Abstract: Three heavy snow processes occurring in Liaoning province on March 4,2007,February 12,2009,and February 21,2017 respectively,were investigated based on the ECMWF (European Center for Medium-Range Weather Forecasts) ERA-interim reanalysis data,convention observations and echo data from the Doppler radar.The similarities and differences of the large scale circulation background,air mass source,vapor source and characteristics of the radar echo parameter of these three processes were studied.The results indicate that the three events are all caused by the influences of the upper-level trough.The upper-level trough cooperated with the surface inverted trough or cold front strengthens the dynamic lift.The air masses from different water vapor sources and the convergence of cold and warm air are the key factors to strengthen the snowfall processes.The reflectivity of the snowfall process is usually below 40 dBz,and the maximum echo depth is less than 10 km.The maximum reflectivity (Z_max) and the average value of the radar reflectivity larger than 15 dBz (Z_mean15) at each level within the thunderstorm correspond very well with the process of snowfall and reveal the intensity change of the systems and the drop conditions of the precipitation particles.It is important to the forecasting of rainfall.