Determination of Illegal Signs of Coal Mining Expansion in Thai Nguyen Province, Vietnam from a Combination of Radar and Optical Imagery

Abstract

Vietnam has diverse and rich natural resources with over 5000 mines and 60 different types of minerals. Mineral activities in Vietnam are licensed by the Government and must ensure to protect the environment and daily activities. However, due to the high profits generated from mining, there have been many illegal mining activities taking place all over Vietnam. One of the causes leading to this situation is the insufficiency in monitoring and managing resources and minerals. Additionally, many mines are located in the forest which is difficult to detect. Therefore, in this study, we adopt Sentinel-1 radar satellite image and Sentinel-2 multi-temporal satellite image to determine illegal coal mining activities at the Minh Tien coal mine, Thai Nguyen province, Vietnam. The research method includes applying Google Earth Engine (GEE) for processing the Sentinel-1 images in both ascending and descending directions and Sentinel-2 optical images to calculate the Normalized Difference Vegetation Index (NDVI) series to control the different trends from 2016 to 2021. NDVI is then employed to mask the areas with vegetation that witness no abrupt changes in the land cover. This NDVI also acts as the basis for collecting samples for Random Forest classification of the Sentinel-2 images. The results from Sentinel-1 and Sentinel-2 combined with MASK from NDVI have determined the expansion area of the Minh Tien coal mine. The results are compared with those published on the website of Thai Nguyen Portal and have a significant similarity.