Water System Analysis, Stationarity, System Dynamic Modeling, Drought, Climate


Hydrological drought is challenging managers of western U.S. snowpack-dependent urban water systems. Snowpack, reservoir storage, streamflow dynamics, and demand are routinely assessed to guide water system management and operations, assuming per-capita demand stationarity. Using the Salt Lake City Department of Public Utilities and two drought scenarios, we investigate water system vulnerability differences between unchanging industry per-capita forecasting methods and dynamic demands driven by hydro-climate-demand relationships. The introduction of dynamic demands estimates a 42% reduction in system vulnerability during supply limiting conditions than the industry methods. These modeled water use behaviors also suggest a reduction in the peak timing and volume (September 2nd, 55MGD vs August 2nd, 89MGD), duration (114 days vs 144 days), and seasonal volume (16,000ac-ft vs 25,000ac-ft) of out-of-district supply requests during extreme drought conditions. By relying on forecasts embedded with per-capita demand stationarity assumptions, large and unlikely system vulnerabilities can misinform management's operational actions.

OpenWaterJournalSubmissionCompilationofEdits.docx (48 kB)
Cover letter of manuscript changes