Precipitation type estimation and validation in China

The results from three methods aimed at improving precipitation type (e.g., rain, sleet, and snow) estimation are presented and compared in this paper. The methods include the threshold air temperature (AT), threshold wet bulb temperature (WBT) and Koistinen and Saltikoff (KSS) methods. Dot graphs are plotted to acquire the threshold air temperature or the threshold wet bulb temperature using daily averaged air temperature, wet bulb temperature and precipitation data at 643 stations from 1961 to 1979 (precipitation types are not labeled in the database from 1980 to present) in China. The results indicate that the threshold AT or WBT methods are not able to differentiate rain, sleet and snow in the most regions in China; sleet is difficult to differentiate from other precipitation types based on the two threshold methods. Therefore, one threshold AT and WBT method was used in this study to differentiate rain and snow. Based on Gaussian-Kriging interpolation of threshold air temperature (T ₀) and wet bulb temperature (T _w), the T ₀ and T _w contour lines and contour surfaces are calculated for China. Finally, a comparison between the KSS, AT and WBT methods are provided in which the KSS method is calculated based on air temperature and relative humidity. The results suggest that the KSS method is more appropriate for water phase estimation than are the other methods; the maximum precision for rain and snow is 99% and 94%, respectively. The AT method performs better than the WBT method when the critical air temperature is 2°C.