Abstract: The surface features in urban areas are very complex and have significant spatial variations.As a result,the subsidence monitoring measures without differentiating surface features can hardly precisely reflect the subsidence characters of urban area since they are influenced by various factors.Previous subsidence monitoring with the PS-InSAR technique is unable to conduct attribute classification for PS monitoring points due to the low spatial resolution of SAR images.In this paper,the high density PS monitoring points in Tanggu research area are firstly extracted using the 1m spatial resolution TerraSAR-X images.Based on various industrial classification standards,the PS attributes for different urban surface features are then divided into different categories to facilitate complex urban subsidence analysis.SAR tomography technique is adopted to deal with the classification error of PS monitoring points for high buildings.The attribute of each PS monitoring point is automatically recognized with spatial analysis and the specific subsidence characters of each kind of surface feature are discussed.As a result,the subsidence rate of soil surface is the highest while the residential area and bridges are the lowest.This preliminary research shows that 1)high spatial resolution TerraSAR-X images make it possible to carry out attribute classification on PS monitoring points and further to carry out subsidence monitoring on different surface feature; 2)it helps us understand the mechanism and hazard of different surface features induced by their own influencing factors, and present us the scientific references to land subsidence monitoring,warning and risk assessment during urban constructions.