Spatiotemporal Evolution of Carbon Emissions and Ecosystem Service Values in Xinjiang Based on LUCC

Abstract

This study is based on time-series land use data of Xinjiang from 2000 to 2022. Using grid tools, bivariate autocorrelation models and other methods, we systematically analyzed the spatiotemporal variation characteristics of land use and ecosystem service value. The results show the following: Firstly, from 2000 to 2022, Xinjiang’s LUCC exhibits differentiated evolution characteristics: cropland, forestland, and built-up land expanded continuously, while the areas of grassland and unused land showed a steady reduction trend, and the area of water bodies showed a fluctuating growth pattern. Secondly, according to the calculation of carbon emissions from LUCC in Xinjiang from 2000 to 2022, the carbon emissions from LUCC have increased significantly, from 27.79 million tons in 2000 to 226.43 million tons in 2022, with built-up land being the main source of carbon emissions, but the continuous reduction in grassland area has led to the weakening of carbon sequestration capacity. Thirdly, from 2000 to 2022, Xinjiang’s ESV shows a fluctuating upward trend, increasing from 1880.528 billion yuan in 2000 to 1894.198 billion yuan in 2022, with grassland and water area being the core contributors to ESV, accounting for over 80% of the total contribution. Fourthly, in terms of spatial distribution, there is an overall negative correlation between the intensity of carbon emissions from LUCC and the intensity of ESV, mainly aggregated as “low–low” and “low–high”, with “high–low” aggregation primarily distributed in the desert areas of the Tarim Basin and Junggar Basin and “low–high” aggregation concentrated in the marginal mountainous areas and oasis regions of Xinjiang. The findings provide a solid scientific basis for the optimization of land use structure, the achievement of carbon emission reduction targets, and the protection of ecosystems in Xinjiang and similar arid regions worldwide.

1. Introduction

Growing carbon emissions and ongoing ecosystem degradation have become prominent issues constraining the sustainable development of human society. The spatiotemporal evolution, spatial correlation, and driving mechanisms of these two processes have emerged as a research hotspot. Among these, the LUCC serves as a critical link between human activities and natural ecosystems. It is not only one of the primary sources of carbon emissions but also a core driver influencing the spatiotemporal pattern of ESV. By altering land cover types and adjusting surface material cycles and energy flows, the LUCC directly regulates regional carbon budget balances and indirectly affects the provision, regulation, support, and cultural services of ecosystems [1,2]. However, existing studies often treat carbon emissions and ecosystem service value (ESV) as independent outcomes of LUCC and analyze them in a unidirectional manner, thereby overlooking their complex spatiotemporal relationships. Particularly in ecologically fragile arid and semi-arid regions, land use change frequently leads to a coupled pattern of carbon source expansion and ecological function degradation. Therefore, clarifying the spatiotemporal evolution and intrinsic interactions between carbon emissions and ESV under the LUCC context is of high theoretical and practical significance for optimizing land use structure and promoting the coordinated advancement of carbon reduction and ecological conservation [3,4].

The arid and semi-arid regions of northwestern China are ecologically fragile and highly sensitive to global climate change. Driven by LUCC, these regions face particularly prominent imbalances in the carbon cycle and degradation of ecosystem services. As a core component of this region, Xinjiang features a vast territory and diverse LUCC types, including deserts, grasslands, forestlands, croplands, and built-up lands. It is not only an important energy base and agricultural production base in China but also a key node in the ecological security barrier of northwestern China. Xinjiang’s unique geographical and climatic conditions characterized by scarce precipitation, intense evaporation, and large diurnal temperature ranges render Xinjiang’s ecosystems inherently vulnerable with poor self-repair. In recent years, with the deepening implementation of China’s Western Development Strategy and the Belt and Road Initiative, the urbanization and industrialization processes in Xinjiang have accelerated, leading to a significant population agglomeration effect. This has resulted in drastic adjustments to the LUCC structure. This not only causes continuous fluctuations in the total regional carbon emissions but also triggers a decline in ESV, a reduction in biodiversity, and an exacerbation of land desertification. These ecological and environmental issues seriously hinder the coordinated and sustainable development of Xinjiang’s economy, society, and ecology. There is an urgent need for targeted research to address this developmental dilemma.

Scholars worldwide have conducted extensive research on the unidirectional relationships between LUCC and carbon emissions, as well as between LUCC and ESV, resulting in a substantial body of research findings. In terms of LUCC and carbon emissions, researchers have primarily focused on the accounting and analysis of driving mechanisms of carbon emissions in various regions and LUCC types. For instance, Liu et al. utilized long-time-series national LUCC data to elucidate the impact of LUCC changes on carbon emissions in China, providing theoretical support for national carbon reduction planning [5]. Some scholars focus on a single region. They explore the spatiotemporal correlation characteristics between LUCC type conversion and carbon emission. They clarify the carbon emission effects of different cover type conversions [6,7]. Other scholars carry out research on carbon emission accounting and optimization paths from different perspectives. These studies enrich the research perspectives in this field [8,9,10]. In the realm of research on LUCC and ESV, scholars often focus on typical river basins and provinces as study areas to explore the impact pathways and spatiotemporal evolution characteristics of LUCC on ESV. For example, Liu et al. took the Yellow River Basin as the research subject, analyzing the driving role of LUCC transformation on ecosystem service values and proposing a collaborative path for ecological protection and LUCC optimization [11]. Additionally, some scholars have focused on areas such as the Miyun Reservoir and Nanjing to explore the spatiotemporal correlations between LUCC and ESV, thereby providing a scientific basis for watershed ecological protection [12,13,14]. Patel et al. focused on agricultural LUCC changes, exploring their associations with agricultural ecosystem service values and agricultural economic growth, thus expanding the research dimensions of a single LUCC type [15].

Although numerous studies have explored the impact of LUCC in different regions on carbon emissions or ESV, there are still key issues in the existing literature that require further investigation. First, most studies remain focused on one-way impact analysis, while systematic investigations into the bidirectional spatial correlation between LUCC-induced carbon emissions and ESV are still insufficient, especially in arid regions. Second, the majority of existing studies rely on administrative units rather than refined spatial units, which cannot effectively reflect the fine-scale spatial heterogeneity of LUCC, carbon emissions, and ESV, thus limiting the accuracy and applicability of research results for policy formulation.

In response to these limitations, this study adopts a gridded analysis approach and divides Xinjiang into 16,777 evaluation units, to systematically explore the spatiotemporal evolutionary characteristics and spatial correlations between LUCC carbon emissions and ESV in Xinjiang from 2000 to 2022. The innovations of this study are summarized as follows: How have spatiotemporal patterns of LUCC-related carbon emissions and ESV in Xinjiang evolved over the past two decades? What is the spatial correlation between these two systems? What targeted implications can be derived for low-carbon development and ecological protection in arid regions? By answering these questions, this study seeks to provide a scientific basis and practical guidance for optimizing regional LUCC structure, advancing carbon emission reduction targets, and promoting ecosystem conservation and restoration in Xinjiang, thereby supporting its coordinated economic, social, and ecological sustainable development. The contributions of this study are as follows: First, it moves beyond traditional unidirectional correlation analyses and reveals the spatial interdependence between LUCC-driven carbon emissions and ESV. Second, it focuses on the whole Xinjiang region, addressing the key shortcomings in the existing literature by providing a large-scale, long-term, and high-resolution gridded analysis for arid zones.

2. Data and Methods

2.1. Overview of the Study Area

Xinjiang is located in the northwest of China, at the heart of the Eurasian continent, with geographical coordinates ranging from 73°40′ E to 96°18′ E and from 34°25′ N to 48°10′ N. It is the largest provincial-level administrative region in China in terms of land area, with the most neighboring countries and the longest land border. Xinjiang comprises 14 prefectural-level divisions (cities and autonomous prefectures). It shares borders with eight countries, including Mongolia, Russia, and Kazakhstan, encompassing a total area of 1.6649 million square kilometers, which is approximately one-sixth of China’s total land area (Figure 1).

Xinjiang’s unique topographical pattern is characterized by the “Three Mountains and Two Basins” formation. From north to south, these are the Altay Mountains, the Junggar Basin, the Tianshan Mountains, the Tarim Basin, and the Kunlun Mountains. The Tianshan Mountains, stretching across the center, divide Xinjiang into southern and northern parts. The Tarim Basin in southern Xinjiang is the largest inland basin in China, with the Taklamakan Desert in its center being the largest in China and the world’s second-largest shifting sand desert. The Junggar Basin in northern Xinjiang is China’s second-largest inland basin. Influenced by its landscape and location relative to the sea, Xinjiang experiences a typical temperate continental arid climate, with scarce precipitation and high evaporation. The average annual precipitation is only 170.6 mm, while the evaporation rate is significantly higher. The region exhibits large diurnal temperature variations, and its overall climatic conditions are relatively harsh.

Ecologically, Xinjiang serves as a crucial component of the ecological security barrier in northwestern China, featuring a unique ecological framework. This includes the Altay, Tianshan, and Kunlun–Altun mountain ranges that form a primary ecological shield, as well as two oasis ecological rings distributed along the Tarim and Junggar Basins. The region harbors diverse ecosystems, such as grasslands, deserts, forestlands, wetlands, and croplands, rich in biodiversity. However, due to its arid climate, Xinjiang’s ecosystems are inherently fragile with low self-repair capabilities. Issues like land desertification and soil salinization are prominent, making Xinjiang a sensitive area to global climate change and an ecologically vulnerable region.

Xinjiang boasts abundant total land resources and diverse LUCC types, as one of the five major pastoral areas in China, while also possessing rich forestlands, croplands, and desert lands. However, regional development within Xinjiang is uneven, with significant disparities between northern and southern regions. Meanwhile, the urban spatial pattern requires optimization: urban agglomerations are predominantly concentrated in a few core areas such as Urumqi–Changji, Kashgar, and Hotan, with an excessive concentration of population, industry, and infrastructure in these hubs, while the surrounding counties and rural areas lack sufficient development momentum and connectivity, which directly affects the rational allocation of land resources. The LUCC structure is undergoing significant adjustments, with conflicts arising from cropland reclamation, expansion of built-up land, and the contraction of grasslands, forestlands, and wetlands. These issues not only affect the regional carbon emission pattern but also have a significant impact on ecosystem service values, making Xinjiang the core study area of this research.

2.2. Research Data

All data required for this study were obtained from publicly accessible databases, ensuring reliability, timeliness, and completeness to meet the research needs. The LUCC data were derived from the CLCD dataset released by Wuhan University [16], a well-validated and widely applied dataset with a 30 m × 30 m spatial resolution [17,18]. The CLCD dataset is based on 335,709 Landsat images available on Google Earth Engine, with an overall accuracy rate of 80%. This ensures its ability to accurately capture the temporal and spatial dynamics of land use in Xinjiang. The study utilized a gridding tool to divide the research region into 16,777 assessment units to examine the relationship between LUCC carbon emissions and ESV. The data related to food are all sourced from the China Agricultural Yearbook, the China Statistical Yearbook and the Xinjiang Bureau of Statistics.

The original CLCD dataset comprises nine LUCC types. In order to align with the research focus on carbon emissions and ESV calculations and simplify the analysis process while retaining the main land cover characteristics, the original nine land types were merged into six core categories: wetland and water were merged into water area, snow/ice and barren were merged into unused land, and shrub and forest were merged into forestland. The final six core LUCC types are cropland, forestland, grassland, water area, built-up land, and unused land.

2.3. Research Methods

2.3.1. Calculation of Land Area by Type

During the data processing stage, based on the ArcGIS 10.8 software platform, the 6 periods of land use raster data (2000, 2005, 2010, 2015, 2020, and 2022) that had undergone reclassification preprocessing were calculated using the Raster Calculator tool (ArcGIS 10.8 software). Subsequently, combined with the gridding analysis tool, the relevant area data of various land use types in the 16,777 evaluation grids in Xinjiang during the research period from 2000 to 2022 were finally extracted and obtained, laying a data foundation for subsequent research and analysis [19,20].

2.3.2. Calculation of Land Carbon Emission

In the existing studies, there is no specific carbon emission/absorption coefficient for land use types in Xinjiang that is widely recognized and applied. Most of the studies on the carbon effect of land use in Xinjiang rely on or refer to the general coefficients calculated based on national or other regional studies [21,22]. Specifically, direct carbon emissions/absorptions from cropland, forestland, grassland, water area, and unused land are calculated using the carbon emission coefficient method, which is a widely accepted and mature approach in land use carbon emission research. In contrast, indirect carbon emissions from built-up land are calculated based on energy consumption data, as built-up land is primarily associated with anthropogenic activities (e.g., industrial production, residential energy use) that indirectly generate carbon emissions, rather than direct carbon exchange with the atmosphere through vegetation or soil. Referring to the research conducted by relevant scholars, LUCC types are categorized into direct carbon emission types and indirect carbon emission types, with different methods used to calculate the total carbon emissions. Direct carbon emissions from cropland, forestland, grassland, water area, and unused land are calculated using the carbon emission coefficient method, while indirect carbon emissions from built-up land are calculated based on energy consumption data. The carbon emission coefficients for cropland, forestland, grassland, water area, and unused land are set at 0.422, −0.612, −0.021, −0.235, and −0.005 (t/hm²) [23,24,25,26]. Negative coefficients indicate that the land type has carbon sequestration functionality, absorbing carbon dioxide from the atmosphere, while positive coefficients indicate that the land type is a carbon source, releasing carbon dioxide into the atmosphere. The specific calculation formula is as follows:

$$C_1={\sum }_{i=1}^n{\alpha }_i{S}_{it}$$

(1)

where i represents different land types, α_i represents the carbon emission coefficient for different land types, and S_it (hm²) represents the area of different LUCC types in region at time t. For built-up land carbon emissions, an indirect accounting method is adopted, drawing on the research approaches of relevant scholars [27,28,29]. Considering the actual energy consumption in Xinjiang, nine major energy sources are selected: raw coal, coke, natural gas, crude oil, gasoline, diesel, kerosene, fuel oil, and electricity. The total carbon emissions from built-up land are calculated based on the consumption data of these energy sources. The calculation formula is as follows:

$$C_2={\sum }_{i=1}^n{{\gamma }_i\beta }_i{E}_{it}$$

(2)

In the formula, γ_i represents the standard coal conversion coefficient for each type of energy, where solid fuels are measured by the weight of standard coal required per kilogram of energy consumption, and gaseous fuels are measured by the weight of standard coal required per cubic meter of energy consumption. β_i represents the carbon emission coefficient for each type of energy; E_it represents the consumption of each type of energy in region at time t.

During the calculation process, the consumption of each type of energy must first be converted into standard coal equivalents before calculating the total carbon emissions. The standard coal conversion coefficients and carbon emission coefficients for each type of energy are detailed in

Table 1. Conversion coefficients and carbon emission coefficients for various types of energy.

Energy	Raw Coal	Coke	Natural Gas	Crude Oil	Gasoline	Diesel	Kerosene	Fuel Oil	Electric Power
Standard Coal Conversion Coefficient	0.714	0.971	1.33	1.428	1.471	1.457	1.471	1.428	0.793
Carbon Emission Coefficient	0.755	0.855	0.448	0.585	0.553	0.592	0.571	0.618	0.404

.

LUCC carbon emission intensity refers to the carbon emission generated per unit area during the LUCC process. It is a key indicator to measure the impact of LUCC patterns on carbon emission. The calculation formula is as follows:

$$C={\sum }_{i=1}^n{\displaystyle \frac{C_i}{S_{it}}}$$

(3)

In the formula, C represents the carbon emission intensity, C_i represents the carbon emission of the i type of LUCC, and S_it represents the area of the i type of LUCC in period t.

In this study, the total energy consumption data of Xinjiang was allocated to 16,777 grid units based on the area weight of built-up land. The specific allocation process is as follows: first, we calculated the area of built-up land in each grid unit and the total area of built-up land in the entire region; second, we determined the area weight of each grid unit by the ratio of the built-up land area of the grid to the total built-up land area of the region; finally, we allocated the regional total energy consumption to each grid proportionally according to the area weight of each grid, so as to obtain the grid-scale energy consumption data and further estimate the grid-level carbon emission intensity.

2.3.3. Calculation of ESV

The ESV coefficients for Xinjiang are determined based on the four dimensions of ESV proposed by Xie et al., namely provisioning services, regulating services, supporting services and cultural services [30]. According to the research of relevant scholars [31,32], one ESV equivalent is defined as 1/7 of the average grain output value per unit in this region. On this basis, the unit ESV of Xinjiang is calculated (

Table 2. ESV per unit land type in Xinjiang from 2000 to 2022 (yuan/hm²).

Functional Types		Cropland	Forestland	Grassland	Water Area	Unused Land
Provisioning services	Food production	2737.390	677.122	553.258	1981.821	12.386
	Grain production	606.933	1560.684	933.107	569.773	37.159
	Water resource supply	−3232.845	809.243	503.713	20,536.618	24.773
Regulating services	Gas regulation	2204.776	5136.219	3212.201	1907.503	161.023
	Climate regulation	1151.933	15,359.111	8901.678	5672.962	123.864
	Environmental purification	334.432	4467.354	2840.610	13,748.882	507.842
	Hydrological regulation	3703.528	9570.543	6729.933	253,276.698	297.273
Supporting services	Soil conservation	1288.184	6250.993	3914.096	2303.867	185.796
	Biodiversity conservation	383.978	478.940	297.273	173.409	12.386
Cultural services	Aesthetic landscape	421.137	5689.477	3559.020	6317.054	173.409
Total		9785.240	52,493.478	33,013.831	311,170.638	1610.229

).

In this case, since the ESV of built-up land is 0, it is not displayed. The average grain yield per hectare in Xinjiang from 2000 to 2022 was 6063.261 kg/hm², and the average grain purchase price in Xinjiang in 2022 was 2.86 yuan/kg. The final calculation yields an ESV equivalent of 2477.256 yuan/hm² for Xinjiang from 2000 to 2022. According to the unit land ESV equivalent for Xinjiang in Table 2, the ecosystem service values for different land types in Xinjiang are as follows:

$${ESV}_i=A_i{U}_i^{ESV}$$

(4)

In this case, ESV_i represents the ecosystem service value of the i land type, A_i represents the area of the i land type, and U_i^ESV represents the ESV_t per unit area of the i land type. The total ESV_t for Xinjiang is calculated as follows:

$${ESV}_t={\sum }_{i=1}^n{ESV}_i$$

(5)

Furthermore, the intensity of the ESV in Xinjiang is calculated as follows:

$${\overline{ESV}}_i={\displaystyle \frac{{ESV}_i}{A_i}}$$

(6)

2.3.4. Sensitivity Index (SI)

To verify the accuracy of the results, the SI of ESV is introduced. This involves adjusting the ESV per unit area by 50% in either a positive or negative direction and observing the degree of change in the SI. If the SI is greater than or equal to 1, it indicates the low reliability of the results; otherwise, the results are considered to be reliable. The calculation formula for the SI is as follows [33]:

$${SI}_i=\left|{\displaystyle \frac{({\mathrm{E}\mathrm{S}\mathrm{V}}_{\mathrm{b}}-{\mathrm{E}\mathrm{S}\mathrm{V}}_{\mathrm{a}})/{\mathrm{E}\mathrm{S}\mathrm{V}}_{\mathrm{a}}}{({\mathrm{V}}_{\mathrm{b}}-{\mathrm{V}}_{\mathrm{a}})/{\mathrm{V}}_{\mathrm{a}}}}\right|$$

(7)

In the formula, V_a and V_b represent the $\mathrm{E}\mathrm{S}\mathrm{V}$ per unit area before and after the adjustment, while ESV_a and ESV_b represent the ecosystem service values before and after the adjustment.

2.3.5. Bivariate Spatial Autocorrelation

LUCC is a core driver affecting regional CE and ESV, and the spatial correlation between the two is a key entry point for revealing the synergistic evolution of human–environment systems. Univariate spatial autocorrelation can only reflect the agglomeration characteristics of the element itself and cannot depict the spatial correlation mechanism between elements. Bivariate spatial autocorrelation analysis, by constructing a spatial weight matrix and combining global and local indicators, can accurately identify the spatial correlation patterns between CE and ESV in Xinjiang, providing a scientific basis for regional low-carbon development and ecological protection decision-making. The global Moran’s I for bivariate analysis is used to measure the overall trend of spatial correlation between the two elements, CE and ESV, within the study area. The calculation formula is as follows [34]:

$$I_{xy}={\displaystyle \frac{n{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{ij}{x}_i{y}_j}{{\sum }_{i=1}^n{\sum }_{j=1}^n{W}_{ij}{\sum }_{j=1}^n{x}_i{y}_j}}$$

(8)

In the formula, n represents the total number of spatial units; W_ij is the element of the spatial weight matrix, indicating the degree of spatial association between spatial units i and j; x_i is the standardized value of carbon emissions for unit i; y_j is the standardized value of ecosystem service value for unit j.

The value range of the global Moran’s I for bivariate analysis is [−1, 1]. When I_xy > 0, it indicates a positive spatial association between CE and ESV, meaning that areas with high carbon emissions tend to be surrounded by areas with high ecosystem service values, or areas with low carbon emissions tend to be surrounded by areas with low ecosystem service values. When I_xy < 0, it indicates a negative spatial association between CE and ESV, meaning that areas with high carbon emissions tend to be surrounded by areas with low ecosystem service values, or areas with low carbon emissions tend to be surrounded by areas with high ecosystem service values. When I_xy = 0, it indicates no significant spatial association between CE and ESV, and their spatial distributions are independent of each other.

3. Results

3.1. Analysis of Changes in Land Area by Type in Xinjiang

Figure 2 and

Table 3. Changes of different types of land area in Xinjiang from 2000 to 2022 (km²).

Year	Cropland	Forestland	Grassland	Water Area	Built-Up Land	Unused Land
2000	61,125.813	14,614.704	392,647.892	8417.654	1146.668	1,153,772.571
2005	66,169.181	16,248.145	391,640.366	9635.414	2025.994	1,146,006.203
2010	76,751.651	17,220.843	387,883.126	9681.672	2999.759	1,137,188.251
2015	86,630.676	17,808.747	381,091.542	10,044.135	3863.093	1,132,287.107
2020	85,996.391	18,206.041	375,387.109	11,068.466	4947.595	1,136,119.701
2022	86,781.550	18,474.974	372,694.512	9592.942	5292.136	1,138,889.189

clearly depict the dynamic changes in the areas of the six LUCC types in Xinjiang from 2000 to 2022: cropland, forestland, grassland, built-up land, water area, and unused land. The analysis is conducted in three dimensions: total area, change trend, and magnitude of increase or decrease. This systematic analysis, considering the ecological and environmental characteristics of the arid and semi-arid regions of the study area and the intensity of human intervention, provides a foundational support for subsequent carbon emission accounting and ESV evolution analysis.

From the perspective of area structure, unused land accounted for the highest proportion among the LUCC types in Xinjiang during the study period, followed by grassland, with built-up land having the lowest proportion. Specifically, the area of unused land remained above 1.13 million km² from 2000 to 2022, making it the dominant land cover type in Xinjiang’s arid areas. The area of grassland, the second largest, stabilized between 370,000 and 390,000 km², forming the most important ecological cover base in the region. The areas of cropland, forestland, and water area decreased in size in descending order, with built-up land being the smallest but showing a significant expansion trend.

In terms of temporal evolution characteristics, there are distinct differences among various LUCC types. Forestland and built-up land showed a consistent growth trend, with forestland area increasing from 14,614.70 km² to 18,474.97 km² and built-up land expanding from 1146.67 km² to 5292.14 km². In contrast, the area of grassland continuously decreased from 392,647.89 km² to 372,694.51 km². Cropland, water area, and unused land generally exhibited a growth trend but experienced fluctuations during the study period. The area of cropland increased from 61,125.81 km² to 86,781.55 km², with a slight decline from 2015 to 2020. The water area increased to 11,068.47 km² from 2000 to 2020 and then experienced a slight decrease from 2020 to 2022.

In terms of the magnitude of increase and decrease, the expansion of built-up land was the most significant, with a growth rate of 361.49%. Cropland and forestland increased by 41.97% and 26.41%, respectively. Grassland experienced a cumulative decrease of 5.08%, being the only land type that continuously shrank during the study period. In contrast, the changes in unused land and water area were relatively minor. Overall, from 2000 to 2022, the adjustment of LUCC structure in Xinjiang exhibited a basic pattern of expansion in built-up land, cropland, water area, and forestland and a contraction in grassland.

3.1.1. Characteristics of LUCC Carbon Emissions

LUCC is a crucial carrier of regional carbon cycling, with different LUCC types exhibiting significant differences in their carbon source and sink, directly determining the regional carbon balance. Based on the LUCC area data and carbon emission accounting methods presented earlier, and combined with the carbon emission accounting results for Xinjiang from 2000 to 2022 in

Table 4. Carbon emission accounting results from LUCC in Xinjiang, 2000–2022 (10,000 tons).

Carbon Emission Type	2000	2005	2010	2015	2020	2022
Cropland	257.951	279.234	323.892	365.581	362.905	366.218
Forestland	−89.442	−99.439	−105.392	−108.990	−111.421	−113.067
Grassland	−82.456	−82.244	−81.455	−80.029	−78.831	−78.266
Water area	−19.781	−22.643	−22.752	−23.604	−26.011	−22.543
Unused land	−57.689	−57.300	−56.859	−56.614	−56.806	−56.944
Construction land	2770.491	4111.728	7557.907	13,146.511	17,756.967	22,547.798
Carbon source	3028.442	4390.962	7881.799	13,512.092	18,119.871	22,914.016
Carbon sink	−249.368	−261.627	−266.458	−269.237	−273.069	−270.821
Net carbon emission	2779.073	4129.336	7615.341	13,242.856	17,846.802	22,643.196

, this study systematically analyzes the evolution characteristics of carbon emissions from LUCC in Xinjiang from three dimensions: changes in net carbon emissions, composition of carbon sources and sinks, and the contribution of various LUCC types to carbon emissions.

Based on the data in Table 3, the net carbon emission from LUCC in Xinjiang showed a dramatic increase from 2000 to 2022, soaring from 27.7907 million tons in 2000 to 226.4319 million tons in 2022, a cumulative increase of approximately 7.15 times, with an annual growth rate as high as 10.2%. This growth rate far exceeds the national average during the same period, suggesting that regional economic activities and changes in land use in Xinjiang significantly disturbed the carbon balance during the study period. The net carbon emissions particularly jumped after 2010, rising from 76.1534 million tons to 220 million tons by 2022, reflecting that Xinjiang entered a period of high-intensity energy consumption and carbon emission growth during this stage.

Carbon emissions from built-up land are the core driving force for the sharp surge in net carbon emissions. The carbon emissions from built-up land stood at 27.7049 million tons in 2000 and soared to 225.478 million tons in 2022, accounting for an extremely high proportion of the total carbon sources. Their contribution rate rose further from 91.3% in 2000 to 99.6% in 2022, almost monopolizing the growth source of regional carbon emissions. This data characteristic profoundly reflects the rapid industrialization and urbanization process in Xinjiang over the past two decades. The construction of infrastructure, the expansion of the energy and chemical industries, and the extensive expansion of urban boundaries in Xinjiang have become the main engines for the growth of carbon emissions. In contrast, the proportion of carbon emissions from other types of LUCC is relatively small.

From the perspective of carbon sinks, the ecological land types including forestland, grassland and water areas exert a certain carbon sequestration effect, yet the growth rate of their carbon sequestration capacity lags far behind that of carbon sources. As the primary carbon sink, forestland saw its carbon sequestration volume rise from −0.8944 tons in 2000 to −1.1307 million tons in 2022, with an absolute increase of merely 0.2362 million tons. The growth in carbon sequestration volume of grassland and unused land is relatively slight. The carbon sink capacity of water areas fluctuates; it dropped after reaching −0.2601 million tons in 2020, which is associated with the shrinkage of water area coverage. The data results show that although Xinjiang’s fragile ecological environment alleviates carbon emission pressure to a certain extent, the potential for incremental carbon sequestration of its ecosystem is limited, making it unable to offset the emission increments brought by industrialization. A comparison of the evolution of carbon source and carbon sink trends reveals a widening gap between them. The total carbon source volume stood at 30.2844 million tons in 2000 and rose to 229.1402 million tons in 2022, an increase of 6.57 times, with built-up land being the absolute dominant force of carbon sources. In the same period, the total carbon sink volume only slightly increased from −2.4937 million tons to −2.7082 million tons, with a growth rate of less than 10%. During the period covered by the research, the total carbon sequestration volume in Xinjiang increased with fluctuations but remained limited in overall scale, failing to offset the rapid growth of carbon source emissions and thus leading to a continuous rise in regional net carbon emissions.

Overall, the formation of the carbon emission pattern from LUCC in Xinjiang during 2000–2022 is the combined result of the carbon source growth dominated by built-up land and the weak carbon sequestration of ecological land. The processes of industrialization and urbanization are the core driving factors. In the future, it is necessary to optimize the LUCC structure, strengthen the control of carbon emissions from built-up land, and improve the carbon sequestration capacity of ecological land to promote the balance of regional carbon budget.

3.1.2. Analysis of Land Use Intensity Change Characteristics

Figure 3 illustrates the spatial evolution of carbon emission intensity from LUCC in Xinjiang from 2000 to 2022. During the period of the study, the spatial pattern and temporal evolution of regional carbon emission intensity show distinct regularities. Overall, the carbon emission intensity of LUCC in Xinjiang consistently maintains a differentiated pattern of “high in the peripheral areas and low in the hinterland”. Extensive Gobi Desert areas, including the Tarim Basin and Junggar Basin, have long maintained carbon emission intensities below 0, which are typical low-emission zones in the region. In contrast, the oases and urban core areas on the northern and southern slopes of the Tianshan Mountains and the Ili River Valley show yellow, orange and even red patches, forming concentrated zones of high carbon emission intensity.

From a temporal perspective, high-value areas were only sporadically distributed in the core areas of a few cities such as Urumqi from 2000 to 2010, with an extremely small spatial scope. From 2010 to 2022, the scope of high-value areas expanded significantly, especially in cities such as Urumqi, Changji, Korla and Kashgar, as well as their surrounding oasis areas. The area of red patches continued to expand, which clearly reflects the evolution process in which the expansion of built-up land and the increase in human activity intensity directly drive the significant rise of regional carbon emission intensity with the acceleration of urbanization and industrialization. At the same time, the carbon emission intensity in the vast Gobi Desert areas always remained at a low level, revealing a high spatial coupling between carbon emission intensity from LUCC and human activity intensity in Xinjiang.

3.2. Temporal and Spatial Change Analysis of ESV in Xinjiang from 2000 to 2020

3.2.1. Temporal Change Trend Analysis of ESV in Xinjiang from 2000 to 2020

Based on the unit ESV in Table 2, the ESV of different land types in Xinjiang from 2000 to 2020 was calculated, as shown in

Table 5. ESV changes of different land types in Xinjiang from 2000 to 2022.

Year		Cropland	Forestland	Grassland	Water Area	Unused Land	Total
2000	Value (billion yuan)	598.131	767.177	12,962.811	2619.327	1857.839	18,805.284
	Contribution Rate (%)	3.181	4.080	68.932	13.929	9.879	100
2005	Value (billion yuan)	647.481	852.922	12,929.549	2998.258	1845.333	19,273.543
	Contribution Rate (%)	3.359	4.425	67.084	15.556	9.574	100
2010	Value (billion yuan)	751.033	903.982	12,805.508	3012.652	1831.134	19,304.309
	Contribution Rate (%)	3.890	4.683	66.335	15.606	9.486	100
2015	Value (billion yuan)	847.702	934.843	12,581.292	3125.440	1823.242	19,312.519
	Contribution Rate (%)	4.389	4.841	65.146	16.183	9.441	100
2020	Value (billion yuan)	841.495	955.698	12,392.967	3444.181	1829.413	19,463.755
	Contribution Rate (%)	4.323	4.910	63.672	17.695	9.399	100
2022	Value (billion yuan)	849.178	969.816	12,304.074	2985.042	1833.873	18,941.982
	Contribution Rate (%)	4.483	5.120	64.957	15.759	9.682	100

.

Based on the data in Table 5, the ESV of different types of LUCC in Xinjiang showed an overall fluctuating upward trend from 2000 to 2022. The ESV of Xinjiang was 1880.5284 billion yuan in 2000 and increased to 1894.1982 billion yuan in 2022, indicating an overall improvement in the regional ecosystem service value. From the perspective of temporal evolution, the ESV of Xinjiang showed a phased characteristic of first increasing and then decreasing from 2000 to 2022. It rose rapidly from 2000 to 2020 and declined after 2020. In terms of structural contribution, the ESV of Xinjiang is highly dependent on grassland and water areas, with their combined contribution rate exceeding 80%, which constitute the core support of the regional ecosystem service value. By type, the ESV of cropland continued to grow during the study period, increasing from 59.8131 billion yuan to 84.9178 billion yuan, and its contribution rate increased from 3.181% to 4.483% accordingly; both the ESV and contribution rate of forestland and water areas showed a fluctuating upward trend, with consistent change characteristics; the contribution rate of grassland ESV fluctuated between 60% and 70%, always occupying a dominant position; although the contribution rate of water areas fluctuated, it remained at a high level for a long time, serving as an important component of Xinjiang’s ESV; the ESV of unused land showed a downward trend, but the decline range was relatively limited.

Table 6. Trends in ecosystem service values (ESVs) of various services in Xinjiang, 2000–2022 (billion yuan).

Service Type		2000	2005	2010	2015	2020	2022
Provisioning Services	Food Production	42.5430	44.2101	46.9632	49.3973	49.1427	48.9379
	Grain Production	47.3960	47.9035	48.3169	48.3770	47.9408	47.7054
	Water Resource Supply	21.3451	22.2777	18.8192	16.0632	18.1261	14.7351
Regulating Services	Gas Regulation	167.2938	169.0283	170.5210	170.8097	169.2986	168.5081
	Climate Regulation	398.0771	400.8646	400.1501	396.2903	392.3781	389.6820
	Environmental Purification	190.2759	192.1679	191.5049	190.4181	190.5569	188.0505
	Hydrological Regulation	548.3727	581.7379	584.9689	593.6543	616.0184	577.4652
Supporting Services	Soil Conservation	194.0719	195.4845	195.8320	194.8062	193.0473	191.9742
	Biodiversity	16.2944	16.5479	16.8790	17.0848	16.9324	16.8732
Cultural Services	Aesthetic Landscape	175.9583	177.3761	176.9143	175.3916	174.2743	172.6180
Total		1880.5283	1927.3542	1930.4309	1931.2518	1946.3755	1894.1982

shows the change characteristics of the ESV of various ecosystem services in Xinjiang from 2000 to 2022. Based on the data, the total ESV of Xinjiang’s ecosystem services showed a trend of first increasing and then slightly declining. The total ESV was 1880.5283 billion yuan in 2000, then gradually rose to the peak of 1946.3755 billion yuan in 2020, and fell back to 1894.1982 billion yuan in 2022. Overall, it was still higher than that in 2000, highlighting the overall growth trend of regional ESV. The ESV of water temperature regulation and climate regulation was the highest, accounting for nearly 50% of the total.

There were significant differences in ESV changes among different types of ecosystem services, and the internal differentiation of supply services was obvious. The ESV of food production showed a trend of first increasing and then decreasing: it continued to grow from 42.5430 billion yuan in 2000 to 49.3973 billion yuan in 2015 and then slightly declined to 48.9379 billion yuan in 2022. The ESV of grain production fluctuated gently overall: it increased slightly to 48.3770 billion yuan from 2000 to 2015 and then gradually decreased to 47.7054 billion yuan in 2022, with a limited overall change range. The ESV of water resource supply showed a fluctuating downward trend, decreasing from 21.3451 billion yuan in 2000 to 14.7351 billion yuan in 2022, which was the sub-item with the most significant decline in supply services.

The ESV trends of each sub-item of regulating services varied. The ESV of gas regulation showed an overall trend of a slight fluctuating increase followed by a decline: it was 167.2938 billion yuan in 2000, increased to 170.8097 billion yuan in 2015, and fell back to 168.5081 billion yuan in 2022. The ESV of climate regulation showed a continuous fluctuating downward trend, gradually decreasing from 398.0771 billion yuan in 2000 to 389.6820 billion yuan in 2022, maintaining a long-term downward trend. The ESV of environmental purification fluctuated gently overall with a slight decrease, dropping from 190.2759 billion yuan in 2000 to 188.0505 billion yuan in 2022, with a small change range. The ESV of hydrological regulation continued to grow from 548.3727 billion yuan in 2000 to 616.0184 billion yuan in 2020 and then dropped sharply to 577.4652 billion yuan in 2022, which was an important sub-item affecting the fluctuation in the total ESV. Among supporting services, the ESV of soil conservation showed a slow fluctuating downward trend: it was 194.0719 billion yuan in 2000 and decreased to 191.9742 billion yuan in 2022. The ESV of biodiversity showed an overall trend of slight fluctuating increase, rising from 16.2944 billion yuan in 2000 to 16.8732 billion yuan in 2022, maintaining a steady growth overall. Among cultural services, the ESV of aesthetic landscape showed a continuous fluctuating downward trend, gradually decreasing from 175.9583 billion yuan in 2000 to 172.6180 billion yuan in 2022, maintaining a slow long-term downward trend.

3.2.2. Spatial Analysis of ESV Changes in Xinjiang from 2000 to 2020

Using the Fishnet tool in the ArcGIS 10.8, the ESV intensity for Xinjiang from 2000 to 2022 was obtained, and a spatiotemporal evolution map of ESV intensity from 2000 to 2022 was created for Xinjiang (Figure 4).

Figure 4 shows the perspective of spatial distribution and the ESV density in Xinjiang presents a significant pattern of “high in mountainous areas and low in basins”. High-value areas are mainly concentrated in mountainous regions such as the Altai Mountains, Tianshan Mountains, Kunlun Mountains and the Ili River Valley. These areas are dominated by forestland, grassland and water areas, with complex ecosystem structures, prominent functions of water conservation and biodiversity maintenance, and the highest ecosystem service value per unit area. Medium-value areas are distributed in the transition zones from mountainous areas to basins, mainly composed of grasslands and desert steppes, with moderate ecosystem service functions. Low-value areas cover large areas of Gobi Desert in the Tarim Basin and Junggar Basin, with extremely low vegetation coverage and the lowest ecosystem service value per unit area.

The temporal evolution perspective indicates the scope of high-value areas shows a slow expansion trend, especially in the northern slope of the Tianshan Mountains and the Ili River Valley, reflecting the gradual improvement in ecosystem service functions in these areas. The distribution of medium-value areas has also expanded, indicating that the ecosystem service value of some transition zones has been improved, while the area of blue desert low-value areas remained basically stable.

This spatial differentiation and temporal change not only reflect the high fragility of the ecosystem in the arid area of Xinjiang but also clarify the key direction of ecological protection. High-value areas need to focus on protecting the forestland, grassland and water area ecosystems to maintain key functions such as water conservation and climate regulation. The expansion trend of medium-value areas suggests that ecological restoration and protection measures can provide support for regional ecological security and sustainable development.

3.2.3. SI Analysis of ESV Intensity

The SI of ESV for different land types in Xinjiang was calculated using Formula (7), and the results are presented in

Table 7. ESV sensitivity coefficients of different land types in Xinjiang.

	Cropland	Forestland	Grassland	Water Area	Unused Land
CS	0.020	0.024	0.493	0.086	0.050

. According to the calculation results, all SI of ESV for land in Xinjiang are less than 1, indicating the high reliability of the results. Among them, the SI of cropland, forestland, water areas and unused land are relatively small, and their land ESV is also relatively small. The sensitivity coefficient of grassland is relatively large, and its land ESV is also relatively large.

3.3. Spatial Correlation Analysis Between LUCC Carbon Emission and ESV Intensity in Xinjiang

To analyze the spatial correlation of ESV intensity in Xinjiang, the ArcGIS Fishnet tool was used to grid the study area.

Table 8. Global Moran’s I of bivariate ESV intensity in Xinjiang from 2000 to 2020.

	2000	2005	2010	2015	2020	2022
Global Moran’s I	0.109	0.139	0.163	0.169	0.171	0.068
P	0.001	0.001	0.001	0.001	0.001	0.001
Z	27.803	48.741	57.163	58.621	59.392	25.093

shows the global spatial autocorrelation between LUCC carbon emission intensity and ESV intensity in Xinjiang from 2000 to 2022. During the study period, the global Moran’s I indices of Xinjiang were all positive, indicating a positive spatial correlation between LUCC carbon emission intensity and ESV intensity in Xinjiang; that is, they tend to agglomerate spatially. All p-values were 0.001, and all Z-values were much higher than the critical value, which indicates that the spatial correlation between LUCC carbon emission intensity and ESV intensity in Xinjiang was very significant during 2000–2022.

Figure 5 clearly reveals the local spatial correlation pattern and its dynamic evolution between LUCC carbon emission and ESV in Xinjiang from 2000 to 2022. From the perspective of spatial correlation types corresponding to colors, the local spatial correlation between LUCC carbon emission and ESV in Xinjiang is mainly dominated by “low–low” clusters and “low–high” clusters. The “high–high” cluster areas are mainly concentrated in ecological barrier zones such as the Tianshan Mountains and Altai Mountains, indicating that these areas have high local LUCC carbon emissions and high ESV in surrounding areas. These areas are rich in natural resources such as forests, grasslands and wetlands, which provide important ecological values for the ecosystem, such as water conservation, climate regulation and soil conservation. At the same time, the spatial overlap between high ecological value areas and high carbon emission areas caused by human activities interference indicates the contradiction and coordination needs between ecological protection and carbon emission control.

The “low–low” cluster areas mean low local carbon emissions and low ESV in surrounding areas, which are mainly distributed in the desert core areas of the Tarim Basin and Junggar Basin. The Tarim Basin and Junggar Basin are typical arid climate zones with scarce precipitation and high evaporation, leading to fragile ecological environments. Meanwhile, due to the low intensity of human activities, the carbon emission level is also low. The “low–high” clusters indicate low local carbon emissions but high ESV in surrounding areas, which are mostly located in the peripheral transition zones of high ecological value areas. This shows that these areas are radiated and influenced by the high ecological service value of surrounding areas, but their own carbon emissions have not yet increased significantly.

From the perspective of temporal evolution trends, the scope of the “high–high” cluster areas slightly decreases between 2000 and 2020. This is more obvious on the northern slope of the Tianshan Mountains. It reflects that with regional development, the carbon emission pressure in high ecological value areas increases. The scope of “low–low” cluster areas decreases to a certain extent. This indicates that the pattern of low ecosystem service value and low carbon emission in the desert core areas is further consolidated. The fragility of the ecosystem and the low intensity of human activities still continue. The distribution of “low–high” and “high–low” cluster areas fluctuates slightly. This shows that the correlation between carbon emission and ESV in ecological transition zones is in dynamic adjustment. Overall, this spatial correlation pattern not only reflects the spatial coupling characteristics between the ecosystem and human activities in Xinjiang. It also provides a spatial basis for the region to formulate differentiated dual-carbon goals and coordinated ecological protection strategies.

4. Discussion

4.1. The Relationship Between Carbon Emission and LUCC

From 2000 to 2022, the land use types in Xinjiang underwent significant changes, with substantial increases in the areas of cropland, forestland, and built-up land, while the areas of grassland and unused land decreased. This finding is consistent with previous studies that have revealed similar LUCC characteristics in arid regions of Central Asia [35]. The dynamic response between LUCC structure and carbon emissions in Xinjiang essentially reflects the universal law of human–environment system evolution. Multiple factors, including policy promotion, population growth, economic development, and ecological environmental demands, are intertwined and collectively shape the abovementioned characteristics of LUCC [36].

From 2000 to 2022, the permanent resident population in Xinjiang increased from 18.4941 million to 25.87 million, and the regional gross domestic product (GDP) soared from 136.5 billion yuan to 17,741.34 billion yuan. The continuous population growth and the rapid economic expansion have significantly increased the rigidity of regional food demand. This demand pressure has further transformed into the driving force for land resource development, leading to an expansion trend in the area of cultivated land. Meanwhile, accelerated urbanization has led to the continuous expansion of urban built-up areas, resulting in a substantial increase in built-up land, which has become one of the core driving forces of regional carbon emission growth. With the rapid expansion of the livestock industry, the number of pigs, cattle and sheep produced in Xinjiang has soared from 24.56 million (heads) in 2000 to 59.8565 million (heads) in 2022. This high-intensity growth of livestock has led to the severe overloading of grassland carrying capacity, breaking the balance between grass and livestock and intensifying the risk of grassland degradation, as well as causing a decrease in the coverage and reduction in the area of grassland vegetation. As an important carbon sink, grassland degradation weakens regional carbon sequestration capacity and indirectly exacerbates the carbon budget imbalance.

Xinjiang exhibits a distinctive pattern characterized by intensive development alongside high ecological vulnerability. Existing studies have shown that if regional land development expansion merely aims at economic growth as the sole objective, it is highly likely to cause the “carbon leakage” problem. That is, the carbon reduction achieved locally will often be offset by the additional carbon emissions generated by the development needs in other regions, ultimately failing to achieve the overall reduction effect [37]. Therefore, carbon mitigation in Xinjiang cannot rely solely on end-of-pipe measures; instead, it requires a fundamental restructuring of the carbon efficiency logic of land use at the source. At the same time, there is a need to formulate policy recommendations that are targeted, feasible, and regionally adaptable. First, the government should establish a carbon capture mechanism grounded in the optimization of the “three-life spaces” (production, living, and ecological spaces), leveraging the spatial heterogeneity of LUCC-driven carbon emissions and ESV. Specifically, in the production space, there is a need to promote low-carbon agricultural practices such as reducing the use of fertilizers and pesticides, adopting conservation tillage, and developing water-saving agriculture. These measures can effectively reduce the carbon emission intensity of farmland and enhance its potential carbon sink capacity. In the living space, there is a need to strictly control the disorderly expansion of urban and industrial land use, implement green building standards, optimize the low-carbon transportation system, and adjust the energy consumption structure by increasing the proportion of renewable energy, thereby reducing indirect carbon emissions. In the ecological space, there is a need to strengthen strict protection and ecological restoration to enhance its carbon absorption capacity and maintain the stability of the regional ecological carbon sink system.

Secondly, there is a need to promote low-carbon infrastructure construction that is adapted to the characteristics and resource endowments of the arid regions in Xinjiang. On one hand, there is a need to prioritize the development and utilization of Xinjiang’s abundant renewable energy resources, such as solar and wind energy, build large-scale solar and wind power generation bases, and promote the integration of renewable energy into the power grid. On the other hand, there is a need to address the water shortage problem that restricts the carbon sink capacity of the ecological space in arid regions, construct water-saving infrastructure for farmland and ecological restoration areas, improve water use efficiency, and ensure the water demand of vegetation growth in forests, grasslands, and water bodies.

Thirdly, there is a need to strictly regulate the disorderly expansion of built-up land and establish a land use quota management system closely linked to carbon emissions. Based on the carbon emission levels of different regions in Xinjiang, there is a need to formulate differentiated expansion quota standards for built-up land.

4.2. Changes in ESV

The total ESV in Xinjiang shows a fluctuating upward trend. Its core contribution comes from grassland and water area. The combined contribution of the two remains above 70% all year round, making them the core carriers of regional ecological security and ecological product supply. This is consistent with the relevant research results of the Xinjiang ESV [38]. As the most widely distributed ecosystem type in Xinjiang, grassland not only provides the material basis for the development of animal husbandry and supports regional economic growth but also plays an irreplaceable role in water conservation, soil and water conservation, climate regulation and biodiversity conservation. Its carbon sink function is an important support for the balance of regional carbon budget. The water area ecosystem provides sufficient water for agricultural irrigation, industrial production and residents’ lives. It also participates in regional water cycle regulation, purifies water quality, alleviates ecological pressure in arid and semi-arid areas, and ensures the stable operation of regional ecosystems.

LUCC exerts profound impacts on ESV through ecological land conversion, functional degradation, and landscape fragmentation. As the largest ecosystem, grasslands play a dual role. They serve as the material foundation for the development of animal husbandry and an important guarantee for regional economic stability, while also being a key barrier against desertification. Besides providing forage, grasslands are crucial for soil stability, hydrological regulation, and climate mitigation. Particularly significant is the carbon sequestration capacity formed by their deep root systems and soil organic carbon storage, which becomes a key buffer factor in the regional carbon budget and can offset the carbon emissions caused by human activities. Fluctuations in the water area intensify water resource constraints, reduce ecosystem resilience, and undermine the stable provision of regulating and supporting services. Meanwhile, the expansion of cropland and built-up land encroaches upon ecological space, exacerbating landscape fragmentation and weakening ecological connectivity, which further amplifies the spatial heterogeneity of ecosystem services. This process represents a common ecological constraint faced by arid regions worldwide during urbanization and agricultural expansion.

In the context of global sustainable development and ecological conservation, maintaining and enhancing ecosystem services in arid regions has become a key international research priority. As regional development continues to increase the demand for water and land resources, the conflict between ecological protection and economic development is likely to intensify. First, regarding spatial governance, it is imperative to rigorously enforce ecological red lines. Key ecological sources must be designated as prohibited or restricted development zones. Implementing the strictest possible land-use regulations is essential to curb the disorderly expansion and fragmentation of ecological land at its source. Second, concerning ecological restoration, strategies must transcend mere increases in vegetation coverage. Based on the principle of landscape connectivity, a network of ecological corridors should be established. Restoration efforts for degraded grasslands and water bodies should prioritize functional rehabilitation and ecosystem resilience rather than simply expanding spatial area. Third, in terms of institutional mechanisms, a robust framework for ecological compensation and the realization of ecosystem service values must be established. By leveraging fiscal transfer payments and green finance, the economic disparity between ecological protection zones and development-intensive areas can be mitigated, effectively alleviating the inherent conflict between conservation and economic growth.

4.3. Spatial Correlation Between LUCC Carbon Emission Intensity and ESV Intensity

According to the results, the carbon emission intensity and ESV intensity of LUCC in Xinjiang exhibit a significant negative spatial correlation, mainly characterized by “low–low” and “low–high” clustering patterns. In “low–low” clusters, the natural ecological baseline is fragile, with low vegetation cover and limited ecosystem service provision. Meanwhile, the intensity of human activities remains low, resulting in relatively low carbon emissions. However, due to the simplicity of the ecosystem structure, the overall service capacity is difficult to improve, forming a coupled pattern of low carbon emissions and low ESV. In contrast, “low–high” clusters benefit from relatively favorable water and soil conditions, which support more stable ecosystem structures and higher levels of ecosystem service provision. At the same time, human activity intensity in these areas remains moderate and has not exceeded ecological carrying thresholds, thereby maintaining a coordinated state of low carbon emissions and high ESV. When human activity intensity is low, carbon emissions remain limited, but ecosystem service provision is constrained by natural conditions. When activity intensity is moderate, and resource use is relatively efficient, a synergistic relationship between improved ecosystem services and controlled carbon emissions can be achieved. However, once development intensity exceeds the ecological threshold, the system is likely to shift toward a degraded trajectory characterized by high carbon emissions and low ESV.

Balancing carbon emission control, ecosystem service enhancement, and regional sustainable development remains a critical scientific issue and practical challenge in arid regions worldwide. In the future, differentiated regulatory mechanisms should be established for different spatial types. For “low–low” clusters (characterized by low ecosystem services and low carbon emissions), policies should prioritize degraded ecosystem rehabilitation and carbon sink enhancement. Furthermore, financial mechanisms like Payment for Ecosystem Services and green subsidies should be introduced to incentivize local stakeholders to transition from extensive land use to intensive ecological-friendly management, thereby catalyzing a shift toward high-level ecosystem service provision. For “low–high” clusters (low emissions but high ecosystem services), the primary policy goal is to consolidate these ecological strongholds. This necessitates establishing strict ecological red lines and access thresholds to prevent industrial encroachment. Policies should enforce a strictly limiting high-intensity development while promoting low-carbon industrial upgrading and eco-tourism. Moreover, institutional integration is crucial. Cross-regional collaborative governance mechanisms should be institutionalized to resolve spatial mismatches between resource supply and demand.

4.4. Research Limitations

Despite the valuable insights provided by this study, several limitations should be acknowledged.

First, the ecosystem service values were estimated using the equivalent factor method, which is a simplified approach based on average coefficients. Although this method is widely applied in large-scale assessments, it may not fully capture the spatial heterogeneity of ecosystem functions under different climatic, topographic, and socio-economic conditions in Xinjiang. The fixed coefficients could lead to uncertainties in representing local ecological processes, especially in highly fragmented oasis–desert transitional zones. Second, this study mainly focused on LUCC structure as the key explanatory factor, while other potential drivers such as climate change, population dynamics, economic development, and policy interventions were not explicitly incorporated into the analytical framework. These socio-economic and institutional factors may exert significant influences on ecosystem services and should be integrated in future studies using multi-variable models. Finally, in the carbon emission accounting process, given the limitations of research capabilities and resources, this study adopted a fixed carbon emission coefficient to conduct the accounting of carbon emissions from land use in Xinjiang. However, the complex geographical characteristics of Xinjiang make the relationship between land use activities and carbon emissions more complex and variable. Therefore, the carbon emission data obtained from this accounting may only to some extent reflect the relative changes caused by the adjustment of land use area in Xinjiang.

5. Conclusions

Based on the analysis of the temporal evolution and spatial characteristics of LUCC, carbon emission and ESV in Xinjiang from 2000 to 2022, combined with the research on their spatial correlation, the following conclusions are drawn:

(1)

From 2000 to 2022, the areas of cropland, forestland, water area and built-up land in Xinjiang showed an overall expanding trend, while those of grassland and unused land showed an overall shrinking trend. Over the 22 years, the area of cropland in Xinjiang increased by 41.9%, the area of forestland increased by 26.4%, the area of water area increased by 13.9%, and the area of built-up land increased by 361.5%. In contrast, the areas of grassland and unused land decreased by 5.0% and 1.2%, respectively.

(2)

From 2000 to 2022, the total LUCC carbon emission in Xinjiang showed a continuous growth trend, with an average annual growth rate exceeding 10%. The carbon emission increment caused by rising energy consumption from built-up land expansion is the most prominent, which established it as the primary regional carbon source. Grassland, forestland and water area are important carbon sink carriers. During the research period, the carbon sink functions of forestland and water area continued to increase, but restricted by the continuous reduction in grassland, the carbon sink function of grassland continued to decrease. In addition, the total LUCC carbon emission in Xinjiang continued to grow from 2000 to 2022, with an average annual growth rate of more than 10%. The conversion of LUCC types and the evolution of carbon sink and source functions are the core driving factors. Among them, the rising energy consumption caused by built-up land expansion leads to the most significant carbon emission increment, becoming the primary regional carbon source. Grassland, forestland and water area are all important regional carbon sink carriers. During the research period, the carbon sink functions of forestland and water area continued to improve, while the grassland area continued to shrink, and the grassland carbon sink function continued to weaken.

(3)

From 2000 to 2022, the LUCC ESV in Xinjiang showed a fluctuating upward trend, increasing from 1880.5284 billion yuan in 2000 to 1894.1982 billion yuan in 2022. In terms of the contribution rate, Xinjiang’s ESV is mainly composed of grassland and water area, with their combined contribution rate ranging from 70% to 80%. The contribution rate of unused land is about 10%, while the contribution rates of cropland and forestland are relatively low.

(4)

From 2000 to 2022, the ESV intensity in Xinjiang showed an overall characteristic of being higher in the west than in the east and higher in the north than in the south, and the overall intensity increased over time. The high-value areas are mainly concentrated in mountains such as the Altai Mountains, Tianshan Mountains and Kunlun Mountains, as well as the Ili River Valley. The medium-value areas are distributed in the transition zone from mountains to basins. The low-value areas cover large areas of Gobi Desert in the Tarim Basin and Junggar Basin.

(5)

The spatial correlation between LUCC intensity and ESV intensity in Xinjiang is significant, showing an overall positive correlation. It mainly forms two patterns: “low–low” agglomeration and “high–low” agglomeration. This spatial correlation provides a scientific spatial basis for formulating differentiated and detailed strategies for land use regulation, carbon emission reduction, and ecological protection in the future. Furthermore, it plays a significant role in promoting the coordination of human–land systems and advancing regional sustainable development in arid areas.