LAUSR

Abstract Many passive microwave remote sensing applications such as land surface temperature, snow water equivalent and soil moisture retrievals need to take into account a soil parameterization to the overall surface signal emission. Soil emission modeling presents large uncertainties when the soil is frozen. In this paper, an empirical retrieval method is presented, specifically for rough frozen soil permittivity estimates at 10.7, 19 and 37 GHz. The method was tested and validated using in-situ passive microwave measurements at incidence angles from 0 to 60° of sub-arctic frozen organic soils in Northeastern Canada. The retrieved permittivity values give an overall RMSE between the measured and simulated brightness temperatures of 4.6 K for all frequencies combined. A sensitivity analysis was conducted on the different soil parameters optimized in this study. This analysis suggests that the accuracy of the retrieved parameters, using the method given here, is of ± 1.00 for the permittivity and ± 0.12 cm for surface roughness. Also, a comparison was conducted between the parameterization used in this study and the one of Wegmuller and Matzler (1999) to estimate the soil contribution to the emitted brightness temperature of snowpacks. An improvement of 66% of the RMSE between the modeled and measured snow brightness temperatures was observed when using the approach of this study compared to the previous work. The method shows great potential to improve the estimation of the frozen soil contribution to the measured passive microwave brightness temperature.