Combining commercial microwave links and weather radar for classification of dry snow and rainfall

Abstract. Differentiating between snow and rainfall is crucial for hydrological modeling and understanding. Commercial microwave links (CMLs) can provide rainfall estimates for liquid precipitation but show minimal signal attenuation during dry snow events, causing the CML time series during these periods to resemble non-precipitation periods. Weather radars can detect precipitation also for dry snow, yet they struggle to accurately differentiate between precipitation types. This study introduces a new approach to improve rainfall and dry snow classification by combining weather radar precipitation detection with CML signal attenuation. Specifically, events in which the radar detects precipitation but the CML does not are classified as dry snow. As a reference method, we use weather radar, with the precipitation type identified by the dew point temperature at the CML location. Both methods were evaluated using measurements from disdrometers located within 8 km of a CML, taken as ground truth. The analysis used data from Norway, including 550 CMLs in December 2021 and 435 CMLs in June 2022. Our results show that the use of CMLs can improve the classification of dry snow and rainfall, presenting an advantage over the reference method. In addition, our research provides valuable insight into how precipitation at temperatures around 0 °C, such as sleet or wet snow, can affect CMLs, contributing to a better understanding of CML applications in colder climates.