Abstract:The droplet spectrum variation characteristics and snow depth forecast of the two snowstorms in 14-15 December 2010 and 20-22 January 2012 are studied by using Thies Clima laser precipitation monitor (TCLPM) and artificial observation data from Danjiangkou Station. The results show that: (1) the TCLPM can automatically identify the precipitation phase state. Combined with surface artificial dense observations, the temperature higher than 0.7℃ is for precipitation phase state, less than 0.7℃ is for sleet, below -0.5℃ is for pure snow. At the same time, when the surface temperature is below 0.5℃, snow begins to accumulate. The surface wind speeds in the two processes are relatively slow, conducive to snow accumulation on the ground; (2) the TCLPM can monitor the variation of droplet spectrum characteristics of heavy snowfall weather, the echo intensity (Z), average diameter (Dm), water content (W), snowfall particle number concentration (N) which increases with the enhancement of snowfall intensity. During the two snowstorm events, there are different levels of positive correlations among Dm, Z, N, W and VSD, better correlation between W and VSD, respectively reaching 0.844 and 0.926; (3) Selecting the first order fitting snow particles in water content W and surface snow rate, the snow rate forecast equation is derived compare surface snow rate forecast (VSDF) in the pure snow stage and the surface snow depth forecasts (SDF) to VSD, SD retrieved with the observation data by TCLPM it is found that they are very close, indicating this method can be used to estimate VSD and SD on the surface snow rate and snow depth, which can reflect the main time period with quick increase of surface snow.