Drivers of Vegetation Cover and Carbon Sink Dynamics in Abandoned Shaoyang City Open-Pit Coal Mines

Abstract

As an important coal-producing region in China, open-pit coal mining in Shaoyang, Hunan Province, has a significant impact on the ecological environment. This study focuses on the three major open-pit mining areas in the city, utilizing remote sensing data from 1998 to 2024. By calculating the normalized difference vegetation index (NDVI) and fractional vegetation cover (FVC), and combining climate factors such as temperature and precipitation with Net Primary Productivity (NPP), this study analyzes the spatiotemporal evolution characteristics of vegetation cover and carbon sinks, and explores the impact of climate and environmental policies on vegetation recovery. The study employed trend analysis and autoregressive integrated moving average (ARIMA) model predictions, which showed that vegetation cover in the mining areas decreased overall from 1998 to 2011, gradually recovered after 2011, and reached a relatively high level by 2024. Changes in carbon sinks were consistent with the trends in vegetation cover. Spatially, the north mining area experienced the most severe vegetation degradation in the early stages, the middle area recovered earliest, and the south area had the fastest vegetation cover recovery rate. Climate factors had a certain influence on vegetation recovery, but precipitation, temperature, and FVC showed no significant correlation. The study indicates that vegetation recovery in mining areas is jointly influenced by mining intensity, climate conditions, and policy interventions, with geological environment management policies in Hunan mining areas playing a key role in promoting vegetation recovery.

1. Introduction

Exploitation without environmental care poses a significant barrier to the sustainable development of mining regions [1,2,3]. The coal mining activities of Shaoyang city, which is the core area of coal resource development in Hunan Province, play an important role in promoting local economic development. However, the ecological problems caused by coal mining have gradually attracted widespread attention. In particular, open-pit coal mining has significantly affected the vegetation ecosystem around mining areas [4,5,6]. As coal mining activities continue, the vegetation cover in the northern, central and southern mining areas of Shaoyang city has generally declined, leading to a series of ecological problems, such as soil erosion, water loss, and deterioration of air quality, posing serious challenges to local ecosystems and the living environments of residents [7]. Consequently, the integration of coal resource exploitation with environmental conservation and the advancement of sustainable development in mining areas have emerged as pressing concerns in Shaoyang city and beyond.

The importance of plant regeneration in the ecological rehabilitation of mining sites is unequivocal. In recent years, the dual impact of climate change and regulatory policies has resulted in dynamic alterations in plant cover and carbon sinks in mining regions. Climatic elements, particularly variations in temperature and precipitation, are pivotal in this process [8,9]. Precipitation directly affects the availability of water for vegetation in mining areas and is a key factor limiting plant growth [10]. Increased precipitation contributes to the resilience of vegetation, especially in arid regions, and adequate precipitation can significantly contribute to the recovery of vegetation. Additionally, temperature affects vegetation growth to a certain extent, especially under extreme temperature conditions, where high temperatures may exacerbate evapotranspiration and lead to plant degradation, especially in arid and semiarid areas [11,12,13]. In addition to climatic considerations, policy forces have significantly influenced the restoration of vegetation in mining regions. In recent years, Hunan Province and Shaoyang city have implemented numerous environmental protection regulations, including the Regulations on the Treatment and Restoration of the Geological Environment of Mines and a pilot initiative for “green mines”, which have compelled enterprises to carry out ecological restoration and take measures such as compensation for vegetation, reclamation and other measures and has achieved remarkable ecological restoration results [14]. Thus, the guiding role of policies, especially under unfavourable climatic conditions, is often effective in compensating for the lack of natural restoration and promoting the recovery and restoration of vegetation in mining areas.

This study utilized remote sensing monitoring technology to thoroughly analyse the characteristics of changes in vegetation cover and carbon sinks in the coal mining area of Shaoyang city, employing fractional vegetation cover (FVC) and net ecosystem productivity (NEP) as primary indicators to evaluate the ecological restoration capacity of the region [15,16]. The change in vegetation cover in the three major mining areas between 1998 and 2024 was systematically analysed by combining the correlations between climate factors (precipitation and air temperature) and vegetation growth as well as vegetation carbon sinks and policy contexts [17,18]. The use of remote sensing data has the advantages of effectively providing large-scale, continuous time series of surface information, is not subject to geospatial limitations, and can comprehensively reflect the spatial and temporal characteristics of the changes in vegetation in mining areas [19]. In addition, correlation analysis with climatic factors can further reveal the impact of climate change on vegetation restoration [20,21]. An evaluation of policy execution can reveal the beneficial impacts of policy-oriented vegetation restoration in mining regions [22]. This thorough analytical method offers insights and a theoretical framework for comprehending the various factors influencing plant restoration in mining regions and their interconnections.

Vegetation restoration in coal mining regions in Shaoyang city is influenced by multiple factors, including mining intensity, climatic circumstances, and policy interventions. Ecological rehabilitation in mining regions faces the dual pressures of the natural environment and human activities [23,24]; nevertheless, the guiding role of policy plays a crucial role in the restoration process [25]. Research has shown that the recovery of vegetation cover can be facilitated by scientific policy regulation, especially in arid and low-precipitation areas, and by artificial restoration measures that can significantly compensate for the lack of natural recovery. Eventually, the vegetation cover and carbon sink can be restored to those of unexploited or lightly exploited stages.

2. Materials and Methods

2.1. Study Area

The research area of this paper (Figure 1) is generally located in Shaoyang city, Hunan Province, which is rich in coal resources, mainly in the mountainous and hilly areas in the north and west. These areas have complex geological formations, and multiple coal seams may exist. There are various types of coal, including bituminous coal and anthracite, which have high industrial utilization value. The three basically similar coal mining areas in this study are located in Shaodong city in northern China, Shaoyang County in central China, and Wugang city in southern China, each covering areas of approximately 184,500 m², 87,300 m², and 209,700 m², respectively. The three open-pit coal mining areas are the main coal-producing areas in Shaoyang city, in which the elevations of the northern and central mining areas are basically the same; the southern mining areas are slightly greater in comparison, but all of them have made outstanding contributions to the coal mining output in Hunan Province.

The research region is situated in the subtropics, typified by a typical mid-subtropical humid monsoon climate, which features four distinct seasons, ample sunlight and warmth, substantial precipitation, and concurrent rainfall and heat during the same season. It is situated inside the South Mountain Range, Xuefeng Mountain Range, and remnants of the Yunnan–Guizhou Plateau, representing one of the four principal forest regions in Hunan. There are 210 species of wood forests, primarily cedrus deodara (Roxb. ex D. Don). The natural soil in the study area is mainly composed of red soil and paddy soil. After coal mining, the soil texture changed from loamy to predominantly clayey, the soil structure became lumpy, and the soil pH decreased from approximately 5 to approximately 3. In addition, the historical mines in the region have left behind piles of waste rock, which have destroyed the vegetation and exposed the grayish-white and grayish-brown rock walls beneath.

2.2. Data Sources

To reduce the errors caused by differences in phenology, remote sensing images with a spatial resolution of 30 m (

Table 1. Remote sensing image sets used in this paper.

Types of Remote Sensing Images	DATE
Landsat5 TM	1998.10.1
Landsat5 TM	2000.5.15
Landsat5 TM	2004.9.15
Landsat5 TM	2006.9.21
Landsat5 TM	2009.10.31
Landsat5 TM	2011.5.30
Landsat8 OLI	2013.10.26
Landsat8 OLI	2015.10.16
Landsat8 OLI	2023.10.22
Landsat8 OLI	2024.10.24

) were downloaded from the USGS website (https://earthexplorer.usgs.gov/, accessed on 14 January 2025) for 20 scenes with good vegetation growth between March and November and processed by atmospheric correction and radiometric calibration operations using ENVI 5.6 software [26,27]. Two perspectives of remote sensing photos from the same temporal interval were amalgamated to create a singular view encompassing the entire study area. The yearly average precipitation and temperature data for the research area were sourced from the National Meteorological Information Centre of China, while the digital elevation model data were acquired from the Geospatial Data Cloud (http://www.gscloud.cn), and the net primary productivity (NPP) data from 2001 to 2023 were obtained from Google Earth Engine (GEE) to support the analysis of this study.

2.3. Methods

2.3.1. NDVI and Vegetation Carbon Sinks

The normalized difference vegetation index (NDVI) quantifies the growth and density of vegetation by examining light reflected in various spectral bands, often red and near-infrared. The calculation is performed via the following formula:

$$NDVI={\displaystyle \frac{NIR-Red}{NIR+Red}}$$

(1)

NIR denotes reflectance in the near-infrared range, whereas Red signifies reflectance in the red light band. The formula constrains the NDVI values to a range between −1 and 1 by levelling the reflectance of the two bands. An NDVI value approaching 1 signifies dense and robust vegetation; when the NDVI value is close to 0 or negative, it may indicate bare soil, water bodies, or other nonvegetated surfaces [28].

The NDVI is effective at reflecting vegetation information, mainly because plant leaves contain a large amount of chlorophyll, which strongly absorbs red light and has a high reflectance of near-infrared light [29].

Vegetation carbon sinks are quantified by net ecosystem productivity (NEP), an important indicator for quantitatively assessing the functioning of terrestrial ecological carbon sinks (Figure 2). Shaoyang city is monitored for changes in vegetation carbon sinks due to soil erosion, and ecological degradation resulting from open-pit coal mining has resulted in less plant cover and reduced production.

2.3.2. Dimidiate Pixel Model

The dimidiate pixel model is a simple and efficient technique for remote sensing estimation. The linear mixed pixel decomposition method is now the most widely employed and uncomplicated method owing to its insensitivity to the impacts of remote sensing image radiation correction [30,31,32]. This method posits that a pixel’s surface comprises vegetated and nonvegetated regions and that the spectral data captured by remote sensing instruments are linearly weighted and synthesized from these two elements, with the weight of each component being proportional to its area within the pixel.

Typically, vegetation cover extraction via the binary search method is calculated on the basis of NDVI data via the following formula:

$$FVC={\displaystyle \frac{\left(NDVI-{NDVI}_{Soil}\right)}{\left({NDVI}_{Veg}-{NDVI}_{Soil}\right)}}$$

(2)

Among these, NDVI_soil represents the NDVI value for the soil portion, and NDVI_veg represents the NDVI value for the vegetation portion. We can assume that the minimum vegetation coverage rate for a given region is 0 and approximate the maximum value as complete coverage [33]. Therefore, the vegetation coverage formula can be changed to:

$$FVC={\displaystyle \frac{\left(NDVI-{NDVI}_{MIN}\right)}{\left({NDVI}_{MAX}-{NDVI}_{MIN}\right)}}$$

(3)

In the formula, NDVI_MIN represents the minimum NDVI value for the region, and NDVI_MAX represents the maximum NDVI value for the region. Since some mixed images may interfere with the results, it is best to select the maximum and minimum values within a certain range as a reference.

This study utilizes remote sensing datasets from April to October, as vegetation growth during this period is relatively robust, allowing FVC_MAX to be approximated as 1 and FVC_MIN to be 0. To mitigate noise interference, NDVI_MAX and NDVI_MIN are typically selected within a certain confidence interval, combined with the NDVI grayscale distribution across different stages of the study area. The 95% and 5% cumulative probabilities of the NDVI are selected as the NDVI_MAX and NDVI_MIN, respectively, for the study area.

2.3.3. Theil–Sen Median Slope and Mann–Kendall Trend Test

Owing to the susceptibility of linear regression analysis to outliers when examining interannual vegetation patterns, the Theil–Sen median slope approach was utilized in conjunction with the Mann–Kendall trend test [34,35,36], and the time series trend of the NDVI was analysed. This approach does not necessitate conformity to a particular distribution and possesses a robust capacity to mitigate outliers or measurement inaccuracies. The formula for the Theil–Sen slope method is presented as follows:

$$\beta =Median ({\displaystyle \frac{X_j-X_i}{j-i}})$$

(4)

In this context, Median() represents the median value, j and i denote the number of time series, and X_j and X_i represent the time series data, which in this paper are NDVI data. If β > 0, the time series tends to increase; if β < 0, the time series tends to decrease.

The Mann–Kendall test is a nonparametric statistical procedure. In contrast to other parametric testing methods, it does not require samples to conform to a specific distribution, is less vulnerable to the influence of outliers, and is more appropriate for ordinal variables [37,38]. The Mann–Kendall test has been widely and efficiently employed in research on hydrodynamic and meteorological trend changes to evaluate the significance of trends in runoff, precipitation, temperature, and other variables. The calculation formula is as follows:

$$\mathcal{S}=\sum _{i=1}^{n-1}\sum _{j=i+1}^{n}sgn\left(x_j-x_i\right)$$

(5)

$$\mathrm{s}\mathrm{g}\mathrm{n}(x_j-x_i)=\left\{\begin{array}{c}+1,x_j-x_i>0\\ 0,x_j-x_i=0\\ -1,x_j-x_i<0\end{array}\right.$$

(6)

$$Z=\left\{\begin{array}{ll}& {\displaystyle \frac{S}{\sqrt{Var(S)}}},(S>0)\\ & 0,(S=0)\\ & {\displaystyle \frac{S+1}{\sqrt{Var(S)}}},(S<0)\end{array}\right.$$

(7)

$$\mathrm{V}\mathrm{a}\mathrm{r}(S)={\displaystyle \frac{n(n-1)(2n+5)}{18}}$$

(8)

In this formula, n is the number of sequences, X_j and X_i are time series data, sgn() represents the sign function, Var(S) is the variance, and Z is the standardized statistic [39]. When n ≥ 8, the test statistic S can be approximated as a normal distribution. In this case, a two-tailed test is used. At a given significance level α = certain value, the value Z₁ − ${\displaystyle \frac{\mathsf{\alpha }}{2}}$ is obtained by consulting a table. When $\left|\mathrm{Z}\right|\ll \mathrm{Z}1-{\displaystyle \frac{\mathsf{\alpha }}{2}}$, the null hypothesis is accepted, i.e., the trend is not significant. When $\left|Z\right|>\mathrm{Z}1-{\displaystyle \frac{\mathsf{\alpha }}{2}}$, the null hypothesis is rejected, i.e., the trend is significant. In this work, the significance level is α = 0.05, so Z₁ − ${\displaystyle \frac{\mathsf{\alpha }}{2}}$ = ±1.96.

2.3.4. ARIMA Model

The autoregressive integrated moving average (ARIMA) model is fundamentally composed of three elements: the autoregressive component (AR), the differencing process (I), and the moving average component (MA). The core concept of the ARIMA model is to leverage historical data to forecast future trends. The label value at a specific moment is affected by both the label values from a prior period and stochastic events occurring during that period. The ARIMA model posits that label values oscillate along a long-term trend, which is determined by historical label values, but the oscillations are affected by stochastic occurrences within a specified timeframe. Furthermore, the long-term trend is not inherently steady [40,41].

3. Results

We examined the primary factors influencing forest restoration by analysing alterations in plant cover within each mining area and adjacent regions. This study involved the processing of Landsat satellite data and NPP data to derive the NDVI, FVC, and alterations in vegetation carbon sequestration (

Table 2. Classification of vegetation cover levels.

Vegetation Cover Value	Vegetation Cover Rating
0–0.2	Extremely low coverage
0.2–0.4	Low coverage
0.4–0.6	Medium coverage
0.6–0.8	High coverage
0.8–1	Extremely high coverage

) as analysis indicators and obtained the spatiotemporal changes, change trends, and correlations between various factors in each mining area [42,43]. These results indicate that the vegetation restoration nodes in the various study mining areas are generally consistent with the mineral resource planning scheme of Shaoyang city, Hunan Province. Therefore, it is believed that the primary factor influencing vegetation restoration is the implementation of local policies. The study results offer a theoretical foundation for vegetation regeneration in mining zones across different locations, facilitating the formulation of relevant policies by national and local authorities to address vegetation restoration issues.

3.1. Analysis of Vegetation Cover in Mining Areas

3.1.1. Temporal Changes in Vegetation Cover

The FVC in the entire study area (Figure 3 and Figure 4) exhibited a pattern of initial decline followed by a subsequent increase between 1998 and 2024. The average FVC in the northern mining area showed an overall increasing trend, whereas that in the southern mining area exhibited a decreasing trend. The alterations in the central mining region were more intricate, featuring both augmentations and reductions.

From 1998 to 2009, vegetation cover in the northern mining area experienced a significant decline, particularly in 2009, with a notable increase in areas with extremely low and low vegetation cover. Starting in 2011, vegetation cover improved, with an increase in areas with high vegetation cover, reaching a relatively high level by 2024. In the central mining area, vegetation coverage showed a declining then increasing trend from 1998 to 2006, continued to decline from 2006 to 2009, reaching its lowest point in 2009, characterized by predominant areas of extremely low and low vegetation coverage. After 2011, the vegetation coverage rebounded, resulting in an expansion of regions with dense vegetation. In the southern mining region, vegetation coverage exhibited a declining trend from 1998 to 2009, culminating in a notable increase in regions characterized by extremely low and low vegetation coverage in 2009. Following 2011, vegetation coverage increased, with areas of high and extremely high vegetation coverage increasing; however, it subsequently diminished in 2023 and 2024. The fundamental alterations in vegetation throughout the three mining regions were predominantly synchronized and influenced by national and Shaoyang city coal resource demands, mining methods, and regional regulatory constraints.

3.1.2. Spatial Analysis of Vegetation Cover

Owing to basic mineral resource management policies, soil and water conservation, and land reclamation plans implemented by Shaoyang city, the vegetation in all three study areas has undergone varying degrees of restoration. The northern and central mining areas have similar elevations. As shown in Figure 5 and Figure 6 and

Table 3. Percentage of vegetation coverage in the central mining area.

Year	Extremely Low (0–0.2)	Low (0.2–0.4)	Medium (0.4–0.6)	High (0.6–0.8)	Extremely High (0.8–1)
1998 percentage	3%	13%	37%	32%	14%
2000 percentage	3%	15%	49%	30%	2%
2004 percentage	15%	17%	10%	33%	26%
2006 percentage	0%	8%	30%	32%	30%
2009 percentage	36%	34%	18%	9%	3%
2011 percentage	27%	27%	24%	11%	11%
2013 percentage	2%	38%	43%	14%	2%
2015 percentage	0%	7%	31%	54%	8%
2023 percentage	1%	6%	38%	45%	9%
2024 percentage	0%	4%	26%	33%	37%

and

Table 4. Percentage of vegetation coverage in the northern mining area.

Year	Extremely Low (0–0.2)	Low (0.2–0.4)	Medium (0.4–0.6)	High (0.6–0.8)	Extremely High (0.8–1)
1998 percentage	4%	13%	23%	39%	23%
2000 percentage	8%	17%	58%	17%	0%
2004 percentage	39%	27%	24%	9%	2%
2006 percentage	20%	24%	41%	15%	0%
2009 percentage	74%	20%	6%	0%	0%
2011 percentage	39%	12%	16%	17%	18%
2013 percentage	14%	27%	32%	19%	8%
2015 percentage	2%	35%	33%	25%	6%
2023 percentage	2%	18%	36%	38%	6%
2024 percentage	0%	19%	29%	44%	8%

, mining activities in the central mining area were not particularly prominent between 1998 and 2006, whereas the northern mining area experienced severe coal mining activities during this period, leading to a decrease in vegetation coverage with no signs of improvement. Between 2006 and 2009, a city-wide mineral survey project was conducted, expanding mining operations in most areas concentrated in the northern and central regions, resulting in large-scale mining geological disasters, soil and water pollution, and vegetation damage. The high vegetation coverage area decreased significantly, with low vegetation coverage reaching 70%. Between 2001 and 2015, there was some recovery. After 2015, Shaoyang city optimized its mining policies and coordinated the allocation of external resources, leading to a gradual recovery of vegetation coverage in the northern and central mining areas, with high-coverage areas accounting for approximately 50% of the total area. The successful implementation of mineral resource planning policies has significantly increased the city’s environmental protection awareness and emphasis on mine environmental restoration and governance during mineral resource development. The southern mining area, located in Wugang city, has well-developed karst terrain. Mining activities between 1998 and 2004 caused significant impacts and harm to groundwater resources, leading to a rise in low vegetation coverage levels (Figure 5 and Figure 6,

Table 5. Percentage of vegetation coverage in the southern mining area.

Year	Extremely Low (0–0.2)	Low (0.2–0.4)	Medium (0.4–0.6)	High (0.6–0.8)	Extremely High (0.8–1)
1998 percentage	1%	6%	23%	49%	22%
2000 percentage	4%	29%	34%	28%	5%
2004 percentage	32%	16%	13%	18%	21%
2006 percentage	1%	16%	34%	28%	21%
2009 percentage	11%	14%	20%	21%	33%
2011 percentage	29%	26%	13%	16%	15%
2013 percentage	10%	28%	33%	24%	6%
2015 percentage	0%	21%	37%	34%	8%
2023 percentage	0%	0%	16%	38%	46%
2024 percentage	0%	2%	16%	40%	42%

). Between 2004 and 2009, the southern mining area faced insufficient coal reserves, resulting in reduced mining activities and providing space for vegetation growth. However, from 2009 to 2011, the local government applied to the national authorities for exploration and mining rights and successfully discovered large coal reserves. During this period, extensive mining activities led to a severe decline in vegetation coverage, with low coverage reaching 54% in 2011. From 2013 to 2024, the local government subsequently recognized the issues arising from mining activities and implemented relevant remediation policies, resulting in a gradual recovery of vegetation coverage.

3.1.3. Analysis of Vegetation Cover Change Trends and Future Predictions

Between 1998 and 2024, there were significant changes in vegetation coverage across the three mining areas in Shaoyang city. As shown in

Table 6. Normalized difference vegetation index (NDVI) change trends in the northern mining area.

SNDVI	Z Value	Trend of NDVI	Percentage
≥0.0005	≥1.96	significant improvement	2%
≥0.0005	−1.96–1.96	slight improvement	23%
−0.0005–0.0005	−1.96–1.96	stable and unchanging	35%
<−0.0005	−1.96–1.96	slightly degraded	27%
<−0.0005	<−1.96	significant deterioration	13%

The S_NDVI values range from −0.0005 to 0.0005. The value of pixels with Z > 1.96 or Z < −1.96 is extremely small, so such pixels are classified as stable and unchanged.

,

Table 7. Normalized difference vegetation index (NDVI) change trends in the southern mining area.

SNDVI	Z Value	Trend of NDVI	Percentage
≥0.0005	≥1.96	significant improvement	3%
≥0.0005	−1.96–1.96	slight improvement	31%
−0.0005–0.0005	−1.96–1.96	stable and unchanging	43%
<−0.0005	−1.96–1.96	slightly degraded	12%
<−0.0005	<−1.96	significant deterioration	11%

The S_NDVI values range from −0.0005 to 0.0005. The value of pixels with Z > 1.96 or Z < −1.96 is extremely small, so such pixels are classified as stable and unchanged.

and

Table 8. Normalized difference vegetation index (NDVI) change trends in the central mining area.

SNDVI	Z Value	Trend of NDVI	Percentage
≥0.0005	≥1.96	significant improvement	8%
≥0.0005	−1.96–1.96	slight improvement	29%
−0.0005–0.0005	−1.96–1.96	stable and unchanging	43%
<−0.0005	−1.96–1.96	slightly degraded	19%
<−0.0005	<−1.96	significant deterioration	1%

The S_NDVI values range from −0.0005 to 0.0005. The value of pixels with Z > 1.96 or Z < −1.96 is extremely small, so such pixels are classified as stable and unchanged.

and Figure 7, the vegetation changes in the three mining areas significantly differed. The southern mining area showed the most pronounced degradation trend, with “severe degradation” accounting for 13% (Z < −1.96). When combined with “mild degradation,” the total degraded area reached 40%, indicating significant ecological pressure on the mining area. The improved areas accounted for only 25% (2% significantly improved + 23% slightly improved), and the stable areas accounted for 35%, indicating a slow vegetation recovery process. In contrast, the northern mining area had a degraded area accounting for 23% (11% significantly degraded + 12% slightly degraded), improved areas accounting for 34% (3% significantly improved + 31% slightly improved), and stable areas reaching 43%, demonstrating better ecological stability than the southern area. The central mining area performed best, with only 20% of the area degraded (1% significantly degraded + 19% slightly degraded), 37% of the area improved (8% significantly improved + 29% slightly improved), and 43% of the area stabilized, indicating significant vegetation restoration effects.

Combining vegetation coverage data, the southern mining area experienced a sharp increase in “high” and “very high” coverage proportions to 84% (38% + 46%) after 2023, possibly related to artificial restoration measures. However, the lag in the NDVI trends indicates that recovery is still in its early stages [44]. In the northern mining area, the proportions of “high” and “extremely high” coverage reached 52% (44% + 8%) in 2024, but the improvement in the NDVI was relatively low, suggesting insufficient improvement in coverage quality. In the central mining area, “extremely high” coverage reached 37% in 2024, and the proportion of significant improvement in the normalized difference vegetation index (NDVI) was the highest (8%), reflecting increased ecological restoration efficiency.

To predict future vegetation coverage levels in each mining area, ARIMA models were used to analyse and calculate data, yielding geometric mean values for the NDVI in each mining area in the future (Figure 8). All three major mining areas experienced vegetation damage during the mining period, leading to fluctuations and declines in vegetation coverage levels. However, through restoration measures and policy support, vegetation coverage levels gradually recovered to higher levels. The ARIMA model does not “actively focus” on human factors, but human factors can be incorporated into the data processing, so the prediction results are considered to have certain reference value, and the subsequent vegetation restoration development can be continuously observed.

3.2. Changes in Vegetation Carbon Sinks

The vegetation carbon sink capacity, as reflected by the NEP data (Figure 9 and Figure 10), reached a value of 345 g·m²·year⁻¹ in the northern mining area during the early stages of mining. As mining activities progressed, the annual average vegetation carbon sink capacity fluctuated between 2004 and 2009, with a decrease in carbon sink levels and an increase in the percentage of carbon sources. Subsequently, it gradually recovered, with high carbon sink levels accounting for more than 50% of the total. The central and southern mining areas showed more pronounced changes, with fluctuations in carbon sink levels between 2000 and 2009. By 2009, the NEP values had dropped to the lowest levels across all stages. However, owing to regional differences in vegetation, the carbon sink levels in these mining areas remained higher than those in the northern area, with only a small increase in low-carbon sink zones. The southern mining area maintained its recovery after 2010, whereas the central mining area experienced a steady increase in carbon sink levels, recovering to the initial mining levels by 2013–2023, with an average annual vegetation carbon sink capacity reaching 478 g·m²·year⁻¹.

4. Discussion

4.1. Contributions of Natural Elements to the Restoration of Vegetation

Vegetation restoration can be affected by temperature and rainfall. According to relevant data analysis, there is a low correlation between natural factors and vegetation coverage, meaning that rainfall and temperature are conducive to vegetation growth and restoration [45]. Figure 11 shows that the correlation coefficient between rainfall and vegetation cover is 0.18, indicating that rainfall has a slight positive effect on vegetation restoration in the study area. However, temperature also affects vegetation restoration. Suitable temperatures may promote vegetation growth, but excessively high temperatures may lead to increased water evaporation, thereby having an adverse effect on vegetation growth [46,47]. A comparison of the two sets of historical remote sensing images in Figure 12 shows that A indicates that after mining, policy measures were implemented to promote remediation and restoration, resulting in good vegetation coverage, while B indicates that no policy measures were implemented to promote remediation and restoration, resulting in significantly insufficient restoration under natural conditions. The correlation coefficient between temperature and vegetation cover in the Figure 11 is 0.41, indicating a positive correlation. However, the negative correlation between temperature and precipitation (correlation coefficient of −0.3) suggests that precipitation may decrease during periods of higher temperatures, which could have a negative impact on vegetation restoration.

Importantly, the associations between the vegetation carbon sink capacity (NEP) and temperature and precipitation are inconsistent. The correlation coefficient between the NEP and precipitation is −0.15, indicating that excessive precipitation is not conducive to vegetation carbon sinks. However, the correlation coefficient between the NEP and temperature is 0.59, with a relatively high value indicating that climatic factors have a relatively strong influence on carbon sinks. The FVC and NEP also exhibit a positive correlation, with a value of 0.81. Therefore, when the impacts of temperature and precipitation on vegetation restoration are considered, it is necessary to balance these two factors during the restoration process to achieve the best vegetation recovery outcomes [48,49].

4.2. The Impact of Land Restoration and Coal Mining Policies

From 1998 to 2024, the open-pit coal mining policy in Shaoyang city, Hunan Province, underwent a transformation process from initial standardization to strengthened supervision, then to green development and sustainable development, and finally to high-quality development. These policy changes reflect the Chinese government’s policy orientation in terms of environmental protection, work safety, and sustainable development, as well as its growing awareness and response to the impact of open-pit coal mining activities. Land restoration measures have played a positive role in vegetation restoration and carbon sequestration [25,50]. Shaoyang city has divided its mining economic zone into two subcategories: the eastern subcategory and the western subcategory. The eastern subcategory includes the northern mining area and the central mining area studied in this research, whereas the southern mining area belongs to the western subcategory. Owing to slight differences in policies between the two subcategories, this study examines alterations in vegetation cover and carbon sequestration across several mining areas and their adjacent regions to determine the principal factors affecting vegetation restoration and carbon sequestration. The increase in vegetation coverage in the northern mining region is correlated with an increase in high-capacity carbon sink zones, which may be attributable to the execution of land restoration initiatives. Moreover, modifications to coal mining policies have impacted vegetation restoration. The execution of these policies mitigates the harm inflicted by coal mining on vegetation and fosters vegetation restoration.

4.2.1. Mineral Resource Extraction Policy

From 1998 to 2010, during the period of extensive development, the mining areas in the central, northern, and southern regions were all dominated by resource exploitation. Unregulated open-pit coal mining in these three areas led to a significant decline in vegetation coverage. The data revealed that the FVC index in the northern mining area sharply decreased, and the area of primary forest in the southern mining area decreased by 20%. During this period, all mining areas relied primarily on the Mineral Resources Law (1996), which only regulated mining activities and did not establish an ecological compensation mechanism. From 2011 to 2017, during the transition period, the northern mining area pioneered the “mine rehabilitation” project to restore the severe damage caused by earlier extensive mining, improve vegetation coverage levels, and increase carbon sequestration capacity. The central mining area relied on the “Shaoyang City Mineral Resources Master Plan (2008–2020),” which strictly adhered to national and municipal mineral resource planning to gradually restore vegetation coverage. The southern mining area, located within the Xuefeng Mountain Ecological Barrier Zone, enforces the “Hunan Province Ecological Protection Red Line” more strictly, closing five coal mines in ecologically sensitive areas. During the 2018–2024 (intensive development period), the three regions implemented the “Mining Geological Environment Governance and Restoration Fund Management Measures” in a differentiated manner. Additionally, the southern mining area was the first among the three mining areas to initiate historical legacy mine restoration projects in 2017, responding to the national call to rapidly restore vegetation and achieve high levels of vegetation coverage (

Table 9. Reference Policies.

Serial Number	Year	Policy	Category
1	1996	Minerals Resources Law	mineral assets
2	2006	Comprehensive Water Pollution Control Plan for the Xiangjiang River Basin	mineral assets
3	2007	Shaoyang City Mineral Resources Master Plan (2008–2020)	mineral assets
4	2009	Notice on the Launch of the Second Round of Municipal and County-level Mineral Resource Master Planning (Xiang Tuozhi Ban Fa [2009] No. 8)	mineral assets
5	2010	Implementation Plan for the Strategic Action for Geological Exploration Breakthroughs in Hunan Province	mineral assets
6	2013	Implementation Strategy for Heavy Metal Contamination Mitigation in the Xiangjiang River Basin (2013)	mineral assets
7	2018	Measures for the Administration of the Mine Geological Environment Restoration Fund	mineral assets
8	2021	Hunan Province Mineral Resources Master Plan (2021–2025)	mineral assets
9	2010	Notice on Further Strengthening Geological Disaster Prevention Work	Rehabilitation of the environment
10	2013	Shaoyang City’s 14th Five-Year Plan for Environmental Protection	Rehabilitation of the environment
11	2015	Perspectives on Expediting the Development of Ecological Civilization	Rehabilitation of the environment
12	2016	Regulations on the Management and Restoration of Mine Geological Environment	Rehabilitation of the environment
13	2018	Perspectives on Thoroughly Enhancing Ecological and Environmental Safeguards and Determinedly Prevailing in the Fight Against Pollution	Rehabilitation of the environment
14	2020	Three-Year Action Plan for Green Mine Construction in Hunan Province (2020–2022)	Rehabilitation of the environment
15	2022	Ecological Restoration Plan for Land Space in Shaoyang City (2021–2035)	Rehabilitation of the environment
16	2023	Notice on Organizing Applications for the 2024 Historical Legacy Abandoned Mine Ecological Restoration Demonstration Project	Rehabilitation of the environment
17	2022	Ecological Restoration and Management Strategy for Abandoned Mines from Previous Eras During the 14th Five-Year Plan Period	Ecological restoration, land resources
18	2023	Ecological restoration + Chinese herbal medicine cultivation project	Rehabilitation of the environment
19	2023	Mining Ecological Protection and Restoration Plan	Rehabilitation of the environment
20	2023	Shaoyang City Shuangqing District Historical Legacy Mine Ecological Restoration Demonstration Project Shaoyang City Shuangqing District Project Performance Target Table (2022–2024)	Rehabilitation of the environment
21	2024	2024 Public Notice on the Completion of Ecological Restoration of Abandoned Mines with Historical Legacy Issues	Rehabilitation of the environment
22	2024	Minerals Resources Law	Rehabilitation of the environment
23	2020	Ecological Protection Project for the Mountains, Rivers, Forests, Fields, Lakes, and Grasslands of Shaoyang County	ecological conservation project
24	2019	Shaodong County Mine Ecological Restoration and Land Reclamation Plan	Land reclamation and ecological restoration projects
25	2017	Wugang City Historical Mine Ecological Restoration Project	Historical Mine Ecological Restoration Project
26	2023	Ecological restoration + oil tea specialty economic project	Rehabilitation of the environment
27	2023	Ecological restoration + Chinese herbal medicine cultivation project	Rehabilitation of the environment
28	2023	Ecological restoration + utilization of waste soil and stone materials + state-owned construction land projects	Ecological restoration + land resources

).

4.2.2. The Impact of Ecological Restoration Policies

The soil layer in the restoration area has been largely damaged, so vegetation restoration is mainly carried out by replanting imported soil. The central mining area implemented the “Shaoyang County Mountain, Water, Forest, Farmland, Lake, and Grassland Ecological Conservation Project” in 2020 to restore vegetation coverage in the mining area. In 2023, based on Shaoyang County’s ecological restoration plus oil tea specialty economic project, oil tea trees were planted to promote the development and strengthening of Shaoyang County’s oil tea industry, supporting the rural revitalization strategy; the southern mining area experienced a significant increase in the “extremely high” coverage area in 2023, while the Wugang city pilot project explored an ecological restoration model combining “ecological restoration + utilization of abandoned soil and rock materials + state-owned construction land,” restoring 428 acres of mine area, generating 2.24 million tons of usable abandoned soil and rock materials through slope reduction, and adding 175 acres of state-owned construction land. Additionally, the southern mining area implemented three-dimensional restoration in accordance with the “Several Opinions on Comprehensively Promoting Green Development in the Mining Industry” (Xiang Zheng Ban Fa [2019] No. 71) issued by the Hunan Provincial People’s Government, establishing a three-tier system of “trees–shrubs–grass,” with the NDVI value gradually increasing and the biodiversity index improving by 1.8 times. The northern mining area adopts an ecological restoration plus traditional Chinese medicine model, transforming abandoned mines into farmland for growing traditional Chinese medicine herbs and advancing the construction of the Lianqiao Medical Special Town. Moreover, in accordance with the “Ecological Restoration and Governance Plan for Historically Abandoned Mines During the 14th Five-Year Plan Period,” the ecological restoration and governance of 19 closed and withdrawn mines have been completed, with 11 cases filed to address the “restoration challenge.” Furthermore, after the 14th Five-Year Plan, the area of high vegetation coverage in each mining area will increase by more than 50%. Refer to Table 9.

The main natural soils in the study area are red soil and paddy soil, which were damaged during the mining period, resulting in a decline in soil fertility. After ecological restoration through soil replacement planting, vegetation coverage increased and gradually returned to the level seen at the beginning of mining, indicating that the restored soil is suitable for vegetation growth and that the policy has been successfully implemented. However, it is also necessary to test the heavy metal content of the vegetation to prevent heavy metal pollution caused by tailings.

5. Conclusions

Changes in vegetation cover and carbon sequestration in three open-pit coal mining areas in Shaoyang city and their driving factors were systematically analysed on the basis of remote sensing data from 1998 to 2024. The results show the following:

(1) From 1998 to 2024, the vegetation cover in the study area first declined but then increased. From 1998 to 2009, mining activities were frequent and severe, causing large-scale vegetation destruction and a decline in vegetation carbon storage. Subsequently, attention began to be given to ecological restoration, and various ecological restoration projects were carried out to restore damaged vegetation cover. Recovery to 2023, the vegetation carbon sink has increased to a high level.

(2) The northern mining area has had low vegetation coverage for a long time and has been most severely affected by mining. The central mining area was the first to show signs of recovery, while the southern mining area, after ensuring a certain level of vegetation coverage, recovered to a high level at the fastest rate among the three mining areas through intensive management, and vegetation carbon sequestration also increased accordingly.

(3) Policy direction is the core external force driving vegetation restoration in mining areas. The significant improvement in vegetation coverage and increase in vegetation carbon sink capacity in central and southern mining areas are closely related to policy implementation, whereas northern mining areas still face issues of inadequate policy enforcement, which requires further observation. This study reveals the critical role of policy interventions at various time points in overcoming the limitations of natural recovery, particularly in areas with insufficient precipitation, where artificial restoration measures can effectively compensate for climatic shortcomings. The synergistic effects of policy implementation, human factors, and environmental factors on vegetation restoration in open-pit mining areas and their surroundings are profound. This study provides valuable references for vegetation restoration projects in other mining areas. It is worth noting that further observation is needed to determine whether the imported soil is compatible with the local environment and whether soil quality has deteriorated.