Assimilation of Airborne Snow Observatory Snow Water Equivalent to Improve Runoff Forecasting Model Performance and Reservoir Management During Warm and Dry Winters

Abstract:

The Airborne Snow Observatory (ASO) NASA-JPL demonstration mission has collected detailed snow information for portions of the Tuolumne Basin in California for three years, 2013 - 2015. Both 2014 and 2015 were low snow years, and 2015 was exceptionally warm and analogous to future years after climate change. The ASO uses an imaging spectrometer and LiDAR sensors mounted in an aircraft to collect snow depth and extent data, and snow albedo. By combining ground and modeled density fields, the ~weekly flights over the Tuolumne produced both basin-wide and detailed sub-basin snow water equivalent (SWE) estimates that were provided to Hetch Hetchy Reservoir operators. The data were also assimilated into an hydrologic simulation model in an attempt to improve the accuracy and timing of a runoff forecasting tool that can be used to improve the management of Hetch Hetchy Reservoir, the source of 85% of the water supply for 2.6 million people on the San Francisco Peninsula. The USGS Precipitation Runoff Modeling System was calibrated to the 1181 square kilometer basin and simulation results compared to observed runoff with and without assimilation of ASO data. Simulated and observed were also compared with observed with both single updates associated with each flight, and with sequential updates from each flight. Sequential updating was found to improve correlation between observed and simulated reservoir inflows, and there by improve the ability of reservoir operators to more efficiently allocate the last half of the recession limb of snowmelt inflow and be assured of filling the reservoir and minimizing ecologically-damaging late season spills.