Monitoring And Prediction Analysis Of Surface Subsidence In Shanghai Urban Area Based On Time Series InSAR Technology

Surface subsidence is one of the main geological disasters affecting the sustainable development of cities,which not only damages urban infrastructure,but also restricts urban economic development.As an international financial center,Shanghai has a large urban construction and population density.With the planning of aboveground and underground projects and fragile geological conditions,there is a potential risk of surface subsidence in this area.In order to grasp the change law of urban geological disasters,it is of great significance to monitor and predict the surface subsidence in urban areas in time.As two main traditional surface deformation monitoring technologies,Global Positioning System(GPS)and ground leveling usually require a long monitoring time,which is difficult to meet the current needs of large-scale surface deformation monitoring.Time series Interferometric Synthetic Aperture Radar(InSAR)technology has the characteristics of all-weather,high efficiency and low cost,which makes it a research hotspot of urban surface deformation monitoring.In this paper,the time series InSAR technology is used to monitor the surface settlement in Shanghai.Based on the monitoring results of surface settlement,the average settlement change trend in the area with serious settlement is predicted.The specific research contents are as follows:1.With 27 scenes Sentinel-1A data and external 90 m resolution Digital Elevation Model(DEM)as the data support,the time series deformation monitoring is carried out by SBAS-InSAR and PS-InSAR technology,and the subsidence rate and distribution of Shanghai urban area are obtained.The two monitoring results are cross verified and analyzed to explore the reliability of the two methods in urban surface settlement monitoring.The results demonstrate that within the study area,the spatial distribution and shape variable level of the monitoring results of the two methods are roughly the same.However,the continuity of SBAS-InSAR is significantly better than that of PS-InSAR,which is in terms of monitoring results,the reliability of the monitoring results of SBAS-InSAR technology is higher than that of PS-InSAR on the whole.2.Based on the subsidence rate and distribution of Shanghai urban area obtained by SBAS-InSAR technology,the temporal and spatial characteristics of the key settlement areas are analyzed and the causes of deformation are discussed.The monitoring results show that the surface deformation rate of Shanghai urban area is mainly concentrated in [-4,4] mm/a during the monitoring period,and the surface deformation is relatively stable as a whole.However,there are three key settlement areas in the northern area of Hongqiao Airport,the area near the foreshore and the area around Haojingyuan.Among them,the Sifangcheng development area near the foreshore has the most serious surface subsidence,with the accumulated subsidence exceeding 50 mm and the maximum subsidence rate of-25 mm/a.The discussion of deformation inducement shows that the settlement funnel has a high spatial correlation with the spatial distribution of residential area,engineering construction and water-rich area.The surface subsidence in Shanghai is mainly affected by comprehensive factors such as engineering construction,urbanization development and geological water-rich conditions.This result provides real-time early warning information for geological disaster prevention.3.A new integrated prediction model of Grey Neural Network is studied by combining the advantages of the Grey prediction model and Neural Network prediction model.The points with large deformation in the key subsidence areas obtained by SBAS-InSAR technology are selected as the characteristic points,and a 40 m buffer zone is constructed to form the characteristic areas of subsidence prediction.A new set of subsidence sequence is generated by calculating the average subsidence of all subsidence points in each phase in the characteristic areas.Three prediction models are used to predict the average subsidence in the later five stages of the characteristic area,and compared with the new subsidence sequence.The results demonstrate that the maximum Root Mean Square Errors of Grey Model,Neural Network Model and Grey Neural Network Model are1.94,1.52 and 0.49 respectively,indicating that the accuracy of Grey Neural Network is better than the other two prediction models.The Mean Absolute Error and Root Mean Square Error of the central urban area are 0.43 mm and 0.15 respectively when the Grey Neural Network model is used to predict the average subsidence of the three characteristic areas,which are less than the other two characteristic areas.It is indicating that the closer to the central urban area,the better the prediction effect,which verifies the applicability of the Grey Neural Network Model to predict the surface subsidence of the urban area.