Feasibility demonstration of using the signal-to-noise ratio observations from geodetic GNSS receivers to retrieve dry snow density

The geodetic Global Navigation Satellite System (GNSS) receiver has been proven to estimate snow depth using the phase change rate of the signal-to-noise ratio (SNR) observations. The snow density (SDE) can be reflected by snow permittivity and, therefore, is theoretically sensitive to the amplitude of the GNSS reflected signal. However, retrieving SDE using the SNR observations is challenging due to the difficulty in extracting the reflected amplitude since it hides in the interference waveform and changes with the satellite elevation angle. This study proposed a novel GSARM model that relates the corrected instantaneous amplitude ratio (<i>α</i>) to the snow permittivity and the resulting SDE. The GSARM-derived SDE is compared with three other data sources, i.e., the PBO-H₂O, the ERA5-Land, and the in-situ measurements over three GNSS sites for six consecutive years. The overall mean RMSD (MAE) values of SDE for GSARM versus PBO-H₂O, ERA5-Land, and in-situ measurements are 0.051 g/cm³ (0.043 g/cm³), 0.052 g/cm³ (0.046 g/cm³), and 0.049 g/cm³ (0.042 g/cm³), respectively. The Root Mean Square Percentage Deviation (RMSPDs) (Mean Absolute Percentage Error (MAPEs)) for the SDE of GSARM versus PBO-H₂O, ERA5-Land, and in-situ are 23.38% (23.19%), 22.80% (19.77%), and 24.20% (23.89%). The findings of this study first prove the feasibility of using geodetic GNSS receivers for SDE retrieval. It also provides supportive information for extending the added-value applications of traditional geodetic GNSS sites and for developing new observation patterns.