Sand rivers — ephemeral rivers with water stored in their sandy alluvial beds — offer a decentralized, low-cost, and underutilized source of water in many arid to semi-arid regions in Africa. This study explores a novel approach to remotely estimate the water storage potential in these systems using satellite-derived dry season total evaporation from the riparian vegetation and sand channels. Sentinel-2 imagery was used to delineate sandy channels and riparian zones across three sub-catchments in southern Zimbabwe, and total evaporation was estimated using WaPOR v3 data. Abstractable water storage was estimated at the locations of 34 in-situ sand depth measurements from literature assuming a specific yield of 0.15 and a rectangular channel shape. Total evaporation was found to be below abstractable channel storage for 27 (79%) of the in-situ observations suggesting remaining unconsumed water in sand river channels. A linear regression was used to estimate abstractable water storage in reaches without depth measurements based on total evaporation. Two methods were used to obtain the slope of the regression and total unconsumed water storage was estimated to be sufficient to irrigate between 3 700 and 6 200 ha across the study area. While field studies are recommended to validate the results, our approach provides a spatially distributed indicator of nature-based water storage potential in sand rivers based on remote sensing. The method presented enables the rapid low-cost identification of priority areas for water access and smallholder irrigation and provides a first approximation for regional planning, in particular for farmer-led irrigation and decentralized water access. ... Read more