For the needs of healthy tourism development, based on the daily meteorological observation data at Lijiang of Yunnan Province, Xishui and Jinping of Guizhou Province, the characteristics of climate comfort degree were analyzed based on human body comfort index caculated by ‘the Golden Section method’ and its classification in the three famous tourist cities. And on this basis the concentration degree of generalized comfort period in the three cities and their complementarity in time were studied. Finally, combined with the relevant disease data, the effect of climatic health care was explored. The results show that the generalized comfort period was longer in Lijiang and Xishui cities, which concentrated from February to November and March to November, respectively, and the climate comfort degree was high and the number of patients with respiratory diseases was low in each month of summer, so the effect of climatic health care was significant in summer. The climate in spring and autumn was comfortable in Jinping County, the climate comfort degree was higher in April, May, September and October, and the incidence rate of diseases was relatively low during the generalized comfort periods. In short, the effect of climatic health care on respiratory diseases was the best in summer in the three places, followed by circulatory system diseases, and the optimal climate comfort periods and health care effects among the three places were complementary in time, which was beneficial to clusters of climatic health care and tourism. The above results could provide some references for the choice of health care and tourist time in the above three places.

Research on highway traffic disaster risk is a new direction for the development of professional meteorological services, and it is an important content of traffic weather forecasting services in future. Based on the ground observation data at 94 weather stations in Shaanxi from 1980 to 2017, the temporal-spatial change characteristics of road icing under different weather conditions were analyzed by using EOF method, etc., firstly. And the risk warning model of road icing in Shaanxi was discussed and established. The results show that there were two centers of road icing in the northwest of Guanzhong to the west of northern Shaanxi and Shenmu of northern Shaanxi, and the ratio of sleet or snowfall road icing was the maximum. The road icing mainly occurred from November to next March in Shaanxi, and the beginning date of road icing generally occurred in mid-to-late October in northern Shaanxi and Guanzhong, while that occurred later in November in southern Shaanxi, but the ending date of road icing mainly occurred in March and April in three regions, and the road icing days in January was the most. After the 1990s, the days of road icing decreased. The risk warning model of road icing was established and applied to business system in Shaanxi, and it could complete automatic production, the prediction effect was better.

Based on the observation data from Shaanxi expressway automatic meteorological stations including the road surface temperature, air temperature, relative humidity, wind speed and cloud cover of reanalysis data of ECMWF from August 2013 to December 2017, the distribution characteristics of the road surface temperature in different seasons and sky conditions were analyzed, and relations between the road surface temperature and meteorological factors were studied, and relevant multivariate regression equations were established. The results show that the road surface temperature had obvious diurnal variation in different seasons. It was easier to freeze on the road surface from 00:00 BST to 08:00 BST after snow. The air temperature was one of the most important factors affecting the road surface temperature. By fitting the measured and caculated values of the road surface temperature, the results show that the model fitted the minimum road surface temperature well in winter with the correlation coefficient above 0.94, and the standard deviation was less than 1 ℃, and the frequency of difference between the observed and mesured temperature ranging from -2 to 2 ℃ was 98%. In addition, the model fitted better when the road surface temperature was 0 ℃.