Parameterization of snowpack grain size for global satellite microwave estimates of snow depth

Abstract

For accurate estimation of global snow depth or snow water equivalent on the Earth's surface using passive microwave instruments, knowledge of the snow pack's physical properties is important. It is known that the bulk snow grain size distribution exerts an important control over the microwave response from snow between 3 mm and 300 mm wavelengths. In the absence of high quality snowpack data at a global scale, we show how the grain size distribution can be estimated using a general empirical model of grain growth. This information is used to parameterize a dense media radiative transfer model (DMRT) to estimate the radiometric response from a snow pack as a function of changing grain size distribution. The DMRT equations are based on the quasi-crystalline approximation (QCA) for densely distributed moderate sized particles in a medium such as a snow pack. The model snow depth estimates from the DMRT are used to calibrate a Special Sensor Microwave Imager (SSM/I) snow depth retrieval algorithm which is based on the brightness temperature difference between 19 and 37 GHz with the SWE. The method is tested using meteorological data from the WMO global network. Results show that using the DMRT model coupled with a grain size model improved estimates of snow depth are obtained.