Abstract:

The Global Navigation Satellite System Reflectometry (GNSS-R) technology has become an important method for estimating the snow water equivalent (SWE). However, when using a snow density model-based method to invert the GNSS-R SWE, the accuracy is often affected by the lack of meteorological observation instruments around the GNSS stations. To address this issue, this study proposed a method that integrated a snow density model established using European Centre for Medium-Range Weather Forecasts Re-Analysis version 5 (ERA5) data and snow depth series derived from GNSS-R to estimate the SWE. This approach aimed to overcome the limitations of insufficient meteorological parameter resolution in traditional GNSS-R SWE estimation. To verify the applicability and effectiveness of the proposed method, experimental studies were conducted at two GPS stations, AB33 and SG27. The accuracy of SWE estimates obtained by constructing snow density models using both SNOwpack TELemetry (SNOTEL) observational data and ERA5 reanalysis data was analyzed. In the region near the GNSS station with fewer SNOTEL stations, the proposed method achieved a higher level of accuracy in SWE estimation compared with traditional GNSS-R SWE inversion methods based on SNOTEL data. Specifically, the snow density correlation improved by approximately 11%, root mean square error (RMSE) decreased by approximately 12%, and SWE RMSE reduced by approximately 24%, effectively mitigating the limitations of traditional methods caused by the spatial resolution of meteorological parameters. Furthermore, the SWE estimates obtained using the proposed method demonstrated a higher level of accuracy than those based on ERA5 reanalysis data, with a correlation increase of approximately 9% and an RMSE reduction of approximately 20%. This further confirms the effectiveness of the proposed method and validates its application potential in the GNSS-R field.

Key words: GNSS-R, ERA5, snow density model, snow water equivalent