Assessing the Sustainability Impact of Land-Use Changes and Carbon Emission Intensity in the Loess Plateau

Abstract

Regional socioeconomic development is intricately tied to reasonable land-use resources. Although many studies have analyzed land-use carbon emissions, there is a lack of analysis of the concept of intensity. Studying the land-use carbon emission intensity (LUCEI) is crucial for shaping effective land management strategies that support the integrated sustainable development of society, the economy, and the environment. This study examines land-use changes on the Loess Plateau (LP) from 2000 to 2020. The coefficient method, spatial autocorrelation analysis, and optimal parameters-based geographical detector model are used to identify and analyze the spatial clustering patterns and influencing factors affecting LUCEI, which provides more in-depth insights for the study of LUCEI. The results indicate: (1) Urban and Grassland areas showed the most significant growth, with Urban areas expanding by 10,845.21 km² and Grasslands by 7848.91 km², respectively. This Urban expansion was mainly caused by the conversion of Grassland and Cropland, while Grassland expansion was primarily attributed to the decline in Barren. (2) The average LUCEI on the LP climbed from 0.38 in 2000 to 0.73 in 2020, indicating a 190.70% growth rate. (3) The spatial pattern of LUCEI remained stable but unevenly distributed, with extensive High-High and Low-Low clusters. (4) Socioeconomic factors had a greater explanatory power for LUCEI in the LP than natural factors. The LUCEI is not driven by a single factor, but by the combined influence of multiple factors. The interaction between nighttime light and population density explained the spatial distribution of LUCEI most strongly, with a q-value of 0.928. The findings underscore the critical role of socioeconomic development in shaping carbon emission dynamics on the LP. By linking LUCEI growth to land-use changes, this study offers concrete scientific guidance for policymakers seeking to balance socioeconomic growth with sustainable land-use practices. Based on these results, we recommend developing appropriate urban development plans that optimize land-use structures, enhance regional carbon sequestration capacities, and fully implement green transition requirements.

1. Introduction

Human activities have increasingly intensified their impact on the Earth’s ecosystems, with the rapid rise in carbon emissions becoming a primary driver of global warming [1,2]. Changes in land-use, a fundamental aspect of human development, now account for roughly one-third of global anthropogenic carbon emissions [3]. The transformation of land-use patterns, particularly the large-scale conversion of cropland into urban areas, not only diminishes crucial carbon sinks but also significantly increases sources of carbon emissions [4,5]. As China emerged as the largest carbon emitter globally, finding a balance between economic growth and environmental preservation has become more critical. In 2020, China announced its goals of attaining peak carbon emissions and becoming carbon neutral, injecting new momentum into global climate governance while presenting new challenges for low-carbon emission reduction efforts [6,7,8]. Economic growth often leads to changes in land-use, directly affecting the carbon emission intensity. The rational use of land, along with the restoration and enhancement of carbon sink functions, will be a core pathway to achieving carbon neutrality [9,10]. Therefore, researching land-use carbon emission intensity (LUCEI) under the framework of sustainable development is of vital importance.

Many previous studies have focused on land-use carbon emissions (LUCE). For instance, Rong et al. [11] examined the spatiotemporal change of LUCE between 2000 and 2018, revealing an upward trend and significant spatial correlations across different regions in China. Yang et al. [12] calculated LUCE in Xinjiang from 2000 to 2020 using socioeconomic and land-use data, finding that urban land expansion was the primary driver of increases in emissions. However, few studies have concentrated on the concept of carbon emission intensity, which refers to the carbon dioxide emissions generated per unit area of land over a specific period. Studying carbon emission intensity is crucial for policymakers to identify areas with high potential for emission reductions and to develop more effective energy-saving and emission-reduction policies [13]. Moreover, previous research has predominantly analyzed data at the provincial or citiey level, often overlooking the county-level dynamics. For instance, Huang et al. [14] took Jiangxi Province as the study area to divide the carbon balance zoning of LUCE in different regions. Yu et al. [15] studied the spatial spillover impacts of LUCE in the Yangtze River Delta urban agglomeration. However, these studies concentrated on broader scales beyond the county level, limiting their ability to precisely capture LUCE in specific areas. County-level studies are essential for developing more targeted and effective regional emission reduction strategies [16,17,18].

Currently, several studies focus on the spatial aggregation characteristics and spatial autocorrelation of carbon emission intensity. For example, Ke et al. [19] utilized spatial Markov chains and exploratory spatial data analysis to examine the LUCEI in China, uncovering significant differences in the spatial patterns of cities with varying clustering types. Xiang et al. [20] employed a centroid shift trajectory and standard deviation ellipse model to examine the spatiotemporal change of carbon emissions from China’s public buildings. In addition, multiple factors affect LUCEI, encompassing both natural and socioeconomic dimensions, and the investigation of these factors remains a critical area of research. For example, Zeng et al. [21] utilized a spatial econometric model to investigate the factors influencing carbon emission intensity from industrial land-use, finding that population density has a significant positive effect. Liu et al. [13] utilized the STIRPAT model to investigate the main drivers of LUCEI. These analytical approaches to spatial patterns and influencing factors have established a robust foundation for this study’s examination of LUCEI. As a systemic concept, LUCEI exhibits spatial heterogeneity, which refers to the marked differences in environmental characteristics, ecosystem attributes, or socioeconomic conditions across various geographic locations. Recently, Song et al. [22] proposed a novel approach for exploring the spatial heterogeneity of influencing factors, the optimal parameters-based geographical detector (OPGD) model. This model, which does not rely on linear assumptions, effectively quantifies the influence of different factors on LUCEI. It has also been widely employed in studies on vegetation change [23], urban thermal environments [24], and ecosystem services [25].

The Loess Plateau (LP) serves as a critical ecological security barrier in China, playing a significant role in achieving the nation’s dual carbon goals. Sustainable land-use in this region is essential for promoting ecological conservation and enhancing carbon sequestration [26,27,28]. The LP is a major grain-producing and energy base in China. However, economic development has led to urban area expansion, resulting in the reduction of grassland and cropland areas and a growing scarcity of land resources [29]. Although the LP has implemented a series of green restoration plans in the past, it still faces the risk of land degradation [30]. Therefore, this study focuses on the LP, with the goal of advancing sustainable land-use.

The workflow of this research consists of several key steps and research questions. (1) How does the land-use change in the LP from 2000 to 2020? (2) What is the change trend of LUCEI during the study period? (3) What are the global and local spatial autocorrelation changes of LUCEI? (4) What are the key factors affecting the spatial distribution of LUCEI, and how do the single-factor and interaction detection between the influencing factors affect it?

2. Materials and Methods

2.1. Study Area

The LP is located in northern China, spanning the middle reaches of the Yellow River. The geographical boundary of the region is 110.87° E–114.56° E and 33.69° N–41.27° N, with an area of 62.46 × 104 km² and an elevation range from 75 to 5149 m (Figure 1). The LP belongs to the temperate semi-humid to semi-arid climate zone, characterized by a distinct continental climate, with annual precipitation typically ranging between 400 to 800 mm. The sustainable development of the LP requires a wide-ranging assessment of economic growth, ecological protection, and social benefits. Through scientific planning and rational use of resources, we will promote the mutual progress of human beings in the LP between the development of production and life and the protection of the natural environment.

2.2. Date Sources and Processing

To conduct this study, we meticulously selected land-use data from the European Space Agency for the period 2000 to 2020, with a 300-m spatial resolution (cds.climate.copernicus.eu/ (accessed on 14 August 2024)). This dataset was chosen for its good temporal and spatial resolution and is widely used, which is crucial for accurately tracking the dynamic shifts and transitions in land-use on the LP. The data were reclassified into six distinct types: Cropland, Forest, Grassland, Water, Urban, and Barren, to facilitate a comprehensive analysis of land-use patterns. Furthermore, to provide a nuanced understanding of the influencing factors on LUCEI, we strategically selected six variables that encompass both natural and socioeconomic dimensions. The selection of these variables is based on their relevance to research objectives and the comprehensiveness and availability of existing data. Precipitation (PRE), temperature (TEM), and nighttime-light (NIG) were sourced from the National Earth Science Data Center (geodata.cn/ (accessed on 14 August 2024)); elevation (ELE) and slope (SLO) were sourced from the Geospatial Data Cloud (gscloud.cn/ (accessed on 14 August 2024)); and population-density (POP) were retrieved from the WorldPop (worldpop.org/ (accessed on 14 August 2024)). All the pre-data processing is cut in ArcGIS 10.8 software, the projection coordinate system is WGS1984, and the county-level data is calculated by using the partition statistical tool.

2.3. Methods

2.3.1. Land-Use Changes Matrix

The main principle of the land-use changes matrix comes from the Markov model, which can reflect the change situation at the initial and final stages of the study area in certain stages, including the direction and area of mutual changes [31,32,33]. This study uses the land-use changes matrix to examine the transformation of land-use in the LP. The formula is as follows:

$$Q_{ij}=\left(\begin{array}{ccc}{Q}_{11}& \dots & Q_{1n}\\ \dots & \dots & \dots \\ Q_{n1}& \dots & Q_{nn}\end{array}\right)$$

(1)

where Q is the area, n is the number of land-use types, and i and j are the initial and final stages of the study.

2.3.2. Calculation of LUCEI

To calculate LUCEI, the first step is to determine the total LUCE. Net carbon emissions from land-use are calculated as the difference between emissions and sequestration [14,34]. Typically, Forest, Grassland, Water, and Barren lands serve as carbon sinks. Based on existing research, their carbon emission coefficients are set at −0.578, −0.021, −0.252, and −0.005, respectively [35,36]. Cropland, on the other hand, functions as both a carbon sink and carbon source, with its coefficient set at 4.595 according to related research [37]. Urban, being centers of human activity and production, are the largest sources of carbon emissions, with a carbon emission coefficient of 65.3 [38]. LUCEI is calculated as the carbon emissions per unit of land area. The formula are as follows:

$$LUCE=\sum E_i=\sum {\gamma }_i\times A_i$$

(2)

$$LUCEI={\displaystyle \frac{{LUCE}_j}{{Area}_j}}$$

(3)

where $E_i$ represents the carbon emissions of the i-th land-use type, ${\gamma }_i$ is the carbon emission coefficient of the i-th land-use type, $A_i$ represents the area of the i-th land-use type, ${LUCE}_j$ represents the LUCE in the j-th region, and ${Area}_j$ is the area of the j-th region.

2.3.3. Spatial Autocorrelation Analysis

Based on the LUCEI of counties in the LP, this study applies spatial autocorrelation analysis to examine the spatial association characteristics of LUCEI. This method is commonly used to identify spatial dependencies, measured using Moran’s Index, which includes both Global and Local spatial autocorrelation [39,40]. In this study, the Global Moran’s I was used to evaluate the overall spatial correlation of LUCEI. To further identify areas with localized carbon emission clusters, Local Moran’s I was utilized to investigate different types of spatial clustering. The formulas are as follows:

$$GI={\displaystyle \frac{n{\sum }_i{\sum }_j{w}_{ij}(x_i-\overline{x})(x_j-\overline{x})}{({\sum }_i{\sum }_j{w}_{ij}){\sum }_i{(x_i-\overline{x})}^2}}$$

(4)

$$LI={\displaystyle \frac{x_i-\overline{x}}{{\sum }_i{(x_i-\overline{x})}^2}}{\sum }_j{w}_{ij}(x_j-\overline{x})$$

(5)

where GI represents the Global Moran’s I, LI represents the Local Moran’s I, x_i and x_j are the LUCEI of counties i and j, n is the total number of counties, w_ij is the spatial weight matrix based on the distance between counties, and $\overline{x}$ is the mean LUCEI.

2.3.4. Optimal Parameters-Based Geographical Detector Model

The LP spans the midstream region of the Yellow River, covering numerous counties. These counties exhibit significant differences in development levels and energy consumption structures, leading to pronounced spatial heterogeneity in LUCEI. The OPGD model was utilized to uncover the underlying driving factors and explore spatial differentiation. This model does not assume linear relationships and can accurately identify the effects of different factors on LUCEI [41,42]. The formula is as follows:

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

(6)

where q represents the explanatory power of influencing factors on LUCEI, ranging from zero to one. A higher q-value implies a better explanatory power of the factor, while a lower q-value suggests a weaker influence. L represents the total number of strata; $N_h$ and $N$ are the sample number and total sample number of stratum h, respectively; ${\sigma }_h^2$ and ${\sigma }^2$ are the variance of the sample in stratum h and the total sample, respectively.

3. Results

3.1. Spatiotemporal Changes of Land-Use Pattern in the LP

Based on the land-use distribution data from 2000, 2005, 2010, 2015, and 2020, the spatiotemporal changes in land-use patterns in the LP are illustrated (Figure 2). Spatially, the LP is mainly characterized by Cropland, Forest, and Grassland. Grassland is concentrated in the northwest, while Forest is mainly in the southeast, largely due to regional differences in precipitation. Cropland is mainly distributed in the plains, areas more suitable for human habitation and agriculture. Barren is mainly located on the northwestern edge, and the spatial distribution of Urban closely mirrors that of Cropland, being concentrated in economically developed regions. Temporally, Urban has expanded significantly, while Cropland has decreased. As urbanization drives economic growth, cities have extended outward from core areas, encroaching on surrounding Cropland, leading to a steady expansion in Urban and a notable reduction in Cropland.

Through land-use change matrices, we determined the area proportions and transitions for various land-use types in the LP over time (Figure 3). In 2000, Cropland, Forest, Grassland, Water, Urban, and Barren land made up 35.86%, 20.84%, 40.27%, 0.27%, 0.36%, and 2.41% of the total LP area, respectively. By 2020, Grassland, Water, and Urban areas had expanded by 7848.91 km², 230.37 km², and 10,845.21 km², respectively. Conversely, Cropland, Forest, and Barren had decreased by 7530.58 km², 7428.14 km², and 3964.78 km², respectively. The most significant growth occurred in Urban areas, largely at the expense of Grassland and Cropland, which were converted into Urban areas by 3876.08 km² and 6071.99 km², respectively. This shift from Cropland to Urban areas is closely tied to the significant economic growth and urban expansion of the LP. Additionally, the reduction in Barren is largely due to its conversion into Grassland, with 4346.39 km² of Barren converted.

3.2. Spatiotemporal Changes of LUCEI in the LP

We categorized the LUCEI in the LP into five classes (Figure 4). Spatially, the Very Low category covers the largest area, mainly found in the central and western regions where large areas of grassland and vegetation exist with minimal human disturbance. The High and Very High categories are concentrated along the northern, southern, and eastern edges of the LP, corresponding to areas with greater economic development levels. Temporally, the average LUCEI in the LP increased from 0.38 in 2000 to 0.73 in 2020, reflecting a 190.70% rise. This indicates that rapid population growth and economic development have substantially driven the increase in LUCEI. In 2000, the proportions of Very Low, Low, Medium, High, and Very High categories were 81.86%, 14.61%, 2.52%, 0.25%, and 0.76%, respectively. By 2020, these proportions had shifted to 63.98%, 23.68%, 6.05%, 2.77%, and 3.53%, respectively. The most significant reduction occurred in the Very Low category, which declined by 17.88%, while all other categories experienced varying degrees of growth.

3.3. Spatial Autocorrelation of LUCEI in the LP

This study employs spatial autocorrelation methods to estimate LUCEI across counties in the LP for the years 2000, 2005, 2010, 2015, and 2020 (Figure 5). The results indicate that the Moran’s I values for these years were 0.124, 0.136, 0.143, 0.151, and 0.136, respectively, indicating an initial rise in clustering followed by a slight decline, with the greatest value in 2015. Additionally, all Moran’s I values were positive, and the p-values for 2000, 2005, 2010, 2015, and 2020 were 0.003, 0.003, 0.002, 0.001, and 0.001, respectively, demonstrating significant positive spatial autocorrelation and noticeable clustering effects of LUCEI.

The results of the local spatial autocorrelation analysis reveal that, while the spatial distribution of LUCEI across counties in the LP remained relatively stable between 2000 and 2020, there was a notable spatial imbalance. High-High and Low-Low clusters were widespread (Figure 6). The High-High clusters remained relatively stable and were primarily centered in counties such as Xi’an, Xianyang, Weinan, Luoyang, and Zhengzhou. As the economic core of the LP, these areas have undergone rapid urbanization and industrialization, resulting in higher land-use intensity and, consequently, higher carbon emissions, thus forming High-High spatial clusters. Conversely, Low-Low clusters were concentrated in counties like Baiyin, Zhongwei, Wuzhong, Yan’an, Tianshui, and Datong. These areas, with favorable natural conditions and limited human activity, exhibited lower LUCEI, resulting in Low-Low spatial clusters. Additionally, many counties surrounding the High-High clusters showed Low-High spatial clusters, suggesting potential future transitions into High-High clusters. Therefore, these areas should be closely monitored in future land-use planning and carbon mitigation strategies.

3.4. Analysis of Influencing Factors of LUCEI in the LP

According to the OPGD model analysis, LUCEI in the LP is influenced by various natural and socioeconomic factors, with varying degrees of influence from each (Figure 7). The results of the single-factor detection analysis demonstrate that the q-values for POP, NIG, SLO, ELE, TEM, and PRE are 0.769, 0.869, 0.152, 0.215, 0.162, and 0.055, respectively. Among these, POP and NIG emerge as the most critical factors in determining the regional distribution of LUCEI, highlighting the leading role of population size and economic development. As urbanization and industrialization advance, the natural landscape is extensively transformed, especially in economically prosperous areas where land-use changes are more pronounced. These activities often consume large amounts of energy and are closely linked to high carbon emissions. Furthermore, population growth typically correlates with increased demand for land resources, whether for housing, industry, or agriculture, all of which contribute, directly or indirectly, to rising carbon emissions. In contrast, natural factors show relatively lower explanatory power, suggesting they play a less significant role in explaining the spatial distribution of LUCEI. Additionally, results from the factor interaction detection reveal that the relationships between factors exhibit significant nonlinear or bi-factor enhancement effects, suggesting that LUCEI in the LP is the consequence of multiple factors working together, rather than being driven by a single factor. The interaction between POP and NIG has the highest explanatory power, with a q-value of 0.928. The combined effect of natural and socioeconomic factors is greater than the impact of individual factors, highlighting the importance of considering these interactions when formulating carbon reduction policies to sustainably reduce LUCEI.

4. Discussion

4.1. Spatiotemporal Change Analysis of Land-Use Change and LUCEI

In the previous analysis, it was observed that Urban and Grassland areas experienced the largest expansions, growing by 10,845.21 km² and 7848.91 km², respectively. To support economic growth during the urbanization of the LP, urban construction land continuously expanded beyond the core areas, resulting in a reduction of cropland. Cropland plays a crucial part in ecosystems, providing essential ecological services and ensuring regional food security. The reduction of farmland can disrupt ecological balance, reduce ecosystem stability and resilience, and diminish agricultural productivity, which may lead to food shortages and compromise regional food security [43,44]. If not properly planned and managed, rapid urbanization and industrialization may pose long-term challenges to environmental, social, and economic sustainability [45,46]. To mitigate these negative impacts, comprehensive strategies are needed, including effective urban planning, cropland protection and restoration, and the promotion of green urbanization and sustainable development practices. Additionally, the study found that the increase in Grassland primarily resulted from the conversion of Barren land, reflecting the positive impact of green restoration initiatives implemented in the LP, which have contributed to improving the region’s fragile ecological environment [47]. The reforestation and greening projects in the LP are critical for promoting long-term development in the region and are key to achieving stability and well-being for the local population [29]. Despite the encroachment of urban development on ecological land, policies have helped compensate for these losses, advancing the ongoing development of the human-land relationship in the LP. The findings highlight the negative impact of Urban expansion on Cropland and Grassland, providing valuable insights for formulating ecological protection and restoration strategies for the region. Through careful land-use planning, ecological balance can be promoted, and regional ecosystem services can be enhanced.

In addition to the changes in the number of land-use change and LUCEI, we also noticed the topic of spatiotemporal heterogeneity in the study. The spatial heterogeneity of regional development has long been a central theme in academic discourse, and addressing these spatial imbalances is essential for promoting sustainable regional growth [48,49]. This research examined the spatial pattern of LUCEI in the LP, revealing significant spatial disparities. Regions classified as High and Very High LUCEI are mainly focused in the southern, eastern, and northern peripheries of the LP, where economic development is more concentrated due to industrial activity, rapid urbanization, and higher population densities. Furthermore, the spatial autocorrelation analysis results reveal a positive correlation in LUCEI, meaning that high-emission areas tend to cluster, forming significant agglomeration effects. This clustering is driven by similar land-use practices, industrial structures, and economic development patterns within the region. Over the past 20 years, the average LUCEI in the LP has risen from 0.38 to 0.73, signifying a 190.70% increase. This trend is closely tied to the region’s rapid population growth and significant economic advancement. As population and economic activities intensify, so too does the demand for land resources, leading to both agricultural expansion and the growth of industrial and urban land [50]. These shifts have directly altered land-use patterns, consequently driving up LUCEI.

4.2. Planning and Management of LUCEI Based on Sustainable Development

This study conducts a comprehensive investigation into the various natural and socioeconomic factors influencing LUCEI, highlighting the significant differences in their impacts. By introducing the OPGD model, this research provides a novel method for analyzing the factors affecting LUCEI, offering a new analytical tool in geography and environmental science to better understand the driving forces behind complex environmental issues. The results indicate that socioeconomic factors play a more effective role in explaining the spatial distribution of LUCEI, with the intensity shaped by the interplay of multiple factors rather than a single cause. Consequently, the planning and management of LUCEI must fully account for the complex interactions between natural and socioeconomic factors. Special emphasis should be placed on socioeconomic governance, ensuring careful consideration on how to implement effective strategies in urban development and human activities to mitigate the rising trend in LUCEI and its associated negative impacts. Existing studies suggest that increasing carbon emission intensity is often accompanied by changes in land-use patterns, such as overdevelopment and improper land-use, which can lead to land degradation, exacerbated soil erosion, and threats to agricultural productivity and food security [30,51]. Therefore, the planning and management of LUCEI under the framework of sustainable development is critically important, as land-use practices directly affect regional ecological balance, economic development, and societal well-being. Proper land-use planning can enhance land-use efficiency, protect the environment, promote green economic development, and achieve harmonious coexistence between humans and land [46].

In view of the above existing problems, we put forward the following more specific suggestions. Firstly, optimizing the land-use structure by increasing the proportion of natural ecosystems, such as forests and wetlands, can enhance regional carbon sequestration capacity, reduce greenhouse gas emissions, and counteract chaotic land development [47]. For instance, the vegetation restoration and reconstruction projects in the LP have significantly improved the region’s vegetation cover and soil carbon sequestration capacity. Secondly, land-use planning and management should adopt scientific methods to improve land productivity and reduce overdevelopment and improper land-use [50,52]. For example, implementing sustainable agricultural methods such as organic farming, crop rotation, and conservation tillage, while minimizing the utilization of pesticides and chemical fertilizers, can reduce the environmental effect of agriculture. Furthermore, at all stages of urban planning, construction, and governance, it is essential to fully implement green transition requirements, advocate for low-carbon planning and design concepts, and strictly enforce urban development boundaries.

4.3. Limitations of Current Research and Suggestions for Future Research

While the current study offers an in-depth analysis of land-use changes and LUCEI in the LP region from 2000 to 2020, it is not without limitations. The geographical scope is confined to the LP area, and although the study spans two decades, it may not be sufficient to capture long-term trends. The singularity of data sources and the assumptions inherent in the model applications limit the universality of the findings. Moreover, while natural and socioeconomic factors have been considered, there may be other variables not included that significantly influence LUCEI. Spatial analysis is primarily conducted at the county level, potentially overlooking detailed changes at smaller scales, and the study does not fully elucidate the causal relationships between variables. Future research could consider expanding the geographical scope, extending the time frame, diversifying data sources, and refining models to enhance their explanatory power. Additionally, incorporating more variables, conducting multi-scale analyses, exploring causal relationships, examining the impact of policy interventions, considering the potential effects of climate change, and engaging in interdisciplinary research could provide a more comprehensive understanding of land-use changes and LUCEI, offering more scientific support for land-use management, carbon emission control, and ecological conservation.

5. Conclusions

In view of the four research questions we have raised above and the gaps in current research, we use LP as the research area and adopt the county scale to analyze the spatial and temporal distribution of land use change and LUCEI. In addition, we also innovatively introduce the OPGD model to analyze LUCEI from the perspective of spatial and temporal heterogeneity. The primary conclusions are as follows:

(1)

The most significant increases in land area were observed in Urban and Grassland types, with expansions of 10,845.21 km² and 7848.91 km², respectively. The expansion of Urban primary resulted from the loss of Grassland and Cropland. The increase in Grassland primarily resulted from the reduction of Barren.

(2)

The average LUCEI in the LP has climbed from 0.38 to 0.73 in the last 20 years, signifying a 190.70% increase. In 2020, the proportion of Very Low, Low, Medium, High, and Very High LUCEI areas accounted for 63.98%, 23.68%, 6.05%, 2.77%, and 3.53%, respectively.

(3)

The LUCEI demonstrated significant spatial positive autocorrelation, indicating a pronounced clustering effect. Although the spatial pattern of LUCEI remained relatively stable, its distribution was uneven, with widespread High-High and Low-Low clustering patterns.

(4)

Socioeconomic factors explained LUCEI on the LP more effectively than natural factors. POP and NIG were the most influential factors in explaining the spatial distribution of LUCEI, with their interaction showing the greatest explanatory power, reaching a q-value of 0.928.

This study is mainly aimed at the existing research gaps to do further analysis, from the perspective of intensity to analyze LUCE, but also from the perspective of spatial and temporal heterogeneity of LUCEI in-depth analysis, which solves the previous research focused on large administrative division scale and spatial and temporal consistency of the hypothesis research, for the future related research provides a good idea. The study believes that it is necessary to carry out urban planning according to local conditions and promote the implementation of reforestation and greening projects. In all stages of urban planning, construction, and governance, low-carbon planning and design concepts must be fully advocated. This research not only provides scientific support for land-use management, carbon emission control, and ecological protection on the LP but also enriches the related theoretical framework, contributing significantly to promoting regional sustainable development. However, this study still cannot comprehensively measure the factors affecting LUCEI. Future research can be further analyzed from the aspects of industrial structure and multi-scenario simulations.