# 3. METHOD USED TO COMPUTE WATER RESOURCES BY COUNTRY

The method used to assess renewable water resources by country was first described in FAO/BRGM (1996). It consists of a set of rules and guidelines leading to the calculation of the IRWR, the total renewable water resources (TRWR), and the country’s dependency ratio.

The method is based on a water resources accounting approach. The TRWR of a country consist of the IRWR plus the external water resources. The IRWR are the amount of water generated inside a country, and the ERWR are the amount of water generated in countries upstream. To avoid double counting, the IRWR is the only variable that can be aggregated for regional or continental assessments.

In order to calculate ERWR, a distinction has been made between natural and actual ERWR. The natural ERWR refer to the amount of water flowing to a country from upstream under natural circumstances. The actual ERWR refer to the amount of external water resources actually available to the country taking into consideration upstream water abstraction and possible agreements with upstream and/or downstream countries.

The TRWR are generally not equal to the amount of water available for use. Therefore, where possible, a compilation has been made of estimations of exploitable water resources per country and included in the country results.

The calculation of renewable water resources is based on long-term averages as available in existing, preferably national, literature. Figure 4 illustrates the hydrological cycle and the components of the country water resources calculations. Figure 5 presents the standard spreadsheet used in the computation and shows the various components of the calculations.

The method consists of the following steps (for each country):

1. Select the most accurate data sources.

2. Assess the IRWR.

3. Assess the natural and actual external water resources entering and leaving the country.

4. Assess the TRWR (actual and natural).

5. Calculate the country’s dependency on external water: the dependency ratio.

6. Ensure consistency between countries by cross-checking inflows and outflows between countries.

The Aquastat computation sheet (Figure 5) is used to calculate: (i) the IRWR; (ii) the ERWR; (iii) the TRWR; and (iv) the dependency ratio in annual averages (km3/year).

## Data sources

The data used in computing water resources originate from multiple sources, including grey literature. Most of the data originate from national sources (priority is systematically given to national sources over international reports). A major part of the data originates from the country surveys carried out for 150 countries within the Aquastat programme between 1993 and 2000. A compilation of the individual sources by country is available on the Aquastat Web site (www.fao.org/ag/agl/aglw/aquastat/main/index.htm).

FIGURE 4: The components of the calculation of water resources by country

Note: P = precipitation, ET = evapotranspiration (sum of the local evapotranspiration and the evaporation from lakes, swamps, etc.), R = total river flow, I = groundwater recharge, QIN = infiltration of surface water to groundwater, QOUT = outflow of groundwater to surface water, SWIN = surface water entering the country, SWOUT = surface water leaving the country, GWIN = groundwater entering the country, GWOUT = groundwater leaving the country, SWPL = part of surface water in border lakes, SWPR = part of surface water in border rivers.

Prior to the calculation, the information collected through the surveys was evaluated, mostly through cross-checking and by comparing the level of detail of the estimates. However, the methodologies behind the estimation of the individual components of the water balance were not always well documented. This made the evaluation more complex.

FIGURE 5: The Aquastat computation sheet

Figure 5 notes

(1) Overlap between surface and ground water = less than 50 % of groundwater recharge; only a small part of the groundwater is drained by the rivers (equal to the low flow of water courses). Most of the groundwater escapes and flow out into the sea, or into sebhat in arid areas. In addition, there is probably some infiltration from surface water.

(2) Exploitable resources according to tunisian sources. The exploitability criteria is probably technico-economical. Another tunisian source (Ennabli, 2000) indicate a lower figure: only 1.9 km3/yr are exploitable.

(3) Tunisia has non renewable resources in the south, estimated to be 0.6 km3/yr, extractable over a period of time (50 years?).

Comment (comparison with modelled data)

FAO/AGLW model to assess internal resources based on rainfall, evapotranspiration, and calibration on flow measurement gives 3.2 km3/yr; so slightly lower than the national figure (9 % difference). The difference is not significant.

Source

BP/JM, 2001 from tunisian sources

Finally, the information in this review stems from multiple sources, with different periods of reference that might vary from country to country. Moreover, the original data were acquired through different methodologies and assumptions, and they might have been extrapolated in time and space. Such variations complicate attempts to determine comparable outcomes.

## Assessing internal renewable water resources

The IRWR are equal to the volume of average annual flow of surface water and groundwater generated from precipitation within the country (Equation 1):

(Equation 1)

where:

R = surface runoff, the total volume of the long term average annual flow of surface water generated by direct runoff from endogenous precipitation;

I = groundwater recharge, generated from precipitation within the country;

QOUT = groundwater drainage into rivers (typically, base flow of rivers);

QIN = seepage from rivers into aquifers.

Although they are linked through the hydrological cycle, surface water and groundwater resources are often computed separately. Therefore, a simple addition of surface water and groundwater leads to an overestimation of the total amount of freshwater resources produced in the country. In this study, the exchange between the surface water and groundwater resource is called overlap. Box 3 illustrates the complexity of surface water and groundwater interdependency in humid and arid countries and the method used to compute overlap.

## Scale and impact of evaporation

Evaporation in rivers, swamps, lakes or large irrigation schemes can have an influence on the estimation of the IRWR (Box 4). In arid areas in large countries, there is a high probability that part of river runoff is lost by evaporation before leaving the country (in some cases, rivers flow to salty depressions where water cannot be put to beneficial use). In such cases, surface water resources should be assessed by measuring runoff upstream of the major loss areas, where they reach their maximum value.

## Assessing external renewable water resources and total resources

The ERWR are equal to the volume of average annual flow of rivers and groundwater entering a country from neighbouring countries (Equation 2):

(Equation 2)

where:

SWIN = surface water entering the country;

SWPR = accounted flow of border rivers;

SWPL = accounted part of shared lakes;

GWIN = groundwater entering the country.

Although inflow from other countries usually consists of river runoff, it can also consist of groundwater transfer between countries in arid regions. However, groundwater transfers are rarely known and their assessment requires good knowledge of the general behaviour of the aquifers. Where groundwater resources estimates are based on groundwater flow as derived from the characteristics of the aquifers and piezometric levels, the calculated flow does not correspond necessarily to renewable resources. There are many cases of transboundary flows that are related to the slow draining of huge groundwater reservoirs with negligible upstream recharge.

 BOX 3 - SURFACE WATER AND GROUNDWATER INTERDEPENDENCY OR OVERLAP Summary of the approaches applied for estimating the overlap: Humid areas In humid areas, IRWR are assessed from available hydrographs (time-series data on measured surface water discharge). For areas where no measurements are available, data is extrapolated over space from areas where data is available. Where necessary, measured data are corrected to take water abstraction into account. In humid areas, the base flow of rivers consists mainly of drainage of groundwater reservoirs. Thus, estimates of surface water resources include a significant part of the groundwater resources. Therefore, the groundwater resources in humid areas have been assumed to be equal to the base flow of the rivers where data are available. Semi-arid areas In semi-arid areas, IRWR are generated mainly from flash-flood events. The groundwater resources are obtained from rainfall infiltration estimates or from analyses of measured groundwater levels/heads in aquifers. The surface water resources are estimated through flash-flood discharge measurements or estimates. Care is required to ensure the correct assessment of the part of surface water flows that recharges the aquifers in order to avoid overestimation of the total water resources. In coastal or very arid areas, a large part of the groundwater aquifers is not drained by the rivers and overlap is therefore relatively small. Examples of relation between surface water and groundwater: Surface water and groundwater base flow in Morocco: The total groundwater resource equivalent to the aquifer recharge is estimated at 10 km3/year, of which 3 km3/year corresponds to the base flow of rivers. Recharge of aquifer by floods in the Islamic Republic of Iran: The contribution of surface water (floods) to aquifer recharge is estimated at 12.7 km3/year, out of the total IRWR of 128.75 km3/year. Surface water and base flow in Kyrgyzstan: The total groundwater resource equivalent to the aquifer recharge is estimated at 13.6 km3/year, of which 11.2 km3/year are drained by the surface hydrologic network, which corresponds to the base flow of rivers. Therefore, the overlap in this case is 11.2 km3/year.

 BOX 4 - INFLUENCE OF EVAPORATION ON IRWR In arid countries, losses by evaporation from wetlands, lakes and rivers reduce the entire flow generated in the country. Losses by evaporation, associated with infiltration, may reduce significantly the surface water flow after it leaves the mountainous part of the basin from where it originates. The losses may reach 100 percent in the case of endorheic basins ending up in evaporative areas called chotts or sebkhas. However, there is a difference between losses in salt-affected evaporative areas, where water does not contribute to any biomass production, and losses in swamp areas, where evapotranspiration produces biomass to sustain aquatic and terrestrial life. In countries belonging to large international river basins, evaporation losses in swamps and in the river itself may exceed the total IRWR. However, it is difficult to assess the actual evaporation in wetlands, lakes and rivers. For large lakes and dams, actual evaporation is approximated as equal to the potential evaporation. For some countries in Africa, a fair estimate of the net evaporation (this is evaporation minus precipitation, and it is different from potential evaporation) was computed and abstracted from the original national source of the total IRWR where it was known that no corrections had been carried out beforehand. This is the case for Sudan, where the national estimate of evaporation in the wetlands is 68 km3/year, which corresponds to double the country’s IRWR and 45 percent of its TRWR. Examples of evaporation losses: In the province of Yaérés, Cameroon, evaporation from swamps amounts to 5 km3/year. From the Aswan Reservoir in Egypt, it is estimated that 10 km3/year is lost by evaporation. In the inner delta of the Niger River, Mali, evaporation is estimated at 33 km3/year, equivalent to 50 percent of the country’s IRWR.In humid areas, actual evaporation from rivers, lakes and wetlands represents a smaller amount of the total IRWR in comparison with arid countries.

In the case of bordering rivers and shared lakes, an arbitrary 50-percent rule was applied to distribute the water evenly between two countries. This rule does not imply any consideration of judgement on possible or effective ways of sharing the resources of a border river. The difficulties encountered in setting these computation rules illustrate the arbitrary aspects of the computation of total water resources for bordering water bodies. In some cases, there are known agreements (Box 5).

This study makes a distinction between actual and natural ERWR. The actual ERWR take into account the quantity of flow reserved by upstream (incoming flow) and/or downstream (outflow) countries through formal or informal agreements or treaties, and possible water abstraction occurring in the upstream country. Therefore, the actual ERWR may vary with time. In extreme cases, the value may be negative when the flow reserved to downstream countries is more than the incoming flow (Equation 3).

(Equation 3)

where:

= volume of surface water entering the country which is not submitted to treaties;

= volume of surface water entering the country which is secured through treaties;

SWPR = accounted flow of border rivers;

SWPL = accounted part of shared lakes;

= volume of surface water leaving the country which is reserved by treaties for downstream countries;

GWIN = groundwater entering the country.

The term treaty is used in a broad sense and does not necessarily imply formal acceptance on both sides of a border about the amount of water to be reserved for each country. Furthermore, treaties cannot always be expressed uniquely in terms of annual flows, and interpretation might be needed for the purpose of this computation. For example, in the Aral Sea Basin, the flow allocation for the individual countries is expressed as a percentage of the actual resources in the basin. Therefore, the amount they receive depends on the amount available, which varies from year to year.

 BOX 5 - EXAMPLES OF SHARED RIVERS 1. Shared river completely submitted to a treaty Nile River transboundary between Egypt and Sudan Total natural discharge (average): 84 km3/year Sharing by treaty: Egypt: 55.5 km3/year,Sudan: 18.5 km3/year,Losses by evaporation of storage: 10 km3/year.2. Shared river partially submitted to an agreement Syr Darya River (Aral Sea Basin in Central Asia) Distribution by agreement defined in 1992: Kazakhstan: 14.5 km3/year,Kyrgyzstan: 4.92 km3/year,Tajikistan: 7.15 km3/year,Uzbekistan: 10.53 km3/year.3. Shared river partially submitted to a treaty Tagus River shared between Portugal and Spain Total natural average discharge: 18.65 km3/year: with a part produced in Spain: 12.2 km3/year,with a part produced in Portugal: 6.45 km3/year.Average actual flow at the Spanish border with Portugal: Actual = 9.76 km3/yearForecast for 2012 = 9.18 km3/year.

Box 6 describes the rules used to calculate the different components of the water resources. They are neither absolute nor universal. They have been selected to represent all the situations in the most realistic way possible. Figure 5 presents the standard computation sheet, in which these rules have been applied for the Syrian Arab Republic.

 BOX 6 - RULES APPLIED FOR COMPUTING ERWR Surface water entering the country (SWIN) The mean annual flow measured or estimated at the border of the transboundary river is accounted for as an external resource for the downstream country. It is not deducted from the resources of the donor country except in the case of an agreed apportionment, i.e. a treaty between the countries. Because of the existence of bilateral and multilateral agreements and upstream water consumption, two categories of external water resources are differentiated: Natural flow corresponding to long-term average flow not affected by or before being affected by upstream consumption. Actual flow corresponding to a given period which takes into account water abstraction from upstream, be it through an agreement or from a factual situation, and/or agreed or accepted commitments towards downstream countries.A particular case is the situation where part of the runoff entering the country originates in the country itself after it has entered and exited a neighbouring country. In such a case, and where the information is available, this flow is deducted from the incoming flow to avoid double counting. Therefore, net inflows are considered over the country borders. For example, the Pripyat River originates in Ukraine and flows to Belarus where it joins the Dniepr River before it enters Ukraine. In this case, the flow of the Pripyat River from Ukraine to Belarus (5.8 km3/year) is deducted from the flow of the Dniepr River to Ukraine (32 - 5.8 km3/year). Flow in border rivers (SWPR) As a general rule, 50 percent of the river flow is assigned to each of the bordering countries. Several situations exist: Where the river exclusively borders the countries without entering any of the adjacent countries nor exiting from them (e.g. the Senegal River between Mauritania and Senegal, the Zambezi River between Zambia and Zimbabwe, and the Prut River between the Republic of Moldova and Romania), the incoming resources are estimated on the basis of the river runoff in the upstream part of the border section. Where the runoff increases substantially from upstream to downstream, the downstream figure is used after subtraction of the part of the runoff generated by the country itself. Where the source of the river is in one of the two countries, the rule applies only for the other country. For the originating country, 50 percent of the contribution from the other country could similarly be considered as external resources where known (e.g. the Samur River, which originates in the Russian Federation before it becomes the border between the Russian Federation and Azerbaijan). Where the river enters one of the two countries after having divided the two countries, it is considered a transboundary river for the receiving country and all the runoff at the entry point in that country is considered as an external resource. The 50-percent rule applies for the other country. Where a treaty exists between the adjacent countries of a river system, the rules applied are those defined in the treaty.Shared lakes (SWPL) Where the lake has an outlet into a river (e.g. Lake Victoria enters the Nile River in Uganda), all the runoff at the entrance of the river is accounted for as external resources for the receiving country. For all the other countries, an equal share of this runoff can be considered as external resources, after having subtracted the country contribution to the lake. Where this results in a negative value, the external resources are considered to be zero for the country in question. Where the river forms the border between two countries, the rule described above for border rivers applies. For lakes without an outlet, the global runoff entering the lake is estimated and shared equally between the adjacent countries, after having deducted the part contributed from the country. Where this results in a negative value, the external resources are considered to be zero for the country in question. Artificial lakes have not been accounted for as the flow reduction is an impact of water development and not a natural phenomenon.Surface water leaving the country (SWOUT) The computation of actual ERWR considers the outflow of surface water only in the case of an agreed apportionment between the upstream and downstream countries. Groundwater entering the country (GWIN) The mean annual estimated groundwater flow entering the country is accounted for as an external resource. Groundwater leaving the country (GWOUT) The computation of ERWR does not consider outflow of groundwater.

## Assessing total renewable water resources and the dependency ratio

The TRWR is the sum of the IRWR and the total ERWR. As with the ERWR, a distinction has been made between the total natural and actual renewable water resources (Equations 4 and 5):

(Equation 4)

where:

IRWR = internal renewable water resources (Equation 1),

ERWRnatural = natural external renewable water resources (Equation 2);

(Equation 5)

where:

IRWR = internal renewable water resources (Equation 1),

ERWRactual = actual external renewable water resources (Equation 3).

In order to compare how different countries depend on external water resources, the dependency ratio is calculated. The dependency ratio of a country is an indicator expressing the part of the water resources originating outside the country (Equations 6 and 7).

percent (Equation 6)

(Equation 7)

where:

IWR = total volume of incoming water resources from neighbouring countries,

IRWR = internal renewable water resources,

= volume of surface water entering the country which is not submitted to treaties,

= volume of surface water entering the country which is secured through treaties,

SWPR = accounted flow of border rivers,

SWPL = accounted part of shared lakes,

GWIN = groundwater entering the country.

This indicator may theoretically vary between 0 and 100 percent. A country with a dependency ratio equal to zero does not receive any water from neighbouring countries. A country with a dependency ratio equal to 100 percent receives all its water from outside without producing any. The indicator does not consider the possible allocation of water to downstream countries.

## Assessing exploitable resources

In some cases, countries have estimated the part of their water resources which is exploitable. Where available, the figures on exploitable resources have been included in the spreadsheet as a reference. In general, not all the natural or actual renewable water resources are accessible because of economic, technical and environmental constraints. Therefore, the exploitable resources may be significantly smaller than the TRWR. In some exceptional cases, the method used to assess exploitable water resources includes return flow from irrigated fields. This may lead to an estimate of exploitable water resources that is higher than the TRWR, e.g. Egypt (Box 7).

 BOX 7 - ASSESSING EXPLOITABLE WATER RESOURCES IN EGYPT This box presents water withdrawal in Egypt by source of water. The actual primary resource corresponds to the actual TRWR (58.3 km3/year). If this estimation alone were compared with the actual water withdrawal (66 km3/year), it would indicate an overexploitation. However, it is not the case here because return flow and infiltration from agricultural fields (secondary resources) represent significant elements in the country’s water balance. Actual primary resources: (groundwater, and surface water, internal and external) 58.3 km3/year Secondary resources: 8.1 km3/year infiltration of irrigation water to groundwater: 4.0 km3/year drainage water and conveyance losses returned to the Nile River 4.1 km3/year Total exploitable water resources (of which withdrawals of groundwater amount to 5.3 km3/year plus reuse) 66 km3/year

Source: Amer, 1999.

### Remarks

TABLE 1: Water resources per watershed in Spain

Note:

Exploitation index (%): withdrawals of conventional freshwater resources (surface and groundwater) over total renewable resources (expressed in %).

Source: CEDEX, 2000.

National-level data provide an idea of the water situation of a country but they hide the local diversity, particularly in large countries. Therefore, it would be preferable to examine each watershed. Table 1 presents watershed data on Spain as an illustration of the variability of water resources situations within a country.

When examining flows of water between countries, cross-checking of transboundary flows is important. This study used matrices showing exchanges between upstream and downstream countries in order to compare inflow and outflow values and to ensure the overall consistency of the computation of country water resources. Matrices allow for a more detailed representation of the flows between countries and avoid double counting of transboundary (external) flows (they do not consider border rivers) (Figure 6).

FIGURE 6: Example of transboundary exchanges in Eastern Europe, km3/year

*: Ukraine receiving: from Europe: 106.150 km3/year. Table water resources: 86.450 km3/year. Difference: 19.7 = 5.8 (Pripyat) + 9.2 (Dnestr) + (2.9 + 6.5)/2 (border Prut and Cisa).