An InSAR-Based Adaptive Method To Delimitate Subsidence Boundary For Guiding The Evaluation Of Building Damage In Coal Mining

Coal is one of the most important energy sources in China,which plays an indispensable role in social and economic development currently.However,coal mining would cause ground surface to move and deform,thereby potentially imposing damage to buildings or infrastructures such as the houses and village roads on the surface.Therefore,it is an effective way to resolve economic disputes between miners and property owners according to objective and true evaluation of mining damage hazards.Nearly all of traditional methods for delimitating mining subsidence boundary rely on in-situ measurements such as mining subsidence and underground mining schemes.If the in-situ measurements provided by the mine party is in doubt,it is difficult to objectively and truly delineate subsidence boundary using traditional methods.Interferometry Synthetic Aperture Radar(InSAR)technique is an active microwave remote sensing technology for detecting ground surface deformations,and it has some distinct advantages,such as high spatial resolution,large spatial coverage,high efficiency,all-weather and dayand-night.InSAR can obtain historical surface subsidence with high spatial resolution based on archived SAR images.Thus,it provides a new opportunity for deliminating mining-related subsidence boundary without in-situ measurements.The existing InSAR methods firstly retrieve mining-related subsidence by neglecting both east and north movements,and then take the subsidence contour of 10 mm as the boundary.However,underground mining activities do not cause only ground surface to subside,but also induce considerable horizontal movements.Therefore,this implies that large errors are possibly be resulted in the estimated subsidence,due to neglecting the considerable horizontal movements.Moreover,this error will definitely have an impact on the subsidence boundary delineated by InSAR.Therefore,these methods cannot take the different errors of InSAR observations in different areas into account,dramatically degrading the reliability of the delimitated subsidence boundary using contour of 10 mm,even can wrongly judge subsidence boundary,and which is easy to have an impact on the evaluation of building damage hazards.In view of this,this thesis is focused on analysing the errors of InSAR-based mining subsidence estimation caused by Neglecting Horizontal mo Vement(NOV)(referred to as NOV method below)components,and delimitating mining subsidence boundary considering the error of InSAR observation.It aims to provide a reference for mining disaster prevention and control,mining damage identification,and guiding the rational application of InSAR in mining deformation monitoring.And it will also help to improve the accuracy of delimitating subsidence boundary based on InSAR observations,and promote InSAR to guide mining damage identification more reasonably.The primary work and innovations of this thesis are summarized as follows:(1)Analysing the errors of InSAR-based mining subsidence estimation caused by neglecting horizontal movement components.Firstly,the error structure of the NOV-estimated subsidence was deduced theoretically.Secondly,a pixel-wise method is proposed to evaluate the accuracy of the NOV-estimated subsidence,and the method is verified by taking Qianyingzi coal mine as the experimental.Then,the error pattern of the NOV-estimated mining subsidence was discussed using the histogram and spatial statistics analyses.Fourthly,we discuss some influence factors(e.g.,SAR imaging view,interferometric coherence,and lithology of overlying rock strata)on the magnitude of the NOVestimated subsidence errors.Finally,influence on the boundary delimitation of mining subsidence is discussed.On the one hand,the final research results show that large errors are possibly caused in the NOVestimated mining subsidence.And the errors can reach 45.1% of the maximum subsidence in this study.On the other hand,the underestimation and overestimation errors of the NOV-estimated subsidence are clustered in a “peak-valley” shape(peak-valley error)in space,and the peak-valley angle is 5°～15°.The delimitated boundary using the subsidence contour of 10 mm insignificantly shrank based on NOV-estimated subsidence.This is mainly due to the “peak-to-valley”error pattern of the NOV-estimated mining subsidence.(2)A method to delineate the mining subsidence boundary is proposed considering the error of InSAR observations.Firstly,the historical subsidence of mining areas was obtained by fusing ascending and descending InSAR displacement observations.Then,the overall accuracy of the obtained historical subsidence was evaluated with the farfield displacements.Following that,confidence intervals were used to guide the selection of the subsidence thresholds,and further to be used to delimitate subsidence boundary.Finally,the feasibility and reliability of this method were tested by simulation experiments and real data from a mining area in Tangshan city,China.The results suggest that the boundary delimitated by InSAR-estimated subsidence considering the InSAR observation errors using a confidence interval of 99% well agree with the reference subsidence boundary.