Abstract: Under the context of global warming,to understand the response of snowfall to climate change,the double temperature threshold method is used in this paper to extract snowfall series,based on daily meteorological data from 43 meteorological stations in the Haihe River Basin from 1960 to 2016. In addition,the spatial and temporal variation characteristics of snowfall are analyzed according to the snowfall levels as determined by the national meteorological departments. The results show that（1） the snowfall at each meteorological station can be effectively estimated at annual scale through the snowfall identification index equation,for which the correlation coefficient in the periodic and verification periods is above 0.90 and the relative error is within ± 5%.（2） In terms of the time distribution,the interannual variation of snowfall at different levels shows a less obvious decreasing trend,but the interannual variation trend of each geomorphic type area is different.（3）In terms of the spatial distribution,the meteorological stations in Taihang Mountain Area and plain areas show a decreasing trend,while those in Yanshan mountain area and coastal area show an increasing trend.（4） By analyzing the correlation between snowfall and temperature and the temperature at different levels of snowfalls,it is found that heavy snow and Blizzard are positively correlated with temperature,while moderate snow and light snow are negatively correlated with temperature,and the temperature of snowfalls in plain areas is the highest and the lowest in Yanshan mountain areas.