Enhancing Snow Depth Estimations Through Iterative Satellite Elevation Range Selection in GNSS-IR to Account for Terrain Variation

Abstract

The multipath effect in Global Navigation Satellite System (GNSS) has become a robust data source thanks to GNSS interferometric reflectometry (GNSS-IR), which provides environment-related features by considering the interference pattern of direct and reflected signals recorded simultaneously on the antenna phase center (APC) of the GNSS receiver. When analyzing the fluctuation of the signal strength, known as signal-to-noise ratio (SNR), SNR metrics such as frequency, amplitude, and phase can be estimated. Frequency is directly related to the reflector height, which can be converted into snow depth. However, traditional GNSS-IR approaches have limitations in retrieving environmental features, as they generally focus on a selected satellite track or use all available tracks together for a common satellite elevation angle, which can result in missing appropriate satellite elevation angle ranges for station-based retrievals, especially for nonplanar surfaces. To address this issue, we proposed an approach that searches for the proper satellite elevation angle range for each satellite track to improve snow-depth retrievals. We analyzed 31-day GNSS data from a CORS station named KARB located in Istanbul, Turkey, including a three-day heavy snowstorm, to prove the performance of the proposed method. The results of the proposed algorithm were compared with the traditional GNSS-IR method results and in situ snow-depth measurements. The initial results indicate a 10.70% increase in the correlation for snow-depth estimation compared to the traditional approach using L2 SNR data. Moreover, when the results were assessed based on median absolute deviation (MAD) threshold values, increasements of 4.93% and 13.84% were obtained in the correlations for L1 SNR and L2 SNR, respectively.