Satellite-based change detection for mining instability

1. Home
2. Publications
3. Online Library
4. Satellite-based change detection for mining instability

Mining operations can generate a large amount of waste material, usually disposed of in tailings dams. Historical as well as recent tailings dam failures have shown that these facilities remain vulnerable and are prone to catastrophic failures. Remote sensing data can be utilized to regularly and remotely monitor the condition of operational and abandoned mining sites. This study aims to detect tailings dam instabilities using high-resolution satellite imagery and evaluate the performance of different remote sensing indexes in change detection for two recent study cases: the 2022 Jagersfontein tailings dam failure in South Africa and the 2019 Córrego de Feijão tailings dam failure event in Brazil. Vegetation and geological indexes derived from images are used to track spatial and temporal change at mining sites. It is shown that remote sensing indexes are good indicators of failure, but some indexes are better than others, depending on the site conditions. For the Córrego de Feijão tailings dam that is located in a vegetated area, the failure resulted in more than 0.1 change in the statistical mode of all remote sensing indexes differences. The Normalized Difference Vegetation Index (NDVI) difference performed best, with the highest change of 0.26. On the other hand, the Jagersfontein tailings dam, located in a non-vegetated area, resulted in a 0.07 change in the statistical mode of the Modified Triangular Vegetation Index (MTVI2) difference but had very little change in the other remote sensing indexes.

9th International Congress on Environmental Geotechnics (ICEG2023)

Resource Mining and Extraction