UC Santa Barbara

One challenge facing water managers is how to supply various downstream claims with conflicting priorities, especially during droughts. This study calculates the total volumes of water released from the Shahid reservoir in the Fars province, Iran, during a 5 year drought period using a standard operation policy (SOP) and genetic programming (GP). The calculated releases show the vulnerability index (VulMax) equals 87 and 82%, the timebased reliability index is 45 and 42%, and the resiliency index reaches 23 and 13% for the SOP and GP, respectively, whereas the volumetric reliability index equals 55% for both methods, demonstrating superior results for the GP method. The releases calculated with SOP and GP are allocated to meet downstream water demands according to their supply priorities. The proportional (Pro) method is used for this purpose considering the available reservoir water is less than the total demands. Compliance with the priority of supplying demands is approached with nonlinear programming in the Pro method. The VulMax of allocated releases obtained with GP for urban, environmental, industrial and agricultural demands equals 5, 50, 60 and 100%, respectively. The results confirmed the Pro method effective in allocating water resources to meet downstream demands that have conflicting water-supply priorities.