Monitoring And Prediction Analysis Of Surface Subsidence In Urban Built-Up Areas Supported By InSAR Technology

With the development of cities,the influence of urban infrastructure construction and groundwater exploitation has made surface subsidence disasters common in cities,and when large area subsidence occurs in urban areas,it will not only bring negative impact to roads,bridges,housing and building facilities,underground infrastructure,etc.,but also threaten the lives of urban residents in serious cases.InSAR technology has replaced traditional settlement monitoring methods as one of the main methods for monitoring urban settlement on a large scale because of its all-weather,all-day,high accuracy and large range characteristics.In this paper,based on the 37-view Sentinel-1A SAR data,PS-InSAR technology and SBAS-InSAR technology are used to invert the surface settlement in the built-up area of Nanchang city and along the metro line,and the monitoring results are verified,analysed and predicted as follows.(1)Using PS-InSAR technology to monitor the settlement in the urban built-up area of Nanchang,the results show that the settlement areas in the urban built-up area of Nanchang are mostly concentrated along the Ganjiang River and there are two key settlement areas,and the settlement formation is caused by the discharge of groundwater to the Ganjiang River and the Fuhe River and the construction of the project.Later,the settlement monitoring results of SBAS-InSAR technology were compared with those of PS-InSAR technology,the results show that the monitoring results of these two technologies were highly consistent,which verified the reliability of the monitoring results of PS-InSAR technology.(2)Based on the settlement monitoring results of PS-InSAR technology,a buffer zone along the metro line was established to extract the deformation rate field along the metro line in Nanchang,and the results showed that the metro lines 2 and 4 in Nanchang passed through two key settlement areas,and the four metro lines had different settlement distributions,and the settlement sections of lines 2 and 4 were longer than those of lines 1 and 3,and the causes of their settlement formation were related to the areas along the metro lines The causes of settlement are related to the areas along the subway lines and their own construction.(3)Based on the 36-period cumulative deformation time series data obtained from the deformation inversion of Nanchang City,a BP neural network model and a wavelet neural network model were used to predict the 36-period cumulative deformation time series data before and after denoising for two major subsidence areas.The results show that the prediction accuracy of the wavelet neural network model is better than that of the BP neural network model in the prediction experiments between the original data and the denoised data,and therefore the wavelet neural network model has a good application prospect in the field of deformation prediction.