Assessing GNSS-PWV and GNSS-SNR as indicators for rainfall detection: a case study for Austria

Severe weather events, like very heavy rainfall, are a significant threat to human lives and infrastructure. Therefore, it is essential to enhance our understanding of such events for improving preparedness and preventing irreparable damage. In the last two decades, GNSS has been widely used in meteorological studies to advance our capability for the analysis of short- and long-term weather phenomena and irregularities. The core objective of this paper is to investigate the variation of GNSS-derived products, namely PWV at 10-minute intervals and SNR on GPS L2C at 30-second intervals, to assess their potential in the detection of threatening phenomena, particularly severe rainfall. To achieve this goal, two GNSS stations located in eastern (LEOP) and western (WOFU) Austria are utilized. The experimental period covers four intense rainfall events in July 2021 and August 2024 at the location of these GNSS stations. Moreover, surface rainfall data and slant-accumulated rainfall derived from weather radar measurements are employed to determine severe rainfall cases and for further correlation analysis with SNR data. The results revealed that PWV is considerably elevated above its baseline ahead of rainfall onset. Nevertheless, the precise lead time of this build-up varies between events. On the other hand, the detrended SNR exhibits significant fluctuations and a pronounced level drop within approximately 30 to 60 minutes preceding rainfall onset. This is most obvious when the satellite signal passes directly through the approaching weather. Correlation coefficients for accumulated rainfall along the signal propagation path and SNR reach values up to 87%. The sign and strength of this correlation are found to depend on the relative geometry and timing between the moving rain field and the satellite line of sight. On the basis of these results, PWV and SNR both show strong potential as real-time rainfall indicators.