Spatial Evolution and Scenario Simulation of Carbon Metabolism in Coal-Resource-Based Cities Towards Carbon Neutrality: A Case Study of Jincheng, China

Abstract

As important energy suppliers in China, coal-resource-based cities are pivotal to achieving the nation’s 2060 carbon-neutrality goal. This study focused on Jincheng City, utilizing the LOW EMISSIONS ANALYSIS PLATFORM (LEAP) model to predict carbon emissions from energy consumption under various scenarios from 2020 to 2060. Then, combined with the Markov-PLUS model to map carbon emissions to land-use types, it evaluated spatial changes in carbon metabolism and analyzed carbon-transfer patterns across different land-use types. The results showed the following: (1) Across all scenarios, Jincheng’s carbon emissions exhibited an initial increase followed by a decline, with the industrial sector accounting for over 70% of total emissions. While the baseline scenario deviated from China’s carbon peaking target, the high-limit scenario achieved an early carbon peak by 2027. (2) High-negative-carbon-metabolism areas were concentrated in central urban zones and industrial parks. Notably, arable land shifted from a carbon-sink area to a carbon source area by 2060 in both the low- and high-limit scenarios. (3) In the baseline scenario, industrial and transportation land uses were the primary barriers to carbon metabolism balance. In the low-carbon scenario, the focus shifted from industrial and transportation emissions to urban construction land emissions. In the high-limit scenario, changes in urban–rural land-use relationships significantly influenced carbon metabolism balance. This study emphasizes the importance of industrial green transformation and land-use planning control to achieve carbon neutrality, and it further explores the significant impact of territorial spatial planning on the low-carbon transition of coal-resource-based cities.

1. Introduction

Accelerated global industrialization and economic development have led to increasing greenhouse gas emissions, and the problem of global climate change is becoming increasingly serious. To mitigate climate change and promote sustainable development, China has set the development goals of achieving carbon peak by 2030 and carbon neutrality by 2060 [1,2]. As concentrated sources of greenhouse gas emissions, cities consume two-thirds of the world’s energy and contribute 75% of carbon emissions [3,4,5]. With the accelerated pace of urbanization, urban carbon dioxide (CO₂) emissions are continuously rising. Changes in land-use patterns have led to an active urban carbon metabolism, but the rise in carbon emissions coupled with limited carbon sinks presents a significant challenge to green urban development [6,7]. Coal-resource-based cities, as major energy suppliers in China, are characterized by resource-dependent economies, marked by high energy consumption and carbon emissions. Under the carbon-neutrality goal, they are facing dual pressures from economic transformation and energy conservation, along with emission reduction [8,9]. Therefore, effectively promoting energy conservation and economic growth in coal-resource-based cities and realizing the green transformation of coal-resource-based cities is not only important to achieve China’s carbon-neutrality target, but also a key part of China comprehensively realizing low-carbon development.

Urban carbon metabolism refers to the dynamic balancing process by which cities absorb gaseous carbon compounds through their carbon-sink capacity within specific spatial and temporal environments [10]. Building on the theory of urban carbon metabolism, scholars have extensively studied the quantification of urban carbon emissions and sequestration. Methods for quantifying carbon emissions and sequestration include model estimation, sample inventory, remote sensing, map estimation, and box observation [11], with model estimation and remote sensing being the most commonly used. Carbon-emission modeling methods are mainly divided into top-down and bottom-up models [12]. Top-down models are based on the national production sector and are suitable for exploring carbon-emission analysis under macroeconomic policy, climate policy, and energy policy, while bottom-up models are more suitable for technical decision-making in energy supply-and-demand forecasting [13,14]. The LEAP model integrates closely with scenario analysis, accounting for factors such as population, economic development, and energy efficiency improvements. It is applicable to medium- and long-term forecasts of energy supply, terminal demand, and carbon emissions under varying development scenarios. Additionally, it can be adapted for long-term energy planning by customizing the model structure, data format, and forecasting methods based on research problem characteristics and data availability [15,16]. It has been widely used in national, regional, and sectoral energy strategy and ‘dual-carbon’ strategy studies. Emodi used the LEAP model to explore Nigeria’s future energy demand and supply and associated GHG emissions from 2010 to 2040 under four scenarios [17]. Nieves used the LEAP model to explore Colombia’s future energy demand and GHG emissions under both positive and negative scenarios for energy demand and GHG emissions in 2030 and 2050 [18]. Liu developed a LEAP–Tourist model with two scenarios and four sub-scenarios to project the peak GHGs from 2017 to 2040 [19]. Remote sensing and map estimation methods are mainly based on the urban land-use pattern, using remote sensing technology to measure the carbon emissions of different land uses, sectors, and industries in order to obtain the carbon emissions per unit area of a specific land type or carbon sink and then combining the urban carbon metabolism pattern with different specific scenarios for optimization [20]. Liang combined carbon accounting with ecological network analysis to depict the evolution of the profile and spatial distribution of carbon metabolism in the Beijing–Tianjin–Hebei urban agglomeration between 2015 and 2019 [21]. Marchi integrated carbon accounting with Geographic Information System (GIS) mapping to visualize the spatial distribution of greenhouse-gas balance results in Grosseto [22]. In-Ae Yeo forecasted the urban energy demand for individual spatial units using cell facility information and an energy GIS database [23].

The high-energy and high-emission development mode of coal-resource-based cities makes them face a more challenging transition compared to other types of resource-based cities. The low-carbon development of coal-resource-based cities under the carbon-neutral target has attracted much attention from the academic community [24]. Most resource-based cities in China have large total carbon emissions, are still in the growth phase, and feature high spatial concentrations. Resource dependency in coal-resource-based cities crowds out technological innovation and restricts industrial diversification, which significantly contributes to the rise in carbon emissions [25,26]. The thriving resource industry requires low-skilled labor, which hampers technological innovation in these cities. Outdated production technologies further escalate resource consumption, thereby increasing carbon emissions [27,28]. Long-term investment of capital in resource-based industries that generate high profits but are highly polluting creates a monolithic industrial structure, which further increases carbon emissions. Furthermore, higher economic growth targets for resource-based cities often significantly reduce carbon-emission efficiency, particularly in those located in central and western China [29]. Li used the Generalized Divisa Index Method to analyze the drivers of industrial carbon emissions in China’s largest coal-producing city and to forecast industrial carbon emissions from 2021 to 2030 [30]. Y Yan analyzed the energy efficiency of 104 resource-based cities in China and concluded that these cities displayed efficiency disparities under different scenarios [31]. Xu argued that the carbon-emission-reduction effects vary across resource cities depending on the type of resources they rely on [32].

In summary, there have been many studies on urban carbon metabolism, mainly in terms of urban carbon-emission accounting and the evolution of carbon metabolism patterns combined with urban-land carbon sequestration, but there is still room for further exploration. Firstly, carbon-emission predictions based on model estimation are rarely integrated with the development trends of land use, making it challenging to predict the future spatial patterns of urban carbon metabolism. Secondly, predictions of carbon metabolism based solely on remote sensing often overlook macroeconomic and policy constraints, resulting in deviations from actual carbon emissions. Thirdly, in addition to the general rules of urban development, resource cities are facing the dilemma of resource depletion, which puts them into the ‘resource curse’, especially the coal-resource-based cities with high energy consumption, and there are fewer studies on carbon metabolism in coal-resource-based cities at this stage.

The main objective of this study was to predict the spatial pattern of carbon metabolism in coal-resource-based cities in China and to explore the weaknesses of future carbon control and sink enhancement in this type of city. The findings aim to inform the development of scientifically sound and rational carbon-peak strategies for the study area. Jincheng, located in Shanxi Province, is China’s largest production base for anthracite coal and coalbed methane. The city’s primary industries are energy and heavy chemicals, with a long-established industrial structure dominated by the coal industry. This makes it a significant consumer of resources and a major emitter of pollutants. As rapid economic development and industrial restructuring continue, the city’s environmental capacity to support further economic growth is nearing its limits, creating an urgent need to reduce carbon emissions and enhance carbon sinks through effective urban planning. Compared with previous studies, this study can provide the following contributions: Firstly, previous studies on carbon metabolism have mainly focused on general city types, while this study fills this gap by taking coal-resource-based cities as the object of study. Secondly, this study combines LEAP-based carbon-emission prediction with land-use modeling and establishes three different scenarios to comparatively study the spatiotemporal development of carbon metabolism in coal-resource-based cities. This approach is an effective attempt to determine the spatial distribution of future urban carbon emissions. Finally, compared to the widely used Future Land-Use Simulation Model (FLUS) and the Cellular Automata–Markov Model (CA–Markov), the Markov–Patch-Generating Land-Use Simulation Model (Markov–PLUS) model incorporates significant improvements. It retains the adaptive inertia mechanism and roulette-wheel competition mechanism from the FLUS model while integrating the Random Forest algorithm to calculate the development probability of each land-use category. This enhancement addresses the limitations of the CA–Markov model in mining land-use transformation rules and simulating patch-level changes in natural land types, such as forests and grasslands, in a spatiotemporal dynamic manner. With higher simulation accuracy and the ability to generate more realistic land-use scenarios, the Markov–PLUS model demonstrates scientific validity for simulating long-term land-use changes [33]. By integrating this model with a carbon metabolism framework, we can explore the spatiotemporal trends of carbon flows over medium- and long-term horizons, providing critical insights for formulating future urban carbon-mitigation strategies.

2. Materials and Methods

2.1. Study Area

Jincheng (longitude 111°55′–113°37′, latitude 35°12′–36°30′) is located in the central region of China, with a total area of approximately 9490 square kilometers and is located in a temperate continental monsoon climate zone. As shown in Figure 1, Jincheng is situated within a plateau basin terrain. The terrain of the whole city is in the shape of a skip and is high in the north and low in the middle and south. Mountains and hills make up 87.10% of the total area, with 58.60% being mountainous and 28.50% consisting of hills. The main land-use type is forest land, covering approximately 38.90% of the total area, mainly in the west and east. Urban construction land and arable land account for 31.00% of the total area, mainly in the central basin. In recent years, prolonged uncontrolled mining activities have resulted in issues such as land and ecological degradation, contributing to an imbalance in the city’s carbon metabolism [34]. Therefore, it is imperative to explore transition pathways to sustain low-carbon urban development amidst pressures from both natural factors and human activities.

2.2. Data Sources

Based on the availability of data and the time period covered by various socio-economic plans, the study used 2020 as the base year and 2021~2060 as the projection period. (1) Economic development and energy consumption data for activity levels, energy intensity, and food production in urban buildings, transport, industry, agriculture, and other related sectors were obtained from the Shanxi Provincial Statistical Yearbook (2001~2021), the Jincheng Statistical Yearbook (2001~2021), and so on. (2) Land-use data were obtained from the global surface coverage GlobeLand30 database (https://www.webmap.cn/mapDataAction.do?method=globalLandCover, accessed on 18 March 2025), with a resolution of 30 m. The data were reclassified into arable land, forest land, grassland, watersheds, urban land, rural settlements, industrial traffic land, and unused land. (3) The DEM data were obtained from the Geospatial Data Cloud Platform (http://www.gscloud.cn/, accessed on 10 July 2024), and the slope data were extracted with the help of ArcGIS, with a resolution of 30 m. (4) The scope of the nature reserve was obtained from the National Geographic Information Resources Catalogue Service System (https://www.webmap.cn/, accessed on 10 July 2024). (5) The spatial distribution data of population, GDP, average annual temperature, and annual precipitation were obtained from the Resource and Environment Science and Data Centre (http://www.resdc.cn/, accessed on 10 July 2024), with a resolution of 1 km. (6) Soil types were obtained from the Harmonized World Soil Database version 1.2 (https://www.fao.org/soils-portal/data-hub/soil-maps-and-databases/harmonized-world-soil-database-v12/en/, accessed on 18 March 2025), with a resolution of 1 km. (7) Location data of primary, secondary, and tertiary roads and government sites were sourced from OpenStreetMap (https://www.openstreetmap.org/, accessed on 10 July 2024). (8) River- and water-system data were obtained from the Tibetan Plateau Science Data Centre (https://data.tpdc.ac.cn/, accessed on 10 July 2024). The spatial resolution of all data was resampled to 30 m in preprocessing.

2.3. Methods

2.3.1. Research Strategy

As shown in Figure 2, we analyzed the urban carbon metabolism pattern in four stages. In the first stage, urban carbon emissions for the three scenarios—baseline, low-carbon, and high-limit—were predicted using the LEAP model. In the second stage, the Markov–PLUS model was used to predict the urban land-use structure for the target year. In the third stage, the area of the land use in the target year was connected with the results of the LEAP model to obtain the land-use carbon-emission pattern, while the carbon-sink coefficient method and the crop carbon-sink estimation method were applied to obtain the land-use carbon-sink pattern, and then spatial superposition was carried out to obtain the spatial pattern of urban carbon metabolism. In the fourth stage, the carbon-transfer characteristics of land use under different scenarios were analyzed and discussed in the future transformation and development strategies of the city.

2.3.2. LEAP Model

Energy demand forecasting is influenced by multiple factors. To ensure accurate LEAP model predictions, this study first established a baseline scenario based on existing policy measures (Table 1). Within the baseline scenario, current emission-reduction measures and their maximum foreseeable potential were incorporated. Additionally, low-carbon and high-limit scenarios were defined as follows. (1) Baseline Scenario: This scenario reflected the current economic development model, following the trajectory of energy consumption and carbon emissions based on the city’s ongoing economic momentum. It did not account for new carbon-reduction technologies or policies. (2) Low-carbon scenario. On the basis of the baseline scenario, we considered the city’s future sustainable economic and social development and energy-saving and emission-reduction potential, with further transformation of the macroindustrial and sectoral structure and a certain improvement in the energy-saving level of the terminal equipment. (3) High-limit scenario. This scenario further explored the carbon-reduction potential across sectors, maximizing the city’s emission-reduction measures, such as large-scale commercial applications of CCUS technology (Figure 3).

Table 1. Setting of key indicators in multiple scenarios.

Scenario Measures	Description of Measures	Baseline Scenario	Low-Carbon Scenario	High-Limit Scenario
Baseline scenario measures	Current policy measures and technology levels	√	√	√
Optimization of the industrial structure	Increase the proportion of the tertiary sector to 50.00% by 2030 and 70.00% by 2060	−	√	√
Conversion efficiency improvement	Thermal power efficiency is projected to increase to 45.50% by 2030 and 47.50% by 2060	−	√	√
Energy efficiency improvement	Projections for the reduction rate of the sectoral energy intensity based on regression analysis of data from 2000 to 2020	−	√	√
Clean energy alternatives	Increase the share of non-fossil energy generation to 80.00% by 2060	−	−	√
Accelerated electrification	The sectoral share of electricity consumption is projected to increase further. By 2060, 80.00% of electricity will be used in the industrial sector and 75.00% in the transport sector [35]	−	−	√
CCUS technology	Scale application of CCUS by 2030. A total of 40.00% of total installed coal power capacity with carbon capture devices and 90.00% of carbon dioxide capture from coal CCUS by 2060 [36]	−	−	√

Table 1. Setting of key indicators in multiple scenarios.

Scenario Measures	Description of Measures	Baseline Scenario	Low-Carbon Scenario	High-Limit Scenario
Baseline scenario measures	Current policy measures and technology levels	√	√	√
Optimization of the industrial structure	Increase the proportion of the tertiary sector to 50.00% by 2030 and 70.00% by 2060	−	√	√
Conversion efficiency improvement	Thermal power efficiency is projected to increase to 45.50% by 2030 and 47.50% by 2060	−	√	√
Energy efficiency improvement	Projections for the reduction rate of the sectoral energy intensity based on regression analysis of data from 2000 to 2020	−	√	√
Clean energy alternatives	Increase the share of non-fossil energy generation to 80.00% by 2060	−	−	√
Accelerated electrification	The sectoral share of electricity consumption is projected to increase further. By 2060, 80.00% of electricity will be used in the industrial sector and 75.00% in the transport sector [35]	−	−	√
CCUS technology	Scale application of CCUS by 2030. A total of 40.00% of total installed coal power capacity with carbon capture devices and 90.00% of carbon dioxide capture from coal CCUS by 2060 [36]	−	−	√

The city’s energy consumption mainly comprises coal, crude oil, electricity, natural gas, and heat. Carbon emissions from energy consumption were estimated using the accounting methodology outlined in the IPCC Guidelines for National Greenhouse Gas Inventories, as shown in Equation (1).

$$CA=\sum E_i·{CV}_i·F_i·a$$

(1)

where CA represents the CO₂ emissions from energy consumption (in tons, t); E_i denotes the annual energy consumption for different energy types (in tons, t, or cubic meters, m³); V_i is the net calorific value of energy type i (in TJ/Gg); F_i represents the emission factor for energy type i; and $a$ is the unit conversion coefficient, which is dimensionless. The calculation results for the 2020 carbon emissions in Jincheng City are shown in Table 2.

Table 2. Accounting for CO₂ emissions in Jincheng City in 2020.

Sector	CO2 Emissions/Mt CO2	Percentage/%
Agriculture	0.31	0.70
Industry	37.05	84.24
Construction	0.20	0.45
Transportation	1.04	2.37
Wholesale and retail	0.52	1.19
Other tertiary industries	1.00	2.28
Residential	3.86	8.77
Total	43.98	100.00

Table 2. Accounting for CO₂ emissions in Jincheng City in 2020.

Sector	CO2 Emissions/Mt CO2	Percentage/%
Agriculture	0.31	0.70
Industry	37.05	84.24
Construction	0.20	0.45
Transportation	1.04	2.37
Wholesale and retail	0.52	1.19
Other tertiary industries	1.00	2.28
Residential	3.86	8.77
Total	43.98	100.00

2.3.3. Carbon-Sink Calculation Model

The carbon-sink coefficient was used to calculate the carbon exchange rate for the land, C_i. The formula used is as follows:

$$C_i=s_i\times k_i$$

(2)

where s_i is the area of land in category i, m², and k_i is the carbon-sink coefficient of land in category i, kgC·m⁻²·a⁻¹ (Table 3).

The carbon sink of arable land mainly originates from crop photosynthesis. Due to the harvesting behavior of agricultural vegetation compared to other land, it is difficult to obtain accurate carbon-exchange-rate results using carbon-sink coefficients for calculation. Therefore, the carbon sink during the growing cycle of crops, C_crops, was estimated to represent the carbon exchange rate of arable land. The results deducted the carbon consumed by crop respiration [37,38]. The formula used is as follows:

$$C_{crops}=\sum _{i=1}^n{\displaystyle \frac{P_i}{H_i}}\times (1-r_i)\times f_i$$

(3)

where n is the type of crop; P_i is the economic yield of the i crop, kg; H_i is the economic coefficient of the i crop; r_i is the water content of the i crop, %; and f_i is the carbon-uptake rate of the i crop, %—the amount of carbon that needs to be taken up by the crop to synthesize a unit of organic matter (dry weight).

Table 3. Carbon-sink coefficients of land * (kgC·m⁻²·a⁻¹).

Land Type	Forest Land	Grassland	Water Area	Unused Land
Carbon-sink coefficient	−0.5810	−0.0210	−0.0248	−0.0005

* The carbon-emission coefficients of forest land and grassland were based on the studies of Fang [39], the carbon-emission coefficients of waters were based on the studies of Duan [40], and the carbon-emission coefficients of unused land were based on the studies of Li [41].

Table 3. Carbon-sink coefficients of land * (kgC·m⁻²·a⁻¹).

Land Type	Forest Land	Grassland	Water Area	Unused Land
Carbon-sink coefficient	−0.5810	−0.0210	−0.0248	−0.0005

* The carbon-emission coefficients of forest land and grassland were based on the studies of Fang [39], the carbon-emission coefficients of waters were based on the studies of Duan [40], and the carbon-emission coefficients of unused land were based on the studies of Li [41].

2.3.4. Markov–PLUS Model

Future land-use projections are key to achieving a convergence between carbon emissions from energy consumption and carbon emissions from urban land use. This study employed the Markov–PLUS model to simulate future land-use patterns: (1) Based on the characteristics of Jincheng, 10 driving factors were selected, including population, GDP, road distance, watershed, DEM, slope, average annual temperature, and average annual precipitation. (2) According to the ‘Jincheng Territorial Spatial Planning (2021–2035)’, areas designated as ecological red-line protection zones, permanent basic farmland, urban nature reserves, nature parks, and rivers were classified as restricted zones. (3) Land-use data from 2010 and 2020 were used to extract land-use expansion, analyze the driving factors of land-use changes, and simulate 2020 land-use patterns using the CA model. Accuracy was validated using the validation module. The land-use data for 2030 and 2060 were predicted step by step. The expansion intensity of land use was expressed by domain weights. (4) The predicted carbon emissions in different scenarios were integrated with land-use structure predictions, resulting in a spatial distribution of carbon emissions. The formula used is as follows:

$${\mathrm{X}}_{\mathrm{a}}={\displaystyle \frac{{\mathrm{x}}_{\mathrm{a}}-{\mathrm{x}}_{\mathrm{m}\mathrm{i}\mathrm{n}}}{{\mathrm{x}}_{\mathrm{m}\mathrm{a}\mathrm{x}}-{\mathrm{x}}_{\mathrm{m}\mathrm{i}\mathrm{n}}}}$$

(4)

where X_a is the domain weight of land type a, the value range being 0~1, where the larger the value, the stronger the expansion ability of the land type; x_a is the difference between the area of land type a in 2020 and that of land type a in 2010; x_max is the maximum value of x_a; and x_min is the minimum value of x_a. The future simulation phase followed the data obtained in this step. Domain weight parameters for each land type were processed (Table 4).

Table 4. Domain weight parameters.

Land Type	Arable Land	Forest Land	Grassland	Water Area	Construction Land	Rural Settlement Land	Industrial and Transport Land	Unused Land
2020~2030 (hm2)	0.00	0.52	0.57	0.69	0.70	0.84	1.00	0.62
2030~2040 (hm2)	0.86	1.00	0.45	0.46	0.40	0.17	0.00	0.48
2040~2060 (hm2)	0.33	1.00	0.24	0.36	0.30	0.08	0.00	0.33

Table 4. Domain weight parameters.

Land Type	Arable Land	Forest Land	Grassland	Water Area	Construction Land	Rural Settlement Land	Industrial and Transport Land	Unused Land
2020~2030 (hm2)	0.00	0.52	0.57	0.69	0.70	0.84	1.00	0.62
2030~2040 (hm2)	0.86	1.00	0.45	0.46	0.40	0.17	0.00	0.48
2040~2060 (hm2)	0.33	1.00	0.24	0.36	0.30	0.08	0.00	0.33

2.3.5. Carbon Metabolism Model

Referring to the research method of Li [42], the net carbon emissions per unit area is defined as the net carbon metabolic density so as to quantify the carbon metabolism capacity of different land spaces and measure the carbon flow and its direction in the process of land-use change with the help of the difference in the carbon metabolic density and the area of land-use transfer. The formula used is as follows:

$$∆W=W_i-W_j={\displaystyle \frac{V_i}{s_i}}-{\displaystyle \frac{V_j}{s_j}}$$

(5)

$$f_{ij}=∆W·∆a_{ij}$$

(6)

where $∆W$ is the carbon metabolism density difference, $W_i$ and $W_j$ denote the net carbon metabolism density of lands i and j, $V_i$ and $V_j$ denote the net carbon emissions of lands i and j, $s_i$ and $s_j$ denote the area of lands i and j, $f_{ij}$ denotes the carbon flow from land j to land i, and $∆a_{ij}$ denotes the area of the land j to land i transfer. In general, $f_{ij}$ > 0 indicates a positive carbon flow, which is indicative of a balanced carbon metabolism, and $f_{ij}$ < 0 indicates a negative carbon flow, which is indicative of a disturbed carbon metabolism.

3. Results

3.1. Analysis of Urban Carbon-Emission Scenario Projections

As shown in Figure 4a–c, under the baseline scenario, carbon emissions in Jincheng peak at 66.53 Mt in 2032, with an average annual growth rate of 4.66%, before decreasing to 43.31 Mt by 2060. In the low-carbon scenario, carbon emissions rise gradually due to improvements in energy efficiency and electrification, peaking at 59.72 Mt in 2029, with an average annual growth rate of 4.48%. By 2060, emissions decrease to 19.69 Mt—significantly lower than in the baseline scenario. In the high-limit scenario, emission-reduction measures result in a continuous decline in end-use energy consumption. Carbon emissions peak at 53.05 Mt in 2027, with an average annual growth rate of 3.44%, and fall to 11.73 Mt by 2060, representing a 72.91% reduction compared to the baseline scenario. Since the baseline scenario assumes current policy measures and technology levels, carbon emissions in Jincheng are projected to continue growing over the next decade. In the low-carbon scenario, industrial restructuring, energy efficiency improvements, electrification, and other measures are introduced, reducing the growth rate and achieving a carbon peak around 2030. In the high-limit scenario, stricter emission-reduction measures lead to an earlier peak around 2027.

The industrial carbon emissions constitute the largest portion of total emissions across all sectors from 2020 to 2060. Under the baseline scenario, industrial carbon emissions reach 56.29 Mt in 2030 compared to 52.21 Mt and 45.32 Mt in the low-carbon and high-limit scenarios, respectively. By 2060, industrial carbon emissions in the baseline, low-carbon, and high-limit scenarios are projected to drop to 32.03 Mt, 14.46 Mt, and 8.85 Mt, respectively. However, industrial carbon emissions consistently account for over 70% of total emissions, primarily due to Jincheng’s large industrial base. Although industrial carbon emissions are reduced under different scenarios, they remain significantly higher than those in other sectors with smaller emission bases. Additionally, under the baseline scenario, residential carbon emissions increase significantly, likely driven by rising living standards, which lead to more frequent and diverse consumption patterns. These changes include shifts in household transport modes, increased appliance usage, and longer operating hours. However, these emissions are effectively controlled under the low-carbon and high-limit scenarios. In the transport sector, as Jincheng serves as a key gateway from Shanxi to the Central Plains, carbon emissions in the baseline scenario do not decrease significantly but are effectively mitigated under abatement measures. Overall, the low-carbon and high-limit scenarios demonstrate effective control of carbon emissions from the industrial, transport, construction, and agricultural sectors through energy-saving and emission-reduction measures.

As shown in Figure 4d,e, the contributions of sub-scenarios to carbon-emission reduction across different time periods were analyzed. In the medium term (2030), clean energy substitution and CCUS technology contribute significantly to emission reductions, accounting for 22.34% and 20.29% of the total reduction, respectively. In the long term (2060), industrial structure optimization and improvements in conversion efficiency play a more substantial role, contributing 38.33% and 22.53%, respectively. This is followed by energy efficiency improvements and CCUS technology, which account for 13.93% and 9.80% of the reductions. Therefore, clean energy substitution and CCUS technology achieve significant short-term emission reductions, while industrial structure optimization and conversion efficiency improvements offer long-term, sustainable reductions. Due to the entrenched nature of Jincheng’s industrial structure, achieving rapid, high-quality industrial transformation is challenging in the short term. However, the introduction of new technologies can quickly reduce carbon emissions. As technological advancements and policy support increase, the city is progressively transitioning toward a high-quality green transformation. The traditional industrial base is gradually being replaced by new technology-driven industries, creating significant opportunities for emission reductions. Consequently, carbon emissions are significantly lowered in the later stages of multi-technology implementation.

In order to further validate the accuracy of the model, and taking into account the availability of data and the reasonableness of using a long-term prediction model, the article compares the predicted data with the actual data and assesses the stability of the model’s predictions. The city’s energy demand in 2020~2022 in the LEAP model was compared with the comprehensive energy-balance table in the 2021~2023 statistical yearbook of Jincheng City, and the MAPE of the three years’ data was <5% [43], which was considered to be a more reasonable effect of the model prediction (Table 5).

Table 5. Accounting for energy demand in Jincheng City from 2020 to 2022.

Year	Total Energy Demand/104tce		MAPE/%
	Actual Value	Baseline Scenario Forecast
2020	1868.80	1817.20	1.38%
2021	1958.00	1924.79	0.85%
2022	1996.60	2032.371	0.90%

Table 5. Accounting for energy demand in Jincheng City from 2020 to 2022.

Year	Total Energy Demand/104tce		MAPE/%
	Actual Value	Baseline Scenario Forecast
2020	1868.80	1817.20	1.38%
2021	1958.00	1924.79	0.85%
2022	1996.60	2032.371	0.90%

3.2. Urban Land Projections

Based on land-use changes in Jincheng from 2010 to 2020, the land-use structure for 2030 and 2060 was simulated using 2020 as the baseline year, with development constraints established according to the Markov–PLUS model principles and previously proposed methods (Table 6).

Table 6. Area of each type of land use.

Land Area (hm2)	Arable Land	Forest Land	Grassland	Water Area	Construction Land	Rural Settlement Land	Industrial Traffic Land	Unused Land
2020	334,808.46	456,640.13	94,573.66	3435.69	10,581.98	31,608.28	10,638.65	125.69
2030	334,770.52	456,402.39	94,455.78	3495.02	10,753.86	31,290.52	11,120.70	123.76
2060	334,712.82	457,224.01	93,074.54	3570.36	11,930.40	29,349.33	12,432.33	118.76

Table 6. Area of each type of land use.

Land Area (hm2)	Arable Land	Forest Land	Grassland	Water Area	Construction Land	Rural Settlement Land	Industrial Traffic Land	Unused Land
2020	334,808.46	456,640.13	94,573.66	3435.69	10,581.98	31,608.28	10,638.65	125.69
2030	334,770.52	456,402.39	94,455.78	3495.02	10,753.86	31,290.52	11,120.70	123.76
2060	334,712.82	457,224.01	93,074.54	3570.36	11,930.40	29,349.33	12,432.33	118.76

In 2030, industrial traffic land, waters, and urban land increased by 482.05 hm², 59.33 hm², and 171.88 hm², respectively, while arable land, grassland, rural settlements, and forest land decreased by 37.94 hm², 117.88 hm², 317.76 hm², and 237.74 hm², respectively. By 2060, forest land, waters, urban land, and industrial traffic land will have increased by 821.62 hm², 75.34 hm², 1176.54 hm², and 1311.63 hm², respectively, while arable land, grassland, and rural settlement land will have decreased by 57.7 hm², 1381.24 hm², and 1941.19 hm². Under China’s arable land protection policy, arable land in Jincheng did not experience significant reductions over time. However, urban and industrial traffic land expanded noticeably, while rural settlement land decreased significantly. According to the land-transfer matrix constructed using GIS (Figure 5), the urban land structure from 2010 to 2020 was primarily composed of arable and forest land. The expansion of urban construction and industrial traffic land encroached on portions of arable, forest, and grassland. Although the overall expansion area was not large, the expansion ratio was significant. The increase in watershed areas may be attributed to the construction of wetland parks.

3.3. Spatial and Temporal Characteristics of Land-Use Carbon Transfer Under Different Scenarios

According to Figure 6, in 2020, urban areas with a carbon-sink density of 0.10 to 0.50 tC·hm⁻² experienced an increase of 68.74 hm² compared to 2010. In contrast, areas with a carbon-sink density of less than 0.10 tC·hm⁻² saw a decrease of 279.19 hm². Overall, the city’s carbon-sink capacity improved during this period. Between 2010 and 2020, the positive-carbon-metabolism compartments primarily consisted of forest land, grassland, and water, while the negative compartments were chiefly industrial traffic and urban land. Notably, excluding arable land, net carbon emissions from the negative-carbon-metabolism compartments, which represent only 10% of the total urban area, were approximately 5 to 15 times greater than the net carbon sequestration from the positive compartments. With the impetus of economic development and accelerated urbanization, urban carbon emissions in 2020 increased by 2.14 times compared to 2010. This period also saw an expansion in the areas characterized by negative carbon metabolism, with high-emission zones growing alongside urban growth. Although arable land functions as a carbon sink, it simultaneously generates significant carbon emissions during cultivation and production, thereby acting as a carbon source. Urban construction land and industrial traffic land are predominant within the negative-carbon-metabolism compartments, and the carbon-emission density of industrial land surged by 1.60 times between 2010 and 2020.

In the baseline scenario for 2030, urban areas with a carbon-sink density exceeding 0.50 tC·hm² are projected to increase by 923.03 hm² compared to 2020. Positive-carbon-metabolism sectors continue to encompass forests, grasslands, watersheds, and under-utilized land, while negative sectors consist of industrial traffic land, urban land, arable land, and rural settlement areas. In 2060, compared to 2030, urban carbon-sink densities in the range of 10.00 to 100.00 tC·hm² will shift toward areas with a density of 0.00 to 10.00 tC·hm², primarily affecting rural settlements. Similarly, areas with carbon-emission densities greater than 500.00 tC·hm² will transition to regions with densities between 100.00 and 500.00 tC·hm², indicating a gradual move towards a balanced urban carbon metabolism. As Jincheng’s economic development continues to expand, the impact of measures aimed at returning farmland to forests, grasses, and lakes has become increasingly less evident in the region’s land-use evolution. At this stage of natural development, there is a limited conversion of land from negative- to positive-carbon-metabolism compartments.

In both the low- and high-limit scenarios, arable land transitions from a negative- to a positive-carbon-metabolism compartment, likely due to enhanced crop production and a reduction in arable land area. This change further influences the development of rural settlements, resulting in a decreased density of negative carbon metabolism within rural land. Spatially, under the low-carbon scenario, the area with negative carbon metabolism is projected to encompass 3867.20 km² by 2030, primarily concentrated in central urban and county regions, which account for 41.04% of the total area. By 2060, this area is expected to shrink to 2996.14 km², representing 31.79% of the total area. In the high-limit scenario, areas exhibiting negative carbon metabolism are anticipated to cover 3571.23 hm² in 2030, decreasing to 933.48 hm² by 2060. These areas are mainly found in urban centers and industrial development zones, characterized predominantly by urban construction and industrial traffic land. In both the low-carbon and high-limit scenarios, the positive carbon flow is primarily attributed to the conversion of high-intensity carbon-emitting land to lower-intensity land, alongside enhancements in the carbon-sink capacity of arable land. With the implementation of various technical measures, the city’s carbon sinks are projected to exceed carbon emissions, playing a vital role in balancing Jincheng’s carbon metabolism.

According to Table 7, Jincheng experienced a negative net carbon transfer from 2010 to 2020, suggesting that land-use changes adversely affected carbon metabolism in the region. The primary source of positive carbon transfer was from industrial traffic land to arable land, while the predominant negative transfer occurred from arable land to industrial traffic land. Notably, the negative carbon transfer was approximately four times greater than the positive transfer, resulting in an imbalance in carbon metabolism.

In the baseline scenario, positive carbon transfer is projected to peak at 341.00 Mt between 2020 and 2030, primarily driven by the conversion of industrial traffic land to forest land (I&T→FL). Negative carbon transfer constitutes only 20% of this total, with rural settlements transitioning to industrial traffic land (RSL→I&T) accounting for 76.31% of that figure. Consequently, the net carbon transfer attains its highest value of 2.76 Mt during this period. Looking ahead to 2030–2040, both positive and negative carbon transfers are expected to decline. The leading contributor to positive carbon transfer remains the conversion of industrial traffic land to forest land (I&T→FL), while negative carbon transfer shifts toward the conversion of rural settlements to urban construction land (RSL→UCL). Between 2040 and 2060, positive carbon transfer continues to decrease, yet the dominant transfer pathways remain stable and unchanged. This indicates that high carbon emissions from the industrial and transportation sectors are significant factors disrupting the balance of urban carbon metabolism. Promoting carbon transfer from these sectors to positive-carbon-metabolism compartments, such as forest land and arable land, can effectively enhance overall positive carbon metabolism in urban areas.

In the low-carbon scenario, positive carbon transfer decreases to 312.20 Mt, while negative carbon transfer declines to 595.60 Mt between 2020 and 2030. The negative carbon metabolism during this period is primarily attributed to the conversion of rural settlements to industrial traffic land (RSL→I&T). From 2030 to 2040, this negative carbon metabolism mainly results from the transition of rural settlements to urban land (RSL→UCL), with its magnitude considerably lower than that observed in the baseline scenario. This suggests that as technology advances and urban areas expand, carbon emissions from urban land use will escalate rapidly. Consequently, in comparison to the baseline scenario, the focus of urban low-carbon transition should shift from managing carbon emissions in industrial and transport sectors to addressing those arising from urban land use.

In the high-limit scenario, both positive and negative carbon transfers are projected to continue declining from 2020 to 2030, leading to a net positive carbon metabolism overall. Net carbon transfers are expected to constitute an increasing share of total transfers. Thanks to technological advancements and mitigation measures, the city’s carbon metabolism transitions from a negative to a positive impact, aligning with the gradual attainment of the carbon-peak target. Concurrently, as long-term carbon emissions decrease, the dominant urban positive carbon transfer stabilizes during the transition from industrial traffic land to forest land (I&T→FL), while the predominant urban negative carbon transfer stabilizes in the shift from rural settlements to urban land (RSL→UCL). Jincheng boasts a significant proportion of forested land with ample carbon-sink capacity, which supports long-term positive carbon metabolism within the city’s spatial carbon dynamics. Overall, there remain disparities in urbanization levels across different regions of Jincheng. As efforts to transform the industrial structure progress, it is also crucial to consider the effects of changing urban–rural relations on urban carbon metabolism.

Table 7. Carbon transfer in Jincheng from 2020 to 2030 *.

Project	2010~2020		Scenario Simulation	2020~2030		2030~2040		2040~2060
	Carbon Transferred (Mt)	Direction		Carbon Transferred (Mt)	Direction	Carbon Transferred (Mt)	Direction	Carbon Transferred (Mt)	Direction
Positive carbon transfer	161.8	+	BS	341.00	+	251.37	+	246.59	+
			LCS	312.20	+	203.05	+	139.36	+
			HLS	263.70	+	164.67	+	97.44	+
Negative carbon transfer	−516.31	−	BS	−65.15	−	−0.56	−	−1.19	-
			LCS	−59.56	−	−0.44	−	−0.71	-
			HLS	−50.31	−	−0.33	−	−0.35	-
Net carbon transfer	−354.51	−	BS	275.85	+	250.82	+	245.39	+
			LCS	252.65	+	202.61	+	138.65	+
			HLS	213.38	+	164.33	+	97.09	+
H1	94.23	I&T→UCL	BS	297.63	I&T→FL	222.23	I&T→FL	286.75	I&T→FL
			LCS	87.28%		88.41%		82.73%
			HLS	272.28		179.57		166.75
Proportion of contribution	58.24%		BS	87.21%		88.46%		82.91%
			LCS	230.00		145.78		80.96
				87.22%		88.53%		83.09%
H2	365.53	UCL→I&T	BS	49.72	RSL→I&T	0.49	RSL→UCL	1.06	RSL→UCL
			LCS	76.31%		88.75%		89.03%
			HLS	45.45		0.37		0.54
Proportion of contribution	70.80%		BS	76.31%		85.65%		61.03%
			LCS	38.38		0.27		0.17
				76.29%		81.21%		48.63%

* Notes: ‘+’ and ‘−’ represent the direction of carbon transfer; H1 represents dominant positive carbon transfer; H2 represents dominant negative carbon transfer. FL: forest land, GL: grassland, WA: water area, UL: unused land, AL: arable land, RSL: rural settlement land, UCL: urban construction land, I&T: industrial traffic land. BS: baseline scenario, LCS: low-carbon scenario, HLS: high-limit scenario.

Table 7. Carbon transfer in Jincheng from 2020 to 2030 *.

Project	2010~2020		Scenario Simulation	2020~2030		2030~2040		2040~2060
	Carbon Transferred (Mt)	Direction		Carbon Transferred (Mt)	Direction	Carbon Transferred (Mt)	Direction	Carbon Transferred (Mt)	Direction
Positive carbon transfer	161.8	+	BS	341.00	+	251.37	+	246.59	+
			LCS	312.20	+	203.05	+	139.36	+
			HLS	263.70	+	164.67	+	97.44	+
Negative carbon transfer	−516.31	−	BS	−65.15	−	−0.56	−	−1.19	-
			LCS	−59.56	−	−0.44	−	−0.71	-
			HLS	−50.31	−	−0.33	−	−0.35	-
Net carbon transfer	−354.51	−	BS	275.85	+	250.82	+	245.39	+
			LCS	252.65	+	202.61	+	138.65	+
			HLS	213.38	+	164.33	+	97.09	+
H1	94.23	I&T→UCL	BS	297.63	I&T→FL	222.23	I&T→FL	286.75	I&T→FL
			LCS	87.28%		88.41%		82.73%
			HLS	272.28		179.57		166.75
Proportion of contribution	58.24%		BS	87.21%		88.46%		82.91%
			LCS	230.00		145.78		80.96
				87.22%		88.53%		83.09%
H2	365.53	UCL→I&T	BS	49.72	RSL→I&T	0.49	RSL→UCL	1.06	RSL→UCL
			LCS	76.31%		88.75%		89.03%
			HLS	45.45		0.37		0.54
Proportion of contribution	70.80%		BS	76.31%		85.65%		61.03%
			LCS	38.38		0.27		0.17
				76.29%		81.21%		48.63%

* Notes: ‘+’ and ‘−’ represent the direction of carbon transfer; H1 represents dominant positive carbon transfer; H2 represents dominant negative carbon transfer. FL: forest land, GL: grassland, WA: water area, UL: unused land, AL: arable land, RSL: rural settlement land, UCL: urban construction land, I&T: industrial traffic land. BS: baseline scenario, LCS: low-carbon scenario, HLS: high-limit scenario.

4. Discussion

The study indicates that the introduction of new technologies significantly contributes to energy savings and emission reductions in Jincheng. However, the city’s reliance on a traditional heavy-industry structure means that carbon emissions continue to be largely attributed to the industrial sector. Exploring strategies for industrial transformation presents a substantial opportunity for the city to achieve carbon neutrality. This transformation requires considerable capital investment, yet current financing channels are limited. It is essential to enhance the willingness of both the government and private investors to support green industrial initiatives. Effective top-level planning and supportive policies are vital for fostering this transformation and upgrading industry standards. In addition to strategic planning, there is a need for improved talent development to enhance specialization in green manufacturing, thereby boosting both efficiency and competitiveness. Prioritizing the green transformation of traditional industries should involve optimizing product structures and re-engineering processes. At the same time, emerging sectors such as new energy vehicles, photovoltaics, and other low-carbon industries should be expanded. Furthermore, the construction of zero-carbon factories and parks must be actively supported. The green transformation of industries is expected to stimulate the growth of energy-saving technologies, enhance environmental protection, and promote sustainable manufacturing, thereby providing new momentum for economic development. It will also drive the advancement of upstream and downstream enterprises, contributing to the optimization and upgrading of the overall economic structure.

Forest land is the predominant land type in Jincheng, functioning as a vital terrestrial carbon sink. The urban ecosystem maintains a stable carbon-sink capacity; however, there is limited potential for further enhancement. The transition of arable land from a carbon source to a carbon sink indicates its significant capacity for carbon sequestration. Nonetheless, high carbon emissions stemming from urban and industrial land use exacerbate the imbalance in carbon metabolism. This imbalance becomes increasingly evident with the adoption of new technologies and the drive for industrial transformation. The spatial relationship between urban land use and carbon emissions is notably close. In land simulations conducted under specified conditions, the spatial changes in urban land use demonstrate relative stability in developmental structure over time, and the trends in spatial differentiation of carbon metabolism density cannot be reversed in the short term. Therefore, urban spatial development should take into account its overall impact, implement mechanisms for coordinated carbon-emission reduction, and monitor the transfer of carbon between different regions. Integrating urban territorial space is essential for analyzing the adverse effects of an irrational spatial structure and understanding the underlying causes of ecological issues, thereby providing critical support for urban spatial planning. During the process of urbanization, it is imperative to rationally plan the development of unused land and the direction of urban expansion, strengthen the internal allocation, and emphasize the carbon-sink function of ecological spaces.

This study has several limitations that warrant acknowledgment. Firstly, the analysis primarily focused on surface-level carbon sources and sinks, omitting considerations of soil-based carbon dynamics. Secondly, the driving mechanisms underlying land-use carbon metabolism in coal-resource-based cities remain unexplored, representing a critical gap in understanding the spatial and temporal patterns of carbon emissions. Additionally, the model predictions incorporated numerous scenario parameters and neighborhood weights, which introduced complexities in mitigating systematic errors. The LEAP model, while robust in policy scenario simulation, exhibits inherent limitations and uncertainties in long-term carbon-emission projections. Its predictions are highly contingent on the accuracy and completeness of input data, such as energy consumption statistics, technical parameters, and economic growth rates, all of which may exhibit significant variability over extended time scales. Furthermore, the model assumes linear or fixed trends in technical efficiency and energy mix evolution, thereby inadequately capturing the impacts of non-linear factors like technological breakthroughs, market dynamics, and abrupt policy shifts. The carbon-emission factors are typically based on current technological levels, failing to account for potential advancements or energy substitutions. Socio-economic variables and policy implementation uncertainties further compound these challenges. Future research should prioritize enhancing simulation accuracy, refining carbon-emission accounting methodologies, and integrating regional planning considerations into scenario settings while improving precision across multiple scales. Lastly, this study focused on analyzing urban carbon-emission trends and their implications for future spatial development, deferring financial analysis of the proposed measures due to the extensive data requirements, including investment costs, operational expenses, policy subsidies, and market fluctuations. These data limitations will be addressed in subsequent studies, leveraging comprehensive datasets and advanced analytical methods to systematically assess the financial feasibility of mitigation strategies and further refine the research outcomes.

5. Conclusions

The study integrates the LEAP and Markov–PLUS models to simulate carbon emissions in urban areas. Taking Jincheng as an example, the study identifies spatial and temporal characteristics of future land carbon metabolism by predicting urban carbon-emission trends from 2020 to 2060 across various scenarios. This analysis aims to guide land-use planning for optimized emission reductions.

Jincheng’s total and net carbon emissions from 2020 to 2060 first increase, then gradually decline, with industrial and transport land contributing over 70% of emissions. In the baseline scenario, Jincheng’s emissions peak at 66.53 Mt in 2032, while the low-carbon scenario projects a peak of 59.72 Mt by 2029. The high-limit scenario achieves its peak earlier, around 2027, at 53.05 Mt. In the medium term, the most significant carbon reductions come from clean energy substitution and CCUS technologies. The highest long-term impact is from industrial restructuring and improvements in processing and conversion efficiency. Due to the rigidity of the city’s industrial structure, high-quality industrial transformation is challenging in the short term. However, new technologies can rapidly reduce emissions, and the traditional industrial base is being gradually replaced by high-tech industries, opening up significant opportunities for emission reductions.

From 2010 to 2020, the main positive-carbon-metabolism compartments were forests, grasslands, and waters, while industrial and urban land formed the negative compartments. Despite covering only 10% of the city’s area, net carbon emissions from negative compartments were 5 to 15 times greater than the carbon sequestration of positive compartments. Under the baseline scenario, limited land is converted from negative- to positive-carbon-metabolism compartments. In the low-carbon and high-limit scenarios, arable land transitions from negative- to positive-carbon-metabolism compartments. The area of negative compartments decreases, particularly in urban and industrial zones, where construction and industrial land dominate. With the implementation of technical measures, Jincheng’s future carbon sink is expected to surpass emissions, playing a key role in balancing the city’s carbon metabolism.

The net carbon transfer in Jincheng was negative from 2010 to 2020, indicating that land-use changes negatively impacted carbon metabolism. In the baseline scenario, the main source of positive carbon transfer is the conversion of industrial traffic land to forest land (I&T→F), and the main source of negative carbon transfer is the conversion of rural settlements to industrial traffic land (R→I&T). In the low-carbon scenario from 2030 to 2040, negative carbon metabolism is primarily driven by the conversion of rural settlements to urban land (R→UR). Compared to the baseline scenario, the city’s low-carbon transition should focus more on reducing emissions from urban land use than from the industrial transport sector. In the high-limit scenario, land-use changes shift from having a negative to a positive impact on urban carbon metabolism, with arable land, forest land, urban land, and industrial land serving as key points of carbon transfer. Changes in urban–rural dynamics also intensify their influence on carbon transfer.