Estimation of snow water equivalent from SSM/I data by using a physical model-based inversion algorithm

The employment of parameter retrieval to the Helsinki University of Technology (HUT) snow microwave emission model is presented. The radiative transfer-based semi-empirical model describes the emission behavior of a homogeneous snowpack as a function of snow water equivalent (SWE), effective grain size and density of snow. The model takes into account the influence of soil surface, forest canopy and atmosphere to spaceborne observed brightness temperature by using empirical and semi-empirical formulas. The comparison of model predictions with independent data shows good correlations. This enabled the development of an inversion technique for the SWE retrieval based on the model. The test results using SSM/I data from boreal forest zone showed SWE retrieval accuracies considerably higher than those obtained with conventional algorithms.