Evaluation of ecological geological environment carrying capacity and analysis of driving mechanisms based on normal cloud model and geodetector model

Abstract

The Luoyang area of the Yellow River Basin, as a typical resource-based city, its special industrial structure and complex geological structure make the ecological and geological environment of the area extremely fragile. In order to realize the sustainable development of the region in this fragile ecological-geological environment, it is necessary to study its Ecological Geological Environmental Carrying Capacity (EGECC) to better serve the regional ecological-geological environment restoration and management work. This study constructs an indicator system encompassing three subsystems: Geological Environment (GE), Social Environment (SE), and Ecological Environment (EE). Based on game theory combination weighting (G1-IEW-GT) and the Normal Cloud model, the EGECC of Luoyang City from 2000 to 2022 was evaluated on a grid scale. Concurrently, a coupled model of geographic detectors and geographically weighted regression was established to recognize the spatial heterogeneity and driving mechanisms of the ecological geological environment system. The findings demonstrate that: (1) The overall EGECC in the Luoyang region of the YRB fluctuates upward, with the eastern area displaying more favorable conditions compared to the western region. (2) Over these twenty-two years, the geological environment system experienced initial deterioration followed by improvement, the ecological environment system remained relatively stable, and social and economic development continuously strengthened. (3) Landform types, elevation, human activity impact index, population density, and geological disaster susceptibility are the main driving elements for the assessment of the ecological geological environment system. Any pair of the 18 variables show varying degrees of enhancement effects on the EGECC. (4) The regression coefficients of the five major driving factors exhibit significant agglomeration characteristics in spatial distribution. Among them, geological disaster susceptibility has a negative driving effect on the EGECC, while landform, human activity impact index, and population density all display bidirectional effects. Overall, this study provides essential guidance for formulating ecological protection and restoration plans for the national territorial space in the Luoyang area of the YRB. Additionally, the evaluation methods established in this study have promotional value and can serve as a reference for the ecological geological spatial management of other resource-based cities.

Introduction

Since the twentieth century, China’s economy has achieved leapfrog growth, urbanization has advanced rapidly, and the population has increased rapidly^1,2. Behind the high-speed development of society is the dependence on large-scale development and utilization of natural resources, which has led to the emergence of a variety of ecological and geological environmental problems, including geological disasters³, resource shortage⁴, water resource pollution⁵, and land resource crises⁶. The Yellow River Basin (YRB) are vital ecological security barriers for the country, playing a crucial role in the national ecological security strategy^7,8. However, due to the ecological sensitivity of the YRB and the over-exploitation of natural resources, eco-geological environment issues in some regions of the basin are particularly severe⁹.

As an important node city in the middle and lower reaches of the Yellow River Basin, the ecological and geological environment of Luoyang City is profoundly affected by the geological background, ecological conditions and economic activities in the Yellow River Basin⁷. The geological structure of the Loess Plateau in the Yellow River Basin makes the Luoyang area susceptible to soil erosion, especially during periods of concentrated precipitation, and the looseness of the loess layer exacerbates soil erosion and sediment loss¹⁰. In addition, the hydrological conditions in the basin determine the spatial and temporal distribution and sustainable utilization of groundwater resources in Luoyang, and the frequent alternation of floods and droughts exacerbates the ecological vulnerability of the area^3,11. As a typical resource-based city, Luoyang has long relied on mineral resource extraction, which has made its ecological and geological environment face severe challenges. This complex geological structure and special industrial structure in the Luoyang area make its ecological and geological environment very fragile, and geological disasters such as soil pollution, groundwater pollution, soil erosion and landslides occur frequently¹². In order to realize the sustainable development of the Luoyang area in the Yellow River Basin in the fragile ecological-geological environment, it is necessary to carry out the study of ecological-geological environmental carrying capacity (EGECC). This will not only help to formulate an ecological protection and restoration plan for Luoyang city’s national land space but also provide theoretical support for the ecological space management of other resource-based cities.

Ecological Geological Environment Carrying Capacity (EGECC) is an extension of the concept of environmental carrying capacity¹³. It refers to the limit of human activities that the eco-geological environment can endure while preserving the stability of its composition, structure, and function^14,15. Research on the eco-geological environment primarily focuses on the interactions between geological and ecological environments, emphasizing the fundamental role of geological factors and the impacts of ecological factors and human activities on the environment^16,17. In recent years, the 3S technologies—Remote Sensing, Geographic Information Systems, and Global Positioning Systems—have been extensively utilized in carrying capacity studies. For instance, Mario Mejia-Navarro at Colorado State University applied GIS technology to a decision-making system for geological disaster and risk assessment, achieving significant results. Zhao et al. (2022) used multi-source data and RS images to evaluate ecological geological environment issues in a typical red soil hilly region mining city from multiple perspectives, including physical geography, mining activities, and geological disasters, and analyzed their influencing factors¹⁵. He et al.¹⁸ took Longkou City, a coastal mining city, as the research area and used the AHP and GIS technology to conduct a synthesized assessment of the eco-geological environment of Longkou City. Zhu et al.⁹ used RS and GIS technologies to evaluate the ecological carrying capacity of the Yellow River Basin in Shaanxi. These studies provide significant theoretical support and practical guidance for evaluating EGECC in the Luoyang of the YRB.

The EGECC evaluation aims to understand better ecological geological environment systems’ changing trends and evolutionary mechanisms. The primary research methods for EGECC include the Analytic Hierarchy Process (AHP)¹⁹, the TOPSIS²⁰, and the Fuzzy Comprehensive Evaluation²¹. However, these methods have certain limitations in addressing the fuzziness and discreteness within systems. In contrast, the cloud model has unique advantages in handling uncertainties within systems. The cloud model not only facilitates bidirectional transformation between quantitative descriptions and qualitative concepts but also analyzes the relationship between discreteness and fuzziness within systems. Peng et al.²² introduced the cloud model into the DPESBRM conceptual model to quantitatively evaluate the water resources environment carrying capacity in Guiyang, demonstrating the model’s applicability and rationality. Xi et al.²³ selected 17 indicators, including strata lithology, fault distance, and geological hazards, using the cloud model-improved AHP to determine the indicator weights and evaluate the EGECC of the mining city of Huangshi. Zhang et al.²⁴ improved the AHP using the cloud model to examine the ecological environment vulnerability of the water source region for the middle route of the South-to-North Water Transfer Project, analyzing its influencing factors. At the same time, analyzing the driving mechanisms behind the evolution of ecological geological environments is also a crucial aspect of ecological geological environment research. It aims to identify the factors that dominate changes in system patterns, thereby serving the restoration and management of regional ecological geological environments. The geographical Detector Model is a new statistical method for detecting spatial dissimilarity and revealing underlying driving forces. It not only facilitates the simultaneous detection of quantitative and qualitative data but also identifies the explanatory power of the interaction between two factors. This method has been applied in the field of carrying capacity evaluation^25,26,27,28. In this research, we integrate the Normal Cloud model with 3S technologies to evaluate the EGECC at the grid scale in the study area. Additionally, leveraging the Geographical Detector Model, we introduce the Geographically Weighted Regression (GWR) Model to elucidate the driving mechanisms of spatiotemporal evolution and spatial heterogeneity of the eco-geological environment in the research region.

This study takes the assessment of the Ecological Geological Environment Carrying Capacity (EGECC) of the Luoyang area in the Yellow River Basin as the starting point, aims to promote the city’s sustainable development, and constructs a complete research system. This study’s scholarly achievements are mostly evident in the following areas: (1) Given the complex geological structure of the Luoyang region and the ongoing impact of long-term mining on the eco-geological environment, the study considers it an ideal area to study human activities’ disturbance of the eco-geological environment. Based on the region’s development characteristics, an EGECC evaluation index system was constructed, encompassing geological, ecological, and social environments. (2) The study integrates subjective and objective weighting methods through game theory, forming a comprehensive weighting model (G1-IEW-GT). On this basis, utilizing the cloud model combined with 3S technologies, this approach takes advantage of the cloud model’s ability to integrate fuzziness and discreteness within the ecological geological environment system, enabling a grid-scale evaluation of the EGECC in the study area from 2000 to 2022. (3) Using the Geographical Detector Model, identify the single-factor driving mechanisms and interactions among multiple factors within the ecological geological environment system. Furthermore, the Geographically Weighted Regression Model is introduced to more precisely explore the spatial variation characteristics of the regression coefficients of the independent variables, deeply revealing the spatial differentiation characteristics of the ecological geological environment system in specific contexts. (4) The EGECC evaluation and classification conclusions offer significant guidance for formulating Luoyang’s ecological protection and restoration plans. They offer theoretical backing for advancing ecological civilization and sustainable development within the YRB. Additionally, the comprehensive evaluation methods established in this study have general applicability, serving as a reference for the ecological space management of other resource-based cities.

Study area and data sources

Study area

Luoyang City (112°16′ E–112°37′ E, 34°32′ N–34°45′ N) is situated in the western region of Henan Province (Fig. 1). It serves as a crucial node city in the middle and lower reaches of the Yellow River Basin and plays a vital role in supporting the economic growth of the Central Plains urban agglomeration. As a typical resource-based city, Luoyang is a significant base for energy, chemical industries, and mining in Henan Province and across China. Luoyang City has a warm temperate continental monsoon climate with an average annual precipitation of 578 mm. Luoyang has a well-developed water system spanning the Yellow River, Yangtze River, and Huai River basins. The multi-year average shallow groundwater reserves are approximately 1.761 billion cubic meters. The city has a resident population of 7.079 million, a GDP per capita of 80,226 Yuan at the end of 2022, and the three industrial structures are 4.5:43.6:51.9.

Fig. 1

Location of the study region.

The overall terrain of Luoyang City slopes from west to east (Fig. 1). Mountainous regions account for 45.51% of the total region, primarily in the western and southern parts. Hilly regions comprise 40.73%, mainly surrounding the mountainous areas, characterized by relatively gentle terrain. Plains cover 13.8% of the area, with the eastern part being a typical alluvial plain, serving as a crucial agricultural production zone. Luoyang lies at the southern edge of the North China Platform, spanning two primary tectonic units: the North China Platform and the Qinling Geosyncline. This region features well-developed faults and complex geological structures, creating favorable conditions for endogenous mineral formation. The Luoyang area belongs to the gold and silver polymetallic metallogenic belt at the southern margin of the North China landmass, with complete mineral types and abundant reserves, and is a critical ore concentration area of Henan Province and even North China Block. However, despite driving rapid economic growth, frequent mining activities have significantly harmed the local eco-geological environment, leading to frequent incidents of soil pollution, groundwater contamination, soil erosion, and landslides^7,12 .

Data sources

Remote sensing data and primary information data are used in this paper. Remote sensing images are Landsat8 OLI and Landsat8 SR data with spatial resolution of 30 m and temporal resolution of 16d provided by the United States Geological Survey. The Normalized Difference Building Index (NDBI) is taken from the Landsat8 OLI. The B4 and B5 bands of Landsat 8 SR are utilized to calculate the NDVI using the GEE cloud computing platform. The B10 band of Landsat8 SR is used to obtain the LST data. All remote sensing data are radiometric calibration, atmospheric correction, and cloud removal. Other indicator data sources are shown in Table 1.

Table 1 Data sources and processing methods for evaluation indicators.

Models and methods

Construct the evaluation indicators system

The study of ecological geoenvironmental carrying capacity focuses on a complex dynamical system composed of three major elements: the geological environment, the ecological environment and socio-economics^1,8,29. These three subsystems have complex interrelationships and action mechanisms, constituting an essential part of the Earth system and jointly determining the face of the Earth, the development of human society, and the state of the natural environment¹. Therefore, the coordination among these three subsystems is crucial for enhancing the quality of the eco-geological environment.

Geological environment subsystem

The geological environment refers to a relatively balanced open system formed by interaction with the atmosphere and hydrosphere, with the lithosphere as the main body, and it is the habitat and activity space on which human beings rely for their survival and development³⁰. In this study, indicators were selected from both terrain landforms and geological structures to reflect the geo-environmental characteristics of the study area²³. The indicators of elevation (X1), slope (X2), and geomorphic types (X7) are used to reflect the terrain landform characteristics of the study region³¹; the fracture zone density (X3), geologic hazard susceptibility (X4) and stratigraphy lithology (X5) indicators are used to reflect the regional geological structural conditions³². In view of the development of the mining industry in the Luoyang area, the density of the mining area density (X6) is selected to reflect the degree of disturbance of the geo-environment by human activities¹⁵.

Ecological environment subsystem

The ecological environment encompasses various natural factors that influence human survival and development, including water, land, biological, and climate resources³³. Therefore, indicators such as river network density (X12) and rainfall (X8) are utilized to delineate the distribution of regional water resources and climatic conditions³⁴. Considering the impact of soil erosion on vegetation distribution and land use patterns, indicators such as soil erosion degree (X9) and land surface temperature (X13) are selected to gauge the degree of land resource degradation and land use intensity^18,32. Normalized difference vegetation index (X10) and biodiversity index (X11) serve to depict the distribution of flora and fauna in the region^35,36.

Social environment subsystem

The social environment is a product of human activities. In this study, Population density (X16) and GDP (X17) are selected to reflect the level of regional socio-economic development directly^1,27. The normalized difference built-up index (X14) was used to reflect the distribution of urban building land³⁶. The human impact index (X15) is a quantitative index that comprehensively considers factors such as land use, nighttime lights, and transportation routes, reflecting the intensity of human interference with the natural environment³⁷. Additionally, the tourism industry plays a driving role in regional economic growth, so tourist attraction density (X18) is selected to reflect the distribution status of regional tourism resources¹⁵.

In summary, based on the response mechanisms of the three subsystems and the development characteristics of the Luoyang area, 18 indicators are selected from the three aspects of geological environment, ecological environment and social environment to construct the EGECC evaluation index system of Luoyang City (Fig. 2).

Fig. 2

Comprehensive evaluation index system of EGECC.

Determine the weight of evaluation index

The assignment of evaluation indicators can be divided into two categories: subjective and objective. The disadvantage of the subjective assignment method is that it is easy to introduce the evaluator’s personal preference and empirical error, which may lead to the loss of objective interpretation of the information contained in the sample data in the assignment results²⁷. While the objective assignment method may not fully reflect the invisible features in the complex system of the study area¹⁵, for example, in the Luoyang area, the complex geological and tectonic conditions and diversified socio-economic factors are difficult to be fully quantified and expressed directly by objective data. Therefore, this study utilizes the empowerment idea of group decision-making, increases the consideration of objective sample data through subjective weight values, and uses objective sample data to control personal subjectivity within a certain range, thus realizing the authenticity, scientificity and credibility of the empowerment results¹.

Determination of subjective weights according to the G1 method

The G1 method, introduced by Guo Yajun³⁸, is optimized and derived from the hierarchical analysis method (AHP) and better inherits the idea of the relative importance of hierarchical analysis. The G1 method neither requires the construction of a judgment matrix nor the performance of a consistency test. It has the advantages of easy calculation and high practicality.

Step 1: Determine ordinal relationship for evaluation indicators

Invite experts to judge the importance of the evaluation index set $\left\{ {D_{1} ,D_{2} , \cdots ,D_{n} } \right\}$ select the most important index as $D_{1}^{*}$, by analogy, an order relation is established, denoted as $\left\{ {D_{1}^{*} > D_{2}^{*} > \ldots > D_{n}^{*} } \right\}$, where $n$ denotes the number of assessment indicators.

Step 2: Determine the relative importance of indicators and calculate of weighting factors

Judgments and scores are assigned based on the relative importance between adjacent indexes. And set the relative importance ratio $f_{m}$ between the evaluation indicators $D_{m - 1}^{*}$ and $D_{m}^{*}$ as:

$$f_{m} = w_{m - 1}^{*} /w_{m}^{*}$$

(1)

$$w_{n}^{*} = \left[ {1 + \mathop \sum \limits_{m = 2}^{n} \left( {\mathop \prod \limits_{i = m}^{n} f_{i} } \right)} \right]^{ - 1}$$

(2)

where $w_{n}^{*}$ is the weight value of the last factor.

Determination of objective weights based on the Improved Entropy Weights (IEW) method

As an objective weighting method, entropy weighting emphasizes the information content between indicators than the subjective weighting method³⁹. The entropy weight approach is vulnerable to the impact of extreme data since it establishes the weight values of indicators depending on the information content in the data. This study improves the traditional entropy approach method using the Hadamard product to avoid errors in weight results due to minor variations in the original data matrix.

Step 3: Construct of a standardized evaluation matrix

Due to the different dimensions of the indicators, it is impossible to evaluate the actual status of EGECC directly. Hence, this study utilizes the Min–Max normalization method to eliminate dimensional discrepancies among indicators and formulates the standardized evaluation matrix Z. This process has been explained in detail in a large number of related studies and is therefore not repeated in this paper^1,15.

Step 4: Calculate of the weights of the indicators

$$H_{i} = - \left( {\ln m} \right)^{ - 1} \mathop \sum \limits_{j = 1}^{m} \left[ {P_{ij} \ln \left( {P_{ij} } \right)} \right]$$

(3)

$$w_{i} = \left( {1 - H_{i} + \varepsilon } \right)/\mathop \sum \limits_{j = 1}^{n} \left( {1 - H_{i} + \varepsilon } \right)$$

(4)

where $H_{i}$ is the information entropy of the indicator $i$; if $P_{ij} = 0$, then define $P_{ij} \ln \left( {P_{ij} } \right) = 0$; $w_{i}$ is the weight value of the indicator $i$, and satisfy $\mathop \sum \limits_{i = 1}^{n} w_{i} = 1$; $\varepsilon$ is an adjustment item, according to the existing literature, it is set to 0.01^40,41.

Determine the combined weights based on the Game-Theoretic (GT) combinatorial weighting models

Game theory focuses on how to maximize benefits through optimal decision-making in situations where multiple factors are in an adversarial or game relationship^42,43. As a theoretically mature methodology, game theory has been widely used in various evaluation methods. This study draws on the ideas of game theory to provide a new solution for the combination assignment of indicators in the evaluation index system of EGECC. Specifically, this study adopts the game theory method to combine and weight the subjective assignment method (G1 method) and objective assignment method (improved entropy weight method). The method aims at Nash equilibrium and seeks agreement or compromise between the subjective and objective empowerment methods, which is a mutually coordinated integration process, i.e., minimizing the deviation between the combined weights and the individual basic weights. The method can scientifically give the optimal combination of subjective and objective weights, which maximizes the benefits of the two assignment methods and greatly improves the scientificity of indicator assignments.

Step 5: Construct the basic weight vector

Suppose there are $L$ kinds of index weighting methods, then the weight vector set of $n$ indicators is $W_{k} = \left( {w_{k1} ,w_{k2} , \cdots ,w_{kn} } \right),{ }k = 1,2, \cdots ,L$.

$$W = \mathop \sum \limits_{k = 1}^{L} \mu_{k} W_{k}^{T}$$

(5)

where $W$ is the combination weight vector set; $\mu_{k}$ is the weight coefficient of the $k$ th weighting method.

Step 6: Solve for optimal combination coefficients

Find a set of optimal combination coefficients to reduce the difference between the combination weight $w$ and each basic weight vector $w_{k}.$

$$min\mathop \sum \limits_{i = 1}^{L} \mu_{k} W_{k}^{T} - W_{k2}$$

(6)

Step 7: Determine the combined weight value

$$W^{*} = \mathop \sum \limits_{i = 1}^{L} \left( {W_{k}^{T} \cdot \mu_{k} /\mathop \sum \limits_{i = 1}^{L} \mu_{k} } \right)$$

(7)

where $W^{*}$ is the weight vector set determined by the combination weighting of game theory.

Evaluation of EGECC based on normal cloud modeling

Cloud modeling is a measurement model that facilitates bidirectional conversion between qualitative data and quantitative data through fuzzy mathematics and stochastic mathematics. Utilizing cloud modeling for evaluation enables the transformation of uncertain linguistic expressions in the evaluation indicator system into quantitative analysis while also reflecting the discreteness and fuzziness of the evaluated objects^24,44. The eco-geological environment system is structurally complex, and the evaluation indicator system exhibits characteristics of fuzziness and discreteness. Moreover, during evaluations, it is necessary to explore the relationship between the internal discreteness and fuzziness of the system³². Therefore, this study leverages the unique advantages of the Normal Cloud model to manage the relationship between quantitative descriptions and qualitative concepts. This approach fully utilizes the cloud model’s advantages of integrating ambiguities and discretization within the ecological-geological-environmental system and conducts a comprehensive evaluation of EGECC.

Step 1: Build the cloud model

Cloud modeling is characterized by three key values: expectation $E_{x}$, entropy $E_{n}$, and hyper-entropy $H_{e}$⁴⁵. Expectation $E_{x}$ represents the central value of the qualitative concept domain interval, serving as the centroid of the cloud droplet group; entropy $E_{n}$ measures the uncertainty of the qualitative concept, indicating the degree of dispersion and value range of the cloud droplets; hyper-entropy $H_{e}$ measures the uncertainty of entropy $E_{n}$, reflecting the cohesion degree of the cloud droplets.

$$E_{{x_{ij} }} = \left( {x_{ij}^{1} + x_{ij}^{2} } \right)/2$$

(8)

$$E_{{n_{ij} }} = \left( {x_{ij}^{2} - x_{ij}^{1} } \right)/2.355$$

(9)

$$H_{eij} = CE_{{n_{ij} }}$$

(10)

where $\left[ {x_{ij}^{1} ,x_{ij}^{2} } \right]$ is the value range of the evaluation factor $i$ with cloud characteristics in the corresponding evaluation grade $j$; the qualitative concept expression is more clearly expressed when the value is less. Therefore, this paper fixes the $C$ value at 0.01 based on the current references²².

Step 2: Draw the cloud map

By employing the cloud generator algorithm, the characteristic values of each corresponding level of the cloud model are transformed into cloud droplets, which are then depicted on a coordinate system, composing a cloud chart. Taking the cloud chart of qualitative indicators alongside the quantitative indicator of fracture zone density (X3) for 2022 as an exemplar (Fig. 3), the presentation of cloud charts for other years and indicators is omitted for brevity.

Fig. 3

The cloud figure: (a) fracture zone density (X3), (b) qualitative index.

Step 3: Calculate the comprehensive evaluation results

$$\mu_{ij} = exp\left[ { - \left( {x_{ij} - E_{{x_{ij} }} } \right)^{2} /2\left( {E_{{n_{ij} }} } \right)^{2} } \right]$$

(11)

$$K_{ij} = \mathop \sum \limits_{i = 1}^{n} w_{i}^{*} \mu_{ij}$$

(12)

where $\mu_{ij}$ is the membership degree of the $i$th assessment indicator at the $j$th evaluation level; $K_{ij}$ is the comprehensive assessment result of the $i$th index at the $j$th evaluation level.

Identification of system evolution driving mechanism based on geographic detector model

The interplay of numerous natural and societal elements leads to the evolution of the eco-geological environment system. Exploring the main driving factors of this system’s evolution in the study region and quantitatively identifying the magnitude of these driving forces can reveal the driving mechanisms behind the evolution of the eco-geological environment system, which will help to serve the regional ecological and geological environmental restoration and management.

The Geographic Detector is a statistical analysis method utilized to unveil the inherent driving forces of the research object. Unlike previous statistical models, its advantage lies in its ability to detect quantitative and qualitative data simultaneously. Moreover, it can detect the driving effect of individual factors and explore the explanatory power of interactions between two factors³³. This study used the Geographic Detector’s Factor Detector and Interaction Detector capabilities to identify the types of interactions between various factors and the degree to which they affect EGECC.

Factor detectors

$$q = 1 - \frac{SSW}{{SST}} = 1 - \frac{{\mathop \sum \nolimits_{h = 1}^{L} N_{h} \sigma_{h}^{2} }}{{N\sigma^{2} }}$$

(13)

where $q$ is the explanatory power of each influence factor on the evaluation results, and $q \in \left[ {0,1} \right]$; $SSW$ and $SST$ are respectively the sum of intra-layer variance and the total variance of the overall area; $h$ means the stratification of variable $Y$ or factor $X$.

Interactive detector

The interaction detector is utilized to recognize the interaction between various factors, that is, whether the explanatory power of the dependent variable will be enhanced or diminished due to the interaction of the two factors. To determine the kind of interaction between the two variables, the core principle is to calculate the explanatory power $P$ of the $i$th and $j$th driving factors about the dependent variable $Y$ and then compare it with the explanatory power of the single driving component (Table 2).

Table 2 Type of interaction relationship.

Revealing system spatial heterogeneity employing a geographically weighted regression model

The Geographically Weighted Regression (GWR) is a local linear regression model that utilizes multidimensional variables to analyze spatial heterogeneity²⁷. Building on identifying influencing factors by geographical detectors, GWR allows for a more precise exploration of the spatial variation characteristics in the regression coefficients of independent variables. As an extension of the ordinary linear regression model, GWR differs from the traditional least squares regression model in that the independent variables’ regression coefficients vary with the samples’ spatial location, offering robust local analysis capabilities for spatial data²⁸. Thus, this study employs the GWR model to reveal the profound connotations of the spatial differentiation characteristics of the eco-geological environment system under specific contexts.

$$y_{i} = \beta_{0} \left( {u_{i} ,v_{i} } \right) + \mathop \sum \limits_{k = 1}^{m} \beta_{k} \left( {u_{i} ,v_{i} } \right) \cdot x_{ik} + \varepsilon_{i}$$

(14)

where $\left( {u_{i} ,v_{i} } \right)$ and $y_{i}$ are the coordinates and prediction variables of the sample point $i$, respectively; $\beta_{0} \left( {u_{i} ,v_{i} } \right)$ represents the intercept of the $ith$ sample point; $\beta_{k} \left( {u_{i} ,v_{i} } \right)$ represents $k$th regression parameter on the sample point $i$; $x_{ik}$ is the $k$th independent variable of sample point $i$; $\varepsilon_{i}$ represents random error of sample point $i$.

Results and discussion

Determination of EGECC evaluation index weights

Drawing from the established tri-dimensional evaluation index system encompassing geological, ecological, and social facets, the spatial distribution map of attribute values for the 18 evaluation indicators in 2022 is depicted (Fig. 4). The modus operandi for subsequent years’ visualization remains akin, thus preempting a verbose repetition within this narrative.

Fig. 4

The spatial distribution of attribute values of each evaluation index of eco-geological environmental system in 2022.

Employing the composite weighting model (G1-IEW-GT) computation methodology, the amalgamated weight values for each evaluation indicator are ascertained (Fig. 5). From the viewpoint of the three subsystems, the geological environment subsystem commands the lion’s share of weight proportion at 0.433, closely trailed by the ecological environment subsystem at 0.359, while the social environment subsystem carries the least weight at 0.208. Zooming into singular metrics, NDVI (X10) and population density (X16) have the most significant weight, clocking in at 0.148 and 0.119, respectively.

Fig. 5

The G1-IEW-GT model was utilized to ascertain the values of the index weights.

Spatial–temporal distribution characteristics analysis of EGECC

According to the normal cloud model constructed above, the evaluation results of the EGECC and the three subsystems carrying capacity in the study area from 2000 to 2022 can be calculated. To further analyze the changes in the eco-geological environmental system in the research area in each period, the evaluation results for 2000, 2011, and 2022 were categorized into five grades of good, better, medium, worse, and bad, employing the spatial analysis tool of ArcGIS and the natural breakpoint method.

Analysis of variation characteristics in subsystem carrying capacity

The characteristics of the spatial and temporal distribution of the carrying capacity of the geo-environmental, ecological, and socio-environmental subsystems for the period 2000–2022 are demonstrated through Figs. 6, 7 and 8.

• 1. Geological Environmental Carrying Capacity (GECC).

From 2000 to 2022, the level of geo-environmental carrying capacity (GECC) shows a dynamic change of “deterioration followed by improvement” (Fig. 6). The red zone (poor geological environment area) was initially scattered. However, due to the frequent occurrence of geological disasters such as landslides and collapses, as well as the superposition of anthropogenic factors such as mineral exploitation and irrational land use, it gradually expanded into a large area of flaky area by 2011. However, since the introduction of the concept of green mine governance, local governments have effectively curbed the further deterioration of the geologic environment by carrying out major ecological protection projects such as the rehabilitation of abandoned mines and open-pit mines³⁷, and the red zone has significantly shrunk into a patchy distribution by 2022. In contrast, the blue zone (better geologic environment area) is mainly distributed in the Iloilo River Plain, where the geological structure is stable. This area remains relatively stable during the study period due to the flat topography covered by sedimentary rocks and balanced groundwater system, while the advantage of geo-environmental quality is more prominent due to less interference from anthropogenic activities. However, localized areas may face potential hazards such as ground subsidence due to water resources development³⁵.

Fig. 6

Spatio-temporal distribution patterns of geoenvironmental subsystem grade.

Overall, the dynamic change of GECC reveals the variability of vulnerability and resilience of the geoenvironmental system in the study area. The geological environment of mining areas and mountains is highly sensitive to external disturbances and relatively slow to recover. At the same time, the plains provide an important guarantee for sustainable development by virtue of their geological stability and environmental adaptability.

• 2. Ecological Environmental Carrying Capacity (EECC).

The ecological environment system in the study area maintained a relatively consistent level from 2000 to 2022 (Fig. 7). Areas with poor EECC are mainly found in the central urban regions in the northern part of the study region. This is primarily attributable to accelerated urbanization, which has disrupted the original vegetation and land resources, causing an imbalance in the ecosystem. Conversely, areas with higher EECC are concentrated in Luanchuan, Song County, northern Luoning County, and northern Xin’an County. The reasons for this are twofold: firstly, these areas contain numerous nature reserves and national forest parks, rich animal and plant species, and good vegetation conditions, which make the soil resources in the area less susceptible to rain erosion. Secondly, the well-developed water systems in these regions provide ample moisture and nutrients to the surrounding soil and create aquatic habitats for various species. Abundant water resources help maintain ecological balance and improve air quality, leading to higher EECC levels in these areas.

Fig. 7

Spatio-temporal distribution patterns of ecological environment subsystem grade.

• 3. Social Environmental Carrying Capacity (SECC).

During the study period, the overall socio-environmental carrying capacity (SECC) of Luoyang City exhibited a notable increase (Fig. 8), which can be primarily attributed to the accelerated growth of the region’s socio-economic strength. The Luoyang region’s gross national product (GDP) increased from 42.28 billion yuan in 2000 to 567.52 billion yuan in 2022, while the resident population rose from 6.332 million to 7.079 million. This reflects a simultaneous enhancement of the economic and population-carrying capacity. This growth trend is closely associated with the sustained optimization of the industrial structure. Luoyang is gradually transitioning towards a new industrialization path, attracting a significant workforce by supporting the development of advantageous and high-tech industries. This is providing crucial support for the stability of the SECC system¹².

Fig. 8

Spatio-temporal distribution patterns of social environment subsystem grade.

However, the enhancement of the SECC in Luoyang is characterized by the presence of notable regional imbalances. Luoyang’s northeastern and eastern regions, including Xin’an County and Yichuan County, exhibit markedly elevated levels of SECC compared to other counties. Notably, Xin’an County, which has a well-developed industrial chain pattern of economic development, has been included in the list of the top 100 counties in the country in terms of county development potential on multiple occasions¹⁷. This highlights the significant role of industrial agglomeration in promoting SECC. In contrast, the SECC of the remaining counties has increased at a slower pace, and the differences in socio-economic development between regions have further widened, thereby restricting the balanced development of SECC in the whole region.

Overall, the dynamic change of SECC in Luoyang City presents a dual feature of regional socio-economic development: on the one hand, SECC as a whole has been significantly enhanced through the optimization of industrial structure and the introduction of new development modes⁴⁶; on the other hand, the imbalance of regional economic development poses a constraint on the further enhancement of SECC. In the future, regional coordinated development and optimal resource allocation should be strengthened, and regional differences should be narrowed to enhance the overall socio-environmental carrying capacity.

Analysis of dynamic change characteristics in comprehensive development levels

Figure 9 illustrates the EGECC grade distribution pattern in the Luoyang area for 2000, 2011, and 2022. The overarching trend indicates a fluctuation in the eco-geological environment system of the Luoyang area from 2000 to 2022, characterized by an initial deterioration followed by subsequent improvement, with a notable inclination towards the eastern region over the western.

Fig. 9

Spatio-temporal distribution patterns of EGECC grade.

More specifically, regions designated as ‘good’ and ‘relatively good’ grades depict a diminishing-then-rising trend in their area coverage. Notably, the area classified as ‘good’ from 5527.23 km² in 2000 to 8173.39 km² in 2022, the area proportion increased as high as 47.92%. Conversely, regions classified as ‘medium,’ ‘worse,’ and ‘very poor’ grades show an opposite development trend. Despite a period of growth during the development process, they all showed varying degrees of reduction compared to 2000. Overall, from 2000 to 2011, a significant degradation trend was observed in EGECC, with an area of degradation amounting to 2169.19 km². However, from 2011 to 2022, there was a sustained enhancement in the comprehensive quality of the ecological geological environment.

Combined with the analysis of each subsystem, it was found that the degradation of EGECC during 2000–2011 was mainly affected by the geological and ecological environment systems. These degradations were mainly concentrated in the southwestern region with higher elevation, including Luanchuan, Songxian and Luoning regions. The underlying geological conditions of the region determine the level of EGECC to a certain extent, and the complicated topography and geomorphology of the southwestern high-altitude areas and the centrally distributed fault zones greatly influence the ecological and geological environmental systems¹⁹. At the same time, due to the abundance of mineral resources in these areas, the high-intensity mineral resource exploration and exploitation efforts have led to the continuous degradation of the ecological and geological environment in the region³². Over the eleven years from 2011 to 2022, the EGECC index grew by an impressive 109%, indicating a significant enhancement in Luoyang’s ecological geological environment quality. Since 2011, relevant departments have persistently promoted key ecological projects and issued the ' Henan Province 14th Five-Year Plan" for Ecological Restoration of Land Space and Construction of Forests Henan.' While the Luoyang Municipal Government has been extensively developing tourist attractions and nature reserves, it has also been continuously advancing the restoration of mining geological environments. Paramount ecological preservation and repair projects, including the treatment of historical abandoned mines in the YRB and the ‘Three Mountains’ rectification campaign, have been successively completed¹⁰, significantly mitigating the damage to the ecology and environmental pollution caused by mining activities.

Analysis of EGECC evolution driving mechanisms

Analysis of single-factor driving mechanisms

By utilizing the factor detector function in the geographical detector model, we revealed the driving effect of each factor on EGECC and its relative importance (Fig. 10). The top five factors with significant driving effects on EGECC are X7 Geomorphic Type (0.602) > X1 Elevation (0.577) > X15 HII (0.545) > X4 Geologic (0.462) > X16 Population Density (0.442). The driving force of all five factors is greater than 0.4, and their driving effect is significant, which is the main factor influencing the EGECC. The application of the factor detector function aims to explore the main driving factors for the evolution of the ecological and geological environmental system in the study area and to identify the magnitude of each driving force quantitatively. The in-depth analysis of the direction and path of action of the five major driving factors and their spatial differentiation will be discussed after applying the Geographically Weighted Regression model.

Fig. 10

Explanatory power value of each driving factor in 2000–2022.

Analysis of multi-factor driving mechanisms

In this study, based on the interaction detection function of the geographical detector model, the interaction relationship between indicators within the ecological-geological environment system was systematically analyzed. The results show (Fig. 11) that there are different degrees of interaction enhancement among the indicators, among which there are 110 pairs of two-factor enhancement types and 61 pairs of nonlinear enhancement types, and no single-factor independent effects are presented. This result verifies the rationality and scientificity of the indicator system constructed in the previous paper from the side. It is noteworthy that the most significant interaction enhancement effects were found between the various indicators and elevation (X1), geohazard susceptibility (X4), geomorphology type (X7), and human impact index (X15) (the factors with interaction enhancement effects greater than 0.5 are labeled in Fig. 11). This finding is consistent with the results of the single-factor detection, which further proves that these four indexes are the key factors driving the changes of EGECC, and reflects the science and effectiveness of the model construction.

Fig. 11

Multifactor interaction-driven results for 2000–2022.

Elevation (X1) and geomorphic types (X7), as representative indicators of topographic and geomorphic features in the study area⁴⁷, had the highest degree of interaction with other factors, with an interaction mean value of 0.63. Among them, the interaction values of elevation (X1) with geomorphic types (X7) and stratigraphy lithology (X5) were 0.67 and 0.66, respectively, which revealed that there was a significant coupling and synergistic effect between terrain undulation and geomorphic types and stratigraphy. In addition, the interaction value between geomorphic types (X7) and geologic hazard susceptibility (X4) reaches the highest value of 0.72, indicating that geomorphic types (X7) modulate geologic hazard susceptibility (X4) particularly. The presence of complex landforms (e.g., mountains, hills, etc.) is prone to accelerate soil erosion and surface runoff, which further induces geohazards such as landslides and mudslides⁴⁸. This result reflects that topographic and geomorphic features, as the basic variables of the eco-geological environment system, have a central position in the carrying capacity of the eco-geological environment.

When Geologic hazard susceptibility (X4) interacted with the indicators, the highest value appeared during the interaction with HII (X15) (0.72). This result indicates that the agglomeration of population and economic activities has a significant amplification effect on Geologic hazard susceptibility (X4). In order to meet the needs of human activities, the original topographic and landform features are often changed (such as mountain excavation, river diversion, etc.). These activities destroy the equilibrium state of the ecological geological environment system¹⁸. In addition, many geohazard-prone areas are precisely the key areas for resource development (e.g., mountainous mineral resource exploitation), so the natural geologic risks are superimposed with anthropogenic disturbances, which significantly aggravate the carrying pressure of the ecological and geologic environments³².

The interaction values of the human impact index (X15) with other indicators all exceeded 0.55, revealing the existence of significant synergistic or coupling effects between anthropogenic and natural factors. This result emphasizes that in the process of ecological protection and resource development, the relationship between human activities and the natural environment must be coordinated, especially in the highly sensitive areas of geological hazards and ecologically fragile areas, to reasonably control the intensity of human development and reduce anthropogenic disturbances, so as to enhance the stability and carrying capacity of the regional ecological environment^1,15.

Spatial heterogeneity analysis of drivers

This study used the geographically weighted regression (GWR) model to analyze the differences in the directions and paths of the first five factors driving changes in EGECC from the perspective of spatial heterogeneity. Since it is difficult for the GWR model to effectively deal with the spatial heterogeneity of qualitative variables (Geomorphic types X7), and the classification of geomorphic type is highly dependent on elevation data (X1)²⁰, and elevation (X1) itself has been identified as one of the top five driving factors. Therefore, this study did not include the geomorphic type (X7) in the GWR model and only analyzed the spatial heterogeneity characteristics of the remaining four factors (X1, X15, X4, and X16). After substituting these four factors into the GWR model, the goodness-of-fit of the model reaches 0.69, which indicates that it has a strong ability to explain ecological and geological environmental changes. The results showed that the regression coefficients of the driving factors showed significant clustering characteristics in the spatial distribution. Among them, Geologic hazard susceptibility (X4) negatively affected EGECC, while the other three factors showed bidirectional effect characteristics.

The elevation (X1) ranges from − 0.31 to 0.27 and exhibits an increasing trend with elevation (Fig. 12a). Regions where topography positively impacts EGECC are mainly concentrated in low-altitude plain areas. The flat terrain in these areas facilitates infrastructure construction, such as buildings and roads, and provides favorable conditions for agriculture⁴. Conversely, most negatively impacted locations are found in hilly and mountainous terrain at higher elevations, and this negative impact intensifies with increasing elevation. In these areas with significant terrain fluctuations, soil erosion is prone to occur under rapid runoff. Additionally, long-term mining activities have disrupted the original soil structure, increasing the risk of geological disasters such as landslides and debris flows²³, which are the main driving factors for the deterioration of the ecological geological environment.

Fig. 12

Regression coefficient of influencing factors: (a) X1, (b) X15, (c) X4, (d) X16.

The regression coefficient of the human impact index (X15) ranges from − 4.80 to 1.63, which is much higher than other factors. From the viewpoint of spatial distribution features (Fig. 12b), the adverse effects of human activities are mainly concentrated in urban central areas and western mining areas. Rapid population growth, urban expansion, traffic congestion, industrial production, mining, and road construction activities in urban areas have altered the original land use structure, causing severe damage to the stability of the eco-geological environment system¹. The intensity of human activities significantly impacts the level of EGECC, making X15 an essential factor influencing the eco-geological environment within the study region. Hence, effectively managing the interaction between human activities and the ecological geological environment system will improve EGECC⁴⁶.

The regression coefficient of geologic hazard susceptibility (X4) ranges from − 1.48 to − 0.13, peaking in the western mining areas and eastern urban areas, both indicating negative driving effects (Fig. 12c). Luoyang’s western mining areas primarily utilize open-pit mining methods, altering the original slope stability through ground and slope excavation, resulting in rock and soil mass deformations. Furthermore, soil erosion issues are exacerbated by wind and water erosion, intensifying collapses, landslides, and other geological disasters⁴⁹. X4’s impact is more prominent in the east urban areas than western regions. Shallow groundwater serves as the primary water supply source in the study area, yet with escalating water demands, groundwater extraction extract continues to rise, leading to a persistent decline in regional groundwater levels. Groundwater overexploitation directly contributes to ground subsidence in urban areas⁵⁰. Although the relevant authorities have strengthened the management of water resources problems and the EGECC situation has improved in most parts of the east, this potential impact still needs to be taken seriously.

The population density (X16) regression coefficient ranges from − 1.40 to 0.63, and its impact on EGECC is generally consistent with X15 (Fig. 12d), indicating a positive correlation between population density and human activity level. The positive impact of population growth is mainly manifested in economically underdeveloped urban fringe areas. In contrast, in central urban areas, the increase in population size has a restraining effect on EGECC. The prosperity of urban economies promotes population inflows, resulting in overly dense urban populations. The rapid urbanization process converts large amounts of natural land into construction lands, such as buildings and roads, which results in various ecological and environmental issues, such as vegetation degradation and loss of biodiversity^15,37. In contrast, urban fringe areas, with lower population densities and weaker interference from human activities, maintain ecosystem functions relatively well. Appropriately increasing population density can help improve EGECC.

Conclusion and policy suggestion

Conclusion

This research introduced 3S technology and selected 18 indicators based on the three-dimensional subsystems of the ecological, geological, and social environment, constructing a set of indicator systems suitable for evaluating EGECC in the Luoyang region of the YRB. Using the G1-IEW-GT model, the weight values for each indication were determined comprehensively, and the EGECC levels were evaluated using the normal cloud model, generating maps of EGECC distribution in the Luoyang region from 2000 to 2022 and because of the Geographic Detector model and the Geographically Weighted Regression model, investigated the driving mechanisms and spatial heterogeneity of the evolution of the Luoyang region’s ecological and geological environment system. The primary outcomes of this research can be summarized as follows:

• 1. The EGECC in the Luoyang region of the YRB has exhibited a fluctuating upward trajectory and a preference for the eastern region over the western region. Specifically, from 2000 to 2011, the EGECC exhibited a significant trend of widespread degradation, with a degraded area of 2169.19 km², while from 2011 to 2022, the ecological geological environment had a consistent enhancement in its overall quality.

• 2. Over the twenty-two years, the geological environment system has undergone a process of degradation and following restoration, while the ecological environment system has remained relatively stable, and the socio-economic growth displayed a consistent pattern of enhancement. In addition, there is a notable discrepancy in the spatial allocation of GECC, EECC, and SECC.

• 3. Geomorphic types, elevation, population density, geologic hazard susceptibility, and human activity impact index are the primary driving factors of ecological geological environment system evolution. Moreover, any two of the eighteen variables exhibit varying degrees of enhancement in driving EGECC. There were 110 pairs whose interaction type was double-factor enhanced and 61 pairs whose interaction type was nonlinear enhanced

• 4. The regression coefficients of the five major driving factors exhibit significant agglomeration characteristics in spatial distribution. The regression coefficient of the human activity impact index is the most significant among them, and it strongly influences the spatial distribution pattern of EGECC within the region. Geologic hazard susceptibility exhibits a negative driving effect on EGECC, while geomorphic types, human activity impact index, and population density all demonstrate bidirectional effects.

Policy suggestion

Given the intricate terrain geomorphological features, complexities of the Luoyang region, and the environmental impacts of mining activities, it is crucial to pursue sustainable regional economic development through Eco-geological environment systematic classification and zoning. Consequently, in order to address the current condition of the eco-geological environment in the Luoyang area, the following strategies are recommended in this study:

Firstly, it is crucial to implement effective measures against geological disasters. During mineral development, it is essential to continuously optimize the exploitation pattern of mineral resources and strictly regulate mining operations’ density and overall scale. In terms of environmental management within mining areas, delineating zones for encouraged, restricted, and prohibited mining activities minimizes the adverse impacts of mining on the ecological environment. Additionally, using remote sensing and other technical means, a timely and dynamic monitoring system should be built to efficiently foresee and monitor the development trend of geological disasters in the area, enhancing the region’s ability to respond to such disasters.

Secondly, it is imperative to bolster the protection and rational utilization of water resources. Focus on protecting essential water sources such as Guxian Reservoir, Xiaolangdi Reservoir, and Luhun Reservoir. To mitigate ongoing declines in groundwater levels, implement strict controls on groundwater extraction, with a particular emphasis on preventing contamination of shallow groundwater and the exploitation of deep groundwater. Through rational water resource management measures, ensure regional water supply security and uphold water quality standards.

Thirdly, protect existing forests and basic farmland and increase the vegetation cover index. Emphasize natural forest protection projects, implementing nurturing and selective logging of protective forests to strengthen the resilience of the ecological geological environment. Fully leverage the soil retention and water conservation functions of forest vegetation to effectively mitigate or prevent soil erosion, thereby reducing the occurrence of landslides and debris flows at their source.

Fourthly, it is crucial to safeguard vital ecological functional zones to uphold ecological security by harnessing the premium ecological resources of the Funiu Mountain area, intensifying the development of ecological conservation zones in the southern districts of Luoyang, including Luanchuan, Luoning, Ruyang, and Song counties, and tailoring development to local conditions, exploiting distinctive resources to cultivate industries (e.g., tourism) compatible with the environmental carrying capacity. Concurrently, bolstering protection and restoration efforts for wetlands like the Yellow River and Yiluo River, executing initiatives engineering for the conservation and sustainable utilization of sites such as the Xiaolangdi Reservoir, Mengjin Yellow River Wetland Reserve, and Luhun Lake National Wetland Park, will actively foster a cluster of wet.