Research on the spatiotemporal evolution characteristics of NDVI in Shendong mining area based on GEE long-time...

Using long-time series remote sensing images to study the spatio-temporal variation characteristics of surface vegetation in mining areas and evaluate the impact of coal development on vegetation coverage has important practical value for ecological protection and sustainable development of mining areas. However, few existing studies simultaneously meet the requirements of long-term and high-frequency Normalized Difference Vegetation Index (NDVI) remote sensing inversion, and few studies simultaneously obtain the spatiotemporal evolution trends and dominant patterns of NDVI, which is not conducive to further research on the laws and mechanisms of the impact of mineral resource mining activities on the ecological environment. This study utilized the Google Earth Engine (GEE) platform to obtain long-term and high-frequency NDVI data from the Shendong mining area. The spatiotemporal evolution characteristics of NDVI in the Shendong mining area were studied using methods such as time series Univariate Linear Regression Equation (ULRE) combined with Empirical Orthogonal Function (EOF). The results are as follows: (1) From 1986 to 2024, the NDVI of the Shendong mining area exhibited a notable fluctuating growth trend. Over the past 39 years, the mean annual NDVI increased by 0.14, with an increase rate of 0.0036 per year, an increase of 54.46%, and an average annual growth rate of 3.18%. The R of the linear fitting equation reached 0.8975, indicating that the overall surface vegetation in the Shendong mining area has shown a significant positive trend over the past 39 years. (2) The annual average NDVI of all 18 coal mines exhibited a clear fluctuating growth trend, with R ranging from 0.8188 (Wujiata) to 0.9118 (Cuncaota). Furthermore, the overall fluctuation trends of the 18 coal mines were largely similar. (3) The five types of NDVI mean evolution trends in the Shendong mining area generally conform to a normal distribution and exhibit strong spatial heterogeneity. The area proportion of the non-significant change region is the largest, reaching up to 55.38%; the area proportion of the extremely significant reduction region is 5.42%, mainly distributed in areas such as construction land and open-pit mining; the area proportion of the extremely significantly increased region is 6.69%, mainly distributed around water bodies, forest land, and valley land; the overall area proportion of the reduction region is 19.32%, while the overall area proportion of the increased region is 25.30%, indicating that the overall surface vegetation in the study area is showing a positive trend. (4) The spatiotemporal distribution characteristics of NDVI in the Shendong mining area from 1986 to 2024 are extremely obvious and single. EOF₁ can represent all the characteristics of NDVI changes in the entire Shendong mining area, and its spatial distribution presents a pattern of "medium coverage as the main, high coverage as the supplement", reflecting the comprehensive results of the interaction between ecological restoration and mining disturbance. This study can provide a reference for the ecological management, restoration, and planning of land and space in mining areas.