Spatially downscaled future food demand for Uganda under climate and social economic changes

Spatially downscaled future food demand for Uganda admin 1 unit, based on IMPACT global economic model.
Methodology:We use high-resolution spatial inputs, a machine learning model, and a matching estimator to downscale national food demand projections generated by a global economic model to subnational level. We suggest that the observed relationships between demand and a range of social-demographic-spatial features can provide a means of downscaling future food demand. We interpret the downscaled food demand at the subnational level as the estimated spatial distribution of a given national level food demand projection.