Dynamics in mountain SNOW water resources by MODEs of climate variability assessed from satellite observations

Satellite information concerning the snow water equivalent (SWE) stored in the world’s mountain ranges is still lacking. This observation gap hinders the accurate estimation of total seasonal water storage in snow. Therefore, the SNOW-MODE project aims to address this gap by improving and developing two satellite retrieval methods to estimate SWE. Firstly, a recently developed empirical change detection algorithm for SWE retrieval from Sentinel-1 (S1) backscatter observations will be thoroughly analyzed and, if possible, improved. Secondly, a snow (Bic-DMRT), soil (Oh) and vegetation (WCM) radiative transfer model (RTM) will be coupled and inverted to estimate SWE using S1 radar backscatter observations and auxiliary snow, soil and vegetation properties. This method will be applied at the point- and grid-scale. The point-scale approach will make use of detailed in-situ measurements and novel tower-mounted radar measurements for RTM development and validation of the retrievals, whereas the grid-scale approach will utilize data generated from a land surface and a snow model. The inclusion of the grid-scale approach allows to investigate whether spatial patterns in SWE can be accurately represented by the S1 retrievals.

Subsequently, both S1 retrieval methods (i.e., change detection and RTM) will be compared over several mountain regions in the Northern Hemisphere (High-Mountain Asia and European and western United States mountains) to assess their uncertainties, validity conditions and main strengths as well as shortcomings. Furthermore, a physics-based snow model (e.g., SnowClim) will also be utilized to simulate snow depth and SWE on a daily basis. To improve the simulation results, the meteorological forcings will be downscaled to a resolution of 500 meter. Further improvements will be aspired by assimilating the mountainous snow depth retrievals (either from the RTM or change detection method) into the snow model. Finally, the generated SWE dataset will be related to modes of climate variability and will be translated into basin-scale water resources availability for society.