Southern Zimbabwe is an illustrative example where (increasing) water scarcity can lead to food- and financial insecurity and hinder socio-economic development. Sand rivers are a nature-based alternative to reservoirs, hosting shallow aquifers in their sandy beds with potential for decentralized and clean water abstraction. However, their extent and storage potential remain poorly understood, limiting awareness and investment. Riparian vegetation, which depends on sand rivers and remains green during the dry season, could serve as a valuable proxy for water storage.

This study uses a remote sensing approach to estimate and map the capacity of sand rivers to store water across three sub-catchments of the Limpopo River Basin (Shashani, Mzingwane, and Shashe), with two methods. Firstly, the sand river and riparian zones were delineated. The minimum storage capacity was estimated and mapped by summing WaPOR v3 evaporation data over the dry season. Secondly, 35 depth measurements were combined with spatial analyses to empirically predict depth and geometric storage capacity for the whole study area.

The results showed that there is a significant water storage potential in the channels. The two estimates can be combined for an unconsumed water availability, showing a significant sustainable potential (totaling 83 x10^6 m3 potentially irrigating 8300ha). Dry season evaporation was consistently exceeded by geometric storage at medium to large evaporation rates, suggesting that these rivers have unconsumed storage potential. The maps suggest that most potential water storage is located in the main river stems (the largest 10% hold 45-55% of water). Most tributares seemed to evaporate all water. Using water for irrigation would directly compete with vegetation. However, some local hotspots were still observed at smaller rivers.

Although significant uncertainties remain and field validation is still needed, the models show promise for wide-scale planning and development. The findings highlight that sand rivers, with rough estimates of their decentralized, cost-effective, and sustainable water storage, can be mapped remotely with minimal effort or field data. This could open up possibilities to offer support to farmers in semi-arid regions and pave the way for farmer-led irrigation initiatives.