Monitoring And Forecasting Of Surface Settlement In Foshan City Based On Time Series InSAR Technology

Surface subsidence monitoring and analysis is an important engineering discipline for the understanding of deformation laws and the assessment of the health of urban ground structures.In recent years,due to the unique soft soil characteristics of the Pearl River Delta region and the rapid development of urbanization,geological disasters have occurred on a frequent basis,However,most of the existing research focuses on the highly developed core cities of the Pearl River Delta.While the surface subsidence problem of the neighboring city of Foshan is becoming more prominent,the relevant research is relatively limited.In order to gain a more comprehensive understanding of the subsidence of the land in the city of Foshan,and predict and prevent potential geological hazards,to better ensure the safety and stability of the surface of Foshan.The emerging time-series InSAR technology is being widely used in urban surface monitoring due to its advantages of all-weather capability,high accuracy and high resolution,which can be used for continuous monitoring of large areas.To this end,this paper uses 42 Sentinel-1A images to invert the surface information of Foshan City from January 2019 to June 2022 using PS-InSAR and SBAS-InSAR technologies,and discusses the deformation induced by the combination of multiple data,establishes the predictive model for key settlement areas,performs accurate analysis,and predicts the future land surface settlement trend.(1)The average deformation rate and temporal shape variables of Foshan city from January 2019 to June 2022 were obtained on the basis of the inversion of PS-Insar technique and SBAS-InSAR technique,integrating PS points.The time series shape variables and spatial distribution obtained by the two independent time series InSAR techniques show a high degree of consistency.Three same-named points are selected to compare the two monitoring results,and correlation coefficients are all above 0.9.The experimental results show that most of the surface of Foshan City,the major deformation rate range of-5～5mm/yr,and the identification of three major settlement areas.(2)For the three major settlement areas,GIS software was used to analyze the temporal and spatial variation characteristics of land subsidence in Foshan City,and combined with natural factors and human factors,the surface deformation characteristics and causes of land subsidence areas in Foshan City were explored.The results showed that land subsidence was closely related to changes in precipitation characteristics,hydrogeological environment and land use type,and that subway and construction engineering were the main causes of land subsidence,and the collapsibility and compressibility of soft soils in Foshan were the internal causes of internal subsidence in Foshan City.(3)In view of the nonlinear and unstable characteristics of land subsidence,the BP neural network prediction model is built for the prediction and analysis of the three major settlement areas in Foshan city.The first 32 periods of data were trained on the basis of the42 periods of data monitored by InSAR technology,and the settlement of the last 10 periods was predicted.The data of the last 10 periods of the study area were predicted according to the model,and the results showed that the model showed high accuracy in predicting the deformation of Foshan,which could well reflect the overall settlement trend of the study area.The results showed that the study area showed a stable state in the last 10 periods.