ESRL/PSD Seminar Series

Abstract

---

Snow is the dominant water storage term in much of the US West. In turn, much forecasting skill can be gained if snow initial conditions (both extent and volume) can be determined at the onset of melt. Data assimilation techniques offer potential to reduce the uncertainty of initial conditions within the forecast system. A quick review of current products and methods available for snow data assimilation will be presented. Focus will then be given to the NOAA-NOHRSC National Snow Analysis (also known as SNODAS), which has been produced since the 2004 water year with a view to providing assistance to hydrologic forecasters. The SNODAS system ingests a variety of satellite data and almost all available ground based observations in order to produce its analysis. As a consequence of this intense data use, there has been almost no objective assessment of the utility of this product. One of the few measures available to evaluate SNODAS is that of streamflow. A distributed hydrologic model (TopNET) is applied in two basins in the Northern Sierra Nevada, namely the North Fork of the American River and the East Fork of the Carson River. By assimilating the SNODAS data into the model, the skill of these data products is assessed against streamflow for the two basins. Results indicate that one can take either a pessimistic or optimistic view of the SNODAS data; this will be discussed.

---