The use of multi-temporal satellite images for environmental assessment in coal mining areas (by example of closed Kizel coal basin)

Abstract

The article describes the possibility of application of multi-temporal Landsat and Sentinel-2 images for environmental monitoring and assessment in coal mining areas. The study area is abandoned Kizel coal basin, which is located in the Perm region, Russia. For the last 15-20 years, the acid mine drainage (AMD) and mine tailings leachate have caused extreme contamination of rivers with iron, alumini-um and heavy metals. Multi-temporal Landsat and Sentinel-2 images for 1987‒2017 and the results of hydrochemical monitoring of surface water bodies for 2006‒2013 are used as initial data. The Acid Mine Water Index (AMWI) is proposed to estimate the AMD-induced river contamination. This index is based on the reflectance values in blue and red spectral bands of Landsat and Sentinel-2 images. The AMWI values are compared with the iron concentration measurements in several streams flowing through of Kizel coal basin. The statistically significant correlation between the AMWI values and iron concentration is found for Yayva river. We have analyzed the temporal changes of AMWI for 1987‒2017, and found that AMWI values for Yayva and Kos'va rivers varied sometime in different directions. These variations can be explained by the difference of AMD regime. Also we outlined some areas (around 20 ha) of degraded soil and vegetation cover from high-resolution images. The largest areas of degraded soil and vegetation occurred in 2006‒2010 in association with the change of AMD regime. After 2010, the degradation process slowed down.