Remote sensing monitoring research for coal fire in Wuda coal mining using ASTER thermal infrared images

Abstract

Coal fire catastrophes in coal field usually lead to subsidence, collapse, landslide, crack of mining land surface, destroy coal resources and threat local environment, residents health et al. Wuda coal field in Inner Mongolia is one of the largest coal fires area worldwide, which has been researched most extensively. Since August 2010, a fire-fighting project in Wuda coal mining was performed, which was designed to be accomplished in 2013 and set 2014 as fire-fighting monitoring period. Currently most researches concentrate on the fierce burning period of coal fire, but ignore the evaluation of extinguishing effect. In order to supervise and evaluate situation of coal fires before and after performing fire-fighting project, we adopted ASTER thermal infrared bands in 2008, 2010, 2013 and 2015 as data resources to retrieve mining land surface temperatures in each year using classic Temperature and Emissivity Separation (TES) algorithm. Self-Adaptive Gradient-Based Thresholding (SAGBT) algorithm was applied to extract coal fire areas in Wuda coal field during four years. By comparing extracted coal fire areas in 2008 with reference coal fire areas based on aerial thermal infrared images in 2009, we found that overlay degrees of both reach 71.4%. The precision of this method would be more higher if acquired times of two kinds of images kept consistency. Based on the same method, coal fire areas during other three years were respectively extracted and used to analyze evolution status of coal fires. The results showed that coal fire areas overall dropped by 37.5% in 2015 than in 2008, which declared fire-fighting activities in Wuda coal field had taken good effects at some extent. However, preservation of 1.2 km2 indicated that more fire-fighting measures need to be adopted by local administrator for drastically extinguishing coal fire disasters.