Investigating the Dynamic Change and Driving Force of Vegetation Carbon Sink in Taihang Mountain, China

Abstract

Vegetation plays an important role in absorbing carbon dioxide and accelerating the achievement of carbon neutrality. As the ecological barrier of North China, the Taihang Mountains are pivotal to the ecological construction project of China. Nevertheless, the dynamic development of the vegetation carbon sink in the region and the impact factors on the sink have not been systematically evaluated. This study employed a comprehensive approach, utilising remote sensing technology and meteorological and topographic data, in conjunction with the net ecosystem productivity (NEP) estimation model to reveal the characteristics of vegetation carbon sinks in the Taihang Mountain, and then revealed the dynamics evolution of the NEP and the inter-annual trend by using Theil–Sen Median slope estimation, the Mann–Kendall test, and the coefficient of dissociation and analysed the driving roles of the influencing factors by using the parameter optimal geographic detector. Our findings suggest that the NEP in the Taihang Mountain area has a clear growth trend in time, the average value of NEP in the Taihang Mountain area is 289 gC-m⁻²-a⁻¹ from 2000 to 2022, and the spatial distribution shows the characteristics of high in the northeast and low in the middle and west, with a gradual increase from the northeast to the southwest; the areas with high fluctuation of NEP are mainly distributed in the areas around some cities that are susceptible to the interference of natural or anthropogenic factors. The vegetation carbon sinks in the Taihang Mountains are influenced by a variety of natural factors, among which the explanatory power of each natural factor is as follows: DEM (0.174) > temperature (0.148) > precipitation (0.026) > slope (0.017) > slope direction (0.003). The natural factor DEM had the strongest explanatory power for NEP changes, and the two-by-two effects of the natural factors on vegetation carbon sinks were all significantly stronger than the effects of a single factor, in which the interaction between DEM and precipitation had the strongest explanatory power; distinguishing from climate change factors, the contribution of anthropogenic activities to NEP changes in more than 90% of the area of the Taihang Mountainous Region was more than 60%, and the driving force of anthropogenic factors on NEP changes in the Taihang Mountainous Region was significantly stronger than that of natural climate change. The contribution of anthropogenic factors to NEP changes in the Taihang Mountains was significantly stronger than that of natural climate change. The results of this study can not only provide a reference for carbon reduction and sink increase and ecological restoration projects in the Taihang Mountains but also benefit the research paradigm of vegetation carbon sequestration in other regions.

1. Introduction

Global warming and climate anomalies represented by the continuous increase in greenhouse gas concentrations, such as CO_2, are threatening the global ecological environment and the economic and social development of mankind [1] and have become an important part of the research on environmental issues around the world. In recent years, China, as the world’s largest carbon emitter, has faced enormous non-traditional security challenges [2,3]. Under the double pressure of sustained and stable economic and social development and ecological environment protection, the Chinese government has carried out the important work of reducing costs and increasing efficiency, promoting the green transformation of economic and social development, and successively introduced a number of environmental policies [4]. In the report of the 20th Party Congress, it was clearly proposed to actively and steadily promote carbon peak and carbon neutrality, adhere to “first to establish and then to break”, implement the carbon peak action in a planned and step-by-step manner, and gradually shift to the “dual-control” system of the total volume and intensity of carbon emissions. In 2024, the National People’s Congress and the Chinese People’s Political Consultative Conference re-emphasised the need to strengthen the construction of ecological civilisation, promote green and low-carbon development, deeply implement the concept of “lucid waters and lush mountains are invaluable assets”, synergistically promote carbon reduction, pollution reduction, green expansion and growth, and build a beautiful China in which human beings coexist harmoniously with nature.

Based on this, the carbon cycle has become a central hotspot as an important element in the study of global climate and environmental change [5]. Terrestrial ecosystems, as an important component of the carbon cycle [6], have increasingly received attention from a wide range of researchers and scholars. As the most important carbon reservoirs in terrestrial ecosystems [7], vegetation and soil are important players in absorbing carbon dioxide and accelerating the achievement of carbon neutrality [8]. However, based on the heterogeneity of geographic environments, there is a wide variety of terrestrial vegetation types, the carbon sink capacity of different regions and vegetation varies significantly [9,10], and their carbon sequestration capacity is affected differently by the environment [11,12]. Therefore, in the process of promoting the global carbon balance, reducing carbon and increasing carbon efficiency, assessing and measuring the capacity of vegetation carbon sinks [13], analysing the spatial and temporal patterns and geographical characteristics of vegetation carbon sinks [9,14,15] and their driving mechanisms [16] are important components of the current carbon cycle research field.

In retrospect, the assessment of vegetation carbon sinks can be traced back to the middle of the last century. After decades of development, vegetation carbon sinks have deepened from initial gross primary productivity (GPP) studies to net primary productivity (NPP) estimates [17]. Many scholars usually use GPP and NPP as the characterisation of vegetation carbon sinks in their studies and analyse the spatial and temporal characteristics of carbon stocks and sinks in terrestrial ecosystems by estimating carbon sources as well as sinks and quantitatively evaluating the impacts of natural factors, including climate, and anthropogenic activities on vegetation carbon sinks [18,19]. Meanwhile, some studies have shown that the dynamics of terrestrial carbon sinks mainly originate from climate change and anthropogenic ecological restoration projects and that large-scale afforestation and plough-back projects will have a significant impact on carbon sinks and further promote the restoration of natural vegetation [20,21]. The interactions between climate change and anthropogenic factors (e.g., ecological restoration projects) complicate the evolution of vegetation carbon sinks [22]. In addition, with the deepening of research on vegetation carbon sink estimation, more and more scholars have used net primary productivity (NPP) and soil heterotrophic respiration to obtain a more accurate net ecosystem productivity (NEP) as an important indicator of ecosystem carbon sinks to characterise vegetation carbon sinks [18,23,24,25].

Overall, in terms of research objects, studies on vegetation carbon sinks are mostly based on a single vegetation type, such as woodland [11], grassland [26,27], wetland [28,29], etc., and most of the existing studies are still mainly focused on forestry carbon sinks [11,30], with the focus of the attention on forest ecosystems, but there is a lack of holistic and comprehensive studies on vegetation carbon sinks. In terms of research content, most of the studies on vegetation carbon sinks show the total amount of carbon sinks in the region [31], while there are fewer studies on the spatial and temporal distribution and evolution of carbon sinks, which inevitably ignore the heterogeneous impacts of spatial and geographic distribution [32,33], and only a single consideration of the role of natural factors on vegetation carbon sinks, ignoring the important impacts of anthropogenic factors. In terms of research scale, many studies have taken the provinces and cities of administrative divisions as research units [34,35] and also focused on river basins [7,36,37,38] and special geomorphological regions, etc. [39], while fewer studies have been carried out on large-scale natural ecological territories, especially the Taihang Mountainous Region, which is the core area of ecological functions in North China. In terms of research methods, remote sensing technology and information systems have been widely used in the dynamic study of vegetation carbon sinks, and many estimation models have been proposed in combination with national and regional meteorological data. The current research mainly applies pixel-by-pixel carbon sink calculation models using remote sensing technology, as well as widely used modified CASA models [40,41] and soil microbial respiration models [42]. Research on model methods is constantly iteratively improving with the deepening of vegetation carbon sinks, but it is not accurate enough and is not yet mature.

As a pioneering demonstration area for the value of ecological product services in North China, the Taihang Mountains are rich in ecological resources, but they are also a poor mountainous area with a high population density. Overall, the ecological environment is more fragile, and economic development is lagging behind. How to fully and rationally utilise the ecological resources of the Taihang Mountainous Region, convert the resources into capital in a sustainable and high-quality manner, continuously and steadily promote the regional carbon balance, and accelerate the realisation of carbon reduction and efficiency is a key issue in the process of regional high-quality development. As an “ecological barrier” in North China, the greening of the Taihang Mountains has also been proposed to be included in China’s 14th Five-Year Plan and will play a key role in achieving the goal of carbon neutrality. At present, the existing studies are more related to the biological and land resources in the Taihang Mountain area [43,44,45], and there are fewer studies on vegetation carbon sinks, and studies on the evolution characteristics and impact driving mechanisms of their vegetation carbon sinks are even less common [46]. Therefore, by using MODIS NPP data, meteorological data such as temperature and precipitation, and the soil microbial respiration model, this paper estimates the vegetation carbon sink in the Taihang Mountain area from 2000 to 2022 through the pixel-by-pixel calculation method with the help of remote sensing technology, aiming to deeply analyse the dynamic characteristics and driving mechanism of its spatio-temporal evolution. The findings of this study offer valuable insights that can facilitate the acceleration of the transition towards carbon neutrality and the establishment of an ecological civilisation within the Taihang Mountain region. Moreover, this study’s results can serve as a paradigm reference for similar initiatives elsewhere.

2. Materials and Methods

2.1. Study Area

The Taihang Mountains are located in North China (latitude 34°35′~40°19′ north, longitude 110°15′~116°27′ east), with an administrative area spanning four provinces and municipalities, namely Beijing, Hebei, Shanxi, and Henan (Figure 1). The mountains extend in the direction of northest–southwest, until the south of Lingchuan County, Shanxi Province, turning east–west, slightly in the shape of “S”. The ridge extends along the border of the three provinces, which is the natural boundary mountain of the Shanxi, Hebei, and Henan provinces, and at the same time, as the natural barrier and geographic demarcation line between the North China Plain and the Loess Plateau, it blocks the cold air and sand from the northwest for the North China Plain and plays a key role in the ecological climate of the whole North China region.

The southeast side of the Taihang Mountains is influenced by the Pacific warm and humid airflow, with high temperature and high precipitation, while the northwest side is influenced by the dry and cold airflow from the northwest, with obvious continental climate, so the northern part of the region has a cooler climate and less precipitation and belongs to the warm-temperate zone with semi-dry and early climate. The average annual precipitation in the Taihang Mountains is around 600 mm, and the average annual temperature is about 10 °C. Its topography is complex and diverse, including canyons, cliffs, peaks, and so on. Habitat conditions are complex and varied, and vegetation types are abundant, among which cultivated vegetation, woodland, and grassland occupy more than 90% of the total area. As shown in Figure 2, the terrain at the periphery of the Taihang Mountain contour is low, so the main vegetation types are dominated by cultivated vegetation, while grassland meadows, coniferous forests, broad-leaved forests, and other vegetation types are mainly concentrated in the interior of the Taihang Mountain area, which is affected by geographic topography as well as DEM.

2.2. Data Sources

The data for the study in this paper mainly include NPP, meteorological, topographic, land use, and other related data (

Table 1. Research data and sources.

Data Types	NPP	Meteorological Data	Topographic Data	Land Use Data
Data Sources	MODIS Terra NPP data from the NASA Data Centre1 (MOD17A3HGF)	Temperature and precipitation data from the National Earth System Science Data Centre2, with a temporal resolution of 1 m and a spatial resolution of 1 km, both spanning the period 2000–2021.	Digital Elevation Model (DEM) data from NASA’s SRTM with a spatial resolution of 30 m; slope gradient and slope aspect data calculated based on the elevation data.	Derived from three phases of land use data for 2000, 2010 and 2020 at a resolution of 1 km from the Centre for Resource and Environmental Science and Data of the Chinese Academy of Sciences.

).

Among them, the MRT (MODIS reprojection tool) tool was applied for the stitching, format conversion, projection transformation, and resampling of NPP data, and then, ArcGIS 10.8 was used to perform operations such as image cropping and the removal of invalid values, which, in turn, resulted in the annual NPP dataset of the study area, with a spatial resolution of 500 m, a temporal resolution of 1 a, and a time span of 2000~2022. Meteorological data were interpolated from station data to raster data using ArcGIS 10.8 software using the inverse distance weight interpolation method (IDW).

In addition, all the data were uniformly projected to the WGS1984/UTM 50N coordinate system, and the spatial resolution was resampled to 1 km, which facilitates the subsequent processing and analysis of the data.

2.3. Model and Method

2.3.1. NEP Estimation Model

Net primary productivity (NPP) is the portion of photosynthetic products fixed by plant photosynthesis minus the portion consumed by the plant’s own respiration. While the ecosystem carbon sink also includes the respiration of soil organic matter, the net ecosystem productivity (NEP) is the net carbon produced by terrestrial ecosystems, which serves as an important indicator of the terrestrial ecosystem carbon sink. When the influence of various disturbance factors is ignored, NEP is used as the net carbon amount of the terrestrial ecosystem to characterise a carbon sink. Its calculation formula is as follows:

$$NEP(x,t)=NPP(x,t)-R_h(x,t)$$

(1)

where $NEP$ denotes the net ecosystem productivity of vegetation, $x$ denotes the picture element, $t$ denotes time, and $t$ is measured in years. Similarly, $NPP(x,t)$ represents the net primary productivity of vegetation in picture element $x$ in year $t$. When NEP(x,t) > 0, it means that the carbon fixed by vegetation is greater than the carbon emitted by soil microbial respiration, which manifests as a carbon sink, and vice versa. Soil microbial respiration estimation models have been widely used in the calculation of regional vegetation carbon sinks. $R_h$ indicates the consumption of microbial respiration in soil. Therefore, the estimation model of soil microbial respiration was also used in this study with the following formula:

$$R_{\mathrm{h}}(x,t)=0.22(Exp(0.0913T(x,t))+\ln (0.3145R(x,t)+1))\times 30\times 0.465$$

(2)

where $T(x,t)$ and $R(x,t)$, respectively, represent the average temperature and precipitation of picture element $x$ in year t.

Through comparative analysis with similar studies on carbon sinks in the Taihang Mountains, the NPP data used in this paper based on the MODIS Terra NPP data (MOD17A3HGF) provided by the NASA Data Center are in the same range as some other scholars’ estimated NPP data. In addition, the spatial distribution of vegetation carbon sinks in the Taihang Mountains estimated by many scholars using the same model is high in the middle west and low in the southeast, and the decreasing trend from west to east is basically the same as the results obtained from the analysis of MODIS data. Therefore, the data used in this paper for the study of vegetation carbon sinks in the Taihang Mountains have high credibility.

2.3.2. Analysis of Inter-Annual Trends

After estimating the NEP of the Taihang Mountain area in each year, the inter-annual trend in NEP in the Taihang Mountain area from 2000 to 2022 was analysed by calculating the median in the long time series through Theil–Sen Median slope estimation and the Mann–Kendall test. Theil–Sen Median regression is a robust linear regression method used to minimise the effect of outliers on the fitting results. Theil–Sen Median regression uses a statistic called median slope for parameter estimation, which provides better resistance to outliers and improves the robustness of the regression model. The specific formula is as follows:

$$Slope=median({\displaystyle \frac{NE{P}_j-NE{P}_i}{j-i}}),2000<i<j<2022$$

(3)

where Slope represents the trend in inter-annual changes in NEP, and ${NEP}_i$ and ${NEP}_j$ are the NEP values for year i and year j, respectively. When Slope > 0, NEP has an upward trend and vice versa. While Mann–Kendall, as a non-parametric test of trends in time series, does not require the measurements to obey the normal distribution, subject to the influence of missing values and outliers, it is suitable for the trend significance test of long time series data, can effectively reduce the influence of nulls, and can be used for the significance test of the trend in the NEP within the study period.

$$\mathrm{S}={\displaystyle {\sum }_{\mathrm{i}=1}^{\mathrm{n}-1}}{\displaystyle {\sum }_{\mathrm{i}=\mathrm{i}+1}^{\mathrm{n}}}\mathrm{sign}\left({NEP}_j-{NEP}_i\right)$$

(4)

$$\mathrm{s}\mathrm{i}\mathrm{g}\mathrm{n}\left({\mathrm{N}\mathrm{E}\mathrm{P}}_{\mathrm{j}}-{\mathrm{N}\mathrm{E}\mathrm{P}}_{\mathrm{i}}\right)=\left\{\begin{array}{cc}1\hfill & {(\mathrm{N}\mathrm{E}\mathrm{P}}_{\mathrm{j}}-{\mathrm{N}\mathrm{E}\mathrm{P}}_{\mathrm{i}}>0)\hfill \\ 0\hfill & {(\mathrm{N}\mathrm{E}\mathrm{P}}_{\mathrm{j}}-{\mathrm{N}\mathrm{E}\mathrm{P}}_{\mathrm{i}}=0)\hfill \\ -1\hfill & ({\mathrm{N}\mathrm{E}\mathrm{P}}_{\mathrm{j}}-{\mathrm{N}\mathrm{E}\mathrm{P}}_{\mathrm{i}}<0)\hfill \end{array}\right.$$

(5)

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

(6)

$${\mathrm{Z}}_{\mathrm{s}}=\left\{\begin{array}{c}{\displaystyle \frac{\mathrm{S}-1}{\sqrt{\mathrm{v}\mathrm{a}\mathrm{r}\left(\mathrm{S}\right)}}},S>0\hfill \\ 0,S=0\hfill \\ {\displaystyle \frac{\mathrm{S}+1}{\sqrt{\mathrm{v}\mathrm{a}\mathrm{r}\left(\mathrm{S}\right)}}},S<0\hfill \end{array}\right.$$

(7)

In the above equation, refer to the NEP values for each year from 2000 to 2022, respectively, as a function of the variance of the random variable. The Mann–Kendall method was used to test the significance of the NEP trend. If the standardised value Z is greater than 0, the series has an upward trend; if it is less than 0, the series has a downward trend. When the absolute value of Z is greater than or equal to 1.64, 1.96, and 2.58, it means that the time series passes the significance test at the confidence level of 90%, 95%, and 99%, respectively.

The discrete coefficients can reflect the degree of dispersion of the observations more intuitively and eliminate the scale, which is convenient for objective comparison. The stability of vegetation carbon sinks in Taihang Mountains from 2000 to 2022 is shown by the discrete coefficient in terms of picture elements. The specific formula is as follows:

$$C_v={\displaystyle \frac{\sqrt{\frac{1}{n}{\displaystyle \sum _{i=1}^n{(NE{P}_i-\overline{NEP})}^2}}}{\overline{NEP}}}$$

(8)

where $NE{P}_{\mathrm{i}}$ denotes the vegetation carbon sink in year $i$, and $\overline{NEP}$ represents the average value of the vegetation carbon sink of each picture element. The smaller the value of the dispersion coefficient $C_v$, the smaller the fluctuation of $NEP$ over time, and the more stable the vegetation carbon sink is; on the contrary, the larger the fluctuation is, and the lower the stability is.

2.3.3. Optimal Parameter-Based Geographic Detector

Geographic detectors, as a statistical method for detecting the effects between factors in multiple spatial units, are widely used in detecting spatial dissimilarities and identifying interactions between factors [47,48]. However, the analysis of traditional geographic detector models is often affected by spatial scale and zoning effects, and the results are often uncertain due to the spatial scale dependence of the process mechanism [49]. Song et al. [50], in order to avoid spatial scale dependence and reduce the influence of zoning effects, established a geographic detector model based on optimal parameters, which can better identify the relevant combinations of spatial scale parameters, greatly reduce the influence of spatial differentiation, and more accurately explain the correlation relationship between variables. Therefore, in this paper, based on parameter optimisation, the driving factor of the spatial divergence of NEP in the Taihang Mountain area is detected by a geo-detector.

The factor detection can detect the degree of influence of the driving factor X on NEP, and the result is measured by q. The specific formula is as follows:

$$q=1-{\displaystyle \frac{1}{N{\sigma }^2}}{\displaystyle \sum _{h=1}^L{N}_h{\sigma }_h^2}$$

(9)

where $L$ is the stratification of variables, that is, classification or partitioning; $N$ and $N_h$ represent the number of units in the entire region and the h-th layer, respectively; ${\sigma }^2$ and ${\sigma }_h^2$ represent the discrete variances of vegetation carbon sinks in the entire region and the h-th layer, respectively. The value of $q$ is between 0 and 1. The larger q is, the stronger the explanation of NEP.

In the actual region, the vegetation carbon sink is not a natural factor acting alone, but often multiple factors acting together, such as temperature and precipitation, precipitation slope direction, slope elevation, and other factors closely linked and acting together. Therefore, in addition to factor detection, this paper further detects factor interactions and assesses whether the influence on vegetation NEP increases or decreases when two closely related driving factors act together or whether the effects of these factors on NEP are independent of each other.

This was assessed by first calculating the q-values of the two factors X1 and X2 on Y: q(X1) and q(X2), respectively, and by calculating the q-values when they interacted (new polygonal distributions formed by a tangent of the two layers of the superimposed variables X1 and X2): q(X1∩X2) and comparing q(X1), q(X2) with q(X1∩X2).

2.3.4. Multiple Regression Residual Analysis

Vegetation carbon sinks are not only affected by natural factors, but the role of anthropogenic activities cannot be ignored. On the basis of the driving analysis of natural factors through a parameter-optimal geographic detector, we then used multiple linear regression and residual analysis to study the impacts and relative contributions of climate change and human activities on NEP changes. The specific method is to use temperature, precipitation, and other data and regression model parameters to calculate the predicted value of NEP and use this predicted value to characterise the relative contribution of climate change to NEP. Using the NEP estimation model, the net ecosystem productivity (NEP) of vegetation in the Taihang Mountains obtained by Equation (1) is the result of the joint action of climate change and anthropogenic activities, so the difference between the NEP calculated by Equation (1) and the predicted value of climate change on the NEP is the impact of anthropogenic activities on the NEP.

With NEP as the dependent variable and temperature and precipitation as the independent variables, a linear regression model is established with the specific formula:

$$NE{P}_{cc}=a\ast temp+b\ast prec+c$$

(10)

where $a$, $b$, $c$ are regression model parameters, and $temp$ and $prec$ are average temperature and cumulative precipitation, respectively.

The temperature, precipitation data, and regression model parameters are used to calculate the NEP prediction ($NE{P}_{cc}$), which is used to represent the impact of climate change on NEP. The difference between NEP and NEP prediction, which is the NEP residual ($NE{P}_{ha}$), is calculated to represent the impact of human activities on NEP. The specific formula is as follows:

$$NE{P}_{ha}=NEP-NE{P}_{CC}$$

(11)

2.4. Selection of Impact Factors

Vegetation carbon sinks are affected by a variety of factors. Topography is an important factor influencing carbon sequestration by vegetation as well as the pattern of spatial differentiation, with significant differences in water and heat distribution between different topographic conditions. In addition, climatic conditions, such as precipitation and temperature, directly affect vegetation growth and distribution. Based on the relevant literature and previous studies and the specific conditions of the Taihang Mountains, we selected five natural factors, namely, temperature, precipitation, DEM, slope, and slope aspect, based on the availability of data, and analysed their impacts on vegetation carbon sinks by combining them with the optimal geographical detector. In addition to natural factors, human activities will also have an important impact on vegetation carbon sequestration, so on the basis of natural factor-driven analyses, further research on anthropogenic factors using multiple regression residual analysis to analyse their impacts should be conducted.

3. Results

3.1. Spatial and Temporal Characteristics of NEP

The mean value of the annual vegetation carbon sink in the Taihang Mountains was obtained by using the NEP estimation model on a term-by-term meta-calculation basis, and the mean value of the vegetation carbon sink in the Taihang Mountains from 2000 to 2022 was about 289 gC·m⁻²·a⁻¹. As shown in Figure 3, the contribution of vegetation carbon sinks in the Taihang Mountains from 2000 to 2022 is mainly concentrated in the large mountainous areas in the central and western parts of the country, with an area of more than 300 gC·m⁻²·a⁻¹ accounting for 75% of the total area, and the mean value of the NEP can reach a maximum of 658.408 gC·m⁻²·a⁻¹, which is located mainly in the mountainous regions of Xinzhou, Datong, and Shuozhou in Shanxi and in the mountain ranges of Jinzhong, China. Combined with the topographic and geographical characteristics, the region has a high topography, and the vegetation is mainly dominated by woodland grass status. In the western part of the Taihang Mountains, the mean value of vegetation carbon sinks is low, and the average value of NEP can be as low as −39.7872 gC·m⁻²·a⁻¹, which exists in the form of carbon sources. This is mainly due to the fact that the western Taihang Mountains, including Shijiazhuang, Baoding, and Xingtai, are low-lying and are part of the Hebei plains, where cultivated vegetation, such as farmland and arable land, occupies a major area, and its contribution to carbon sinks is relatively low compared to that of woodland and grassland. Overall, the vegetation carbon sinks in the Taihang Mountains show a spatial pattern of high in the centre-west and low in the northeast, gradually increasing from the northeast to the southwest. Combined with geographic features, regional elevation and vegetation type were found to be important factors in the spatial heterogeneity of the vegetation carbon sink.

From the time series of NEP in the Taihang Mountains (Figure 4), the mean value of vegetation carbon sinks increased significantly from 222.19 gC·m⁻²·a⁻¹ in 2000 to 357.78 gC·m⁻²·a⁻¹ in 2022. Especially before 2005, the growth trend in NEP was obvious, thanks to important measures such as China’s vigorous reforestation programme and the construction of nature reserve projects at the end of the last century. Since then, economic development has been running smoothly, alongside environmental protection. Along with the “lucid waters and lush mountains are invaluable assets” concept, “ecological civilisation construction”, “Carbon Peaking and Carbon Neutrality Goals”, and other strategies that have been put forward one after another, the carbon sink of the vegetation in the Taihang Mountain area has also reached a new peak. Taken together, during these 23 years, the vegetation carbon sink in the Taihang Mountains as a whole has shown a fluctuating upward trend, and the capacity of vegetation carbon sequestration has continued to rise steadily.

3.2. The Changing Trend in NEP

Since 2000, the spatial and temporal changes in vegetation carbon sinks in the Taihang Mountains have been obvious. Over time, the vegetation carbon sink in the Taihang Mountains as a whole shows an increasing trend (Figure 5). Among them, the area of regions with an average vegetation carbon sink value of more than 300 gC·m⁻²·a⁻¹ increased from 20% in 2000 to 60% in 2022, especially Zhangjiakou, Taiyuan, Linfen, and Jincheng, which had the most obvious changes in growth, indicating that the growth environment of vegetation in those regions had been improved, the types of vegetation had become more abundant, and the amount of carbon sequestered by vegetation had been significantly increased. In addition, the area of low-carbon sinks in the eastern Taihang Mountains has been shrinking, with more than 50% of the area with an average vegetation carbon sink value of less than 250 gC·m⁻²·a⁻¹ in 2000, and less than 20% of the area with low-carbon sinks by 2022. Over time, the vegetation carbon sinks in Fangshan in Beijing and Baoding and Xingtai in Hebei within the Taihang Mountains region have increased significantly, while other parts of the region, including Shijiazhuang, Handan, Anyang, and Xinxiang, are influenced by geographic and topographical factors, and the vegetation types are mostly cultivated vegetation, with plains, farmland, and cropland dominating, so that this part of the region contributes to the carbon sinks to a relatively low degree.

Based on the Theil–Sen Median slope estimation, the inter-annual trend and spatial distribution of significance of vegetation carbon sinks in the Taihang Mountains were obtained from 2000 to 2022 (Figure 6), and the proportion of areas with positive NEP values accounted for more than 80% of the total area, indicating that the vegetation carbon sinks in the Taihang Mountains have shown an overall increasing trend over the 23 years. Among them, the areas with significant inter-annual trends were concentrated in the northeastern and western mountain range areas of the Taihang Mountains. The Theil–Sen Median analysis and MK test results were superimposed into five categories (

Table 2. Table of Mann–Kendall test results.

Slope	|Z|	Trend Type	Trend Feature
Slope > 0	≥2.58	2	Significantly increased
	≥1.96	1	increased
	<1.96	0	Stable and unchanged
Slope < 0	≥1.96	−1	decreased
	≥2.58	−2	Significantly decreased

), Based on the Mann–Kendall test to visualise the inter-annual trend of significantly changing NEP, it is more intuitive to obtain the trend of significant increase in the overall NEP in the Taihang Mountainous Region, and almost the whole region of the vegetation carbon sink had a significant growth trend; only a small part of the peri-urban areas bordering Beijing, Shanxi, Hebei, and Henan provinces and cities were significantly affected by the anthropogenic factors, and the fluctuation of NEP was large, the stability was poor, and the trend of the decrease was more significant.

The spatial distribution of the coefficient of dispersion of vegetation carbon sinks is shown in Figure 7, with the exception of a few areas, all of which are less than 1. The proportion of areas with coefficients of dispersion of less than 0.1, 0.1–0.3, and 0.3–0.5 in the total area of the vegetation area is 5.9%, 91.6%, and 1.9%, respectively, which suggests that the overall fluctuation of the region’s vegetation carbon sinks over time is relatively small, and the stability of the region is relatively high. In terms of spatial distribution, the coefficient of dispersion of vegetation carbon sinks is high in the northeast and low in the west-central direction, and the high fluctuation areas are scattered, mainly in the border of Beijing, Shanxi, and Hebei provinces, the junction of Shijiazhuang, Jinzhong, and Yangquan, and the periphery of Jincheng, Xinxiang, and Jiaozuo, which have a low stability of vegetation carbon sinks and are vulnerable to the influence of natural or human activities. By comparing Figure 3 and Figure 7, it can be seen that these regions with lower stability also have lower NEP, and there is no obvious trend of reversal to a positive direction in the spatial and temporal evolution.

3.3. Driving Analysis of Influencing Factors

3.3.1. Impact Analysis of Natural Factors on NEP

From the results of single factor detection (Figure 8), the degree of influence of each natural factor on vegetation carbon sinks from 2000 to 2022 is as follows: DEM (0.174) > temperature (0.148) > precipitation (0.026) > slope (0.017) > slope aspect (0.003). The p-values of the models are all less than 0.01, i.e., the results are all significant at the 0.01 level. Among them, DEM has the strongest explanatory power for the spatial and temporal distribution characteristics of vegetation carbon sinks, followed by temperature and precipitation, and the least explanatory power for topographic slope aspect. Vegetation carbon sinks show great differences depending on the spatial distribution of vegetation, and DEM is an important factor that directly affects the spatial distribution of vegetation; therefore, DEM directly affects the habitat of vegetation and then affects the spatial distribution of vegetation carbon sinks. Suitable temperature and precipitation, on the other hand, are important bases for the survival and growth of vegetation. The q-value of precipitation is small, 0.026, and in combination with the climatic characteristics and vegetation type characteristics of the Taihang Mountainous Region, the climate in North China is drier, the average annual rainfall is smaller, and the vegetation in this region is also mainly dominated by grassland meadows and coniferous forests. In addition, the explanatory power of slope gradient and slope aspect on vegetation carbon sinks is weak. As the slope increases, the more difficult it is for water to collect, the higher the risk of erosion to the vegetation, and the stronger the erosion and scouring of the soil by the water flow, which will affect the coverage of the vegetation. Slope aspect mainly affects the soil texture by influencing the light time and then indirectly affects the growth of vegetation, which has a weaker effect compared with other factors.

From the detection results of the interactions among natural factors, the explanatory power of natural factors on vegetation carbon sequestration is significantly stronger than that of single factors in both pairwise interactions. As shown in Figure 9, the NEP natural factor interactions basically show a non-linear enhancement relationship, i.e., the explanatory power of the pairwise interaction of influencing factors on NEP is greater than the sum of explanatory power of single factor effects. For example, the explanatory power of the non-linearly enhanced interaction of temperature and precipitation is 0.299 (Figure 10), which is greater than the sum of the explanatory powers of the single-factor effects of 0.148 and 0.026 (Figure 8). In addition, the interactions of the two sets of factors, DEM and air temperature and DEM and slope, show a bilateral enhancement relationship, i.e., the explanatory power of the pairwise interaction of influencing factors on NEP is greater than the explanatory power of each single factor. For example, the explanatory power of the bilaterally enhanced interaction of temperature and DEM is 0.217 (Figure 10), which is greater than the one-factor explanatory power of 0.148 and 0.174 for temperature and DEM.

Combining Figure 8 and Figure 10, it is found that the vegetation carbon sinks in the Taihang Mountains are influenced by a variety of natural factors and that this influence is not determined by a single factor alone but rather by the combined effects of multiple factors. The interaction between DEM and temperature and DEM and slope is a bilaterally enhancement relationship, indicating that when the regional DEM is supplemented with suitable temperature and slope, it will be more conducive to the positive growth of the vegetation carbon sink. The interactions between DEM and precipitation and temperature and precipitation show a significant and strong non-linear enhancement, with explanatory powers of 0.308 and 0.299 for the vegetation carbon sink, respectively. In addition, there are strong correlations between the natural factors shown in Figure 10; especially, the correlation between precipitation, air temperature, and DEM is more obvious. Although the interaction between slope aspect and other natural factors is a non-linear enhancement relationship, its correlation is weak, and because its own explanatory power on vegetation carbon sinks is only 0.003, this also makes the combined explanatory power of the interactions with the other factors not strong.

Overall, the influence of natural factors on vegetation carbon sinks is more complex, manifesting multiple interactions of factors. Based on the analysis of the parameter-optimal geographic detector, NEP is mainly affected by the interaction of DEM with precipitation and temperature, and DEM plays a dominant role in the single-factor interaction, while the interaction of DEM, temperature, and precipitation acting as a dominant factor in affecting the vegetation carbon sinks when the interaction of the factors is explored. The reason for this is that the Taihang Mountains are located in northern China, where the climate is dry and the region is relatively arid, so an increase in precipitation has a more significant effect on the growth of vegetation when DEM and air temperature are appropriate.

3.3.2. Analysis of Climate Change and Human Activity Drivers of the NEP

Based on Equation (3), the inter-annual trend in vegetation carbon sinks in the Taihang Mountainous Region from 2000 to 2022 can be estimated, and combined with the residual analysis of multiple regression, Equations (10) and (11) are used to calculate the linear trend rate of $NE{P}_{cc}$ and $NE{P}_{ha}$. Its value Slope > 0 indicates that climate change or human activities have a facilitating effect on vegetation carbon sinks, and vice versa, there is an inhibiting effect. In order to better evaluate the driving effect of the two factors on NEP, the effects were classified into six levels in this paper (

Table 3. Grading of impacts of climate change and human activities on NEP.

Slope (NEP)	≤−2	−2–0	0–2	2–4	4–6	≥6
Degree of Impact	Moderate inhibition	Slight inhibition	Slight contribution	Moderate contribution	Clear-Cut contribution	Extremely obvious contribution

).

Further visualising the extent of their impacts yields Figure 11, where the effects of climate change and human activities on NEP in the Taihang Mountain area vary considerably and show great spatial heterogeneity. Combined with Figure 11, about 72.8% of the areas in the Taihang Mountains show that climate change has a slight contribution to the change in NEP impacts, which can also be expressed as an insignificant effect; about 20% of the areas show that climate change has a moderate contribution to NEP, which is mainly concentrated in Taiyuan and Jinzhong of Shanxi Province, which have higher terrain and are subject to the significant effect of DEM factors. In contrast to climate change, the impact of anthropogenic activities on NEP is significantly different and is basically “Lean to One Side”. More than 95% of the area shows a moderate to strong contribution to NEP, and of these, more than 50% of the total area significantly and very significantly contribute, especially in the northern and western parts of the Taihang Mountains, where anthropogenic impacts are extremely pronounced.

By dissecting the hierarchical spatial distribution of the impacts of climate change and human activities on NEP, further differentiating the dominant drivers of NEP changes in the Taihang Mountains, calculating and visualising the contribution of climate change and human activities to NEP changes, i.e., using NEP_cc and NEP_ha, respectively, and comparing them with the values of carbon sinks of vegetation obtained from the NEP-based estimation model, we obtained the spatial distribution of the contribution rates shown in Figure 12.

Overall, the contribution of climate change to NEP changes in the Taihang Mountains is much smaller than that of human activities. The contribution of climate change to NEP changes in the Taihang Mountains is concentrated in the 0–40% range, which accounts for more than 90% of the total area of the Taihang Mountains, with 51.6% and 42.5% of the area in the 0–20% and 20–40% ranges, respectively, contributing to climate change. In addition, the area with a climate change contribution rate of more than 60% is very small, less than 1%, and more scattered, mainly located in Yangquan, Jincheng, and Jiaozuo.

Distinguishing from climate change factors, the contribution of anthropogenic activities to NEP changes within more than 90% of the Taihang Mountains area is more than 60 per cent, indicating that the contribution of NEP changes is basically driven by anthropogenic activities. Among them, the areas with a contribution rate of more than 80% are mainly concentrated in the northern part of the Taihang Mountains, a large area of the mountain range line; these areas have high altitude and high slope, and the vegetation type is basically dominated by grassland meadows and coniferous forests, etc. And the effect of large-scale afforestation and ecological restoration projects by human beings is extremely effective in influencing the role of the NEP.

Focusing on the Taihang Mountains, compared with the influence of natural factors, anthropogenic activities play a more significant role in driving vegetation carbon sinks, and anthropogenic activities drive changes in vegetation carbon sinks mainly through changes in the type of vegetation or land use in the region, which indirectly affects the amount of vegetation carbon sinks. The spatial and temporal evolution of land use types in the region is also an important perspective for analysing the impacts of anthropogenic activities in the Taihang Mountains, including large-scale afforestation, desertification control, and other ecological restoration projects. Comparing the spatial distribution of land use and cover types in the Taihang Mountains at ten-year intervals (Figure 13 and Figure 14), the overall change in vegetation types is not obvious enough, but the area of urban land use has increased over the past 20 years, mainly in the areas bordering Beijing, Shanxi, and Hebei Provinces, the junction of Shijiazhuang, Jinzhong, and Yangquan, and the periphery of Jincheng, Xinxiang, and Jiaozuo, etc. The increase in urban land use has deepened the impact of human activities on NEP, resulting in the low stability and high volatility characterising NEP changes in the region.

4. Discussion

4.1. Characteristics of Spatial and Temporal Evolution of Vegetation Carbon Sinks

This study reveals the spatial and temporal evolution of the vegetation carbon sinks and the inter-annual trend in the vegetation carbon sinks in the Taihang Mountainous Region from 2000 to 2022. The mean value of the vegetation carbon sinks showed a significant increase in more than 20 years, and the vegetation carbon sinks were more fluctuating, especially over time, which is basically in line with the previous studies by some scholars [16,51,52,53]. In addition, this result is also consistent with some other scholars’ studies on the spatial and temporal evolution of vegetation carbon sinks in other regions, such as the Yellow River Basin [38], Northwest China [10,54], and Southwest Karst Landscape [20,39]. In addition, Zhang et al. emphasised that terrestrial ecosystems experienced a shift from carbon sources to sinks accompanied by an increase in carbon stocks [55], further corroborating the validity and accuracy of the findings of this study. In addition, the Taihang Mountain area is a temperate region geographically, but its vegetation carbon sequestration capacity is increasing, and the vegetation carbon sink increasing trend is more consistent with the tropical Hainan region, subtropical Fujian, and Southwest Yunnan, and other regions. Lei (2024) and other scholars have studied the tropics in the Hainan region for the primary productivity of vegetation carbon sinks. In this region, the carbon sequestration capacity of vegetation was stronger, showing an overall fluctuating increasing trend, but the spatial difference was obvious, showing a spatial distribution pattern of decrease from the central and southern part of the country to the surrounding coastal areas [56,57]. Xu (2022) and other scholars, on the other hand, analysed the primary productivity of vegetation carbon sinks for the subtropical southern Minnan region, which showed an increasing trend in time and was significantly different from the Taihang Mountainous Region in space, relying on the Yangtze River Basin and the Minjiang River Basin, which showed the characteristics of a river as the main trunk and decreasing to the periphery [58,59]. And distinguishing from tropical and temperate regions, many northern regions of China, such as northern Inner Mongolia, are close to the frigid zone, with less grassy vegetation, and the sequestration capacity of their vegetation carbon sinks is significantly lower [60]. But the northern part of Heilongjiang, despite its high dimensionality, has a vast area of forests, and the sequestration capacity of its indicator carbon sinks is similar to that of the temperate tropics.

Combined with Figure 4, the analysis of the growth rate of vegetation carbon sinks reveals that the trend of its growth has become slower and slower in recent years, which suggests that the carbon sequestration potential of vegetation in the region may be difficult to increase significantly in the short term over time. Similar conclusions have been obtained by Xu and Lyu et al. [7,20], and reasonable explanations have been given for the regional vegetation types and land use changes. In addition, in 2001, the northern part of China suffered from drought, and the NEP value of the study area was low; in 2003, the study area received abundant precipitation, which resulted in a high NEP value and improved the vegetation growth. In the three years after 2004, the Taihang Mountains experienced a series of natural disasters, especially severe floods and torrential rains, and Jincheng City in Shanxi Province suffered from severe forest fires during this period, which resulted in huge losses of vegetation and wildlife, leading to the fluctuation and decline of NEP in the three years after 2004. Combined with Figure 13 and Figure 14, various land cover types have tended to be stable, and vegetation such as grassland meadows, coniferous forests, and bushes in the central and western parts of the region have limited contributions to carbon sinks, while farmland and arable land in the western part of the Taihang Mountains are affected by anthropogenic factors, and their contributions to carbon sinks are also relatively stable; in some areas, they even exist in the form of carbon sources.

Some low-stability and high-volatility carbon sink areas are mostly distributed around cities (Figure 7), which are more susceptible to anthropogenic activities, as confirmed by the studies of many scholars, and in the Loess Plateau [61], the Northwest Arid Zone [54], and the Southwest Mountainous Region [39,62]. Anthropogenic activities play a major role in the changes in vegetation carbon sinks. Combined with the geomorphological features and ecosystem characteristics, Zhou et al. [61] proposed that human activities have the greatest influence on the time scale changes in NEP in the Loess Plateau and each ecological subregion, with relative contributions greater than 50%. On the time scale, ecological restoration projects in the region have been continuously conducted. In addition, part of the southwestern mountainous areas have a special karst landscape; Zhu et al. suggested that these areas have low ecosystem stability, are ecologically more fragile, and have less stable vegetation carbon sinks, which are more susceptible to the impacts of human activities [39,62]. Based on the spatial and temporal changes in NEP, targeted management strategies are needed. It is suggested that, based on the relevant geographical and topographical characteristics, the Chinese government should take the lead in establishing vegetation nature reserves in these low-stability and high-volatility areas and implementing the return of farmland to forests, so as to achieve a key breakthrough in the utilisation of carbon sinks.

4.2. Driving Analysis of Carbon Sink Factors

The results of the attribution analysis of natural factors showed that the positive correlation between NEP change and DEM was the strongest, NEP was not determined by a single factor only but by the complex effect of multiple factors, and DEM supplemented by precipitation and temperature was the main driver of NEP change in the Taihang Mountainous Region. The results of some scholars were basically the same, although there were differences in the study areas [34]. In contrast, studies on regional NEPs such as the Minjiang River [63] and the Weihe River [7] have yielded different results, in which factors such as precipitation and land use type change act as the dominant factors for changes in vegetation carbon sinks and have even concluded that vegetation carbon sinks show a certain negative correlation with temperature, emphasising that elevated temperatures inhibit the increase of vegetation carbon sinks [7,36,37,64,65]. There is significant heterogeneity in the role of vegetation carbon sink changes in different regions of DEM, temperature, and precipitation, and Wang et al. [66], in their study on the dynamics of phenology on the Tibetan Plateau, proposed that the spatial and temporal response pattern of primary net productivity of vegetation is controlled by local climatic and topographic conditions. Zhao et al. [67] studied the carbon sink potential of vegetation communities in temperate monsoon climatic zones and proposed that factors such as temperature and precipitation are kept in the average range and that the composition and diversity of species are the most important factors influencing the potential of carbon sinks. In addition, scholars such as Wei [16] and Yang [68], in their studies on vegetation carbon sinks in Shaanxi and Gansu, suggested that the effect of precipitation is more significant than that of air temperature and that vegetation carbon sinks show a bimodal pattern of change with differences in elevation. Based on this, the degree of influence of each natural factor on vegetation carbon sinks varies significantly in some plateau areas, temperate monsoon climate plains, and some arid areas in Shaanxi and Gansu, and the dominant natural factors show heterogeneity due to different topographic and environmental conditions in the localities.

In addition to the influence of climate change, which includes natural factors such as temperature and precipitation, the influence of anthropogenic factors is more strongly expressed. This is basically consistent with the results of Xu et al. [20] on the carbon sink drivers in the karst region of southwest China. In regions with obvious traces of human activities, the impact of climate change on the dynamics of vegetation carbon sinks is much smaller than that of human activities. However, some scholars have come to the opposite conclusion: Zhou et al. [61] explored the spatial variation in NEP on the Loess Plateau and concluded that the spatial distribution is strongly influenced by meteorological factors, while Ge et al. [51] and Yan et al. [52] concluded that the NEP in parts of mainland China and the northern grassland areas are more affected by climate change than anthropogenic factors. Through a more detailed analysis of the causes of the differences, Dong et al. [69] analysed the influencing factors of terrestrial carbon sinks on the Tibetan Plateau and suggested that anthropogenic factors have a weaker influence on NEP changes than natural factors and that the differences are due to more unique regional scales and geographic environmental conditions. In addition, Li and Zhou et al. [21,61] suggested in their study of vegetation carbon sinks in the Loess Plateau that the Loess Plateau region suffers from severe soil erosion and that the instability of climatic factors, especially precipitation and solar radiation, tends to trigger a negative effect on carbon sinks, and climatic factors have a greater influence on vegetation growth at the spatial scale. You and Yan et al. [52,70], in their studies on the carbon sinks of grassland vegetation in western Inner Mongolia, suggested that the vegetation types of northern grasslands are relatively homogeneous and that the effects of temperature and precipitation on vegetation growth are stronger than those of anthropogenic activities. Xu et al. [20] detailed the important impacts of human activities such as large-scale afforestation and ecological restoration projects in their study of carbon sink drivers in the karst regions of southwest China. Based on this, it is further confirmed that the unique regional scale and geographic environmental conditions are the important reasons for the dominant differences between natural and anthropogenic factors and that human activities such as large-scale afforestation and ecological restoration projects obviously have more far-reaching impacts on the mountainous areas and karst geomorphic zones.

By analysing the contribution of human activities and climatic factors to NEP in the Taihang Mountains, in general, the contribution of the influence of human activities is much larger than that of natural factors such as temperature and precipitation. However, through Figure 12, it can be seen that there are large differences in the influence factors of vegetation carbon sinks in different regions of the Taihang Mountains. Combined with the regional topographic map and the administrative division map of the Taihang Mountains, in a few regions of the Taihang Mountains, such as Yangquan, Jinzhong, Changzhi, etc., the contribution of climatic factors to the NEP is more than 60% or even more than 80%.

Combining the land cover types and topographic features, it is found that this part of the region has higher elevation, and the land use types are dominated by forests, grasslands, shrubs, etc., which are richer in biodiversity and have stronger ecosystems and environmental carrying capacity, and are less affected by human activities than the lower elevation areas in the eastern Taihang Mountains.

On the other hand, in the lower altitude areas of the eastern Taihang Mountains, human activities are more complex, including rural population migration, urbanisation, and other anthropogenic activities, which make the regional vegetation type more homogeneous, and the area of urban land is increasing, which makes the impact of anthropogenic activities on the NEP much larger than that of climate change.

4.3. Shortcomings and Future Improvements of This Study

This paper describes the influence of natural factors on vegetation carbon sinks with the help of a parameter-optimal geographic detector. Although it can reduce the spatial scale dependence of traditional geographic detectors to a certain extent, the discretisation method of spatial data and the quantitative study of classification number are still insufficient and have not formed a rigorous research system. In addition, focusing on the geographic detector itself, its interaction detection is limited to two-factor, and it cannot obtain the comprehensive effect of three-factor and five-factor in depth. Therefore, the results obtained in this paper through this factor detection and interaction analysis—the interaction of DEM, precipitation, temperature, and other factors on the interpretation of vegetation carbon sinks—still has some limitations.

In future studies, the quality of the data and the accuracy of the research model will be further improved, and the effects of other driving factors on the carbon sink of vegetation in the Taihang Mountains will be further analysed.

5. Conclusions

This paper combines remote sensing technology and meteorological, topographic, and land use data to visualise the dynamics of vegetation carbon sinks in Taihang Mountains from 2000 to 2022 with the help of multiple modelling methods and uses a parameter-optimal geographic detector and multivariate regression residual analyses to carefully analyse the driving mechanisms of its influencing factors, obtaining the following conclusions:

• From 2000 to 2022, the trend of NEP growth in the Taihang Mountains was obvious, and the average value of NEP in the Taihang Mountains was 289 gC·m⁻²·a⁻¹, showing the characteristics of high in the northeast and low in the west and centre of the country in the spatial distribution, with a trend of gradual increase from the northeast to the southwest. Combined with geographic features, regional differences in DEM and vegetation types were found to be important factors for the spatial heterogeneity of the vegetation carbon sink.
• Over the 23 years from 2000 to 2022, the vegetation carbon sink in the Taihang Mountains region as a whole showed an increasing trend, and the area of the eastern low carbon sink area evolved over time and shrank. In addition, the vegetation carbon sink in the Taihang Mountains region had less fluctuation over time and was more stable.
• Vegetation carbon sinks in the Taihang Mountains are influenced by a variety of natural factors, of which the explanatory power of each natural factor is as follows: DEM (0.174) > temperature (0.148) > precipitation (0.026) > slope (0.017) > slope direction (0.003). The explanatory power of natural factors on vegetation carbon sequestration is significantly stronger than that of single factors in both pairwise interactions, with the interaction between DEM and precipitation having the strongest explanatory power. In addition, compared with climate change factors such as temperature and precipitation, the contribution of anthropogenic activities to NEP changes within more than 90% of the Taihang Mountain area is above 60%, indicating that the contribution of NEP changes is basically driven by anthropogenic activities and that human activities drive NEP changes in the Taihang Mountain area more significantly. The contribution of anthropogenic activities to NEP change is significantly higher than that of climate change, showing a strong promoting role.

Combined with the above conclusions, the vegetation carbon sinks and land use in the Taihang Mountains have changed considerably in more than 20 years, and anthropogenic activities have seriously disturbed them. Therefore, the protection of high carbon sink areas from human activities should be strengthened in the future development process, consolidating the existing benefits of ecological carbon sinks, such as adding ecological buffer zones at the borders of large-scale high carbon sinks to reduce the interference of human activities in high carbon sinks. Meanwhile, reducing carbon sources can be achieved by adjusting the spatial layout of green space within the construction land, increasing green space coverage, and effectively enhancing the carbon sink capacity of urban green space. In addition, in future research, we should further improve the quality of data and the accuracy of the research model, expand the analysis of the influence of other driving factors on the carbon sink of vegetation in the Taihang Mountainous Area, and analyse more comprehensively the natural and anthropogenic driving factors of the carbon sink of vegetation in the study area, so as to further innovate the research on the methodology and modelling aspects.