A study on a combined retrieval algorithm for falling snow and snow on the ground using passive microwave observations

Snow plays an important role in hydrological and climatic cycles of Earth. Monitoring snow falling and snow on ground can be beneficial for estimating water and hydropower supplies during spring melt, for improving the capability of forecasting weather and natural hazards, and for understanding the Earths climate system where surface albedo changed by falling snow can make a significant impact. For last decade, many studies have been seeking the methods to retrieve parameters related with snow on the ground (e.g. snow depth and snow coverage) or to retrieve parameters related with falling snow (e.g. snowfall rate) using the space-borne passive microwave remote sensing observations. Several retrieval algorithms have been discussed in the literature. Calculations of electromagnetic scattering properties of snow crystals have been mainly focused in previous studies on falling snow while main attention has been focused on parameterizations of multiple scattering to consider correlated scatters (e.g. dense medium theory) in snow on the ground. However, a physical model combining falling snow and snow on the ground has not been challenged yet although they are strongly linked by hydrological process and in passive microwave observations. For example, change in surface emissivity of snow on the ground can cause large uncertainty for retrieving precipitation rate by affecting observed microwave brightness temperatures at frequencies between 89 GHz and 183 GHz, especially in dry environment like polar region. At the same time, unknown amount of falling snow makes it difficult to validate a land surface model predicting the snow depth. In this study we investigate the possibility of combining electromagnetic models for falling snow and snow on the ground and retrieval algorithms in order to improve the estimation of geophysical parameters of both snow conditions. We also evaluate improvement of each algorithms performance by comparing retrieved parameters before and after applying the combined technique.