IGAD Groundwater Information System
World Bank Logo Project Logo Water Unit Logo
Facebook X LinkedIn Instagram YouTube

GW Recharge PREVIEW

Groundwater recharge scenarios for the IGAD region at 1 km resolution, computed with the Thornthwaite–Mather soil water balance method developed for the HoA-GW4R Joint Regional Study.

Preview version — provisional scenarios, not a forecast.

Period

Rainfall scenario

Display

Map

Click on the map to read the value at a point.
Loading…

Model coverage by country

Share of each country’s surface (area-weighted). “No PET” areas are mostly deserts and sparsely vegetated land where the MODIS potential evapotranspiration product is not computed; “No soil” areas are mainly lakes.

About these maps

Method and data

Monthly Thornthwaite–Mather soil water balance (method of Dr.-Ing. S. Legesse, consultancy under the HoA-GW4R programme). Inputs: CHIRPS rainfall, MODIS potential evapotranspiration and OpenLandMap soil water properties, with a constant effective root depth of 0.5 m and a depletion fraction of 0.5, as in the training notebook. The soil water state was computed monthly from 2012 to December 2025, then advanced month by month to the last observed month (currently October 2026) with observed CHIRPS rainfall and climatological potential evapotranspiration. January–September use CHIRPS v2; later months use CHIRPS v3 preliminary.

Scenarios

For the months not yet observed (currently November–December 2026) the monthly rainfall climatology is multiplied by a single factor (dry ×0.7, normal ×1.0, wet ×1.3). These factors are arbitrary placeholders, not a probabilistic forecast: they should not be read as the outlook of any forecasting centre.

Limitations
  • Values are model estimates, not measurements of groundwater flux, and carry the uncertainty of the input data and parameters.
  • All water surplus is assumed to recharge: surface runoff is not separated, so recharge is likely overestimated in very wet or steep areas. Values are not capped.
  • Where the Dry and Normal maps look identical, all the recharge comes from observed October rainfall: the November–December rainfall of those scenarios does not add any further recharge in the model. Recharge only appears once the soil is full, so it reacts very unevenly to the amount of rain.
  • Areas without data (grey) are not modelled; they are not necessarily areas of zero recharge.
  • Resolution is about 1 km; local conditions (geology, land use, irrigation) are not represented.

Implementation in GWIS: IGAD Water Unit. Funded by the World Bank.