Abstract for: Land–Water Allocation, Yield Stability and Policy Trade-Offs under Climate Change: A System Dynamics Analysis

Climate change is intensifying drought risk and agricultural water scarcity, making it harder to balance yield stability, irrigation efficiency and farm income in grain-producing regions. Existing studies often optimise crop structure or irrigation separately and seldom assess how pricing and subsidy policies perform dynamically when climate stress and absolute water scarcity interact. We develop an integrated ML–SD–NSGA-II framework for a rice–wheat irrigation system in Hubei, China. Machine learning generates future daily climate series, system dynamics simulates soil moisture, yield, water availability, and farm returns at dekad scale, and NSGA-II searches crop-area and irrigation strategies under land, water, and policy constraints. Compared with the observed baseline, Pareto-efficient strategies improve irrigation water productivity by about 14% without relaxing land or water constraints. A profit-oriented strategy raises net revenue by about 6% but increases yield deviation, whereas a knee-point strategy keeps profit near baseline while improving water efficiency and reducing instability. Under extreme drought, policy instruments provide only limited buffering. The findings show that coordinated adjustment of cropping structure and irrigation allocation is more effective than relying on single policy instruments alone. Water pricing and subsidies can modestly improve allocation under normal conditions, but structural biophysical scarcity dominates under severe drought. The framework provides a transparent tool for ex ante testing of adaptive land–water governance strategies. For translation