Multistage Stochastic Optimization for Semi-arid Farm Crop Rotation and Water Irrigation Scheduling Under Drought Scenarios

Extreme weather events such as droughts have posed a significant risk to the agricultural economy of the semi-arid region in the American Southwest. To address the potential drought scenarios, which impact the precipitation and water availability, a data-driven multistage stochastic optimization model is constructed for crop rotation and water irrigation scheduling, to maximize the expected farmers' profits over a planning horizon. The optimal decisions will be made for crop rotations, deficit level for water irrigation, crop yield response, and multi-method irrigation system scheduling. To overcome solving the multistage stochastic large-scale mixed-integer optimization model with the exponentially growing number of scenarios, we employ the stochastic dual dynamic integer programming (SDDiP) method. Numerical experiments and sensitivity analysis on drought scenarios are performed to validate the proposed approaches in a case study in Arizona.