Assessing the Impacts of Historic and Pre-Historic Droughts on a Modern-Day Reservoir System Serving Phoenix, Arizona, USA

Approximately 40% of the water demand of the Phoenix, Arizona metropolitan area in the Southwest United States is met by the Salt River Project (SRP) through a supply of water that approaches one million acre-feet annually from a system of reservoirs along the Salt and Verde Rivers. Climate change concerns have led to studies that typically produce a range of percentage change in mean runoff from the river basins for a future 30-year climatological period. Such findings do not incorporate temporal evolution or the high year-to-year variability that characterizes precipitation and runoff within this semi-arid region. Analysis of the cumulative effects of climate variability upon the reservoir system is essential to understanding impacts on water availability. Through consultation with SRP personnel we developed a reservoir simulation model that incorporates the parameters and operational guidelines embodied in their long-term drought planning.

To test the robustness of the reservoir system we passed two runoff series through the model – one derived from the instrumental record (1898-present) and one from a reconstruction of pre-historic stream flows for the Salt and Verde rivers using tree-ring analysis (1361-2005). The instrumental record includes three significant droughts, while the pre-historic data broaden the range of possible hydrological conditions on the basins by revealing extended periods of drought far greater than in the instrumental record. Tree ring data overlapping the instrumental record suggest that peak values in high and low inflow years may be understated. The study therefore also explored cumulative response sensitivities to accentuating years within drought periods. The results show that operational guidelines communicated by SRP and built into the model provide measures of resilience that can avoid reservoir system depletion under a wide range of flows and temporal variability. The reservoir system appears well-matched to the arid watersheds on which it relies. Of course, simulation modeling responds only per its assumptions and there are clearly variables not yet exercised which would force the system into a critical condition. These include curtailment of supplemental groundwater pumping during times of drought. However, results thus far indicate that the SRP reservoir system will confront and is capable of surviving periods of significant flow deficit.