Advancing the representation of reservoir operations in SWAT to evaluate alternative water management strategies: A case study in the Omo River Basin, Ethiopia

Agricultural systems are complex coupled natural-human systems in which farmers’ water and land management decisions are influenced by, and in turn influence, their land quality and economic resources through changes in erosion, nutrient retention, and income. These dynamic interactions are further influenced by transient climatic stressors that can lead to tipping points in agricultural productivity. Water infrastructure development may enable farmers to adapt to climate change by transitioning to more productive irrigable agriculture. However, water infrastructure development may also negatively affect ecological resources and inhibit water access for downstream users. Similarly, farmers’ adoption of fertilizer may improve agricultural benefits, but degrade downstream water quality. A useful tool for analyzing such agroecological tradeoffs is the Soil and Water Assessment Tool (SWAT), which can model irrigation, fertilizer application, and reservoir operations, although options for simulating reservoirs remain fairly limited. This study advances the representation of reservoir operations in SWAT by integrating alternative, closed-loop, multi-reservoir operating policies optimized using Evolutionary Multi-Objective Direct Policy Search into the model. This enables water managers to evaluate the tradeoffs of more complex water management options while allowing for a better spatial visualization of impacts than traditional reservoir models. As a case study for this integration, we consider the Omo River basin of Ethiopia. Recent controversial dam construction along the Omo River will create opportunities for hydropower and irrigation, but has already proven detrimental to indigenous people and aquatic wildlife depending on natural flows and sufficient water levels in Lake Turkana downstream. Comparing our advanced reservoir operations with SWAT’s existing operating options, we find a wider range of policies for managing these conflicting stakeholder objectives that better compromise across them. This also aids in the identification of policies that are more robust to a range of heterogeneous future scenarios, including changes in both human actions and climate conditions. This shows promise for informing integrated water resources management in the Omo and other river basins.