By using the daily temperature from Jan 1955 toApril2008, thewintermean temperature, cold days, cold accumulated temperature and climate abrupt change aswell as periods ofwinter temperature in Shijiazhuangwere analyzed. The variation of extremely cold(warm) days in winter and the distribution of severe cold (warm) winters of Shijiazhuang under the background of the warming winterwere discussed also. The results show that thewinter temperature especially the lowest temperature increased obviously in the recent53 years, and the cold days and the cold accumulated temperature reduced obviously. W inter temperature changed abruptly in the 1990s. Nearly twenty days shortened inwinter in recent53 yearsmainly for ahead of time ofwinter termination. The extreme cold days decreased evidently and yearly variation of the extreme high temperature days increased inwinter. The severe coldwinter occurred before the 1980s, and the severewarm winter occurred from the later1990s to the early 2000s.

The air quality dynamic prediction system in Yinchuan city is given out simply，and the prediction result and error about the system are analyzed in detail. Results show that the system can forecast pollution meteorological condition in 24 hours ，and the prediction result corresponds to monitoring result well，the prediction accuracy for PM10 and SO2 is 61% and 92%，respectively；the average absolute error of predicted value for PM10 is relatively less in winter，summer and autumn，but it is contrary to that of SO2；Considering the influence of blow sand condition on predicting accuracy of the system，the blowing sand coefficient in the model can be adjusted progressively according to different time and seasons，thus it can improve prediction accuracy of the model for air quality in windy and dusty season effectively，so as to be adapted to the special climate and environmental condition in Yinchuan city.