Forecasting the Effects of Snow Drought on Streamflow Volumes in the Western U.S.

TitleForecasting the Effects of Snow Drought on Streamflow Volumes in the Western U.S.
Publication TypeConference Proceedings
Year of Conference2019
AuthorsSturtevant, Joshua T., and Harpold Adrian A.
Conference Name87th Annual Western Snow Conference
Conference LocationReno, NV
Keywordsclimate change, NRCS, snow drought, streamflow, Water supply forecasting

Mountain snowpack supplies critical water resources to natural ecosystems and downstream populations, particularly in semi-arid regions such as the Western U.S. Extremely low snowpack, or snow drought, can arise from inter-annual climate variability but is worsened by long-term declines in snowpack. Snow drought has negative implications for water availability as earlier, slower, and smaller snowmelt fundamentally changes runoff patterns including runoff efficiency and groundwater recharge rates. We explored the implications of snow drought on streamflow in the Western U.S. through the investigation of historical (1980-2014) observational data from the U.S. Geological Survey and Natural Resource Conservation Service for 59 watersheds. Our primary focus was evaluating the accuracy of April-July streamflow volume forecasts using the statistical method of Principal Component Regression (PCR). Early results indicate that forecast errors are highest during snow drought years, but that these errors vary across basins and are affected by precipitation received after the forecast issue date. We also show that forecast errors have a strong positive correlation with the inter-annual variability of runoff efficiency. Our objective is to refine statistically-based forecasting methods by introducing non-linearities into the forecast equations or subsetting the data into dry and wet years in order to improve streamflow volume forecasts during snow drought years.