Three-dimensional time-varying large surface displacements in coal exploiting areas revealed through integration of SAR pixel offset measurements and mining subsidence model

REMOTE SENSING OF ENVIRONMENT

卷:240

Article

Chen, Bingqian^1,2,3,4  Li, Zhenhong^2,3  Yu, Chen²  Fairbairn, David²  Kang, Jianrong¹  Hu, Jinshan¹  Liang, Liang¹ 

[1] Jiangsu Normal Univ, Sch Geog & Geomat & Urban Rural Planning, Xuzhou 221116, Jiangsu, Peoples R China

[2] Newcastle Univ, Sch Engn, Newcastle Upon Tyne NE1 7RU, Tyne & Wear, England

[3] Changan Univ, Coll Geol Engn & Geomat, Xian 710054, Peoples R China

[4] Hunan Prov Key Lab Coal Resources Clean Utilizat, Xiangtan 411201, Peoples R China

关键词: Synthetic aperture radar;    Deformation monitoring;    Coal mining;    Pixel offset tracking;    Time series analysis;    Time-varying displacement;   

DOI  :  10.1016/j.rse.2020.111663

来源: Elsevier

【 摘 要 】

Underground mining activities usually induce large surface displacements thus causing serious safety hazards and potential ecological damage. The capability of conventional Interferometric Synthetic Aperture Radar (InSAR) to monitor tectonic movements, volcanic eruptions and city subsidence has been fully demonstrated, but its application to mining subsidence is limited because of the failure caused by localized surface displacements with strong spatial gradients. In this paper, a new method is presented to utilize SAR pixel Offset Tracking (OT) with a single pair of SAR images to resolve three-dimensional (3D) large surface displacements caused by underground coal mining. Coal mining subsidence theory is utilised to analytically separate the vertical and horizontal components. This method is applied to the Daliuta coal mining area in Shaanxi Province, China, where a dense GPS network is available. Results show the RMS differences of OT derived displacements against GPS in both horizontal and vertical directions are in the sub-centimeter level. In addition, a prediction of mining-induced ground movements is performed with the Support Vector Regression algorithm and RMS differences of 12.4, 13.1 and 14.4 cm are observed compared to GPS, in the vertical, easting and northing directions respectively. The framework demonstrated in this paper is not only able to derive the evolution of the 3D large surface displacements with multi-temporal SAR images in a single-geometry, but also has a potential for short-term predication, which can provide early warnings and promote strategic decision-making for engineering management in the process of coal mining.

【 授权许可】

Free