Snow Retrieval Algorithm for Passive Microwave Remote Sensing Using Dense Media Radiative Transfer Theory

The Quasi-crystalline Approximation and the Dense Media Radiative Transfer Equation (QCADMRT) have been applied to active and passive microwave remote sensing of snow [1] [2]. The salient features of the model include weaker frequency dependence than Mie scattering for the same grain size and a stronger forward scattering than Rayleigh phase function. We use passive microwave satellite data from AMSR-E to develop retrieval algorithms for the snow water equivalent (SWE). The passive retrieval algorithm makes use of the brightness temperatures at 18.7GHz and 36.5GHz and a least-squared optimization algorithm, where the snow grain size and snow depth are the two variables in the cost function. The retrieval algorithm has been applied to Alaska and Western United States. The results of retrieval algorithm are validated with SNOTEL ground measurement data from the USDA Natural Resources Conservation Service. The performance of SWE retrieval is beyond the heritage algorithm.