Land Surface Temperature(LST)is a key process parameter of global change,and its importance is being increasingly recognized.With the increase of remote sensing data sources,HJ-1 Bsatellite data has short return cycle and large sensing width is used for land surface temperature retrieval.This paper is based on the HJ-1B/IRS thermal infrared data,using the methods of Radioactive Transfer Equation(RTE),Qin Zhi-hao mono-window algorithm(Qins)and generalized single-channel algorithm(JM&S)to retrieve LST in Nanjing,analyzed the results of the three algorithms and use the same phase MODIS land surface temperature product to analyze temperature retrieval results.The results show that three algorithms have consistency.Compared with the MODIS temperature products,the LST has the closest retrieval results with it and lower than 2.1K when RTE is used;Qinsretrieval result was only slightly lower 3.87K;and retrieval result of JM&S significantly lower than 5.96K for the MODIS temperature products.Combined with the land use type map to analyze the spatial distribution characteristics of the surface temperature of Nanjing,it can be seen that the difference between highest temperature of construction land and the lowest temperature of water is 4.1K when RTE is used;the difference between construction land and water is 4.38K when Qinsis used;the difference between the highest temperature of construction land and the lowest temperature of water is 2.15K when JM&S is used.The temperature results of the RTE and Qinscan reflect the temperature difference among the different land use types and its contribution to the urban heat island.