Research On DS-InSAR Phase Spatiotemporal Filtering And Its Application Of Deformation Monitoring In Abandoned Mining Areas

With the westward shift of China’s coal energy center and the transformation and upgrading of the eastern resource-based cities,the surface subsidence and instability of abandoned goaf caused by underground coal mining have led to severe challenges and threats to the sustainable development of resource-exhausted cities.Time-series InSAR technique,as a powerful means of surface deformation monitoring in abandoned mining areas,is difficult to fully reflect the surface deformation characteristics of goaf due to the insufficient number and uneven distribution of observation points.As for distributed scatters,the number of monitoring points increases significantly,however,the phase stability is still affected by factors such as spatio-temporal incoherence.Therefore,as the core foundation of DS-InSAR technique for surface deformation monitoring,the phase optimization estimation of distributed scatters has become an important prerequisite for the fine deformation monitoring in urban abandoned mines.In order to solve the problems of insufficient spatial continuity constraint of phase optimization and further improve the accuracy of surface deformation monitoring in urban abandoned mines,the thesis gives some improvement methods based on in-depth study of DS-InSAR technique and combined with the surface deformation characteristics of mining areas.The phase optimization filtering incorporating the region-growth approach,the phase optimization based on spatio-temporal total variation regularized robust low rank tensor decomposition,and the wavelet decomposition-based atmospheric noise separation method for cumulative deformation signals are proposed,which have realized the phase quality improvement and atmospheric delay compensation in abandoned mining areas,improved the reliability of the interpretation of deformation monitoring results,and provided technical support for the sustainable development of cities.The main work and research results are as follows:(1)The research background of surface subsidence in abandoned mining areas is outlined.The principles,processes and shortcomings of InSAR,D-InSAR and time-series InSAR analysis techniques are summarized.The DS-InSAR technique is thus elicited and the statistical characteristics of distributed scatterers and the current main phase optimization algorithms are presented.(2)Based on the comparative analysis of commonly used statistically homogeneous pixels methods,a region-growing phase optimization filtering algorithm incorporating deformation information is proposed to address the shortcomings of distributed scatterers in the spatio-temporal dimensional continuity constraint.In order to fully take into account the phase characteristics of distributed scatterers in the spatio-temporal dimension,a spatio-temporal total variation regularized robust low rank tensor decomposition was proposed.The optimization effect of the distributed scatterers phase is evaluated qualitatively and quantitatively evaluation using simulated data and real data.The results show that both proposed algorithms can effectively reduce the effect of incoherence noise,which is useful for surface deformation monitoring in urban abandoned mining areas and provides a new idea for surface deformation refinement acquisition under different monitoring density environments.(3)The surface deformation monitoring is carried out from December 2018 to December 2021 at abandoned mining areas in Xuzhou city.The performance of the two proposed phase optimisation algorithms is verified by comparison.A study on the separation of atmospheric phase noise based on deformation time series signals is carried out using wavelet analysis.The spatial and temporal evolution of surface deformation in a typical abandoned mining areas is analysed in depth,and the accuracy and reliability of the monitoring results are verified by using ground data and PS-InSAR results.The results show that the improved distributed scatterers phase spatio-temporal optimization filtering method can effectively reduce the influence of phase noise,which is helpful for fine surface deformation information acquisition in urban abandoned mining areas and provides theoretical and technical support for the transformation,upgrading and sustainable development of resource-exhausted cities.There are 43 figures,10 tables and 114 references.