International Journal of Applied Earth Observations and Geoinformation (Dec 2021)

Average-DInSAR method for unstable escarpments detection induced by underground coal mining

  • Xin Yao,
  • Yiping Chen,
  • Donglie Liu,
  • Zhenkai Zhou,
  • Veraldo Liesenberg,
  • José Marcato Junior,
  • Jonathan Li

Journal volume & issue
Vol. 103
p. 102489

Abstract

Read online

A catastrophic rock avalanche from a tableland escarpment occurred at the Pusa village, Guizhou Province, southwest (SW) China, causing 35 fatalities and huge economic losses. The steep slope lies in the Longtan Formation coal-bearing shale of Permian, which is widely distributed in SW China. It was overlaid by brittle superstrata in Triassic and followed by gently anticline tectonic movement in Cenozoic, thus forming large-scale tableland escarpments with an “upper brittle, lower ductile” structure. Affected by underground coal mining activity at the base, this escarpment has become unstable and prone to failure. In order to further clarify the geological conditions and other influence factors for the Pusa landslide, we propose a newly improved multiple Differential Interferometric Synthetic Aperture Radar (DInSAR), named average-DInSAR, to detect the displacements on escarpments in a broad region. Extensive experimental results show that there existed obvious pre-failure displacements on the swarming escarpments, evidencing their unstable state, which were verified by field inspection. The spatiotemporal correlation analysis suggests that this abnormal deformation is probably induced by underground coal mining in the vicinity. Further confirming that the special geological conditions and nearby coal mining activity were responsible for the 2017 Pusa rock avalanche. Our study also demonstrates that the average-DInSAR method is simple and effective, which can overcome low coherence and noise of DInSAR, especially for shorter X- or C-band SAR data. Application of proposed method would permit to detect displacement before slope failure with higher re-visiting frequency, thus helping define early warning strategies for landslides in area with similar geological conditions.

Keywords