LAUSR

Rare-earth mining has caused extensive damage to soil, vegetation, and water, significantly threatening ecosystems. Monitoring environmental disturbance caused by rare-earth mining is necessary to protect the ecological environment. A spatiotemporal remote sensing monitoring method for mining to reclamation processes in a rare-earth mining area using multisource time-series satellite images is described. In this study, the normalized difference vegetation index (NDVI) is used to evaluate the mining impact. Regression analysis is conducted to relate the HJ-1B CCD and Landsat 5/8 data to reduce the NDVI error related to sensor differences between different datasets. The analysis method of NDVI trajectory data of ground objects is proposed, and areas of environmental disturbance caused by rare-earth mining are identified. Pixel-based trajectories were used to reconstruct the temporal evolution of vegetation, and a temporal trajectory segmentation method is established based on the vegetation changes in different disturbance stages. The temporal trajectory of the rare-earth disturbance points is segmented to extract features in each stage to obtain the disturbance year, recovery year, and recovery cycle and evaluate the vegetation recovery after rare-earth mining disturbance. We applied the method to a stack of 20 multitemporal images from 2000 to 2019 to analyze vegetation disturbance due to rare-earth mining and vegetation recovery in the upper reaches of the Guangdong-Hong Kong-Macao Greater Bay Area, China. The results show the following. (1) Mining industry in the study area experienced rapid expansion before 2008, but growth slowed since the policies implemented by the government since 2009 to restrict rare-earth mining. (2) The continuous influence to the land caused by rare-earth mining can last for decades; however, the reclamation activities shorten the recovery cycle of mining land from 5 to 3 years.