Research on surface subsidence information extraction method based on high phreatic coal mining area

Accurate， rapid and efficient access to coal mining subsidence information is great significance for land reclamation in mining areas， especially for coal mining areas in high phreatic areas， which are likely to form large-area surface subsidence and non-watershed areas.The work of reclamation is even more serious.Taking Baodian Coal Mine， a high phreatic mining area， as the research object， combining and improving the water body and non-water body extraction methods， the surface subsidence information of the entire mining area was obtained.On the basis of Modified Normalized Difference Water Index (MNDWI)， an Enhanced Modified Normalized Difference Water Index (E-MNDWI) was proposed for the characteristics that the edge of coal mining subsidence was easy to be confused with water body information.The Landsat 8 data was used to extract the subsidence water area through E-MNDWI.The Sentinel-1A data (Sentinel-1A)was used to extract the subsided non-watershed area by Small Baseline Subset Interferometric Synthetic Aperture Radar (SBAS-InSAR).Finally， Kriging interpolation was performed to obtain the 10 mm contour of the subsidence of the mining area.The results showed that the accuracy of extracting subsidence water by E-MNDWI was high， and the Kappa coefficient was 85.07%.The western part of the mining area has basically reached a stable state， and the surface subsidence in the east was more obvious， and the southern surface was slightly elevated.During the monitoring period， the maximum surface subsidence rate of the mining area was 41.69 mm/a， and the maximum sub-sidence was 85.16 mm.As of 2017， the subsidence area caused by coal mining in the mining area was 10.1 km2， of which the subsidence water area was 4.6 km2 and the non-water accumulation area was 5.5 km2.A number of reference points were selected to verify the extraction results of the subsided non-watershed area， and the coefficient of determination R2 was 0.92.The use of multi-source and multi-temporal data， combined with a variety of methods to obtain mining subsidence information， could provided a theoretical basis for ecological restoration and land remediation in coal mining cities， and provide a new idea for the rapid and efficient monitoring of surface deformation in coal mining subsidence areas.