Monitoring And Prediction Analysis Of Urban Surface Subsidence Based On Time-series InSAR Technology

It is of great practical significance to monitor land subsidence,due to the great harm to urban development and the safety of human life and property caused by land subsidence.Traditional monitoring methods are time-consuming,labor-intensive,low-efficiency,highcost,at the same time,it is incapable of achieving large-scale measurement.Time series InSAR technology has gradually become an important method for urban surface monitoring,because of its advantages of all-weather,wide monitoring range,high precision and acquisition of time series surface subsidence.This paper used two different time series InSAR techniques to monitor the surface deformation of urban areas in Hangzhou and Wenzhou,and to verify and analyze the monitoring results.The main contents and results are as follows:(1)The surface deformation information was inverted,using the 25 scene Sentinel-1A images covering the Hangzhou area based on the PS-InSAR and SBAS-InSAR methods.As a results,between January 2019 and January 2021,it is demonstrated that the average annual deformation rate in the study area ranges from-27.5mm /a to 28mm/a.The land subsidence area is mainly distributed in Xiaoshan area and Jianggan area;The overall surface morphology in the study area tends to be stable,and the average annual deformation rate is concentrated in-5～5mm/a.The uneven ground settlement in Hangzhou is mainly related to high-rise building load,urban development projects and industrial activities.(2)The surface deformation was monitored from May 2019 to May 2021,in the urban area of Wenzhou using 60 scene Sentinel-1A data based on the SBAS-InSAR technology.The results suggested that during the observation period,the average annual deformation rate in the Wenzhou area is-35～21mm/a,and there is significant subsidence in some areas.The subsidence is mainly distributed in the densely populated plain areas in the south of the Oujiang River,and the largest subsidence area is located in the area along the Oujiang River.Combining the measurement data and PS-InSAR method monitoring results to verify its monitoring results,it is found that the three are consistent.For the subsidence area,three representative areas were selected for detailed analysis.The deformation information is combined with groundwater data and engineering monitoring data to investigate the correlation between surface deformation and groundwater utilization and engineering construction in Wenzhou City.(3)Based on the monitoring results of surface deformation in Wenzhou and Hangzhou,the subsidence prediction model with BP neural network principle was established.Taking the deformation data of the first 50 periods of monitoring results of Wenzhou as a sample,the BP neural network model was trained to predict the surface deformation information of the next 10 periods.The same method is used to predict the deformation of Hangzhou city.The results showed that the model has high accuracy in predicting the surface deformation of Wenzhou and Hangzhou city,which proves the applicability of the model in the prediction of surface deformation.Subsidence prediction is made for unknown data in later period.