Flexible policies aimed at irrigated agriculture are essential to adapt to climate change. Despite the importance of this goal, little published work has conceptualized, formulated, developed, and applied an integrated optimization framework for irrigated agriculture to guide adaptation to climate-related water stress. This research addresses the question: how can water management plans for irrigated agriculture be designed to minimize economic losses caused by adapting to climate-induced water stress? The study answers this by developing an optimization approach that identifies water use patterns to minimize farm income losses during water shortages, considering three water shortage sharing programs. An optimization model, calibrated using positive mathematical programming, is applied to replicate historical land use while adapting to future water supplies that deviate from the historical pattern. The analysis focuses on two irrigated regions in North America’s Rio Grande Basin, illustrating land use, water use, and cropping patterns that minimize regional farm economic losses to shortages. These losses are assessed under three water-sharing strategies: intercrop and interdistrict trading (IIT), intercrop and intradistrict trading (IRT), and no trading (NT). The results demonstrate that IIT yields an average economic gain of $2.824 million per year, while IRT results in an average gain of $2.600 million per year compared to NT. These findings offer valuable insights for water managers, scientists, stakeholders, and policymakers tasked with developing irrigation management strategies in arid regions facing future water supply challenges. The methods developed and results shown here highlight a path forward, using scientific, economic, and policy innovations to strengthen agricultural livelihoods in regions facing uncertain water availability.