Evaluation of the effect of ecological restoration in mineral resource cities and analysis of driving factors

Abstract

The rapid expansion of the mining industry, while propelling economic growth, has concurrently inflicted severe damage on local ecosystems. The challenge of harmonizing the exploitation of mineral resources with the preservation of ecological environments is critical, particularly in densely mined regions, where achieving sustainable development is an urgent issue that demands immediate resolution. In this study, the Remote Sensing Ecological Distance Index (RSEDI) was developed based on the Google Earth Engine (GEE) platform, employing data from the Landsat series. Five indicators—greenness, humidity, dryness, heat, and black particulates—were selected and integrated using the distance function method. The city of Dexing was selected as the study area, with research conducted at both the municipal and township levels. The results illustrated that: (1) Between 2000 and 2020, Dexing City’s ecological quality demonstrated a trend of initial decline followed by improvement. The region, as a whole, showed a pattern of enhanced greenness and moisture levels, along with reductions in dryness, temperature, and the spatial extent of atmospheric indices. (2) From a spatial perspective, the areas of ecological damage within the study region were predominantly located in the central and western mining zones. The shift in the focus of ecological improvement initially moved northeastward, then northwestward, and finally southwestward. In contrast, the migration pattern of the focus for ecological deterioration followed an opposite trajectory to that of ecological improvement. (3) Dynamic monitoring and analysis of RSEDI during different periods revealed that the ecological quality of Dexing City underwent an initial deterioration followed by an improvement. Overall, the ecological environment in Dexing City is progressively improving. (4) The results of the study provide valuable insights into the assessment of ecological restoration in mineral resource-based cities and contribute to the formulation of regional sustainable development strategies.

Introduction

Since the 20 th century, the world has witnessed a significant surge in population and rapid economic development. Concurrently, this growth has led to widespread vegetation degradation, land deterioration, a dramatic reduction in biodiversity, and water pollution. These issues collectively exert a substantial negative impact on the global ecological environment^1,2,3. Since the inception of its economic reforms and open-door policies, China has accomplished notable economic achievements, wherein the development of mineral resources has played a key role. However, while propelling socio-economic growth, such exploitation has inevitably inflicted considerable damage on the ecological environment, posing a severe threat to human health and constraining sustainable regional development^4,5. Mining activities frequently result in land occupation and degradation, leading to soil erosion, vegetation damage, and ecosystem disruption⁶. Ecological restoration in mining areas is recognized as a crucial measure for achieving environmental protection, sustainable resource utilization, and regional sustainable development. It holds indispensable importance and necessity for mitigating the negative impacts of mining operations on the environment and for restoring and preserving ecosystems^7,8,9. Restoration of degraded ecosystems in mining areas is observed to enhance land value, increase carbon credit potential, improve water protection and quality, and elevate ecosystem services¹⁰. To monitor and address ecological issues related to mining activities promptly and effectively, it is crucial that rapid and accurate assessments of both the mining operations and subsequent site restoration efforts are conducted.

In the field of regional ecological restoration research, the United Nations General Assembly passes the “UN Decade on Ecosystem Restoration” resolution in 2021¹¹. This resolution explicitly underscores the critical role of ecosystem restoration in achieving the Sustainable Development Goals¹². Ecosystem restoration is prioritized as a primary global initiative, aimed at conserving biodiversity, preventing desertification and land degradation, and mitigating the effects of climate change. Consequently, numerous ecosystem conservation and restoration projects are initiated by governments both domestically and internationally to mitigate environmental degradation and advance the Sustainable Development Goals^13,14,15,16. For instance, in 2000, a 9,000-hectare project was initiated in northwestern Minnesota, USA, aimed at restoring tallgrass prairie previously disturbed by gravel mining activities. Ecological restoration measures were implemented from the perspectives of landscape hydrology, native vegetation, and habitat. As a result, 1,240 hectares of wetlands were restored, and 8,100 hectares of prairie were replanted, leading to improved water quality and the reestablishment of native vegetation¹⁷. A series of supportive policies and measures were implemented in Iceland with the aim of promoting ecological restoration of degraded rangelands and enhancing land management practices. These strategies effectively facilitated the recovery of damaged ecosystems, maintained their integrity, and contributed to sustainable development in the region¹⁸. Since 1978, China executed various large-scale ecological projects, including the Three-North Shelter Forest Program, ecological restoration of abandoned historical mining sites, and the conversion of farmland to forest initiatives. Significant progress was made in regional ecological restoration and in advancing sustainable development as a result of these efforts^19,20,21.

In the construction of the ecological evaluation index system in mining areas. In the 1960 s, international efforts began to focus on ecological and environmental assessment research. Following the enactment of the United States’ National Environmental Policy Act, which first established the environmental impact assessment system, several countries including Germany, Australia, Sweden, and Japan subsequently enacted relevant legislation^22,23,24. As an ecological endeavor grounded in ecological theory, the assessment of ecological environments in mining areas represents a systematic project that integrates multiple disciplines and scales. Currently, there is no unified standard system recognized both domestically and internationally^25,26. Typically, scholars select one or multiple evaluation indicators that are scientifically sound, systematic, and concise to construct an assessment framework. This framework is then utilized to conduct either quantitative or qualitative evaluations of regional ecological environmental quality^27,28,29. The core of environmental assessment lies in the evaluation index system, which encompasses two key components: The selection of indicators and the assignment of their corresponding weights are determined³⁰. An evaluation index system is typically established by scholars using the Pressure-State-Response (PSR) framework^31,32. In the selection of indicators, emphasis is currently placed on the environmental impact of mining operations and anthropogenic disturbances. Primary indicators typically encompass aspects of mine resources, ecological environment, social factors, and economic elements. Secondary indicators include variables such as vegetation cover, waste emissions, landscape patterns, topography, mine development, and hydro-climatic conditions. Tertiary indicators are further detailed, including geological hazards, landscape fragmentation, physicochemical properties of vegetation, soil characteristics, water bodies, and population density^33,34,35. Currently, research on ecological environmental evaluation indicator systems for mining areas primarily focuses on the following aspects: indicator systems for assessing the carrying capacity of mine resources, ecological vulnerability in mining regions, ecological security of mines, and the health of ecosystems within mining zones^36,37,38. The ecological environmental evaluation indicator systems for mining areas are generally characterized by a large number of indicators, clearly defined structural hierarchies, and systemic features.

In terms of the weighting of indicators for evaluating the ecological environment in mining areas. The determination of weights for ecological environmental evaluation indicators is categorized into subjective and objective weighting methods. Subjective weighting relies entirely on the intentions of decision-makers for weight allocation, which tends to be less objective. In contrast, objective weighting methods utilize the relationships within raw data and apply certain mathematical principles to ascertain weights, resulting in outcomes that are independent of human subjective judgment and possess greater objectivity. However, these methods do not reflect the varying degrees of importance that decision-makers may assign to different indicators. The integration of subjective and objective weighting methods for determining weights, ensuring both respect for expert opinions and consideration of the objective information from raw data, is a focal point of current research³⁹. A novel weight optimization combination, integrating the Analytic Hierarchy Process (AHP) and Entropy Weight Method (EWM) to blend subjective and objective weighting, has been employed for ecological environmental assessment in the coal mining base of Northern Shaanxi. The evaluation outcomes were compared with field survey data, revealing a general consistency in the conclusions⁴⁰.

With the advancement of science and technology, satellite remote sensing now offers extensive area coverage, long-term, and multi-scale surface information. As a result, the analysis of remote sensing data is increasingly becoming a principal method for the assessment of ecological environmental quality in mining areas⁴¹. In the context of mining area ecological assessments, remote sensing technology primarily relies on data products or ecological inversion indicators to reflect the quality of regional environments influenced by multiple factors. Considering the unique characteristics of iron mining areas, the Iron Mining Remote Sensing Ecological Index (IM-RSEI) has been developed to assess regions with intensive iron ore exploitation⁴². The PSR framework has been employed, considering the extent of coal dust pollution in mining regions, to construct a Coal Mine Ecological Index (CMEI) for the extraction of environmental data from a particular coalfield. The overall accuracy of this extraction is commendable⁴³. The RSEDI is designed for the assessment of both temporal and spatial shifts in Urumqi city’s ecological quality. This model is advantageous due to its objective weighting and convenience, overcoming the one-sidedness of single-factor evaluations⁴⁴.

Currently, a significant volume of research is dedicated to the remote sensing surveillance of ecological environments within mining zones. However, the consideration of mining characteristics, such as blasting, excavation, and transportation, which generate significant amounts of dust and fine particulate matter, is relatively limited. These particles can remain suspended in the air and be inhaled by individuals, posing not only health risks to humans but also causing severe environmental pollution. Research indicates that the dispersion of dust and particulate matter not only adversely affects the atmospheric environment but can also, under certain conditions, lead to explosions and fires^45,46,47.

In summary, while there is considerable research on regional ecological environment monitoring using remote sensing technology, the consideration of mining characteristics inherent to mining areas is minimal. This paper, drawing upon the RSEI framework, introduces an index of black particulates matter and employs a distance function method to construct the RSEDI index. The study focuses on Dexing City, a representative mining-resource-based city. This paper aimed that: (1) Utilizing the GEE platform, a scientific construction of the RSEDI index is achieved through the integration of long-term, multi-source remote sensing data; (2) The temporal dynamics of ecological quality in Dexing City from 2000 to 2020 were reconstructed and analyzed, identifying the spatial autocorrelation and dynamic shifts in the migration of ecological quality; (3) The influence of mining operations on the ecological quality of urban areas is exhaustively investigated and assessed. This research offers a foundation for the formulation of sustainable development and ecological rehabilitation strategies in cities reliant on mineral resources.

Materials and methods

Study area

Situated in Jiangxi Province’s northeastern region, China, Dexing City encompasses an area of 2,101 km², as depicted in Fig. 1. The local climate is characterized by warmth and abundant rainfall, placing it within the humid monsoon region of the mid-subtropics. This area is distinguished by its extended frost-free period and significant diurnal temperature variation, typical of mountainous microclimates.

The research area is rich in mineral resources, and its underground contains abundant treasures, among which copper reserves are extremely rich, while gold, silver and other metal mines have a long history, with more than 1,300 years of mining and metallurgical history, and are still an important base of non-ferrous metal minerals in China. Owing to a history of extensive mining activities, numerous legacy mines have been abandoned, resulting in severe damage to the local ecological environment. Owing to the extensive and unregulated mineral exploitation practices in the past, numerous legacy mines persist, causing significant damage to the local ecological environment. Prolonged unsustainable mining practices result in severe soil erosion within the mining areas, while heavy metal concentrations in sections of the Le’an River frequently exceed safety standards. The soil in parts of these regions is contaminated by heavy metals, significantly impacting local industrial and agricultural productivity as well as disrupting the daily lives of residents. The excessive concentrations of copper, lead, and zinc ions, in particular, cause substantial deterioration of both water quality and soil conditions. The combined effects of these factors severely impair the ecosystem services in Dexing City, including water resource conservation, soil retention, and biodiversity protection. In 2018, the People’s Government of Dexing City established a specialized team to undertake ecological restoration of legacy mines. Over the past five years, the local government has invested over 200 million yuan, successfully completing the ecological restoration of all 48 legacy mines. The measures and policies implemented have effectively supported the sustainable development of both the economy and society in Dexing City, resulting in a comprehensive improvement of the local ecological environment.

Fig. 1

Location map of Dexing city.

Data collection and processing

This study utilized the GEE platform for the collection and processing of remote sensing imagery. The study conducted an analysis of temporal and spatial dynamics in Dexing City by coding in JavaScript on a cloud platform. Utilizing the Landsat series of remote sensing images as the data source, as shown in Table 1, images from the years 2000, 2008, 2013, and 2020 were extracted. Five indices—greenness, wetness, dryness, heat, and atmospheric conditions—were selected. The RSEDI was then constructed using the distance function method (Fig. 2). Moreover, the Moran’s I index for global spatial autocorrelation, along with the Local Indicators of Spatial Association (LISA), were utilized to examine the spatial autocorrelation of the ecological environment. Additionally, the center of gravity migration model was applied to illustrate the spatial migration trends of ecological quality within the study area.

The cloud platform for remote sensing big data, exemplified by GEE, exhibits distinct advantages in ecological monitoring. Leveraging the platform’s vast repository of remote sensing data and robust cloud computing capabilities, it significantly reduces the time required for image processing. The application of this platform in long-term terrestrial ecological monitoring research has become increasingly mature⁴⁸. The study utilized the Landsat SR T1 image dataset integrated on GEE, which has undergone preprocessing steps such as geometric and atmospheric corrections, effectively spanning the study’s temporal series. Images from May to October across the years were subjected to cloud removal and selection processes, excluding images with extensive disturbances within the study area. Additionally, gaps caused by minor cloud interference were addressed by applying a median composite of images from preceding and succeeding periods (Table 1).

Fig. 2

Working framework for data processing.

Table 1 Information on data sources in the study area.

Methodology

RSEDI model construction

In this article, ecological indicators are selected based on the United Nations Sustainable Development framework and a review of the literature. Vegetative characteristics are considered the most focal point and extensively studied factor in ecological monitoring research within mining areas. Vegetation indices, which are derived from the combination of different spectral bands of remote sensing data, serve as indicators of vegetative health. Among these, the NDVI is recognized as the optimal method for analyzing vegetation in mining regions. Besides, factors such as soil moisture content, the degree of desiccation of buildings, and surface temperature also play essential roles in assessing the ecological quality in mining regions^49,50. Soil moisture is utilized in monitoring land degradation in mining areas and stands as one of the key factors in studying ecological environmental changes. Within an optimal range, soil moisture supplies essential water for plant growth without inhibiting root respiration. An increase in surface moisture can promote plant development, making this indicator closely related to vegetation health⁵¹. The study area contains bare land and construction sites, which lead to the “desiccation” of the surface, thereby causing a degree of ecological degradation and negatively impacting the quality of the ecological environment. Surface temperature holds significant importance for the survival of humans and other life forms. Excessive surface temperatures accelerate soil moisture evaporation, resulting in declining soil fertility, which impacts root development in plants. Consequently, vegetation growth and the evaporation cycle of surface water resources are directly affected⁵². During the extraction and processing phases within mining areas, fine particulate matter is generated, posing a certain impact on the atmospheric environment and human health²⁴. Hence, it is essential to incorporate atmospheric indicators into the assessment system for the ecological quality of mining areas.

The principal method of coupling multiple ecological indicators into a single variable involves assigning weights to each indicator component based on specific rules, followed by a weighted sum. However, due to the distinct ecological characteristics of different regions, each area presents its unique ecological issues, land use types, socio-economic features, and so forth. Therefore, significant variations exist in the ecological environment features across different areas. The typical human-determined weighting method often fails to accurately reflect the unique characteristics of regional ecological issues⁵³. Moreover, assigning weights to various indicators through subjective judgment introduces a degree of arbitrariness among the weights of the indicators. Consequently, utilizing fixed weights to evaluate different regions often renders the accuracy and specificity of ecological assessment results susceptible to human factors during the evaluation process⁵⁴. When multiple indicators collectively determine the ecological status, ascertaining the significance of a particular indicator to the overall environmental system, and assigning an appropriate value to it, can be challenging. To address this issue, the present study developed a RSEDI model tailored for the study area, employing a distance function that integrates greenness, wetness, dryness, heat, and black particulates. The calculation formulas for each component indicator are provided in Table 2.

Table 2 Formulas and descriptions of component indicators.

Within the framework of the constructed model, the spatial point exhibiting the poorest ecological quality is characterized by the maximum values of the heat, dryness, and black particulates, coupled with the minimum values of the greenness and wetness indicators. The evaluation of ecological quality across the space is achieved by calculating the distance from each point to this point of worst quality. A shorter distance, or a lower index value, indicates a decline in ecological quality, while a greater distance signifies an improvement in ecological quality⁵⁹. The calculation is formulated as follows.

$$~RSEDI=\sqrt {{{\left( {GI - G{I_{\hbox{min} }}} \right)}^2}+{{\left( {WI - W{I_{\hbox{min} }}} \right)}^2}+{{\left( {DI - D{I_{\hbox{max} }}} \right)}^2}+{{\left( {LI - L{I_{\hbox{max} }}} \right)}^2}+{{\left( {BI - B{I_{\hbox{max} }}} \right)}^2}}$$

(1)

Where GI, WI, DI, LI, and BI represent the greenness, wetness, dryness, heat, and black particulates of a specific pixel, respectively. The advantage of this index lies in its determination of the distance from each pixel to the point of poorest ecological quality based on the pixel’s own index values. This approach assesses the quality of the ecological environment at that point, circumventing inaccuracies that may arise from arbitrarily assigned weights to the various indicators. The indices are normalized within a 98% confidence interval to mitigate the effects of seasonal variation, resulting in values ranging from 0 to 1. Utilizing the method of natural breaks, these values are segmented into five equidistant categories representing ecological quality: poor (0–0.2), fair (0.2–0.4), moderate (0.4–0.6), good (0.6–0.8), and excellent (0.8–1).

Spatial auto-correlation analysis

Spatial autocorrelation serves as a crucial index for examining whether a significant correlation exists between a given feature or attribute of an area and its neighboring spaces. The interactions between the attribute characteristics of a spatial reference unit and those of its neighboring spatial units are revealed, a method commonly utilized in the examination of spatiotemporal pattern evolution⁵⁹.

The calculation of spatial autocorrelation encompasses both the global Moran’s index and the local Moran’s index, primarily analyzing the “global” and “local” spatial clustering aspects of the RSEDI⁶⁰. The Moran’s I coefficient, which ranges from − 1 to 1, reflects the degree of correlation in attribute values among spatially adjacent units. Values greater than 0 indicate a clustering pattern, equal to 0 suggests a random pattern, and less than 0 signifies a dispersion pattern. However, Moran’s I does not reveal the distribution of regional hotspots and coldspots. Consequently, the LISA are employed for spatial clustering and outlier analysis to assess the spatial heterogeneity of the RSEDI.

The Global Moran’s I index is calculated using the following formula:

$$I = \frac{{\mathop \sum \limits_{i = 1}^m \mathop \sum \limits_{i = 1}^m {w_{ij}}\left( {{B_i} - \overline B } \right)\left( {{B_j} - \overline B } \right)}}{{{S^2}\mathop \sum \limits_{i = 1}^m \mathop \sum \limits_{i = 1}^m {w_{ij}}}}$$

(2)

$${S^2}=\frac{1}{m}\sum {\left( {{B_i} - \overline {B} } \right)^2}$$

(3)

where I is the global Moran’s I index with a threshold of [−1, 1], I < 0 indicates negative correlation, I = 0 indicates no correlation, and I > 0 indicates positive correlation; w_ij is the weight coefficient, and w_ij takes the value of 1 only when i and j are neighboring, and takes the value of 0 otherwise; B_i and B_j are the ecological environmental quality indexes at i and j in the study area, respectively. S² is the sum of the squares of deviations of variable B.

The local Moran’ s I index is calculated as:

$${I_i}=\frac{{\left( {{B_i} - \overline {B} } \right)}}{{{S^2}}}\sum\nolimits_{{j=1}}^{m} {{w_{ij}}\left( {{x_j} - \bar {x}} \right)}$$

(4)

where I_i is the local Moran’s I index statistic; the remaining variables have the same meaning as the global autocorrelation variables. LISA clustering has five classes of local spatial aggregation maps, High-High (H-H), Low-Low (L-L), Low-High (LH), High-Low (H-L) and No Significant.

Center of gravity migration model analysis

The centroid migration model is frequently utilized across various domains such as economics, demography, and urban development. The ecological quality centroid is defined as a coordinate data point within the entire study area where the quality of the ecological environment is balanced in all directions⁶¹. Examining the ecological quality centroid of a study area aids in understanding the developmental direction and balance of the ecological environment. The centroid migration model not only represents the central location of distribution but also reflects the spatial evolution of a specific element. Furthermore, by examining the shifts in centroids, the temporal and spatial clustering and migration characteristics of the element can be analyzed⁶². The spatial variability of ecological quality, characterized by the RSEDI model, can be described through the centroid migration process, which delineates the overall trends and spatial characteristics of changes⁶³. The formula for calculating the centroids of various ecological quality levels in the study area for year t is as follows:

$$X = \frac{{\mathop \sum \limits_{i = 1}^n {D_{ti}}{X_i}}}{{\mathop \sum \limits_{i = 1}^n {D_{ti}}}}$$

(5)

$$Y = \frac{{\mathop \sum \limits_{i = 1}^n {D_{ti}}{Y_i}}}{{\mathop \sum \limits_{i = 1}^n {D_{ti}}}}$$

(6)

Where X and Y denote the longitude and latitude of the center of gravity, respectively; D_ti is the area of the ecological environment of grade i in year t; X_i and Y_i are the latitude and longitude of the geometric center of the ecological environment of grade i, respectively; and n is the total area of the ecological environment of this grade.

Results

Ecological quality monitoring on time scales

Utilizing the GEE, the mean values of the Greenness Index (GI), Wetness Index (WI), Dryness Index (DI), Heat Index (LI), Black particulates Index (BI), and the RSEDI were computed for the study area, as detailed in Table 3. The findings revealed that from 2000 to 2020, Dexing City’s RSEDI demonstrated a trend of an initial decrease followed by a rise, with its ecological status being categorized as moderate. Analysis of the component indices revealed that the GI generally increased, except for a decrease of 0.11 (14.23%) in 2008. The GI values rose by 0.159 (19.34%) and 0.003 (0.36%) in the subsequent periods, resulting in a cumulative growth of 6.30%. The decline in the GI value in 2008 was primarily attributed to excessive mining and accelerated urbanization. The subsequent rise in vegetation cover was closely associated with the execution of policies, including the conversion of farmland to forest and grassland in 2010, as well as the ecological restoration projects in mining areas. The WI demonstrated a close correlation with vegetation, generally mirroring the trend of the GI. From 2000 to 2020, the DI initially rose and subsequently declined, experiencing an overall decrease of 9.67%, which was related to a series of policies implemented by the local government. The LI and the BI were inversely correlated with vegetation, showing declines of 4.55% and 1.74%, respectively, over the same period.

Evidently, between 2000 and 2020, the study area exhibited an overall trend of increased greenness and wetness, while dryness, heat, and atmospheric indices decreased.

Table 3 Mean values of the component indicators and RSEDI from 2000 to 2020.

Spatial scale monitoring of ecological quality

Changes in ecological class distribution

To enhance the quantification and visualization analysis of the RSEDI, this study derived classifications of the RSEDI index based on its spatial distribution map (Fig. 3).

Statistical analysis revealed that in the year 2000, the majority of Dexing City was classified as having ‘Good’ ecological quality, covering an area of 822.33 km² (39.14%) of the city. This was followed by a ‘Moderate’ classification, which constituted an area of 541.43 km² (25.77%); In 2008, the RSEDI category was ‘Moderate,’ encompassing an area of 600.47 km². The areas classified as ‘Excellent’ and ‘Good’ measured 341.62 km² (16.26%) and 536.60 km² (25.54%), respectively, marking decreases of 4.45% and 13.60% from the year 2000. Conversely, the regions categorized as ‘Fair’ and ‘Poor’ expanded by 9.79% and 5.45%. By 2013, the most prevalent ecological condition remained ‘Moderate,’ with its area peaking at 804.68 km², the largest during the four periods studied. The proportion of areas rated as ‘Good’ rose from 25.54–26.12%, while those rated as ‘Fair’ and ‘Poor’ witnessed a decline, dropping from 18.77% and 10.85–18.59% and 7.31%, respectively. This suggests an improvement in the ecological environment quality between 2008 and 2013. In 2020, the ‘Good’ category became the leading RSEDI class, with an area of 755.52 km² (35.96%), and the ‘Poor’ category reduced to 101.69 km², accounting for just 4.84% of the total area. The proportion of areas rated as ‘Excellent’ and ‘Good’ saw significant increases from 2013, rising by 3.83% and 9.84%, respectively.

In summary, Over the past two decades, it is apparent that the study area initially suffered significant degradation. However, following a series of mine rehabilitation initiatives, there was a substantial improvement in the quality of the ecological environment.

Fig. 3

Spatio-temporal dynamics of ecological quality and percentage of area with different RSEDI status in Dexing City from 2000 to 2020.

Examining the spatial transition trajectories (Fig. 4), it is evident that there is frequent interconversion among the various ecological categories of the RSEDI. The changes in ecological quality primarily involved transitions between two levels. Degradation predominantly occurred through shifts from ‘Good’ to ‘Moderate,’ ‘Moderate’ to ‘Fair,’ and ‘Fair’ to ‘Poor.’ Conversely, improvements were mainly characterized by transitions from ‘Moderate’ to ‘Good,’ ‘Moderate’ to ‘Excellent,’ and ‘Fair’ to ‘Excellent.’ Between 2000 and 2008, the greatest transition intensity was observed from ‘Good’ to ‘Moderate’ areas, with a transfer rate of 11.40%; the smallest transition occurred from ‘Poor’ to ‘Excellent’ areas, with a rate of 0.46%, indicating a trend towards ecological degradation. From 2008 to 2013, the pattern of transition largely mirrored the previous period, with the highest rate of change still from ‘Good’ to ‘Moderate’ areas, reaching 11.40%, suggesting further ecological deterioration within the study area. Between 2013 and 2020, a significant improvement in ecological quality was noted, with the highest transfer rate observed from ‘Moderate’ to ‘Good’ areas, at 14.64%. Additionally, ‘Poor’ and ‘Fair’ areas began to show a certain percentage of transition into ‘Good’ and ‘Moderate’ categories, with rates of 3.58% and 6.54%.

Fig. 4

Transfer Relationships by Levels of RSEDI in Dexing from 2000 to 2020.

At the township scale, between 2000 and 2020, fluctuations in ecological conditions were observed across all townships (Fig. 5). Overall, stability was the predominant trend, characterized by an initial deterioration followed by subsequent improvement. Between 2000 and 2008, a trend of varying degrees of ecological degradation was observed across all townships. Notably, Zhangcun Township, Sizhou Town, and Xiangdun Sub-district experienced significant declines in their RSEDI, with decreases of 0.149, 0.126, and 0.096, respectively. The proportion of poor-quality areas in these regions increased by 13.72%, 25.4%, and 19.16%, respectively. In contrast, Xingangshan Town witnessed an improvement in ecological conditions, with a decrease in the proportion of poor-quality areas by 6.84%. This divergence may be attributed to the rapid development occurring in some townships during this period, particularly those in proximity to mining sites where concentrated populations and frequent mining activities led to increased areas of exposed land. From 2008 to 2013, the ecological environment of Zhangcun Township improved, with the area proportion of the lower grade regions decreasing from 15.2 to 1.27%. Meanwhile, the ecological issues in Wancun Township intensified, with its RSEDI dropping from 0.284 in 2008 to 0.234 in 2013. The ecological quality in Sizhou Town and Xiangdun Sub-district remained a concern, with degraded areas comprising over half of the total area, recorded at 59.59% and 50.55%. Subsequently, local authorities implemented a series of environmental management measures.

By 2020, a continuous improvement in the ecological environment was evident in most towns, especially in Xiangdun Sub-district, Huangbai Township, Wancun Township, and Sizhou Town in the southwest of Dexing City, as well as in Xingangshan Town and Kuanda Township in the northeast. Additionally, the concentrated areas of the Dexing Copper Mine were primarily situated in Sizhou Town and Yincheng Sub-district. In 2015, driven by policy initiatives, Dexing City focused on implementing measures for soil and water conservation, vegetation restoration, and ecological rehabilitation of mines. These efforts contributed to the improvement of the ecological quality in Sizhou Town, Yincheng Sub-district, and their surrounding areas.

Fig. 5

RSEDI ecological rating statistics by township from 2000 to 2020.

Changes in spatial center of gravity migration

Utilization of the centroid shift in Dexing City allows for a macro-level analysis of spatial variations in ecological quality within the study area, thereby facilitating the assessment and determination of spatial balance and trends in ecological quality. The centroid migration map illustrating the changes in the RSEDI grades from 2000 to 2020 (Fig. 6) was constructed based on the positions of various centroids.

The distribution of RSEDI centroids generally exhibited high consistency with the spatial pattern of RSEDI itself, with the migration routes of the centroids effectively reflecting the spatial distribution changes of RSEDI over time. The ecological centroids across the years consistently concentrated around Huaqiao Town. The results indicated that, from 2000 to 2020, the migration of the centroids representing ecological improvement first moved northeast, then shifted northwest, and ultimately changed towards the southwest. In contrast, the centroids denoting ecological degradation migrated in a direction entirely opposite to that of the improving ecological environment.

Examining the distances involved, the spatial shifts in areas experiencing ecological degradation, as measured by the RSEDI, were considerable. Over the past two decades, the total migration distance amounted to 16.10 km, with a significant relocation span of 7.67 km between the years 2000 and 2020. Notably, the years 2000, 2008, and 2013 saw substantial shifts, coinciding with periods of significant mining area development and change. This can primarily be attributed to the rapid urbanization and industrial mining growth from 2000 to 2008, accompanied by continuous land-use transformations, consequently causing centroid shifts. On the other hand, areas exhibiting ecological improvement under RSEDI saw more modest spatial changes, maintaining a generally stable trend, with an overall migration distance of 10.58 km over two decades, and a 2.02 km shift between 2000 and 2020. The changes in spatial location related to ecological improvement are closely linked to the ecological restoration and management of mines in the southwest region initiated since 2015.

Fig. 6

Shift in ecological center of gravity.

Monitoring of dynamic changes

In this study, based on the grading of the RSEDI, changes in the RSEDI over various years were detected through difference analysis, thereby examining the dynamic shifts in the ecological quality of Dexing City. By employing the reclassification tool in ArcGIS, the RSEDI values from 2000 to 2020 were categorized into three levels: improvement, constant, and deterioration. This classification facilitated the creation of an ecological grade change monitoring map (Fig. 7), and the area proportions of different grades were statistically analyzed (Table 4).

Between 2000 and 2008, factors such as excessive mining and accelerated urbanization led to the highest proportion of environmental degradation in the study area, covering 982.01 km² (46.74%), predominantly in the northwestern region, including Xiangtun Street, Huangbai Township, Zhangcun Township, Huaqiao Town, Raoer Town, and Longtoushan Township. The area where the environment remained unchanged accounted for 555.08 km². Areas of ecological improvement, chiefly located in the northeastern part of the study area, were significantly smaller than those experiencing degradation, totaling only 563.91 km² (26.84%), and were found in Xingangshan Town and Bandao Township.

During the period from 2008 to 2013, the study area exhibited significant ecological heterogeneity. Notably, the southern and northwestern regions experienced a marked improvement in their ecological environment, while the northeastern areas faced severe ecological issues, primarily concentrated in Xingangshan Town and Bandao Township. Despite the continued prevalence of degraded areas over those experiencing improvement, the disparity between the two began to narrow, with respective proportions of 36.64% and 32.68%.

Between 2013 and 2020, a significant enhancement in the ecological environment was observed within the study area. The degraded areas were primarily concentrated in the central region, covering 423.14 km² (20.14%). Conversely, the areas of improvement exceeded those of degradation, totaling 859.31 km² (40.90%). During this period, driven by policy initiatives, Dexing City intensified its efforts in ecological management and protection. Additionally, the region focused on activities such as soil and water conservation, vegetation restoration, and ecological rehabilitation of mines, which collectively contributed to the enhancement of the ecological quality.。.

An analysis of the three preceding periods reveals that from 2000 to 2020, Dexing City’s ecological environment underwent a transition from initial deterioration to subsequent improvement. Overall, the ecological condition of Dexing is gradually ameliorating. However, over the past two decades, the central and western mining areas continue to suffer from significant ecological damage. In recent years, despite the local government’s implementation of a series of ecological restoration projects, the environment has not yet returned to its state from the year 2000.

Fig. 7

Dexing remote sensing ecosystem change monitoring map.

Table 4 Classification of the magnitude of change of RSEDI in dexing City.

Spatial autocorrelation analysis of ecosystems

This study investigated the spatial relationships of RSEDI at a 1 km×1 km grid scale, revealing the existence of spatial autocorrelation in RSEDI. Utilizing GeoDa software, the analysis employed both global spatial autocorrelation and local spatial autocorrelation indices to examine the spatial autocorrelation of RSEDI variations (Fig. 8). The Moran’s I coefficient for the 2000 RSEDI is 0.885 and Z = 5.22, Z > 1.96, P = 0.002, P < 0.01; the Moran’s I coefficient for the 2008 RSEDI is 0.834 and Z = 5.31, Z > 1.96, P = 0.002, P < 0.01; Moran’s I coefficient for RSEDI 2013 is 0.821, and Z = 5.19, Z > 1.96, P = 0.002, P < 0.01; Moran’s I coefficient for RSEDI 2020 is 0.780, and Z = 4.68, Z > 1.96, P = 0.002, P < 0.01. In summary, the Global Moran’s I values for the four time-period images of Dexing City are consistently greater than 0, with P-values less than 0.01, indicating statistical significance. This suggests that the RSEI in the study area exhibits a strong spatial positive correlation and high spatial clustering. From 2000 to 2020, there is a declining trend in the Global Moran’s I, with the highest value observed in 2000, indicating that spatial clustering was most pronounced in that year.

Fig. 8

RSEDI Moran scatterplot for Dexing City from 2000 to 2020.

The spatial distribution pattern of ecological environmental quality in Dexing City was further delineated by analyzing the local spatial correlation of ecological environmental quality within the study area. The LISA significance map (Fig. 9) was created using Moran’s scatter plot, illustrating the distribution. In the study area, the predominant features were High-High (H-H) and Low-Low (L-L) clusters, manifesting positive spatial correlations. Variations in these clustering characteristics were closely tied to the development of urbanization in the region. In 2000, areas characterized by High-High (H-H) clustering predominantly occupied the western and northeastern parts of the study area. These regions, where forestry was the prevailing land use type and human disturbance indices remained low, exhibited relatively high ecological environmental quality. Conversely, regions exhibiting Low-Low (L-L) clustering were primarily located in the southwestern part of the study area and adjacent to roadways. These areas were characterized by extensive development and poorer ecological environmental quality. By 2008, the Low-Low (L-L) areas had concentrated in the eastern sector and the central region, which is rich in mineral resources. During this period, frequent mining activities and the expansion of urban and rural land use led to significant spatial clustering differences. In 2013, the central and southwestern regions exhibited characteristics of High-High (H-H), whereas the southeastern regions were identified as Low-Low (L-L). This indicates that an ecological restoration project was implemented between 2008 and 2013, leading to an improvement in the ecological environment of the Dexing Copper Mine located in the central region. The areas in the southeast, primarily designated for urban construction, represented the low-value clusters of the RSEDI. In 2020, spatial aggregation patterns resembled those observed in 2013, with the ecological quality of the Dexing Copper Mine area showing sustained improvement.

Additionally, the ecological quality among different units was found to influence each other, consistently exhibiting a trend of either High-High (H-H) or Low-Low (L-L). Conversely, the patterns of high-low and low-high clustering were not prominently observed. Consequently, in the subsequent development planning of the Dexing mining area and its socio-economic context, it is imperative to thoroughly consider the results of cluster analysis on ecological environmental quality. Areas classified as having “excellent” ecological environmental quality should be prioritized for protection, allowing them to exert a radiating influence that enhances the ecological environmental quality of surrounding areas. Conversely, regions categorized as having “poor” or “fair” ecological environmental quality require focused remediation efforts to improve their ecological conditions, thereby also positively influencing the environmental quality of adjacent areas. Over time, the proportion of high-high clustering areas has increased, rising from 10.36% in 2000 to 15.84% by 2020; this denotes a growing trend in High-High (H-H). In contrast, Low-Low (L-L) have exhibited a decreasing trend, with their proportion declining from 10.88% in 2000 to 7.17% in 2020. The variations observed in High-High (H-H) and Low-Low (L-L) clusters indicate that the degree of ecological quality clustering within the study region was not constant, attributable to the interplay of natural and anthropogenic factors.

Fig. 9

LISA clustering map of RSEDI in Dexing City from 2000 to 2020.

Discussion

Impact of mining activities on ecological quality

The proliferation of mining operations directly precipitates a decline in vegetation cover, which consequently impacts the quality of the regional ecological environment⁶⁴. This study extracted mining activity zones in Dexing City from 2000 to 2020 (Fig. 10), with the main mining areas concentrated in the central-western region, including the Yinshan and Dexing Copper Mines. Analysis of the statistical results revealed that over the past two decades, the overall trend of RSEDI in the Dexing mining area showed a decline followed by an increase. Specifically, between 2000 and 2013, the RSEDI decreased from 0.474 to 0.443, a reduction of 6.54%. From 2013 to 2020, the RSEDI rose from 0.443 to 0.498, an increase of 11.04%, which generally aligns with the ecological environmental evolution patterns in Dexing City. Notably, the RSEDI inversion results for 2008 showed an anomalous increase, primarily due to the commencement of reclamation projects in the mining areas by the government during this period, with extensive tree planting likely influencing the RSEDI inversion. It can be concluded that mining activities are the primary factor influencing the ecological quality of Dexing City.

From the perspective of area changes, between 2000 and 2013, the proportion of areas classified as ‘poor’ and ‘fair’ in the Dexing mining zone increased by 1.38% and 5.47%, respectively, while the proportion of areas rated as ‘excellent’ also saw an increase of 1.23%. This indicates that while the scale of mining operations in the Dexing mining area is progressively expanding, the local government has concurrently initiated measures to remediate the ecological damage caused by mineral extraction. From 2013 to 2020, the ecological quality of the Dexing mining area showed continuous improvement, with the proportions of areas classified as ‘excellent’ and ‘good’ increasing by 9.05% and 6.83%, respectively. However, the proportion of areas rated as ‘poor’ also rose by 2.32%. This indicates that despite the significant effects of restoration projects in mining areas, the ecological damage caused by mining activities remains severe and warrants attention. In summary, Dexing City has managed to balance mining impacts with regional environmental protection effectively, serving as a reference for sustainable development in similar mining cities^65,66.

Fig. 10

Changes in the ecological environment of the Dexing mining area from 2000 to 2020 and the statistics of each ecological class.

Social policy factors

Between 2000 and 2020, the ecological quality of Dexing City transitioned from a state of deterioration to one of improvement. This shift occurred as the concept of sustainable development gained traction across the nation, leading to the introduction of numerous environmental protection policies⁶⁷. Measures specifically targeting mining pollution were progressively implemented. Governments at various levels and social organizations began to actively embrace and act upon environmental conservation principles.

Due to prolonged mining development and overexploitation, mining activities in the region resulted in severe ecological and environmental issues. The local government prioritized these concerns and established a specialized task force for the rehabilitation and rectification of abandoned mines and historical mining pits. Investigations revealed a total of 358 historical mining pits and 48 abandoned mines in Dexing City. Consequently, documents such as the “Dexing City Plan for the Assessment of Rehabilitation Effects on Historical Mining Pits and Abandoned Mines” were formulated. These initiatives aimed at restoring and improving the environmental conditions of disused mining sites, as well as protecting and managing local resources. From 2019 to 2021, a total of 110 million yuan was invested in Dexing City for the reclamation and remediation of 47 abandoned mines, covering an area of approximately 5420 acres. This resulted in an increase of 168 hectares in mine ecological restoration, with the restoration ratio reaching 220%. By 2023, the ecological restoration of all 48 historic legacy mines had been completed. Additionally, the local government systematically planned and designed innovative models such as “Mine Ecological Restoration plus Rural Complex” and “Mine Ecological Restoration plus Mine Cultural Tourism”. These initiatives have created new highlights for industrial and mining tourism tailored to Dexing’s characteristics, effectively stimulating local economic growth, vigorously advancing rural revitalization, and actively exploring pathways for green transformation and development. On one hand, active collaboration was pursued with local enterprises to develop a new model for mine ecological restoration, which was characterized by government leadership, policy support, social participation, developmental governance, and market-oriented operations. This initiative attracted a total investment of 1.087 billion yuan and targeted a construction area of approximately 15,975 acres. On the other hand, the local government proactively engaged in green cultural and tourism activities, creating unique projects that showcased the region’s copper mining heritage. These measures not only resolved the ecological and environmental issues associated with the historically abandoned mines in Dexing City but also significantly promoted the socio-economic development of the region.

Ecological quality improvement suggestions in dexing

As a region abundant in mineral resources, Dexing City experiences significant ecological impact due to mining activities, even as these activities contribute to local economic growth⁶⁸. Therefore, the necessity of implementing ecological management in mining areas is multi-faceted. Based on the findings from prior research on ecological quality and the analysis of social factors, the following recommendations are proposed to enhance the ecological environment of Dexing City in the future:

(1) Conceptual Innovation Leading to Green Mine Reclamation.

Guided by the development philosophy of “harmonious coexistence between humans and nature”, Dexing Copper Mine shifts its focus from merely re-greening land to restoring entire ecosystems. By optimizing ecological restoration plans and adjusting mining and waste management schedules, it concentrates on rehabilitating the permanent fixed slopes created by mining activities. This approach aims to achieve comprehensive ecological restoration, transforming connected patches into continuous greenery wherever possible.

(2) Technological innovation to promote the construction of green mines.

Through the reliance on technological innovation and the implementation of standardized management practices, a dual approach of advancing both technological and managerial innovations is adopted. A range of new technologies and processes is promoted and applied, leading to a multi-faceted win that encompasses economic, ecological, and social benefits. The application of energy-efficient mining technologies is vigorously promoted to enhance energy conservation and emission reduction, thereby achieving cleaner production. Environmental protection measures are strengthened to foster sustainable development, and mining areas are landscaped in accordance with garden-style mine standards. Existing geological environmental issues and potential hazards are actively managed and mitigated.

(3) Changing the land use pattern and improving the land utilization rate of the mine treatment area.

In recent years, Dexing City has experienced rapid growth in construction land area, accompanied by the presence of abandoned factories and idle mines. This has led to deforestation and mining activities, resulting in inefficient and unsustainable land use practices. There is an urgent need to enhance land use efficiency. Future land use planning in Dexing City should focus on reduction strategies, revitalizing idle construction land, and minimizing new land allocation. Abandoned mines should be ecologically restored to a green state through restoration technologies, facilitating cultural and tourism projects in the reclaimed areas. This approach aims to simultaneously boost both ecological and economic benefits.

Advantages and limitations

Since the RSEI model was proposed, it has been widely used in the fields of regional ecological evaluation and ecological change monitoring due to its advantages of strong comprehensiveness and objectivity⁶⁹. The model employs principal component analysis to reduce dimensions and extract information from various ecological indicators, which inevitably results in a loss of some data information⁷⁰. Moreover, the significance of the composite evaluation function may become unclear when the factor loadings of the principal components are both positive and negative, leading to reduced clarity in naming. In certain instances, the computational process using PCA can become quite complex, especially with large datasets or a high number of variables⁷¹. Consequently, based on the principles underlying the construction of the RSEI and taking into account the characteristics of air pollution in mining areas, this paper employed the distance function method to construct the RSEDI model. The Distance Function Method, commonly utilized in fields such as mathematics, physics, and engineering, addresses problems by calculating distances between points, points and surfaces, and surfaces themselves⁷². This approach preserves the integrity of data while simplifying computational processes and enhancing efficiency. It is also straightforward to program and apply. Additionally, the method can be readily extended to higher dimensions or more complex spatial structures⁷³. Overall, it is well-suited for solving problems involving spatial relationships. Through the analysis of remote sensing imagery, the ecological quality evolution in Dexing City from 2000 to 2020 is found to align with the actual distribution of its ecological conditions. Furthermore, employing models such as the centroid migration model and spatial autocorrelation, this research conducts an in-depth examination of ecological quality changes at both the municipal and township scales.

Although the RSEDI was developed effectively to capture the features of cities dependent on mineral resources, it is not without its uncertainties. This model, which employs remote sensing inversion to gather information on various indicators, is based on the spectral characteristics of surface coverings and is unable to penetrate vegetation. This limitation leads to a disregard for the actual state of the ground surface. The presence of “same matter, different spectrum” and “same spectrum, different matter” phenomena in multispectral inversion introduces certain inaccuracies. Furthermore, the ecological characteristics unique to each study area necessitate differing focal points in ecological environment assessments. Considering the unique environmental characteristics of various regions, future modifications to the component indices of the RSEDI model may be contemplated. The aim is to enable a more scientific and precise assessment of the ecological environmental quality across different areas.

Conclusions

In recognition of the atmospheric pollution attributes unique to mining areas, this paper introduced a novel remote sensing index: the RSEDI, utilizing a distance function method. Employing the GEE platform, the spatiotemporal evolution of the ecological quality in Dexing City over the past two decades was monitored and studied. The findings indicated:

• (1) From a temporal perspective, the overall trend of the RSEDI in Dexing City from 2000 to 2020 exhibits an initial decline followed by a rise. This trend is characterized by two distinct phases: a period of urbanization and active mineral development from 2000 to 2008, and a period of stable urban and mining activities from 2008 to 2020.

• (2) On a spatial scale, it is observed that the centroid of ecological environmental enhancement within Dexing City, over the period from 2000 to 2020, initially advances northeastward, subsequently veers northwestward, and currently trends southwestward. In contrast, the centroid migration of ecological environmental degradation follows a pattern that is diametrically opposed to that of improvement. Additionally, the migration of ecological environmental centroids is closely linked to the local mineral resource development policies.

• (3) The dynamic monitoring and analysis of the RSEDI across different periods reveal that Dexing City’s ecological quality initially deteriorated before showing signs of improvement. On the whole, Dexing’s ecological environment is progressively ameliorating. However, over the past two decades, significant ecological degradation persists in the central and western mining areas. The quality of the ecosystem has not yet returned to the levels observed in the year 2000.

• (4) An analysis using the Global Moran’s I reveals that the spatial distribution characteristics of ecological environment quality exhibit a positive correlation. This distribution is indeed influenced by open-pit mining activities. Local Moran’s I analysis reveals a distinct clustering pattern in ecological environmental quality. Following ecological restoration projects, the High-High (H-H) in 2020 were predominantly located in the central and southwestern mining areas, while the Low-Low (L-L) were mainly concentrated in the southeastern urban construction zones.

Due to the limitations of the current study, future research will necessitate the targeted design and refinement of an ecological assessment indicator system for cities rich in mineral resources. A deeper investigation into the quantitative relationship between mining activities and regional ecological quality is also required. Furthermore, the selection of ecological environmental assessment indicators appropriate for the study area must be conducted in a comprehensive and objective manner, with efforts to minimize errors during data processing. This will enhance the precision and objectivity of ecological environmental quality assessments in ongoing research endeavors.