Economic representation of agricultural activities in water resources systems engineering

Water demands reflect water users' decisions in uncertain and often complex water systems, yet the driving forces behind these decisions are frequently neglected in water management engineering models. Traditional engineering modeling approaches represent economic water demands as fixed targets with pre-defined priorities and are often limited in capturing the behaviour of water users when affected by water management plans and operations. Despite the variety of solutions proposed in the fields of water resources engineering and resource economics, the understanding and simulation of economic water demands is still limited given the multiple decisions faced by water users and the uncertainty in water supplies. This dissertation presents modeling approaches for water management that borrow concepts and methods from resource economics combined with stochastic, linear, and quadratic optimization for simulating agricultural water decision and demands. The first approach is implemented with a two-stage stochastic quadratic programming model to simulate water, irrigation technology and conjunctive use decisions calibrated to real operations and marginal conditions. The model maximizes the net expected benefit of permanent and annual crop production with probabilistic water availability, and results demonstrate users' willingness to pay for increased water supply reliability and improved water management with conjunctive use operations. A second approach employs water demand functions to drive water decisions in a regional water system modeled with network flow programming and variable groundwater costs. Results demonstrate how users' decisions on supply sources and local management (conjunctive use) are influenced by water availability and price. The two-stage quadratic model is developed in the General Algebraic Modeling System (GAMS) and solved with MINOS. The network flow system model is developed in MODSIM.