Improving Streamflow Forecasts in the NOAA National Water Model Using Observational Constraints on Snowpack Albedo and Snow-Covered Area from STC-MODSCAG (Extended Abstract)

The NOAA National Water Model (NWM) is a physically-based modeling system which simulates major hydrologic processes across the conterminous United States (US). Difficulties in accurately simulating snowpack states hinder the NWM’s ability to provide high-quality streamflow forecasts, particularly in snow-dominated western US mountains. Errors in snowpack simulations propagate into streamflow time series as errors in both magnitude and timing of peak streamflow in snow-dominated basins.

We imposed observation-based constraints on simulated fractional snow-covered area (fSCA) and snowpack albedo in the NWM using remotely sensed data and investigated the impacts on simulated snow states and streamflow over the Upper Colorado River Basin. We identified a set of parameters that influence the relationship between snow depth and fSCA (the snow depletion curve) and seasonal snowpack evolution. For each parameter, we derived spatially-distributed values using 15 years of data from STC-MODSCAG (Spatially and Temporally Complete MODIS Snow-Covered Area and Grain Size), which provides daily estimates of fSCA, snowpack albedo, and other variables at ~500 m spatial resolution. When implemented into the NWM’s snow model, the derived values tended to shift simulated streamflow peaks lower and earlier, often improving agreement with observed streamflow. Results from these experiments will help to improve streamflow forecasting for water management and inform NWM data assimilation strategies with model parameter uncertainties.