Monitoring And Analysis Of Land Subsidence In Mining Area Based On Time Series InSAR Technology

Coal has been mined in China for thousands of years,and it has an extremely important status in China's energy structure.Large-scale coal mining has promoted China's economic development,but a series of ecological environment and surface subsidence problems caused by it have become increasingly serious.With the continuous advancement of the working face,ground fissures,subsidence funnels,collapse pits and landslides will form on the surface of the mining area,causing damage to roads,pipelines and civil construction in the mining area.In addition,the secondary disasters caused by mining subsidence have brought great harm to the life and property of the mining area and industrial and mining plants,and also caused frequent geological disasters in the surrounding areas of the mining area,and related problems have become increasingly prominent.Therefore,the monitoring and analysis of surface subsidence in mining areas is particularly important.InSAR(Interferometry Synthetic Aperture Radar)technology has been widely used in earth observation because of its advantages of all-day,all-weather,large range,high accuracy,non-contact,low cost,continuous monitoring,and easy repeated observation.Unlike traditional leveling and GNSS surveys,the results of InSAR monitoring have spatial continuity,overcofning the difficulties of large leveling field workload,sparse GNSS points,small monitoring range,and difficulty in reaching ground surveys.However,InSAR technology is susceptible to factors such as spatiotemporal loss of coherence,orbit error,external DEM data error,atmospheric delay,and other factors,and coal mining subsidence generally has the characteristics of rapid development and large settlement,which makes the use of InSAR technology to monitor land subsidence There are still some problems to be solved urgently.In response to these problems,this paper takes Shendong Mining Area and a certain mine in Shendong Mining Area as the research area,and uses time series InSAR technology as the means to monitor and research the mining subsidence area.The main research contents and results are as follows:(1)Taking Shendong mining area as the research object,collecting 30 scenes of Sentinel-1A data covering two adjacent orbits of part of Shendong mining area,and using interferogram stacking technology(Stacking-InSAR)to conduct large-scale deformation detection in the research area.The average settlement rate and standard deviation of the study area were obtained,and the ground settlement area was analyzed using GIS tools.The experimental results show that Stacking-InSAR technology can monitor large-scale deformation of mining areas,and the monitoring results are relatively reliable.(2)Using the 22 scenes of Sentinel-1A data collected from the research area,and using the small baseline subset InSAR technology(SBAS-InSAR)for data processing,the annual average deformation rate and time series cumulative settlement of the research area were obtained.Combined with the data of the mining area,the analysis of the mining line profile and characteristic points,it is shown that the spatial distribution,mining footage and mining conditions of the mining settlement area and mining face are in good agreement.Using the collected ground level data of the mining area to verify the timing results of the small baseline technique,the correlation coefficient between the two was calculated to be 0.922.Finally,after the revised model corrects the SBAS-InSAR results,the correlation coefficient between the two is increased to 0.992,indicating that the revised model can well optimize the monitoring results of the SBAS-InSAR technology.(3)The gray model and support vector regression model are used to predict the settlement of the mining area monitored by the small baseline set technology and the two models are improved respectively.The improved model combination is used to predict the settlement of the mining area and the estimated accuracy of the model The evaluation shows that the improved grey-support vector regression model has greatly improved prediction accuracy and prediction stability,and meets the prediction accuracy,and can be used in actual production.