Mountain areas provide disproportionally high runoff in many parts of the world, and here we quantify for the first time their importance for water resources and food production from the viewpoint of the lowland areas downstream. The dataset maps the degree to which lowland areas potentially depend on runoff contributions from mountain areas (39% of land mass) between the 1960s and the 2040s.

Water resources index W quantifies the potential dependence of the world's lowland areas on water resources originating in mountain areas upstream.

The data cover the timeframe from the 1960s (1961–1970) to the 2040s (2041–2050) in decadal steps. Data for projections from the 2010s onwards are available for three scenario pathways (SSP1-RCP4.5, SSP2-RCP6.0, SSP3-RCP6.0) and show median results from 5 CMIP5 GCMs (GFDL-ESM2M, HadGEM2-ES, IPSL-CM5A-LR, MIROC-ESM-CHEM, NorESM1‑M). The files are GeoTIFF formatted and in a regular raster of 5’×5’ (arc minutes in WGS 1984 coordinate system)

The values of W can be classified using the following ranges:
W ≤ -2 → Essential but vastly insufficient
-2 < W < -1 → Essential but insufficient
-1 ≤ W < 0 → Essential and sufficient
W = 0 → No surplus from mountains
0 < W ≤ 1 → Supportive
1 < W < 2 → Minor
W ≥ 2 → Negligible

The values of W are rounded to four decimal places and limited to a range of -1110 to 9998. Values falling outside of that range are set to the nearest limit.

he following flag values apply to W:
-5555 indicates that there is no water balance surplus from the mountain area upstream, but a lowland water balance surplus;
-6666 indicates that there is no water balance surplus from the mountain area upstream, and a lowland water balance deficit.

Mountain areas and oceans are NODATA, large ice shields are omitted (Greenland: NODATA, Antarctica: not covered in extent).