Impacts of Satellite-derived Fractional Snow-Covered Area Observations on Operational Streamflow Predictions via Direct Insertion

Abstract

Snowmelt is a primary driver of spring and early summer streamflow in the western United States. Improved predictions of snowmelt-driven streamflow benefit a wide variety of users. In this study, the snow model used in the National Weather Service's hydrologic operations, SNOW17, is run with and without consideration of fractional snow covered area (fSCA) observations from the National Aeronautics and Space Administration's MODerate Resolution Imaging Spectroradiometer (MODIS). Because computationally frugal methods are desirable in an operational environment, the updating scheme evaluated is a simple direct insertion method. Resulting predictions of snowmelt-driven streamflow for water years 2000 to 2010 are compared to observed flow and a control simulation (using the model without snow cover input). Results indicate that use of MODIS fSCA in SNOW17, with no adjustments, via direct insertion, degrades the streamflow predictions, compared to control simulations. Future research directions include advanced data assimilation and use of different snow models.