Abstract: In order to evaluate the spring tea frost disaster with meteorological data, tea tree canopy temperature observation data and the meteorological observation data of tea plantations in 2009 and 2010 were used to analyze the relations of the lowest temperature differences of the tea tree canopy, the weather patterns and the micro meteorological factors. A fitting model of different tea tree canopy temperatures was built to evaluate the spatial distribution of the lowest temperature of the tea tree canopy in the frosting process combined with the intensive automatic weather station data. The results showed: the weather type was the main factor that affected the lowest temperature difference between the lowest air temperature and the lowest temperature of the tea tree canopy, while the lowest difference was also affected by the wind, humidity and other micro meteorological factors, but the lowest temperature difference between the lowest air temperature and the lowest temperature of the internal leaves in the tea tree canopy was less affected. The relations among weather type, micro meteorological factors and the lowest temperature difference between the lowest air temperature and the lowest temperature of the tea tree canopy were non-linear, and the supporting vector machine model could achieve a better imitative effect. As for the same kind of tea tree canopy, the imitative effect of the lowest temperature in external canopy would be superior to that in the internal. The weather station data, GIS technology and the fitting model of the lowest temperature difference between the lowest air temperature and the lowest temperature of the tea tree canopy were utilized to evaluate the spatial distribution of the tea frosted.