Abstract: Evaluation of urban thermal comfort is the important means of assessing the quality of thermal environment for cities, as well as provides the scientific basis for improving urban thermal environment. The temperature-humidity index (THI) is an important indicator of thermal comfort indexes. In this paper, taking Nanjing City, capital of Jiangsu Province as the study area, we analyzed the impacts of landscape patterns on the dynamics change of thermal comfort. Four Landsat images of 1994, 2000, 2010 and 2013 were used to retrieve the land surface temperature (LST) and normalized moisture Index (NDMI). And the modified THI (MTHI) of Nanjing was calculated through LST and NDMI. The paper investigated the dynamic changes of thermal comfort in Nanjing City from the macro and micro scales during 1994 to 2013. Landscape metrics were calculated to explore the changes of various landscape patterns based on the different thermal comfort levels. The correlation between MTHI in virtue of 2.5 km fishnet and the composition and structure of the urban landscape was quantified, which can clarify the relationship between urban thermal comfort and urban landscape patterns on microscale. The results showed that the impacts of landscape patterns on thermal comfort in Nanjing has been increasing from 1994 to 2013 from the macro perspective, which was getting worse. The compositions and structures of different landscapes exert different impacts on urban thermal comfort from the micro perspective, among which cropland, forest, especially water can relieve the urban thermal comfort. The more proportion and greater concentration of cropland, forest and water body, the better they can improve the urban thermal comfort. Conversely the thermal comfort got worse when there were more aggregative built-up areas.
Key words：Urban thermal comfort; The modified temperature-humidity index (MTHI); Landscape metrics; Dynamic changes; Nanjing city