The “agriculture resources sensitivity” represents the agriculture potential in 2010. This potential is measured by the availability of land and food production per capita, and the main thread to agriculture land, represented by desertification risk. The index results from the second cluster of the Principal Component Analysis preformed among 16 potential variables. The analysis identify four dominant variables, namely “potential rain-fed food production per capita”, “cropland crowding”, “desertification index” and “topographic resources availability”, assigning respectively the weights of 0.29, 0.29, 0.27 and 0.15. Before to perform the analysis all the variables were log transformed to shorten the extreme variation and then were score-standardized (converted to distribution with average of 0 and standard deviation of 1; all variables with inverse method except “desertification index”) in order to be comparable. The 5 arc-minute grid “potential rain-fed food production per capita” of 2007 was gathered from FAO GeoNetwork and then sampled at 0.5 arc-minutes. It was multiplied by crop land occurrence dataset, extracted from the FAO Global Land Cover-SHARE dataset of 2014 and divided by population grid in order to compute the per capita values. The 0.5 arc-minute grid “cropland crowding” of 2010 was produced dividing the crop land occurrence (FAO Global Land Cover-SHARE) by the population. The 0.5 arc-minute grid “desertification index” of 2000 was measured in terms of number of months recording values less than 0.75 of the ratio between precipitation (current monthly average) and potential evapo-transpiration (PET, current monthly average). Data of precipitation and PET were gathered from Worldclim and from CGIAR Consortium for Spatial Information, respectively. Finally the 0.5 arc-minute grid “topographic resources availability” was produced within the ClimAfrica project based on SRTM DEM of NASA. The “potential rain-fed food production per capita” measures the availability of food in a certain area produced with subsistence techniques. Cells with low food production are sensitive to climate change impacts because the low input agriculture (dominant in Africa) may not produced sufficient food quantities to support the local populations. The “cropland crowding” is an indicator that assess the availability of crop land hectares per 1,000 people. Sensitive areas are where few crop lands are shared by large population. The “desertification index” assesses the climatological risk of a certain area to be subjected to desertification due to lack of rainfall. Such areas are more sensitive to lost crop land and thus food production quantities due to climate change impacts. The “topographic resources availability” is the percentage of each cell with slopes equal to or lower than 15 %. Landscapes strongly dissected contain less land with agriculture values than plain landscapes. The scarcity of agriculture land may increase the fragility of a system because unable to increase to crop surface to cope with climate change impacts.

This dataset has been produced in the framework of the “Climate change predictions in Sub-Saharan Africa: impacts and adaptations (ClimAfrica)” project, Work Package 4 (WP4). More information on ClimAfrica project is provided in the Supplemental Information section of this metadata.