|Long Distance Water Transfer: A Chinese Case Study and International Experiences (UNU, 1983)|
Commission for the Integrated Survey of Natural Resources, Academia Sinica, Beijing, China
THE TERM "water resource" in this chapter indicates the average yield of a basin over a number of years. "Surface water resource" is a general term for the stream runoff in a basin: this is somewhat different from the concept of surface runoff used in hydrology. "Groundwater resource" refers to that portion of the shallow groundwater which actively participates in the water cycle. Strictly speaking, it refers to that portion which is not duplicated in stream runoff. This is based on the premise that, if the formation of runoff is considered, surface water and groundwater are originally one organic whole. They both originate from precipitation and each is an integral component of the water cycle. The difference is only one of form. Therefore, when evaluating the total water resources of a river basin, we should act according to the principle of the water cycle and combine surface and groundwater in an integrated water balance framework. Each should not be calculated separately and mechanically, and then the results added to derive the total water resources of a river basin, as this inevitably results in considerable double counting.
In keeping with the above basic principles, the following steps must be followed in evaluating the water resources of a relatively large river basin:
(1) Have a relatively comprehensive understanding of the basin's water yield
conditions and runoff formation processes. In particular, one must keep in mind
the interchange between surface and groundwater.
(2) Divide the basin into several computational units according to the surfacegroundwater recharge relationships. Determine which components should be included in the water resources of the basin. Particular attention must be paid to separating out those districts where surface and groundwater might be double counted.
(3) Select different balance equations in light of the specific conditions of each component of the resources.
(4) Compare the calculation results with actual data.
The following is a concise explanation of the method of calculation using the basin of the Hai and Luan He as an example.
Features of the Computational Basin
The Hai-Luan basin lies between 112° and 120°E longitude and 35° and 43°N latitude. It has an area of 319,000 km², of which 60 per cent is mountainous and 40 per cent is plains. Two major mountain ranges, the Yan Shan and the Taihang Shan, lie across the basin from northeast to southwest in the shape of an arc. The western and northwestern portions of these ranges is a high plateau with several intermontane basins such as the Yingbei, Xinding and Changzhi in Shanxi Province. To the east and southeast lies the North China Plain. Since the hilly transition zone is quite narrow, the mountainous and plains regions are connected almost directly (Figure 1). The plains are level with a slope of roughly 1:10,000. The intermontane basins in the upper reaches of the basin and the plains in the lower reaches have relatively thick Quaternary deposits and are important agricultural regions. About 84,700 km² are cultivated in the plains and 28,700 km² in the mountains. These are 66 per cent and 15 per cent respectively of the total land area.
Climatically the basin is divided into the semi-humid North China warm temperate zone and the semi-arid Inner Mongolian temperate zone. The mean annual rainfall of 557 mm is one of the least among the coastal areas of eastern China. Because of the influence of the strong southeast monsoon, 70 to 80 per cent of the year's precipitation is concentrated in the four-month flood season from June through September. There are significant interbasin disparities in precipitation due to the influence of the arcuate mountains mentioned above. The southern and southeastern windward portions of the ranges are wet, with a maximum precipitation up to 800 mm while the northwest is dry, with a minimum of only about 300 mm per annum.
The Characteristics of the Calculated Basin Stream Runoff Formation and the Mutual Recharge Relationships Between Surface and Groundwater
In general the rivers in the basin can be classified into two types, mountainous and plains. Mountainous rivers can be further divided into their upper reaches in the mountains and lower reaches in the plains.
Because they flow through mountainous regions, the upper reaches of the mountainous rivers are erosion channels with steep gradients and deeply cut beds. Except for some individual karst regions, the basin boundaries for surface and groundwater are generally the same. In terms of water balance, they are closed basins with perennial flow. Stream flow during the flood season comes mainly from the surface runoff formed by rainfall, while dry season stream flow is fed primarily from subsurface runoff resulting from the direct seepage of precipitation during the flood season.
The gradients of the depositional lower reaches of the mountainous rivers flatten abruptly and the river beds broaden and become shallow. Sometimes a channel lies partially or entirely above the land surface along either bank and must be restrained by dikes. In these stretches, the river generally cannot receive groundwater recharge from either bank. On the contrary, the river recharges the areas along its banks through seepage, irrigation diversions and floodwater breaches. That is to say, the mountain runoff recharges the groundwater along both banks in the process of crossing the plain.
Most of the plains rivers are artificially excavated channels for the drainage of excess surface water during the flood season. They have low gradients and broad, shallow beds. After the flood season, they never receive groundwater recharge and usually become dry. In terms of the water balance, they are intermittent flows without closure, i.e., that portion of flood season precipitation which seeps into the ground usually cannot flow back into the river after the flood season but is stored in aquifers whose phreatic level is lower than the river bed.
River Runoff and the Composition of Water Resources in the Basin
According to the preceding analysis, mountain river runoff represents the total water resources of the mountainous regions because it includes all of both the surface and groundwater runoff. Plains river runoff cannot represent all the water resources of the plains because it is limited to that portion of plains precipitation which produces surface runoff and excludes the portion which recharges the groundwater. Therefore there should be three components of water resources in the basin:
(a) mountain river runoff;
(b) plains river runoff; and
(c) the direct recharge of plains groundwater by atmospheric precipitation.
Estimation of Basin Water Resources
Mountain river runoff and plains river runoff can usually be obtained through statistical analysis of observed runoff data. The specific computational steps will not be discussed here.
The direct recharge of plains groundwater by atmospheric precipitation, unlike stream runoff, cannot be obtained through direct measuring methods.
This is especially true with such a large recharge area (128,000 km²). Therefore we must use an indirect method, viz., a hydrological balance approach. The steps of computation are as follows:
Balance equation. Since the plains rivers of the basin are intermittent without closure, the elements in the hydrological balance should be related to each other as in the following equations:
for natural conditions (i.e., no groundwater involvement). Here, nearly all recharge from precipitation is eventually exhausted through evaporation; and
Determination of balance elements
(a) Precipitation. Data on precipitation in the basin are plentiful
and an isoline map of long-term average precipitation depth has been drawn up,
from which the value for the plain (P) may be obtained directly: 573.0 mm.
(b) Surface runoff. An isogram of long-term average runoff depth was drawn on the basis of the above-mentioned stream runoff data. Rs' measured from this map, is 57.6 mm.
(c) Evapotranspiration. There is a paucity of experimental data on evaporation in the plain. There are only five years of observation experiments carried out between 1961 and 1965 at Dezhou, Shandong Province, by the Institute of Geography of the Chinese Academy of Sciences. Furthermore, these measurements include evaporation of the phreatic water.
Consequently, the value of E for the entire plain is determined by comparing climatic and physiographical factors which influence evapotranspiration in the mountains and in the plains. This comparison reveals: first, that the combined effect of climatic factors on potential evaporation is basically the same in the mountains and the plains; and second, by analyzing hydrological balance for 64 river basins in the mountains, we discover that there is a clear relationship between precipitation and evapotranspiration in basins with basically the same physiographical characteristics. Moreover, as shown in Figure 2, there is a certain regularity in the distribution of data points even in different kinds of basins, i.e., the data lie to the right for those basins which have good evapotranspiration conditions and to the left for those which do not. Also, since the groundwater is generally fairly deep in the mountainous areas, evapotranspiration
usually represents only that of the zone of aeration and does not include
Using the evapotranspiration of the mountainous areas in this way to
ascertain that of the plains provides a basis for analysis reflecting conditions
where the utilization capacity is strong and the phreatic water is relatively
deep (e.g., 4 m or more below the surface). Considering the physiographical
conditions of the plains, they are generally more prone to evapotranspiration
than the mountainous areas. Finally, using the precipitation and
evapotranspiration data of the Dezhou experimental stations for reference, we
chose the line relating P and E for the leeward slopes which are fairly close
physiographically to the plains. The relevant equation is approximately:
Specifically, if we let P = 573 mm, we derive E = 470 mm. This figure may be too small, but it is unlikely to be too large.
(d) Groundwater runoff depth. Substituting the values determined above for precipitation, surface runoff and evapotranspiration into balance equation (2), we obtain Rg = 45.4 mm. This amounts to 5.82 km³. This is the amount of direct recharge into the plains groundwater from atmospheric precipitation.
Hydrological Balance of the Basin
Once the water balance elements have been determined, we can draw up a hydrological balance for the entire basin. The results are presented in Table 1.
From this table we can acquire a general impression of the interrelationships between the various hydrological elements in the basin. Of the long-term average precipitation of 177.49 km³, about 81 per cent (143.32 km³) is consumed in evaporation and the remaining 34.17 km³ goes to form runoff. Of the latter, mountain river runoff constitutes 61.3 per cent (20.95 km³), plains river runoff 21.7 per cent (7.40 km³) and atmospheric precipitation recharge of the plains groundwater 17 per cent (5.82 km³).
The Sources of Plains Groundwater Recharge
There are two main sources of plains groundwater recharge: direct recharge from atmospheric precipitation and recharge by the mountain runoff as it passes across the plain. We have already explained how to estimate the former. Here we will introduce a method of estimating the latter by means of the basin surface water balance.
Period covered. The recharge value which we seek here is a relatively longterm average. In view of the large-scale construction of reservoirs in the mountainous areas of the Hai He basin begining in 1959 and completed in most eases in 1961, we have chosen the subsequent eleven years(l962-1972)as our basic computational period. This period reflects the impact of existing water control facilities on the utilization of surface water and its recharge to plains groundwater. Furthermore, the average annual runoff in the mountainous regions of the Hai He in these eleven years was very close to the long-run average, so our calculations may also represent long-term average conditions. The balance was computed for each year and then the eleven-year mean was calculated.
Balance equation: (all units are in m³/annum)
R = Rm + Rp = Cm + Cp + Er + DV + Gp + R0 (3)
where R = total runoff:
Rm = runoff in mountainous areas
Rp = runoff in the plains
Cm = net water use in mountainous areas
Cp = net water use in the plains
Er = reservoir evaporation
DV = change in reservoir storage
Gp = recharge of plains groundwater from surface water
R0 = discharge into the sea
The determination of balance elements. Observation data are generally available for Rm, Rp, DV and R0. Cm can be obtained through investigation and experimentation. Er can be calculated from water surface evaporation data and the area of the reservoirs. Cp refers to the net amount of mountain runoff which is used for plains irrigation after it has been regulated by the reservoirs. It is computed by multiplying the actual year-to-year beneficially regulated reservoir volume by the effective utilization factor of the canal systems in the downstream irrigation districts. The beneficially regulated volume of water can be derived from the statistical observation records for reservoir operation and the effective canal system utilization factor can be obtained from observation data in the existing irrigation districts.
Once the above seven elements have been determined, the recharge of plains groundwater from surface water Gp can be determined from balance equation (3). The value so derived is approximately 6.0 km³ for the Hai He basin. If we add the Luan He and the 5.82 km³ direct replenishment from atmospheric precipitation from Table 1, we obtain a total replenishment of groundwater of approximately 12.5 km³.
Table 1 Long-term Annual Hydrological Balance of the Hai-Luan Basin
|Region||Basin||Drainage area||Precipitation||Runoff||Evapotranspiration||Precipitation recharge to groundwater|
|Hai He basin total||264,617||147.3||557||22.6||86|
|Luan He basin total||54,412||30.2||556||5.8||105|
Evaluation of Accuracy of Results
The accuracy of the surface water computations is the easiest to handle because we have streamflow measurements. The general level of accuracy is especially high for mountain stream runoff because of the density of the station network and the relatively long data records. The accuracy of groundwater components is harder to grasp because of the lack of observed runoff data. Between 1973 and 1978, however, one agency carried out a large-scale investigation of the actual extraction of groundwater on approximately half of the plains area (about 62,000 km²). These survey figures show that no matter whether viewed from the five-year average or from the wet year of 1977, the above groundwater source calculations using the water balance are all quite satisfactory and very close to the actual conditions.
Because of the way work was divided in the past, it was common practice to treat surface and groundwater resources separately and then to add the two together to estimate the total water resources of a basin or region. This kind of evaluation method is open to question as it double counts a considerable amount of water. For example, in the Hai-Luan basin, the groundwater of the intermontane basins and that portion of the groundwater of the North China Plain in the lower reaches which is recharged by surface runoff are in actuality already included in the mountain river runoff and should not be counted again. The question, however, is how to get rid of the duplicated portion. In the case of the intermontane basins this is handled relatively simply by disregarding groundwater. The main problem is in the plains where atmospheric precipitation and surface water recharges are mixed together and are difficult to separate. This chapter has proposed using two kinds of water balances to estimate these two types of recharge. This approach makes feasible the integrated evaluation of surface and groundwater resources in a river basin and allows the rationally extractable reserves of plains groundwater to be estimated hydrologically.