Research On Surface Deformation Monitoring Method Of Mining Area Based On Domestic L-band Differential Interferometric SAR Image

Coal occupy a leading position in China’s energy structure,providing an important force for the rapid development of the national economy.Meanwhile,the large-scale Mining and utilization of coal has also caused a series of severe issues on the ecological environment of the surface and the production and life of people,such as local ground subsidence,cracks and collapse,landslides,and destruction of buildings and structures.Therefore,high-precision monitoring of surface deformation caused by mining in mining areas has very significant practical significanceSpace-borne SAR system has technical advantages such as all-weather,all time,strong penetration ability,and wide monitoring range,and is widely used in deformation monitoring in mining areas.L-SAR satellite,as the first distributed SAR satellite constellation used for deformation measurement in China,has the characteristics of short temporal baseline and orbital tube,high spatial resolution,high correlation,and kinds of polarizations.This thesis takes Datong Coal Mine in Shanxi Province as an experimental area to conduct in-depth research on the deformation monitoring capability and accuracy of L-SAR satellite in mining areas,and improves some key technologies based on the characteristics of mining area deformation.The main research work and achievements are as follows:(1)The current development status of spaceborne SAR systems was reviewed,with a focus on introducing the L band SAR satellite(L-SAR)and corresponding parameter information.The basic principles of D-In SAR/Multi-Temporal In SAR,Offset-tracking methods,and the current research status in deformation monitoring of mining areas at home and abroad were elaborated.(2)In order to solve the problems of domestic SAR data processing and autonomous production,Land SAR,the latest domestic SAR software,are used to test L-SAR data.A modification scheme was developed to solve the problems of incorrect imaging parameters such as the Doppler polynomial coefficient of L-SAR data,center scene geocentric radius,time per azimuth line,SAR parameters of platform state,etc.,and developed and improved L-SAR data query,interference pair automatic screening,multilook calculation,coherence mask and other functions,and then complete the subsequent L-SAR data suitability analysis and deformation algorithm verification based on these functions and modifications.(3)In response to the current issue of In SAR research mainly relying on foreign data,long satellite revisit cycles,and difficulty in obtaining large-scale and high-precision deformation information in mining areas,12 L-SAR images covering the Datong mining area in Shanxi Province from December 2022 to March 2023 are used in this thesis to study the suitability of domestic L-SAR satellites under different In SAR/SAR algorithms and different levels of deformation conditions in mining areas.This provides reference for the research and application of deformation monitoring methods in mining areas based on domestic L-SAR satellites.The results show that:(1)the differential interferograms of the mining area obtained based on L-SAR data is more coherence than Sentinel-1A data,and the deformation detection ability of the large gradient mining area is also stronger.(2)D-In SAR has unique advantages for the monitoring of nonlinear deformation in mining areas,as it can achieve centimeter level monitoring of surface deformation in mining areas.And the cumulative D-In SAR based on L-SAR data achieves a complete mining subsidence basin monitoring which maximum subsidence is 3.135 meters with high accuracy of 16.5mm.(3)The SBAS-In SAR calculation results based on L-SAR data are basically consistent with the Leveling monitoring results in small deformation areas,and have millimeter level monitoring accuracy.The root mean square error of the SBAS-In SAR results with the short-term spatiotemporal baseline is 3.7 mm,while the root mean square error of SBAS-In SAR with the long-term spatiotemporal baseline is 3.9 mm,meeting the accuracy requirements for deformation monitoring in mining areas.However,the error of SBAS-In SAR results will gradually increase with the increase of mining area deformation level.(4)The Offset-tracking technology based on L-SAR data has significant advantages in monitoring low coherent large gradient deformation in mining areas.The root mean square error of the monitoring results is0.280 m,and the root mean square error of the Multi-Temporal Offset-tracking monitoring results is 0.158 m.However,all of them have underestimated the defomation of the(4)A adaptive window homogeneous point selection method based on the linear fitting coefficient is proposed to address the issue of the deformation results of single view complex SAR images being easily affected by decoherence noise,and the traditional DS-In SAR homogeneous point selection method being prone to aliasing in large gradient deformation areas.This method distinguishes large gradient deformation regions through the linear fitting coefficient,and sets windows for selecting homogeneous points of different sizes based on this.The experiment utilized L-SAR images covering a coal mine in Datong,Shanxi Province from December2022 to March 2023 as the data source.The fixed window homogeneous point selection results and phase optimization results validated the method proposed in this thesis to reduce the occurrence of interference stripe aliasing in deformation areas,while also ensuring that the differential interferograms have high phase quality as a whole.This thesis includes 47 figures,7 tables,and 99 references.