Analysis of Spatio-Temporal Dynamics and Determinants of Land Use Carbon Emissions in Jiangsu Province, China

Abstract

This study investigates the changes in carbon emissions resulting from land use in Jiangsu Province, China, over the past two decades. It constructs a land use transition matrix to analyze spatio-temporal dynamics, applies the carbon emission coefficient method to calculate land use carbon emissions, and employs the Geographically and Temporally Weighted Regression to investigate the determinants of land use carbon emissions. The findings reveal the following: (1) From 2000 to 2020, construction land in Jiangsu Province, China, exhibited a significant increasing trend, primarily due to the conversion of cultivated land. Meanwhile, the area of other land use types decreased. (2) The primary sources of carbon emissions are construction land and cultivated land. Between 2000 and 2020, the total carbon emissions from land use in Jiangsu Province, China, showed a clear upward trend. However, between 2015 and 2020, these emissions gradually stabilized. (3) Economic development level and consumption level have significant positive effects on land use carbon emissions, while the technological level has a significant negative effect.

1. Introduction

In recent decades, rapid economic development has led to a significant increase in greenhouse gas emissions, making global climate change and carbon emissions a major concern worldwide. As the world’s largest carbon emitter [1], China proposed at the 2015 Paris Climate Conference the goal of striving to peak carbon dioxide emissions by 2030 [2]. To implement the “Dual Carbon Goals” (peaking carbon emissions by 2030 and achieving carbon neutrality by 2060), the 2020 Central Economic Work Conference emphasized the need to accelerate the optimization of industrial and energy structures, improve land use allocation, promote nationwide greening initiatives, and advance green, low-carbon, and circular development to a higher level. These efforts aim to achieve synergistic progress between land resource utilization and the dual carbon goals.

Land, as the foundation of human survival and development, is an indispensable and irreplaceable component of terrestrial ecosystems. Recent research has revealed that land use and management practices are significant contributors to greenhouse gas emissions [3,4,5]. Land use changes not only alter surface cover types and ecosystem structures but also profoundly impact soil carbon sequestration and carbon cycling mechanisms within the atmospheric system [6,7,8,9]. Different land types exhibit distinct capacities for carbon sequestration and release, and land use changes have a direct and substantial influence on carbon cycling processes.

Land use change is a primary driver of carbon accumulation [10,11]. Research by scholars on land use carbon emissions primarily includes the estimation of land use carbon emissions, the analysis of spatio-temporal variations in land use carbon emissions [12], the study of influencing factors of land use carbon emissions, and the prediction of land use carbon emissions [13], among others. A more traditional approach to estimating land use and land cover carbon emissions is through the view and Bookkeeping Model. This model integrates land use and land cover change data with carbon density data, determining land areas in different regions and assigning initial biomass and soil carbon densities to various ecosystems. It then simulates disturbance activities using statistical data to alter carbon densities, thereby calculating carbon pools [14,15]. While this method considers the impact of forest changes on land carbon emissions and sequestration, it neglects the feedback between CO₂, climate, and land use carbon emissions. The Bern Carbon Cycle–Climate Model divides grid cells into portions corresponding to different land types, simulating forestry activities and the feedback between land and the atmosphere to redistribute CO₂ within the Earth system, thereby quantifying the impact of land use on carbon fluxes [16]. The most commonly used method is the emission factor approach based on the IPCC guidelines, which quantifies the carbon stocks of different land types to determine their annual net carbon fluxes. Specific values of annual net carbon fluxes are then applied to the areas of corresponding land cover categories to derive carbon emissions for each land cover type [17,18,19]. This study employs this method to calculate land use carbon emissions and further analyzes their spatio-temporal dynamics.

The methods for studying influencing factors are diverse. The STIRPAT model incorporates the concept of ecological elasticity and integrates multidimensional variables such as urbanization and industrial structure, enabling the assessment of nonlinear relationships between population, economy, technology, and their impacts on CO₂ emissions and carbon footprints [20]. Multiple Linear Regression (MLR) is an efficient model for verifying the relationship between carbon emissions and various urbanization parameters, considering the influence of human activities and multiple factors in the urbanization process on land use carbon emissions [18]. While MLR is simple in form and easy to understand and implement, its assumption of a linear relationship between dependent and independent variables is overly rigid, and it ignores spatial relationships between data points. The log-averaged D-type exponential decomposition (LMDI) model decomposes carbon emissions to identify influencing factors, analyzes the contribution of each factor to carbon emissions, and explains the reasons for changes in carbon emissions [21]. Traditional models in previous studies have overlooked spatial heterogeneity, making it difficult to reveal regional differences in influencing factors. The Geographically Weighted Regression (GWR) model fits local regression models for each spatial location and introduces a spatial weight matrix, allowing regression coefficients to vary with spatial location. This enables the analysis of the driving effects of various influencing factors by incorporating spatial information on land [22].

The existing literature on land use carbon emissions has primarily focused on influencing factors such as urbanization and economic development. This study employs the Geographically and Temporally Weighted Regression (GTWR) to comprehensively examine the impacts of multiple factors, including the economy, population, land use, industry, technology and consumption on land use carbon emissions. Compared to the GWR model, the GTWR model not only incorporates spatial location effects but also introduces a temporal dimension, applying spatio-temporal weighting to the relationships between explanatory and response variables. This improvement significantly enhances the model’s ability to capture the dynamic changes of variables over time and space, allowing for a more accurate reflection of the relationships among variables under different spatio-temporal contexts. It is, therefore, more suitable for analyzing correlations in geographic phenomena. Through the integrated application of the land use transition matrix, the carbon emission coefficient method, and the GTWR model, this study systematically investigates the spatio-temporal evolution of land use structure, the spatial and temporal variations in land use carbon emissions, the analysis of driving forces, and corresponding decision-making support. This comprehensive approach offers a more complete framework for studying land use carbon emissions in contrast to existing research that often focuses on only one of these aspects.

With the rapid development of the global economy and continuous population growth, the conversion of ecological land and the expansion of urban construction land have intensified significantly. Large areas of land have been allocated for urban development and industrial use, resulting in substantial changes in land use patterns. These anthropogenic activities have not only altered land use types but also disrupted the dynamic balance between carbon sequestration and emissions. Benefiting from a mild climate, fertile land, and a solid economic foundation, Jiangsu Province, China, has experienced rapid economic development over the years, which has led to high carbon emissions and increasingly intensive land use. The development of industrial production, construction, and related fields has generated substantial demand for land and ecological resources, accelerating the exploitation and utilization of land resources and leading to land use migration and transformation. As an economically powerful province in eastern China, Jiangsu’s significant scale of energy consumption and carbon emissions makes the low-carbon transformation of its land resources critical for achieving the province’s overarching strategic goals of carbon peak and carbon neutrality. Based on the current land use situation in Jiangsu Province, the following research goals are proposed: (1) reveal the spatial-temporal dynamics of land use structure, (2) analyze the spatial and temporal differences in land use carbon emissions, and (3) explore the determinants of land use carbon emissions. Furthermore, the study aims to clarify the sources and pathways of carbon emissions, providing a scientific basis for formulating land use policies and carbon reduction measures, thereby achieving a win–win scenario for economic development and carbon reduction.

The main contributions of this study are: (1) The study introduces a comprehensive methodological framework that combines a land use transition matrix, carbon emission coefficient method, and the Geographically and Temporally Weighted Regression (GTWR) model. This framework systematically investigates the spatio-temporal dynamics of land use structure, quantifies carbon emissions, and analyzes driving factors across both spatial and temporal dimensions. (2) The research identifies construction land expansion (primarily converted from cultivated land) and economic development as the dominant drivers of carbon emissions in Jiangsu Province, China. It quantifies the significant positive impacts of economic growth and consumption levels alongside the mitigating role of technological advancement. These findings offer actionable insights for optimizing land use planning, accelerating energy structure transitions, and promoting green industrial development, thereby supporting Jiangsu’s “Dual Carbon Goals” and providing a scientific basis for regional low-carbon policy formulation.

2. Materials and Methods

2.1. Overview of the Study Area

Jiangsu Province is located in the central part of the eastern coastal region of mainland China and is a key component of the Yangtze River Delta (Figure 1). It spans 30°45′ N to 35°08′′ N and 116°21′ E to 121°56′ E. The province covers a land area of 107,200 square kilometers, a sea area of 37,500 square kilometers, and a coastline stretching 954 km. The terrain of Jiangsu Province is predominantly flat, generally characterized by higher elevations in the north and south, lower elevations in the middle, and a gradual slope from west to east. The landscape is primarily composed of plains, with low mountains and hills mainly concentrated in the southwest.

Jiangsu Province has long been dominated by energy-intensive industries, with the manufacturing sector maintaining rapid growth. Energy consumption is primarily based on coal, making it one of the largest provinces in China in terms of energy consumption and carbon emissions. Carbon emissions from high-energy-consuming industries such as building materials, chemicals, and steel are increasingly rising.

2.2. Research Data

The data overview is presented in

Table 1. Data overview.

Data Type	Data Sources	Spatial Resolution	Temporal Resolution
The land use data	https://zenodo.org/records/8176941 (accessed on 20 January 2024)	30 m	1 year
The administrative boundary vector data	http://bzdt.ch.mnr.gov.cn/ (accessed on 20 January 2024)	/	1 year
The data on energy consumption	(1) https://www.jiangsu.gov.cn (accessed on 20 October 2024) (2) https://www.cnki.net/ (accessed on 20 October 2024)	/	1 year

. The land use data originate from the 30 m annual land cover datasets, and their dynamics in China are created by the research team led by Yang and Huang from Wuhan University, based on Google Earth Engine and Landsat data [23]. Five time points were selected: 2000, 2005, 2010, 2015, and 2020. Referring to the national standard of China, “Current land use classification” (GB/T 21010-2017) [24], the land use types are reclassified into cultivated land, forest land, grassland, water area, construction land, and unused land. The reclassification was performed using ArcGIS 10.7.

The vector data of Jiangsu Province’s administrative boundaries are produced based on the standard map with the map review number GS(2019)1822 downloaded from the Standard Map Service Website of the China Ministry of Natural Resources. The boundary of the base map remains unmodified.

The data on energy consumption in Jiangsu Province were estimated following the”2006 IPCC Guidelines for National Greenhouse Gas Inventories” and the “General rules for calculation of the comprehensive energy consumption” (GB/T 2589-2020) [25]. Socioeconomic development and energy consumption statistics were sourced from the “Jiangsu Statistical Yearbook”, “Jiangsu Informatization Yearbook”, “China Urban Statistical Yearbook”, and statistical yearbooks of various cities in Jiangsu Province. After collection, the data were organized. For years with missing values, interpolation methods were applied to complete the dataset.

2.3. Research Methodology

This study constructs an analytical framework for the spatio-temporal evolution and driving factors of land use carbon emissions, following a chain of ‘data fusion, model construction, factor analysis, and decision support’. The technical workflow is illustrated in Figure 2.

2.3.1. The Land Use Transition Matrix

The land use transition matrix is a method used to describe the mutual conversion of land types. It quantifies the area changes and directions between different land use types by analyzing the transformation of various land use categories within the same region over different time periods, resulting in a two-dimensional matrix [26,27]. The mathematical expression is as follows:

$$\mathrm{S}=\left(\begin{array}{ccc}{S}_{11}& \cdots & S_{1n}\\ ⋮& \ddots & ⋮\\ S_{1n}& \cdots & S_{nn}\end{array}\right)$$

(1)

where S represents the area of land use, and n represents the number of land use types.

2.3.2. Carbon Emission Coefficient Method

The carbon emission coefficient method is a technique used to estimate the emissions of greenhouse gases such as CO₂. Carbon emissions from land use are categorized into direct and indirect carbon emissions [28]. Direct carbon emissions refer to carbon emissions generated due to changes in land types and land management methods. Indirect carbon emissions, on the other hand, are calculated based on statistical data related to human production and activities (such as energy consumption data, industrial data, and sectoral data) and represent the carbon emissions generated by human activities associated with different land use types.

In this study, the carbon emissions from cultivated land, forest land, water area, grassland, and unused land are estimated using the direct carbon emission coefficient method. The calculation formula is as follows [29]:

$$Q=\sum q_i=\sum A_i\times {\lambda }_i$$

(2)

where $Q$ represents the total carbon emissions. For the i-th land use type, $q_i$ denotes the carbon emissions (or absorption) generated. $A_i$ is the area of that land use type, and ${\lambda }_i$ is the corresponding carbon emission (or absorption) coefficient, where a negative value indicates absorption and a positive value indicates emissions.

For carbon emissions from construction land, the indirect carbon emission coefficient method is adopted. This method indirectly estimates carbon emissions by calculating the carbon emissions generated from energy consumption during the utilization of construction land. Specifically, the consumption of energy sources such as petroleum, natural gas, and coal is converted into an equivalent amount of standard coal, which is then used to indirectly estimate the carbon emissions from construction land. The calculation formula is as follows [29]:

$$Q={\sum }_{i=1}^n{\alpha }_i{\beta }_i{\gamma }_i$$

(3)

where $Q$represents the carbon emissions from construction land; $n$ denotes the total number of energy types; for the i-th energy type, ${\alpha }_i$ represents its consumption; ${\beta }_i$ is its standard coal conversion coefficient, and ${\gamma }_i$ is the carbon emission coefficient of the energy type calculated in terms of standard coal.

2.3.3. The Geographically and Temporally Weighted Regression

The Geographically and Temporally Weighted Regression (GTWR) model represents an advancement that has evolved from the Geographically Weighted Regression (GWR) model and panel data models. Panel data models analyze data by combining time series and cross-sectional observations (i.e., multiple periods of data from different individuals, regions, or units). They focus on temporal variation and individual-level differences but typically do not account for spatial effects. In contrast, the GWR model allows regression coefficients to vary across spatial locations. The GTWR model further extends this variability to both temporal and spatial dimensions, enabling coefficients to dynamically adjust over time and space. As a result, the GTWR model offers a more refined and accurate means of characterizing the intricate spatio-temporal relationship between variables and the dependent variable [30].

By incorporating both temporal and spatial dimensions into the regression framework [31], the GTWR model enables researchers to conduct a more precise analysis of the driving factors behind land use carbon emissions. This model assigns specific weights based on the characteristics and contributions of various influencing factors, effectively reflecting the combined effects of spatial and temporal elements. The corresponding expression is as follows [32]:

$$y_i={\beta }_0\left(u_i,v_i,t_i\right)+{\sum }_{k=1}^p{\beta }_k\left(u_i,v_i,t_i\right)x_{ik}+{\epsilon }_i$$

(4)

where $y_i$ represents the explained variable at the i-th sample point; $u_i$ denotes the longitude position of the i-th sample point; $v_i$ represents the latitude position of the i-th sample point; $t_i$ represents the time coordinate of the i-th sample point. ${\beta }_0\left(u_i,v_i,t_i\right)$ constitutes the spatio-temporal coordinates of the i-th sample point. ${\beta }_k\left(u_i,v_i,t_i\right)$ represents the constant value at the i -th sample point, and it is a fixed constant in the model; $x_{ik}$ serves as the k-th regression parameter value in the geographically and temporally weighted regression for the i-th sample point. $x_k$ represents the value of the i-th independent variable (free from interference), which is the value of each quantitative index at the point. ${\epsilon }_i$ is the random error.

3. Results

3.1. The Evolution Process of Land Use

3.1.1. Characteristics of Quantitative Changes in Land Use

Table 2. Land use area and proportion in Jiangsu Province from 2000 to 2020.

Year	Unit	Cultivated Land	Forest Land	Grassland	Water Area	Unused Land	Construction Land	Total
2000	Area (km2)	77,851.13	2034.06	36.49	11,697.82	7.19	10,345.07	101,971.77
	Proportion (%)	76.35	1.99	0.04	11.47	0.01	10.15	100.00
2005	Area (km2)	75,423.09	1936.46	29.37	12,616.84	2.86	11,963.15	101,971.77
	Proportion (%)	73.96	1.90	0.03	12.37	0.00	11.73	100.00
2010	Area (km2)	73,207.04	1883.17	37.61	12,590.63	2.52	14,250.80	101,971.77
	Proportion (%)	71.79	1.85	0.04	12.35	0.00	13.98	100.00
2015	Area (km2)	70,706.80	1696.36	11.80	12,342.91	1.75	17,212.14	101,971.77
	Proportion (%)	69.34	1.66	0.01	12.10	0.00	16.88	100.00
2020	Area (km2)	70,385.61	1646.72	3.74	11,318.03	1.11	18,616.55	101,971.77
	Proportion (%)	69.02	1.61	0.00	11.10	0.00	18.26	100.00

reflects the land use situation and the changing trends in Jiangsu Province from 2000 to 2020. The dominant land use type in Jiangsu Province is cultivated land, accounting for approximately 70% of the total land area. From 2000 to 2020, the areas of cultivated land, forest land, grassland, and unused land generally showed a downward trend. The area of water areas first increased and then decreased, while the area of construction land exhibited a consistent upward trend.

3.1.2. The Land Use Transfer Matrix

In this study, GIS technology was employed to conduct spatial overlay analysis on land use maps corresponding to different year nodes. Map algebra was applied to calculate the transition values of land use maps during the period between two specific year nodes. The results are presented in

Table 3. Land use transition matrix of Jiangsu Province (2000–2005) (km²).

Year 2000	Year 2005
	Grassland	Cultivated Land	Construction Land	Forest Land	Water Area	Unused Land	Transferred-Out
Grassland	21.35	10.68	3.48	0.53	0.36	0.10	15.15
Cultivated Land	7.71	74,730.32	1724.86	112.83	1275.39	0.01	3120.8
Construction Land	0.00	10.36	10,111.95	0.00	222.76	0.00	233.12
Forest Land	0.09	200.07	12.30	1821.27	0.34	0.00	212.8
Water Area	0.15	470.59	109.80	1.82	11,115.16	0.30	582.66
Unused Land	0.08	1.07	0.77	0.00	2.83	2.44	4.75
Transferred-in	8.03	692.77	1851.21	115.18	1501.68	0.41

,

Table 4. Land use transition matrix of Jiangsu Province (2005–2010) (km²).

Year 2005	Year 2010
	Grassland	Cultivated Land	Construction Land	Forest Land	Water Area	Unused Land	Transferred-Out
Grassland	19.69	4.76	3.51	0.69	0.14	0.58	9.68
Cultivated Land	17.28	72,295.30	2157.15	136.02	817.29	0.05	3127.79
Construction Land	0.00	3.51	11,838.02	0.00	121.63	0.00	125.14
Forest Land	0.08	181.97	9.94	1744.12	0.34	0.00	192.33
Water Area	0.54	721.22	241.56	2.34	11,650.78	0.41	966.07
Unused Land	0.01	0.28	0.63	0.00	0.45	1.48	1.37
Transferred-in	17.91	911.74	2412.79	139.05	939.85	1.04

,

Table 5. Land use transition matrix of Jiangsu Province (2010–2015) (km²).

Year 2010	Year 2015
	Grassland	Cultivated Land	Construction Land	Forest Land	Water Area	Unused Land	Transferred-Out
Grassland	9.79	8.69	17.56	1.42	0.03	0.11	27.81
Cultivated Land	1.85	69,663.42	2711.41	90.48	739.87	0.00	3543.61
Construction Land	0.00	0.52	14,180.75	0.00	69.52	0.00	70.04
Forest Land	0.01	271.94	6.63	1604.09	0.50	0.00	279.08
Water Area	0.05	761.99	294.97	0.37	11,532.89	0.36	1057.74
Unused Land	0.10	0.24	0.81	0.00	0.09	1.28	1.24
Transferred-in	2.01	1043.38	3031.38	92.27	810.01	0.47

and

Table 6. Land use transition matrix of Jiangsu Province (2015–2020) (km²).

Year 2015	Year 2020
	Grassland	Cultivated Land	Construction Land	Forest Land	Water Area	Unused Land	Transferred-out
Grassland	2.94	5.24	2.37	1.15	0.01	0.09	8.86
Cultivated Land	0.59	68,917.53	1272.39	116.51	399.79	0.00	1789.28
Construction Land	0.00	2.22	17,144.06	0.00	65.86	0.00	68.08
Forest Land	0.19	164.50	2.46	1528.89	0.32	0.00	167.47
Water Area	0.00	1295.81	194.62	0.16	10,852.02	0.30	1490.89
Unused Land	0.02	0.32	0.66	0.00	0.04	0.72	1.04
Transferred-in	0.8	1468.09	1472.5	117.82	466.02	0.39

. Subsequently, a more in-depth analysis of the land use change process was conducted [33].

From 2000 to 2005, the total areas of cultivated land, grassland, forest land, and unused land decreased, while the total areas of construction land and water areas increased. The most significant change was observed in the area of cultivated land, primarily due to its conversion into construction land for production and construction activities. The conversion of cultivated land into water areas ranked second.

From 2005 to 2010, the total areas of cultivated land, forest land, water areas, and unused land decreased, while the total areas of construction land and grassland increased. The most remarkable change was the conversion of cultivated land into construction land. Forest land and water areas were mainly converted into cultivated land. Although the area of grassland increased, most of the increase came from the conversion of cultivated land. The area of water areas decreased, but the overall change was not significant.

From 2010 to 2015, the total areas of cultivated land, grassland, forest land, water areas, and unused land decreased, while the total area of construction land increased, mainly through conversion from cultivated land.

From 2015 to 2020, the total areas of cultivated land, grassland, forest land, unused land, and water areas decreased, while the total area of construction land increased. Most of the transferred-out cultivated land still flowed to construction land, followed by water areas and forest land. The transferred-out areas of forest land, grassland, and water areas also decreased. The slowdown in the reduction rate of green land was closely related to the implementation of policies such as the transfer of farmland back to forests, environmental protection, and ecological restoration.

Based on the above analysis, from 2000 to 2020, construction land in Jiangsu Province showed a significant increasing trend, mainly transformed from cultivated land. The area of cultivated land decreased, the area of water areas first increased and then decreased, showing an overall downward trend, and the areas of grassland, forest land, and unused land also decreased.

3.2. Spatio-Temporal Evolution of Carbon Emissions from Land Use

3.2.1. Direct Carbon Emissions from Land Use

Referring to existing research findings [34,35], the selected carbon emission coefficients for each land use type are presented in

Table 7. Direct carbon emission coefficients of land use.

Land Use Type	Cultivated Land	Forest Land	Grassland	Water Area	Unused Land
carbon emission coefficient (t/km2)	45.9	−60.5	−2.1	−25.4	−0.5

.

3.2.2. Indirect Carbon Emissions from Land Use

The indirect carbon emission coefficient method was used to calculate the carbon emissions generated by construction land. Based on the “2006 IPCC Guidelines for National Greenhouse Gas Inventories” and the “General rules for calculation of the comprehensive energy consumption” (GB/T 2589-2020), the standard coal conversion coefficients and energy carbon emission coefficients are shown in

Table 8. Carbon emission parameters for diverse energy sources.

Energy	Coefficient for Conversion to Standard Coal (kgce/kg)	Carbon Emission Coefficient (t/t)
Raw coal	0.7143	0.7559
Coke	0.9714	0.8550
Crude oil	1.4286	0.5857
Petrol	1.4714	0.5538
Kerosene	1.4714	0.5714
Diesel	1.4571	0.5921
Fuel oil|	1.4286	0.6185
Liquefied petroleum gas (LPG)	1.7143	0.5042
Coal gas	0.3571 kgce/m3	0.3548
Natural gas	1.1–1.33 kgce/m3	0.4483
Electricity	0.1229 kgce/kw·h	0.2132

.

3.2.3. Changes in the Total Carbon Emissions from Land Use in Jiangsu Province

Based on the land use status and energy consumption data of Jiangsu Province from 2000 to 2020, the carbon emission coefficient method was applied to calculate the carbon emissions. The specific results are presented in

Table 9. Land use carbon emissions in Jiangsu Province from 2000 to 2020 (10⁴ Tons).

Year	2000	2005	2010	2015	2020
Cultivated Land	357.34	346.19	336.02	324.54	323.07
Construction Land	6503.05	12,562.78	18,664.45	21,536.05	21,990.21
Forest Land	−12.31	−11.72	−11.39	−10.26	−9.96
Grassland	−0.0077	−0.0062	−0.0079	−0.0025	−0.0008
Water Area	−29.71	−32.05	−31.98	−31.35	−28.75
Unused Land	−0.00036	−0.00014	−0.00013	−0.00009	−0.00006
Total Carbon Emissions	6818.36	12,865.2	18,957.09	21,818.98	22,274.57

.

As shown in Table 9, from 2000 to 2020, the carbon emissions from land use in Jiangsu Province increased year by year. Among them, carbon emissions from cultivated land decreased year by year, while carbon emissions from construction land gradually increased. The growth rate of carbon emissions in Jiangsu Province varied significantly across different time periods. Generally, it showed a trend of rapid increase initially, followed by slower growth, with the growth rate gradually slowing down. By 2020, carbon emissions from land use were approaching stability. Construction land is the primary source of carbon emissions, accounting for more than 95% of the total carbon emissions in Jiangsu Province. Forest land and water areas are important carbon sinks in Jiangsu Province, but their overall carbon absorption showed a downward trend.

The main reasons for the annual increase in carbon emissions from land use in Jiangsu Province are the substantial expansion of construction land, increased energy consumption, and intense carbon emission activities in production and construction. In the early stages, the urbanization process was relatively rapid, and economic development based on land use was highly dependent on energy consumption. In the later stages, with the rapid development of science and technology and the implementation of a series of low-carbon and emission–reduction policies, the growth rate of carbon emissions from land use slowed down.

3.2.4. Changes in Carbon Emissions from Urban Land Use

For the carbon emissions from land use in each city, the carbon emission coefficient method was also adopted to obtain the data for 2000, 2005, 2010, 2015, and 2020. The results are shown in Figure 3.

During the study period, the carbon emissions from land use at the municipal scale in Jiangsu Province generally showed a steadily increasing trend. Carbon emissions increased from 2000 to 2015, with the growth rate gradually slowing down. From 2015 to 2020, carbon emissions tended to stabilize, and some cities even witnessed a downward trend. In the initial stage of the research, due to the rapid economic development in Jiangsu Province, the demand for resources in production and daily life increased, leading to a rise in energy consumption and, thus, an increase in carbon emissions from land use. In the later stage of the research, influenced by national policies, enterprises accelerated their transformation towards green production, resulting in a slowdown in the growth rate of carbon emissions.

From a spatial distribution perspective (Figure 4), the carbon emissions from land use at the municipal scale in Jiangsu Province generally exhibit a pattern of higher values on both sides and lower values in the middle. In northern Jiangsu, Xuzhou, in particular, has relatively high carbon emissions. Xuzhou serves as a crucial coal production area and power base in East China. As a resource-based city, it is endowed with abundant coal reserves and a developed heavy industry sector. In southern Jiangsu, the high-level economic development and rapid economic growth have led to a continuously increasing demand for energy. Moreover, with a relatively high level of urbanization, a dense urban population, and substantial demands for residents’ production and daily life, the carbon emissions have been further augmented.

3.3. Analysis of Driving Factors for Land Use Carbon Emissions

In this study, the GTWR model was employed to investigate the carbon emissions from land use in Jiangsu Province. Specifically, data pertaining to the geographic locations and various influencing factors within the jurisdiction of each city were meticulously collected and systematically organized into a panel dataset. Before proceeding with the regression analysis, multicollinearity diagnostics were performed on the selected influencing factors using SPSS Statistics 26. The results (

Table 10. VIF for influencing factors.

Influencing Factors	VIF
Population size	2.303
Economic level	3.539
Industrial structure	1.345
Land use structure	2.318
Technological level	6.800
Consumption level	7.567

) showed that the Variance Inflation Factor (VIF) values for the explanatory variables were all less than 10, indicating that there was no multicollinearity among the selected variables. Subsequently, the data underwent Z-Score standardization. This process effectively normalizes data of varying magnitudes onto a uniform scale, thereby ensuring comparability among different data points. It also eliminates potential analytical biases and obstacles that may arise from discrepancies in data magnitude, thereby enhancing the accuracy and robustness of the regression analysis.

The R², AICc, and RSS of the GTWR model were selected as evaluation metrics for model performance (

Table 11. Model evaluation metrics.

Model Metrics	GTWR Model Metric Parameters
R2	0.85
AICc	94.85
RSS	9.43

). A higher R² value indicates a better model fit, while lower AICc and RSS values suggest superior model performance.

3.3.1. Selection of Influencing Factors

Drawing on the research findings of previous scholars and considering the specific context of Jiangsu Province, this study selects six key indicators—population size, economic level, industrial structure, land use structure, technological level, and consumption level—as the explanatory variables. Meanwhile, carbon emissions from land use are designated as the explained variable for the geographically and temporally weighted regression analysis. The results are shown in Figure 5, Figure 6, Figure 7, Figure 8, Figure 9 and Figure 10.

1. Population size: It refers to the total number of people in a region within a certain period [36]. An increase in population will bring more demands for food, clothing, housing, and transportation, and carbon emissions will change accordingly.

2. Economic level [37]: It is represented by the GDP of each city. With the development of society and the economy, on the one hand, the demand for energy use by humans increases, intensifying carbon emissions. On the other hand, humans’ awareness of environmental protection gradually strengthens, leading to an increased use of clean energy and a reduction in carbon emissions.

3. Industrial structure: It is represented by the proportion of the GDP of the secondary industry [38]. Compared with the primary and tertiary industries, the secondary industry focuses on various industries or products and has a closer relationship with carbon emissions from land use.

4. Land use structure: It is represented by the proportion of construction land. There are more commercial and industrial activities on construction land, resulting in a relatively high carbon emission rate. Different land use structures can lead to different carbon emission values.

5. Technological level [39]: The number of patent authorizations is used to represent the technological level. An improvement in the technological level can continuously enhance production efficiency. While increasing production, it also raises energy consumption. However, when technological innovation is applied to energy conservation and emission reduction, it will promote low-carbon development.

6. Consumption level [40]: It is represented by the total retail sales of social consumer goods. The improvement in the material living standards stimulates people’s consumption of products, thus increasing the quantity of consumer goods and public goods. The manufacturing of these products further leads to an increase in carbon emissions.

3.3.2. Analysis of Driving Factors

• Population size

During the research period, the impact of population size on carbon emissions was mainly positive. With the expansion of the population, urban sprawl led to a sharp increase in urban construction land. Meanwhile, the demand for transportation kept rising, and the demand for energy in production and daily life increased significantly. As a result, energy consumption increased, leading to a growth in carbon emissions.

The spatial influence of population size on carbon emissions from land use presents a pattern characterized by a stronger impact in the northern regions and a weaker one in the southern regions. In northern Jiangsu, the pace of economic development is relatively slow, and the capacity for resource allocation is relatively weak. With the expansion of the population size, the demands for aspects such as food, clothing, housing, and transportation escalate, resulting in a relatively more significant positive impact on carbon emissions. Conversely, southern Jiangsu is more economically developed and has a more robust capacity for resource allocation. A certain degree of population agglomeration effect promotes the development of clean energy and the improvement in technological processes. This effectively transforms the growth of the population size into a driving force for development [41]. Consequently, the increase in population size has a relatively smaller impact on carbon emissions.

2.

Economic level

The economic impact on carbon emissions from land use generally demonstrates a trend of transition from a positive effect to a negative one. In the initial stage of the research, with the rapid growth of the GDP in the study area, energy consumption increased, which exerted a substantial positive influence on carbon emissions from land use. However, when the urban economy develops to a certain level, environmental pollution issues receive increasing attention. The government formulates corresponding policies to promote the carbon reduction process and encourages the development and utilization of clean energy. Simultaneously, enterprises carry out green production to reduce carbon emissions. As a result, the economic level begins to have a negative impact on carbon emissions.

The influence of the economic level on carbon emissions follows a pattern of being higher at both ends and lower in the middle. In southern Jiangsu, rapid economic development drives high energy demands in residents’ production, daily life, and overall economic activities, leading to substantial carbon emissions. In contrast, Xuzhou in northern Jiangsu, which is primarily based on heavy industry, consumes a large amount of energy and consequently generates high carbon emissions. Yancheng, however, experiences the least impact on carbon emissions from its economic level. This is largely due to Yancheng’s extensive coastline, which accounts for 56% of Jiangsu Province’s total, providing abundant clean energy resources such as tidal and wind energy. Yancheng has been an early adopter of a low-carbon economy and has established a solid foundation in this area, effectively reducing carbon emissions and minimizing the impact of economic development on carbon emissions.

3.

Industrial structure

The impact of industrial structure on carbon emissions is generally positive, exhibiting a trend of first strengthening and then weakening over the research period. This pattern is primarily driven by the high energy consumption of secondary industries, especially heavy industries, which significantly contribute to carbon emissions. During the study period, frequent industrial activities in Jiangsu Province led to a gradual increase in the demand for coal and other energy sources, thereby driving up energy consumption and carbon emissions. However, as national policies began to emphasize green development and carbon emission control, efforts were intensified to promote sustainable industrial practices, which gradually mitigated the positive impact of industrial structure on carbon emissions.

Among these cities, the carbon emissions in Suzhou, Nantong, and Wuxi are particularly influenced by their industrial structures. In Suzhou and Wuxi, the manufacturing industry constitutes a significant proportion of their industrial structures, leading to relatively high energy consumption and emissions. However, in recent years, both Suzhou and Wuxi have been actively promoting industrial transformation and supporting environmentally friendly industries, which has gradually reduced the impact of their industrial structures on carbon emissions. In contrast, Nantong’s industrial structure exhibits a pronounced trend towards heavy industrialization, with a large share of traditional high-energy-consuming industries, resulting in higher carbon emissions. Meanwhile, the carbon emissions in Suqian and Huai’an are less affected by their industrial structures. Huai’an, as an emerging industrial city, primarily focuses on light industry, which contributes to its relatively low carbon emissions. Suqian, on the other hand, mainly focuses on agriculture and the high-end textile industry, both of which are associated with relatively low carbon emissions.

4.

Land use structure

The impact of land use structure on carbon emissions shows a relatively stable positive effect. During the research period, with continuous economic development, there have been ongoing industrial and economic activities on construction land, and the population is increasing. The proportion of construction land has been steadily rising, which has a clearly positive impact on carbon emissions [42].

The impact of land use structure on carbon emissions presents a pattern of being higher in the northeast and lower in the southwest. Xuzhou and Lianyungang are significantly affected. In Xuzhou, construction land is mainly used for the development of heavy industries, energy extraction industries, etc., resulting in high carbon emissions. Lianyungang mainly develops industries with high carbon emissions, such as mining and shipbuilding, and these industries have witnessed rapid growth in recent years. Therefore, the proportion of construction land in these two cities has a substantial impact on carbon emissions. Nanjing is relatively less affected. As the provincial capital, Nanjing has strict land use planning and management measures. It implements land use policies oriented towards low-carbon development, resulting in relatively low carbon emissions.

5.

Technological level

The technological level exerts a clearly negative impact on carbon emissions from land use. Ever since the Fifth Plenary Session of the 18th Central Committee of the Communist Party of China put forward the incorporation of green development into the overall development strategy of China, the nation has been unwaveringly committed to the path of prioritizing ecological protection and green development, propelling the comprehensive green transformation across various economic and social sectors. There has been a progressive increase in enterprises’ awareness of green development, leading to an accelerated transition towards green production. With the continuous advancement in technology, energy utilization efficiency has been constantly enhanced, and the growth rate of carbon emissions has decelerated.

The impact of technological level on carbon emissions shows a pattern of being higher in the west and lower in the east. Among them, Xuzhou and Nanjing are most affected by the technological level. As the capital city of Jiangsu Province, Nanjing has a relatively high technological level, which provides effective technical support for promoting green upgrading and low-carbon transformation of enterprises, thus further reducing carbon emissions. Xuzhou is a typical resource-based city, and its energy structure mainly relies on coal. Through technological innovation, energy utilization efficiency can be significantly improved, promoting industrial transformation and upgrading and reducing carbon emissions.

6.

Consumption level

The impact of consumption level on carbon emissions is distinctly positive and demonstrates an increasing trend year by year. The development of the economic society has spurred the growing material demands of residents. This, in turn, promotes the development of the manufacturing industry, resulting in an upsurge in energy consumption and thereby exerting a positive influence on carbon emissions. During the research period, with the rapid economic development, the living demands of residents have been continuously escalating. As a consequence, the impact of consumption levels on carbon emissions has been on the rise year by year.

The impact of consumption level shows a pattern of being higher in the southwest and lower in the surrounding areas. Nantong is less affected. The main reason is that Nantong focuses on implementing the energy efficiency improvement project, promoting advanced and applicable energy-saving and low-carbon process technologies, equipment, and products, increasing the output rate and recycling rate of energy and resources, and thus reducing carbon emissions. Nanjing, Zhenjiang, and Changzhou are more significantly affected, and the impact is gradually intensifying. These cities have a relatively high level of economic development, and residents have a large demand, which promotes the development of the manufacturing industry, increases energy consumption, and further contributes to the increase in carbon emissions.

4. Discussion

4.1. Recommendations for Low Carbon Development

In the current context of “carbon peak and carbon neutrality,” exploring land use changes and carbon emission factors from the perspective of land use types is of great significance for optimizing the land use structure and reducing carbon emissions. Taking into account the impact of diverse land use types on carbon emissions in Jiangsu Province and considering the contributions of various influencing factors to carbon emissions, the following recommendations are proposed to optimize the land structure and reduce carbon emissions in Jiangsu Province [43].

4.1.1. Optimize the Energy Structure

At present, the energy consumption structure in Jiangsu Province is primarily based on coal, with a high degree of dependence, resulting in a relatively high level of carbon emissions. Given that Jiangsu Province is economically developed and has a high level of technological capacity, it should accelerate the development and utilization of clean energy to optimize its energy structure. Lianyungang, Yancheng, and Nantong are coastal cities with abundant clean energy resources such as offshore wind energy and tidal energy. These cities should speed up the construction of offshore wind power stations [44]. Each city should take local conditions into account, encourage and lead the development of photovoltaic power generation, and promote the construction of photovoltaic power plants. By building a clean marine–land integration power generation system, the dependence on coal can be reduced, and the energy structure can be optimized.

4.1.2. Promote the Green Development of Industries

In Jiangsu Province, the secondary industry occupies a relatively high proportion, and the majority of its sectors are energy-intensive. A substantial number of industries are characterized by high carbon emissions. Promoting the green development of industries represents an effective approach to reducing carbon emissions. The government ought to augment technical support and policy assistance for industries. It is essential to clearly define carbon emission thresholds and impose severe penalties on enterprises that transgress these thresholds. Enterprises, on their part, should increase capital input, proactively engage in technological upgrades and energy-saving transformations, enhance energy utilization efficiency, and thus contribute to the reduction in carbon emissions. Leading enterprises in the realm of energy conservation and emission reduction should assume exemplary roles. They should organize cooperation among enterprises, give full play to their respective advantages, and intensify the exchange of technologies and experiences [45].

4.1.3. Reasonable Planning of Construction Land

With the advancement in urbanization, the construction land area in Jiangsu Province has been continuously expanding. Since construction land functions as a carbon source, its planning has a profound impact on carbon emissions. The government should meticulously review newly added construction land and firmly safeguard the ecological protection red line. Meanwhile, it is crucial to tap into the potential of existing construction land by promoting the spatial-composite utilization of the original construction land and encouraging the mixing of spatial functions. Moreover, tailored assessment standards should be developed for different regions, and real-time monitoring of carbon emissions from various uses of construction land should be implemented. These emissions should be incorporated into comprehensive assessment indicators for a more holistic evaluation [46].

4.2. Limitations and Future Research Directions

Although this study has achieved certain outcomes, there remain several limitations in the research process. Firstly, the spatial resolution of land use data significantly influences the accuracy of land use classification. The 30 m resolution land use data employed in this research may inadequately capture micro-level land use variations. Future studies should prioritize the exploration of higher-resolution datasets for land use classification to enhance the precision of analytical outcomes.

Furthermore, this study primarily concentrates on provincial and city-level scale analyses in Jiangsu Province, China, with limited examination of land use carbon emissions at the county level. As a result, it is difficult to fully understand the internal heterogeneity and characteristics of different sub-regions. Future studies should strengthen research on land use carbon emissions at county and township scales and promote coordinated investigations across regions. This would provide more precise scientific support for the formulation of targeted emission reduction policies at the local level.

5. Conclusions

This study conducts an in-depth investigation into carbon emission changes induced by land use in Jiangsu Province, China, over the past two decades. By constructing a land use change model, applying the carbon emission coefficient method to develop a quantitative estimation framework, and utilizing the Geographically and Temporally Weighted Regression (GTWR) model for analysis, the study systematically and comprehensively examines the carbon emission effects of land use from three perspectives: (1) the spatio-temporal dynamics of land use structure, (2) the spatial and temporal variations in land use carbon emissions, and (3) the determinants of land use carbon emissions. The findings reveal the following:

(1) From 2000 to 2020, the land use pattern in Jiangsu Province, China, underwent significant changes. Construction land experienced rapid expansion, increasing by 8271.48 square kilometers over the two-decade period, primarily through the conversion of cultivated land. During the same period, cultivated land continuously declined, while the water area initially expanded and then contracted, exhibiting an overall downward trend. Additionally, the areas of grassland, forest land, and unused land also showed a decreasing trajectory.

(2) Land use carbon emissions in Jiangsu Province have shown a steady upward trend, primarily driven by the substantial expansion of construction land. This expansion has led to a sharp increase in the demand for production and living resources, as well as in energy consumption, thereby intensifying carbon-emitting activities. However, the growth rate of carbon emissions has gradually slowed over time. In the early stages, rapid urbanization and economic development were highly dependent on energy consumption. In the later stages, the optimization of the energy structure and the implementation of environmental protection policies effectively curbed the acceleration of carbon emissions.

(3) An analysis of the factors influencing land use in Jiangsu Province from 2000 to 2020 reveals significant temporal variations in their effects. The impact of the economic development level shifted from a positive driver to a negative constraint, influenced by policy interventions and the growing emphasis on green development concepts. Population size exhibited both positive and negative effects across the study period, reflecting disparities in regional development levels and resource allocation capacities, with notable spatial heterogeneity. Industrial structure, consumption level, and land use structure emerged as prominent positive drivers, with their influence progressively intensifying over time. In contrast, technological advancement exerted a negative effect, attributed to improvements in energy efficiency and the expansion of clean energy utilization.

Based on the current development status of Jiangsu Province and in light of the research findings, this study provides scientifically grounded recommendations for energy conservation, emission reduction, optimized resource utilization, and the achievement of sustainable development goals.