Abstract: Global navigation satellite system reflectometry (GNSS-R) technology has become an important supplement to traditional snow depth measurement. In this study, the GNSS-R technique was used to obtain snow depths near four GPS stations in Alaska during 2012–2018. In combination with snow depths from the Canadian Meteorological Centre (CMC) model, and taking snow depth data from the Plate Boundary Observatory (PBO) H2O project team as reference, variations in snow depth obtained using different methods on different time scales were analyzed. The capability of GNSS-R-derived snow depth was taken as an independent dataset to evaluate the performance of the CMC model data. Results showed that long-term snow depths from GNSS-R, CMC, and PBO all exhibit obvious and consistent periodic variation. Typically, the GNSS-R-derived results are more accurate than the CMC data in detecting interannual variation in snow depth. Both GNSS-R and CMC can capture the monthly variation seen in the PBO data for each station, although the accuracy and stability of the GNSS-R results are generally better than those of the CMC. In comparison with the CMC results, seasonal variation in GNSS-R-derived snow depth is more consistent with the PBO data. For the four studied stations, GNSS-R-derived snow depth accuracy is higher in spring and winter when snow depth is larger, and slightly worse in autumn when snow depth is smaller. Overall, GNSS-R is proven effective for evaluation of the accuracy of CMC-simulated snow depths, and the evaluation effect is generally better in winter and spring than in autumn.

Key words: GNSS-R, CMC, snow depth, Alaska