Global modeling of water availability, water use, and water quality


The WaterGAP model (Water - a Global Assessment and Prognosis) has been developed at the Center over the past decade with a first implementation in 1997. The overall aim is to investigate current and future world-wide water availability, water use and water quality. Whereas water availability and water use have been implemented, a water quality module is currently under development and constitutes the next major goal. In particular, Water-GAP is concerned with the various impacts of global change on water availability and water demand, and to determine the development of water stress conditions on different spatial and temporal scales. Further research subjects are the variations of the water balance components on the hydrological large scale, and the future development of extreme conditions - such as floods and droughts.


To date, WaterGAP is being applied in a number of different projects. The model combines a hydrological module for the determination of global water resources and water availability, and a water use module to quantify water consumption from different economical sectors, including a sub-model for an assessment of global irrigation requirements. Until now, the following results can be reported for the hydrological module WGHM of WaterGAP

  • In order to enhance WGHM towards an integration of the hydrologically relevant processes of lake and wetland retention and evaporation, a new Global Lakes and Wetlands Database has been created (Lehner and Döll, 2004).
  • The introduction of a sub-grid variability in the snow sub-model of WGHM has improved the representation of snow related processes, especially the dynamics of snow accumulation and snow melt in the model (Schulze and Döll, 2004).
  • For a better reproduction of river discharge and in order to improve the model with respect to flooding processes, the formerly constant flow velocity is now considered as a function of actual discharge and the hydraulic conditions of the river, expressed as the roughness and drop of the river bed and its geometrical conditions (Schulze, Hunger and Döll, 2004). 

The water use module in WaterGAP is subdivided into a number of sub-models according to the different water use categories. With respect to these sub-modules, the following has been accomplished:

  • Domestic water use: The basic approach of this sub-model is to first compute the domestic water use intensity [m³/cap-year] and then to multiply this value with the number of population. Further, it is assumed that changes in water use intensity can be expressed by structural and technological changes. The structural changes are represented by a sigmoid curve which indicates the relationship between water use intensity and income (Alcamo et al. (2000, 2003a, b). In the former model version, historical data from Shiklomanov (2000) were used to derive the sigmoid curves for 26 regions. In the new version of this sub-model, sigmoid curves for around 170 countries have been derived, with data mainly provided by national and international statistics.
  • Water use for electricity production: The objective of this sub-model is to compute water use for producing electricity. Here, the model simulates the amount of water withdrawn and consumed for cooling purposes (Vassolo and Döll, 2005). The former version of this model has been updated by using the newly available World Electric Power Plants Data Set of the Utility Data Institute (UDI, 2004). Additionally, the model has been validated for the base year 2000 with country data on electricity production and the volume of water used for cooling purposes in thermoelectric power plants.
  • Water use in the manufacturing industry: In order to assess the great diversity of industrial processes and the variety of input and output specifications, this sub-model distinguishes between six manufacturing sectors. Data on water withdrawal per sector are available for many countries from national or international statistical agencies for the base year 2000. It is assumed that future trends of water with-drawals for the manufacturing industry can be estimated by multiplying the water intensity of the base year with the future trend of the driving force. In this approach, the sector-specific gross value added (GVA) is used as driving force. The sector-specific water use intensity for the base year is obtained by dividing the sector-specific water use for the base year by the GVA. Technological improvements are considered by means of a sector-specific technological change factor. One of the main overall results reported by the model output is water stress which indicates the amount of pressure put on water resources and aquatic ecosystems by the various users. Water stress is estimated as the ratio of withdrawals to availability (w.t.a.). The higher the w.t.a.-ratio, the more repeatedly the water in a basin is used and the higher is the possibility of the resource to be polluted or depleted. As opposed to the w.t.a.-ratio which indicates changes in water quantity, the concept of "return flows" is used to assess changes in water quality. Return flows are defined as the difference between water withdrawals and consumption and thus give a first estimation of the amount of wastewater discharged back into the terrestrial water cycle. This constitutes the beginning of the development of a water quality sub-module and has resulted in a Global data set on wastewater treatment. A global data set has been developed that provides information on the percentage of population connected to a public or independent sewage system as well as the wastewater treatment technology. With this information, the amount of return flows which has undergone a wastewater treatment can be assessed.


The main task is to develop a WaterGAP 3 version with the major objectives

  • to increase the model's spatial resolution from 30-minutes (1/2°) to 5-minutes (1/12°). This enables a better validation of the model output with regional data and facilitates further developments in the description of hydrological processes in the model. Further, the higher spatial resolution allows to carry out detailed analyses of water-related problems on the regional scale.
  • to revise, in accordance with the advance in the spatial resolution of the model, the underlying data-bases on climate input parameters and natural conditions, and make them available to the appropriate spatial scale of the new model version. For example, in a joint project with WWF, a new Global Drainage Direction Map in 5-minute resolution is currently under development.
  • to develop a new water quality sub-model to determine nutrient and pollutant fluxes in different pathways which will allow to combine water quantity with water quality analyses.