Inversion of a passive microwave snow emission model for water equivalent estimation using airborne and satellite data

Snow Water Equivalent (SWE) is an important variable in climate and hydrological studies in northern latitudes. To date, remote sensing tools, especially in the microwave domain, have shown their potential for monitoring this parameter. However, operational use at the regional and global scales needs further improvements. Strong uncertainties remain, mainly due to snow metamorphism as well to the presence of vegetation cover. The objective of this paper is to use the semi-empirical Helsinski University of Technology (HUT) snow emission model to estimate SWE with a minimum number of ancillary ground information. Non linear inversion techniques were used to compute SWE considering snow grain size as a free parameter, and initializing the inversion process with rough values of snow depth and density. Results were validated by comparing SWE estimations derived from airborne and spaceborne 19.35 and 37.0 GHz brightness temperatures to ground based SWE measurements over Central Canada. We showed that remotely sensed SWE estimations, retrieved with an average root mean square error of 24 mm, permit the monitoring of the seasonal snow pack evolution, as well as interannual variation (tested over 11 years), useful for hydrological applications.