Abstract: Thermal spectrum mapping is a kind of methods that utilizes thermal spectrum fingerprint of the road to reflect the spatial variation of road surface temperature at winter night under different weather conditions.In this paper,a spatial variation model of thermal spectrum fingerprint under three weather conditions from Taohuayuan to Heishui road section was established based on the data of road surface conditions and road surface temperature at winter night at the road section from Taohuayuan to Heishui in Baomao highway in 2015.The characteristics of thermal spectrum fingerprint under three weather conditions and their correlation,similarity and reliability,were analyzed.The results show that there is a clear difference in the change range of thermal spectrum fingerprint under different weather types at the road section from Taohuayuan to Heishui in Baomao highway in 2015.The standard difference of road surface temperature is an important indicator to distinguish the types of thermal spectrum fingerprint.The similarity degree of thermal spectrum fingerprint under different weather types is high,with similarity coefficient in a range from 0.78 to 0.92.There are strong correlation in the thermal spectrum fingerprints under different weather types,with the significant level of 0.01.The similarity and correlation degrees of thermal spectrum fingerprint under the same kind of weather are higher than those under different weather types.The pattern of spatial variation in thermal spectrum fingerprint can be used to effectively describe the change in surface temperature of regional road networks in winter night.It provides a reference for meteorological station planning in highway and winter road management and control decision.