(introduction...)

Matrices relate activities to environmental components so that the box at each intersection can be used to indicate a possible impact. The term "matrix'' does not have any mathematical implication, but is merely a style of presentation.

The matrix can be used to identify impacts by systematically checking each development activity against each environmental component. If it was thought that a particular development activity was to affect an environmental component, a mark is placed in the cell at the intersection of the activity and the environmental component. A matrix analysis can systematically identify potentially important effects demanding more careful attention or analysis or focus attention on important possible effects that might otherwise be overlooked. Matrix is thus an extension of the basic checklist.

There are three types of commonly used matrices:

• descriptive matrices;
• symbolic and presentation matrices;
• scaled/weighted or numeric matrices.

4.3.1 Descriptive matrix

In descriptive matrices, short written descriptions are used. For each phase of a project, that is, site selection, construction, operation, and closure (if applicable), the various impacts associated with each activity are defined and short descriptions are provided. There is no scaling or quantification of these impacts. Table 4.1 presents a structure of the descriptive matrix for a quarry.

4.3.2 Symbolized matrix

A most noteworthy matrix presentation essentially improves the communication between the impact analysts, decision makers, and the public. In the symbolized matrices, symbols are used to capture the understanding of the impacts.

Environmental impacts may be described by words such as "important'' or "significant''. These subjective, qualitative words are difficult to deal with because their interpretation depend on cultural values and specific circumstances. Even when quantitative data are available, they must be gauged against some standard and often there is none or at least none widely accepted. There are, however, some useful guides for ranking impacts or assessing impact assessment.

There are several factors that must be taken into account when assessing the significance of an environmental impact arising from a project. The factors are interrelated and must not be considered in isolation. For a particular impact some factors may carry more weight than others, but it is the combination of all the factors that determines significance.

An example is to use abbreviations and scales, e.g., S for short term and L for long term or 10 to denote a very high order of the impact and 1 to denote almost negligible impact, etc. In this way, a symbolized matrix becomes a combination of descriptive and numeric scales. There are, however, some useful guides for the grading or classification of impacts. These are listed below:

Table 4.1 Environmental matrix for a quarry

Phase	Development action	Social	Physical	Biological
Planning	Consents, district plan, EIS timetable	Law, regulation, public participation, employment, land values, alternative sites, justification, risks and anxieties, cultural/historical	Location of access road	Water table effects on adjacent land
Engineering design	Design of quarrying plan, resolve environmental factors, evaluate options	Landscape effects	Design of quarry, restoration plan	Design of drainage system, including sediment traps to protect water quality in the river
Construction	Access road, drainage system, site crushing plant, energy supply, traffic discharge, stormwaters, silt, sewage, site staff facilities	Cultural/historical, safety, noise, vibrations, effect on farm animals	Disposal of stripping, stability, nuisance, landform noise, blasting, drilling	Sedimentation, surface water pollution
Operation	Stripping overburden, drilling, blasting, excavation, crushing, loading, traffic, review and adjust environmental measures	Landscape, farm animals, noise/vibration/dust, emissions, safety/risk, staff facilities, working environment	Landscape effect	Progressive restoration plan, sedimentation, surface water pollution
Termination	Remove plant, check stability, replace topsoil, plant ground cover, maintenance	Safety, landscape restoration		Maintain and monitor sediment traps

• Sign of the impact
Positive or negative. This is unfortunately not that simple.

• Magnitude
This is defined as the probable severity of each potential impact. Will the impact be irreversible? If reversible, what will be the rate of recovery or adaptability of an impact area? Will the activity preclude the use of the impacted area for other purposes? The answer to these questions may be difficult and may have to be speculated on a subjective basis. The size often depends on the source release, mitigation measures adopted, if any, and the assimilative capacity of the receiving environment, etc.

• Type of change: reversible or irreversible
Irreversibilities always command attention because they signal a loss of future options. Species extinction, severe soil erosion, and habitat destruction are examples of irreversible changes. Pollution of groundwater is often irreversible because of its slow movement. Urbanization of agricultural land is virtually impossible to undo once the land use trend has begun.

• Prevalence
This is defined as the likely eventual extent of the impact as, for example, the cumulative effect(s) of a number of stream crossings. Each one taken separately might represent a localized impact of small importance and magnitude but a number of such crossings could result in a widespread effect. Coupled with the determination of cumulative effects is the remoteness of an effect from the activity causing it. The deterioration of fish production resulting from access roads could affect subsistence fishing in an area many miles away, and for months or years after the project activity has ceased.

• Duration and frequency
The significance of duration and frequency is reflected in the following questions. Will the activity be long term, short term, or sporadic? If the activity is intermittent, will it allow for recovery during inactive periods?

• Risk
This is defined as the probability of serious environmental effects. To accurately assess the risk, both the project activity and the area of the environment affected must be well known and understood.

• Importance
This is defined as the value that is attached to an environmental component in its present state. For example, a local community may value a short stretch of river for bathing or a small swamp for hunting. Alternatively, the impacted component may be of regional, provincial or even national importance.

• Mitigatability
Are solutions to problems available? Existing technology may provide a solution to a siting problem expected during construction of an access road, or to bank erosion resulting from a new stream configuration.

• Understanding
For example, if an access road is to cross a stream and the assessor does not know the extent of use of that stream (for fish spawning, fish migration, subsistence fishing, river transport, etc.), then the impact would be classed as unknown. Similarly, the nature of the river crossing (ford, bridge, ferry, or causeway) may not yet have been planned and so the significance of the environmental impact of that crossing is therefore unknown.

The assessment of significance is best done by holding at least two group discussion meetings involving interdisciplinary expertise.

The most frequently used presentation of a comparison of alternatives is a matrix, in which +, 0 and - show how each alternative affects the different environmental aspects. This can be a useful way to provide a quick overview of the differences between the alternatives.

• the impacts of each alternative are evaluated against a reference (usually the existing situation); or
• for each alternative, it is shown how it contributes to the environmental objectives; or
• the impacts of each alternative are compared with the preferred alternative.

+ + and - - give extra possibilities for differentiation.

In each case it is important that the significance of the symbols is properly defined. If necessary, a reference should guide the reader to more ample information (Table 4.2).

Table 4.2 Example of plus/minus matrix on the theme "Why does the proposed activity improve the soil condition?"

	Existing situation	Proposed activity	Process alternative	Environmentally most friendly
Air	0			0
Soil	0	+	+	+
Surface water	0		0	0
Waste	0	+	+	+
Noise	0	-	-	-
Safety	0	(-)	(-)	(-)
Nature	0	0	0	0
Energy	0	-	0	+
Costs	0	-	-	-

Legend: - deterioration compared to the existing situation; + improvement; 0 no difference; (-) insignificant deterioration.

Table 4.3 Numeric matrix lending a quick but factual overview

	Lead concentration in air (μg/m3)	NOx emission (μg/m3)	Noise level at periphery area (dB(A))	Hazardous waste produced tons/year
Alternative 1	0.8	35	55	2674
Alternative 2	0.7	20	50	2350
Background level existing situation	1.0	80	43	-
Predicted background level 2005	1.2	110	45	-
Environmental standard	1.5	100	55	-

4.3.3.1 Simple numeric matrix

Simple numeric matrices are useful to derive facts to assist in showing the degree of impact or help in making a comparision. Table 4.3 presents a simple numeric matrix used for assessing the alternatives at an IEE level.

Numeric, ordinal, and interval scaled evaluations are given by numerical scores. Leopold et al. (1971) use a scale of 1 to 10 to score two impact attributes - significance and importance. Fischer and Davies use one score on a scale of -5 to +5 to indicate both positive and negative degrees of impact. An interscaled impact matrix has been attempted by Ross (discussed under mathematical matrices).

The second option is to lay down separate matrices describing each impact characteristic. For example, one can make a matrix presentation to show only the size attribute impacts, while another shows the nature of the impact, etc. It is also possible to use some order of shading or colour coding to draw attention to some of the critical cells. Against the apparent disadvantage of handling more matrices, the advantage of improved communication, as well as improved focus, lies in such separate presentations. We can call this approach a thematic matrix approach.

To get a quick overview of the results of an EIA, a comparison of alternatives can be presented as a graph, a map, a diagram, or some other kind of picture. Colours or shades of grey can enhance the presentation (Table 4.4).

The advantage of using colours is that they can be combined in a matrix with factual information. Shades of green can show the positive effects, and shades of red, the negative. This gives an indication of which environmental compartments can be expected to contain problems and which alternatives are most promising.

Table 4.4 Matrix using colours

Objectives	Indicator	Route 1	Route 2	Route 3
Limitation of car kilometres	Amount of car kilometres	293.700	297.700	293.700
Limit expansion of infrastructure	Kilometres new motorway	7	10	7
Nature conservation	Dissection of ecological zones	yes	no	Mitigation partly possible

4.3.3.2 Scaled matrices

Weighted-scale matrices typically use a scale of 1 to 10 to score two impact attributes, significance and importance. One of the most popular scaled matrices is the "Leopold matrix'' named after Dr Luna Leopold of the US Geological Survey who developed it in the early 1970s. All development activities are listed across the top of the matrix and all environmental components that might be impacted are listed down the side. A Leopold matrix attempts to assign numerical ratings of magnitude and importance so that the completed matrices for alternative sites or technologies could each be added and compared. In the original Leopold matrix, scores from a 1-10 scale can be assigned to describe the importance and magnitude of individual impacts. Importance refers to the significance of an impact and the magnitude of its scale and extent. Leopold-type matrices are easy to use and are perhaps the most widely employed and successful of all EIA methods. Figure 4.2 shows a portion of a Leopold matrix, as used for the comparision of alternatives.

Another approach is the environmental impact matrix, with and without mitigation (from Biswas and Agarwal, 1992). This is a conventional technique for summarizing environmental impacts utilizing the matrix method. Initially, the predicted impacts are converted into an ordinal scale (ranking) of impact severity, as in the following example.

Severity	Impact score
No impact	0
Negligible	1
Minor (slight or short term)	2
Moderate	3
Major (irreversible or long term)	4
Severe (permanent)	5

A positive sign denotes a beneficial impact, while a negative sign denotes an adverse impact.

Figure 4.2 Site comparison matrix for a quarry

A significant value (weighting) is attached to each environmental component (independent of the predicted levels of impact) based upon some expert or consensual (Delphi) system. Individual impact scores can then be calculated as the product of impact severity and significance. These may be summed by row and/or column to gauge the net impact of the project on a particular environmental component or, conversely, the net effect of a single project activity on the environment as a whole.

In this way, project alternatives can be systematically compared, and possible mitigation measures can be explored. In addition, this method can draw attention to the most significant impacts in the matrix, as revealed by individual cell scores. This procedure can also be used to identify negative impacts on environmental components that surpass a critical threshold. Such instances will have to be addressed through mitigation or project alternatives (Tables 4.5 and 4.6)

Table 4.5 Environmental impact matrix without mitigation

		Impacting actions
Environmental parameters	Importance value	Premining phase				Operational phase							Impact score
		A	B	C	D	E	F	G	H	I	J	K
Air quality	100	-1	-1			-2		-2			-1		-700
Water resources	75		-1				-1			-1			-225
Water quality	100					-1	-2	-1		-1			-500
Noise and vibration	75	-1	-1	-1	+1	-2		-1			-1		-450
Land use	150	-3	-1		+1	-2		-1					-900
Forests and vegetation	150	-4			+1								-450
Wildlife	50	-2			+1	-1					-1		-150
Human settlements	75	-1	+1						+1				+75
Health	100				+1	-3						+1	-100
Infrastructures and support services	50									+2	+1	+2	+250
Employment	50	+1	+1			+2				+1			+250
Places of tourist or archaeological importance	20												0
Total	1000	-1350	-275	-75	+525	-1000	-275	-525	+75	-25	-175	+200	-2900

+ sign shows beneficial impact; - sign shows adverse impact.

A, land acquisition and transformation; B, civil works construction; C, erection of mechanical and mining equipment; D, green belt formation; E, mining operations including CHP; F, disposal of liquid effluent; G, disposal of solid wastes on land for reclamation; H, housing provision: I, provision of water, sewage, electricity, and other civic amenities; J, transportation; K, medical facilities.

Table 4.6 Environmental impact matrix without mitigation

		Impacting actions
Environmental parameters	Importance value	Premining phase				Operational phase								Impact score
		A	B	C	D	E	F	G	H	I	J	K	L
Air quality	100	-1	-1		+1	-1		-1			-1		+1	-300
Water resources	75		-1				-1			-1				-225
Water quality	100						-1	-1					+1	-100
Noise and vibration	75	-1	-1	-1	+1	-1		-1			-1			-375
Land use	150	-3	-1		+1	-1		+1					+2	-150
Forests and vegetation	150	-4			+1								+4	+150
Wildlife	50	-2			+1	-1					-1		+1	-100
Human settlements	75	-1	+1						+1					+75
Health	100				+1	-1						+1		+100
Infrastructures and support services	50									+2	+1	+2		+250
Employment	50	+1	+1			+2				+1				+250
Places of tourist or archaeological importance	25													0
Total	1000	-1350	-275	-75	+625	-375	-175	-125	+75	+75	-175	+200	+1150	-425

+ sign shows beneficial impact: - sign shows adverse impact.

A, land acquisition and transformation: B, civil works construction: C, erection of mechanical and mining equipment; D, green belt formation; E, mining operations including CHP; F, disposal of liquid effluent; G, disposal of solid wastes on land for reclamation; H, housing provision: I, provision of water, sewage, electricity, and other civic amenities; J, transportation: K, medical facilities: L, land reclamation.

4.3.4 The component interaction matrix

The component interaction matrix (CIM) developed by Ross was first used in an EIA of five alternative sites for the transshipment of lumber on the Nanaimo estuary, British Columbia. The uniqueness of the area under consideration prompted an investigation of secondary impacts in an attempt to present the full implications of the project proposals.

In a CIM, the environment is represented by a list of environmental components, arranged along both horizontal and vertical axes. Direct dependencies between the components are identified and marked as "1'' in the appropriate cells. Interdependencies up to the nth order (i.e., all higher order dependencies) can be determined by the use of a matrix powering procedure.

The Canadian CIM above used 21 environmental components, and 120 first-order dependencies were identified. Matrix manipulation (powering) was performed until fifth-order dependencies had been discovered. From the information revealed by the powering process, a minimum link matrix was derived. All cells of the original CIM were used to contain integer values denoting the length (in terms of intervening nodes) of the shortest linkages connecting the two components. A disruption matrix was also formulated in which the impacts of each project alternative on all primary dependencies were scored on an ordinal scale from 0 to 3.

Provided that the initial identification of dependencies is explicit, the values (derived by mathematical procedures) in the minimum link matrix are substantive. The processes of matrix multiplication are not complicated, but they are tedious for large matrices and would normally require the use of a computer. It is unfortunate that, while the minimum link matrix can indicate the existence and length of a linkage between any two components, the structure of these linkages is not exposed.

The results of the component interaction analysis were not readily incorporated into the overall assessment of the transshipment project. In fact, the ad hoc assessment report of the five Nanaimo Port site proposals made little use of the results displayed in the component interaction, minimum link, and disruption matrices.

The CIM has been reviewed by Clark et al. (1981) and Bisset, but has not received much positive comment. The minimum link matrix is useful as a means of communicating the complex structure of the environmental systems likely to be affected by a project.

4.3.5 Advantages of the matrix approach

• A matrix presentation has a better structure framework than the checklist approach. In fact, it makes a summarized analytical presentation of the project and environment-related checklists.

• Matrix structure allows for speculation of impact characteristics, albeit in a subjective way. This provides a gradation in the impacts, thereby providing a focus for further studies, verification, or discussions. It also helps in making priorities on some mitigation measures which are estimated to alleviate the impacts speculated.

• It presents an easily understood summary of a large number of primary impacts.

• It is a generalized but well defined approach, forcing a comprehensive consideration of environmental components and primary impacts.

• It is an easily performed process which can specify the overall character of a project early in the design phase.

• In an extended form, the method can include information about many impact attributes, and clarify the assumptions supporting the assessments.

• Matrices have low resource requirements.

4.3.6 Limitations of the matrix approach

Despite the elegance of matrix presentation, there are certain limitations which need to be addressed.

• Unless weight-scaled impact scores are used, the comparison of many project alternatives is difficult.

• Scaling the multitude of scores contained in a matrix is also not a tractable proposition, as the ability to independently replicate the method is undermined by a dependence on highly subjective judgments.

• The impact characterization step of the matrix involves subjective prediction as well as assessment.

• There is little opportunity for quantification. However, it is possible to accommodate further detailing in the matrix presentation if prediction/evaluation techniques are separately used.

• While developing matrix structure, it becomes apparent that higher order impacts are not accounted for using this approach.

For example, impacts propagate from one component to another and are not necessarily linked directly with the project activities. In the case of a thermal power plant, waste emissions alter the air quality and the altered air quality in turn affects crops, public health, or materials. A water resources project upstream of a river mouth entering the sea, alters the fresh river flow into the sea and this in turn changes the saline zone of the river mouth. This change in the saline zone influences the marine life feeding near the saline wedge, which influences the income of fishermen as well as the marine ecosystem in general. Both these examples question our rudimentary understanding of impacts. This implies that impacts on the nth environmental component can be due to simultaneous and/or successive changes in the other interlinked components.

There is a lot be learned from this improved understanding of impacts.

• In the case of component-component or secondary impacts, the project activity specificity ends. In other words, if a particular project activity alters a particular component, then, regardless of the project activity, this changed component would affect the linked component. For example, if the temperature of the water of the river is raised above a certain threshold (by any activity) then the fish life in the river would be affected.

• Impacts have non-linear relationships and due to the participation of more than one component in some cases, there is the possibility of a delay in their realization, especially in terms of time. Again, delayed impacts do not mean that the "size'' of the impact is attenuated. It is possible that the size can be bigger, especially if there are processes such as "biomagnification'' or if the receiving environment is fragile (e.g., mangrove ecosystems).

• The matrix style needs to be expanded to allow for component-component interactions. This is technically possible by writing a matrix adjacent to another and so on but it can become rather clumsy if there are multi-component (or multi-order) impacts. One may need here a presentation style which allows one to depict the interconnections in a causal style. Network presentation, discussed later, is perhaps a better choice.

• Writing a single matrix for infrastructure or spatial projects becomes rather difficult. For a thermal power plant, for instance, the impact of waste emissions on air quality depends on whether the region under consideration is mostly downwind or not. If a region (or portion of the neighbouring environment) is beyond a hill, then the waste emissions from the power plant almost get screened. Similar arguments would hold for describing the impacts upstream or downstream of a water resources reservoir. In other words, the impact association attempted in the matrix style assumes homogeneity or isotropy in the region, which is not the case in most situations. This may call for writing more than one (maybe five or six) matrix presentations for a project, describing specific situations happening in the spatial elements. This leads once again to technical as well as communication difficulties. Use of geographical information systems (GIS) coupled with impact assessments methodology becomes an attractive alternative.