Snow depth monitoring with GNSS reflectometry: Results from Antarctica and selected geodetic ground stations

The permanent and seasonal snow covers are an important element of the global hydrological cycle and have substantial influence on global climate. Currently around 10% of the Earth’s land surface is covered by glaciers, ice caps and snow cover. Snow and ice cover play important role in the Earth’s climate by reflecting solar radiation and thus decreasing the average Earth temperature. Glaciers and ice caps participate in a positive feedback loop in the Earth’s climate. By contracting due to increasing temperatures, they reflect less solar radiation, further contributing to the global temperatures increase.

Using the single antenna ground-based GNSS Reflectometry (GNSS-R) method for snow depth estimation is an emerging application. A new technique for snow depth measurement using the phase changes in the observed SNR data, rather than the height estimates, is validated in a GNSS-R setup in Antarctic station Neumayer III. The new technique shows improved characteristics to the classical single antenna ground-based GNSS-R snow depth determination method. The validation is done in an environment of constant snow accumulation. The results from new technique show high correlation of the de-trended datasets between the GNSS-R and in-situ snow buoy measurements of 0.85. The de-trended classical height estimations of the SNR show lower correlation to the snow buoys of 0.60.

A screening of the International GNSS Service (IGS) global network shows, that snow depth observations are possible in only 7 of the 506 available stations. The main limitations on the stations are the local topography and climate. The snow depth observations from these seven stations are compared with the ERA5 snow depth estimations, local measurements and climate normals. The analysis of the data for station Visby, following the new GNSS-R analysis technique, shows very high correlation of 0.91 and low RMSE of 2.26cm, while the classical GNSS-R estimation has RMSE of 2.48cm and ERA5 shows RMSE of 4.2cm when compared to local meteorological observations.