| Popis: |
Irrigation water use can be significantly affected by crop prices and production costs yet policy makers continue to rely on static analyses that do not properly account for uncertainty in these influences. This is particularly problematic in developing economies which are largely dependent on agriculture but facing severe water constraints. We model irrigation water demand uncertainty in two West Bank agricultural districts by executing a deterministic optimization algorithm (mathematical program) repeatedly using Monte-Carlo simulation. When use is highly constrained as seen in the Jenin district, very different decisions would be made if additional water were available, as illustrated by the water demand curve for the Jericho district. We show how this modeling approach can be used to improve the robustness of policy and infrastructure planning and to explore trade-offs between the resilience and effectiveness of such policies under contrasting demand scenarios. Introduction It has long been recognized that systems analysis techniques used to evaluate water allocation problems are not responsive enough to adequately reflect the true concerns of decision makers (Rogers and Fiering, 1986). One of the primary reasons cited is the insufficient treatment of uncertainty in modeling (Tsur and Dinar, 1997; Harou et al, 2009). As simplifications of reality, significant characteristics of a system are represented as parameters and decision variables and the mathematical relationships among them. Parameters are average or calibrated values that best represent the conditions upon which decisions are made. In truth however, parameter values are not precisely known and small changes can lead to dramatic changes in model outputs (Harou et al, 2009). Water allocation policies and infrastructure must be effective over long planning periods due to their associated expenses and required effort. Robust and resilient planning helps ensure the system performs as intended under a wide range of environmental and economic conditions. Robust interventions are those that can be applied to a set of possible scenarios or outcomes (Watkins and McKinney, 1997). |
