Monitoring the impact of coal mining on vegetation cover changes over the last four decades in the coalfields region of Ermelo using Landsat earth observation data

Abstract

Mining poses a major threat to grass communities and environmental systems. Hence, it is necessary to determine the state of vegetation cover around mines. This study compared three sensors (WorldView-2, Sentinel 2 and Landsat 8) to map grass cover within a mining landscape. The study established that the WorldView-2 image, with 88.5% accuracy, was superior in classifying grass communities. Results also established that Erogrostic gummiflua, Monocymbium ceressiiformer, Seriphium plumosum and Mixed-grass communities were most dominant within the study area. The study also sought to determine the value of Worldview 2 bands and Normalized Difference Vegetation Index on Landsat data to determine grass communities and other land use and covers. The results established reliable accuracies in mapping grass communities and other land use/ land covers. The results demonstrate the value of the above-mentioned data set in detecting grass community changes in a mining area