Land Subsidence Monitoring Based On Time Series InSAR And Parameter Inversing Of Groundwater-Subsidence Prediction Model

With the large-scale urban construction and excessive groundwater extraction,the water level of cities has dropped,which caused a series of geological disasters such as serious surface settlement.To strengthen urban subsidence monitoring and establishing groundwater-land subsidence prediction model is an effective measure to prevent the occurrence of surface subsidence disaster.Using In SAR(Interferometric Synthetic Aperture Radar,In SAR)technology to monitor land subsidence of large scale urban can provide large-area,high-resolution and high-precision timing series subsidence monitoring results,which provided a large number of measured data for parameters inversing of ground subsidence prediction model.This thesis systematically analyzed the principle,processing flow and main error sources and sensitivities of conventional of In SAR and time series In SAR technologies.In view of the atmospheric effect in time-dependent In SAR surface subsidence monitoring technology,two kinds of atmospheric signal separation algorithms were studied,such as conventional spatio-temporal filtering and smoothing spline filtering.Atmospheric filtering was performed using conventional spatio-temporal filtering method and smoothing spline filtering method,respectively,to separate the atmospheric signal and deformation information.The atmospheric and deformation results obtained by the two methods were compared and analyzed.The experiment results shows that the filtering parameters of the conventional filtering method were fixed,which makes the filtering result uncertain under the condition of non-uniform sampling status,and smooth spline filtering can also effectively separate the atmospheric and sedimentary signals,furthermore,when the sedimentation rate is large,compared with the conventional spatio-temporal filtering method,this method can improve the accuracy of ground subsidence monitoring.Aiming at the parameter Inversion Problem of Groundwater-Ground subsidence prediction model,the correlation between model parameters and subsidence values wass analyzed,and the consolidation coefficient was obtained by LS(Least Square,LS)inversion method under In SAR land subsidence monitoring data's constraint.The surface settlement of elasticity and inelastic caused by the compressive deformation of aquifers in ground subsidence were separated by continuous wavelet transform,and elastic release coefficient and inelastic release coefficient were obtained by the inversion of ground subsidence prediction model.The parameters of groundwater volume variation were retrieved according to the volume strain formula under free stress in Biot's consolidation theory by using In SAR data.Taking the In SAR land subsidence monitoring in Tianjin as an example,the deformation of the land subsidence was obtained by the In SAR(Small Baseline In SAR)method with short baseline set.The relevant hydrogeological information such as groundwater level,groundwater system and basement structure based on the ground settlement value of PS point was interpreted.In SAR surface subsidence value is used to inverse the parameters of ground subsidence model.In SAR surface subsidence value is used to inverse the parameters of ground subsidence prediction model.The groundwater-ground subsidence prediction model of Tianjin is established.Modflow,ground subsidence prediction software was used to simulate the subsidence.The simulation results were compared with subsidence measurements the correctness of the model parameter inversion method was verified,and experiments show that the ground subsidence error is less than 10 mm,the maximum residual error of 15 mm.The main research results are as follows.(1).two kinds of atmospheric signal separation algorithms were studied,such as conventional spatio-temporal filtering and smoothing spline filtering.Atmospheric filtering was performed using conventional spatio-temporal filtering method and smoothing spline filtering method,respectively,to separate the atmospheric signal and deformation information.The atmospheric and deformation results obtained by the two methods were compared and analyzed.The experiment results showed that the filtering parameters of the conventional filtering method are fixed,which makes the filtering result uncertain under the condition of non-uniform sampling status,and smooth spline filtering can also effectively separate the atmospheric and sedimentary signals,furthermore,when the sedimentation rate is large,compared with the conventional spatio-temporal filtering method,this method can improve the accuracy of ground subsidence monitoring.(2).By studying the establishment of ground subsidence model and the theory of parametric inversion,the establishment process of three-dimensional seepage-ground subsidence prediction model of groundwater was analyzed,and the definitive solution conditions and convergence of coupled partial differential equations were discussed,and the correlations between several important model parameters in the model such asgroundwater level,elastic release coefficient,inelastic release coefficient,consolidation coefficient,extraction volume and settlement were analyzed.The contribution and sensitivity of each parameter to the settlement model were explained.(3).A inversion method of elastic and inelastic water release coefficient of aquifer based on In SAR technology was proposed.The deformation mechanism of confined aquifers was studied.Based on the vertical In SAR time series deformation components and water level variation factors,continuous wavelet transform(CWT)was used to separate the periodic deformation signal components and long-term trends,and hydrologic elastic and inelastic time series deformation information were respectively obtained,isolated signal components were used to estimate elastic storage coefficients,inelastic skeletal water release coefficients,and compaction time constants.thus accurate elasticity,non-elastic water release coefficient were calculated,which provide more accurate parameters,which can provide more accurate parameter basis for the subsequent hydrogeological research.The least square method was used to establish the constraint coefficient of consolidation coefficient under the In SAR time series data source,and the inversion method of the aquifer consolidation coefficient based on In SAR timing technique was proposed.Combined with the volume strain model under free stress in Biot's consolidation theory,the sensitivity of In SAR to groundwater exploration under different conceptual models of hydrogeology was discussed.The vertical component of satellite's perspective component vector was incorporated into the Biot model so as to reach the volume strain integral equation to solve convergence,groundwater volume change rate of the purpose of the inversion.(4).In order to verify the accuracy and practicability of the In SAR-based inversion method of groundwater-subsidence prediction model proposed in this thesis,taking Tianjin subsidence area as an example,SBAS-In SAR technology was applied and the main subsidence area and single hopper area were monitored.and the qualitative and quantitative methods were used to interpret the land subsidence characteristics,the groundwater level,the groundwater system and the geological structure of the basement based on the PS point.Based on the proposed inversion method above,the elastic release coefficient,inelastic release coefficient,recovery of production and consolidation coefficient were inversed for groundwater-land subsidence prediction model,and the groundwater flow was simulated by the software of Modflow,subsidence was simulated by the Interbed-Storage package.First,a single-point parameter of the initial water level geology was used as the initial parameter to simulate a subsidence value corresponding to the In SAR time series and then obtained by In SAR inversion,The reliability of the groundwater-ubsidence prediction model was verified by comparing with the measured level and the In SAR precision.Finally,a comprehensive evaluation method based on AIC(Akaike's Information Criterion,AIC)model of land subsidence parameters and settlement accuracy was put forward.