horizon. Soil Phases are determined by deviation from the norm, for example a stony phase.

5.2. Soil moisture

The soil moisture content of a soil is of primary importance. Soil moisture is expressed either in percent by weight (Pw) or volume (Pv) . The relations are:

Pw = (Ww - Wd/Wd) × 100 and (5.1)
Pv = ( Vw / Vs) × 100 (5.2)

respectively, where the subscripts w and d are wet and dry samples and Vs is the volume of the sample.

Percent weight of water is the most common (gravimetric) determination and is found by using samples obtained from the field. Samples can be taken with shovels, augers or soil sampling cores. Samples are best transported for immediate drying in electric ovens at 105° C in sealed cans to prevent moisture loss; this temperature removes all moisture, without driving off other volatile matter.

Balances should weigh to the nearest 0.1 g and samples should each weigh at least 100 am. With a typical weight of moisture of the sample being 20 g, the accuracy of measurement will thus be approximately 0.5 %. Several samples will be needed and areas used continually for sampling may suffer. Soil heterogeneity can be a problem, though as discussed previously, the manner in which samples are collected determines the extent of this difficulty.

The usefulness of volumetric determinations of soil moisture content lies in their easy conversion to surface units. This conversion allows comparison with rainfall and irrigation applications, although percent volume determinations are not usually obtained by direct sampling. The volume of water from soils could be estimated by determining the weight of moisture and converting to volume assuming a specific gravity of 1.0 for water, though the volume of soil is difficult to measure. Field determination of soil moisture on a volume basis is normally found using a neutron probe. This is method discussed later.

5.2.1 Soil Moisture Potential

The soil moisture potential (SMP) represents the thermodynamic energy status of a soil and is conventionally expressed in units of bars (1 bar - 106 dynes cm-2). Two particularly important specific points of soil moisture conditions are field capacity and wilting point. Field capacity is the condition whereby moisture is retained after the gravity drainage of a saturated soil is complete and the soil moisture tension is equal to one third of an atmosphere. Wilting point is the condition beyond which plants can no longer extract water and is taken to be a tension of 15 atmospheres. Soil moisture between these two points is regarded as that available to plants.

Total SMP is composed of three components:

Potential energy due to the force of gravity, osmotic potential and capillary potential. The latter is by far the most important and is assumed to be more or less equal to the total soil moisture potential. Equipment is available to measure this variable in the field.

Since many occasions arise when either the soil moisture content (SMC) or capillary potential can be measured, a relation between them is desirable; this could be used to describe the SMP. Unfortunately, a unique relation does not exist for most soils and the moisture content depends not only on capillary potential, but also on previous soil moisture history. The effect of this is that two main soil moisture relations exist (see Figure 5.3), one for the dewatering of the soil (the soil moisture retention curve) and another for rewetting. These soil moisture/suction conditions are called the "hysteresis loop" and it is often represented in graphical form.

In many situations it is suitable to refer only to the dewatering branch of the hysteresis loop since this has the most profound effect on plant growth. It is obtained and applied under drying conditions. In very many cases these relations are unique to the soil, though some generalisation is permitted and related to field (and sometimes project) conditions. In some soils the two branches of the hysteresis loop may be relatively close and it can be assumed that they are the same for all practical purposes, though this is not usually the case. Table 5.2 gives typical soil moisture values for various soil types.


Figure 5.3: Typical Soil Hysteresis Loop


Table 5.2: Typical Moisture Values for Various Soil Types

Table 5.3 below gives a field guide for judging how much of the available water has been removed from different soils.

Equipment for the measurement of SMP, with the exception of the neutron probe, usually consists of a material that is placed in the soil to reach equilibrium with the soil moisture and as such measures capillary potential, not SMC. Care should be exercised in the use of this equipment since the hysteresis behaviour makes the step from capillary potential to SMC problematic.

The soil retention curve is sometimes called the soil characteristic and is found by laboratory analysis. Tensiometers can be used in the field, but the limit of the suction pressure that they exert is very low (about 0.8 bar), whereas the use of a pressure plate apparatus in the laboratory gives a much wider range and water content can be found by weighing the sample at each stage of dewatering. The soil moisture retention curve can then be plotted (Soil moisture in % versus suction in bars), using the data points.

With regard to sampling, spatial variability in soils is the norm and the representative nature of sites that are selected will be limited. Microtopography, runon, runoff, the lateral flow water within horizons and land use are some of the factors that affect soil moisture variability. A number of sites will be needed within a "homogeneous " area and the extent of study will depend on the aims and resources of a particular project. Knowledge of the degree of soil variability is in itself a useful tool in assessing the place of water harvesting and agrohydrological research in the agricultural agenda. A survey at the outset of a project, that is as comprehensive as resources permit, will usually be extremely rewarding. Where possible, work that has been undertaken previously by soil surveys and land use planning organisations should be consulted.


Table 5.3: Field Guide for Judging Available Soil Moisture

5.2.2 In situ Methods of Soil Moisture Measurement

a. Tensiometers
Tensiometers are used to measure capillary potential, the sensing elements are usually porous ceramic membranes or pots. Usually a water and mercury manometer is attached to these membranes to measure potentials, though dial type gauges can be used. The manometers are housed to prevent weather damage (especially sunlight and high temperature) and a set may have perhaps six or ten sensors, each placed at different depths, to cover the possible range of plant rooting.

Various manufactures of tensiometers are available and the instruction manual should be followed carefully. The tensiometer must be saturated with water to work properly and is installed after the membrane has been boiled to remove gases, filled with boiled water and transported to the field wrapped in wet rags or in a container of water. To protect against damage the clear plastic tubing, sealed into the porous pot and placed in a plastic pipe of suitable internal diameter (usually about 2 cm) and length, is lowered into a pre-prepared hole. To ensure good hydraulic contact with the soil, some of the excavated soil from the hole is mixed with water and poured in as a slurry, to act as a seating. Careful back-filling of the hole with the soil is necessary to avoid a depression at the surface. Time must elapse before the slurry dries out, which if below the rooting zone, may take several weeks. Figure 5. 4 shows the installation of mercury manometer tensiometers


Figure 5.4: Mercury Manometer Tensiometers

The tubing is charged using boiled (degassed) water with a large syringe to prevent air bubbles. As the porous pot is in contact with aerated soil however, it is likely that bubbles in the tubing may be a recurrent problem. Modest housing fixed to metal poles with a hinged door for access and painted white to reflect heat is adequate. Installations at every 20 cm depth are suitable.

Points to note:

- Tensiometers are used for monitoring moist soils because of their narrow range of sensitivity.

- Boiled water is always used for filling and flushing.

- Theoretically, tensiometers cannot measure negative pressures greater than one atmosphere, but in practice their limit is less than this, about 0.8 of one atmosphere, greater tensions will encourage air to enter the system.

- In very arid conditions, appreciable volumes of water may be passed through the membrane and may affect soil conditions.

- Tensiometers are subject to thermal variations and it is best if readings can be made during early morning, several time a week, depending on conditions.

- Adequate regard to routine monitoring and rainfall conditions must be given for remote field sites. Flushing with boiled water to remove air bubbles (which make the manometer operate incorrectly) must be part of the routine.

- Vacuum gauges may be used instead of mercury manometers. They are more robust, but may be less sensitive.

b. Electrical Resistance Method

This method of measurement involves the use of blocks of porous material, usually gypsum (calcium sulphate), though sometimes units of fibreglass or nylon construction are preferred. The block material will tend toward a potential equilibrium with the surrounding medium. They are placed in good contact with the soil and the electrical resistance of the block gives an estimate of soil moisture content. Gypsum blocks are best buffered against saline soils, although they are less sensitive to changes at high moisture contents and generally deteriorate more quickly than the fibreglass and nylon alternatives. The blocks come provided with electrical connections in the form of wires or coaxial cables.

Before installation they should be saturated and a resistance reading taken for reference. They are installed in the same manner as tensiometers, but a shallow horizontal trench should be dug with a slope away from the blocks, in which the wires can be laid, to avoid water being directed downwards them and resulting in incorrect readings. Though of simple construction and relatively cheap, the main disadvantage of these devices is their relative insensitivity to changes in the high soil moisture content range.

Points to note:

- They can be used in much drier soils than tensiometers.

- Gypsum blocks operate best at tensions 1 to 15 atmospheres (drier soils).

- They will last for 2 or 3 seasons.

- Nylon and fibreglass operate best at tensions of less than 2 atmospheres, but they are more expensive and are sensitive to salts.

- The materials are relatively cheap, but difficulties include unobserved deterioration while underground, attack by chemicals with consequent errors and a possible drift of the calibration curve.

- More than one unit can be used per installation hole, but great care must be exercised during emplacement.

- The blocks must be calibrated in the laboratory.

Calibration

Two methods of laboratory calibration are in general use. The first is to place the resistance block in a small container surrounded by soil that is initially saturated and to allow the soil to dry out gradually, reaching various levels of moisture content, each of which is determined by weighing the whole system. At each level, a reading of electrical resistance is also taken. The second method is preferable, but depends upon the availability of a pressure plate or pressure membrane apparatus. The resistance block is placed in a pad of soil in the pressure apparatus which is best equipped with the facility of electrical connection through its wall, thus allowing continuous monitoring of the block. The soil is initially saturated and when equilibrium is reached (water no longer flowing out) a resistance measurement is taken. The block is brought progressively, in steps, to various levels of desorption and the measurements of electrical resistance are repeated.

Where accurate and comprehensive field soil moisture measurements are to be made, the blocks should be calibrated for both wetting and drying, to overcome the problems of hysteresis. Soils of different textural characteristics must each be calibrated separately.

c. Neutron Probe Method of Soil Moisture Determination
Neutron probes emit fast neutrons from a small radioactive source (typically 50 - 100 millicurie mCi ). The neutrons are slowed when they encounter hydrogen atoms in the soil and these slow neutrons are registered by a boron trifluoride detector integral to the instrument. Unaffected fast neutrons are not detected. The detection of slow neutrons is amplified and counted by a rate-scaler. As water is the main source of hydrogen atoms in the soil (though organic matter, boron and chlorine will also slow the neutrons), the number of slow neutrons reaching the detector is a function of soil moisture content. The neutron probe has a limited diameter of detection of about 20 - 30 cm.

The neutron probe consists of the main body which houses the cable, clamps, rate-scaler and handles. The source and detector are lowered on the cable into an aluminium access tube in the ground (aluminium being more or less transparent to neutrons). Readings are taken at required depths. The base of the probe body usually has a shield to prevent inadvertent irradiation when the cable is wound in. Figure 5.5 below shows the main components of a neutron probe.

Access tubes

The observation holes that are used for neutron probes are made permanent by casing them with a thin walled material, usually aluminium. Aluminium is almost transparent to neutrons, is durable and can be purchased in varying diameters. The latter is a significant practical point as it is important to select diameters of tubes that give the practical minimum air gap between the probe and the access tube wall, while allowing unimpeded access to the source and detector. This air gap will affect the instrument's count-rate.

Commonly, a single piece of tube is used, with the upper end protruding above the soil surface. This not only aids location, but in many cases is used as a seat for the probe. It is essential to match the outside diameter of the access tube with the inset of the neutron probe or the diameter of the tubes will have to be modified at a later date. Example (but not exclusive) diameters are: tube outside 44.45 mm, thickness 2 mm, internal diameter 40.45 mm probe diameter 38.00 mm. The surface exposed portion of the access tube should be kept to an acceptable minimum but as

it is usually used as seat for the probe body, this portion should be adequate for the purpose. To prevent dirt and water from entering, a rubber bung is fitted. Covering this with a soft drinks can gives extra protection. Usually, condensation is not a problem, butt where it is, small sacks of desiccant should be hung in the tube.


Figure 5.5: Main Components of a Typical Neutron Probe

Installation

The installation of access tubes can be one of the most difficult and time-consuming aspects of neutron probe field work. Hard, dry soils, gravel, concreted layers and high bulk densities make observation hole excavation problematic, despite the development of mechanical installation systems. In fortunate circumstances, hand excavation by auger is possible. The main criterion of effectiveness of installation is the snugness of fit of the access tube within the hole. This depends not only on the method of installation, but also on the type and wetness of the soil. All sites around the access tube should be protected from trampling and soil compaction by the use of wooden frames or palettes, upon which the operator can stand. Vegetation around the tube should not be disturbed if it constitutes part of the actual environment.

Installation by hand:

Continually hand-auger from within the tube in depth stages of 20 cm, remove the material, then push the tube down. This should ensure a good fit. The tube will shear off a little soil material as it is placed, but the undersized hole gives snugness of fit. Some workers excavate an oversized hole and after the tube is installed, back-fill with the soil material. Whichever method is used, tropical and sub-tropical soils are often too hard for hand boring. Ingenuity of the individual worker is often required to solve the problems of installation, which may be quite different from site to site.

Mechanical installation:

One example of a field-tested mechanical installation device is given below, it was developed from a geological sampler, capable of excavating to 25 m. It is illustrated in Figure 5.6.

An initial hole is made to about 0.5 m using a guide tube and sledge hammer. Then solid steel string sections, joined by threaded joints, push down a 0.6 m cutting tube head under the force of a hydraulic jack-hammer. Depending on the soil, the excavation may proceed in stages of 20 -50 cm. It may be necessary for the cutting tube to be locally manufactured, to the same diameter as the access tubes. The cutting tube is retrieved by the use of a ball clamp and hydraulic ram which operate from the hammer compressor, pouered by an 11 hp petrol motor. The soil is removed from the cutting head by screw augers, though this can be a difficult task as the soil is highly compacted. Breakages tend to occur in the string joints and the cutting tubes need to be sharpened frequently. They often how after prolonged use.

Location

The precise site location of neutron probe access tubes should be considered carefully. The radius of detection is small and combined with soil spatial variability, can lead to unrepresentative sampling. Moreover, the variability of surface water infiltration, sometimes due to soil type, but more usually due to local runoff distribution, can be very great. Soil moisture in low areas generally penetrates further and a water table may be present in such locations. It is advisable not to plumb the water table, to avoid submersion of the probe. If field crops are under investigation, access tubes to 2 m, possibly 3 m, will be adequate. Much deeper access tubes will be necessary to study deep drainage, but there is no value in placing access tubes to great depth unless this is warranted; it merely increases the cost and time involved in installation, monitoring and analysis, to little purpose. Where the interest is to monitor the variability of soil moisture due to pronounced topographic features, access tubes should be located on or within the features. Where a field-wide representation is required, access tubes should be located on a predetermined grid, so that random sampling operates. The depth of access tubes should also be considered carefully where there is topographic variability.


Figure 5.6: Acces Tube Instalation Equipment

Neutron probe access tubes are relatively cheap to install and the temptation to over-monitor must be resisted; large quantities of raw data will be produced which must all be quality checked and analysed. Moreover, sites require constant monitoring on a routine basis and additional site visits after important rainfall events will be necessary. It is important not to over-stretch resources of time, transport and manpower. Access tubes that have been installed but which cannot be monitored, or backlogs of data that cannot be processed not only create a sense of frustration, but are a waste of time, money and effort that could be used elsewhere.

Operation

Standard counts are used as a routine method of checking the performance of neutron probe equipment. A variety of faults that cause a drift in the count-rate under identical conditions may be found, and standard counts should be performed before each time the probe is used and, preferably, after. An important use of standard counts is calibration between instruments, thus allowing their interchangeable use, a great advantage when planning any monitoring schedule. There are two types of standard count:

Water barrel counts are taken with the probe seated on an access tube section which can be conveniently mounted in a 200 litre oil drum. The drum should be thoroughly cleaned, water-proofed and filled with clean water. A three-barred support welded to the top edge of the barrel and holding the tube section in a vertical position is suitable. A length of tube, perhaps 25 cm, should extend out of the water. Ten counts of approximately one minute each are adequate.

Shield counts are taken with the probe retracted into the shield of the unit and with the unit standing on its carrying case. This method of checking the instrument is especially useful at field sites where water barrel facilities cannot be maintained, though during this calibration the same prevailing conditions are not so easy to replicate. Ten shorter duration counts, perhaps 15 seconds each, are sufficient.

The equipment will vary according to manufacturer, but certain aspects of operation are common. Care should be taken in packing and in transit. The battery supplies will need constant recharging at the design rates and cannot be expected to last more than a year or so; a good supply of spare batteries is essential. The probe cable is susceptible to damage with use as it carries the weight of the probe and detector unit and is continually being reeled in and out. Spares should always be available. Jamming of the source and detector in the access tube may occur and exacerbate the problem of cable damage.

Modern probes usually have depth indicators integrated with the rate-scalers. On the whole, rate-scalers that require a manual record of count-rates are best to purchase, although some types with integral memories are available as alternatives. The latter are perhaps are less robust and more susceptible to damage. If they malfunction data may be lost and the equipment is rendered unusable for some time. Neutron probes, though generally very reliable, are complex instruments and local repairs are often impossible. It is best to have one complete unit available as a spare.

Protection to the operator must be given by emphasising that while the radioactive source used in the instrument is very small, any exposure to radiation constitutes a needless risk. Radiation badges that are pinned to the clothing of the operator should be purchased, to monitor any exposure to radiation. In developing countries, it is likely that these will have to be sent overseas to be evaluated. It is also well worth remembering that the transport of radioactive material, in many countries, is restricted by certain laws. Although these may not pose problems for routine movement from site to site, they may require special customs clearance procedures to be effected. Before purchase is complete, the appropriate local authority should be consulted so that paths may be cleared and frustrating delays in obtaining the equipment can be avoided.

Calibration

Neutron probe count-rates need to be calibrated, as both individual instruments and soils behave differently. For example, high bulk densities give higher count-rates for the same soil moisture content. Neutron probes are supplied with calibration curves as a guide, but these are constructed in the laboratory and are seldom appropriate for the soils under investigation. Calibration curves developed by other researchers may be of interest, but generally the same problems apply. Calibration must always be undertaken if soils are suspected as being sufficiently individual to necessitate this. Different manufactures of instrument will greatly affect count-rate/soil moisture relations.

Field Calibration

In essence, calibration consists of taking count-rates at specified depths in the soil profile; for example at 10, 40, 70, etc. cm depths. Soil samples are then taken at these depths for soil moisture and bulk density analysis. It is important to note here, that compared to deep readings shallow readings may not be accurate nor consistent, because of the loss of fast neutrons through the soil surface. As a consequence count-rates tend to be lower and separate calibration curves should be constructed for the top 30 cm or so of the soil profile. Calibration is a relatively time-consuming business and due the nature of the method, will destroy the further usefulness of the access tube location. Proceed as follows:

- An access tube is installed, or an existing tube is used.

- Two wooden pegs are sunk 1 m either side of the tube and made level using a board and level. The soil surface is scraped level and the height above the board noted.

- The soil moisture profile is monitored at 10, 40, 70 cm below the soil surface, first with the probe at the height that readings are usually taken above the soil surface with the instrument sitting on the access tube (where applicable), then repeated after the tube has been tapped level with the soil surface. Several readings at each depth should be averaged, to minimise random errors.

- Duplicate cores taken close to, but on opposite sides of, the tube are removed with their mid point at 10 cm depth. At the same time a duplicate pair of cores are taken in the same way, at right angles to the first pair.

- Excavate a 1 m diameter soil pit accurately to 20 cm below the original surface and using the level board make it so for 35 cm radius around the tube.

- The tube is then sawn off, tapped down or replaced with a shorter one (the former has the advantage of less soil disturbance), then the 40 cm depth is re-monitored at +5 and 0 cm above the soil surface.

- Ten cm of soil is removed and the original 40 cm depth re-monitored.

- Core samples are taken once more.

- A new soil surface is located at SO cm below the original and the last three steps repeated for the original 70 cm depth.

- The soil samples should be analysed for volumetric water content and bulk density. The analysis of particle size distribution links calibration to soil textural type.

- Deeper horizons can be calibrated if it is felt that a need exists.

Each calibration, carried out as above, yields 3 points for the 10 cm calibration and 2 points for the 20 and 30+ cm calibrations. It is probably best to limit the number of such calibrations to one per day, to ensure good working practice, though two are possible and should enable soil moisture relations at all research sites to be established.

Calibration is necessary at the wet, intermediate and dry parts of the calibration curve, where seasonal variations in rainfall lead to extreme differences in soil moisture content. This can be undertaken after continual rain (or artificial wetting), after a period of drying and at the end of the dry season. Figure 5.7 shows an example of the field calibration of neutron probe data.


Figure 5.7: Field Calibration of Neutron Probe Data

Cross Calibration of Different Manufactures of Neutron Probe

It is quite possible that different makes of neutron probe may be used by the same project or associated projects. In this case it is preferable to compare water barrel and access tube counts rather than to repeat the procedures of field calibration for both types of instrument. Cross calibration may be regarded as essential and is undertaken as follows:

Count rates are taken in the water barrel access tube at 2 cm intervals from above the water level to 25 cm below its surface, then at 5 cm intervals to 35 cm. Averages of duplicate counts are taken for each instrument. In field access tubes, first one probe is used then the other, and this is repeated for average readings. Usual probing depths, for example every 10 cm, are used. Where instruments show accumulated count-rate totals, these must be converted to count-rate. Figure 5.8 show the 30+ cm calibration curve for some semi-arid soils, with Count -Rate/ Water Barrel Rate (the Count-Rate Ratio, R/RW) versus Moisture Volume Fraction. Figure 5.9 shows a graph comparing the countrates of one make of neutron probe with another.


Figure 5.8: 30+ cm Calibration Curve for 10 Semi-arid Sandy Loam Soils


Figure 5.9: Comparison of Count-Rate Ratios for Two Manufactures of Neutron Probe

d. Time Domain Reflectometry (TDR)

TDR measures soil moisture content by utilising large differences in the dielectric properties of soils, water and air. Methods exploiting these soil/water properties have been in development for several decades, but it is only very recently that TDR has become a convenient, practical tool for soil moisture investigation.


Figure 5.10: Diagrammatic Layout of TDR equipment

The advantages of TDR equipment over the neutron probe are that it can be logged continually, it is relatively non-destructive and does not utilise radioactive materials. Figure 5.10 shows a diagrammatic layout of the equipment and Figure 5.11 the idealised waveform of a wet soil.


Figure 5.11: Idealised TDR Waveform of a Wet Soil

Principles of Operation

A pulse generator creates a fast rise-time voltage pulse (about 10-10 seconds) which passes through a transmission line to a balun transformer designed to achieve maximum transmission into the soil. The pulse then passes to the waveguide (a pronged "fork" or "probe") in the soil. An impedance mix-match causes part of the pulse to be reflected back to the instrument, while part is propagated to the end of the waveguide, from where it is then reflected back to instrument. The analysis of the waveform (change in voltage) of the pulse that is reflected from the ends of waveguide is the key to measurement. The interpretation of the waveform is somewhat subjective.

Point C shows the rise-time of the pulse (usually 10-10 to 30-10 seconds), the point A represents the location of the balun transformer, the distance between C and A depends upon the length of the transmission cable. Point B is the reflection from the end of the waveguide. The travel time 't' is the time in nanoseconds for the pulse to pass from A to B and back. Pulse attenuation, the loss of magnitude of the signal, increases with moisture content. The calculation of the apparent dielectric constant, Ka, (approximately 80 for water, 1 for air and 2-4 for soil minerals) is made by:

Ka = (t c/L)2 where (5.3)

t = travel time
c= speed of light (29.979 cm ns-1) and
L= length off the waveguides in cm.

TDR instruments can be "multiplexed", utilising a variety of waveguide lengths set to log continuously on a series of channels at a given time interval and controlled by a small computer. However, it may take half an hour to obtain a set of measurements and any rainfall during the logging cycle could give confusing results if the logging cycle were short. Continuous monitoring opens many opportunities for the detailed examination of moisture changes, especially at shallow depths, but generates a large amount of data.

Much development has concentrated on waveguide ("probe") form. There are two main types:

Unbalanced waveguides have three prongs, the inner prong, for example, carrying a positive charge and acting as the centre of a coaxial cable, the two outer prongs carrying half a negative charge and acting as the shield.

Balanced waveguides have only one outer prong and the lack of shielding is compensated for by a balun transformer acting as a shield. They are regarded as more sensitive and less destructive, but are more expensive and can hinder interpretation by inducing interference on the waveform.

Practical considerations

The sampling volume of the instrument is very small (radius of sampling within about 1 mm of the waveguide surface) and may be greatly effected by the near-surface environment, though research into these effects is as yet inconclusive. Waveguides are (usually) placed horizontally at various depths, avoiding the penetration of soil layers with different characteristics. This often necessitates the excavation of a trench behind the locations of the waveguides.

The length of the waveguide prongs is an important consideration. Usually they are between 0.05m and 1.0m long. The suitability of length is determined by the attenuation characteristics of the soils under examination. Long prongs offer a greater volume of sampling, but are more difficult to install and are not suitable for highly conductive soils. Spacing between prongs of the waveguide is usually no more than ten times their diameter, with an upper limit of 5 cm. Wider spacing leads to difficulties in the interpretation of the (greatly attenuated) waveform. Increasing the transmission cable length decreases resolution and a maximum of 30m is recommended.

The effects of soil mineralogical composition on TDR is not fully understood. Research has indicated that heavy clay soils may give anomalous results and clay content; bound water and bulk density have been shown to influence calibration. The dielectric constant of a soil is temperature dependent and therefore diurnal variations may be significant, especially near the soil surface. The influence of organic material is not yet fully understood, though increasing organic content tends to reduce the calibration slope of Ka versus water content. Increasing the bulk density of soils of a given water content decreases the dielectric constant of such soils, and the installation of the waveguides should disturb the soil as little as possible. The reduction of air gaps due to installation may be achieved by wetting the soil before insertion, though the characteristics of soils will determine the effectiveness of the installation. TDR is unsuitable for use in stony soils, and in soils that swell and shrink. Waveguides should not become the focus for cracking and enhanced percolation, nor be sited cross distinct soil boundaries. Soils high in iron and titanium (and possibly aluminium) minerals have enhanced dielectric constants. A guide to operational difficulties in various soil types is given below.

Vertisols: Vertical installation should not be undertaken. They may have a low dry bulk density due to cracking. Organic content may be high; electrical conductivity may effect attenuation.

Entisols: Relatively unstructured soils with little organic matter.

Aridisols: Conductivity may be affected by salts and high temperatures close to the soil surface.

Mollisols: Grassland soils, often with high organic contents in the upper horizons.

Alfisols: No obvious problems.

Ultisols: High content of iron and/or aluminium may exaggerate moisture readings.

Oxisols: Low bulk densities may reduce the slopes of calibration. Temperature effects may be seen and possible influence of oxides which increase conductivity.

Spodosols: Sharp density changes down the horizon. A high organic matter content in the A horizon may lead to the underestimation of soil moisture content.

Histisols: High in organic matter and will need separate calibration. Bulk densities tend to be low.

Calibration

Calibration is undertaken for values of Ka versus volumetric water content and may be done in the laboratory of field.

Laboratory calibration:

Research in the past has concentrated on the water content of substitute materials rather than soil, for example vermiculite, glass beads and washed sand. More recently soils have been used and indicate separate calibration curves for organic and mineral soils. However, as for the neutron probe, laboratory calibrations do not accurately represent the spatial variability of such factors as texture, bulk density and organic content.

Field Calibration:

Field calibration may involve the measurement of soil moisture by lysimetery, neutron probes and gravimetric methods, though sampling and oven drying gravimetric methods are standard. Sampling may be undertaken in a manner similar to that described above, for neutron pobes. At water contents of greater than 20%, laboratory calibrations were seen to overestimate moisture content. Differences between neutron probe and TDR measurements have been found to be large (10 - 20%) and may reflect the different sampling volumes of the two instruments. These results indicate that the use of TDR should be carefully examined and TDR may not be suitable for water balance studies, though its more precise and accurate measurements can be effectively used in rooting zones and areas of marked changes in water content over small distances.

Waveform Interpretation

Interpretation determines the value of Ka that is derived from the waveform. An example of a Ka calculation is given below, with the use of Figure 5.12.


Figure 5.12: Hypothetical Waveform

From equation 5.3, Ka = (6 × 29.979/20)2 = 80.89 (the value of Ka is dimensionless)

Figures 5.13 (a), and 5.13 (b) show the idealised waveforms of air, water.

Figure 5.14 shows how the waveform changes with increasing and soil moisture content.


Figure 5.13 (a): Waveform, Air


Figure 5.13 (b): Waveform, Water


Figure 5.14: Waveform, Different Soil Moisture Contents

Below are example waveforms for different soil conditions that may be encountered. Figures 5.15 (a) shows the waveforms of a waveguide installed vertically through a wetting front and (b), a dry zone over a wet zone.

Soils with increasing percentages of iron oxides display a waveform of decreasing amplitude and increasing travel time. Relatively dry, clay mineral soils frequently exhibit a large amount of background noise which makes interpretation difficult, while wet clays often exhibit large pulse attenuations, probably due to high pore water electrolytic concentrations.


Figure 5.15 (a)


Figure 5.15 (b)

Figure 16 shows the effect of increasing salinity. The travel time is not changed, but the magnitude of the pulse decreases as the salinity, and conductivity, of the soil increase.


Figure 5.16: Waveform, Increasing Salinity

e. Capacitance Probe

Capacitance probes, like TDR, have only recently become practical field instruments for measuring soil moisture. They also measure the dielectric constants of soils, like TDR, but are relatively low cost. They are usually utilise an access tube, but recent designs will allow direct insertion into the soil and readings are usually taken from a hand-held meter. It is necessary to calibrate these probes for individual soils.

5.3 Infiltration

Infiltration is the process whereby water on the soil surface percolates downwards. Infiltration rates represent the speed at which this percolation occurs and are expressed in mm in-l. The maximum rate for any given soil condition is termed the "infiltration capacity". The controls on infiltration rates are many: soil texture, cavities and impermeable layers, vegetation cover, air spaces, soil wetness and topography all may be influential. Land use and cultivation can also be extremely important as they affect the quantity of suspended material in surface water which in turn influences rates of infiltration, because the suspended material blocks pore spaces and increases runoff. Measured rates of infiltration lump the effects of all these influences together.

Runoff is the proportion of rainfall that does not infiltrate, at least immediately, but infiltration rates vary greatly with time, especially during rainstorms when the soil becomes progressively wetter and as rainfall intensities vary. Thus the supply rate of water to the soil surface and the rate at which it infiltrates are never constant. Despite the fact that infiltration rates represent a gross generalisation of soil/water behaviour, they are often important components of hydrological and runoff models.

Extensive field trials have shown that infiltration rates decrease with time, in the general form:

I = (aTn + b) where (5.4)

I = infiltration rate
a, b and n are constants and
T= Time elapsed

Thus infiltration rates are exponential. As the rate of infiltration decreases it approaches, and sometimes achieves, a terminal value. For clay soils, the value of 'b' in equation 5.4 may be almost zero, while for sandy soils it will be much greater; soil texture plays an important part in the determination of infiltration rates. Infiltration rates are used not only to estimate the likelihood of runoff, but are also quoted as a general soil characteristic, but it should be noted that the infiltration rates of soils are notoriously spatially variable, even over distances of a few metres.


Figure 5.17: Variation of Infiltration Rates Over Small Areas

Variability is important because it can be associated with microtopography in fields and cultivation practices which alter the location of topsoils. For example Figure 5.17 shows the variation of infiltration rates of 20 tests undertaken over an area only 15 m square, on an apparently uniform soil. Variation will affect the redistribution of runoff at the local scale and will be influential in determining the soil moisture that is available to plants. Important differences between rates on cultivated and uncultivated land are likely to be seen.

Figure 5.18 shows example infiltration curves for different soil textwal types: sandy, loamy and clay soils.


Figure 5.18: Example Infiltration Curves for Sandy, Loamy and Clay Soils

Three methods of measuring infiltration are discussed below.

5.3.1. Equipment and Methods of Measurement

a. Double Ring Infiltrometers

Double ring infiltrometers consist of two concentrically placed rings, both filled with water. The rate of infiltration into the ground within the inner ring is measured, while the water in the outer ring provides a buffer to ensure the direct downward movement of water below the inner ring.

Infiltrometers can be purchased, but as they are simple to manufacture to specific requirements, this may be preferred. They can be made by using metal water pipe cut to a suitable length and given a bevelled, sharpened edge at one end. Example dimensions are: length 30 cm, diameter 60 cm, the smaller ring should be of the same length, but half the diameter, also sharpened. The two different sized rings are inserted into the ground, the depth of insertion will depend on the hardness of the soil. A more or less constant head of water is maintained at a measured and marked level (10 cm is suitable) above the ground surface. Water is poured into both rings until it reaches this level and then another 1 litre of water is added. When the water in the inner ring falls to the marked level another litre is poured in, to compensate for infiltration. If infiltration is slow, only 0.5 litre need be added; the water in the outer ring is also kept to the level and generally it will take one or two hours to reach a constant (terminal) rate of infiltration. Figure 5.19 shows the installation of the double ring infiltrometer.


Figure 5.19: Double ring Infiltrometer.

The accuracy of the data collected from infiltrometers may be affected by the insertion of the rings, which can alter the physical characteristics of the soil. Soils with macropores, burrows etc. may show extreme rates of infiltration and prove to be unsuitable for study by this method. Crusted soils can be prepared by cutting the crust with a razor blade and inserting the rings through the cuts. The gap between the ring and soil may be sealed with gypsum paste or hydraulic cement.

The major difficulty with data that are obtained by the double ring method however, is that they do not represent infiltration under rainfall/runoff conditions: rainfall impact may increase or reduce infiltration at the soil surface; rain storm intensities vary greatly and the standing heads of water used by these infiltrometers are not usually representative of real conditions. Runoff flows away and is not impounded. The spatial variability of infiltration capacities is common and results may be relevant only to very small areas. It is important to note, however, that data from double-ring infiltrometers are widely used and it may be essential to collect this information if comparisons between a range of sites and soils are to be made. The dates' main value is as a reference for comparison rather than the provision of absolute, true infiltration rates, though these may be of direct application to irrigation practice. To overcome the unrealistic results that double ring infiltrometers often provide, sprinkle infiltrometers have been developed. An approximate replication of natural rainfall can be obtained by the use of sprayers and sprinklers. Sprinklers can be complex systems that simulate duration, rate, drop size etc., but even the smallest versions of these instruments are not really portable and often require large volumes of water to operate. Knapsack sprayers are low cost alternatives that can be easily transported and when used carefully, provide relatively good data.

b. Knapsack Spray Infiltrometer

Spray infiltration tests using knapsack sprayers improve upon the results from ring infiltrometers by minimising the effect of a standing head of water and by applying the water as a spray. Further emulation of the rainfall/ infiltration process is not attempted. Although this method is only a rudimentary attempt to simulate rainfall it does provide an improved technique which is portable, easily replicated and inexpensive. Typically, a quadrat is sprayed at a designated rate which is reduced upon the evidence of standing surface water. Details of the method are as follows:

An area 1 m square is marked and within it a 50 cm × 50 cm × 5 cm deep wooden quadrat is placed centrally, pushed 2 -3 cm into the ground. The central quadrat is divided into quarters. Wind shields should be provided around the site if necessary. The sprayer is best fitted with a fan nozzle to provide a wide, even spray. The 1 m² quadrat is sprayed with an even application every 30 seconds. The period of pumping to prime the sprayer and the duration of application are regulated (for example pumping and application for 5 and 10 seconds respectively) to give a similar intensity spray each time. When water is seen standing on two or more of the central quarters, the next spray is omitted. This may be continued for an hour or more, or until a recognisable uniformity of application indicates stability of the infiltration rate. The amount of water delivered is quantified by repeating the spray test procedure into a measuring vessel before and after the test. The total volume of water that can be applied during the test will depend upon the nozzle aperture, but depths of 40 mm can easily be achieved. Variations on this method may be designed to account for local conditions.

This method, though not representing true rainfall conditions, is relatively simple and easy to replicate and the data obtained discriminate clearly between soils of different texture. The rates also resemble those that might be expected under rainfall conditions. Figure 5.20 shows a vertical view of equipment layout used in running tests with a knapsack sprayer.


Figure 5.20: Knapsack Infiltration Test Equipment (Vertical View)

Testing may be undertaken on cropped, range or fallow land. The simulation of ploughing effects can be achieved by digging the soil over; this may be necessary in regions that have distinct dry and wet seasons. In these areas testing is most suitably undertaken during the dry season, when the influence of rainfall and high levels of antecedent soil moisture are not evident. It is likely that fields will not be cultivated at this time.

c. Sprinkler infiltrometers (Rainfall Simulators)

Unlike double-ring infiltrometers, sprinklers are not limited to the study of infiltration rates. They are often used to investigate such influences on the rainfal/runoff process as soils, slopes and tillage practices and may be used to measure rates of soil erosion. It is convenient however, to discuss this type of equipment here. Sprinklers attempt to simulate the process of rainfall, while allowing a control over the amount, intensity and drop size of applications in a manner that is not possible with natural rainfall. Though they represent rainfal/runoff conditions more realistically than ring infiltrometers and portable knapsack sprayers, sprinkle infiltrometers also have limitations. An important question to ask when considering sprinkler design is which of the main rainfall characteristics should be simulated most closely: for example drop-size or terminal velocity? Variations in intensity or uniform applications? Difficulties exist even in measuring the characteristics of the natural rainfall that is being simulated.

Design

Sprinkler design can be esoteric, and "production models" are not common. Most sprinklers cannot be considered at all portable, they are too large and even small sprinklers may need compressors, pumps or a mains water supply to operate. Large, boom type sprinklers are usually confined to agricultural research stations where there is a plentiful supply of water and where it is preferable to maintain tight control over the soil, slope, cover and tillage variables under investigation. As runoff and soil loss are related to rainfall kinetic energy per unit area, this is a useful parameter by which to make comparisons of sprinklers. In general two types of rainfall simulations are adopted.

The first type are those using nozzles which most easily reproduce a drop size distribution akin to natural rainfall, but which have complex intensity reducing systems.

The second use drop-formers and are simpler in construction, but the drops do not reach terminal velocities until falling 5 m or more.

It is far beyond the scope of this book to describe the many individual types of sprinklers that have been developed, often for particular research purposes, and which are not easily nor commercially available. A comparative list of such equipment is given in Part 1 of monograph no. 9 of the American Society of Agronomy and Soil Science Society of America (1986). Most sprinklers of the "portable" kind apply water to relatively small areas, usually about 1 m². The problem of spatial variability of soil characteristics is therefore often as great with these devices as it is with ring infiltrometers and knapsack sprayers, and the extrapolation of results beyond the locality of application needs careful consideration.

The Type F infiltrometer has been used in the USA on larger plots of approximately 2 m × 4 m in size and is not regarded as portable. the type FA operates over a smaller area. The manner of operation to obtain infiltration data is recommended as follows:

- First, several calibration runs are undertaken with the test area covered by a waterproof sheet, to measure the simulated rainfall application rate.

- A test run is then started with the sheet removed and continued until the rate of runoff becomes constant (as does the rate of infiltration).

- The analytical run is started when application ceases and runoff stops, but before any recovery of the infiltration capacity has occurred. The rate of infiltration is therefore constant throughout the run. Runoff is measured. Any difference between the application minus infiltration and runoff is due to depression storage and detention storage.

- The effects of these on the runoff process may be investigated for various land conditions, if required.

The relation between runoff and infiltration data is discussed further in chapter 8, Data Analysis

Manufacture

The most important components of sprinklers are the nozzles that control the characteristics of the water that is applied to simulate rainfall. The testing of suitable nozzles on an individual basis was undertaken by the Ministry of Agriculture, Harare, Zimbabwe as part of research into the SLEMSA soil erosion model (see chapter 3), over 1 m² areas. Of the various nozzles tested, several performed well and details are given in the Ministry's Research Bulletin no 25 (1980). The construction details of a mobile sprinkler taken from this bulletin are given Appendix C.

Equipment costs

All costs of locally made equipment are approximate. The costs of raw materials and especially labour are highly variable from country to country, but a good idea of cost magnitude can be gained from the figures quoted below. The costs of manufactured equipment are based on 1993 prices. Shipping, agents' fees and fluctuations in exchange rate cannot be taken into account.


Table

Appendix C: Soils and soil moisture

Appendix C1: Construction details of a mobile sprinkler system


Figure


Plan on A - A


Windshield frame


Plan on driving dolly


Fig. 1: IH12 nozzle assembly details


Fig. 2: Size distributions for IH12 nozzle

(introduction...)

Catchment characteristics interact with variable patterns of rainfall and determine the character and size of runoff volumes and peak flows. This is true for both natural catchments where human activity is absent or unimportant and runoff plots upon which tillage or other agricultural treatments are being tested. Generally, there is a hierarchy of influence imposed by different characteristics, but this hierarchy is often difficult to sort out and understand. For example, where a catchment has high slopes and lime vegetation, slope will play a major role in determining the runoff regime. This regime would tend to exhibit high runoff proportions; rapidly increasing flows to high peaks and equally rapid falls. Were this catchment to be of a more linear form and were its vegetation cover to increase, then peak flows would be smaller but more prolonged and total runoff volumes would probably be less. Were the slope lower, runoff would probably be less.

In the case of catchments with low slopes, the effects of vegetation cover and microtopographic features often exert a stronger influence over runoff than the overall land slope. Local slopes are often relatively high and they may direct runoff either into basins where it can infiltrate or to channels by which it can easily leave the catchment. Heavy textured soils tend to give a higher proportion of runoff. Soil textures are related to slope as well as to parent material, and the climatic regime under which the soil formed will often have been a determining factor of the soil textural type. Where human interventions have been imposed the natural conditions of a catchment may have been altered radically; grazing, tree-felling and clearance are obvious examples. Agricultural techniques; ploughing, bunds and microcatchments are introduced to reduce runoff and usually they do, but the removal of natural vegetation and badly managed systems can have the opposite effect. Below, the main catchment characteristics and their influence on runoff are discussed.

6.1 Natural vegetation

Natural vegetation can be very important in determining runoff amounts; in many instances it is the most important influence of all, after rainfall. Areas bare of vegetation can lose more than 40% of seasonal rainfall through runoff and for intense, individual storms the loss can be much greater. Areas with dense grass cover and tree canopy cover can retain as much as 99% of the rainfall that reaches the ground. Vegetation reduces the energy of raindrops making them less erosive and intercepts rainfall which is then re-evaporated. Thus natural vegetation works against the occurrence of runoff in several ways. The same can be said of crops, but most crops provide only temporary cover and their densities, especially at ground level, rarely attain that of natural vegetation. Examples of increased runoff, soil erosion and subsequent land degradation due to the removal of natural vegetation, are common throughout the world and the literature. Consider the data presented below in Table 6.1, which compares runoff from different rangeland catchments of the same size catchments, but with various densities of vegetation cover.


Table 6.1: Comparison of End of Season Vegetation Cover and Seasonal Runoff

Note that no account is taken of other factors that influence runoff production and that the coefficient of correlation between runoff amount and vegetation cover is 0.91.

6.1.1 Measuring Vegetation Cover

Plant biomass represents the total quantity of vegetation over a given area at any time and may be variable both within and between seasons. It might be expected that the quantification of biomass is the best indicator of vegetational influences on runoff. However it is probably not the most practical index for runoff studies, because the quantification of biomass is extremely time-consuming; a large number of samples must be taken and mapped in detail and size/mass relations must be determined by the destructive sampling of trees. Generally, the assessment of total plant biomass is unlikely to be relevant to agrohydrological and water harvesting projects.

The form of vegetation; leave-shape, density, branching pattern, etc., is highly variable between species and groups of plants. Although these differences are implicit within the classification of plant species, their effect on rainfall/runoff relations are very difficult to quantify. Research into commonly-occurring trees (and crops) has been undertaken, but the results of this work is understandably limited in its applications. Moreover, biomass and vegetation cover are usually very closely correlated and the use of vegetation cover as a proxy for biomass in runoff analysis, is a legitimate substitution.

Figure 6.1 show example correlations between biomass and vegetation cover and vegetation cover and runoff.


Figure 6.1: Biomass and Vegetation Cover


Figure 6.1: Vegetation cover and Runoff

In contrast to biomass measurement, there are rapid methods of quantifying the areal extent of total vegetation cover and even though effects due to vegetation type are not always accounted for, this index makes a good indicator of the influence that vegetation can have on runoff.

Vegetation cover assessments may be undertaken on a frequent basis to study its effect on runoff, almost storm by storm. Alternatively, assessments may made only a few times each season, to understand its role in the production of runoff over longer periods. The latter case is most common, because the variation in influence of vegetation cover is not dramatic in the short term, except where wholesale removal is involved. Vegetation cover is not closely correlated to other factors that influence runoff (except perhaps seasonal changes in rainfall and temperature), but may considerably alter the soil moisture status by evapotranspiration. The relative independence of vegetation from other variables makes it a suitable factor for use in regression analysis. On the other hand because it does vary with time, unlike factors such as slope, soil type and catchment size, vegetation cover can provide an extensive range of data points for individual catchments. It lends itself well to and is often used in, runoff modelling. In general it has been recognised that the amount of vegetation cover present is a more influential factor than the type of cover. The pattern of spatial distribution of vegetation cover may also be very important.

For agrohydrological purposes, there are two conventional methods of cover measurement, though the collection of aerial photography and satellite remote sensing data are also discussed below.

a. Quadrats.

A quadrat is a defined area. For field purposes, quadrats are usually permanent sampling areas retained throughout the season, within which the extent of vegetation cover is assessed. Prior to the field visit that will install them, a suitable number of quadrats is decided upon and these are placed on a site map using a fixed grid pattern. The quadrats are laid out in this predetermined, regular manner to overcome subjective bias and attain a random sampling of the area. No strict percentage sampling is required, though the more quadrats, generally the better. Ten 2m × 2m quadrats to sample 1 ha (10,000 m²) would be adequate. The quadrats are then subdivided into four sectors to facilitate accurate assessment. The quadrats can be made easily by using steel rods driven into the ground with perimeters defined by nylon rope or string. Each quarter of the quadrat is individually assessed by eye for percent total ground cover of live and dead vegetation. The overall estimate of cover is made as an average of all sectors and all quadrats. In cases of natural vegetation, any trees are included in the assessment. Assessments should be made as frequently as is feasible throughout the season, though the rate of growth will largely determine the need for inspection The estimates are to some extent subjective and it is a good idea to compare those made by different field staff, under the same conditions of cover.

This method is suitable for small areas and can be completed quickly, but projects that need to quantify cover accurately on large plots and small catchments can utilise a rapid method that is detailed below.

b. Wheel Point Method

This method is based on the simple equipment shown in Figure 6.2.

A bicycle handlebar is fitted with an extended fork assembly. The extension is made long enough to allow the passage between them of strong, sturdy spokes. The spokes, made of 5 - 10 mm diameter mild steel, are welded to a supporting plate which in turn is fixed to the axle. The forks can be any convenient length, but it is advantageous that when the observed, marked fork completes a revolution, it travels an easily recorded horizontal distance, for example 1.0 or 1.5 m. Versions with longer or shorter spokes can be made, according to whether the areas to be covered are small or large, to maintain a sufficient number of data points per unit area of catchment.

As an example, consider a plot 100 × 40 m in extent. A tape measure is stretched across one of the longer sides of the plot, 5m from the end. The apparatus is held with the marked spoke at the start of the tape and then walked along using the tape as a direction indicator.

If the marked spoke hits a bare area on touching the ground, this is called out. If it touches a vegetated area on the ground this (and if required the type and species of plant) is called out. A second person notes the call. The tape is then moved on 10 m and the process is repeated, until the whole plot has been covered, the last transect being 5 m from the other end of the plot.

This procedure gives approximately 600-700 data points for each hectare that is surveyed and takes about one hour. Less frequent sampling by using more-widely spaced transects is permissible in areas where the vegetation cover is relatively uniform. This method is also easily adapted for larger catchments and can be used in difficult and wooded terrain, with practice.


Figure 6.2: Wheel Point Apparatus

c. Aerial Photography

Aerial photographs and, more recently, satellite imagery can play an important role in the assessment of many aspects of agrohydrology, vegetation cover being one of them. Clearly, this method is inappropriate for small runoff plots, but for natural catchments it can be very useful. Large areas can be viewed quickly and catchments that are otherwise difficult to survey on the ground (those with dense tree cover, or that are inaccessible) can often be mapped much more effectively and cheaply. Additional information on surface flow routes, areas of flooding, land use, microtopography and agricultural features can also be obtained at the same time. The simplest methods of obtaining and using aerial photographs are discussed here.

Aerial photographs are used to compile maps and are often available from local survey departments. The main advantage of this is that once obtained, no further effort is needed before assessment can begin. There are, however, some serious drawbacks:

- Aerial photographs are often restricted material in many parts of the world and you may be refused them.

- When available, they are often at a scale of 1:50,000 or smaller. This is often unsuitable for detailed mapping.

- Enlargements can be made, at conventional scales, for example 1 :10,000 or 1:5,000. These are much more useful, but facilities for enlargement may not be available.

- They are almost always in black and white panchromatic format, which is poorly suited to vegetation studies. - In areas with marked seasonal differences, they will almost certainly be taken during the dry season when conditions for photography are best, but little information is available on vegetation or crop cover.

- Photographs for mapping purposes are not taken frequently and different sets of photos may be decades apart, ground conditions may have changed radically since they were obtained.

If suitable orthodox photographs can be obtained, fine, but it is well worth considering obtaining your own. This is much simpler and cheaper than may be expected and has several advantages:

- Photographs can be obtained at the most useful scales. The use of slides allows a range of scales to be obtained.

- Colour or infra-red photographs can be obtained (though the latter film may be difficult to buy and have processed).

- If slides rather than prints are taken, these are very useful for projection, mapping and conversion to prints.

- They can be taken at critical times during the season.

- Particular sites or areas can be selected.

It is unlikely that the precision of scale and lack of distortion of map survey photographs can be equalled, but in most cases these are of minor importance compared to the advantages listed above. The general conditions to obtain good quality photographs economically are as follows.

Any light aeroplane (2-3 seat) can be used. Enquire if a glass panel can be easily inserted into the floor to give a vertical view or if this modification has been made previously. If not, a door will have to be removed and a wind shield fitted (this is not unusual, but vertical photographs will be more difficult to take). Plan the most economical route to all sites and submit a flight plan to be discussed with the pilot. As a guide, three sites situated within a 50 km radius of the airport can be covered in little more than one hour.

A good 35 mm single lens reflex camera (through-the-lens viewing is essential) is adequate. The type of lens is a point of preference and the aims of the photography will play an important part in the choice, because although the focal length of the lens will determine photographic scale and is technically important, the ease of use in the confined space of the cabin, the ability to work rapidly and the need for different scales may be paramount.

A 70 - 210 mm focal length zoom lens will probably be suitable for most occasions since it gives approximately × 1.4 to × 4.0 magnification. This flexibility of magnification means there is no need to change the lens to cover different sized areas efficiently. Unless very small areas are to be studied, a 35 - 150 mm zoom would also be suitable and in this case slightly wide angle views can also be obtained. Another advantage that zoom lenses have is that their magnifications obviate the need for the aircraft to change altitude.


Figure 6.3: Scale of Vertical Photograph Over Flat Terrain

In a small plane this can take a long time and can add considerably to the cost when several sites are being photographed at different scales. It is important to remember, however, that with zoom lenses, the exact focal length currently in use may be unknown and the scale of the photograph cannot be calculated, unless ground reference points of known dimensions are available. Sometimes it is best to preview the area to be photographed and tape the focal length of the lens in a fixed position with adhesive tape. It is not necessary thereafter to be continually manipulating the lens and the tape prevents it from accidentally sliding out of position when held vertically downwards.

Lenses of single focal lengths overcome these problems, but time must be allowed to change them and only a limited range can be used. As sunny conditions will undoubtedly prevail, film and shutter speeds are not usually a problem. A large depth of field is not needed, so wide aperture stops can be used to give high shutter speeds. To prevent blurring due to vibration, 1/500 th or 1/1000 th of a second exposures are recommended. Fast films (ASA 400 and above) should not be necessary and may not be available nor be easily developed. They tend to be grainy when enlarged.

Films should be at hand and clearly marked with date and location. Ground location markers may be necessary for site identification. At 2,000 - 3,000 feet (650 - 1000 m) above ground level, a good operating altitude for light aircraft, strips of white paper about 30 cm wide and 10 - 20 meters long are clearly visible. If they are set to known lengths, they make good ground reference markers for obtaining scales.

Photographs taken over terrain of widely varying altitudes exhibit varying scales and tilted photographs have nonuniform scales.

Table 6.2 below gives a guide to ground coverage with various altitudes and focal lengths . This is the actual area on the ground that will be captured by a 35 mm negative or slide diapositive of size 25 mm × 36 mm.


Table 6.2: Ground Cover Area for Different Altitudes and Focal Lengths

The largest area covered in the table above is between five and six square kilometres. This is the size of a small catchment, but details on the ground are not easy to see.

A mosaic of photographs, or a continuous transect of frames that cover a large area but which also show fine detail are possible, but not easy to obtain. Transects can be planned on maps and air speeds calculated so that photographs may be taken at counted time intervals, without taking account of the view below. In practice, pilots find it difficult to keep a straight course with only a visual marker on the horizon and airspeeds vary due to wind. Drifting causes further problems. To some extent trial and error must play a part, but care and acute observation must be exercised to obtain reasonable coverage using transect flight paths.

d. Satellite Remote Sensing

During the last two decades or so, satellite imagery has become more widely used for water resource projects, among others. The importance of such imagery cannot be overstated, but the area of satellite image analysis is a very complex one and can only be covered here, very briefly.

The three main factors that dictate the usefulness of satellite imagery to a project are:

Orbital parameters

These define the potential repeat period for the coverage of an area. For example the polar orbiting NOAA satellites can obtain imagery at least once per day per satellite. The Landsat satellites have a repeat period of about two and a half weeks. The altitudes of various satellites are also greatly different and will affect ground resolution and size of coverage.

Sensors:

Satellites, their orbits and sensors are designed for particular purposes. For example, Landsat satellites were designed for terrestrial research, Seasat for oceanographic study and Metsat for meteorological investigation. Different sensors are used to give the best results within a particular environment and may have restricted use outside that environment. Visible, infra-red, near infra-red and micro-wave (radar) sensors are commonly used, each of which is most suited to a particular application.

Resolution:

The size of an object that can be detected from a satellite, depends upon the resolution of the sensor, this may vary from a few metres, or even less, to several kilometres. It will also depend on the kind of sensor that is deployed and the spectral characteristics, shape and surroundings of the object that is viewed. In general, the area of coverage is smallest when resolution is finest, but in all cases coverage is "regional".

Imagery comes in two formats; hard (usually photographic) copy and computer compatible tapes (CCTs). The former may be colour (a combination of bands) or black and white (single band) and is relatively cheap and easy to work with. It will be purchased in a form that has been geometrically corrected for changes in satellite velocity, altitude, attitude and for Earth rotation and curvature. CCTs must be viewed using special computer facilities, desk-top versions of which are now widely available. These images can be extensively processed and enhanced and are the source of hard copy images. They and the equipment to process them are usually very expensive, though research institutions can in some cases, gain the image material for no, or little, cost.

Vegetation cover assessment is commonly undertaken using satellite imagery and the physical characteristics of catchments, their soil moisture status and hydrology can also be studied. However, the selection of satellite, imagery and waveband; the selection and utilisation of techniques for analysis is extremely complex and specialist literature should be consulted.

6.2 Interception

Interception can only be loosely defined as a catchment characteristic as it is the combined effect of several influential factors such as rainfall, climate and vegetation cover. However, in other respects it falls conveniently into this chapter and so is discussed here.

Losses from interception, the rainfall that collects on vegetation and is re-evaporated, can be highly variable and depends mostly on vegetation type (size, shape and disposition of leaves and branches); rainfall amount, intensity and drop size; wind speed, temperature and eddying. Interception is difficult to measure, especially for crops. It can be attempted by placing rain gauges under vegetation either randomly to sample average interception, or by the selection of specific target areas. In wooded catchments, rain gauges should be attached to tree trunks to assess stem flow, as in Figure 6.6 below, but with multi-stemmed vegetation this is very difficult.


Figure 6.6: Stemflow Measurement on trees

In Figure 6.6, in addition to free-standing gauges under the canopy, a peripheral collector is wrapped around the trunk to direct flow into a single rain gauge that is covered.

Empirical work has led to estimates of losses of 10 - 20% of seasonal rainfall and deduced storage capacities of 0.8 to 1.5 mm of rain per storm. Equation (6.1) describes an empirical interception relation and Table 6.3 gives examples for various crops for a 25 mm rainfall.

I = (Si + Etr) (1 - e-kP) where (6.1)

I = total interception
Si = storage capacity per unit of the area
E = evaporation rate
tr = duration of rainfall
P = amount of rain k = 1/ (Si + Etr)
e = base of natural logs.

In terms of runoff studies, the situation regarding interception is even more complex. It is usually lumped with rainfall storage due to ponding and infiltration for runoff modelling purposes, where it is assigned a purely notional value.

Table 6.3 Interception Losses from a 25 mm Rainfall

Crop

Height (m)

Interception (mm)

Maize

1.8

0.8

Cotton

1.2

8.4

Tobacco

1.2

1.8

Small grains

0.9

4.1

Meadow Grass

0.3

2.0

Alfalfa

0.3

2.8

6.3 Catchment size, slope and topography

6.3.1 Catchment Size and Land Slope

Catchment size is an important influence on absolute values of runoff amount and peak flows and is an essential parameter in runoff formulae that predict these hydrological characteristics. The determination of catchment size will be straightforward in most cases. Runoff plots are usually bounded by bunds or galvanised metal sheets that prevent runon from outside the proscribed catchment area. Natural catchments will usually be defined by clear patterns of drainage and topographies that show the limits of a catchment area. In some cases these details will be available from topographic maps, in others aerial photography may be the most suitable source of information. In general, the size of a catchment that is monitored will be limited by the practicalities of the natural or artificial controls that can be used as flow measuring sections, the aims of the project and the resources that can be invested in obtaining runoff data. Catchment size is not a good indicator of percent runoff; influences such as land use, soil type and slope are more important, but in terms of absolute values catchment size is very important. It is unfortunate that a simple proportional reduction or increase of runoff cannot be deduced from the size of a catchment, even where catchment conditions are ostensibly the same ( see the section on slope and microtopography below). To illustrate the difficulties in making assumptions on runoff proportion and catchment size, Table 6.4 gives percent runoff for large catchments, R2 0.12 and is not significant.


Table 6.4: Relation Between Catchment Area and Runoff


Figure 6.7: Catchment Size versus Runoff from Experimental Plots

The scale of these catchments is larger than is often studied for agrohydrological research, but Figure 6.7 shows a graph of catchment size versus percent runoff, the data for which were obtained from experimental plots and catchments sited in and around farmers' fields. These plots are divided into three groups with similar catchment conditions, to remove any influence that different conditions could exert on runoff. The conditions are crop (squares); rangeland (triangles) and fallow (circles). The R2 of the analyses were 0.108, 0.066 and 0.602 respectively and none of the relations were significant.

Suitable Catchment Sizes for Runoff Plots

a. Plots Representing Farmers' Field Conditions

In many cases, it is important to collect data on the actual losses of rainfall, as runoff, from farmers' fields. These data show whether such runoff is important and if so, provide the information to design preventative measures. Observations of runoff which do not involve actual measurement are notoriously misleading and anecdotal evidence to estimate runoff amounts should not be used. Runoff channels and other evidence do not provide accurate information on volumes and frequencies and no decisions should be made on the basis of their observation

It is important at the outset of runoff plot experimentation, to define the most appropriate size of plot. This size will depend on several factors, but the most important is that it should be representative of actual field conditions. The use of very small plots has several advantages; many replicates can be built, they are easy and cheap to instrument, and they occupy only a small portion of any research area. It is unlikely, however, that a plot that is only 20 square metres in extent, for example, can be used to represent the runoff regime of a farmer's field. The actual dimensions and shape of the any runoff plot are best determined by the aims of the research agenda, the finance and equipment that are available, the remoteness of the site etc., but it is essential that the following considerations be made:

- The plot should include representative field topography, so that within the plot, the overall land slope of the field should be included. Slopes influence the velocity of runoff and will affect opportunities for it to infiltrate and overwhelm ploughed ridges. Because runoff velocity increases by the square root of slope, small differences in slope between plots will not lead to large differences in runoff velocity or amount. Low overall land slopes greatly increase the storage capacity of ploughed ridges and bunds (see chapter 7 on water harvesting for details), thereby reducing the possibility of runoff.

- Within the plot, the microtopography (the small-scale ups and downs and ploughed ridges and furrows) of the field should be included. This is especially important in flat areas where microtopographical features may have local slopes greatly in excess of the overall land slope and may be very important in inducing runoff. The redistribution of this local runoff (which may constitute net runoff from the field) will be determined by the size, pattern and distribution of microtopography. This can exist as basins and mounds or ridges and channels, the former could be expected to impede runoff, the latter to assist its passage to the field margins.

- Ploughed ridges and furrows will inevitably leave the contour at some point and encourage water movement to low-lying areas. This should be taken into account when plots are being planned and runoff should not be impeded by the artificial boundaries of the plot.

- Another important reason to include representative rnicrotopography is its potential to indicate changes in soil texture and nutrient status. Differences in infiltration rates, water holding capacity, soil depth and soil chemical characteristics may be present, resulting in a local variation of runoff production and crop performance. The inclusion of microtopography within runoff plots will not only influence the physical processes of runoff, but will also allow agronomic sampling procedures to assess more accurately, the effect that these have on crops.

- It is important to note that although in land-levelled fields natural microtopography may not be evident, residual soil variability will still be present and may have an important influence on crop growth. Plots that are used to measure runoff from farmers' fields should cover at least 10% of the total area, more where fields are less than 5 ha in extent. A 30 cm H flume will have an adequate capacity to cope with flows from plots of around 0.5 to 1 hectare. Plot length should exceed 80 m where field-scale runoff is to be defined and plots should be representative of field slope and topographic conditions. They should be ploughed and planted according to the farmer's usual methods. Where similar plots are used to measure runoff from naturally vegetated areas, a representative cover should be included. Very bare plots of 0.5 hectare may be expected to give flows close to the capacity of a 30 cm H flume and a larger instrument may be preferred.

b. Within Field (Small-Scale) Runoff Plots

Plots built to estimate runoff on small-scale water harvesting and tillage schemes are much simpler than those built to represent farmers' field conditions. They are usually smaller in dimension than any microtopography that may be present.

In these instances, it is usually not difficult to place plots to measure runoff on any slope that is desired. Edge effects can be influential and it is important that boundaries do not channel runoff to the collection tank in an unrealistic manner. Rain falling directly into impermeable gutters, drains, etc. should be taken into account.

Runoff will exploit very small elevation differences and sheet flow is quickly converted into channel flow. If the aim of the experimentation is to promote the even redistribution of runoff to the crop rooting zone, this is an important fact to note.

Ploughed ridges and furrows play an important part in influencing runoff in these circumstances and dead furrows may be a consequence of ploughing technique. They can store a considerable amount of runoff (typically about 500 litres or 0.5 m³ per 10 m length) and their location can make a significant difference to runoff measurement, especially for small runoff events.

It should be noted that such small plots may not behave as on the research station if they are transferred and installed as extensive systems on farmers' field, where pronounced microtopography may exist. The importance of placing runoff plots in full knowledge of the effect of microtopography on runoff measurement cannot be overstated.

In the first case (location Figure 6.8) average seasonal percent runoff from the mounds was measured as 29.0 %, while the runoff from the crop plot (marked on Figure 6.8) and which measures 100 m × 40 m, was only 4.5 % on average, over three seasons. Slopes of the microtopography were about 5%, of the large plot about 0.5%.

In the second case (location Figure 6.9), local runoff due to microtopography, from the ridges to the channels, was in excess of 15% whereas average runoff from four, 100 m × 40 m plots located on farmers' fields, but not shown in Figure 6.8, ranged from 1.7% to 4.5% over three seasons. Slopes from ridges to channels ranged from about 3-8%,

large plots slopes were approximately 1%. If, in such cases, the results of runoff measurement from the small plots were extrapolated to estimate net runoff values from the whole field, they would lead to a gross over-estimation.

In practical terms this over-estimation might lead to the supposition that the prevention of runoff was of paramount importance and costly (to the farmer in terms of labour input for reward from increased yields) control measures might be implemented. Where rainfall amounts are regarded as marginal for crop production, these results might also suggest that additional supplementary water should be obtained by water harvesting. The apparent runoff efficiencies of 15 - 29% indicate a high runoff efficiency, and it might be expected that an extra 100 - 125 mm per season could be provided on the basis of a 1:1 crop to water harvesting area ratio. The actual runoff efficiencies of around 2 - 4% for the larger plots show that this is not the case and 10 mm might represent the realistic supplement that would be available for crops (ratio 1:1), unless the harvesting to crop area ratio was very large.

Figure 6.10 shows a typical simple installation for the measurement of runoff from field microtopography.


Figure 6.10: Installation to Measure Runoff from Field Microtopography

c. Natural Catchments

Natural catchments are usually larger than those that are artificially defined for the purposes of runoff measurement. They frequently include areas with different land slopes, soil textures, vegetation and microtopography. In areas with abrupt changes in geology, different densities of stream networks are often exhibited. Natural catchments are, therefore, more difficult to characterise than artificially bounded catchments. For the purposes of study they may have to be divided into subcatchments each with a more homogeneous nature. Runoff may then be measured at locations to include each of these relatively homogeneous areas.

6.4 Field orientation

Field orientation is particularly important with regard to water conservation measures that may be attempted on agricultural land. Fields are often defined according to convenience, exploiting useful land marks such as the position of roads and access, rivers and natural features. They are rarely oriented with runoff losses and methods of runoff prevention in mind. Figure 6.11 shows a typical semi-arid agricultural landscape.

A study of the photograph and the features of drainage shows that most fields are oriented so that one corner is at the highest topographic elevation. Few of the boundaries are parallel or at right angles to the overall land slope and the natural drainage. In practical terms, this means that when a farmer ploughs, he or she will always plough such that ridges and furrows provide channels that encourage runoff. To plough along the contour would necessitate a start in the highest or lowest corner, and ploughing for very short distances. The length of ploughing would gradually increase until the full diagonal width of the field was attained, then the distances would decrease until the farmer eventually reached the opposite corner from where he or she had started. This would be a very difficult and inefficient exercise from the viewpoint of ploughing, but cultivation would be on the contour, disregarding local variations, and would inhibit the natural flow direction of runoff. If contour bunding were to be practiced, similar difficulties would be encountered.

The boundaries set for fields, in areas of agricultural activity where the effects of urbanisation are small, are often those of roads and tracks. The directions of these roads and tracks are not often exactly along the contour. They remove natural vegetation, cross natural drainage systems, redirect runoff and concentrate it into ditches, under culverts and bridges. This leads to the disruption of the natural drainage, the concentration of flow and in many cases, serious problems of soil erosion.

The problems of field orientation are complex. They involve land ownership, the freedom of access and many other social issues, as well as a consideration of the physical environment and the behaviour of drainage. The field studies of most projects will be sited upon land that is already allocated and used for farming, so few opportunities for the implementation of new allocations will exist. However, the influence of field orientation is an important factor to note when field sites are being selected and where the opportunity exists, serious consideration should be given to the siting of new fields with a favourable aspect to natural drainage. The problems of contour cultivation and the effects of local microtopography on such practices are discussed in more detail in chapter 7, Water Harvesting.

6.5 Antecedent soil moisture conditions

Antecedent soil moisture conditions strongly influence the rate at which rainfall infiltrates into the soil and contribute to the processes of runoff production. Soil moisture levels at any time are the result of a combination of several factors, mainly: the time elapsed since the last rainfall; the rainfall amount and intensity; the climatic conditions that have prevailed since rainfall; the type and stage of development of vegetation and soil texture and depth. Soil moisture levels can be highly variable both between and within periods of a particular meteorological activity. A high degree of spatial variability of soil moisture conditions may also be encountered.

Soil moisture levels can be estimated by accounting procedures that balance the infiltration of rainfall against losses by drainage and evapotranspiration. The calculation of evapotranspiration (Et) by different methods is discussed in Chapter 8. A commonly used accounting procedure derives an Antecedent Precipitation Index, by the application of an estimated factor for Et losses on previous rainfall. It is generally assumed that the rate of reduction of the soil moisture reserves is logarithmic, the rate falling as the availability of water decreases. The mechanisms by which antecedent soil moisture effects runoff are highly variable from soil to soil, but the general assumption equates higher proportions of runoff with higher levels of soil moisture. This reflects the behaviour of infiltration rates under increasingly moist conditions. Figure 6.12 illustrates changes in antecedent soil moisture according to rainfall.


Figure 6.12: Change of Antecedent Soil Moisture Levels as shown by the Antecedent Precipitation index

It is important early in a study to determine the precision to which antecedent soil moisture needs to be measured or calculated. In situ measurements can be time-consuming and calculations of Et usually necessitate the collection of a wide range of meteorological information (see Chapter 8). General indicators of soil moisture status may be adequate in some instances, but for use in, for example, regression analysis against event runoff data, estimates of actual values are necessary.

6.6 Other catchment influences

a. Geology

Among the other influences on runoff that can be important is the geological nature of the area under study. It affects runoff in three main ways:

1. Lithology. Particular rock types that are exposed at the surface of the ground can have a profound influence on runoff, but generally large areas of exposed rock are not common. Impermeable rock surfaces such as granite, gneisses, shales etc. can produce very high percentages of surface flow and may be locally important as sources of runoff. In semi-arid areas, and where these rocks are highly fractured, they not only lead to rapid runoff from their impermeable surfaces, but can also provide ground water that prolongs stream flow beyond the normally short period of flash-flooding. Permeable lithologies such as limestone and porous sandstone can limit surface flow to brief periods, only attained when ground water levels are extremely high. Perennial springs may occur where they overlie impermeable layers. According to its permeability, the geology of an area will determine surface drainage density, stream channel length and catchment shape.

2. Soils. The most widespread effect on runoff, of geology, is through the type of soil that it engenders. Granite, sandstones and quartzites produce sandy soils with relatively few nutrients and high rates of infiltration. Shales and basic igneous rocks usually give rise to relatively impermeable clays, though the humidity of the climate will determine the processes of weathering and erosion, and the type of soil that is subsequently formed.

3. Topography. Topography is also the result of geology and climate which determine land form, slopes and local microtopography.

b. Stream Density

Stream density is an index of the concentration of a drainage network within a catchment area. It will not be an important factor when small runoff plots are studied, but is used frequently as an independent variable in regression analysis for natural catchments. It should be noted that larger runoff plots, for instance those greater than about 0.5 hectare especially where they constitute a portion of a large field or a natural catchment, may well exhibit a stream network, though in arid and semi arid regions stream flow will be ephemeral. Such networks exhibit themselves as microtopography and may be influential in determining the runoff efficiency of the catchment, as they intercept sheet flow and channel it to the catchment outlet. The measurement of discharge by sheet flow alone is more difficult to achieve, as it is prone to retention by vegetation, ponding and subsequent infiltration. In origin, stream density is closely related to structural geology, Ethology, slope and climate. In general terms the greater the density of a stream network, the greater the percentage runoff for any given rainfall, because stream channels conduct runoff efficiently they lead to high, sharp peaks and rapid recessions. Relatively complicated systems of stream hierarchy are used to derive a stream density index in hydrological analysis, but they are somewhat beyond the scope of this book. A simple but appropriate index that can be used for regression analysis is length of stream per unit area (e.g. km km-2), though the correlation of such a stream index with other catchment characteristics must be considered before use in regression.

c. Human Factors and Agriculture

Human influences on runoff can be very great and they work at many scales. The wholesale destruction of huge forested areas has led to disastrous flooding and extreme environmental degradation. A list of the most important influences that affect the amount of runoff from the land may include:

Urbanisation
Deforestation
Overgrazing
Dam building
Canalisation
Reduction of flood plains
Draining of marshes and swamps

Arable agriculture appears less harmful, perhaps, but its effect on the hydrological nature of much of he world's surface is profound.

Equipment costs

All costs of locally made equipment are approximate. The costs of raw materials and especially labour are highly variable from country to country, but a good idea of cost magnitude can be gained from the figures quoted below. The costs of manufactured equipment are based on 1993 prices. Shipping, agents' fees and fluctuations in exchange rate cannot be taken into account.

Item

Typical Approximate Cost in $ US

2 - 3 Seat light aircraft

per hour

300 - 500

4 - 5 Seat with global navigation

per hour

400 - 700

(sufficient to cover 2 -3 sites within about 30 km radius)

Wheel point apparatus

1

30 - 60

7.1 Water harvesting

Background

The use of water harvesting for crops and the design and construction of channels, ridges and bunds are discussed in this chapter. Other applications such as the collection of water for domestic and animal watering are widespread and may be more easily achieved, but are peripheral to the coverage of this book. As well as providing supplementary water for an established crop, efficient water harvesting systems may allow an early start to the season in areas where farmers cultivate land away from their main residence and where they are dependent upon a supply of domestic water obtained from rainfall. Similarly, water harvesting can be used to promote early cultivation where farmers plant in soils with high bulk densities and soil wetness promotes easy ploughing and a moisture store for germination.

A convenient classification of water harvesting systems has in the past been regarded as somewhat problematic, a situation that may have more to do with the relative newness and dispersed nature of research into these methods of supplementing water availability, than with any inherent difficulty. This section on water harvesting covers three types of system that are classified on the basis of the scale of water movement and catchment size. As these two factors are the most important influences in determining the design, construction and the operation of the systems, this seems logical. It is inevitable that systems of one classification will grade into another and too much emphasis should not be laid on a strict adherence to the categorisation presented here or elsewhere, since all are to a greater or lesser extent arbitrary. Similarly there is no intent to enter into confusing discussions of water harvesting terminology, most of which is also arbitrary, though it is recognised that a clear and agreed definition of terms would be beneficial.

The importance of water harvesting lies in whether it can be used to improve agriculture or not. The categories used in this chapter are:

Micro systems - very small scale methods of concentrating runoff over very small distances, perhaps less than one metre, but which involve the construction or use of a catchment and which are not methods of encouraging in situ infiltration.

Meso systems - systems that redistribute runoff usually over metres or tens of metres, but which always use runoff gathered within the area of normal field boundaries.

Macro systems - systems that use runoff from outside normal field boundaries.

Water control and harvesting systems that provide supplementary water for crops have been used in widely different areas of the world, perhaps since 2,500 BC or earlier. This guide does not present an analysis or description of these systems because they are unimportant, but because documentation is more complete elsewhere than could be achieved here. These systems were used in the past in the Middle East, Nonrth Africa, India and other areas which are now regarded as under- or undeveloped. In many instances it appears that water harvesting provided a vital economic basis for the existence of the societies that practiced it, though these societies no longer exist in their past form and water harvesting has been largely abandoned. Water harvesting, of several different kinds from large to small scale, is currently practiced in areas of the developed world, such as the USA and Australia, but in forms that are inappropriate for the economic and social conditions that prevail in under-developed countries which, arguably, are in greater need of increased agricultural production. The financial and technological inputs to water harvesting in developed countries are impossible for poor countries to match. This chapter concentrates on methods of harvesting water that do not exclude poorer areas of the world.

7.1.1 Considerations for the Implementation of Water Harvesting

In many respects, successful water harvesting is a very complex mix of climate, soils, technology, social organisation and economic factors. Some of these are listed below, with plus and minus aspects, so that an unquestioning credulity will not be automatically assumed toward this form of farm practice. Under many circumstances successful water harvesting is very difficult to achieve.

Factor

Aspect

Climate:

Encourages water conservation and useful concentration around plants

May only operate effectively during large storms when it is not needed


Replenishes soil moisture reserve after periods of drought

Rainfall is unpredictable and totally uncontrollable

Soils:

Helps reduce erosion and soil loss

Induces erosion by runoff concentration


Increases soil moisture, plant growth

Increases leaching in


already nutrient



organic material and



biological activity



poor soils. It can cause



destruction and



water logging.


Crops and

Encourages crop and other vegetation

Water harvesting catchments work least

Natural vegetation:

growth and thereby provides favourable microclimatic conditions, grazing, fuel .

efficiently when covered by the vegetation that they encourage

Human activity:

Can provide, food, water and money for poor farmers

Restricts other activities, demands labour, fertiliser and time

There is litlle doubt that in some circumstances water harvesting can improve crop performance, but whether this improvement is widely attainable or, in some cases a priority, is another matter. In the general sense, it is useful to recognise that the technology of particular water harvesting systems cannot be applied haphazardly, many depend upon the physical constraints of availability of suitable materials at a precise locality and it is important to recognise that every system, to function properly, must be designed for the field upon which it is to work. An attention to detail is essential.

Rainfall

On the whole, the tropical and sub-tropical semi-arid regions are those expected to benefit most from water harvesting. They have generally low rainfall totals and long periods of drought. The supplementation of water for crops appears to be an obvious advantage.

Long term values of annual average rainfall have been proposed as a pointer to areas that should benefit from water harvesting, but actually such values are of little use. Rainfall in semi-arid areas is highly erratic and notoriously unpredictable. Large coefficients of variation exist for annual rainfall and the intensity-duration relations of individual storms. In many regions, for example Southern Africa, the start of rains is associated with complex weather systems entering from oceanic areas, and in this instance bimodal distributions of rain within the season are determined by the interaction of such systems with the Inter Tropical Convergence Zone. False starts and early closes to the rainfall season are common. In contrast, the start of the rainy season in West Africa is more reliably predictable.

Although relatively long term records of daily rainfall are available in such regions, the variability of rainfall distribution inevitably results in difficulties of prediction when the analysis of data is undertaken. High levels of spatial variability are also a problem. The fact that over many seasons an averaging effect occurs and totals for locations within a region may be similar, is of no comfort to a farmer who experiences such a dramatic shortfall of rain that crops fail. It is important to remember that farmers draw readily from experience and a negative experience with a water harvesting system that fails to work because of an unfortunate period of weather, may kill any interest for good.

The spatial and temporal variability of rainfall make the accurate assessment of water harvesting systems difficult to achieve from the results of short term projects. They make the prediction of extreme rainfall events difficult and demand that structures should be capable of dealing with large variations in the level of runoff. A level of over-design may be preferable, despite the extra cost. These estimates of extreme rainfall are translated to extreme runoff events, but runoff has a highly localised character; between regions, areas, fields and even within fields. The usefulness of any additional water supplied by this runoff will also vary according to regional climatic conditions; for instance whether rain falls in summer or winter.

Topography and Soils

Land form and topography are strong influences on the success of water harvesting. They influence, by interaction with other factors, the proportion of runoff that will occur from a given rainfall. They determine the manner in which overland flow collects, how it is distributed, the size and density of stream channels and thereby runoff velocities and peak flows. Although steep slopes give more runoff, other things being equal, the velocity of runoff increases (and in this sense the opportunity for infiltration decreases) by only the square root of slope. Increases of slope in low slope areas have a proportionally greater effect than similar increases in areas that have high slopes. The effects of microtopographic slopes is very important and was discussed in chapter 6.

Steeper slopes necessitate greater earth working to provide storage, but more water is retained over a smaller area. With structural limitations imposed by the nature of bunds and ridges, the horizontal distance between them is shorter on land with high slopes. Problems of channel erosion and over-topping with the subsequent destruction of bunds and ridges is a greater problem in areas of high slope. A more technical discussion of these matters is given in section 7.2, with recommended slope/area/bund lengths provided in Appendix D1.

Soil Me also exerts influence over water harvesting. Soil textural properties will determine rainfall infiltration and runoff production. Soil textures also control the structural capabilities of bunds (see section 7.2). Soil depth and soil texture will determine the extent of the soil moisture reserve; the amount of water that soils can retain for crop use. There is sufficient evidence that infiltration in very sandy soils is extreme and that any water added simply passes beyond the rooting zone of crops by downward percolation. The soil nutrient status will control crop yield, should the limiting factor of water availability be overcome. Leaching and subsequent nutrient loss, in the commonly poor soils of arid and semi-arid regions, can pose a serious problem that is exacerbated by the addition of harvested water. Waterlogging, in reality a combination of poor gas exchange conditions and nutrient leaching, has been seen to adversely affect crop growth even in medium textured soils. A considerable amount of work has been undertaken to solve the problems of soil crusting in the semi-arid tropics. Crusting acts as an inhibitor of seed germination and as a promoter of runoff. However, soil crusting appears to be as much a function of tillage practice as of inherent soil characteristics. Soils that are tilled to a modest extent, such as those used for subsistence agriculture, are much less likely to suffer than those which are well worked on commercial farms or the research station. In some cases where enhanced runoff is desired, crusting has been regarded as beneficial.

Vegetation

Vegetation cover is an extremely important factor in determining the runoff efficiency of a catchment, though authors attribute very different efficiencies to different covers and vegetation cover thresholds. Vegetation cover slows runoff velocities, encourages infiltration and inhibits soil erosion. On the other hand vegetation intercepts rainfall leading to its subsequent re-evaporation. Vegetation cover, retained on shedding areas to reduce soil erosion can be a serious problem, providing a vigorous seed bank for weeds and promoting the incursion into the crop area of invasive species. Once established, weed species that are difficult to eradicate will incur increased expenditure and difficulty for the farmer.

Agricultural and Social Influences

Although in many respects these two factors cannot be regarded as separate, some agricultural practices may be looked at from the technical viewpoint only. The different water harvesting systems described below are aimed at specific crops. Some are most suitable for tree production, which depending on the locality, may be fodder trees, fuel or fruit trees. Some systems may provide grazing for stock while others are most suited to arable agriculture. Crop types will depend upon climate and tradition, though the traditional crops of semi-arid areas are millet, sorghum and (to a lesser extent) maize.

Improved husbandry is necessary to exploit any advantage of increased water supply. The careful timing of planting is essential to exploit optimum conditions of soil moisture. This care may take different forms in different localities; dry planting to await predictable rains, planting after a certain date or a combination of planting after a certain date when sufficient rain has fallen, in other areas.

Weeding is particularly important at the early stages after germination, to avoid excessive competition. However, the amount and quality of weeding will depend upon the manner in which crops are sown. Weeding is much easier and more effective in row-planted crops, but this practice is by no means universal and in some regions may traditionally be undertaken only by farmers with access to draught power. Thus certain farmers may be at a great advantage compared to others and a socio-economic factor may play an important part in what appears to be a straightforwardly agricultural problem.

Plant population densities, in general, are reduced by the constraints of water harvesting technologies. In dry years this is an advantage, but in wetter years the full potential yield may not be realised and may limit the popularity of water harvesting. The tillering capacities of crop varieties can be adapted to take advantage of improved soil moisture conditions, or setbacks of the main growth due to exceptionally dry periods. Sorghum is especially suited to water harvesting since it can withstand both temporary inundation and drought. Inter or relay-cropping, whereby short maturity crops are introduced to take advantage of exceptionally good conditions are an important bonus that water harvesting systems can provide. Integrated systems, which promote the cropping of annual and perennial species, trees etc. may be suitable, but as yet have not been extensively tested.

Social and economic factors may be the most important constraints to the successful adoption of water harvesting systems, but integration of these systems into population groups may provide the basis of long term acceptance and benefit. In Malawi, soil conservation has a long history of practice and this provides the foundation of water control and harvesting, for many reasons. It has been promoted by an active extension group, but benefits for the farmers have been accrued by their own exertions. Some of the main reasons for the development of control and harvesting systems in Malawi can be summarised as follows:

Land slopes are high in much of the country and many years of trial and error have developed systems of water harvesting that are often individually tailored to a farmer's land out of sheer necessity. Such technologies are essential to conserves soil and water in these circumstances. In many areas the production of food crops for domestic consumption and the local market is crucial in a country of low economic development; the production of cash crops is a possibility while other sources of income are very few. Farmers live on or close to their land, in many areas soils are relatively deep and fertile and have a high water holding capacity. The addition of organic matter does not involve high costs of transportation and mulching with residues is frequently observed. Markets are widespread and transportation distances and costs are low. Produce is of high intrinsic value (maize, fruit, vegetables).

Botswana, with its erratic and marginal rainfall, widespread cultivation at the subsistence level and low rural incomes appears to be a very suitable area for the implementation of water harvesting schemes. In these respects it is a relatively typical sub-Saharan, semi-arid agricultural setting. However, various social and economic circumstances mitigate against any increased input to agriculture. Distance are long and therefore transport costs are very high. There are few markets other than the government purchasing agency in the capital, Gaborone, and the main crop is low value sorghum, with some millet. Soils are generally nutritionally poor and thin. Research has shown that manure spread at less than 10 tonnes per hectare is of little value and its transport is costly. Rainfall is highly erratic.

In Botswana, the traditional organisation of family life necessitates the need for three homes. A winter season village home is occupied while the fields are bare fallowed. At the beginning of the wet season, the male component of the family take the cattle (the traditional source of wealth) to the drier west, once ploughing has been completed. The female members travel to the arable land once it is thought that enough rain has fallen to permit ploughing. They stay there throughout the growing season to farm. The distances between these localities may be hundreds of kilometres.

Such physical and social circumstances do not lend themselves to the development of a more intensive agriculture based on a ready supply of water made available by water harvesting. The risks are high, the inputs great, the rewards are low and the social status of arable agriculture is negligible. Within the fastest growing economy in Africa, the opportunities exist for a farmer to earn more in cash in one month on a building site than in a good year on the land. Money is more attractive than sorghum. Botswana is by no means typical of sub-Saharan Africa, despite its annual average rainfall of about 500 mm, but serves as a good example of the strong influence that socio-economic factors can have on the suitability of water harvesting systems in agriculture.

Examples in the Text

Many texts on water harvesting systems concentrate on their construction and land form aspects and there is a dearth of agronomic and soil moisture data to assess the value, the total or partial success or failure of these systems, beyond the fact that they can actually be built. The underlying expectation that prompts the use of water harvesting is that a lack of available water for crops is the crucial factor that limits yield. The situation is rarely so straightforward. This is particularly true in arid and semi arid regions where labour and financial inputs can only be small, because returns are low. The rainfall regimes of these regions are notoriously unpredictable. Moreover, the availability of water for crops may be severely limited during dry periods within a season that has a more than average rainfall. The rainfall of one season may be too low to grow crops, whereas the next season may be so wet that soils are badly leached and crops are physically destroyed by the surfeit.

For this reason, research into an example system from each category of water harvesting is described at the end of each section. These examples give comprehensive field data on rainfall, runoff, soils, slopes, crop yields and, when available, soil moisture. The number of examples is limited by the availability of comprehensive information. However, this information is provided so that a personal assessment of typical systems can be made, bearing in mind that the successful capture and delivery of water to a designated area is only the first stage in growing crops. Apart from the obvious facts that systems need to work and be economic in construction, they need also to provide an environment within which plants can be easily sown, germinate and develop. They must also be viable from the farmer's point of view, so the difficulties experienced in the operation of these examples are also discussed.

7.1.2 Micro Water Harvesting Techniques

Almost all farmers practice these techniques, each time they plough. They plant crops in furrows surrounded by plough ridges that direct any runoff down the slope of the ridge sides towards the crop. Even farmers who broadcast seed and then plough achieve, effectively, the same result. Farmers who plough on the contour do this most effectively of all.

a. Tied Ridge and Furrow (TRF) System

Tied ridge and furrow systems encourage runoff by increasing the size of the ridges and extending the catchment area adjacent to crops. An attempt is also made to give advantageous degrees of slope to the ridge side. The runoff that is collected within the furrow is then retained by a smaller secondary ridge that is placed at right angles to the primary ridges and "ties" them together. Thus the dual action of encouraging runoff while overcoming its redistribution by local microtopography is achieved.

Construction

The construction of the ridges may be by hand labour or tractor and ridger, depending on the availability and cost of either alternative. The ridges may be constructed with indifference to the contour in low slope areas, but where slopes are greater than 2%, they are constructed as close as possible to the contour. A frequently-used ridge to ridge spacing is 1.5 m, with ridges built to 0.75m or so. The ties are hand-dug with a mattock or similar implement to 30-50 cm height, spaced according to the gradient, though distances of between 5 and 10 m are commonly used. Such microcatchments are extremely effective in retaining all the runoff of storms of 75 mm or so, and dimensions can be changed to suit particular circumstances.

The practical considerations of dimensions are important. When built by tractor, the spacing should be such that the tractor wheels run along the tops of the ridges during subsequent passes, to compact them and thereby improve runoff shedding, while avoiding compaction of the furrow and lower portions of the ridges where the crops will be sown. In most cases a ridger designed for the purpose is used to easily achieve the desired ridge height, rather than, say, a mouldboard plough.

Planting may be done in the furrow, or just above the bottom of the side of the ridges. The latter is frequently recommended because during ridge construction, top soil is removed and planting in the bottom of the furrow may be into subsoil. In addition, planting in the side of the ridge avoids any danger of waterlogging, especially during early stages of development, while allowing the crop root zone to develop adjacent to the area of enhanced soil moisture. However, soil compaction may arise during the ridging process, with attendant difficulties in the early stages of germination.

Modified systems of the basic TRF system have been used. An example is the wide ridge bed which places ridges at 1.5 - 2.0 m apart, with a flat bed between. The tractor wheelings are located at the base, not the top of these ridges. The central portion of the bed is planted with two crop rows. This system exploits the water harvesting and retention aspects of the TRF system to a lesser extent, but exploits the advantage of having the crop planted into top soil, while still benefiting from concentrated runoff at the root zone. In both systems, runoff usually travels very short distances.

Figures 7.1 (a), (b) show the TRF and wide bed systems, respectively.


Figure 7.1: (a) Tied Ridge and Furrow (b) Wide Ridge Bed

It is important to recognise that TRF systems provide a lower planting density than traditional ploughing methods. This is advantageous where water is short and crops must share it, but the natural consequence is a low per unit area compared to many traditional systems, when conditions are good. As TRFs also represent a significant economic input for the farmer, it may be important to grow higher profit crops (maize instead of millet or sorghum, for example), at least in part, to recoup some of this increased expenditure.

Modifications to orthodox bunds have led to the development of the "W" shaped catchment (in section) formed by alternating wide and narrow ridges, the former acting as shedding areas, the latter being used as the planting area. Inter-ridge distances are dependent upon wheel spacing. Planting densities maybe reduced in this, as other systems, but plants are lifted above the area of greatest saturation.

Tied Ridge and Furrow Example:

The following example from Botswana, is cited in detail for several reasons. The most important is access to comprehensive data over two seasons at several sites. The second reason is that Botswana represents some of the most marginal agricultural land in any semi-arid region, with poor soils and a low and erratic rainfall. Agriculture is exclusively subsistence farming. TRF systems that were adopted came from reasonably successful trials in Zimbabwe (see below) and Malawi, where experimentation on vertisols and medium textured soils had shown not only that the system was excellent at retaining water, but also gave increased yields.

Example 1. Botswana:

Location:

SE Botswana, approximately 24° 30 S, 26° 00' E.

Rainfall:

AAR approximately 500 mm, but variable between 200 and 900 mm

Soils:

Loamy sand/ Sandy loam

Slopes:

0.5 - 2.0 %, marked microrelief up to 1.0 m above general field level

Season 1: Rainfall, 692 mm for total season.

Method

2 × 75 m strips were ploughed on.. On one a TRF was installed, on the other flat bed planting was done for comparison. These strips were aligned to cross the marked microtopography. Crop was sorghum (var. Segaolane) in both cases.

Results: Crop and Yields

The TRF crop failed to germinate and was replanted with maize (Kalahari Early Pearl) and thinned to 0.2 m (3300 plants ha-l). Growth / yield monitoring was therefore limited to within the strip, comparing performance of higher areas with and without ties, low areas with and without ties and a small flat bed transect planted at the same time as the TRF maize.


Table 7.1: TRF Maize production (areas + and - ties) and TRF versus Flat Row Planting

Comparisons showed that for high areas with ties, dry matter was greater than areas with no ties. The same was true for cob production, though the crop did not reach maturity.

Plants in TRF were more vigorous overall.

Comparisons with the small flat bed transect maize show that one of the main problems of TRF (and other water conservation systems) is that inherently low crop densities limit yield. Data are provided in Table 7. 1.

Soil Moisture

Tensiometers were emplaced below the TRF system to monitor soil moisture behaviour.
Figures 7.2 (a ) to (d) show soil profile data at the end of the season (April) after rainfalls of 26 mm (24 th) and 16 mm (26 th). The effective rooting depth was 250 cm below ridge and furrow by 21st April, prior to rainfall. The infiltrated wetting front after the rain had reached 170 cm under the furrow 6 days later, whereas under the ridge it had not attained 70 cm by then. Lateral redistribution of wetting enable it to reach 130 cm by 11th May. The driest part of the soil surface was the side of the ridge, which may be regarded as expected, compared to the furrow and the flat ridge top.

This observation has important implications for planting in this position, which is preferred for reasons explained in the text above, and is consistent with difficulties encountered in early establishment of the crop.

Season 2:

Rainfall, Site 1 505 mm for the whole season

Sites 2 and 3 657 mm for the whole season

Method

Ridges were made or re-made at three sites as for the previous year and the comparison between TRF and traditional flat bed planting continued. Both sorghum and maize were planted, but replanting at site 3 was necessary 2 months later because of poor establishment and at site I 10 weeks after original planting because of destruction by a violent storm. A deep-ripping comparison was also carried out within the TRF sorghum treatment to assess the value of increased water percolation.

Results

With the exception of the ripped plots, the sorghum results are not presented because of late planting/poor grain filling. Analyses are compared within sites only.

Tillage was seen to make no significant difference to grain yield at any site although yields were greater for flat bed than TRF at sites I and 2. The reverse was true at site 3.

Total dry matter was significantly less for TRF at sites I and 2, no significance at 3, the trend was TRF flat bed.

Population densities confuse the issue (flat bed has a × 2 difference in row spacing from TRF) at sites 1 and 2, though the better performance of TRF at site 3 seemed due to better growth in individual plants as densities were similar. Growth monitoring indicated that TRF plants suffered a setback to early development, but recovered later. No significant difference found between the ripped and non-ripped rows. Table 7.2 gives components of yield.


Figures 7.2 (a) to (d) Contours of Hydraulic Head Beneath Tied Ridges and Furrows After 42 mm Rain


Table 7.2: Components of Yield for Maize at All Sites, TRF and Flat Bed Planting

Root studies showed that there were no statistically significant differences in root development between treatments, but several trends could be seen. Generally there were more roots at depth in the flat beds than TRF.

TRF roots tended to be concentrated near the soil surface and especially below the furrow, rather than below the plant perhaps due to ridge compaction and the greater availability of water in the furrow area.

Conclusions for Both Seasons

In a total of 18 trials, TRF did not out-perform the traditional flat bed systems on the basis of yield, but it is important to note that during the period of trials, rainfall was greater than the average annual rainfall (AAR) and no "dry" season was experienced.

The system proved very effective at preventing runoff, but difficulties in crop establishment and development were found consistently, despite evidence of improved soil water availability.

Evidence showed clearly that the mechanical and construction aspects of this kind of water harvesting are not difficult to apply, but their successful application does not guarantee improved crop yield.

It is important to bear in mind that there are complex soil/crop/water relations inherent in water harvesting systems and it should not be assumed that they will work, simply because they prevent runoff and concentrate it in the crop root zone.

In general, the system was not popular with farmers who disliked its intensive, high input nature. The results of this work should be compared with that from Zimbabwe, below, which is in marked contrast.

Example 2. Zimbabwe

The trials of TRF in Botswana were stimulated by the relative success of its application by the Agricultural Research Station at Chiredzi in low veldt Zimbabwe. Work in the period 1983-85 had found good results from TRF systems at various sites with high clay content soils. Results were not so good on lighter soils, as for the Botswana example above, and soil infertility and poorer water holding capacities were suspected. The results of later research, for the period 1986-1991, are summarised below.

Methods

Three land forms, TRF at 1.0 and 1.5 m spacing, and 1.0 spacing on the flat and three levels of fertility were used. Sorghum, maize and cotton were the crops.

1 986-87 Fertility levels were zero; 8 t ha-1 manure + 50 kg ha-1 N; 8 t ha-1 manure + 200 kg ha-1 8:14:7 NPK.

1987/88 to 1989/90
Fertility levels were zero; 100 kg ha-1 NPK + 50 kg ha-1 N top dressing; 200 kg ha-1 + 50 kg ha-1 top dressing.
These levels are referred to as low, medium and high.

1990/91
Fertility levels were increased to 4:
zero; 25, 50, and 75 kg ha-1 + 150 kg ha-1 NPK. These levels are referred to as 1, 2, 3 and 4.

Planting was done after at least 15 mm of rain, plants were thinned to 22,000 ha-1 (maize) and 44,000 ha-1 (sorghum).

Results

In all seasons there were significant differences in yield between TRF and flat bed at most sites.

These differences were greater in years of low or poorly distributed rainfall.

The 1.0 m spacing TRF performed better than the 1.5 m, and this was attributed to a greater loss of top soil in 1.5 m ridge construction.

The trend was for increased yield with increased fertility. This trend was stronger for maize than sorghum.

Poor rainfall distributions sometimes reduced the advantage of fertiliser applications.

A summary of results is given in Tables 7.3 and 7.4.


Table 7.3: Effects of System and Fertility 1986-87 and 1987-88, Grain Yield (kg ha-1) of Sorghum


Table 7.4: Effects of System and Fertility, 1988-89, 1989-90 and 1990-91 Grain Yield (kg ha-1), Maize

7.1.3 Meso Water Harvesting Techniques

These systems are constructed and operate within the field and do not receive important amounts of water from outside.

a. Zay

"Zay" are shallow pits dug into the soil, usually about 25 cm in diameter and 10 cm or so deep. Soil fertility and structure are enhanced by placing organic matter, usually grass and/or manure, mixed with earth into the pits. Termite activity commonly reduces the organic material to a state whereby it can be readily exploited by crops and an improvement in infiltration may be achieved due to their burrowing activity. The remaining earth is used to construct a small bund around the pit on the down slope side. They are staggered about 1m apart. This is a revived technique practiced in Burkina Faso to rehabilitate degraded land and is used in conjunction with stone bunds which reduce runoff velocities. Sorghum and millet are the usual crops. The labour input is large.


Figure 7.3: "Zay"

b. Contour Bunds or Ridges

Contour bunds are used to prevent runoff and soil erosion and supplement soil moisture for crops, often, though not exclusively, in high slope areas. Where water retention is of primary importance, ties are used to prevent any loss of water by lateral flow. In cases where erosion control is more important and where increased soil moisture is a bonus, the ridges are constructed with a slight gradient (usually about 0.5%) to allow controlled drainage and render runoff velocities non-erosive. Ridges can be broken to provide drainage and thereby, rudimentary water control.

The size and spacing of the bunds is dependent upon land slope, the practical limitations on bund height and the desired area of control. Construction may be by manual or mechanical means and the soil is excavated up slope of the bund which is under construction. Excessive depth of extraction must be avoided or the loss of top soil sufferes. Water naturally accumulates adjacent to the bund, where the top soil is removed. In areas that suffer from inundation, the crop is planted on the side of the ridges to overcome temporary waterlogging

Bund construction is widespread in Malawi, especially in the Highlands region, where it is used to control runoff from high slopes and reduce soil erosion; individual farmers modify the system according to their own particular needs. It is rarely observed in the Lower Shire Valley area where slopes are generally 1% or less. In the Baringo District of Kenya, contour bunds are not completely tied, but have small bunds that extend up slope, to reduce water loss.

One of the main problems with the implementation of contour bunds is the presence of microtopography which can lead to complex arrangements being necessary. Although a compromise can be reached by increasing the bund height at low lying locations, the natural tendency for runoff to collect in these areas increases the risk of over-topping, though it does allow for a simpler alignment of ridges. Once over-topping occurs, serious erosion can take place and the increased runoff volume imposes a threat to all down slope areas.

Figure 7.4 shows a simple plastic tube level that is easy and cheap to manufacture. It is now extensively used for the laying out of the contour system


Figure 7.4: Plastic Tube Levelling Instrument

c. Hoops (Demi-lunes) and Trapezoidal Bunds

Like Zay, the hoop system could be considered a macro or off-field measure, as external runoff may enter it, but probably most water is captured from local runoff. The harvesting structures are crescent shaped bunds that enclose an arable area, though in Kenya they are used for land rehabilitation and fodder production.

Construction may be undertaken with the dug furrow that provides the bund material excavated downslope, thereby retaining all topsoil within the hoops. In other instances the furrow is dug on the inside of the hoop, thus increasing water storage. Material moved is in the order of 35 - 50 m³ for each hoop, depending on size and slope.


Figure 7.5: Hoops and Trapezoidal Bunds

Usually, the semicircular bunds, about 30 cm high, are between 2 to 10 metres across and may be placed in lines or staggered to manipulate the catchment to crop area ratio. These ratios are usually estimated as between 4:1 to 12:1, depending on hoop density. Adequate distance is left for surplus runoff to pass between the hoops. The open arms of the crescents face up slope. They are reportedly liable to breakage with large runoff events, though this may be avoided by reducing the catchment area.

Trapezoidal bunds, as used in the Turkana District of Kenya are very similar in the manner of operation to hoops and demi-lunes, but are of a larger construction, though scale depends largely upon land slopes. The main bunds may be 60 cm high and 6 m in width, the tapered arm tips 120 m apart and 40 from the main basal bund. The main bund has a freeboard of about 30 - 40 cm, with the enclosed area filled with runoff when sited on low slopes. Construction is estimated to involve the movement of about 400 m³ of soil on a 1% slope.

d. Diamond Shaped Basins

This system is often regarded as a micro catchment system. It consists of diamond shaped microcatchments oriented with one corner up slope. The opposite corner is excavated to form an area of water concentration, where the crop is planted. The crops are usually trees grown for fruit/nuts (Israel) or animal fodder (Kenya) which are situated in the down slope, excavated corner.

In some cases, V shaped catchments are used, thus saving labour and allowing the inflow of water from external sources. Dimensions vary, but sides 5 - 10 m long are usually constructed, depending on local rainfall conditions. Some research has been undertaken in The Negev Desert on optimal shapes and densities of implementation, but the diamond/V system remains most common.


Figure 7.6: Diamond Shaped Catchments

e. Strip Tillage

Strip tillage is used for erosion control, in conjunction with vegetation cover manipulation and grassed bunds in some countries, though the advantages of strip tillage as a water harvesting system have been mooted.

Strip tillage is another contour system and in some respects is similar to the contour bund system described above. With strip tillage crops are planted in strips along the contour, downslope of a shedding area. Labour inputs are reduced, because ridges are not constructed. The natural land slope is used to shed runoff and strips may be made as wide as it thought suitable, no top soil is removed from the planting zone, so that the problems associated with this activity are avoided. Shedding strip to crop ratios of 1:1, 1:2 or 1:4 are typical, with a cropped strip of 5 m. As such, strip tillage represents a medium input system with no need for special equipment, but requires more management than most traditional systems. It is a system that is also open to manipulation on a seasonal basis, whereby fast growing crops can be planted on the shedding areas, should a season provide adequate rainfall.

Strip Tillage Example: Botswana

Location: SE Botswana 24° 30' S, 26° 00' E

Soils: Loamy sands/Sandy Loams

Slope: 2 - 3%

Work was initially undertaken in farmers' fields on a randomised block basis, but it was quickly realised that strip orientation relative to the overall land slope and microtopography could not be achieved precisely enough to regulate surface flow. Two further approaches were made to study the system:

1. Strip catchments on the crop station using bounded plots.
2. A final experiment looked at runoff redistribution within crop areas from shedding strips.

1. Bounded Plots

1989-90 and 1990-91

Method

Runoff:

Runoff was measured from 2 replicates of 5 m, 10 m, 20 m shedding areas. Two 20 m shedding areas with 5 m crop strips were also used to measure runoff through the crop. Multi-slot dividers were used.

For the 1990-9 season, two replicates of 5 m strip with 5 m crop and 10 m strip with crop, were added to the runoff experiment.

Agronomy:

Two replicates of 5 m strip with 5 m crop and 10 m strip with 5 m crop were planted separately from the runoff experiment. and the 20 m strip with 5 m crop were also monitored for crop performance. Figure 7.3 shows the details of plot layout.


Figure 7.7 Layout of Strip Tillage Shedding Catchments

A dead furrow, the natural consequence of ploughing was placed at the top of the crop to act as a sink for runoff and the shedding strips were kept weed-free throughout the season. In both seasons fertiliser (2:3:2) was added at rate of 400 kg ha~, to overcome any spatial differences in soil fertility. Six rows, 0.75 m apart were hand-dug and planted with sorghum beginning 0.5 m from the dead furrow. The central 4 rows were harvested when the crop was mature. Runoff was measured on duplicate plots. Control plots with no runoff were added to the experiment during the 1990-91 season.

Results

Table 7.5 gives the mean runoff values for the two seasons.


Table 7.5: Mean Runoff from Strip Tillage Shedding areas (% of rainfall)

The smallest runoff strips were the most efficient, but volumes were small. On average, the cultivated areas gained the following percentages of seasonal rainfall: 5 m + 3%; 10 m + 3 %; 20 m+ 7%. The effect of individual rainfall/runoff events may be more illustrative of the usefulness of the runoff supplements than seasonal averages. When these are studied, the following points can be noted:

Only the 20 m strips contributed large individual volumes to the crops equal to between 10 and 45 mm or about 2 to 10 days' water lost to evapotranspiration, during the summer growing season.

Almost half the runoff passed through the crop strip and a "cascade" effect could, potentially, result. The construction of bunds down slope of the crops would be advantageous.


Table 7.6: Components of Yield for Final harvest

The even redistribution of runoff to the crop was difficult to achieve.

A large labour input was necessary to keep the strips weed-free and maintain runoff efficiency. Despite differences in received water, analysis of variance showed no significant differences in any component of crop yield, on a planted unit area basis in either season.

Water did not appear to be limiting in either season, though yields seemed to be smaller in the wetter 1990-91 season.

Figures 7.8 (a) to (d) show rainfall against runoff on an individual event basis.


Figures 7.8 (a) to (d) Rainfall versus Runoff on an Event Basis, Strip Tillage 1989-90 and 1990-91

Edge effects were seen in 1989-90 (dry) where the marginal rows gave significantly more yield.

When yields on a field basis were considered, a great difference was seen between the performance of the three crop to strip ratios. Results showed that the 5m strips would need to double yields to equal those of the non-runon control, the 10 m × 3 and the 20 m × 5. This seems to be a serious disadvantage, especially in favourable rainfall years, of many water harvesting systems which inevitably control overall field population densities.

2. Redistribution of Runoff on Crop Areas

The distribution of harvested runoff is an important issue for all but the smallest systems. Runoff flows along channels and collects in low areas, its natural inclination is not to redistribute itself evenly for the benefit of crops. This can lead to problems of uneven crop growth, differential waterlogging and nutrient leaching.

Method

Ten planted, 10 × 5 m experimental plots were installed with different-sized runoff strips placed up slope to provide runoff. The runoff areas were 0, 10, 25, 50 and 100 m², covered with plastic to ensure maximum shedding. Fertiliser (2:3:2) was applied at 400 kg ha-1 and 20 rows of maize were sown 0.5 m apart. Crops were harvested before maturity because of lodging due to infestations of ants in the first season and because of late sowing in the second season.

Rainfall for the two seasons was about 10% below and 10% above average respectively: 1989-90 - 449 mm and 19909 1 - 572 mm.

Results

No significant differences were seen between plots for any season . But a trend of inverse production of dry matter with runoff area was seen. Physical damage and yellowing was seen on areas with the greatest runon. Stunted and yellow plants were seen closest to the runoff areas. Growth distributions supported the hypothesis that too much water damaged plants. Plants recovered noticeably during dry spells.

7.1.4 Macro Water Harvesting Techniques

Off field-water harvesting systems have one great potential advantage over smaller systems: they are capable of exploiting much larger amounts of runoff by utilising much greater catchment to crop area ratios. This does have a concomitant problem: larger flows require more secure control, because their destructive capabilities are considerable.

Bund, channel and dam structures form the major components of such systems which fall into two general categories: runoff collection from broad, flat catchments by the intervention of stone or earth bunds, and the utilisation of water from ephemeral stream channels. In some cases, hybrids may evolve.

a. Runoff Collection from Broad Flat Catchments

The interception of what is essentially sheet overland flow necessitates the construction of large bunds, which concentrate runoff to an area of cultivation. These structures are usually aligned for the most part, on the contour.

b. Stone Contour Bunds and Lines

As currently practiced in Burkina Faso and Niger, stone bunds are discontinuous lines of stones, piled to extend perhaps 10 - 20 m, with a height of about 30 cm and laid in a trench to aid stability. Their action is to reduce runoff velocity by means of their permeability rather than to block the flow of runoff. This reduction in velocity encourages infiltration, reduces erosion and increases the deposition of suspended material. Unlike earth bunds, they allow the passage of water and are not so easily washed away. Given suitable material they are easy to construct and need no special equipment but a simple levelling tube. Farmers may depend on mechanised transportation of the material to site. Their location is usually on low slope areas in cultivated fields or on highly degraded land under rehabilitation. They are often used in conjunction with zay.

Rock dams are a logical development of stone bunds and are used in stream channels, often where soil erosion is a problem. The extremities of the dams extend beyond the channel on to the surrounding land to prevent lateral erosion, giving an overall length of 50 - 300 metres. The gully part of the dam may be well over 1 m in height, but the extensions are usually lower, with a width of 2 - 4 m. The down slope side of the bund is usually built to a 1:2 gradient (vertical:horizontal) while the up slope side is 2:1, giving stability to the structure. The largest stones are used on the outside and the inner portion of the bund is infilled with smaller grade material.

The collection of soil debris up slope of the dam can be considerable and crop yields of 1.9 t ha-1 upslope of the dam have been reported, though the areal extent of such yields is not declared. The main technical considerations in addition to the gradients of construction and the use of larger stones for the outer casing, are laying the foundations in a trench .

c. External Runoff with Enclosed Crop Areas

The collection of runoff from external sources, in a way similar to trapezoidal and hoop systems can be used to exploit large external catchments and to water more extensive areas of crop. Diversionary bunds are built diagonally to the contour, to collect runoff from an up slope area. This runoff is directed by the bunds to the crop area. It is usual to enclose this area with bunds, to prevent the loss of the harvested water. Examples of this general type are the " Caag" and "Gwen" systems of Somalia and the "Teras" of Sudan.

The redistribution of water within the cultivated area may be achieved by ridges and furrows, but in cases where runoff volumes are large, internal bunds and spillways are used. The bunds retain the water up slope until the first section of the crop area is filled, when the spillway freeboard level is attained the runoff passes over to the next section. Eventually the whole of the cultivated area is filled and any surplus water is allowed to drain via spillways positioned in the most down-slope enclosing bund.

Bunds are made of earth, sometimes with stone cores and the spillways and adjacent areas are made with stone or cement blocks to prevent erosion. Catchment and crop areas may be defined to suit the locality and farming practice.

An example of this general system is given below.

Off-field Water Harvesting Example: Botswana

Location: SE Botswana, 24° 30 S, 26° 00 E

Soils: Loamy sands/sandy loams

Slope: 0.5 - 1.5%

Off-field water harvesting involves the collection of runoff from a source external to the field under cultivation, the control and direction of this runoff and its subsequent redistribution over the crop area.

During the season 1986/87, an off-field water harvesting system was installed at Kgapamadi, approximately 15 km north of Gaborone by the International Sorghum and Millet Collaborative Research Program (INTSORMIL). The main aims of the research were to establish whether yield improvements could be obtained by increasing soil moisture availability and to estimate the reliability of receiving agronomically useful runoff.

Method

The water harvesting system consisted of a 0.5 m high earth bund that intercepted runoff from a shallow natural drainage channel and conducted it to a runon area of approximately 0.5 ha, enclosed by 0.3 m earth bunds. Spillways in the bunds determined flooding depth and allowed controlled drainage of excess runon. Bunds were built by a tractor-drawn ridger and manual labour. Soils in the runon area were classified as Chromic Luvisol with an argillic B horizon. Sand, silt and clay contents were 75, 11 and 14%, respectively in the top 40 cm; and 70, 9 and 21% between 40 and 150 cm. The average soil water holding capacity was 12%, by volume and soil depth exceeded 150 cm. Figure 7.9 shows a plan of the water harvesting field, channels and bund structures and immediate surroundings.

Agronomic data were collected for the seasons 1986-7, 1987-88 and 1988-89. Runoff was measured by a different project for the 1988/89, 1989/90 and 1990/91 seasons by 0.90 m H-flume and water level recorder. Daily rainfall and intensity data were also collected. Because the overlap of hydrological and agronomic/soil moisture data is limited to one season, they are treated, to a large extent, separately.

Hydrology

Table 7.7 gives runoff intercepted and measured by the flume, which are the maximum Bows available for water harvesting. However, it is unlikely that these flows could be transferred to the runon area without some small losses due to infiltration in the transmission channel.

The potential amounts of additional soil moisture provided by the runon were very large in most cases. Four, two and six events for each season respectively, were measured as large enough to more than fill the soil profile. The timing of these runoff events is also important; in each season runon volumes were sufficient at the beginning of the season (October and November) for flooding to be practiced by the farmer. Having a deeply wetted soil profile in early season has been shown to be highly advantageous.

On average, rainfall in the range 10-15 mm was sufficient to give some runoff, though it was clear that larger rainfalls (>20 mm) were needed to produce useful runoff amounts. The likelihood of receiving one 20 mm daily rain was calculated from historical rainfall data to be 82% during November-December, 90% during December-January and during grain-fill, 94%.

The precise area of the catchment was difficult to calculate as the smallest scale maps available only provided 15m contour intervals, which indicated an area of about 400 ha. From field observations this was felt to be a gross overestimate. The use of 1:7,000 air photos indicated a catchment between 40 and 100 ha, with a variable contributing area, depending on storm size.

Overall percent runoff was low only a few percent, sometimes less, of storm rainfall.


Figure 7.9: Off Field Water Harvesting Scheme


Table 7.7: Runoff Received by the Water Harvesting Scheme, 1988-89 to 1990-91

Agronomy

The agronomy experiments were undertaken in co-operation with the farmer and cultivation consisted of mouldboard ploughing after harvest and before planting. Sorghum variety "Segaolane", (Sorghum bicolour (L) Moech) was planted 5 Nov. 1986, 7 Dec. 1987 and 2 Nov. 1988 (replanted 3 Jan.), in the runon area. Control plots were planted on the same days in the field, except for 1986, when it was planted 23 days later. All plots were weeded and birdscaring was employed as necessary. At maturity, sorghum plants were harvested, counted and threshed with grain weights adjusted to 12.5% moisture. Fertiliser effects were evaluated each season. During 1986-87 two fertiliser experiments were established after emergence in a low area where crop growth was poor. In separate experiments sorghum and maize (Zea mays, var. "Kalahari Early Pearl") were grown at 0 and 83 kg ha-1, in a 2x2 factorial replicated three times. Single row plots 10 m long were used.

In 1987-88 in the runon area, a combination of 10 t manure ha-1 and 30 kg P ha-1 was applied to a single 700 m² area. Yield was measured on six 20 m² plots in the fertilised area and three 40 m² plots in the unfertilised area. Two 40 m² plots in the field were used as controls. In 1988-89 an experiment was conducted with two treatments, no fertiliser and a compound supplying 45 kg N, 30 kg P and 15 kg K ha-1. Three replicates were made in the runon area, two on the field. The fertiliser was broadcast onto 250 m² plots immediately prior to sowing. Table 7.8 gives details of the size and number of experimental plots.


Table 7.8: Number and Size of Plots, Runon Area and Traditionally Managed Area

Results

Sorghum grain yield was greater in the runon area for the seasons 1986-87 and 1987-88 (Table 7.9). Rainfall was 36% below the long term average in the first season and 11% above in the second, but it was poorly distributed in both seasons. During 1986-87 (Fig. 7.10) only 48 mm fell between 46 and 103 days after planting (22 Dec. to 26 Feb.) and in 1987-88 (Fig 7.11) only 13 mm was received between 16 and 63 days after planting (23 Dec. to 8 Feb.). The 2 Nov. planting of the 1988-89 season was replanted on 3 Jan. and the lack of yield differences probably reflects favourable rainfall (Fig. 7.12) and weather conditions after replanting.


Table 7.9: Sorghum Grain Yield in Runon and Control Plots

In 1987-88 and 1988-89, soil moisture was monitored to 150 cm at 20 cm intervals by neutron probe and because of the effects of runoff redistribution by microtopography, the access tubes were placed 5 m apart at low, middle and high areas, in both control and runon areas. Soil moisture levels measured during 1987-88 illustrate the differences between the field and runon areas (Fig. 7.9). Sorghum used more water in the runon area than in the control during the 44 day intra-seasonal drought and this was associated with greater root depth and mass. In the runon area, for 1986-87, P fertiliser significantly increased yields (P < 0.05), but N did not. In 1987-88, P combined with manure and water harvesting gave the highest yields. No comparable results were obtained for 1988-89 because of stalk borer infestation.


Figure 7.10: Rainfall 1986 -87 Season (F indicates water harvesting flood dates)


Figure 7.11: Rainfall 1987-88 Season


Figure 7.12: Rainfall 1988-89 Season

The off-field water harvesting system proved in general to be successful. Problems of up-slope runon, not associated with the ephemeral stream exploited for water harvesting, occurred. Protective bunds 30 cm high proved inadequate because of microtopography, but despite the high level labour inputs, the farmer was convinced that the system was economically viable.

Feasibility and Practical Considerations for Off-Field Water Harvesting Using the Botswana Example

A wide range of factors must be taken into account in the location, design and operation of off-field water harvesting systems, which will ultimately determine their success or failure. In many respects off-field water harvesting is more complex, but potentially more rewarding than on-field harvesting, because much larger volumes of water are available. The main practical aspects of off-field water harvesting are discussed below, in the light of experience with the scheme described above.

Location and Opportunities

Several important observations were made using a simple non-stereoscopic survey of air photographs of the area around the water harvesting scheme and practical knowledge of how the system worked:

- An aerial survey of 250 km² around Gaborone showed that in SE Botswana, the shallow ephemeral water courses such as that used for the water harvesting system described above were common and that at least in theory, considerable opportunities existed for the adoption of such schemes.

- The number of farmers that could use water harvesting systems was limited to those with fields located in a suitable position, usually low-lying in the landscape. The majority of fields were not located in valley bottoms and for the farmers of these fields, water harvesting of this kind is not an option, though the exploitation of up slope runoff could be possible.

- However, many fields share the potential for the use of runoff and while this is not a problem at present, the unregulated use (and possibly disposal) of runoff may bring different farmers into conflict with one another. The water rights aspect of the interception and use of runoff in Botswana, are at present not clear. Legislation covering water harvesting rights on agricultural land is non-existent.

- In parts of the area with steeper slopes and larger catchments, water courses were seen to be incised. In such circumstances, considerable difficulty could be encountered in obtaining runoff, as the channels were 1-2 m deep and overall land slopes were shallow.

- In many cases diversion bunds or channels would need to be outside the farmer's field if the stream were to be exploited. They could be placed on common land in some instances.

Design Problems of successfully harvesting water can be illustrated by the scheme described above, even though this was a favoured location.

- The area flooded was limited to 0.5 - 1 ha, of a total field area of about 5 ha. Most of the field was at a higher elevation than the channel and runoff could not be directed onto it. To increase the floodable area to 2 or 3 ha, using a channel slope of 0.5%, the take-off point for collected runoff would be 200 m outside the farmer's field. This limitation is one that will be met at many sites.

- The design of the water harvesting system was very simple. The diversion ridge was not difficult to make, though access to mechanical draught power was necessary and the farmer had to acquire the rudimentary skills of bund and channel alignment.

- The exact design of each system would need to be individually prepared.

- Early preparation for the onset of the rains and the maintenance of structures throughout the wet season were necessary.

Operation

In the example above, runoff was introduced into the runon area simply by breaking a section of the bund and allowing water to flow in. When the farmer regarded the runon as adequate, the breach was repaired. No mechanical devices were used (for example sluice gates), but the presence of the farmer was essential.

A study of 15 hydrographs of the larger runoff events for 1988-89 to 1990-91 gives a good idea as to the time of day (or night) that large storms take place and the time-distribution of flow, which together dictate the farmer's opportunity to avail him/herself of the runoff. Figure 7.13 shows example hydrographs with low peak flows (less than 50 1 s-1). Low peak flows are more manageable and present a reduced risk of channel and bund erosion, but more of the runoff must be harvested to provide an adequate supply of water for crops.

Note that the total period of flow is similar in all three cases (about 24 hours) despite the differences in peak flows. Note that the duration of high flows are also similar (about 6 hours for the period when flow is greater than half the maximum peak). These durations are important from the viewpoint of opportunities for the farmer to operate the system.


Figure 7.13: Water Harvesting Scheme: Low Peak Runoff Hydrographs


Figure 7.14: Water harvesting Scheme: High Peak Flow Runoff Hydrographs

Figure 7.14 shows the hydrographs with high peak flows (greater than 2001 s-1). Note that the overall duration of flow is similar to the low peak hydrographs, as are the durations of maximum flow. The recession curves are somewhat steeper.

Of the 15 events, 5 started in the morning, from 03:00 to 10:00 (3 of which occurred between 3 and 4 am.) and 10 started between 15:00 to 23:00 (3 between 9 and 11 pm). This poses some difficulty for farmers who have to be aware of heavy rain, assess that runoff is sufficient, operate the harvesting system and estimate when adequate runon has been collected, in the dark. The development of a more automated system, perhaps with sluice gates is indicated. Spillways were tested, but did not work well.

The average duration of flow was 20 hours, the longest duration 40 hours and the shortest 7.5 hours. These periods appear to give adequate time for the farmer to undertake the necessary action to harvest water. However, because the catchment area is small, the flows peak rapidly (average time 3 hours) and the hydrograph recessions contain only about half of the total flow volume. A farmer who does not act promptly will only have access to a rapidly decreasing supply of runoff. Experience has shown that protection from other sources of up slope runoff is important. The largest runoff event which occurred for the 1990-91 season (from rainfall of 49.6 mm on 25.11.90) caused substantial damage to the field and crop. Runon from up slope sources washed away much of the crop and several contour bunds. Runoff in the channel was sufficient to cause erosion (removal of soil to a depth of about 30 - 40 cm), such that the channel bed was lower than the area usually flooded.

The balance between using harvested water and preventing the damage from runoff, is difficult to achieve. The system of flood prevention that was used, field perimeter bunds about 30 cm high, was not adequate to prevent runon when heavy rain occurred. Runoff tended to concentrate in low microtopographical areas. However, the farmer was convinced of the value of water harvesting. Field observations showed that an increased water supply to crops can cause yellowing and poor growth. The reason for this appeared to be the leaching of nutrients from relatively infertile soils. The agronomy experiments showed greatly increased yield with manure and fertiliser applications. Traditional applications of fertiliser in Botswana are very low, indeed many farmers do not use them. When water availability is not a limiting factor, soil fertility can be.

It is advantageous that soils in low-lying areas where water harvesting can be practiced, tend to be heavier than usual. Soils must be deep to retain the water as soil moisture. Sandy textured soils not only have problems of poor nutrient status, but also allow the deep drainage of runon, with little benefit for the crop. Successful crop growth early in the season can pose difficulties. Crops can attract pests simply by being the only crops in the area. Successful farming will demand increased management, labour and money to protect crops from stalk borer, aphids and birds.

Weed growth is enhanced by favourable growing conditions and more time and labour will be needed to control it. Unfortunately, row planting is not commonly practiced in Botswana, broadcast crops are more usual. Row planting, when practiced competently, gives better controlled planting densities, more even crop stands and facilitates weeding. Access to planters, draught power and the development of row planting skills would greatly enhance the farmer's ability to exploit harvested water and increase crop yield.

Conclusions

The implementation of water harvesting techniques for increased and sustained crop production on a national scale will need a range of favourable preconditions for success. They are listed below, in an approximate order of priority:

1. Farmer enthusiasm and commitment.

2. A suitable socio-economic background, especially a profitable market (be that increased food security or an increased cash income).

3. A co-operative communal framework.

4. A favourable soils environment.

5. Access to basic draught power and operating skills.

6. Simple but efficient agrohydrological designs

7. Access to fertilisers and pest control

8. Competent and effective research and extension planning.

7.2 The design of bunds, channels and other field structures

The success of water harvesting schemes and the productive marriage of hydrology and agriculture depend on the identification of suitable systems for the regional or national social-economic-farming environment and the correct application of known engineering principles in the field. In this section aspects of the latter are covered.

Badly-designed and implemented systems cause more problems than they solve, interfering with natural drainage, promoting soil erosion and causing the destruction of crops. On land that has been developed already, it is useful to assume that all structures; roads, drains and field boundaries have been laid out with a total disregard for topography and drainage, because this will usually be the case. They should also be regarded as moveable. Catchments are natural features that are constituted of smaller subcatchments; fields are some of their components. In planning water harvesting and water conservation over large areas it essential to recognize the varying scales of catchment areas and their hierarchy. Field layouts aim to create artificial subdivisions of natural catchments into smaller units. A number of points should be considered:

- All layouts should effectively reduce erosion.
- Any redirection and discharge of water should not cause erosion elsewhere.
- Concentrations of water should be kept as small and travel as slowly as possible.
- As far as possible, water should be allowed to follow its natural route.
- Access roads and tracks should be planned as an integral part of any large layout.
- Structures should interfere as little as possible with natural drainage and farm operation.
- Future development should be borne in mind.

In general, it is sensible to first describe agricultural constructions at the small scale, adjacent to the crops that are being grown. Large structures, which are planned to deal with large, exceptional flows, can be described thereafter. There are two main types of structure which operate in different ways.

Contour Structures: are aligned parallel to a master ridge on the contour. They are designed to divide the catchment into smaller subcatchments and maximize infiltration. They do not aid in the discharge of runoff and may be tied to ensure that this control is effective. The greatest danger to these structures is that they fill and over-top.

Graded Structures: may be ridges or drains that are aligned at a slight slope and help in the discharge of runoff at low, non-erosive velocities. They prevent runoff from taking the shortest and fastest route downslope. Generally they are used in association with natural waterways.

7.2.1 Channels and Waterways

a. General Design Considerations

The design of channels that conduct water at non-erosive velocities is a common feature of agrohydrological practice and involves the basic channel hydraulic formula Q Va, discussed in the section on stream gauging. Manning's formula is commonly used determine channel design and several factors need to be considered:

Channel Size: larger channels carry more water than small channels, on the same slope.

Channel Shape: channels of the same cross-sectional area, but of different shapes will carry different amounts of water. Friction reduces velocity, so designs usually aim at reducing frictional resistance. The unit used to measure the effect of shape is the Hydraulic Radius of the channel (R):

R = a/w where (7.1)

R is the hydraulic radius
a is the cross-sectional area of flow and
w is the wetted perimeter of the channel

Generally, the smaller the value of R, the lower the velocity of flow. Channel Gradient: as the bed gradient increases, so does velocity. Channel Roughness: channel roughness is a factor that determines frictional resistance.

Manning's open channel formula is:

v = R0.667s0.5/n where (7.2)

s is the slope of the charnel (m m-1)
n is the roughness coefficient (dimensionless)
R is the hydraulic radius (m) and v is velocity of flow (m s-1)

Design Velocity

The design velocity will depend upon the erodibility of the channel lining. Channels may be lined with earth, vegetation or artificial materials. Generally, vegetation, especially grass cover, is recommended for waterways to restrict velocity and to prevent erosion. High retardance species are recommended for use wherever they are available. However, it is difficult to compare the effect of different vegetation types. Vegetation varies from region to region and although much work has been done in countries such as the USA on retardance classifications according to plant species, these plant species may not be present in other areas. Grass and vegetation types may be divided into the following groups or classes of retardance, as presented in Table 7. 10.


Table 7.10: Retardance Classes for Vegetation

Note: condition and vegetation height are important factors in influencing retardance, as can be seen from Table 7.10 and although "good" conditions are tabulated, in real life the extents and conditions of cover are often "poor". In this case a value of Manning's 'n' = 0.040 is commonly used for vegetated channels.

It is also important to note that flow depth has a strong influence on retardance. With medium-height grass vegetation, Low flow (depth of water 10 cm) values of 'n' will be in the approximate range 0.20 to 0.50 whereas Intermediate flow (depth sufficient to just submerge vegetation) values will be around 0.30. However, when the vegetation is completely submerged, values of 'n' drop rapidly to the range 0.030 to 0.040. Bare channels have permissible velocities according to soil texture, though the encroachment of vegetation is common. Table 7.11 gives examples of permissible velocities for unvegetated channels.


Table 7.11: Permissible Velocities for Earth Channels

Permissible velocities for soil types with medium and very good grass cover are given in Table 7.12


Table 7.12: Permissible Velocities ( in m s-1) for Channels with Grass Cover


Table 7.13: Approximate Values of Manning's 'n' for Various Channel Conditions and Materials

Calculations of flow velocity from Manning's formula necessitate the estimation of the roughness coefficient 'n'. Table 7. 13 gives values of 'n' for various channel conditions and the various artificial materials that may be used to line water ways, as well as values of 'n' for dug earth channels: The calculations of velocity can be made from formula 7.1, or nomograph, Figure 7.15, may be used to read off velocity values.


Figure 7.15: Nomograph for use with Manning's Equation

Designs should be made to account for the lowest level of retardance that is likely to be encountered in the rainfall season, according to changes in vegetation growth.

Design Section

Most commonly, channel cross-sections are parabolic or trapezoidal. Triangular sections should be avoided unless they are on very low slopes (< 2 %) and are well grassed, as they concentrate flow and may cause erosion.

Over time, natural channels tend to a parabolic cross-section and trapezoidal channels, by the natural processes of sedimentation' tend to this form also. As a result, their capacity will decrease. It is useful to remember that if tractors and similar equipment are to cross the channel, then the slope of the sides should not be greater than 1:4, to allow access. Figures 7.16 and 7.17 give the details of hydraulic radii (R) and cross sectional area of Trapezoidal and Parabolic channels, according to their dimensions.


Figure 7.16: Trapezoidal Channels (Metric dimensions)


Figure 7.17: Parabolic Channels (Imperial dimensions)

Channel Slope

Flow velocities increase as slopes become steeper, and the danger of scouring and destructive erosion becomes more likely. It is important to recognize that to reduce erosion on steeper slopes, channels can be made wider and shallower to spread the flow and thereby increase the effective retardance of the lining. On shallow slopes channels can be made deeper and narrower with less risk, and narrow channels have the advantage of occupying less land.

b. Methodology of Channel Design

The procedure for determining the design of channels at any location point is as follows:

- Determine the maximum discharge from the procedures in Chapter 2. Remember that the important factor is peak design flow .

- Estimate the channel roughness ( 'n' ) from the details given in Tables 7.10 and 7.13 above. Remember that values of 0.0025 and 0.04 are usually used for earth channels and poor cover grassed channels, respectively.

- Calculate the actual gradient in m m-1 or set a design gradient that will be used in construction.

- Select from Tables 7.11 or 7.12 above, the maximum permissible velocity for the channel design.

- Find the hydraulic radius from Manning's formula (7.2) or from the nomograph, Figure 7. 15.

- Calculate the cross-sectional area for the maximum estimated or design peak flow / velocity

- Using the cross-sectional area and hydraulic radius, find the top width and depth.

- Add the freeboard.

- Assess whether the channel design dimensions are acceptable in the proposed location.

Gradient, channel roughness and velocity are open to variation if the dimensions are unsuitable. Various components of Manning's formula can be found according to which are known and which are not.

Controlled and Uncontrolled Gradients Two sets of circumstances will occur with regard to gradients:

1. Where gradients can be controlled, i.e. selected more or less at will, the velocity, lining and shape dimensions can be used to determine the appropriate gradient.

2. Where gradients are uncontrolled, for example natural streams following a course at right angles to the contours and which it is wise to utilise, the gradient is pre-determined and therefore termed "uncontrolled". In these circumstances it is necessary to determine the velocity according to the channel lining that is present. If time and resources allow, the type of lining can be imposed to select a preferred velocity. Velocity is used with peak flow to determine cross-sectional area, the shape and dimensions of the channel.

It is useful to remember that increases in the permissible velocity of a channel can be achieved by lining with vegetation, increasing the vegetation cover or using artificial lining materials. Grass cover for channels has the advantage that it often already exists in natural waterways. In this case it is best to calculate channel width and estimated peak flow with and without cover and if possible, to extend and improve this cover. Grass is cheaper to install and maintain than an artificial surface.

Design for Catchments

To aid correct design, catchments should be divided into subcatchments. Where the outflow of each subcatchment joins the main stream, the peak flows (known from previous data or calculated from the methods described in chapter 2) should be added together to determine the ever-increasing flow volume. Estimates of channel design should then be carried out at each point. The lowest value of permissible velocity for each section of channel should be taken as the greatest permitted value for that section.

Select and specify the points at which along the channel, the cross-section is to be designed, taking into account the subcatchment dispositions and:

- For all points determine the maximum runoff for these points.
- Select and specify the lining.
- Determine the most erodable soil type.
- Measure the slope of the channel bed segments that include these points.
- Determine the permissible velocity for the slope/roughness/vegetation.
- Determine the channel dimensions according to velocity, slope and runoff.

Any increase in dimensions, above and beyond the increase due to drainage from the increased area of catchment (for example a drop in velocity and therefore capacity), should be allowed for.

Example Design

An example pro-forma sheet for the design of channels is given below:


Figure 7.18: Example Channel Construction Pro-forma Sheet

Establishment and Maintenance of Vegetation

The value of establishing good conditions of growth in vegetated waterways should be recognised and undertaken where it will mean the effective control of erosive velocities. Fertility should be improved with available nutrients and any seed mixtures should include quick-growing annuals as well as hardy perennials for permanent protection. Any material that can be used to stabilise the soil while plants grow ( such as mulches), will be of value.

Channels are depressions frequently filled with water and as such often retain good vegetation during the early dry season. The use of waterways for excessive grazing and stock routes, as well as tracks should be severely discouraged. Care should be exercised when equipment crosses waterways and runoff discharge from terraces and bunds can do serious damage if not properly sited and controlled.

Vegetation should be mown or carefully grazed to stimulate root growth. Repairs should be made when necessary and velocities controlled to allow the addition of nutrients by sedimentation without the smothering of growth. Revegetation is an important and useful process.

7.2.2 Storm Drains

Storm drains form an important part of the overall control of water in a farm system. They may be used to prevent damaging inundations of runon or may safely relocate runoff from large rain storms, by controlling gradients and velocities to prevent erosive flows.

Storm drains are usually lined with unprotected soil, thus reducing the permissible velocities of water. Their controlled (i.e. selected) gradients allow flexibility in design. In some cases, where low slope arable land meets a steep upslope area, very steep channel gradients may be necessary in the alignment.

The general method for the design of storm drains is as follows:

- Design either for the same dimensions throughout the channel length, or in sections and subcatchments and increasing flow as discussed earlier. The latter choice may take several iterations of design before the most appropriate is found and will depend to a large extent on locality and a balance of cost.

- Estimate flow for the design storm, for example a 10 year return; calculate the flow from the methods in chapter 2.

- Determine the maximum permissible velocity for the most erodable soil in the channel (or channel section).

- Choose the most suitable gradient for the location and purpose of the drain.

- Calculate the design depth; construction depth (design depth + a freeboard of minimum 50 cm), design width and construction width (design width + width of freeboard).

Note that each storm drain should be designed for individual circumstances, but if drains merge, then the section by section approach to design must be followed.

Example Design An example pro-forma sheet for the design of storm drains is given below (Figure 7.19).

7.2.3 Bunds

Bunds are ridges built within a field to allow, and control, the flow of draining water. They are placed with shallow gradients just off the contour and their locations need to be surveyed. Usually the vertical distance between bunds will be kept constant though the horizontal distance between them will vary with the slope of the land. In some special and rare cases of uniform slope, parallel bunds ma,, be used. The vertical distance between bunds and the within-field spacing (which determine the amount of runoff to be controlled), will be calculated according to slope and soil texture, and will be determined before the bunds are designed.

Bunds can be:

a. Narrow-based or Ridge bunds - which are formed by hand and are narrower in both channel and ridge than

b. Broad-based bunds - which are made by whatever machinery (animal or engine powered) is available.

In form the two types have the basic size relations:

Ridge bunds

bank width (b) = 0.75 freeboard (f) = 45 cm

Broad-based freeboard

bank width (b) = 0.95 t(f) = 30 cm

where t = the channel width in metres.


Figure


Figure 7.19: Example Pro-forma Sheet for Storm Drain Construction

The relative merits of narrow and broad-based bunds are listed below in Table 7.14


Table 7.14: Advantages and Disadvantages of Bund Types

Appendix D 1 gives diagrams that can be used to calculate bund lengths for different catchment areas and slopes.

General Method of Design

In many respects, this procedure is similar to that for earth channels, though in general bund designs cannot vary so widely as those of channels.

- It is assumed that all bunds in a field catchment will be of the same design as the largest bund, which will have the largest catchment area.

- Use the longest bund and the average distance from its neighbour, to calculate the catchment area.

- Calculate the runoff peak from this area for (as a realistic example) the I in 10 year storm.

- Determine the lightest and most erodable soil type and select the maximum permissible velocity.

- Use the surveyed, design gradient of the field layout.

- Using the table in Appendix D 2 with the appropriate velocity, gradient and capacity select the suitable width and depth of bund channel.

- If machinery is expected to cross the bund a wide, shallow broad-based design will be needed.

- Add the freeboard.

Example Design

An example pro-forma sheet for the design of bunds on a controlled gradient is given below:


Figure


Figure 7.20: Example Pro forma for Bund Construction

Standard Gradients and Standard Bund Design

Sometimes, standard gradients are used for field layouts. The maximum permissible velocity is found from the soil texture, the gradient is known and the design is calculated to carry maximum runoff. In other cases, it is more convenient to specify a standard bund design for all situations, this is especially so when only standard farm machinery is available. A standard bund design is shown below in Figure 7.21


Figure 7.21: Example of Standard Bund Design

In Figure 7.21 the bund is capable of carrying 0.09 m³ s-1 at a gradient of 1:1000 or 0.25 m³ s-1 at a gradient of 1:200. If runoff is expected to exceed this capacity, the gradient may be made steeper, to increase the velocity to that which is the maximum permissible; or the vertical intervals between the bund may be decreased to reduce the catchment area (and thereby the volume and peak flow of runoff). Alternatively, more waterways can be interpolated, thus reducing bund length.

Parallel Bunds

In some cases parallel bunds can reduce management inputs, but there are difficulties and in general parallel bunds are not recommended. Slopes must be shallow and free from microtopography; inter-bund distances must be reduced and the suitability of each bund must be assessed individually in the field. Velocities must be restricted and on the whole the process is time-consuming and complex.

7.2.4 Roads

Roads fall into two main categories:

Crest roads:

In this case, rainwater sheds to both sides of the camber of the roadbed and will travel away from either side in response, thus ensuring that the road does not lie wet.

- The minimum width of a cambered road should be 1.5 m either side of the centre line (total 3 m).

- Shallow V shaped side drains should be planned and the material used to add to the construction of the road.

- Mitre drains should lead off from the side drains at frequent intervals and especially at low points. The gradient used should be 1 in 50, with the head turned upslope to intercept any road side drainage.

- Wheel ruts may prevent lateral drainage from the road where it is on a gradient. Gentle depressions across the road at locations of the mitre drains.

Cross-Slope Roads:

Where possible these roads should be integrated with the bund layout of land, by selecting particular broad based bunds and developing them as road lines. Such roads are developed to the extent to allow one-way traffic.

In areas without bund layouts, roads should be designed in the same way, but with sufficient up slope channel capacity to ensure that over topping does not occur.

7.3 Surveys, marking out in the field and construction

The detailed methodology of surveying and the use of survey equipment is a large area of study and it is assumed for the purposes of this section that a rudimentary knowledge of maps and levelling equipment has already been acquired. If not, it is recommended that project staff consult texts that deal with surveying to familiarize themselves with basic procedures.

Marking Out

Marking out in the field should be as accurate as possible, to ensure that the designs operate as planned. Survey equipment is used to mark out from the large to small scale, in a sequence such as the one below.

Water ways: the boundaries of fields and catchments
Crests
Storm drains
Bunds: the boundaries of bund catchments
Marker and master rows: the guides for row and micro catchment construction
Other required features

When the exact positioning of these features is known, the dimensions of bunds, water ways, etc. can then be pegged.

Equipment
Dumpy/quickset level and recording book
Levelling staff
Ranging rods
Chains and arrows
Compass and record book
20 m + strong string
Pegs

Conventional symbols

Centre of crest

First/ last peg of graded line

Intermediate peg

red

roads

blue

water ways

yellow

survey stations

red + yellow

bunds of storm drains

7.3.1 Methods of Marking Out

Field catchments

The natural topography of an area is the first to receive consideration for the survey of field catchments, because natural features define the field catchments into which all other subsequent catchments are integrated.

- Main crests then subsidiaries are marked out.

- Natural waterways are marked, first main channels then subsidiaries.

- Where field and stream catchments coincide, the bund lengths should be acceptable, where this is not the case and bunds are too long, the catchments must be subdivided by roads and artificial water ways to form smaller field catchments.

- Where suitable condition exist, details should be transferred on to the ground from aerial photographs (use transparent film) upon which the planned layout is set, otherwise the area must be levelled. In many cases an intermediate situation will exist.

Crests

Crests are the lines of maximum elevation and are the next features to be marked. Access roads to undeveloped areas should follow these crests as much as possible.

- Crests are marked on the plan layout (use transparent film) in increments along the approximate alignment, noting any levelled points.

- Along the approximate crest line on the ground, level at right angles at 10 m intervals to determine the true crest and peg.

- Proceed forward 50 m and repeat with guide pegs between levelled positions as necessary.

- An office plot is drawn, smoothing the outline and the final line is transferred to the ground and is best cut, to overcome problems of disappearing pegs.

Drainage lines

The same method is followed, to establish the lowest points between crests. Often drainage is obvious on photographs and the ground.

- Peg the centre line of obvious channels by eye.

- Level if the channel is indistinct and peg the centre line.

- If the starting point of the channel is clear, swing the levelling staff down stream on the string, the highest reading gives the lowest point.

- If the channel is not obvious from the photograph use this method so long as a start point is available.

- If not, a grid survey must be undertaken and the channel defined.

- Natural water ways may have to be extended up slope to ensure that they intercept drainage from all the proposed bunded areas.

Subdivision of catchments with roads and waterways

Decide from field, map and/or photographs, where the longest contour line lies, from crest to channel.

- Calculate the maximum desired bund length from bund design (length should not exceed 400 m on light soils or 500m on heavier soils) and if the length is excessive, divide it by 3, to allow an alternating arrangement of crest and water ways.

- Locate new crests and water ways at the third intervals (see Figure 7.22 below).

- Extend intermediate roads and artificial waterways to meet the natural features. There must always be a water way between two crest roads.

- Water way widths are then marked according to design specifications, with pegs.

- Where designs join, a gradual, smooth transition must be planned.


Figure 7.22: Subdivision of Excessive Bund Lengths with Roads and Waterways

Bund Catchments

Once catchment and field areas have been defined, they are subdivided into smaller artificial areas by graded bunds.

The design of bunds has been discussed and guidelines to appropriate bund lengths and bunded areas are given in Appendix D1. Bunds can be set out according to vertical height or horizontal distance between them. In the case of maximum permissible vertical height, this relates to; slope and the danger of increased velocity and soil type and the associated danger of erosion. The steeper the slope and/or more erodable the soil, the closer bunds are set together .

The general formula for the determination of the Vertical Interval, the maximum permissible height between bunds, is:

VI = S + f/ 6.5 where (7.3)

VI is in metres
S = Slope in %
f is a factor related to soil and bund type and for the following different conditions has a value of:

Sand, Loamy sand, sandy loam

f = 4

Sandy clay loam, Clay loam, Sandy clay

f = 5

Clay, Heavy clay

f = 6

Subtract (-) 1 from 'f' for fine "rained sands, limited permeability above 1m, soil crusting, row crops steeper than 3%. But the minimum value of f = 3.

Add (+) 1 to 'f' for well drained soils, tillage that encourages infiltration and reduces surface detention. But the maximum value of f= 7.

For Narrow-based bunds controlling peak flows of up to 0.23 m s-1, f is 3 or 4. For Broad-based bunds f can be from 3 to 7.

VI increases with increasing slope, but decreases with the erodibility of soils. Where high annual rainfall and/or poor agricultural practice are encountered, VI should be reduced to 0.8 of the calculated value, and to 0.6 of the calculated value if a parallel bund system is used.

The general formula for the Horizontal Distance is

HD = VI × 100/% Slope (7.4)

For all practical purposes the horizontal distance as measured on the field (i.e. up or down the slope) and the actual horizontal distance can be taken as equal. HD is not usually used for spacing the bunds, but is needed to calculate the catchment area up slope of the bunds, and subsequently for the calculation of runoff and for channel design purposes.

Percent slope:

The measure of % slope to determine VI can be taken as the average slope of the bund catchment found by continuous measurement while pegging out. This accounts well for variations of slope if these are frequent on a field and leads to bunds varying continuously in HD. Alternatively, on a relatively uniform field the minimum slope can be used to find the VI. This means that on parts of the field with higher slopes, the density of bunds is actually greater than is needed for conservation, but time is saved in pegging and measuring.

Field Layouts of Bunds

The critical factor is to ensure that velocities of drainage do not cause erosion, and the following must often be considered:

1. The best line for a bund must be defined so that obstructions (e.g. termite hills) do not interfere.

2. Where a pre-determined gradient (e.g. 1:250) has been selected, calculations of design must ensure that velocities do not exceed permitted velocities for the soil type.

3. If a standard bund shape has been selected, the gradient alone can be used to determine safe volumes and velocities for the soil.

4. A particular depth and gradient can be selected at the maximum velocity of the soil type, and channel width is then adjusted to account for expected flow.

Generally, gradients for within-field bunds vary between 1:250 and 1:1000. Obviously the lower gradients are more suitable for lighter soils. Catchment areas do not usually exceed 1.5 ha on sandy soils and 3.0 ha on clay soils and channels are not usually more than 3 m wide and 0.30 m deep.

Pegging Bund Lines

The pegging of bund lines is outlined below, even though such field survey practices are generally beyond the scope of this book. It is included here for two main reasons. First, it is likely to be encountered very frequently and second, it provides a good example of the general practice without entering into the details of field surveying. The method is as follows:

- Peg from the highest point so that if pegging is interrupted, runoff from upslope will do no damage.

- Inter-peg distances are 15 m on uniform land, 7.5 m on uneven land, so use string of these lengths.

- Locate the start point by measuring the average slope from the high point (or previous bund line) then calculate the VI to the next start point.

- This is recorded and added to the high point (or previous bund line) reading to give the new desired reading.

- Move the staff down the crest of field edge until the desired reading is found.

- This position is then pegged as the new start point.

- If the HD is pre-determined then it is simple to measure or convert to VI for the particular slope.

- Check the slope at various points below each bund line and remember that the slope refers to the catchment that will drain to the next bund down.

- Peg at desired or standard gradients.

Remember to check the grade and direction of flow before construction.

- Increase the grade by 0.6% over the first 15 m near the crest and waterway to help intercept any roadside or channel runoff.

- Normally peg from crest to waterway except for storm drains (the reverse procedure), where bunds must end opposite at waterways for cross access and around obstructions.

- Draw pegging diagram with grades, lengths etc.

Obstructions

In some areas obstructions will be found. In such cases the bunds should be placed at least 3 m away from termite hills, for instance, to avoid tight kinks in the line. It may be necessary to try a different grade to avoid the obstruction, pegging back from the 3 m distant location until the original peg line is met, but care should be taken. In some cases, excavation of the obstruction will be possible. It must be stressed that termite hills are very common in some areas and must be incorporated carefully to avoid bund failure.

When the lines are complete, kinks can be smoothed by moving low pegs uphill (NOT high pegs downslope). At least two consecutive pegs must remain in place undisturbed between moved pegs.

As soon as possible, the pegging lines should be ground-marked by hand or shallow disc cut. Any soil should be thrown downhill, so not to interfere with construction.

Contour Ridges Ploughed ridges and furrows are commonly used to prevent runoff and improve infiltration. Marker rows are first set out:

- Using a level, mark contour lines every 20 -30 m down the slope.

- Peg at 15 m or less

- Do not peg across drainage lines.

Master Rows and Microcatchments

Master rows are used to ensure that ridges have gradients no less than that of the main bund that determines their location.

Alternatively, they can be used to restrict runoff when cross-tied. A master row is pegged out between bunds either parallel to the upper bund where bunds converge or parallel to the lower bund where they diverge. Figure 7.23 shows how master rows are positioned in relation to bunds. Note that during pegging the string used to determine pegging positions is kept at right angles to both bunds.

Where a master row runs into a lower bund, move at right angles to the master row and start a new row at the upper bund. Ground mark the row. Microcatchments are made by cross-tying at the construction stage, to no more than two thirds of the ridge height. When pegging is complete, it is essential to draw a field diagram. This allows the farmer to assess accurately the area of land under cultivation, it provides construction details of waterways etc. and shows how much land is lost to obstructions, roads, bunds and other non-cultivated areas.


Figure 7.23: Pegging a Master Row

7.3.2 Construction

It is essential that the construction of water conservation structures proceeds in the correct sequence. In the case of simple water harvesting structures, such as Zay and Demi-lunes, construction is straightforward, but it is becoming increasingly recognised that to be successful, water harvesting schemes must be viewed as features integral to the landscape.

The overall control of runoff and its safe redistribution and disposal are also critical in preventing destructive water movement. This will quickly become self-evident in areas where extensive water control by conservation structures is attempted.

Waterways, Drains and Bunds

These are constructed according to permissible velocities and unless the design grass cover has been achieved, erosion will take place when runoff occurs. For this reason waterways must be constructed one or two seasons before other works. Careful planning is needed to integrate this time scale into project activities.

Subsoil exposure can be a serious problem. Grass cover will not be encouraged and every effort must be made to improve condition for vegetation growth. This can be done by spreading back top soil, including mulches, manure or fertiliser. Where possible, the creeping species of the waterway are best replaced by grasses with a bunching habit at the edge to restrict spreading onto the field. Water when necessary, if possible.

Waterways should never be used as roads or serious erosion will take place. Grazing should be controlled to encourage spreading growth.

The danger of obstructing the drainage from bunds into a waterway must be recognised, especially where the sides of a waterway are used by vehicles. This will lead to dangerous over-topping of the bunds and render the waterway useless, and must be avoided at all times.

Storm drains are constructed after waterways have been stabilised with vegetation and before bunds have been made.

Bunds are only constructed after waterways and storm drains, and construction starts at the top of the field, working downwards. If they are not, the drainage they cause worsens a difficult situation. It may be very important to explain this to farmers, who often see bunds as the total solution to water control.

Costs and Equipment

The equipment with which construction is undertaken will depend upon the locality, the economic status of the farmers and the inputs from the project and other external sources. The most appropriate equipment must be determined carefully according to local circumstances and the ability of farmers to sustain the inputs once project support has been removed.

Frequently, the costs involved in the construction of water harvesting and conservation systems are difficult to quantify. The data on costs are very approximate. Often labour costs are not accurately accounted and the cost of long term maintenance is not considered. Moreover, both costs and benefits are highly variable locally, for instance some farmers weed thoroughly and do not regard such inputs as being anything "extra", whereas other farmers do. Byproducts of farm production such as stover and residue grazing may not be accounted for, nor even exploited.

The financial costs of water harvesting structures in particular can be highly variable, perhaps from US$ 50 - US$ 1000 per hectare. Where mechanised transport is used to import rocks for bunds, and where tractors and bulldozers are used to move earth, costs are always very high compared to small hand made structures. However, the mechanised systems may allow water control over a larger area and permit a more fully integrated approach to water harvesting. This can bring benefits in erosion control at the large scale, but these benefits may not be recognised or accounted for.

Hand Construction:

Hand tools and manual labour will limit the size of structures that can be built and the scale of activities. However, the advantages of cost and sustainability should not be overlooked. Hand tools can provide a flexibility of operation that is not attainable by machinery and both ridge-based and broad-based bunds are best finished by hand. In many cases this equipment will be the most suitable and the control of such labour does not impose any serious technical difficulties, beyond limitations of scale.

Ploughs:

Figure 7.24 shows the construction of a waterway by disc plough. The plough should be correctly set. Once the soil is pulverised it should be allowed to settle, encouraged by water (or rain) if possible. An increase in tractor speed will increase bank height. More discs are best suited to large straight structures, fewer discs for bunds. Other ploughs can be used.

Other Machinery:

Blades: only powerful machinery can move soil with a blade and the use of such equipment is rarely warranted, except for large structures.

Scoops: animal drawn scoops can be very useful for bund construction, after soil has been loosened by ploughing. Alternatively scoops can be moved by tractor.

Various ridgers, trenchers and ridge tying units have been developed to suit local needs.

Figure 7.25 shows the construction of bunds using a disc plough.

Numbers indicate position of equipment for each round. Apply water after each two phases, continue until required section is achieved. Final section is smoothed with a blade terracer or by hand.


Figure 7.24: Construction of Waterway with Disc Plough

Numbers indicate position of equipment at each round


Figure 7.25: Using a Disc Plough to Construct Bunds

Equipment costs

All costs of locally made equipment are approximate. The costs of raw materials and especially labour are highly variable from country to country, but a good idea of cost magnitude can be gained from the figures quoted below. The costs of manufactured equipment are based on 1993 prices. Shipping costs, agents' fees and fluctuations in exchange rate cannot be taken into account.

Item US

Quantity

Typical Approximate Cost in $

Abney level complete

1

200 -300

Automatic level complete

1

600 -800

Levelling staff

5 m

50 -100

Appendix D1: Bund dimensions for various areas, slopes and soil types

Bund Length - Catchment Area versus Slope SAND V.I = % S + 3 / 2 (VI in feet)

Note: 1 metre 3.281 feet

1 hectare 2.47 acres


Figure


Figure


Figure


Bund Spacing - Horizontal Distance versus Slope (feet)

Conversion Factors:

1 foot 0.3048 metre

1 inch 2.540 centimetre

1 acre 0.4047 hectare

(introduction...)

Although runoff information is most important to the hydrologist, it cannot be treated in isolation and many of the methods of analysis used on runoff data are commonly applied to other types of information. The analysis of all data is extremely important because it can be used to understand why and how processes happen, though many statistical methods ignore the understanding of behaviour and focus upon probabilities, fitting data to particular distribution patterns, the form of which can be defined. Knowledge of both the processes that take place and statistical methods used to predict hydrological behaviour should be sought. A knowledge of statistics is essential in data analysis, and in this chapter a deliberate attempt has been made to include a comprehensive explanation of statistical methods while at the same time avoiding over-detailed statistical theory. The intention is to help in the selection of the correct techniques of analysis and explain how they are used.

The desire to understand the complex inter-relation of hydrological processes and, later, the temptation to ignore this complexity for statistical methods has led to three main approaches to the study and treatment of hydrological data.

Deterministic Hydrology: assumes that certain influences determine the passage of rain to its ultimate destination, and that the physical environment is responsible for the presence and level of importance of these influences. No concept of probability is involved, although where deterministic models are used, their accuracy is defined by statistical parameters.

Probabilistic Analysis: defines the chance of particular values of data occurring. It is independent of time and sequences of events do not carry any significance.

Stochastic Methods: recognise that probability plays a large part in the nature of hydrological data, but also that the sequence of the data also carries significant information. In many instances, this type of analysis may be regarded as a combination of the two other methods.

Hydrological data are stochastic in reality, but it is easier to deal with them, mathematically at least, if they are regarded as probabilistic or deterministic. Stochastic data represent a time series that may be viewed in discrete periods or continuously. The daily, monthly or yearly flow hydrographs are good examples.

An understanding of hydrological processes enables a pattern to be fitted to runoff events that have numerical values. This pattern-fitting is used to predict hydrological behaviour and is often referred to as "modelling", though the term is often over-used; a simple regression that states that 20 mm of rainfall on a given catchment will produce 100 m³ of runoff under particular conditions will often be referred to as a statistical "model".

It is not surprising that hydrological analyses and models have become progressively complex. The attempt to derive models that can be generalised for use in many geographical regions, yet at the same time give accurate results, demands complexity, but could be regarded as self-defeating. In addition, many hydrological data must by analysed in a manner that cannot be regarded as "modelling" in any mathematical sense. These analyses are essentially the treatment of data to provide further information or to render basic data into more useful forms.

It is important however, to evaluate carefully the needs that are required for a particular project. The simple treatment of hydrological information, collected in the area for which results will be applied, can be far more useful to agricultural development than complex, imported models generalised from parochial results.

For many projects the cost of data collection (both in money and time) and the duration of records will impose severe limitations on the analysis of data and the development of models. For example, the simple linear regression of daily rainfall against runoff not only provides a causal relation, the accuracy of which can be quantified, but in the context of long-term daily rainfall records (which are usually available in most countries) can give an acceptable probability basis upon which to plan field layouts and runoff control. The data collected from many projects may not be adequate to allow a more sophisticated treatment. The following chapter concentrates on the important analyses that should be considered when agrohydrological information is being studied.

This chapter is subdivided into two main sections:

8.1 Statistical Methods and Data Analysis

This section deals with the statistical methods that are essential techniques in the analysis of data and provides examples of particular statistical methods to ensure that the general forms of analyses which are described, are fully understood. The data under study and which is used to provide examples are hydrological data. However, these statistical methods are applied to other information, most commonly rainfall information.

8.2 Non-Statistical Analysis of Data

This section discusses methods by which data that are not amenable to statistical treatment are prepared and studied; often these treatments are concerned with changing basic information into a more useful form. Sections on the analysis of hydrological, rainfall and meteorological, evapotranspiration and sedimentation data are included.

8.1 Statistical methods and data analysis

There are several kinds of hydrological data that may be collected. The quality of these data will be determined by various errors.

Random Errors: cannot be avoided, though good practice keeps them to a minimum. They are assumed to have a statistically normal distribution, that is that there are as many low values as high values distributed around the mean, and that there are more values close to the mean than at the high and low value extremes.

Systematic Errors: usually show an increasing or decreasing trend, for example reduced water depth measurements due to the progressive stretching of a measuring cable.

Non-homogeneity: is not an error as such, but effects the values of time series data in a progressive way. It may be due to changes in catchment characteristics; for example progressive de-forestation, causing a trend of increased runoff.

The term "population" is often used in statistical analysis and is used to describe the variable values that are under consideration, the sample population may be regarded as a selected group of those values that are used.

8.1.1 Elementary Statistical Properties

Mean, Median and Mode

The arithmetic mean is a widely understood idea. It is easily calculated and is the usual method of describing the "central tendency" of a group of data. With distributions that have a small sample of very high or very low values, the mean may be greatly influenced by these values and may not accurately reflect the central tendency of the data, a situation often found with runoff and rainfall data which often have a few, very high values.

The median value is that which falls exactly in the middle of a range of values and not is affected by the weighting of a few extremely small or large values. Hydrological data often takes this type of distribution and therefore the median is frequently used.

The mode is the value in a group of data that occurs most frequently and used in extreme value asymptotic functions.

Standard Deviation and Coeffcient of Variance

The standard deviation is a measure of dispersion of values of a variable "x" around the mean value. It is the square root of the mean-squared deviation of individual measurements from their mean and is defined by the equation for unbiased standard deviation, which is preferred:

xxxxxxxxxxxxxxx(8.1)

where x = mean of x
It may be regarded in a non-statistical sense as the "average" value of dispersion of values from the mean.

Figure 8.1 below shows a set of time series data with the arithmetic mean and standard deviations. One standard deviation + and - the mean represent the 68% level confidence limits.


Figure 8.1: Elementary Statistics of Time Series Data

The standard deviation is denoted by d. The standard deviation of a sampling distribution is the standard error. The standard error of the mean =

xxxxxxxxxxxx(8.2)

and standard error of the standard deviation is =

xxxxxxxxxxxxx(8.3)

The variance is the square of the standard deviation. The coefficient of variation is a dimensionless measure of the dispersion around the mean and is defined by:

Std dev, -Sx/

xxxxxxxxxxxx(8.4)

Statistical computer programs and many electronic calculators give these parameters as a matter of course, though they may be calculated manually if necessary.

Skewness

The lack of symmetry of a distribution around the mean is called skewness. The skewness of the population is defined as:


xxxxxxxxxxxxxx(8.5)

The unbiased estimate of skewness from the sample is given by:


xxxxxxxxxxxxxxxxx(8.6)

8.1.2 Correlation Analysis

Correlation analysis measures the association of one variable with another; and is often used to define associations between several variables. The table that defines these relations is called a correlation matrix and is a standard part of the output of statistical analysis computer packages. The variation of both (or all) variables is not fixed by the observer nor by control and unlike regression analysis, correlation does not pre-suppose a causal relation between the variables. Because no inference of causality is present, correlation does not allow the prediction of values of one variable from another (as does regression), by the use of a formula that describes the correlation. However, correlation defines association and has a practical application; it is used to select variables that are suitable for multiple regression analysis.

For example, it may be required to obtain an equation by multiple regression (using previously collected data) that will predict runoff amount, given that values for rainfall amount and catchment physical characteristics are available. Table 8.1 illustrates this. A number of variables, obtained from catchments in Belize, Central America, were used in multiple correlation analysis were used in this example.


Table 8.1: Correlation Matrix of Hydrological Variables

It can be seen that pairs of variables such as catchment area (AREA) and mainstream length (MSLE); stream frequency (STRFQ) and the soil/slope index (SOS); Average main stream slope (AMSSL) and basin slope (BSL), are highly correlated. In regression, variables used as "independent" variables should be independent of each other and should not be significantly correlated. It is not be appropriate to use highly correlated variables, such as those in the example pairs, together in the same regression analysis. The inclusion of correlated variables does very little to improve the quality of the resulting equation.

The numerical value that describes relations between variables is the "correlation coefficient", 'r' and relates to the sample data. Correlations can be positive or negative. As the number in the sample becomes large the distribution of the standardised variable "t" can be used to test the significance of the correlation (Student's "t" test).


Table 8.2: gives data for testing the significance of correlation between variables. They may also be used for testing significance in regression analysis.

Correlation Significance Examples:

a. SOS (soils/slope index) and STRFQ (stream frequency in terms of number of streams per square kilometre) have: a correlation coefficient, r = 0.961, n, the number of data points (or "sample size") = 15 degrees of freedom (d.f) = n - 2 = 13

From Table 8.2, variables with a correlation coefficient greater than 0.760 are significantly correlated to the 99.9% level (0.001) and these variables would not be suitable for simultaneous inclusion in regression analysis. Indeed, even a significance level of 90% (0. l)would justify their mutual exclusion.

b. SPFN (floodplain area) and RSMD (a soil moisture deficit index) have: r, = 0.158, n = 15,d.f = 13

From Table 8.2, the relation is not nearly significant even at the 0.1 (90.0% ) level and these variables would be suitable for simultaneous inclusion in regression analysis. Standard textbooks on statistics provide a more detailed background on the "t" distribution and limiting conditions for its use.


Table 8.2: Table for Testing the Significance of the Correlation Coefficient 'r'

8.1.3 Regression Analysis

Methods Usually Adopted to Estimate Flow Volumes

a Simple Regression

Regression analysis is a widely used method to derive formulae that define the relation between two variables and unlike correlation analysis, admits this relation to be causal. Very many of the predictive or relation equations found in the literature and that link hydrological parameters are in fact regression equations.

Regression analysis is often seen as a simple x-y graph, but with distributions of real data, it is rare that all the data points fit exactly on the line of best fit drawn through them. An even balance of points on, above and below the line is sought, with the line drawn to minimise the dispersion of points around the line. The method of "least squares" is used to draw such lines, and although regression and the least square for each point can be calculated manually, computer programs that provide a best fit in this way are so common that manual calculation is almost obsolete. A standard textbook on statistical techniques will provide examples of these manual computations, if desired.

The values of the independent variable are plotted on the x axis (in this case rainfall) while the values of the dependent variable (runoff) are plotted on the y axis. The initial step is to plot a scatter diagram of values of y on x. This gives a visual idea of whether a significant relation between the two variables may exist, but does not define it. Statistical computer packages (and sometimes graphics packages also) are widely available to plot scatter diagrams and undertake regression. They will also provide numerical values for equation 8.1, evaluate the significance of the relation and quantify residuals (the errors in the fit of the regression line) and confidence limits. The form of the equation of simple regression is:

y = a + b (x) where (8.7)

y = runoff
x = rainfall
a = intercept value
b = gradient of the graph (a and b are known as regression coefficients)

The selection of variables for regression is based on a reasonable assumption that a variation of the independent variable will cause a variation in the dependent variable. A good example is the relation between rainfall and runoff; it is most important in the runoff process and the starting point for most runoff analyses. Figures 8.2 (a) and (b) show examples of linear regression using daily rainfall and runoff values from two rangeland runoff plots of 0.4 ha extent. The details of the relations are given.

(a) y = 1.0497x - 6.30

(b) y = 0.1266x +1.16

n = 44

n = 10

R2 = 0.83

R2 = 0.34

P<0.001

P>005

The relation is significant to the 99.9% level

The relation is not significant to the 95% level

In the case of graph (a), a significant linear relation may be assumed from the distribution of the data points. When the regression analysis is complete this significance is verified to the 99.9% level. This indicates that there is a less than a one in one thousand chance that the distribution of these data points is due to chance and indicates a strong causal relation between the amount of rainfall and the amount of runoff. The value of R2 (the coefficient of determination) indicates that the variation in rainfall explains 83% of the variation in runoff.

No definition of the influence of other catchment factors which determine the amount of runoff from a plot such as this (soil texture, slope, vegetation cover etc.) can be made, other than that, combined, they account for the remaining 17% of variation.

The distribution of data points in graph (b) may appear to show a significant relation, but the value of R2 is 0.34 and testing shows that this is not significant to the 99.9% level, nor even to the 95% level. There is more than a one in twenty chance that the relation is explained by chance, rather than the influence of rainfall amount on runoff amount. The selection of significance levels is one of subjective choice. As a hypothetical example, the relation of graph (b) may be significant at the 50% level, but this admits that chance is as likely as not to be responsible for the apparent relation. In general, the 95% level is the lowest accepted, though the convenience of the 68% level of confidence being equal to the mean + / - one standard deviation results in its occasional use.

The availability of data for analysis is an important consideration in experimental planning and the collection of data. Analyses with very few data points need very high values of R2 to assure significance, since the number of degrees of freedom (d.f.) are = n - 2 and with few data, values of d.f. are very small. Graph 8.2 (b) has only 10 - 28 degrees of freedom and its R2 would need to be 0.87 to give P< 0.001 (99.9%), whereas graph (a) with 42 degrees of freedom would only need an R2 of 0.48 or so. Where runoff data are expected to be sparse, for example in arid and semi-arid climates, a single runoff plot that typifies one kind of environment will not be adequate to provide a sufficient number of data for useful analysis, unless it is monitored for many seasons. The data for graph (b) represents the collection of information over three seasons; runoff was rare, some data were lost and the dataset of 10 values is really inadequate for use in regression analysis.


Figures 8.2 (a) and (b) Regression of Daily Rainfall Versus Runoff

A shortfall of data for large rainfall/runoff events is a particular problem. Such events are naturally infrequent, but are usually of great interest. If the influence of rainfall on runoff is the case under study, there is strong justification for the use of many replicate plots of the same soil, slope, vegetation cover, etc., that are located together and which experience the same rainfall. The data from such replicates could be combined into one dataset and used as though they came from one plot.

The independent variable under study (in this case rainfall amount) should be allowed to range as widely as possible, while the values of all other influences (for example catchment size, slope and vegetation cover) should kept under tight control. This approach requires extensive replication and the cost of construction, instrumentation and monitoring must be balanced against the increase of data that is achieved. The input to collect sufficient data during a project's time, may exclude experimentation in other areas of interest.

Data from catchments that are only generally similar may also be combined so that the range of physical circumstances to which the equations can be applied may be widened, but the accuracy of such equations in predicting the response of the dependent variable will be lessened. Simple linear regression can be undertaken between many factors, though most combinations are likely to involve runoff as the dependent variable.

Goodness of fit

The Chi-square goodness of fit test can be used to decide whether a line fitted by one method of fitting is more accurate than that of another method; for example whether the least squares technique is more accurate than fitting a regression line by eye, or whether one theoretical statistical distribution gives a closer fit than another.

The Chi-square parameter is defined as

xxxxxxxxxxxxxxxxx(8.8)

Where two different methods of fitting are used, the method with the lowest value of x2 shows lowest dispersion of points around the line drawn and is the most suitable. The use of this test between different regression fits is not so important because of increased computerisation. However, it is widely used to decide whether one statistical distribution (for example Pearson Type III) fits the data better than another (for example the Normal distribution).

For instance, if a Pearson Type III distribution has a X2 value of 4.56 and a Normal distribution has a X2 value of 5.04 for rainfall data, the differences of fit are smallest for the Pearson distribution and it is to be preferred. Moreover, given that the degrees of freedom for the data 'v' (=n -1) is 10, then neither of the distributions can be rejected at the 0.05 (95%) level, a , because the critical X2 value for v = 10, a = 0.05 is 18.31 and the X2 values for both distributions are less than this critical value. Critical values of X2 for various v and a can be obtained from statistical tables in Appendix E.

Confidence Limits

As discussed above, the value of the results obtained from regression can be qualified by testing for the significance of the relation; the higher the significance, the stronger the relation and the more useful the equation will be to predict future runoff amounts. Confidence limits define the range of probable values of any estimate made from a regression equation, and as stated previously, those most commonly used are the 95% confidence limits. These limits are positioned above and below the regression line (+ and - ) and diverge as the values of x (or more correctly [x - mean x] ) increase and are therefore parabolic, but symmetrical above and below the regression line. Where a log transformation of data has been applied, (see below) the confidence limits will not be symmetrically distributed around the regression line. For example, a hypothetical logarithmic regression equation may give the value of runoff 'y' for rainfall 'x' to be in the range, y = 135.6 m³ + 6.8, - 5.3 m³. Thus the value of y will lie in the range 142.4 to 130.3 m³, with a confidence of 95%.

Confidence limits for a range of y values (sometimes called control curves) may be may be calculated manually by the use of a range of x values, but this is an unduly laborious process when large datasets are being analysed. Standard statistical packages usually include confidence limit plotting as a basic feature, though the quality of graphical output varies a lot.

The importance of confidence limits, and also the importance of adequate detail can be seen from Figure 8.3. The logarithmic regression of rainfall against runoff for this graph (in log10 form) is P < 0.001 and therefore highly significant. However, the dispersion of points around the mean is relatively large and the 95 % confidence limits are therefore widely spread. The details of the relation of this graph are:

n = 41
log (a) = -3.26, (se = 0.68)
b = 2.09, (se = 0.28)
R2 = 0.59

The relation is significant to the 99.9% level.

The catchment from which these data were obtained was located at a site in SE Botswana with a resident gauge reader/observer, but only 41 data points could be used for the analysis out of a total of 45. Four data points were lost due to equipment malfunction and silting of the flume, though it is unlikely that 4 extra values of rainfall and runoff would have improved the regression greatly. The main reason for the small dataset, collected over three years, was the semi-arid climate of the region which resulted in infrequent rainfall and an average of only 15 runoff events each year. As can be seen from the graph, it would be foolish to estimate runoff from rainstorms greater than 20 mm or so, because the confidence limits are so wide and only six plotted points represent rainfall equal to or greater than 20 mm However, if interest lay in low rainfall/runoff values, the analysis would be satisfactory and the removal of the higher values would improve the relation of a re-plotted regression and more reliable estimates of low flows could be obtained.

Usually the interpolation or extrapolation of high runoff values that are of interest, however. It must be stressed that in this case the problem lies not with the analysis, but with the limited amount of available information. Serious consideration must be given to experimental planning, so that adequate data are collected.


Figure 8.3: Regression of Rainfall versus Runoff with 95% Confidence Limits

Transformations of Data for Regression

Hydrological data frequently do not exhibit a linear form when plotted, and it is common practice to transform the data before regression is attempted, so as to arrive at linearity. Also, transformation may stabilise variance of the data and render the points along the regression line more homogeneous. A number of transformations are available, but by far the most common is the transformation of the data to the logarithms of the values before regression, such that the equation of the relation becomes:

y = axb, that is

log (y) = log (a) + b [log (x)]

(8.10)

y is = antilog log(y)

Research has shown that other modifications of variables (for example the use of (Rainfall) 2 versus runoff) can give improved results in obtaining a descriptive equation.

Isolated data points that conform very poorly to an otherwise good linear relation (these points are called "outliers") can be removed from the analysis if an obvious reason for their inaccuracy can be detected. For instance, errors that commonly cause such outliers may be silting of the flume, inaccurate gauging of a river, a change in the rating curve that has not been corrected or allowed for, or the incorrect setting of a water level recorder. It is important to refer to the original source of information (field data sheets for example) to identify errors, or gather explanatory notes on unusual occurrences.

b. Multiple Linear Regression

It is useful that regression can predict a response in dependent variable y, from the changes in a number of independent variables. For example runoff amount may be estimated from the values of rainfall, land slope, soil texture, soil moisture, etc., by the use of a single equation. The form of the equation for linear multiple regression is:

y = a + b1x1 + b2x2+......... bkxk where (8.11)

x1........xk represent the different independent variables.

Values are entered into the statistical database for the dependent variable y and independent variables x1, x2,..........xk.

The independent variables should not be significantly cross-correlated. The quality of the total regression is described by the coefficient of multiple determination, R2 (the square root of R2 is the multiple correlation coefficient) and values of the standard error of prediction are often given by computer packages. Degrees of freedom for testing for significance of the regression are calculated by using (n - k ) -1, where n is the sample size and k is the number of x independent variables.

It is important to recognise that for each x variable that is added to the regression equation, a degree of freedom is lost. Where the number of data are small, the addition of independent variables can actually be counter-productive, especially if a variable does not improve the quality of regression to any great extent. The transformation of data, for example to a logarithmic base, is also a common preparation for multiple regression analysis. The methods of prediction of values of y is similar to that of simple regression.

Typical computer printout information is shown below using examples from the variables in Table 8.1, the correlation matrix.

Dependent variable, Runoff,

y = 2.69 × 10-4 + Area3.698 + Soil Moisture46.029 + Stream Frequency-7. 441

R2 of the equation is 0.893

Standard error of the estimate of the equation is 0.383, the standard error of the coefficients are, respectively, 1.024, 11.784 and 1.777 with 95% confidence limits of - 83 to + 483.

The data of the equation were transformed to log 10 before regression and runoff is in m³.

Regression analysis is most usually adopted for the prediction of runoff amounts or volumes. Theoretical distributions are most commonly used to estimate peak flows and such distributions are described below.

8.1.4 Statistical Probability Distributions

These distributions are adopted to estimate flows, their probabilities and return periods. The return period of a flow is the reciprocal of its probability. In particular, knowledge of peak flows is essential in the successful planning of water harvesting and field systems and in this context it is the size and frequency distribution of peak flows that are often of paramount importance. The return periods and sizes of flows involved will be determined according to the aims of a project, but most analyses must overcome the almost universal difficulty of working with short periods of data. For example, studies show that 80% of estimates of the 100 year flood, based on records of 20 years, are overestimates. Fortunately, planning for such long return periods is rare in agrohydrology and many techniques of analysis have been developed with the problems of short records in mind.

The analyses applied to predicting flow data are primarily statistical. Two basic questions can be asked with regard to risk and design:

- What is the probability (p) of a flow Q being exceeded during the design life L?
- What is the flow Q which has a selected probability (p) of being exceeded during the design life L?

The study of probabilities attempts to answer these two questions.

Probability Paper

Before looking at the various distributions that can be applied to runoff data, it is important to consider the manner in which these distributions are actually plotted as graphs.

Probability paper is used to plot, manually, cumulative probability (x axis) against a variable (y axis) and is designed so that the data will fall on a straight line, if it actually conforms to the selected distribution. Different types of distributions require different types of paper and such a plot is used as a convenient guide to the interpolation and extrapolation of variables, probabilities and return periods.

In the distributions discussed below Extreme Value Type I and normal distributions are plotted on Gumbell - Powell paper (y axis rectangular, x axis Type b; EV III plotted on Weibull / log. extremal paper (y axis log, x axis Type I) and log-normal distributions are plotted on log-normal paper (y axis log, x axis normal probability).

Where probability paper is not available, the probability scales can be constructed from the equation:

x= mean x standard deviation K, (8.12)

The value of K with corresponding return period Tr can be obtained from the tables provided below, with the discussion on different distributions (note log-normal Pearson Type III with skew = 0). A rectangular scale of K is drawn and the corresponding value of T is transferred to the x axis of the paper. The probability value p is the reciprocal of Tr. The plotting positions according to Weibull or Gringorten are use for EV1 and normal distributions. Weibull is most commonly used for the annual maximum series.

It is likely that the data points will not fit exactly on a straight line and a line of best fit may be fined by eye or regression analysis (the method of least squares).

There is an obvious advantages in the interpolation and extrapolation of probabilities, where runoff data are found to conform to theoretical statistical distributions and any flow can be described by the parameters of the distribution. Many distributions have been studied to discover any such conformity. Hydrological data is usually highly skewed and not evenly distributed about the mean; rather there are usually very many small values and a few, very large values. This has generally precluded the successful use of the Normal distribution, as data in this type are distributed uniformly around the mean. The use of the Log-normal distribution has been widespread in the past, because when transformed, the logs of peak flow magnitudes are commonly seen to be normally distributed. The Log-normal distribution is a special case of the Pearson Type III distribution, described below, with skew equal to zero.

It is not surprising that no particular distribution is universally appropriate for the fitting of hydrological data, though the log-Pearson Type III and the Gumbel Type I (EV1) distributions have been adopted for flood study in the USA and UK, respectively. These two distributions are discussed below with worked examples, though the details of the statistical theory is only given in outline.

Most frequency functions can be generalised to the form: X =

xxxxxxxxxxxxxxxxx + K sd x; where (8.13)

X is a flood of a specified probability

xxxxxxxxxxxxxxxxx is the mean of the flood series
sd x is the standard deviation of the series and
K is a frequency factor defined for each specific distribution and is a function of the probability of X

a. Log-normal Distribution

This distribution has been used historically, as a suitable distribution for flood flows. It is a transformed normal distribution with the variate data transformed to logarithmic values.


Figure 8.3: Log Normal Plotted on Probability Paper

The probability density is given by:


xxxxxxxxxxxxxxxxxx where (8.14)

y = ln x,
m is the mean = emy + dy2/2
d is std dev. = m(edy2 - 1) 0.5
M is the median = emy
CV is coefficient of variation = (edy2 - 1)0.5
Cs is the coefficient of skewness = 3CV + CV3

When data are plotted on log-probability paper, a straight line occurs only for one value of Cs ( 1.139). Curved plots of data indicate the need to modify Cs. Figure 8.3 shows a plot of discharge versus probability. To widen the opportunity for fitting data to a defined distribution, more complex treatments of data have also been studied.

b. Log-Pearson Type III Distribution

This one of a series of distributions. The data are converted to logarithms and the mean is computed using equation 8.15 as a basis:

Mean

xxxxxxxxxxxxxxxx (8.15)

The standard deviation is given by

(8.16)
end the skew coefficient by

(8.17)
The value of X for any probability level is computed from log x = log x + K·sd log x. (8.18)


Table 8.3: Values of K for the log-Pearson Type III Distribution

The Pearson Type III is a skew distribution, bounded on the left like many hydrological distributions. The skew parameter allows flexibility in fitting it to datasets and when the skew is 0, it is identical to the semi-log distributions used commonly in the past for hydrological analysis. Table 8.3 gives values for K for the log-Pearson Type III distribution which are used in the calculations of the worked example.

c. Gumbel (Extreme Value) Type I Distribution

The EV1 distribution, like the Pearson Type III, is one of a family of distributions, and its form parameter is equal to zero. As is explained in the Flood Studies Report (NERC, 1975), the distribution of maximum values selected from a data set approaches a limiting form when the size of the size of the data set increases. If the initial distributions within the dataset are exponential (see Peaks Over Thresholds, below), the Type I distribution results. The form of the distribution is given by:

p= 1 -e(-e)-y where (8.19)

p is the probability of a given peak being equalled or exceeded, e is the base of natural logs and y is a reduced (i.e. standardised) variate, a function of probability. Thus

x =

+ (0.7997y - 0.45) sd x (8.20)

In this case, the term in brackets in equation (8.20) is equal to the Pearson term K, that is the frequency factor K(Tr) is equal to - 0.45 + 0.779y (Tr).

Table 8.4 gives terms of the Gumbel distribution, for the calculation of flow probabilities.


Table 8.4: Gumbel Type 1 (EV 1) Distribution

Worked Examples

Pearson Type III Distribution:

A hypothetical set of log peak flow (=log q), annual maxima data from a small agricultural catchment when analysed, give the following values. Find the 5 and 25 year return period peak flows:

Mean of the logs of q,

= 3.087
(not log of the mean)

Standard deviation
of the logs of q, sd log q = 0.981

Skewness coefficient G = 0.0390

The 5 and 25 year peak flow (q5) is found by the following procedure:
From Table 8.3, K5 = 0.855 and K25 = 1.610

Therefore from equation 8.18, log q5 = 3.087 + ( 0.981 × 0.855) = 3.926
q5 = antilog 3.926 = 8,429 1 s-1 (8.4 m³ s-1)
From equation 8.15, log q25 = 3.087 + (0.981 × 1.610) = 4.666
q25 = antilog 4.666 = 46,388 1 s-1 (46.9 m 3 s-1)
Gumbel (EV 1) Distribution

A set of hypothetical annual maxima data from a field catchment give the following values. Find the 10 and 100 year peak flows:

Mean peak flow, q = 1,234 1 s-1
Standard deviation q = 434 1 s-1

From Table 8.4, K10 = 1.31 and K100 = 3.14

Therefore from equation 8.20 q10 = 1,234 + (1.31 × 434)=18031 s-1
and q100 = 1,234 + (3.14 × 434) = 2597 1 s-1

8.1.5 Extreme Value Series

It has been found that in many cases, the whole set of data of hydrological events in a water year (that is the beginning of the wet or dry period of one year to the same time the next) need not necessarily be used in analysis. The largest or smallest values in a particular time period may be analysed instead, and often this time period is selected so that one flow per year is used; the Annual Series. The annual series is called the Annual Maximum Series where the largest flows are used or the Annual Minimum Series when lowest flows are analysed. As the time period increases, the data become less inter-dependent. In regions with discrete seasonal variations of flood, time periods of a few months may render the data independent and the annual maximum series is widely used.

Where a base level of flow is selected so that only floods which exceed this base are used for analysis, the data are said to form a Partial Duration Series. This type of series will be discussed below, but there is really little difference between it and the annual maximum series, except that the high level of the base in the latter excludes all flows but the greatest each year.

Annual Maximum Series (AMS)

The annual maximum series is a special case of the extreme value series and for return periods of 5 years or more it is often suitable, though the design of structures and the nature of a project will also influence whether the annual maximum series should be chosen. For instance a dam or bridge may not only be affected by the largest flood, but also by the second largest and other flows associated with a flood period. The annual maximum series takes no account of these other runoff events and therefore may not be suitable for use. Alternatively, a culvert may be washed away by the one large flow, but can easily and cheaply be repaired and the AMS may provide a suitable method of analysis. The annual minimum series may be used where low flows are under consideration.

In Table 8.5 the AMS is used to illustrate the estimation of size and probability of peak flows. The difficulties of small datasets may be acute when using the AMS, as only the single largest annual (or seasonal) peak is used in the ranking. The data may also be plotted on a log-probability graph with peak flow values on the log., y, axis. The form of the data is a straight line and this renders the extrapolation of higher return period flows, relatively simple.

The Flood Studies Report (1975) is a major work on flood probability analysis and recommends the following, alternative relations for the probability and return period terms in Table 8.5:

p = (m - 0.44 )/ (n + 0.12) and Tr = (n + 0.12)/ (m - 0.44) (8.21), (8.22)

These relations define "plotting positions" where p is the probability, m the rank of the flow, n the total number of items in the rank and Tr is the return period in years. Several such plotting position formulae have been tested (the examples shown in Table 8.5 are according to Weibull) and used. Those in equations 8.21 and 8.22, known as Gringorton's formulae, attribute longer return periods to higher floods in the series.

The difficulty experienced in using many methods to extrapolate from short spans of data is due to the estimation of the tail of the distribution from values not included in this tail. The annual maximum series is especially vulnerable to error since it discards most of the available data. The use of theoretical distributions, in particular when used in conjunction with partial duration series, are attempts to overcome these problems, though ultimately long periods of records are the best basis for probability estimates.


Table 8.5: Annual Maximum Peak Flows

8.1.6 The Problems of Short Records and Partial (Duration) Series

A partial series is made of all peak flows that exceed a selected threshold, the AMS being one particular case. In other instances, a number (usually from about two to five runoff events each year) are included, so that a larger range of data is selected from periods of short records. For this reason the partial series may also be known as the Peaks Over Threshold (POT) model. Although considerations have been made as to whether the exceedences should occur within a water year or a season, these considerations are less vital than the type of distribution of flood magnitudes, which influences the estimation of high return periods for given flows.

Several models have been studied, but the simplest is presented here. It assumes a random distribution of exceedences in any year or season and an exponential distribution of the magnitudes of these exceedences. These assumptions combine to give a Gumbel Type 1 distribution.

The problems of short records and the need to estimate flood flows for relatively small return periods is the situation usually faced by agrohydrological projects. The Flood Studies Report investigated the use of partial (duration) series to overcome the problem of short records and the work that it reports is of value, though the use of partial data is relatively well known.

a. Annual Exceedence Series (AES)

This series is a particular type of partial duration series, where the dataset is obtained by setting the level above which flows are included is such that the number of flows is equal to the number of years record. In real terms this means that some years will provide more than one flow, while some years will provide none at all. In this way, the occurrence of several high flows in one year will be taken into account, unlike the annual maximum series. The relations between these two series (and indeed the annual maximum and partial duration series in general) are discussed below. In the example above on the suitability of series selection and the destruction of bridges, dams and culverts, the place of the annual exceedence series can be seen.

b. Peaks Over Threshold Model

For very short return periods (Tr), values obtained from the annual maximum series and the POT model differ appreciably, when longer periods are considered, they are very similar. However, the basic question as to which is the most suitable method of estimation to use, is still open to question. For example, the annual maximum series, by limiting exceedences to one peak per year, may actually bias the sampling distributions by not taking into account, say the second or third largest flood on record. On the other hand, the POT model is often seen to work better for small return periods when only one exceedence a year is used. For longer return periods, a greater number of exceedences a year seem more suitable. Below is a list of ratios of return periods for the POT and annual maximum series:


Table 8.6: Return Period (Tr) POT versus Annual Maximum Series Return Period (years):

The simplest manner in which to abstract the data from the full set is to decide upon a number of flows per season to be analysed which exceed an (as yet) unknown threshold and from the relations given below, calculate the threshold.

The formula that gives return period flow is:

Q(T) = q0 + B (ln $ + ln T) (8.23)

Q = the peak flow of return period T
q0 = the threshold value
B = the gradient of the distribution
$ = the number of exceedences

The sample size is noted as M, the minimum flow of the sample is qmin, and the mean is q.

The linking equations of the parameters are:

q0 = qmin - (B/M) and (8.24)
B = M ( q - qmin) / (M - 1) (8.25)

The sampling variance, var. QT is given by:

var. QT = B2/N [(1- ln $ -ln T )2 / N$ -1] + (ln $ + In T)2 where (8.26)

N = the number of years of record. The procedure for using the POT model is shown below, by worked example.

Worked Example of the POT Model

As stated earlier, the assumption is made that the magnitudes of the peaks are distributed exponentially, however in some instances this may not be the case, therefore the first step in approaching the use of the POT model is to verify this exponential assumption. The data in Table 8.7 were obtained from a 4,000 m² rangeland runoff plot with a land slope of about 2%, sandy loam soil and a vegetation cover in the range 55-65%. It is important to note that no major changes in the catchment characteristics took place during the period of record.

From the total set of peak flow data, a number of the highest flows are selected, such that M, the total number of peaks selected = $ (exceedences per year) × N (the number of years of data). For this example $ = 3 and $ = 5 exceedences per year were used to illustrate by result, any variation due to the use of different exceedences. They were abstracted from data collected over only 3 years.

Thus, a total of 9 and 15 peaks were used in each case, though there is no statistical evidence to suggest that a better result is obtained from using a larger numbers of exceedences. It is important to note that like the annual exceedence series, no regard is given to whether or not the peaks occur in any particular year, it could be possible, but unlikely, that all data were drawn from the same year.

The values of the reduced variate y are calculated from:


xxxxxxxxxxxxxxxxx (8.27)

such that when N, the number of years from which the data are drawn = 5, then:

y1 = 1/5 = 0.200; y2 = 1/5 + 1/4 = 0.450.........y5 = 1/5 + 1/4 + 1/3 + 1/2 + 1/1 = 2.283, etc.


Table 8.7: Data for POT Model, Exceedences 3 and 5 per Year

The peak flows in litres per second and variate y are plotted as in Figure 8.4, both distributions appear closely exponential.

The calculation of Q10 and Q25 peak flows for $ 3 and $ 5 are shown below.

For $ 3 and using equation 8.25

For $ 5 and using equation 8.25.

B = 9 (38.53 - 29.90) / 8 = 10.83

B = 15 (33.15 - 21.80) / 14 = 12.16

Using equation 8.24,

Using equation 8.21,

q0 = 299-(1083/9) = 28.70

q0 = 21.80-(12.16/ 15) = 20.99

Using equation 8.23, the 10 and 25 year peak flows are:

Q10 = 28.70 + 10.83 (ln 3 + ln 10) = 65.5 l s

1Q10 = 21.80 + 12.16 (ln 5 + ln 10) = 69.4 l s

Q25 = 28.70 + 10.83 (ln 3 + ln 25) = 75.5 l s

1Q25 = 21.80 + 12.16 (ln 5 + ln 25) = 80.5 l s


Figure 8.4: Plot of Peak Flow Q versus Reduced Variate y

Stochastic Analysis

The term "stochastic" is frequently met with in hydrology and although the term is used in statistics to define the data as being governed by the laws of chance (synonymous with terms such as "random" or "probabilistic") in hydrology "stochastic" refers to time series that are partially random and may be regarded as a treatment of data halfway between probability analysis and deterministic modelling.

Although the occurrence of events is regarded as being random, the order in which the events occur is regarded as carrying significance, unlike probability analysis which is concerned only with size and number of events. Stochastic hydrology is useful for design and decision-making in hydrology, since it is assumed that although future runoff events are not known, they will have the same statistical properties as historical records. Stochastic methods tend to deal with cycles of events and the generation of possible future flows. It is assumed that the statistical properties of runoff do not change with time. Two approaches are used; either data are aggregated from monthly data and combined to give annual results or seasonal/annual data are disaggregated to provide monthly flow values.

The general form of the stochastic modelling equation on a monthly basis is:

Xi + 1 = Xj + 1 + Bj ( X1 - Xj ) + Ti Sj + 1 (1- Rj2)0.5 where

Xi + 1, Xi = generated flows during the (i + 1)th and ith months
Xj + 1, Xj = mean flows during the (j + 1)th and jth months
Bj = least squares regression coefficient based on Bj = RjSj + 1/ Sj
Ti = normal random variate with mean zero and variance one
Sj +1, Sj = standard deviations of flow during the (j+1)th and jth months
Rj = correlation coefficient between jt and (j+1)th months

Although stochastic analysis is especially widely used in generating long sets of records from shorter periods, the amount of information needed for such complex analysis is nevertheless considerable. Annual flow volumes, or even monthly results are often of little value to water harvesting and agrohydrology in arid and semi-arid areas, though stochastic analysis is often used when reservoir inflows are under study.

8.2 Non-statistical analysis of agrohydrological data

8.2.1 Runoff Analysis

The Runoff hydrograph

In humid regions, runoff is composed of contributions from groundwater and, when it rains heavily, surface flow. Groundwater enters the stream channel when the water table adjacent to a stream is at a higher elevation than the surface of the stream and a slow, but continual seepage occurs. In arid and semi-arid areas, where the ground water table is usually very deep, the opportunity for seepage is rare and stream flow is usually the concentration of runoff coming directly from the land surface. In some cases however, even in relatively arid areas, sufficient water from the high stages of river flow may enter riparian deposits and be released slowly from this temporary storage to extend river discharge beyond that supported by direct, surface contributions. In the case of small runoff plots and field sized catchments, where stream channels are not found, runoff is composed only of surface flow.

Where ground water and surface flow are combined, their separation by use of the flow hydrograph is a routine, but important step in analysis. Where there is no groundwater contribution, the whole flow volume is attributed to surface flow. A great deal of research has been undertaken to define different hydrograph types and what they represent in terms of the runoff process, rainfall, the variability of source areas, etc. Figure 8.5 shows a simple diagrammatic representation of a runoff event, the flow hydrograph, and defines its components. Alternative methods of separating ground water and surface flow are shown, though to some extent each is arbitrary. If the hydrograph is plotted with a logarithmic y axis, the curve usually breaks into three sections, each section represented by a straight line component of the whole. Conventionally, these are regarded as the contributions from groundwater, interflow (sub-surface but not deep) and surface flow, though this probably is a rather simplified view of true conditions.


Figure 8.5: The Flow Hydrograph and Its Components

Unit Hydrographs

Despite the variation in runoff due to the complexity of catchment characteristics and rainfall amounts, intensities and distributions, the Unit Hydrograph method of synthesising runoff hydrographs for particular rainfall amounts is widely used. By constructing a basic runoff hydrograph from a known and conveniently selected rainfall amount, runoff hydrographs can be synthesised for any rainfall amount. The method is not appropriate for small runoff plots or very small catchments less than about 2 hectares in area.

The unit hydrograph is defined as the surface runoff hydrograph generated from a unit depth of rainfall distributed over the catchment area, occurring during a specified period of time. The most suitable period of time will depend on the typical flow duration and size of catchment. The unit hydrograph is best obtained from a storm of reasonably uniform intensity (and therefore usually of short duration), desired duration and large volume. A single storm peak is preferred, but is not essential. The unit hydrograph ceases to be applicable when the catchment area is so large that it is not covered by a single storm, and in such circumstances the catchment must be divided into subcatchments that are treated separately.

Usually several unit hydrographs are obtained from a number of storms and combined, to "average" the effect of different rainfall patterns.

Figures 8.6 and 8.7 illustrate the method of unit hydrograph separation from a single storm. The first step is to separate the base flow component from direct (surface) runoff. The volume of direct runoff is calculated, for example in m³ and then converted into runoff depth, i.e. the depth of runoff if the volume were to be spread evenly over the catchment, for example in cm. The ordinates of the direct flow hydrograph are divided by this runoff depth, which then defines the unit hydrograph for 1 cm of direct runoff.

The average unit hydrograph, which gives a more generalised hydrograph shape and size, is drawn as an interpolation to conform to the overall shapes of several individual unit hydrographs, using the average peak flow and average time to peak as guides to the outline of the graph. It should not be surprising that the unit hydrograph may not show precise linearity when used to construct flows for a wide range of volumes, since the recession of flow depends, to some extent, on peak flow. However, estimates of flow can be good, and the unit hydrograph can be regarded as the typical hydrographic form of a particular catchment. Figure 8.8 shows the use of the unit hydrograph in construction of the hydrograph from a complex rainstorm.


Figures 8.6 and 8.7 Derivation of a Unit Hydrograph

Conversions to other unit hydrograph duration periods can be made, where these are integral multiples of the basic unit hydrograph, by simple superimposition of a series of the unit hydrographs.

Where time periods are not whole multiples of the unit hydrograph time period, S-hydrograph techniques can be used. These techniques provide a flexible method of obtaining a wide range of hydrographs for different durations, but they are complex and time-consuming. Although they are in widespread use and are important in the understanding of orthodox river behaviour, they are marginal to agrohydrological research and water harvesting applications. The selection of a unit hydrograph that will allow a convenient permutation of durations is not usually difficult to achieve. Most textbooks on hydrology cover the subject in detail. An Instantaneous Unit Hydrograph, which is independent of duration, can also be derived.

Various empirical formulae have been developed to enable a Synthetic Unit Hydrograph to be defined. These have related catchment characteristics to hydrograph form, but the application of these empirical formulae to define a synthetic unit hydrograph tends to be of limited value. If unit hydrographs are desired for ungauged streams, it is preferable to obtain them by the use of flow data from adjacent catchments. The theoretical bases of these variations of unit hydrograph theory may be addressed by reference to a standard hydrological textbook such as the Handbook of Applied Hydrology; see chapter 1.

A flow hydrograph provides two especially important values: the runoff volume and the peak flow, but in most cases interest will also be taken in the duration of the rising and falling limbs; the time after rainfall starts that runoff

begins and the overall duration of flow. For the practice of water harvesting these characteristics of flow have practical implications and are important in understanding when peak flows have passed; when flow diversions should be made; whether a recession flow will continue or rapidly decline, whether all flow is direct or whether sub-surface contributions can be relied upon to deliver further opportunities to supplement the water availability to crops.


Figure 8.8: Construction of a Flow Hydrograph using the Unit Hydrograph

The construction of water harvesting and field systems assumes that sufficient quantities of runoff water will be available to be moved safely to crops, using well designed structures that can control and manipulate specified peak flows. The study of runoff volume data in agrohydrology anticipates these activities by trying to answer two main questions. The first and most important is, will the amount of runoff that can be collected be agriculturally useful? The agricultural dimension is very important and "useful" will be determined by geographical location and agricultural practice; crop type and climate. However, the clear presentation of simple hydrological analyses will allow this question to be answered according to the conditions that prevail in a particular locality.

The second question, which factors most control the production of runoff and how? is concerned with an understanding of the runoff process. The quantification of the relationships between the components of these physical processes is important, so that predictions of future behaviour can be made. The analysis of rainfall and catchment information is essential to the study of runoff volumes.

8.2.2 Rainfall Analysis

The analysis of rainfall information may use the statistical distributions and probability procedures that have been discussed previously in this chapter. Regression analysis is important in understanding how rainfall amounts and intensities affect the production of runoff. These characteristics of rainfall may also be used in multiple regression in association with other influential factors. In addition, treatment of basic rainfall data to render them useful for a variety of applications is an important stage of analysis. The analysis of rainfall information concerns two types of data; those obtained from a single point and those extrapolated or interpolated from a number of points to estimate rainfall over an area. These data may involve a single event or may be directed at the study of how precipitation changes with time.

General Characteristics of Rainfall

Arid and semi arid regions, where water harvesting is likely to be practiced, tend to experience great spatial variability of rainfall; rainfall intensities also tend to be highest during the first half of a storm. It has been found that the characteristics of rainfall in these areas is essentially independent of locality and may be similar in widely different geographical regions. The standard deviation from the long term mean in such areas is very high, the average annual rainfall varying between by perhaps 35% to 200% of the mean, compared to temperate climates where standard deviations are more typically only 10 - 20%.

Missing Data and the Adjustment of Records

Individual Data Points

Missing point rainfall data are not uncommon and may be estimated in three ways, each of which use information that is taken from stations close to that which has no record. The three alternative methods of estimation are:

Averaging the same daily values from three adjacent stations; Estimating from an isohyetal reconstruction of the missing day's rainfall Using a weighted ratio.

In the first case, the simple arithmetic average may not provide an accurate result where, for example, topographic variation is great and where rainfall values are influenced by such variation. However, if the annual rainfall totals of the adjacent stations are within 10 % of the station with missing data, this method is regarded as suitable.

An isohyetal estimate of a station's missing rainfall is easy to achieve where an adequate density of stations allows it and where the spatial variability of rainfall is low. Knowledge of local conditions can provide added accuracy to such estimates and the technique of isohyetal construction is described below.

In some areas, the spatial variability of rainfall and/or an insufficient number of rainfall stations may preclude the production of an isohyetal map, in which case the latter of the three options is to be preferred. The missing value may be obtained from an equation linking the rainfall experienced by adjacent stations:

Dm (missing daily rainfall) = 0.333 [D1 (Anm/ An1) + D2 (Anm/ An2) + D3 (Anm/An3)] (8.28)

where:

D is the daily rainfall
An is the annual total rainfall
the subscripts 1, 2, and 3 refer to the adjacent stations and m to the station with missing daily data.

Adjustment of records

The change in location or exposure of a raingauge can effect long-term records. The Double Mass Curve technique may be used to correct such defective records, by the comparison of data from the queried station with those from adjacent gauges. Two important considerations should be taken into account; that records from at least ten stations are needed by which to make a comparison; and the longer and more homogeneous the records of these stations, the more successful the correction.

In the case of example Figure 8.9, the values of the queried station must be corrected downwards by a factor of 1357/1535 = 0.89, after 1987.

Maximum accuracy is gained by the comparison of double mass data within the base stations' dataset and the elimination from the set of any that show large changes of slope. Minor differences can be ignored, or more than one parallel slope line can be used to account for true differences. The method is not recommended for daily or individual storm values.

This method is frequently used in the same way to compare and correct the flow records of stream gauging stations.

Accumulated Annual Rainfall of Queried Station (mm)


Figure 8.9: Double Mass Comparison of Rainfall

Rainfall Depth Over Areas from Point Measurements

The point measurement of rainfall at a site is usually adequate to describe rainfall on an area basis when small catchments and runoff plots are used, though the high spatial variability of some locations can be extreme. As catchment area increases, especially in those regions with a large inherent spatial variation of rainfall, it becomes increasingly important to convert rainfall data collected at several points into a rainfall depth over the area.

Several methods can be used to extend rainfall depths at points to areas. The arithmetic mean of several stations can be used, but this usually encounters limitations of spatial distribution and does not weight any variation in network densities. The Thiessen Polygon and isohyetal methods are the two commonly used alternatives to the arithmetic mean and are described below. The Thiessen Polygon method assumes that values of rainfall amounts at a point can be extended half way to the next station. Polygons are constructed around each station and the area of each polygon is then used to weight the rainfall value at the centre of each . The geometrical construction of this method is illustrated in Figure 8.10 as is the area-based weighting applied to the value of rainfall at each station.


Figure 8.10: Thiesson Polygon Method to Calculate Rainfall Over an Area (Daily Rainfall)

One disadvantage of this method is that the polygons must be re drawn each time data from a station is missing, or if a station is removed from the network. Figure 8.11 illustrates the method of isohyetal construction.

The Isohyetal Method uses station data to construct isohyets of equal rainfall. Interpolations between station values are made according to knowledge of topography and climatic regime. The average rainfall depth is then calculated by adding the incremental volume between adjacent pairs of isohyets weighted by area as in Figure 8.11. This method has the advantage that calculations are made according to knowledge of the climate and topography of a catchment. Unlike the Thiesson method which relies solely on geometric construction, variation in rainfall amount will reflect changes in altitude and proximity to other meteorological influences.


Figure 8.11: Isoheyetal Method

Depth -Area -Duration Analysis

DAD analysis is used to determine the greatest precipitation over different areas and for different durations. Regional and seasonal comparisons can be made, though the analysis is only applied to storms that are expected to approach maximum values. Several centres within a storm may have the greatest DAD values and these can be compared to identify the greatest for various selected areas and durations. The procedure is often used, but is relatively complex and will be of limited value in many water harvesting situations unless large catchments are under study. The procedure is therefore outlined briefly. Depth-area-duration data may be available from local Meteorological services.

Conversion of the 24 hour ("daily") rainfall to a time period more suitable to hydrological purposes is needed, often the 6 hour rainfall amount, so that aggregation into various time periods of, for example, 6, 12, 18, 24 hours, etc., is possible. Comprehensive recorded rainfall data for the storm are needed. Mass curves of rainfall are drawn (accumulated rainfall versus time). Several total storm depth are curves are then prepared by dividing the total storm rainfall map into major centres, where there are more than one, and determining the total size of the area and the average rainfall depth within the area which is enclosed by successive isohyets around each centre.

A time breakdown is then needed for these depth-area curves by weighting the rainfall of each station by the ratio of its Thiessen area within the isohyets, to the total area within the isohyets. A time distribution of the total storm rainfall within each isohyet is found. Values are plotted as area versus total rainfall, with different lines for different durations. Smaller areas receive largest rainfall amounts for a given duration, and rainfall amounts generally increase with duration, but regional differences can be very great.

Areal Reduction Factor (ARF)

Rainfall depth tends to decrease as area increases. In many instances, this reduction cannot be assessed by the use of a comprehensive network of raingauges, because such a network does not exist. A simple factor that can be applied to point rainfall for a specified duration and return period, and which converts point rainfall to areal rainfall is called an areal reduction factor (ARF). Research has shown that such a factor, for a specified area and duration, does not vary greatly with return period and this aspect can be ignored for practical purposes.

A simple ARF can be obtained by selecting the maximum areal rainfall event for a catchment. The rainfalls (R1) at each station are noted. In some cases these will be the same as the maximum annual rainfall event (R2 for a specified or desired duration, D), but in others they will not. The ratios of R1/R2 are noted and mapped for each year and an areal mean of R1/R2 is found. The mean of a number of years' means is then calculated and this gives the ARF for a known area A and duration D.

In many cases, comprehensive rainfall records will not be available to allow areal estimates of the mean ARF by mapping, but the method is adaptable to the circumstances most commonly found in developing countries: medium term (> 10 years) daily rainfall records from a limited number of stations. In general it is better to analyse local data in a simple way than import relations obtained from other world regions. A straightforward adaptation of the method described above, to obtain an ARF, is illustrated below by an example. The data originally used for the adaptation came from Belize, in Central America.

The 1 day rainfall with a 5 year return period was selected as desirable for the following reasons: the 1 day rainfall is commonly available, it is suitable for use since the 1 day and runoff-producing storm rainfall are often very similar, and they can be regressed and their precise relation determined; the 5 year return period is one widely applicable to farm practice, though the 10 year return could also be justified. Eleven rainfall stations (maximum station data 22 years) were used in the analysis, over an area of approximately 2000 km².

According to their long term average annual rainfall, the rainfall stations were divided into three groups, each located within one of three AAR bands of 1400 - 2000 mm, 2000 - 2800 mm and > 2800 mm per year. This information was obtained from isohyetal maps. The R1 and R2 values were listed for each station and year, and the ratios of R1/R2 were obtained. The arithmetical average of the ratios was then calculated for each station. The overall average ARF for each rainfall band was then obtained from the stations within it, by weighting according to the length of station records.

The area-reduced, 1 day 5 year point rainfall for each station was then found by listing daily rainfall in an annual maximum series and applying the appropriate band ARF. These then gave an average, area reduced value of the 1 day, 5 year rainfall for the catchments under study, the contribution from each station being weighted according to area as defined by the Thiesson method. Where a catchment was composed of areas lying within different isohyetal bands, the different ARFs were applied, also by weighting in proportion to area.

Rainfall Depth and Intensity - Frequency Relations

Where sufficient records exist, the relations between rainfall intensity or total rainfall amount, duration and return period are often established, for use in practical applications concerning runoff, such as the theoretical calculation of peak flow values. One of two methods are usually applied, but reasonably long-term intensity gauge records are needed for both.


Figures 8.12 (a) and (b) Rainfall Intensity-Duration and Depth-Duration Frequency Graphs

In the first method a particular time interval is selected for convenience; the 5, 10 or 15 minute interval is often used, and the maximum depths and mean intensities are calculated for this time interval. Generally, mean intensities will be highly variable within the time period and will become less hydrologically significant as the time period increases. In the second method, the highest rainfall intensity for any duration is the criterion for interpretation. Rainfall depth by segments of the intensity trace are classified to determine hydrological response. Typical relations for return period intensity and duration for a particular station are illustrated in Figures 8.12 (a) and (b). Locations with similar relations may be mapped where there is sufficient information.

Theoretical Distributions

Rainfall data may be analysed according to the methods which apply theoretical statistical (usually Extreme Value) distributions to the data, as discussed in section 8.1 using runoff data. Return periods and probabilities for different rainfall amounts can be obtained. The Pearson Type III distribution (with varying skewness coefficients) has been found to fit semi-arid data to a satisfactory extent in many cases. The validity of different distributions may be compared or tested using the Chi-square test, described in section 8.1.

Unlike temperate climates which usually have a normal distribution, rainfall distributions in semi arid regions tend to be positively skewed, that is they have many more rainfalls smaller than the mean, than greater than the mean. The mode is less than the value exceeded in 50% of years (the median) and the mode and median are less than the mean. The mean is of limited use in describing the central tendency of data in these areas and the mode is often used as the statistic to describe the most frequently experienced conditions best. Associated with the skewness of such data is the difficulty in obtaining estimates of mean rainfall without large errors, especially from short records.

The annual maximum series and partial duration series are also used to evaluate extreme rainfalls and their probabilities by ranking and the imposition of selected thresholds.

Alternatively, the probability of daily rainfall amounts may be studied by grouping rainfall according to amount, for example rainfalls greater than 1 mm, 5 mm, 100 mm etc. and calculating their probabilities. Table 8.8 below gives an example.


Table 8.8: Probability Distribution of Daily Rainfall Occurrences

Bi-modal distributions of rainfall (two main peaks in the rainfall season) are also often evident, due to different meteorological conditions that prevail. In these circumstances it may be necessary to study rainfall statistics on a short period basis, so that the parent population of rainfall events accounts for this bi-modality.

8.2.3 Evaporation and Evapotranspiration

The use of evaporation and evapotranspiration data is important to runoff and agricultural studies. The rate at which runoff is produced will often depend on the existing wetness of the soil, which in turn is highly inversely correlated to evaporation and evapotranspiration losses. Plant growth is intimately bound to soil moisture availability and the stage and rate of crop growth will be affected as well as be influenced by evapotranspiration

8.2.3.1 Free Surface Evaporation

Evaporation is most commonly measured for agricultural purposes using evaporation pans (chapter 4) and the US.

Weather Bureau 'A' pan is now the world international reference pan type. Although other pan designs are met with, most organisations use 'A' pans and methods of estimating evaporation in the following section concern this type of pan.

The evaporation (E0) from a free surface is impossible to measure when surface areas are very large; however coefficients which quantify the relations between evaporation losses from 'A' pan data and water bodies have been arrived at. These vary according to location and seasonal and short-term changes in weather. They can be used to obtain evaporation estimates from streams and other water bodies. For example if 'A' pan losses are 2 mm day-1 and the evaporation coefficient is 0.70, the daily free surface evaporation from a distribution channel 2 km long and 5 m wide will be:

fdse = E0 = 2000 × 5 × 0.002 × 0.70 = 14 m³ day-1

Losses from seepage, percolation etc., are not accounted for. In a similar way, reservoir storage losses can be estimated, usually by using monthly values of the pan coefficient.

The measurement of climatic variables is essential in establishing pan coefficients, and in most countries some information on these coefficients will be available. Temperature, wind speed and duration, barometric pressure and the nature of the container all effect evaporation measurements to various degrees. Table 8.9 gives some examples of long-term annual 'A' pan coefficients and examples of seasonal variation at Khartoum (Sudan) and Lake Hefner (USA).


Table 8.9: Example 'A' Pan Coefficients

Empirical Formulae

Many empirical formulae have been developed to estimate evaporation, based on Dalton's Law; the basic driving mechanism of these equations is the difference in vapour pressure between water and the atmosphere. The greatest problem with empirical formulae is the difficulty in measuring the components of the equations in a manner that can be related realistically to the dynamic processes that lead to evaporation.

Dalton's equation is E = C ( es - ed) where (8.29)

E is the rate of evaporation
C is a constant
es is saturated vapour pressure at the temperature of the water surface in mm of Hg.
ed is actual vapour pressure of the air (es × relative humidity) in mm of Hg.

The constant C in equation 8.29 has been described as:

C = (0.44 +0.073 W) (1.465 -0.00073p) where (8.30)

W is wind speed in km hr-1 at 0.15 m
p is atmospheric pressure in mm of Hg at 0°C.
E is in mm day-1 and reservoir evaporation = E × 0.77

Alternatively, the constant C for shallow ponds and evaporation pans has been evaluated as:

C = 15 + 0.93 W (8.31)

and for small lakes and reservoirs as:

C = 11 + 0.68W (8.32)

with W as for equation 8.30.

The Water Balance method of estimating evaporation from reservoirs, compares changes in storage with a balance of known in flow and out flow. This method appears to be simple, but seepage losses are extremely difficult to calculate or measure accurately. Precipitation onto the reservoir can also be a complicating factor. Energy Balance methods are analogous to water in flow and out flow, balancing all the energy components in the evaporation process. However, although they have been tested more rigorously in recent years, difficulties still remain with instrumentation and their use is not widespread.

Evaporation from Soil Surfaces

The evaporation of water from soil surfaces is a more complex process than that from a free water surface. Although during periods of saturation, these processes may be very similar, saturated conditions rarely last for long and the evaporation rate drops rapidly as soil moisture levels decrease; evaporation from soils at less than field capacity may even be regarded as generally unimportant. Soil evaporation losses are defined by free energy and the free energy required by plants to take water at wilting point is less than 0. 1% more than at saturation. In addition, evaporation from soils depends on the nature of the soil; its texture, chemistry, organic content, vegetation cover and depth. The losses from soils and vegetation are usually combined and treated as one process, termed evapotranspiration, though in regions where vegetation cover approaches zero or plants become dormant at certain seasons of the year, it may be desirable to monitor evaporative losses of soil moisture.

8.2.3.2 Evapotranspiration

Transpiration is the process whereby plants lose water from their leaf stomata and is essentially the same as evaporation, though not from a free water surface. Losses are proportional to the diameter of the stomata, but not their area, as is true for a perforated membrane. The rate of transpiration is essentially governed by the difference between vapour pressure under the stomata and that of the atmosphere, the number of stomata per unit area being variable with species and climatic conditions. Evapotranspiration (Et), sometimes called 'consumptive use', is the evaporation of water from all sources combined. The term 'potential evapotranspiration' (Ep) defines conditions where water availability is in no way limiting. 'Actual evapotranspiration' (Ea) attempts to define realistic conditions, whereby rates fluctuate according to the availability of water and changes in climatic conditions.

Evaporation Pans

Because the soil water availability conditions that allow potential evapotranspiration are not often met with, empirical studies have attempted to relate 'A' pan evaporation values to actual evapotranspiration values. Such relations should be used with caution, though most experimental data show them in the form:

Crop (Actual) Evapotranspiration Ea = k E0 (Pan Evaporation.) (8.33)

where k is a coefficient according to crop and stage of growing season.

Table 8.10 gives values of k for a variety of crops during different stages of growth. Values take some account of incomplete shading, but crop density, soil variability, wind profile etc., can make significant differences and values should be only regarded as a guide.


Table 8.10: Coefficient k to be Multiplied by 'A' Pan Evaporation to give Actual Evapotranspiration

Empirical Formulae

A wide range of empirical formulae have been developed to calculate evapotranspiration. The most commonly used are discussed below.

a. Blaney-Criddle Equation

This relatively simple equation estimates consumptive use when water availability is not a limiting factor. Like many of the empirical equations for evapotranspiration it is most suitable for conditions immediately after rainfall, irrigated conditions or as a climatic descriptor:

Monthly Ep (in inches) = kF where (8.34)
F=(t × p)/100
k is the annual, seasonal or monthly consumptive use coefficient (for different crops)
p is the monthly % of daytime hours of the year, occurring during the period
t is the mean monthly temperature in °F

The greatest difficulty in applying the equation is the determination of the crop factor 'k', which varies not only with crop type, but also with climate and growing season. Examples are given below in Table 8.11, which relate to specified crop growing season lengths.


Table 8.11: Crop Coefficients (k) for the Blaney- Criddle Equation

Table 8.12 gives daytime hours percentages for various latitudes.


Table 8.12: Daytime Hours Percentages ( 100 p) for the Blaney-Criddle Equation

Thornthwaite Equation

This equation is based on an exponential relation between monthly mean temperature and mean monthly consumptive use, based on experience gained in the central and eastern states of the USA. It is widely-applied, but tends to be less satisfactory in regions that undergo frequent short-term changes in temperature and relative humidity.

Monthly Ep in mm = 16(10t/I)a where (8.35)

t is the mean montthly temperature °C

I is a temperature efficiency index (and is equal to the sum of 12 monthly values of the heat index 'i' which= (t/5)

1514 for each month of the year and where t is the mean monthly temperature in ° C).

a is a cubic function of the annual heat index ' I ', which can be obtained from Table 8.13 or from



a = 6.75 × 10-7 I3 - 7.71 × 10-5 I2 + 1.792 × 10-2 I + 0.49239 (8.36)


Table 8.13: Values of 'a' in the Thornthwaite Equation

It is necessary to adjust the calculated rates of evapotranspiration, because monthly durations are not equal and the number of hours of evaporation in a day will vary with latitude and season. The adjustment factors are given in Table 8.14.


Table 8.14: Adjustment Factors for Thornthwaite Values of Ep

Penman Equation

Penman's equation is the most complete theoretical approach to estimating potential evapotranspiration. The collection of data for many of the meteorological variables described in chapter 4 is directly attributable to the application of this equation. The equation is probably the most widely used empirical formula and shows that the consumptive use of water is inseparable from the level of incoming solar energy. In effect, the Penman equation is a combination of a measure of the drying power of the air and an estimate of available net radiation.

Penman's Equation is of the general form:

Ep in mm day -1 = [D / d (Rn) + Ea]/[(De/d) + 1] where (8.37)

D is the slope of the saturated vapour pressure curve / temperature curve at mean air temperature in mm Hg °C-1

Rn is the net solar radiation [Ri (1 - r) - Rb] where

Ri is the radiation reaching the earth's surface in cals cm-2 min-1 and is = Rs (a + bn/ N) where

Rs is incoming radiation in terms of mm of water evaporated day-1, 'a' and 'b' are latitude constants

N is the maximum possible duration of bright sunshine in hours, at the location

r is the reflectance (albedo) of the surface which is a ratio of reflected radiation / incident radiation

Rb is the long wave back radiation

d is the psychometric constant, 0.49 mm Hg ° C-1 or 0.66 mb ° C-1

Ea is the mass transfer (Dalton's equation component) and is = f (u) (es - e) where

f(u) is a function of wind speed in m s-1

es is the saturated vapour pressure at air temperature at the evaporating surface in millibars

e is the vapour pressure of the atmosphere above in millibars

Penman's equation has been modified for various conditions and locations of latitude, with different values for the various numerical factors included in the equation. The example of Penman's modified equation below is given with constants according to a geographical location in the central USA, with constants and variables for °C, potential evapotranspiration is in mm day-1.

Ep = D / D + d[Rs (1-r) (0.22+0.55( n/N))0 - D /D + d [d Ta4 (0.56-0.091ed1/2) (0.10+0.90 (n/N))] + d/D + d[(0.175+0.0035 u) (ea- ed)] (8.38)

(This equation is in wide use elsewhere, with only minor modifications of the numerical constants given above.) where,

Ep, d, D, r and N are as above in equation 8.37

n is the actual number of sunshine hours recorded

Rs is radiation at the top of the atmosphere and relates to time of year and latitude

d is the Stefan-Boltzman constant, 2.01 × 10 9

Ta4 is black body radiation at mean air temperature (= mean daily air temperature + 273 °K in mm of evaporation)

u is the wind run at 2m in km day-1

ea is saturation vapour pressure at mean daily temperature (millibars)

ed is mean vapour pressure (millibars)

Table 8.15 Coefficients a and b

Latitude (° N and S)

a

b

54

0.21

0.55

36

0.23

0.53

24

0.28

0.49

13

0.26

0.50

3

0.26

0.44

Table 8.16 Typical Albedo Rates

Surface

Albedo

Close growing crops

0.15-0.25

Bare land surfaces

0.05 - 0.45

Forest

0.15

Water

0.05

(The Tables of the Smithsonian Institute provide a comprehensive guide to albido rates)

Care should be taken that variables and constants are those appropriate for measurements in ° C or ° F when applying versions of the equation that may have been developed locally. Tabulated values for the components of Penman's equation are available to assist calculation and are given below. They are used with an illustrative example of Penman calculations.


Table 8.17: Values of d Ta4 for Different Temperatures


Table 8.18: Values of D/D + d and d/D + d for Different Temperatures

Table 8.19 gives values of Rs for different latitudes.

Table 8.20 gives values of possible sunshine hours for different months and latitudes.

Figure 8.13 gives saturation vapour pressure for temperature values °C and °K.


Table 8.19: Values of Rs for Different Latitudes in mm day-1

Mid Monthly Radiation on a Horizontal Surface in mm of Water Day-1 Evaporated


Table 8.20: Mean Possible Sunshine, N

To calculate maximum duration of sunlight for any month multiply 12 × 30 × coefficient

Local variations of Penman's equation are often developed after a comprehensive study of local climatic variables. The example below is for the semi arid and arid conditions found in southern Africa, the value of x being 0.5 for open surface evaporation and 1.0 for evapotranspiration.

Ep = D /D +d[Rs(1-r) (0.25 + 0.50 ( n/N)] - D /D +d[d Ta4 (0.32 - 0.42ed) (0.30 + 0.70 (n/N))] + d/D + d[(x + 0.0062 u) (ea - ed)] (8.39)

In this equation it is interesting to note that the term (0.32-0.42ed1/2) is always negative, thereby adding to the calculated evapotranspiration obtained from the overall formula.

Penman's equation was developed for estimates of Ep from short grass under humid conditions. These conditions are frequently not met with, though they may be simulated by the presence of wet soil after rain, shallow water bodies or irrigated conditions.

In arid and semi arid areas, localised variations in crop cover may lead to areas of crops being surrounded by hotter, drier conditions, and is sometimes referred to as the "oasis effect". In these circumstances more energy is available for evapotranspiration than indicated by measured incoming solar radiation, leading to increased and widely different localised rates of evapotranspiration.


Figure 8.13: Saturation Vapour Pressure with Temperature

A worked example of a Penman calculation is given below:

Worked Example

Substituting into equation 8.38:

Ep = 0.753 [14.6 (0.75) (0.22 + 0.55 (0.65))] - 0.753 [16.34 (0.56 - 0.09 ( 4.77)) (0.10 + 0.90 (0.65)]+ 0.247 [(0.175 + 0.0035 (125) (35.0 - 22.75)] = 0.753 [6.324] - 0.753 [1.463] + 0.247 [5.534] = 5.03 mm day-1

In addition to the difficulty of equating consumptive use to actual crop use, a difficulty shared with all empirical ED equations, one particular problem with Penman's equation is the requirement of data for a wide range of measured variables. Some work has been undertaken to relate, through regression analysis, the major components of Penman's equation with values of Ep calculated from the full equation. For example Rs which is tabulated and u the windrun which is easily measured, versus Ep could be used. Ten day mean values of evapotranspiration are often used rather than daily values and Ep does not vary greatly between locations with similar prevailing weather conditions.


Table

In basin resources studies the problems of localised variation is not great, but in cases where localised values of actual evapotranspiration are needed, the situation is often more difficult. For example the response by plants to reduced water availability may lower actual rates of evapotranspiration from more than 60% of potential to 10% or less. The problems involved in the application of empirical formulae to estimate actual water losses by evapotranspiration go a long way to explain the continued measurement of these losses by lysimetry and other field methods.

The FAO is currently undertaking a review of methods for the calculation of evapotranspiration and is expected soon to pronounce on the methods it regards as most appropriate.

8.2.4 Sedimentation Data Analysis

The most common analysis of sedimentation data is a regression relation against runoff, often called a sediment-rating curve. The sedimentation factor may be sedimentation concentration, Co or sediment load, Qs. Such a relation may be applied to long term flow records and produces a sediment-duration curve. The best fits are given by plotting log 10 discharge against log 10 sedimentation and the best correlations are obtained by using load rather than concentration as the sediment factor. The form of the equation is:

Qs-aQb

thus log Qs = b (log Q) + log a (8.33)

The coefficient a has no particular range of values but is related to width of the channel in the form, the sediment per unit width of channel given by qs = a Qb /W.

The coefficient b however, has a lower limit, where b = 1 the concentration is constant and independent of Q, where b > 1 the concentration increases with Q. In practice, b is often from 1.4 to 2.8. The scatter of data is often wide, perhaps representing an order of magnitude at the 90% confidence limits. This may be due to random errors of sampling, laboratory procedure or the fact than the quantity of sediment in transport is related to many other physical variables than discharge alone. Figure 8.14 shows a graph of log Qs against log discharge Q.

In such a case, the regression line of the log values is often ignored, or rather replaced by a second regression line. This second regression is positioned to pass through the means of the actual values of Qs and Q, rather than the means of the log values. The line is drawn parallel to the original line, but now passes through the arithmetic values of the means, and through the mean Qs - mean Q intercept. This produces a weighted estimate of Qs for given Q and tends to reduce the large errors otherwise found.

Sample sediment concentration may be plotted in a similar way, with concentration x discharge giving the data points for Figure 8.14. It should be remembered that sediment estimates based on such relations are subject to errors in the relation and the uncertainty as to whether the relation can be applied to other periods of measurement.


Figure 8.14: Sediment Discharge against River Flow

Extrapolation

The extrapolation of such relations is difficult, because the value of the coefficient, b, falls with higher discharge levels and the relation is in reality non-linear on log-log plotting. Studies have shown that the sediment carrying capacity of discharge decreases with increased discharge after the concentration has reached 100 kg m-3. The physical basis of this observation is not clear. Methods of extrapolation are available according to absolute rates of sedimentation, but these may be extremely involved, using combinations of water level and grain size groups, a range of water level/discharge relations, a topographic survey of a test reach of the river, water temperature etc.

The FAO paper 37, "Arid Zone Hydrology" gives full details of the procedure (the Einstein method) which is not relevant to the scope of this handbook. It is however, the only method recommended by the FAO for the theoretical determination of sediment load and the augmentation of empirical information. In the same chapter, this paper also outlines other methods that can be used to estimate the silting of dams by sedimentary deposition.

First approximations of total sedimentation according to volume of discharge have been developed as regression for semi arid and arid regions, and while these cannot be used for design purposes, they have proven useful in generalising the order of sediment load, see Figure 8.15.

The development of regression relations is likely, as for other such relations, to be limited by geographical locality, soils, hydrological and climatic regimes etc., and the translocation of formulae from one area or even stream to another is likely to be unsuitable for any purpose other than approximate estimation.

In Figure 8.15, the form of the regression is Vs = aVb, the coefficient values are:


Figure 8.15: Sediment from Short Grassland and Scrub

Appendix E: Data analysis

Appendix E1: Critical values of the chi-square distribution


Figure


Figure