AUTHOR=Xie Ruiheng , Liao Chunhua , Luo Xiao , Guo Haifeng , Huang Zequn , Peng Weiying TITLE=Research on road surface temperature characteristics and road ice warning model of ordinary highways in winter in Hunan province, central China JOURNAL=Frontiers in Earth Science VOLUME=11 YEAR=2023 URL=https://www.frontiersin.org/journals/earth-science/articles/10.3389/feart.2023.1251635 DOI=10.3389/feart.2023.1251635 ISSN=2296-6463 ABSTRACT=

The study of road surface temperature (Ts) characteristics in winter and the early warning method of road icing is of great significance to reduce traffic accidents and improve transportation efficiency. Using the hourly observation data of Hunan traffic meteorological stations from December 2020 to February 2022, this study analyzes the winter Ts characteristics of ordinary roads in Hunan Province, and uses the Logistic regression model to establish the temperature threshold for icing of ordinary roads in the province. So as to build a road icing early warning model hierarchically. The results show that the Ts in southern Hunan is relatively high, the Ts at most stations is above 10 °C, and the low Ts area is in western Hunan, and the stations below 8 °C are mostly distributed in this area. This may be due to the higher altitude in western Hunan. In terms of diurnal variation, the lowest value of average Ts and air temperature (Ta) in Hunan Province in winter both appeared at 7:00 Beijing Time (BT), while the highest value appeared at 15:00 BT, and the average Ta is always lower than the Ts. The temperature variation on the bridge surface is more pronounced. When the Ta is lower than −2.5 °C, more than 70% of the sites have a rapid increase in the risk of icing; and when the Ta is lower than −5°C, nearly 87% of the sites have a risk level of 4, which means icing risk is extremely high. Furthermore, combining the warning model with thermal spectrum mapping can improve the spatial resolution of the warning model and also solve the problem of lack of observations in some areas.