In arid and semi-arid regions with agriculture-dependent economies, water demand often conflicts with ecosystem needs, putting long-term pressure on the Environmental Water Supply (EWS). This study proposes a novel scenario-driven decision support system (DSS) that evaluates complex water allocation strategies under realistic hydrological conditions. The DSS framework uniquely integrates forecast-informed inflow scenarios with a physically-based Water Evaluation and Planning (WEAP) hydrological model, refined by combining the Soil Moisture (SM) and MABIA (MAitrise des Besoins d'Irrigation en Agriculture) modules and a multi-objective calibration method. We applied this framework to Iran's Lake Urmia basin (LUB), which has experienced a sharp water-level decline in recent years. Using a Multi-objective Evolutionary Algorithm based on Decomposition (MOEA/D), we assessed a range of water allocation scenarios, including dynamic Peak Environmental Supply (PES), crop pattern changes, and deficit irrigation strategies. Our analysis shows that optimal strategies can increase EWS to the lake by 31% while simultaneously boosting agricultural profit by 26%. Shifting PES to late winter-early spring yields a robust compromise, raising EWS while keeping agricultural supply and unmet demand within acceptable limits. The DSS further ranks water allocation scenarios and clarifies how outcomes shift as priorities move from agricultural-driven to environmental-driven, or toward a balance of both. As climate change continues to exacerbate water scarcity and variability, this integrated, scenario-driven modeling approach provides a transferable and practical tool for identifying water-management strategies that enhance agricultural profitability while supporting the ecological recovery and long-term ecosystem function of shrinking lakes.