Abstract: With global warming and rapid urbanization, urban climate is considered to be one important factor impacting urban ecological environment, and the most obvious feature of urban climate is the surface urban heat island(SUHI). The accurate division of rural reference area is critical in evaluating the intensity of SUHI in mountainous cities by urban-rural dichotomy. Therefore, a comprehensive method is proposed based on multi-source satellite remote sensing data to solve this problem, and the spatiotemporal variation of the SUHI in Chongqing metropolitan circle from 2001 to 2020 are analyzed. The results show that the spatial distribution of rural reference area obtained by the comprehensive buffer method is obviously different from that by the traditional buffer method, and the main cause is the limitation of the elevation difference between urban and rural reference area. There are a large number of "false SUHI" pixels in the SUHI product estimated by the traditional buffer method, which makes the surface urban heat island regions obtained by the traditional buffer method significantly larger than that by the comprehensive buffer method. The spatiotemporal distribution results show that the regions affected by the SUHI are concentrated in the core area of Chongqing, Changshou, Fuling and nearby regions, and the cold island are concentrated in the high-altitude areas in the southeast of Chongqing and some mountains in the core of Chongqing. The interannual variation of the SUHI shows that the proportion of areas above strong SUHI level increases with time from 2001 to 2020, and the SUHI is strong in summer and weak in winter. The proportion of strong SUHI area increased from 0.5% in 2001 to 2.95% in 2020, with an average annual growth rate of 0.12%. The proportion of intense heat island area increased from 0.21% in 2001 to 0.78% in 2019, with an average annual growth rate of 0.03%. The change of SUHI is closely related to urban development, and has a significant positive correlation with the economic driving factors such as the total population of the main city and metropolitan area, urban population, GDP, total energy consumption, urban built-up area and civil vehicle ownership. The correlation coefficients are 0.91, 0.94, 0.95, 0.94, 0.90 and 0.96, respectively, indicating that the driving factors of urban economy play an obvious role in promoting the change of SUHI intensity in the main metropolitan area.