In hydrological investigations, modeling and forecasting of snow melt runoff requires timely information about snow properties and their spatial variability. The liquid water content in snow pack is an important parameter. Previous study1 has indicated that the fully polarimetric C-band synthetic aperture radar (SAR) is capable to estimate the free liquid water content-snow wetness-in the top layer of a snow pack quantitatively. The objective of this study is to evaluate the capability of a radar system with measurements of the dual frequency (C-band 5.3 GHz and Ku-band 13.4 GHz) and of the dual-polarization (VV and HV) in estimation of snow wetness based on the numerical simulation. We have established C-band and Ku-band radar wet snow data-base by using second-order radiative transfer backscattering model. The data-base covers the most possible wet snow physical properties and surface roughness conditions. Using this data-base, an inversion algorithm has been developed for snow wetness retrieval. The newly developed algorithm mainly involved two steps: 1) decomposing the surface and volume scattering signals using depolarization factor, and 2) using each scattering component (surface or volume backscattering signals) to estimate snow wetness.
In this paper, we evaluate the capability of a multi-scattering microwave emission model that including the Dense Media Radiative Transfer Model (DMRT) and AIEM to simulation of dry snow emission with Matrix Doubling approach. We compared the predictions of this model with the ground experimental measurements. The comparison showed that our snow microwave emission model agreed well with the experimental measurements. In order to develop retrieval snow properties: snow depth or snow water equivalence (SWE) retrieval algorithm, we carried out the sensitivity test between the emission models with the different scattering-order: the zeroth-order, the first-order and the multi-scattering models. The results indicated that the multi-scattering effects have to be taken into account in the snow emission model, especially for large grain size. Due to the complexity of the multi-scattering model, we developed a parameterized inversion model using our multi-scattering emission model with a wide range of snow and under-ground properties for algorithm development purpose.