Identification And Monitoring Of The Loess Landslide By Using Of High Resolution Remote Sensing And InSAR

Loess is widely distributed in China, which occupies 0.64 million square kilometer in 12 provinces, among which 0.4 million square kilometer is loess plateau, accounting for one fifth of the national arable land and one sixth of the people live in this place. Besides, it is an important base of oil and gas resources. But in the region, topography is complex, the geological environment is increasingly fragile, and geological disasters frequently occurred, such as land subsidence, ground fissures, collapse, landslide and mudflow. Each year almost a third of the national geological disaster occurred there, which seriously restricts the living security in the region, urbanization construction and energy development. Therefore it is urgent to carry out the early identification of loess disasters, precise monitoring, forecasting and effective prevention and controlling in a timely manner. This paper selects typical loess landslide as the research object, that is Jingyang southern plateau in Shaanxi province, and takes both high resolution remote sensing data and InSAR technique to do the theoretical and experimental research of early identification and monitoring of loess landslide.In this paper, due to the characteristics of loess landslide of Jingyang southern plateau, such as wide distribution, long history, frequent occurrence and small and uneven single landslide, I carry out the recognition research of the loess landslide over a large area, and analyze the landslide types and the characteristics in time and space domain. At the same time, as for the characteristics of individual landslide, such as small spatial extent, occurrence, sudden, difficult to accurately capture the landslide onsaet and accelerate point, I propose the research and test of high precision monitoring of typical landslide deformation, based on the high resolution InSAR technology.The detailed research contents and main research achievements are as follows:1) I developed multi-temporal optical remote sensing technology to map the landslide inventory, and scenes of 32 optical images are chosen from January 28, 2003 to February 6, 2016, and finally multi-temporal landslide inventory maps and landslide inventory maps for different types over Jingyang southern plateau landslides group are achieved. Based on the multi-temporal landslide inventory map, it can be noted that the loess landslide occurred at March to April in each year, which is mainly associated with irrigation and melting after winter, while have no direct relationship with the rainfall. Further, can be found, it can be found that the loess landslide occures in group, for example, along from Jiangliu to Miaodian village, with less than 10 km distance, 64 landslides occurred with much high frequency. Closely analysis shows the landslides evolution is two phases. The first one is mainly irrigation flow slide, the second phase is flow slide or sliding loess landslide, which is the main loess landslides types.2) Then, key technique process and parameter settings of monitoring and identification of small scale loess landslide, is studied by using high spatial-temporal resolution SAR data, including the setting of multi-look number and coherent threshold, the correction of phase unwrapping error and DEM error, and the selection of high quality interferograms, which ensuring the reliability of InSAR deformation time series estimation.3) Finally, the advantages of InSAR technique, that is high resolution and short revisiting period of TerraSAR data, we systematically monitor and analyze the Xingyuan landslide for pre-, co- and even post-slide stages. The accurate pre-slide time series deformation is firstly obtained, and is analyzed by combining with the underground water level data characteristics of the pre-slide time series deformation. Moreover, the starting and accelerating time point of this landslide is successfully uncovered, which have significant guideline for prediction and forecast of the any other similar loess landslide. By analyzing intensity before and after landslide, we accurately achived Xingyuan landslide accumulation area and source area, which is consistency with the field survey results. Lastly, the post-slide surface change is calculated from the interferograms obtained after landslides with InSAR technique, which shows high corretation with the real changes of the surface.