A Microwave Radiance Assimilation Study for a Tundra Snowpack

Recent studies have begun exploring the assimilation of microwave radiances for the modeling and retrieval of snow properties [1—6]. At a point scale, and for short durations (1 week), radiance assimilation (RA) results are encouraging [5, 6]. However, in order to determine how practical RA might be for snow retrievals when applied over longer durations, larger spatial scales, and/or different snow types, we must expand the scope of the tests. In this paper we use coincident microwave radiance measurements and station data from a tundra site on the North Slope of Alaska. The field data are from the 3 Radiobrightness Energy Balance Experiment (REBEX-3) carried out in 1994-95 by the University of Michigan [7, 8]. This dataset will provide a test of RA over months instead of one week, and for a very different type of snow than previous snow RA studies. We will address the following questions: How well can a snowpack physical model (SM), forced with local weather, match measured conditions for a tundra snowpack?; How well can a microwave emission model, driven by the snowpack model, match measured microwave brightnesses for a tundra snowpack?; How well does RA increase or decrease the fidelity of estimates of snow depth and temperatures for a tundra snowpack?