Surface precipitation phase (SPP) discrimination at the ground level (rain or snow) is a key step in numerous meteorological and hydrological applications. Previous studies have undertaken this by comparing surface observations, such as air temperature, relative humidity and wet-bulb temperature, with concurrent present weather observations of the precipitation phase to derive thresholds for discrimination purposes. The first objective of this study was to examine schemes for precipitation phase discrimination at the ground level, using data interpolated from a network of automatic weather stations covering an area of complex terrain. The second objective was to combine the SPP interpolated fields with precipitation estimates from single-polarisation weather radar, which provide precipitation occurrence information but not precipitation phase type, to obtain a real-time SPP product. Finally, the third objective was to evaluate the role of citizen science and crowd sourced observations in the monitoring of snowfall events with SPP schemes. Results from nine cold seasons (Oct-May) indicated that out of the seven SPP schemes tested against 7,702 quality-controlled present weather observations in Catalonia (NE Spain), those including information on air saturation conditions provided the best results. A wet-bulb temperature threshold of 0.7 °C produced the best discrimination for snow vs no snow, with a Pierce skill score of 0.77. Finally, the SPP product was used with two case studies, demonstrating its added value and pending challenges for real-time applications.

https://doi.org/10.1016/j.jhydrol.2020.125780