Modeling of water availability for food system transformation in Upper Offin Sub-basin and Mankran Micro-watershed of Ghana: a baseline study

The increasing population and changes in the climate in Africa demand a more sustainable approach to water usage for improved food and water security in the region. One of the key users of water, agriculture serves as the primary livelihood in Ghana, with a growing focus on cocoa production. To effectively implement sustainable water management strategies, it becomes imperative to conduct hydrological studies, including water balance components and water quality at sub-national and watershed scales. This would assist decision-makers in the proper planning and interventions for agriculture. This study aims to quantify and evaluate the hydrological response of the Upper Offin sub-basin and Mankran micro-watershed under baseline conditions. Upper Offin was selected because of its competing land uses of cocoa and mining, and Mankran was targeted as the CGIAR regional integrated initiative for Transforming Agrifood Systems in West and Sectral Africa (TAFS-WCA) is co-designing a landscape management plan for the area. The Soil and Water Assessment Tool (SWAT) model was first calibrated and validated at the Adimbera gauging station (Upper Offin sub-basin) using observed streamflow data from 2001 to 2011, considering Mankran as one of the SWAT sub-basins. After several iterations of the selected seven parametersthat include mainly channel and groundwater flow, the SWAT model reproduced the observed flow with reasonable performance. The sensitivity analysis depicted that channel and groundwater parameters were markedly the most sensitive in the region. Evapotranspiration accounts for the largest share of the water cycle, with a mean annual rainfall of 72% and 74% for the Upper Offin and Mankran watersheds, respectively. The mean annual surface runoff and percolation were below 5% for both watersheds. Also, the mean annual percolation for Upper Offin and Mankran were 15% and 17% of the rainfall and the mean annual sediment yield was 0.68 t/ha and 0.37 t/ha, respectively. The SWAT model successfully captured the hydrological responses in the study areas, providing a reliable quantification of surface runoff, percolation, and sediment yield under baseline conditions. Utilizing SWAT in this context was essential for assessing the potential impact of future supplementary irrigation interventions, evaluating the effectiveness of water management strategies, and monitoring changes in hydrological processes over both spatial and temporal scales.