Monitoring ecosystem restoration of multiple surface coal mine sites in China via LANDSAT images using the Google Earth Engine

Abstract

The restoration of surface mining (open‐cast) sites is key to meeting global ecosystem restoration targets. With the improving of data availabilities and technologies, it has become possible to expand restoration monitoring from single to multiple mine sites on a large scale. Based on the MODIS global disturbance index (MGDI), this study proposes a mine landscape restoration index (MLRI), by coupling the LST and EVI to simultaneously monitor the restoration of multiple mine sites. Restoration areas were identified by MLRI time‐series analysis and classified into significant consistent increase (SCI) and significant anti‐consistent increase (SAI) areas. The restoration effects of 46 surface coal mine sites located in the ecologically fragile northwestern region of China from 2000 to 2019 were assessed based on 3,675 LANDSAT images from the Google Earth Engine. Results show that the MLRI was effective at identifying restoration areas and processes, and this effectiveness was validated by high‐resolution images and field investigations of mine samples. The overall percentage of restored area for mines that started mining before 2000 was 55.01% ; for mines that starting mining after 2000, 33.68%. According to the differences of SCI and SAI area percentages, 46 mine sites were classified into three clusters, with 13, 11, and 22 mine sites, respectively. The mine sites with high restoration percentage are located mainly in Hailar and Shanbeimengnan regions. This study provides a new approach for monitoring the restoration effects of multiple mine sites and informs government managers about developing mine restoration programs and sustainable mining development plans.

Greening or browning? The macro variation and drivers of different vegetation types on the Qinghai-Tibetan Plateau from 2000 to 2021

“…Meanwhile, this period is the maximum time range for which MODIS NDVI datasets are currently available. The Mann-Kendall test and Sen’s slope trend analysis were combined to identify trends and magnitudes of vegetation greenness change ( Wang et al., 2021 ; Liu et al., 2022 ). As a nonparametric statistical test, the Mann-Kendall test does not require the data to be normally distributed, only that they are independent ( Li et al., 2020 ).…”

“…The Mann-Kendall test is also useful for removing noise from time series ( Zhou et al., 2020 ; Zhang et al., 2022 ). It is widely used to determine whether processes such as climate, hydrology and vegetation greenness change are undergoing natural fluctuations or have a definite trend of change ( Wang et al., 2021 ; Cai et al., 2022 ). Therefore, the Mann-Kendall test was chosen to determine the significance of trends in vegetation dynamics.…”

“…As a fundamental component of terrestrial ecosystems, the distribution and condition of vegetation communities often affect the qualities of animal habitats ( Stein et al., 2014 ; Seibold et al., 2019 ). Therefore, changes in vegetation communities may have important impacts on the response to global climate change, the conservation of biodiversity and the sustainable development of human society ( Hansen et al., 2013 ; Wang et al., 2021 ). At the national, regional and global levels, a series of ambitious ecosystem restoration targets have been proposed to address multiple major socio-environmental challenges, such as global change and biodiversity loss, and thus improve people’s livelihoods ( Strassburg et al., 2020 ).…”

“…From reginal to global scale, researchers have chosen NDVI to indicate the greenness status of vegetation such as forests ( Hilker et al., 2014 ), grasslands ( Miao et al., 2021 ) and tundra ( Myers-Smith et al., 2020 ), etc. An increase in NDVI indicates a greening of the vegetation greenness and, conversely, a decrease in NDVI indicates a browning ( Myers-Smith et al., 2020 ; Wang et al., 2021 ). Therefore, time series analysis of NDVI can characterize the changes in vegetation over time and thus provide a basis for further exploration of the drivers of changes in vegetation dynamics.…”

Automated mapping of Portulacaria afra canopies for restoration monitoring with convolutional neural networks and heterogeneous unmanned aerial vehicle imagery

“…This monitoring is often limited to costly techniques that are time- and labour-intensive when hundreds to thousands of hectares need to be covered. Fortunately, recent technological advances have provided innovative techniques to overcome this challenge in restoration ( Almeida et al, 2021 ; Chen et al, 2021 ; Wang et al, 2021 ).…”