Introduction

Ecological geology is a comprehensive study that integrates geology, ecology, engineering geology, hydrogeology, pedology, geomorphology, environmental science and their respective sub-disciplines. Drawing on the theories and methods of geology, it combines the ideas and concepts of ecology to investigate the ecological geological environment of mountains, rocks, soils, plant communities, groundwater, and surface water within the framework of Earth system science. It explores the ecological geological issues and effects generated under the influence of modern geological processes1. Russian scholars Trofimov and Korolev have conducted in-depth research on the subject of ecological geology internationally. Trofimov 2,3,4,5 focused on studying the research scope and content of ecological geology, different types of ecosystems, disciplinary categorization and position, as well as the connections and distinctions between ecological geology and related fields such as ecology, environmental geology, and engineering geology. Korolev6 delved into the definition, framework, classification of ecological geology, and its interplay with environmental geology.

In China, there is still a lack of consensus on the concept, research subjects, evaluation methods, and other aspects of ecological geology. According to Nie7, ecological geology is an interdisciplinary field that examines geological mechanisms, processes, and environmental conditions related to ecological issues. It combines ecological spatial distribution patterns, ecological issues, and their change patterns with geological processes to comprehensively gather information on ecology, soil, water, weathering shells (or vadose zones), and rocks above and below the ground. This research identifies and diagnoses major ecological issues and their causes, scientifically evaluates ecological conditions, and proposes natural-based solutions for ecological protection and restoration. On the other hand, Yuan8 focus on the constraint effect of surface geological processes on surface ecological development. They explore the tasks and goals of ecological geological investigation from the perspective of Earth system science and suggest that ecological geology falls under the category of ‘geology + ’.

The evaluation of vulnerability plays a pivotal role in the field of ecological geology investigation. Nonetheless, a uniform assessment methodology is presently absent. Commonly used evaluation techniques include the comprehensive index method, analytic hierarchy process (AHP), fuzzy comprehensive decision analysis integration with AHP, principal component analysis (PCA), and 3S technology methods. Each method has its own strengths and weaknesses. For instance, AHP is straightforward and practical for calculations, while PCA helps eliminate subjectivity in determining indicator weights.

These methods have been widely utilized in ecological vulnerability assessments, leading to the creation of popular research models like the sensitivity–resilience–pressure (SRP) model9,10,11,12, the pressure-state-response (PSR) model13,14,15, and the exposure-sensitivity-adaptability (VSD) model16,17.

The SRP model is constructed around the concept of ecosystem stability, aiming to encompass the main driving factors of ecosystem vulnerability. It also captures the overall characteristics of ecological vulnerability, making it suitable for assessing the ecological vulnerability of specific research areas18. It has been successfully applied in various regions of China including the Qilian Mountain area19, the Bailong River Basin in Gansu Province20, the Shiyang River Basin in Gansu Province21, and the Southwest Karst Mountainous Region22. Due to the enhanced ability of the SRP model to capture the essence of vulnerability, this study has opted to employ the SRP model as the ecological geological vulnerability assessment model.

Situated in the northern part of Shanxi Province in China, Northern Shanxi Province holds a strategic ecological position as a key area for soil conservation on the Loess Plateau and the upper reaches of the Yongding River, a vital water source conservation zone. However, the region faces ecological geological challenges such as environmental issues from open-pit mining, land salinization, and desertification. Therefore, assessing the ecological geological vulnerability in Northern Shanxi Province is crucial for improving the regional ecological environment, enhancing ecosystem functions, and promoting harmonious development between humans and nature. Hence, this study undertakes an assessment of the ecological geological vulnerability present in Northern Shanxi Province as the focal point of investigation.

Currently, there is limited research on the ecological geological vulnerability of Northern Shanxi Province. Previous studies have only conducted ecological vulnerability assessments in specific areas such as Sanggan River Basin23, Wutai Mountain region24, Datong City25, and Hunyuan County26. However, there is a lack of comprehensive analysis on the overall vulnerability of the region and the influence of geological factors on vulnerability. Simultaneously, in current vulnerability assessments, numerous scholars solely rely on singular evaluation methods such as AHP or PCA, thereby leading to limitations and uncertainties in the assessment outcomes.

The research area Northern Shanxi Province was selected for this study. Using SRP model, 11 evaluation indicators were scientifically chosen based on a regional ecological geology survey. The ecological geological vulnerability was assessed through AHP and PCA, establishing the ecological geological vulnerability index (EGVI) and the ecological geological vulnerability comprehensive index (EGVCI) for the first time, to uncover the distribution patterns of regional ecological geological vulnerability. The EGVI index is calculated based on multiple indicators, and the EGVCI index is calculated based on the EGVI, representing the average value of the vulnerability of a region.

The main scope of this study is: (1) to provide a more scientifically robust vulnerability assessment method by using SRP model and integrating AHP with PCA; (2) to identify the overall characteristics, spatial distribution, and driving factors of vulnerability in Northern Shanxi Province; (3) to offer data support and a scientific foundation for the integrated ecological protection, restoration, and comprehensive management of ‘mountains-water-forests-farmlands-lakes-grasslands’ for Northern Shanxi Province.

Description of the study area

The study area is situated in the northeastern part of Shanxi Province in China (38°30′ ~ 40°21′N, 111°51′ ~ 114°31′E), encompassing Datong City, Shuozhou City, and Xinzhou City (including Ningwu County, Fanshi County, Yuanping County, Wutai County, and DaiXian County), with a total area of approximately 36,176 km2 (Fig. 1).

Fig. 1
figure 1

Location and Digital Elevation Model (DEM) of study area (*the figure was drawn with QGIS 3.34.12 (https://www.qgis.org/download/) and CorelDRAW Graphics Suite 2023 (https://dl.cleverbridge.com/1103/rxrj06-29866-C0BB8CB7-397645995/CDGS2023.exe/)).

Full size image

Situated in the transition zone between the northeastern edge of the Loess Plateau and the Taihang Mountains in China, the study area spans an elevation range of 586 to 3047 m. The terrain is categorized into seven types: loess hills and terraces, loess gullies, loess hill gentle slope sand areas, bedrock mountain areas, earth-stone mountain areas, mountain front tilt plains, and alluvial plains. Climatically, the region is classified by Hengshan Mountains, with areas to the north falling under the temperate climate subzone and areas to the south categorized as warm temperate. The prevailing climate is semi-arid, characterized by significant diurnal temperature fluctuations, low precipitation, sand-laden winds, and abundant sunshine hours. The average annual temperature ranges from approximately 6.5 to 7.5 °C, with an average annual rainfall of 345 to 588 mm. The frost-free period typically lasts 120 to 140 days, and the average wind speed is around 3.0 m/s.

The ecological background of Northern Shanxi Province is relatively fragile, facing ecological geological issues such as environmental issues from open-pit mining, land salinization, and desertification. Due to its significant ecological strategic position to the upper reaches of the Yongding River, assessing ecological geological vulnerability in this region and proposing targeted ecological protection and restoration suggestions are of great theoretical and practical significance.

Material and methods

Data acquisition and processing

The data utilized in this study comprises DEM elevation data, meteorological data, land use type data, and socio-economic data.

The DEM elevation data was obtained from Geospatial Data Cloud of Computer Network Information Center, Chinese Academy of Sciences (https://www.gscloud.cn/) at a spatial resolution of 12.5 m. Rainfall and evaporation data were acquired from the annual average rainfall and annual average evaporation spatial interpolation data provided by Resource and Environmental Science and Data Platform, Chinese Academy of Sciences (https://www.resdc.cn/). The improved Night Light Index (NLI) data, soil type data and mineral exploitation intensity data were sourced from Resource and Environmental Science and Data Platform, Chinese Academy of Sciences (https://www.resdc.cn/). Land use type data was interpreted based on the 2022 Landsat-8 OLI imagery, while EVI data was calculated using Landsat-8 OLI imagery. Engineering geological lithology, soil parent material, water conservation capacity, and soil erosion intensity data were all derived from the author’s survey and research findings (Table 1). Finally, all indicator data was resampled into 250 m resolution raster data for subsequent analysis and evaluation.

Table 1 Description and source of evaluation indicators.
Full size table

The Enhanced Vegetation Index (EVI) is computed using the Eq. (1) 27 as shown below:

$$EVI = G \times \frac{{\rho_{{{\text{NIR}}}} - \rho_{{{\text{Red}}}} }}{{\rho_{{{\text{NIR}}}} + C_{1} \times \rho_{{{\text{Red}}}} - C_{2} \times \rho_{{{\text{Blue}}}} + L}}$$
(1)

where: \(\rho_{{{\text{NIR}}}}\) represents the near-infrared band reflectance; \(\rho_{{{\text{Red}}}}\) denotes the red band reflectance; \(\rho_{{{\text{Blue}}}}\) indicates the blue band reflectance; G is a gain factor, set at 2.5; C1 stands for the atmospheric-corrected red band calibration parameter, set at 6.0; C2 signifies the atmospheric-corrected blue band calibration parameter, set at 7.5; L represents the soil-adjustment factor, set at 1.0.

The water conservation capacity is determined by applying the water balance equation Eq. (2) 28 as shown below:

$$TQ = \mathop \sum \nolimits_{i = 1}^{j} \left( {P_{i} - R_{i} - ET_{i} } \right) \times A_{i} \times 10^{3}$$
(2)

where: TQ stands for the total water conservation capacity (m3), Pi represents precipitation (mm), Ri denotes surface runoff (mm), ETi signifies evapotranspiration (mm), Ai indicates the area of ecosystem type i (km2), where i represents the i type of ecosystem in the study area, and j represents the total number of ecosystem types in the study area.

The soil erosion intensity is assessed using Chinese soil loss equation (CSLE) Eq. (3) 29 as shown below:

$$A = R \cdot K \cdot L \cdot S \cdot B \cdot E \cdot T$$
(3)

where: A represents annual average soil loss (t hm-2 a-1); R is rainfall erosivity (MJ mm hm-2 h-1 a-1); K is soil erodibility (t h/(MJ mm)); L and S are dimensionless slope length and slope steepness factors; B, E, and T represent dimensionless factors of biological-control, engineering-control, and tillage practices respectively.

Methods

Select evaluation model and indicators

When conducting a comprehensive assessment using multiple variables, it is typically necessary to select appropriate indicators and construct composite indexes for evaluation purposes. It is generally believed that an indicator is a metrics used to assess a particular state or characteristic of a system (natural, ecological, geological, economic, and social, et al.), either qualitatively or quantitatively. An index is a comprehensive quantitative statistic, typically derived through a mathematical method from multiple indicators or variables.

Based on the ecological geology survey conducted by the authors, Northern Shanxi Province belongs to the semi-arid to arid area of northern China. The main ecological issues within the region include soil erosion, environmental problems caused by open-pit mining, land salinization, and land desertification. Therefore, the assessment of ecological geological vulnerability primarily focuses on these ecological issues.

Following the scientific, holistic, targeted, and feasible principles in selecting indicators, slope and soil erosion intensity were chosen as evaluation indicators for soil erosion issues. For environmental problems caused by open-pit mining, mineral exploitation intensity and NLI were selected as evaluation indicators. Considering the semi-arid to arid climatic background, average annual precipitation and water conservation capacity were included as evaluation indicators. Taking into account the impact of geological factors on ecological vulnerability, engineering geological lithology and soil parent material were selected as evaluation indicators. Considering the influence of surface vegetation on ecological vulnerability, land use type and EVI were chosen as evaluation indicators.

The SRP model is a comprehensive assessment model constructed based on the stability of ecosystems. With an inherently unstable framework structure, it exhibits sensitivity to external disturbances and may evolve in a direction unfavorable to its own development due to a lack of coping mechanisms19. It is widely applied in ecological vulnerability assessment research.

In consideration of the ecological geological background and typical ecological issues in Northern Shanxi Province, 11 indicators have been selected to develop SRP model to conduct ecological geological vulnerability assessment. These indicators include slope, average annual precipitation, engineering geological lithology, soil type, soil parent material, soil erosion intensity, land use type, EVI, water conservation capacity, mineral exploitation intensity, and NLI. The SRP model comprises three components: ecological sensitivity, ecological resilience, and ecological pressure. Ecological sensitivity is represented by 7 indicators, while ecological resilience and ecological pressure are each represented by 2 indicators (Table 2).

Table 2 Ecological geological vulnerability assignment indicator system for Northern Shanxi Province.
Full size table

Based on the determination of eleven evaluation indicators (slope, average annual precipitation, engineering geological lithology, et al.), the ecological geological vulnerability assessment of Northern Shanxi Province was conducted using AHP and PCA methods in GIS. The technical roadmap of this study can be observed in Fig. 2.

Fig. 2
figure 2

Technical roadmap of the study (*the figure was drawn with CorelDRAW Graphics Suite 2023 (https://dl.cleverbridge.com/1103/rxrj06-29866-C0BB8CB7-397645995/CDGS2023.exe/)).

Full size image

The AHP involves several key steps: evaluating different indicators and grading; calculating the weights for each indicator. The main steps of PCA evaluation involve: standardizing and assigning values to the indicators; determining the principal components. Then, both of AHP and PCA calculate the EGVI and EGVCI indexes‌ for a comprehensive assessment. The evaluation results of EGVI and EGVCI from both methods are comprehensively compared to assess the overall characteristics, spatial distribution, and driving factors of the vulnerability. Finally, based on the characteristics of the vulnerable regions and the ecological problems, targeted suggestions for ecological protection and restoration are put forward for regional ecological management.

Analytic hierarchy process (AHP)

The Analytic Hierarchy Process (AHP) is a multi-level weight decision-making method known for its systematic, hierarchical, and flexible characteristics30. The AHP decomposes problems into constituent elements (indicator selection) and aggregates indicators into different levels of combination based on their interrelationships, influences, and hierarchical relationships, forming a multi-level analytical structural model.

The AHP method is an exploratory method that decomposes complex problems layer by layer and finally uses multiple indicators for evaluation. The key to the AHP lies in the selection of indicators and the determination of indicator weights. Choosing different evaluation indicators, or the same indicators but with different weights, will result in different evaluation results. Therefore, the evaluation of AHP is influenced by human subjectivity.

AHP method has been extensively utilized in vulnerability assessments across different regions of China, including the Qaidam Basin in Qinghai Province31, Sichuan Province32, source region of Yangtze River33, karst mountain areas in Guizhou Province 34, and Gannan region of Jiangxi Province35. Simultaneously, AHP has also found extensive application in disaster risk assessments in Greece36,37.

The AHP involves several key steps: initially, evaluating different indicators and grading; then, constructing a judgment matrix to determine the relative importance with pair-wise comparisons, conducting consistency checks of the judgment matrix, and calculating the weights for each indicator; ultimately, calculating the EGVI and EGVCI indexes‌ for a comprehensive assessment.

  1. (1)

    Indicator grading assignment

The evaluation of the AHP necessitates the grading and assignment of values to the indicators. The slope index is classified according to the ‘Comprehensive control of soil and water conservation—general rule of planning (GB/T 15,772–2008)’ of China; the soil erosion intensity index is classified according to the ‘Standards for classification and gradation of soil erosion’ (SL190-2007) of China; the mineral exploitation intensity index is classified according to the grading method of annual coal production in loess hilly mining areas38; average annual precipitation, EVI, water conservation capacity, and NLI index are classified using the ‘Natural Breaks (Jenks) method’ in GIS (Table 3).

Table 3 Grading assignment standards of ecological geological vulnerability for 11 indicators.
Full size table

Qualitative indicators (engineering geological lithology, soil type, soil parent material, land use type) are graded and assigned mainly through field investigations. The impact of engineering geological lithology on vulnerability is mainly reflected in the impact of rock weathering resistance on soil thickness; the impact of soil type on vulnerability is mainly reflected in the difference in soil nutrient status; the parent material is the initial source of soil formation and the impact of soil parent material on vulnerability is mainly reflected in the difference in the nutrient status of soil parent material; the impact of land use type on vulnerability is mainly reflected in the difference in ecosystem service functions. According to field investigations combined with research analysis, the value assignment of qualitative indicators is completed in GIS (Tables 2 and Table 3).

The impact and level of each indicator on the regional ecological geological vulnerability are determined, and single indicator evaluations are conducted (Figs. 3 and 4a–e).

Fig. 3
figure 3

Vulnerability evaluation of single indicator by AHP for Northern Shanxi Province (*complete content of the legends for (c, d and e) explained in Table 3; the figure was drawn with QGIS 3.34.12 (https://www.qgis.org/download/) and CorelDRAW Graphics Suite 2023 (https://dl.cleverbridge.com/1103/rxrj06-29866-C0BB8CB7-397645995/CDGS2023.exe/)).

Full size image
Fig. 4
figure 4

Vulnerability evaluation of single indicator and ecological geological vulnerability (EGVI) assessment by AHP for Northern Shanxi Province (*complete content of the legends for (a) explained in Table 3; the figure was drawn with QGIS 3.34.12 (https://www.qgis.org/download/) and CorelDRAW Graphics Suite 2023 (https://dl.cleverbridge.com/1103/rxrj06-29866-C0BB8CB7-397645995/CDGS2023.exe/)).

Full size image
  1. (2)

    Indicator weight determination

Determining the weights of indicators is crucial in the evaluation of the AHP. Based on field investigations in the research area, the study focuses on the environmental issues stemming from open-pit mining and key ecological geological concerns like land salinization. The research assesses the impact levels of different indicators on regional ecological geological vulnerability, establishes the relative importance of these indicators.

The specific steps of AHP are as follows: compare all indicators in pairs, build a judgment matrix based on relative importance; use the consistency ratio (CR) to perform consistency test on the judgment matrix. When CR is less than 0.1, it is considered that the judgment matrix passes the consistency test; calculate the weight of each indicator based on the judgment matrix.

The consistency ratio (CR) of the judgment matrix below 0.1 indicates satisfactory consistency. By conducting consistency testing, the weights of each indicator are determined as outlined in Table 4.

Table 4 Weight of 11 indicators by AHP.
Full size table
  1. (3)

    Comprehensive assessment

The ecological geological vulnerability index (EGVI) for Northern Shanxi Province is assessed using Eq. (4) as shown below:

$$EGVI = \mathop \sum \limits_{i = 1}^{j} C_{i} \cdot W_{i}$$
(4)

where: EGVI stands for the ecological geological vulnerability index, Ci represents the vulnerability assignment of the indicator i, Wi represents the weight of the indicator i, j indicates the total number of the indicators.

In the realm of ecological geology, the ecological geological vulnerability comprehensive index (EGVCI) is utilized to assess the overall vulnerability of specific areas like eco-geological zones and river basins, standing for an average vulnerability value of a region and aiding in regional management and governance. EGVCI is assessed using Eq. (5) as shown below:

$$EGVCI = \mathop \sum \limits_{i = 1}^{j} P_{i} \cdot \frac{{A_{i} }}{S}$$
(5)

where: EGVCI stands for the ecological geological vulnerability comprehensive index of a evaluation region, j represents the total number of vulnerability units in the evaluation region, Pi represents the vulnerability index value (EGVI) of the vulnerability unit i, Ai represents the area of the vulnerability unit i, S indicates the total area of the evaluation region.

Based on field surveys and regional ecological geological background characteristics, the ecological geological vulnerability for Northern Shanxi Province is categorized into 5 levels, as outlined in Table 5.

Table 5 Classification criteria of ecological geological vulnerability for Northern Shanxi Province.
Full size table

Using GIS, the EGVI index for Northern Shanxi Province is computed by considering the evaluation results of 11 individual indicators and their respective weights by Eq. (4). Based on the classification of EGVI in Table 5, the ecological geological vulnerability assessment diagram by AHP for Northern Shanxi Province was obtained (Fig. 4f). The ecological geological vulnerability assessment diagram (Figs. 4f and 5) is the distribution map of the EGVI index, and its value of each pixel represents the EVGI value.

Fig. 5
figure 5

Ecological geological vulnerability (EGVI) assessment diagram by PCA for Northern Shanxi Province (*the figure was drawn with QGIS 3.34.12 (https://www.qgis.org/download/) and CorelDRAW Graphics Suite 2023 (https://dl.cleverbridge.com/1103/rxrj06-29866-C0BB8CB7-397645995/CDGS2023.exe/)).

Full size image

Principal component analysis (PCA)

The PCA can be utilized to assess the vulnerability of the ecological environment by replacing original multidimensional variables with a small number of comprehensive variables through linear transformation. This method helps to objectively determine weight, avoiding subjectivity and arbitrariness39. It has been widely applied in vulnerability assessments in various regions of China such as the Oroqen region in the Inner Mongolia40, Danxia Indus Nature Reserve in Nanxiong District in Guangdong Province41, Qianjiang City in Hubei Province42, Guiyang City43, and at the national scale44. Simultaneously, PCA has been widely utilized in the assessment of ecological vulnerability in terrestrial ecosystems45 and social vulnerability to geological disasters46.

The main steps of PCA evaluation involve: identifying the impacts and levels of influence of various factors on regional ecological geological vulnerability, standardizing and assigning values to the indicators; constructing a correlation matrix among multiple indicators, calculating the eigenvalues and eigenvectors of this matrix to extract the principal components; finally computing the EGVI and EGVCI indexes‌ for comprehensive assessment.

  1. (1)

    Indicator standardization and grading assignment

Positive indicators are those where the value of the indicator is positively correlated with ecological vulnerability. The larger the value of the indicator, the higher the vulnerability; negative indicators are those where the value of the indicator is negatively correlated with ecological vulnerability. The larger the value of the indicator, the lower the vulnerability.

In this study, positive indicators (slope, soil erosion intensity, mineral exploitation intensity, and NLI) and negative indicators (average annual precipitation, EVI, and water conservation capacity) are defined in Table 2.

Positive indicators (slope, soil erosion intensity, mineral exploitation intensity, and NLI) and negative indicators (average annual precipitation, EVI, and water conservation capacity) were standardized using the min–max standardization method47 by Eq. (6) and Eq. (7) respectively.

$$X_{{^{i} }}^{\prime } = {{\left( {X_{i} - X_{i\min } } \right)} \mathord{\left/ {\vphantom {{\left( {X_{i} - X_{i\min } } \right)} {\left( {X_{i\max } - X_{i\min } } \right)}}} \right. \kern-0pt} {\left( {X_{i\max } - X_{i\min } } \right)}} \times 10$$
(6)
$$X_{{^{i} }}^{\prime } = {{\left( {X_{i\max } - X_{i} } \right)} \mathord{\left/ {\vphantom {{\left( {X_{i\max } - X_{i} } \right)} {\left( {X_{i\max } - X_{i\min } } \right)}}} \right. \kern-0pt} {\left( {X_{i\max } - X_{i\min } } \right)}} \times 10$$
(7)

where: \(x_{i}^{\prime }\) represents the standardized value of indicator i, xi represents the original value of indicator i, ximax and ximin represent the maximum and minimum values of indicator i respectively. The range of the standardized values is 1 ~ 10.

For qualitative indicators (engineering geological lithology, soil type, soil parent material, and land use type), the grading and assignment method follows the same criteria as presented in Table 3.

  1. (2)

    Principal components determination

Utilizing ‘Principal Component’ tool in ArcGIS software, a correlation matrix was constructed between multiple indicators. The eigenvalues and eigenvectors of the matrix were calculated, and the principal components were outputted. With a cumulative variance contribution rate greater than 85%, it was determined that the number of principal components for this analysis is 6. The characteristics of each principal component are detailed in Table 6.

Table 6 Characteristics of principal components.
Full size table

Utilizing the factor loading matrix of the principal components (Table 7), one can derive linear combination expressions between the 6 principal components and the original 11 indicators, resulting in the output of each principal component.

Table 7 Factor loading matrix of principal components.
Full size table
  1. (3)

    Comprehensive assessment

When evaluating by PCA, the ecological geological vulnerability index (EGVI) for Northern Shanxi Province is assessed using Eq. (8) as shown below:

$$EGVI = \mathop \sum \limits_{i = 1}^{j} a_{i} \cdot F_{i}$$
(8)

where: EGVI stands for the ecological geological vulnerability index, αi represents the contribution rate of the principal component Fi, Fi indicates the i principal component, calculated according to the linear expression of the principal component, and j is the total number of principal components.

According to Eq. (8), the EVGI by PCA is assessed using Eq. (9) as shown below:

$$EGVI = 0.3138 \cdot F_{1} + 0.2526 \cdot F_{2} + 0.1124 \cdot F_{3} + 0.0835 \cdot F_{4} + 0.0695 \cdot F_{5} + 0.0605 \cdot F_{6}$$
(9)

Utilizing ‘Raster Calculator’ tool in ArcGIS to compute the EVGI index, categorize it based on the EVGI index grading standards (Table 5), and produce the ecological geological vulnerability assessment diagram by PCA for Northern Shanxi Province (Fig. 5).

Results

Distribution characteristics of ecological geological vulnerability

The evaluation conducted using AHP (Table 8) reveals that the EGVCI index for Northern Shanxi Province is 3.97. The area of slight vulnerability covers approximately 4931 km2, representing 13.63% of the total area; while the mild vulnerability area encompasses around 22,666 km2, accounting for about 62.66%. Collectively, regions with moderate to extreme vulnerability represent about 23.7% of the total area.

Table 8 Area statistics of ecological geological vulnerability grading compared AHP with PCA in Northern Shanxi Province.
Full size table

Based on the results of PCA (Table 8), the EGVCI index for Northern Shanxi Province is 3.95. The area of slight vulnerability is around 6908 km2, making up about 19.09% of the total area; while the mild vulnerability area is approximately 20,666 km2, accounting for about 57.13%.

The combined proportion of moderate to extreme vulnerability is about 23.8%.

The integrated analysis indicates that the average value of the EGVCI index by AHP and PCA for Northern Shanxi Province is approximately 3.96, with an overall mild vulnerability level of the regions (Table 5). This suggests that for the entire region of Northern Shanxi Province, the overall vulnerability is not high, and ecological restoration efforts should primarily focus on natural recovery. Specifically, in ecologically problematic areas primarily caused by natural factors such as forest and grassland degradation and soil erosion, the ecological restoration approach should emphasize natural recovery. Conversely, in areas affected by anthropogenic factors like mining activities and land salinization, the ecological restoration strategy should prioritize artificial restoration, with natural recovery playing a supporting role.

In terms of spatial distribution, the regions of moderate to extreme vulnerability mainly distributed in the basin-mountain transition zones and the western low-mountain areas, such as Xinrong District, Huairen City, Youyu County, Shanyin County, Pinglu District, Fanshi County, Ningwu County, Yuanping City, and Wutai County.

Driving factors of ecological geological vulnerability

In this study, the top 6 indicators in AHP are: soil erosion intensity, water conservation capacity, mineral exploitation intensity, EVI, land use type, and average annual precipitation. PCA reveals that slope and soil erosion intensity have higher loadings in the principal component F1, mineral exploitation intensity in the principal component F2, soil type in the principal component F3, land use type in the principal component F4, soil erosion intensity in the principal component F5. The leading factors identified are: slope, soil erosion intensity, mineral exploitation intensity, soil type, and land use type.

The study area is located in the loess hilly area, characterized by high soil erosion and severe water and soil loss issues. Therefore, soil erosion intensity is considered a driving factor in vulnerability, representing the extent of regional water and soil loss. Open-pit mining (coal and iron ore) in Northern Shanxi Province has led to significant ecological damage, making mineral exploitation intensity another crucial factor in vulnerability, reflecting the state of regional mining development. The function of water conservation plays a critical role in vegetation restoration and soil erosion control, particularly vital in semi-arid to arid regions. Thus, water conservation capacity is included as a primary controlling factor in vulnerability, representing the region’s water conservation status. Land use type significantly constrains vegetation distribution in the region, and changes in land use type also impact soil and water conservation as well as water conservation functions. Slope plays a crucial role in soil erosion and land use practices. EVI (vegetation coverage) can indicate the quality of the ecological environment, directly influencing the changes in soil, water, and other environmental factors. Therefore, land use type, slope, and EVI are considered primary controlling factors in vulnerability.

Considering the weightage ranking of indicators from both AHP and PCA, and the semi-arid to arid climate of Northern Shanxi Province, the driving factors of ecological geological vulnerability in the region are believed to be soil erosion intensity, mineral exploitation intensity, water conservation capacity, land use type, slope, and EVI.

Recommendations for restoration and management of vulnerable regions

The research area is situated in the loess hilly region of northern China, characterized by a semi-arid to arid climate. Additionally, it is affiliated with the coal resource development zone and serves as an integral component of the northern agricultural-pastoral transitional belt.

Soil erosion issues are mainly concentrated in the western loess hilly areas of the research area, including Xinrong District, Youyu County, and Pinglu District. The terrain in this region is undulating, the soil is loose, vegetation cover is sparse, and soil erosion is intense, resulting in a landscape of gullies and ravines. Mining activities are relatively active in the area, leading to ecological problems such as vegetation destruction, geological disasters (landslides, collapses, subsidence et al.), and land occupation. Open-pit mines mainly consist of coal mines and iron mines, with coal mines mainly located in Datong City and Shuozhou City, and iron mines mainly in Xinzhou City. Open-pit mining has a severe impact on the ecological environment. Land desertification is primarily found in the northwest part of the research area, including Shuocheng District, Pinglu District, Youyu County, Zuoyun County, Xinrong District, Yanggao County, and Tianzhen County. The region features undulating terrain, gentle slopes, loose soil mainly composed of sandy loess, a dry climate, strong winds leading to sand accumulation, and low vegetation cover. The loose soil surface is directly affected by wind erosion, resulting in land desertification.

Based on the ecological background and main ecological issues of the research area, the ecological conservation and restoration objectives can be summarized as curbing regional soil erosion, promoting environmentally friendly mining methods in mines, and enhancing regional water conservation capacity. Targeted ecological protection and management recommendations are proposed for addressing the main ecological issues in different vulnerability areas.

  1. (1)

    Extreme vulnerability region

The extreme vulnerability region is predominantly located in Pinglu District, Shanyin County in Shuozhou City, the low mountain area in the northwest of Yungang District in Datong City, mountainous areas in Hunyuan County, eastern mountainous areas of Ningwu County in Xinzhou City, and western mountainous areas of Wutai County, totaling approximately 147 km2 (Fig. 4f). The primary ecological geological issues in these regions include environmental problems caused by open-pit mining, land desertification, and soil erosion.

The following strategies and recommendations for ecological restoration and management are proposed in response to these ecological issues: strengthening ecological restoration of historical abandoned mines, improving ecological restoration and governance of waste disposal, subsidence areas, and mining sites. Additionally, promoting integrated management in the loess hilly areas in the western region through measures like gully consolidation, terracing for soil conservation, afforestation on abandoned slopes, and sand-fixation with irrigation-grass integration to control soil erosion and land desertification. It is also important to carry out soil improvement in moderation, advance soil erosion control, and enhance the protection and management of water conservation areas in the eastern and southern mountainous regions.

  1. (2)

    Severe vulnerability region

The region of severe vulnerability is mainly distributed in Shanyin County in Shuozhou City, Yungang District, Tianzhen County, the southern mountainous areas of Hunyuan County in Datong City, the eastern mountainous areas of Ningwu County, and the western mountainous areas of Wutai County in Xinzhou City, covering an approximate area of 562 km2 (Fig. 4f), with the major ecological geological issues of environmental problems resulting from open-pit mining, soil erosion, and forest degradation.

The subsequent strategies and suggestions for ecological restoration and management are posited as a response to the aforementioned ecological challenges: enhancing the development of a protective forest system focusing on windbreak and sand fixation, while progressively increasing vegetation coverage. This approach aims to preserve the stability of the regional ecosystem, enhance regional water conservation capabilities, and prevent the deterioration of forest, shrub, and grassland ecosystems, as well as the decline of ecological functions.

  1. (3)

    Moderate vulnerability region

The region of moderate vulnerability is mainly concentrated in Pinglu District in Shuozhou City, Xinrong District in Datong City, southeastern part of Tianzhen County, and Hunyuan County to Guangling County, covering an area of approximately 7870 km2 (Fig. 4f), with the major ecological geological issues of soil erosion, land salinization, and water environment problems.

The following strategies and recommendations for ecological restoration and management are proposed as a reaction to the previously mentioned ecological issues: in mountainous and hilly areas, focus should be on soil erosion control, adjusting the agricultural, forestry, and animal husbandry industry structure, and promoting green forestry and animal husbandry. It is also recommended to adapt crop varieties to local conditions and develop dryland agriculture.

In the central part of Datong Basin, it is important to control underground water extraction, prevent underground water funnels and land subsidence, and manage groundwater quality. Efforts should be made to enhance the overall water quality of the Sanggan River Basin and gradually eliminate inferior Class V water bodies. Additionally, actively promoting the reclamation and transformation of saline-alkali wasteland can improve land utilization efficiency.

Discussion

Evaluation results of the vulnerability

The present study utilized AHP and PCA to compute EGVI and EGVCI indexes for vulnerability assessment. Findings revealed that: (1) under both methods, the value of EGVCI index for Northern Shanxi Province were similar. The AHP calculated EGVCI as 3.97, while PCA yielded 3.95, both indicating a mild level of vulnerability (Table 5); (2) the total area proportions of moderate to extreme vulnerability are comparable between the two methods (Fig. 6, Table 8). The proportion of the area of moderate to extreme vulnerability in AHP is 23.71%, while it is 23.78% in PCA; (3) the spatial distribution of regions with severe vulnerability and extreme vulnerability remains consistent in both assessment methods. These results highlight the rationality of employing both AHP and PCA in this evaluation, thereby ensuring the credibility of the assessment results.

Fig. 6
figure 6

Area proportion of vulnerability grading compared AHP with PCA (*the figure was drawn with CorelDRAW Graphics Suite 2023 (https://dl.cleverbridge.com/1103/rxrj06-29866-C0BB8CB7-397645995/CDGS2023.exe/)).

Full size image

Ye48 utilized 17 indicators to assess the ecological vulnerability of various prefecture-level cities in Shanxi Province, concluding that Shuozhou City’s ecological environment is at a moderate level of vulnerability. Meanwhile, Shi49 employed the Analytic Hierarchy Process, selecting 12 indicators to evaluate the ecological vulnerability of prefecture-level cities in Shanxi Province, determining that Shuozhou City’s ecological environment is at a slight level of vulnerability, and Datong City at a moderate level. In comparison to their findings, the evaluation results of this study exhibit a certain degree of consistency.

In terms of spatial distribution, the research reveals that areas with higher ecological vulnerability are primarily located in the hilly loess region in the western part of the study area. This finding is consistent with the study on the hilly loess region50, where regions with higher ecological vulnerability are situated at middle to low altitudes, characterized by severe soil erosion. Soil erosion is identified as a key factor influencing ecological vulnerability.

Currently the assessment of vulnerability in a region lacks a standardized method for validation. This deficiency is believed to stem from the absence of rigorous mathematical proof at a large scale (regional level) and all-round dynamic monitoring data. Presently, our verification of the vulnerability assessment results primarily relies on point-by-point investigations. Specifically, we select regions ranging from moderate to extreme vulnerability, where multiple survey points are established to conduct on-site ecological issues investigations. The main focus is to determine whether significant ecological issues exist in regions with higher vulnerability in the assessment results, thus using the findings from on-site surveys to validate the accuracy of vulnerability assessment results. The vulnerability assessment findings were corroborated through on-site field validation, revealing that areas classified as moderate to extreme vulnerability exhibited varying degrees of ecological issues such as soil erosion, mining activities, and land desertification. The alignment between the assessment outcomes and empirical observations was notably strong.

Driving factors of the vulnerability

The present study conducted an analysis of the ecological background, prevailing ecological issues, and underlying causes of vulnerability in the research area. By employing AHP and PCA, it was determined that the driving factors of ecological geological vulnerability in Northern Shanxi Province included soil erosion intensity, mineral exploitation intensity, water conservation capacity, land use type, slope, and EVI.

Through the utilization of various assessment methods and selection of different evaluation indicators, distinct outcomes regarding the driving factors of vulnerability may arise.

Yang51 conducted an assessment of ecological vulnerability for loess area in the southern of Northern Shanxi Province, identifying vegetation coverage, precipitation, slope, and human activities as the main driving factors. In a similar study, Mao23 focused on the Sanggan River Basin in the central region of Northern Shanxi Province, utilizing PCA with 16 indicators for ecological vulnerability assessment. The primary driving factors identified in his study included soil erosion intensity, land use type, vegetation coverage, precipitation, and slope. A comparative analysis with their findings reveals a certain level of consistency in this study.

The integrated analysis suggests that the ecological vulnerability of Northern Shanxi Province results from the combined influence of natural and anthropogenic factors. The semi-arid to arid climate, low precipitation levels, distinctive loess hilly terrain (slope), and soft soil texture have led to ecological deficiencies (EVI), weak water conservation capabilities (water conservation capacity), and severe soil erosion issues (soil erosion intensity) in the region. Human activities (mineral exploitation intensity, land use type) have further exacerbated the degradation of the ecological environment. Therefore, considering soil erosion intensity, mineral exploitation intensity, water conservation capacity, land use type, slope, and EVI as the driving factors of vulnerability holds significant practical implications.

Uncertainty of the vulnerability evaluation methods

In the current field of vulnerability assessment, both AHP and PCA are widely used methodologies; however, they exhibit certain limitations. Many scholars tend to rely on a singular approach when conducting vulnerability assessments, leading to relatively high uncertainties in the assessment outcomes.

The process of calculating indicator weights in AHP involves inherent subjectivity. Even with the same set of indicators, different scholars may assign varying weights, resulting in potential discrepancies in evaluation outcomes. To mitigate such subjectivity, a thorough analysis of the primary ecological issues within the research area is necessary. By identifying relevant variables that represent these ecological issues, one can assess the impact levels and relative importance of each indicator more accurately.

PCA in applications emphasizes the thorough assessment of factor information contribution. The evaluation outcomes are heavily influenced by the quality and accuracy of the original data factors. Hence, it is crucial to prioritize the quality and accuracy of the original data pertaining to significant factors. If the principal component factor loading is negative, the interpretability of the principal component linear combination function diminishes. In such cases, careful consideration is necessary when interpreting the practical significance of the principal component linear combination, and additional analytical methods should be employed for validation.

It is noteworthy that many scholars utilize PCA to determine the number of principal components based on the cumulative variance contribution rate exceeding 85%. This study reveals variations in evaluation results depending on the chosen number of principal components, as shown in Table 9.

Table 9 Area statistics of ecological geological vulnerability grading under different number of principal components with PCA in Northern Shanxi Province.
Full size table

From Table 9, when the number of principal components is 6, the cumulative variance contribution rate is 89.22%. At this point, the EGVCI value is 3.95, and the proportion of moderate to extreme vulnerability areas is 23.78%; When the number of principal components is 8, the cumulative variance contribution rate is 95.58%. The EGVCI value at this stage is 4.16, and the proportion of moderate to extreme vulnerability areas is 29.82%; With 11 principal components, the cumulative variance contribution rate reaches 100%. The EGVCI value is 4.30, and the proportion of moderate to extreme vulnerability areas is 35.63%. As the number of principal components increases, the interpretability of vulnerability information also increases, along with the rise in EGVCI value and the proportion of moderate to extreme vulnerability areas.

The above findings indicate that varying the number of principal components selected leads to distinctive outcomes. As the number of principal components increases, the focal point of regional vulnerability gradually shifts towards areas with higher values.

Currently, there is no uniform standard or specification for assessing ecological geological vulnerability. Due to the multidisciplinary nature of indicators in ecological geological vulnerability assessment, which involve ecology, geography and geology, et al., different scholars selecting various evaluation indicators for the same region may also yield diverse assessment outcomes. This study emphasizes that when conducting vulnerability assessments, it is essential to start from the primary ecological issues of the study area, select relevant variables that can represent these ecological issues as evaluation indicators, and then proceed with the vulnerability assessment. Furthermore, employing multiple evaluation methods for conducting vulnerability assessments often yields more scientifically sound and reliable conclusions compared to relying on a singular approach.

Limitations of this study

Vulnerability assessments inherently embody a degree of uncertainty. Firstly, vulnerability assessment indicators span various fields such as geography, ecology, and geology et al., leading to potential discrepancies when evaluated by experts from different disciplines who may choose divergent indicators and exhibit inconsistencies in calculating indicator weights. Secondly, some regional data are derived from remote sensing imagery or spatial interpolation, and while the quality of the data may be assured, errors in the original data are bound to influence computational outcomes. Lastly, vulnerability assessment results are typically validated solely through on-site investigations at multiple survey points, lacking dynamic validation through continuous monitoring data.

In future research, it is imperative to enhance the evaluation indicator system for ecological vulnerability. For instance, in regions situated within the agro-pastoral transitional zone of northern China, facing ecological pressures from agricultural cultivation, pastoralism, and mineral resource exploitation, considerations may include incorporating indicators such as grazing intensity or livestock numbers. Secondly, emphasis should be placed on validating the results of vulnerability assessments and forecasting future trends in regional ecological vulnerability.

Conclusion

This study applied AHP and PCA to analyze and evaluate the ecological geological vulnerability in Northern Shanxi Province based on the SRP model. The findings indicate that:

  1. (1)

    The average value of the EGVCI index by AHP and PCA for Northern Shanxi Province is approximately 3.96, with an overall mild vulnerability level of the regions. This suggests that for the entire region of Northern Shanxi Province, the overall vulnerability is not high, and ecological restoration efforts should primarily focus on natural recovery.

  2. (2)

    Considering the weightage ranking of indicators from both AHP and PCA, and the semi-arid to arid climate of Northern Shanxi Province, the driving factors of ecological geological vulnerability in the region are believed to be soil erosion intensity, mineral exploitation intensity, water conservation capacity, land use type, slope, and EVI.

  3. (3)

    Targeted ecological restoration and management recommendations are proposed for various ecological problem areas as follows: mining areas should strengthen the ecological restoration of historical abandoned mines, focusing on geomorphic restoration and enhancing vegetation coverage; eastern and southern mountainous areas should maintain regional ecosystem stability, improve regional water conservation capacity; western mountainous and hilly areas should strengthen construction of protective forest systems focusing on windbreak and sand fixation, soil conservation, and promote integrated management such as gully control and terrace construction in loess hilly areas, slope land conversion to forest, and sand fixation and grass planting in sandy areas; the central Datong Basin should control underground water extraction reasonably, focus on groundwater quality management, and saline-alkali land treatment.