A model-based groundwater recharge zone mapping for food security: A case study of Notwane sub-catchment in Botswana

The understanding of groundwater recharge occurrence in drylands is central to water resources management for various uses. This study uses Remote Sensing and GIS techniques to understand where groundwater recharge occurs, and its implications for water and food security in Notwane River Basin located in the Botswana drylands. WetSpass, a distributed hydrological model was applied to map the potential groundwater recharge zones. Crop yield was predicted using the Commonwealth Science and Industrial Research Organization (CSIRO) precision weighing system. Model inputs were land use, soil texture class, topology, slope, groundwater level and catchment hydro meteorological patterns from 1987 to 2017- all sourced from satellite images. Image based classification was done to map the land cover changes in the catchment using ILWIS 30 software. Model outputs were evapotranspiration, surface runoff and groundwater recharge zone maps. The results of image-based land cover classification showed an increase of Settlements/buildup area (22.35%), grassland (5.24%) and a decline in forest cover (3.64%), agricultural land (22.23%) and bareland (3.16%). The results indicate that high recharge zones are associated with low surface runoff in rural, forested areas with sandy soils and the opposite is true for urban, buildup with clay soils. CSIRO predicts yield estimation of up to 2.037 x 103 tonnes of drought resistant maize or sorghum annually using 1100 x 106 L of the available 517.32-434.32 mm/year and 532.64-426.50 mm/year potential surface runoff and groundwater recharge, respectively. Runoff and potential recharge in Notwane sub-catchment suggest an existence of water resources worthy to be explored for food security in water scarce drylands.