Policy-Driven Land Use Optimization for Carbon Neutrality: A PLUS-InVEST Model Coupling Approach in the Chengdu–Chongqing Economic Circle

Abstract

In the context of global “dual carbon” objectives, land use dynamics exhibit a strong correlation with regional carbon storage. Facing significant ecological–economic conflicts, the Chengdu–Chongqing Economic Circle in western China necessitates multi-scenario modeling of carbon storage. This research integrates the PLUS model (simulation accuracy Kappa = 0.84) and InVEST model to project land use and carbon storage trajectories under natural development (NDS), urban development (UDS), carbon peak (CPS), and carbon neutrality (CNS) scenarios from 2030 to 2060, leveraging historical data from 2000 to 2020. The results show the following: (1) The study area is dominated by forest land and cultivated land (accounting for more than 90%). From 2000 to 2020, cultivated land decreased, and construction land increased; construction land continued to expand under all future scenarios. (2) Carbon storage showed a trend of first increasing and then decreasing, reaching 4974.55 × 10⁶ t in 2020 (an increase of 4.0 × 10⁶ t compared with 2000). The peak carbon storage in the CPS scenario reached 5015.18 × 10⁶ t, and the overall spatial pattern was “high around and low in the middle”. (3) The CPS achieved a carbon peak through intensive land use and ecological restoration, and the CNS further strengthened carbon sink protection and promoted carbon neutrality. Constructing a multi-scenario coupling model chain provides a new method for regional carbon management, which has important guiding significance for the low-carbon development of the Chengdu–Chongqing Twin Cities Economic Circle.

1. Introduction

Against the backdrop of increasing attention to global climate change and sustainable development, land use change and its impact on carbon storage has become a crucial research area. With the frequent occurrence of extreme climate events around the world, governments and international organizations have increased their investment in carbon emission reduction and climate adaptation strategies. At the 75th United Nations General Assembly in 2020, China announced that it would strive to achieve a peak in carbon dioxide emissions by 2030 and achieve carbon neutrality by 2060. Land use change directly affects the carbon fixation and release process by changing the vegetation cover, soil organic carbon content, and ecosystem carbon sequestration capacity, which, in turn, has a profound impact on the global carbon cycle and climate change [1]. Therefore, in-depth exploration of the impact mechanism of land use change on carbon storage is of great significance for formulating scientific carbon emission reduction strategies and achieving sustainable development goals.

Existing research has widely explored the effects of land use change (LUC) on carbon storage through diverse modeling techniques. Although models such as CLUE-S, FLUS, and CA-Markov are commonly used for land use simulation [2], the transition matrix approach utilized herein provides unique benefits for analyzing the specific characteristics of the Chengdu–Chongqing Economic Circle. Unlike purely spatial models, the transition matrix provides quantifiable metrics of land type conversions, which is particularly valuable for (1) tracking urbanization-driven transformations between agricultural/forest lands and built-up areas and (2) establishing baseline conversion probabilities for scenario projections. Global-scale meta-analyses (e.g., Huang et al.’s work on SOC impacts [3]) and national studies (e.g., Liu Xiaojuan’s FLUS-InVEST integration for China [4]) have demonstrated that transition matrices effectively capture dominant LUC pathways. This approach complements spatial modeling tools like InVEST (used by Imran et al. for mountain forests [5,6]) by explicitly quantifying conversion magnitudes between specific land classes, which is crucial for carbon storage calculations where forest-to-urban transitions yield different impacts than grassland-to-cropland dynamics. Regionally, while studies in eastern China (e.g., Beijing–Tianjin–Hebei [7]) have combined transition matrices with ecosystem service indices, western regions remain understudied. Our methodology advances prior work (e.g., Wu Dan’s PLUS-InVEST application [8]) by (1) incorporating initial land type proportions to weight transition probabilities, addressing the “initial scale neglect” limitation of conventional matrices, and (2) integrating topographic and socioeconomic drivers specific to the Chengdu–Chongqing context. This hybrid approach balances the quantitative rigor of transition analysis with spatial explicitness, offering improved precision for this rapidly urbanizing basin–mountain system, where carbon storage shows pronounced core–periphery gradients [8].

In 2011, the State Council officially approved the “Chengdu-Chongqing Twin Cities Economic Circle Regional Plan” [9]. In 2016, the National Development and Reform Commission and the Ministry of Housing and Urban-Rural Development jointly issued the “Chengdu-Chongqing Urban Agglomeration Development Plan”, which clearly stated the construction of a national-level urban agglomeration that leads the development and opening up of the western region. On 3 January 2020, the Sixth Meeting of the Central Financial and Economic Commission clearly proposed to promote the construction of the Chengdu–Chongqing Twin Cities Economic Circle and form an important growth pole for high-quality development in the west. On 16 October of the same year, the Political Bureau of the CPC Central Committee also held a meeting to review the “Outline of the Construction Plan for the Chengdu-Chongqing Twin Cities Economic Circle”. In 2022, the General Offices of the People’s Governments of Chongqing and Sichuan Province issued the “Action Plan for Promoting the Market Integration Construction of the Chengdu-Chongqing Twin Cities Economic Circle”. In 2023, the Reform and Opening-up Special Working Group of the Joint Office for Promoting the Construction of the Chengdu–Chongqing Twin Cities Economic Circle issued the “Key Tasks for Promoting the Market Integration Construction of the Chengdu-Chongqing Twin Cities Economic Circle in 2023”. This series of strategic decisions is aimed at expanding market space, optimizing and stabilizing the industrial chain and supply chain, and building a new development pattern with the domestic circulation as the main body and the domestic and international circulations mutually promoting each other.

PLUS (Patch-generating Land Use Simulation Model) is a land use change simulation tool developed by the HPSCIL@CUG Laboratory of China University of Geosciences. It has shown good applicability in many large-scale regional studies [10]. For example, Zhu used the PLUS model to simulate the ecological risk of the Chengdu–Chongqing Twin Cities Economic Circle (20.6 × 10⁴ km²) [11], and Yang applied it to the assessment of the ecosystem service value of the Guanzhong Plain Urban Agglomeration (10.7 × 10⁴ km²) [12]. These studies show that the PLUS model can effectively support the simulation and prediction of large-scale land use dynamics.

This study uses the PLUS model to make multi-scenario predictions of future land use dynamics in the Chengdu–Chongqing Twin Cities Economic Circle. The PLUS model incorporates multiple driving factors to dynamically simulate future land use dynamics in the Chengdu–Chongqing Twin Cities Economic Circle using historical datasets. This framework enables precise prediction of land use transitions under diverse carbon peak and carbon neutrality scenarios, offering robust support for analyzing carbon storage dynamics and advancing related policy objectives. This is its significant advantage in this study. The input parameters of the model include the elevation, slope, aspect, annual average temperature, average rainfall, GDP, population density, and distance to major transportation routes. These data are spatially processed and standardized using GIS technology. The PLUS model simulates the spatiotemporal dynamics of land use change by integrating the Markov transition probability matrix and CA-Markov model with the land use suitability atlas.

The Chengdu–Chongqing Economic Zone is located in the upper reaches of the Yangtze River and has superior ecological resources. It is the most densely populated, industrially prosperous, and innovative region in western China. However, rapid urbanization has led to the expansion of construction land, and large amounts of forest land, agricultural land, and grassland with strong carbon sequestration capacity have been occupied, resulting in a significant reduction in regional carbon storage and serious damage to ecosystem service functions [13]. This phenomenon not only affects regional sustainable development but may also have an adverse impact on the ecological barrier function of the upper reaches of the Yangtze River. Therefore, scientific prediction of future land use dynamics and their impact on carbon storage is of great significance for formulating reasonable land use planning and carbon reduction strategies.

Based on the coupling method of PLUS and InVEST models, this study uses land use data from 2000, 2010, and 2020 to make multi-scenario predictions on future land use dynamics and carbon storage in the Chengdu–Chongqing Twin Cities Economic Circle. Compared with existing studies, the objectives of this study are mainly reflected in (1) accurately planning the carbon peak path of the Chengdu–Chongqing Twin Cities Economic Circle in 2030 and the carbon neutrality realization plan in 2060, clarifying the changing trends and key emission reduction nodes of carbon storage at different stages; (2) promoting the green and low-carbon transformation of industries in the Chengdu–Chongqing Twin Cities Economic Circle, using the carbon peak and carbon neutrality scenario research, exploring and cultivating emerging low-carbon industries, and achieving a win–win situation for economic and ecological benefits; and (3) strengthening the coordination and cooperation among cities within the Chengdu–Chongqing Twin Cities Economic Circle in responding to climate change and promoting the all-around linkage of regional resources, industries, and policies based on the research results of carbon storage dynamics.

2. Materials and Methods

2.1. Study Area

The Chengdu–Chongqing Twin Cities Economic Zone is located in the upper reaches of the Yangtze River and in the hinterland of the Sichuan Basin (Figure 1). It spreads outward with Chengdu and Chongqing as the dual cores. It is the most developed and most potential urbanized area in western China. This region plays a crucial role in implementing China’s Yangtze River Economic Belt strategy and advancing the Belt and Road Initiative. It is the fourth pole of China’s economic growth and the strategic hinterland of the inland economy. It has the unique advantage of connecting the southwest and northwest and connecting East Asia with Southeast Asia and South Asia. Specifically, the Chengdu–Chongqing Twin Cities Economic Zone is connected to Beichuan, Pingwu, Guangyuan, Bazhong, and Wanyuan in Mianyang, Sichuan, in the north; to some districts and counties of Ya’an and the three autonomous prefectures of Aba, Ganzi, and Liangshan in the west; to Zhaotong in Yunnan and Bijie and Zunyi in Guizhou in the south; and to some districts and counties of Chongqing and Enshi Autonomous Prefecture in Hubei in the east. It covers 113 districts and counties in 15 prefecture-level cities in Sichuan Province, such as Chengdu and Mianyang, and the central urban area of Chongqing City and 27 districts and counties such as Wanzhou and Fuling, with a total area of about 18.66 × 10⁴ km² [9]. In 2024, the Chengdu–Chongqing Economic Circle will achieve a regional GDP of CNY 8719.3 billion, accounting for about 6.9% of the national GDP [13]. China’s western region has always been regarded as a population, industrial, and commercial center with great potential and has always been the backbone of China’s economy. However, due to the rapid economic development and fast-paced urban construction in recent years, severe challenges have been posed to the ecological environment, making the shortage of domestic land increasingly serious. As a result of this conflict, the region’s ecosystem has been severely damaged, and it is also facing tremendous pressure to maintain local carbon emissions and respond to global warming.

2.2. Data Source and Processing

The study selected the DEM, slope, aspect, climate factors (temperature and rainfall), socioeconomic factors (GDP and population density), and traffic accessibility (distance from main roads) as driving factors because they can comprehensively reflect the synergistic effects of natural geographical constraints, human activity intensity, and regional development patterns on land use change, which is in line with the evolution characteristics of the “mountain–city” complex system of the Chengdu–Chongqing Twin Cities Economic Circle. The data mainly come from the Geospatial Data Cloud, the China Science Resources and Environmental Science Data Center, the WorldPop Global 100m Population Grid Data, and the National Geographic Information Resource Directory Service System. In order to ensure the accuracy and consistency of the data, this study strictly processed and verified the data. The specific steps include data download, format conversion, reclassification, clipping, and normalization, which can clearly demonstrate the entire process and methods of the research (Figure 2).

In this study, the land use data of the Chengdu–Chongqing Twin Cities Economic Circle during 2000–2020 were used. Indicators such as the DEM, slope, aspect, average temperature, average rainfall, GDP, population density, distance from traffic hubs, and distance from highways are regarded as driving forces affecting land use (

Table 1. Specific sources of data on driving factors of land use change.

Land Use Drivers	Data Name	Data Source
Natural factors	Elevation	Geospatial Data Cloud (https://www.gscloud.cn)
	Slope	Extraction based on DEM
	Slope aspect
	NDVI	Geospatial Data Cloud (https://www.gscloud.cn) is based on Landsat series remote sensing data calculation
	Average annual temperature	Chinese Science Resources and Environmental Science Data Center (https://www.resdc.cn)
	Average annual precipitation
Social factors	GDP	Chinese Science Resources and Environmental Science Data Center (https://www.resdc.cn)
	Population density	worldpop Global 100 m population grid data (https://www.worldpop.org/)
Traffic factors	Distance from highway	National Geographic Information Resource Directory Service System (https://www.webmap.cn/); ArcGIS Euclidean distance analysis
	Distance from main road
	Distance to railway
	Distance to the train station
	Distance to secondary roads
	Distance to third-level road
	Distance to river

). By applying the PLUS model in a large-scale space with a resolution of 100 m, this study aims to evaluate its effectiveness and accuracy in simulating land use in the Chengdu–Chongqing Twin Cities Economic Circle in 2030 and 2060. By using both InVEST and Plus models, we can study the temporal and spatial development of carbon emission storage in the Chengdu–Chongqing Twin Cities Economic Circle from 2000 to 2060. In terms of data processing, this study first reclassified and clipped the land use data for the three periods of 2000, 2010, and 2020 to ensure the consistency and comparability of the data. Then, the Markov model was used to calculate the transition matrix of land use types, and future land use dynamics were predicted based on the matrix. In order to improve the prediction accuracy, this study also normalized the driving factors to ensure that the influence of each factor in the model is consistent.

2.3. Research Methods

This study uses the coupling method of the PLUS model (patch generation land use simulation model) and the InVEST model (ecosystem service assessment model), combined with the land use transfer matrix and multi-scenario simulation, to systematically analyze the land use dynamics in the Chengdu–Chongqing Twin Cities Economic Circle and its impact on carbon storage. First, based on the land use data from 2000 to 2020, the PLUS model is used to simulate the dynamics in land use patterns under different future development scenarios (natural development, urban development, carbon peak, and carbon neutrality), and the land conversion law is quantified through the transfer matrix. Subsequently, combined with the carbon storage module of the InVEST model, the dynamic dynamics of carbon storage under different scenarios are calculated to reveal the impact mechanism of land use dynamics on regional carbon balance. The study uses multi-source data (such as climate, topography, socioeconomics, etc.) to drive the model to ensure the accuracy of the simulation and provide a scientific basis for the low-carbon land space planning and carbon neutrality path optimization of the Chengdu–Chongqing Twin Cities Economic Circle.

2.3.1. PLUS Model

In the Land Expansion Analysis Strategy (LEAS) module of the PLUS model, a random sampling mechanism is used to optimize computational efficiency. At the same time, a random forest algorithm is used to process high-dimensional data and solve the multicollinearity problem between driving factors, thereby effectively exploring the transfer rules between land use types and calculating the development probability of various types of land use [14]. Neighborhood weight settings are used under different scenarios (

Table 2. Neighborhood weight values under different scenarios.

	Agriculture	Forest	Grass	Water	Construction	Unused Land
2030NDS	0.48	0.48	0.48	0.5	0.52	0.48
2030UDS	0.25	0.25	0.25	0.3	0.75	0.25
2030CPS	0.7	0.7	0.7	0.7	0.3	0.3
2030CNS	0.75	0.75	0.75	0.75	0.25	0.3
2060NDS	0.5	0.5	0.5	0.5	0.5	0.5
2060UDS	0.2	0.2	0.2	0.2	0.8	0.2
2060CNS	0.3	0.3	0.3	1	0.7	0.4
2060CNS	0.7	0.7	0.7	1	0.4	0.45

). Its mathematical expression is as follows:

$$P_{i,k(X)}^d={\displaystyle \frac{{\sum }_{n=1}^{M}I[h_n(X)=d]}{M}}$$

(1)

where $\mathrm{X}$ is a vector of driving factors; hn(X) is the predicted land use type calculated when the number of decision trees is n; I[∙] is the indicator function of the decision tree; $\mathrm{M}$ is the number of decision trees; $\mathrm{d}$ is 0 or 1, where 1 means that other land use types are transformed into land use type k, and 0 means that there is no transformation to land use type k; and $P_{i,k(X)}^d$ is the probability of the increase in land use type k at spatial unit i when d is 0 or 1 [15].

2.3.2. Land Use Transfer Matrix

The land use transition matrix is an important method for analyzing regional land use dynamics, which can quantitatively describe the mutual transformation relationship between different land types. This method is based on historical land use data, constructs a transition matrix on a time series, and reveals the dynamic characteristics of land use dynamics and their spatial distribution laws [16]. Transfer matrix settings are used under different scenarios (Figure 3). The mathematical expression is:

$$S_{ij}=\left[\begin{array}{ccccc}{S}_{11}& S_{12}& S_{13}& \cdots & S_{1n}\\ S_{21}& S_{22}& S_{23}& \cdots & S_{2n}\\ S_{31}& S_{32}& S_{33}& \cdots & S_{3n}\\ ⋮& ⋮& ⋮& ⋮& ⋮\\ S_{n1}& S_{n2}& S_{n3}& \cdots & S_{nn}\end{array}\right]$$

(2)

where $\mathrm{S}$ is the land area; $\mathrm{n}$ is the number of land use types; and $\mathrm{i}$ and $\mathrm{j}$ are the land use types at the beginning and end of the study period, respectively.

Single land use dynamic refers to the change in a certain land use type in the study area within a certain study period, which is used to indicate the change speed and change range of different land use types within a certain period [17]. Its expression is:

$$K={\displaystyle \frac{U_b-U_a}{U_a}}\times {\displaystyle \frac{1}{T}}\times 100\%$$

(3)

where $\mathrm{K}$ is the dynamic degree of a certain land use type during the study period; and ${\mathrm{U}}_{\mathrm{a}}$ and ${\mathrm{U}}_{\mathrm{b}}$ are the quantities of a certain land use type at the beginning and end of the study period, respectively.

Comprehensive land use dynamics refer to the change speed of the entire land use type within a certain time range in the study area. It is an indicator that describes the regional differences in the change speed of land use types and can reflect the comprehensive impact of social and economic activities in the region on land use dynamics [18]. Its expression is:

$$S=({\sum }_{n=1}^m(\mathsf{\Delta }{S}_{i-j}/S_i))\times 1/t\times 100\%$$

(4)

where ${\mathrm{S}}_{\mathrm{i}}$ is the total area of land use type i at the beginning of monitoring; $\mathsf{\Delta }{\mathrm{S}}_{\mathrm{i}-\mathrm{j}}$ is the total area of land use type $\mathrm{i}$ converted to other land use types from the beginning to the end of monitoring; and $\mathrm{t}$ is the time period. The land use dynamics $\mathrm{S}$ reflects the land use change rate of the research sample area corresponding to the time period t.

2.3.3. Scenario Design

According to China’s commitment to increase its national voluntary contributions at the 75th United Nations General Assembly in 2020, more powerful policy measures will be implemented to strive to peak carbon dioxide emissions by 2030 and achieve carbon neutrality by 2060. Focusing on the carbon neutrality goal of the Chengdu–Chongqing Twin Cities Economic Circle, according to the strategic planning requirements of policy documents such as the Outline of the Construction Plan for the Chengdu–Chongqing Twin Cities Economic Circle, the Construction of the Twin Cities Economic Circle, the Development Plan of the Chengdu–Chongqing Urban Agglomeration, the Notice of the People’s Government of Sichuan Province on Issuing the Implementation Opinions on the Red Line of Ecological Protection in Sichuan Province, and the Chongqing Land Space Ecological Protection and Restoration Plan (2021–2035), forest land, grassland, cultivated land, and water areas are strictly regulated and restricted from being converted into construction land, aiming to achieve ecological environmental protection by restricting the expansion of construction land [19]. In order to comprehensively assess the uncertainty of future land use dynamics and carbon storage in the Chengdu–Chongqing Twin Cities Economic Circle, this study covers various development possibilities from different angles by setting a natural development scenario to reflect the conventional development state, an urban development scenario to focus on the impact of urban expansion, a carbon peak scenario to fit the phased policy goals, and a carbon neutrality scenario to focus on the long-term strategic direction. The four scenarios are natural development scenario (NDS), urban development scenario (UDS), carbon peak scenario (CPS), and carbon neutrality scenario (CNS) (

Table 3. Scenario descriptions.

Scenario Type	Scenario Introduction
NDS	(1) NDS is a simulation of urban agglomeration land use change from 2000 to 2020, without mandatory constraints, and in accordance with the current urbanization development model. (2) It continues the historical trend, does not set restrictions on the conversion of various types of land use, and does not involve government and market intervention. It is the basis for the simulation of urban agglomeration land use change considering other constraints.
UDS	(1) UDS prioritizes construction. It is assumed that the probability of conversion of unused cultivated land, forest land, and grassland to construction land increases, and the probability of conversion of construction land to other land types decreases, so as to reflect the characteristics of land use change in the process of rapid urbanization. (2) The probability of converting unused cultivated land, forest land, and grassland to construction land increases by 20%, and the probability of artificial surface transfer to other landscape types except cultivated land decreases by 30%.
CPS	(1) CPS encompasses ecological red line control, strictly controlling the growth of construction land, promoting economical and intensive use, and reducing the encroachment on cultivated land and other ecological land. (2) It encourages the conversion of unused land to cultivated land and ecological land to improve the regional carbon sequestration capacity. The transfer ratio of cultivated land to forest land and grassland will increase by 20%; the transfer ratio of forest land and grassland to cultivated land will decrease by 20%, and the conversion ratio to construction land will decrease by 50%; the transfer ratio of water areas to cultivated land will decrease by 50%, and the transfer ratio to construction land will decrease by 70%; the transfer ratio of construction land to forest land, grassland, and water areas will increase by 30%; the transfer ratio of unused land to forest land and grassland will increase by 20%; and the transfer ratio to construction land will increase by 60%.
CNS	(1) This involves carbon tax + compensation. On the basis of CPS, this will further strengthen the protection of carbon sinks such as forest land, grassland, and water areas and strictly control the expansion of construction land and the encroachment of ecological land. (2) This will increase regional carbon sinks through measures such as returning farmland to forest and grassland and achieve the goal of carbon neutrality. The proportion of cultivated land transferred to construction land is reduced by 60%; the proportion of forest land, grassland, and water areas transferred to construction land is reduced by 30%; the proportion of construction land transferred to cultivated land is increased by 20%; the proportion of transfer to forest land, grassland, and water areas is increased by 15%; the proportion of unused land transferred to cultivated land is reduced by 15%; the proportion of transfer to forest land, grassland, and water areas is reduced by 10%; and the proportion of transfer to construction land is increased by 60%.

).

2.3.4. InVEST Model

The InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs) model is a tool for evaluating ecosystem service functions, in which the carbon module is specifically used to estimate the carbon storage of surface ecosystems. The model calculates the carbon storage and its changing trend in the study area by inputting carbon density data and land use change data of different land use types. The InVEST model has been widely applied in global studies due to its parsimonious parameters and computational efficiency. Carbon storage estimation involves four aspects: surface biocarbon, underground biocarbon, soil biocarbon, and dead biocarbon [20]. Based on relevant research results and field survey data domestic and overseas, this study determined the carbon density values of different land use types in the Chengdu–Chongqing Twin Cities Economic Circle (as shown in

Table 4. Carbon densities of various land types in the Chengdu–Chongqing Economic Zone (t/ha).

	Aboveground Carbon Density	Underground Carbon Density	Soil Carbon Density	Death Carbon Density
Agriculture	38.7	80.7	92.9	1
Forest	55.56	144.87	206.45	3.5
Grass	29.3	52.9	135	1
Water	21.4	73.1	113	1
Construction	3.3	87.3	115.3	0
Unused land	22.6	136.9	171.8	0

). Then, combined with land use change data [21], the carbon module of the InVEST model was used to estimate the future carbon storage of the Chengdu–Chongqing Twin Cities Economic Circle. The carbon storage calculation formula is:

$$C_{\mathit{\text{i-total}}}=C_{\mathit{\text{i-above}}}+C_{\mathit{\text{i-below}}}+C_{\mathit{\text{i-dead}}}+C_{\mathit{\text{i-soil}}}$$

(5)

where $\mathrm{i}$ is the $\mathrm{i}$-th land use type; ${\mathrm{C}}_{\text{i-above}}$ is the aboveground carbon storage of the i-th land use type; ${\mathrm{C}}_{\text{i-below}}$ is the underground carbon storage of the i-th land use type; ${\mathrm{C}}_{\text{i-dead}}$ is the dead organic carbon storage of the i-th land use type; ${\mathrm{C}}_{\text{i-soil}}$ is the soil carbon storage of the i-th land use type; and ${\mathrm{C}}_{\text{i-total}}$ is the total carbon storage of the i-th land use type [22].

By using the carbon module, researchers can accurately estimate carbon storage in surface ecosystems. The accuracy of this estimate is affected by many factors. The forest resource survey in Chongqing, the average carbon density of various land use types in different climate zones, and the research results of Baoxing County, Sichuan, all provide a reliable basis for obtaining carbon density data in the Chengdu–Chongqing Twin Cities Economic Circle so as to better evaluate the carbon storage of surface ecosystems [23]. The research results show that Zhu Wenbo, Guo Jingjing, Chuai Xiaowei, and others took into account the carbon densities of water bodies and construction land in their studies in the Qihe River Basin of Taihang Mountains, the Yangtze River Basin, and Jiangsu Province so as to better reflect the differences in local carbon storage [24,25,26]. Among them, the national level of aboveground carbon density values of cultivated land and grassland was obtained based on the relevant research results of Li Kerang, Fang Jingyun, and others [27,28]. The aboveground carbon density of forest land is derived from the existing research of Wang Shaoqiang, Xie Xieli, and others [29,30]. In addition, the carbon density of construction land and unused land is obtained based on the existing research results of Chuai Xiaowei, Tang Lin [26,31]. The specific carbon density values of the Chengdu–Chongqing Economic Circle can be found in Table 4.

3. Results Analysis

3.1. Dynamics in Land Use from 2000 to 2060

3.1.1. Spatiotemporal Characteristics of Land Use Dynamics

This study uses the PLUS model to simulate the land use dynamics in the Chengdu–Chongqing Twin Cities Economic Circle, with a simulation resolution of 100 m and a coverage area of approximately 18.66 × 10⁴ km². To evaluate the simulation accuracy of the model, the land use data of 2000 and 2010 were used as the benchmark to simulate the land use in 2020 and 2030, respectively, and the real data of 2010 and 2020 were used for verification. The accuracy analysis results show that the Kappa coefficient of 2020 simulated based on the data of 2000 and 2010 is 0.84, and the Kappa coefficient of 2030 simulated based on the data of 2010 and 2020 is 0.91, both exceeding the reliability threshold of 0.75. The overall accuracy is 0.90 and 0.91, respectively, indicating that the PLUS model has a good simulation effect on the land use dynamics in the Chengdu–Chongqing Twin Cities Economic Circle at a resolution of 100 m and is suitable for the prediction of future land use scenarios. To further verify the significance of the simulation results, this study used paired sample t-test to statistically analyze the land use dynamics under different scenarios. The results showed that the land use dynamics between the scenarios were significantly different (p < 0.05), indicating that the scenario design can effectively reflect the land use change trends under different policy objectives.

Land use in the Chengdu–Chongqing Economic Circle is dominated by agricultural land, which is widely distributed in the central plain area. Forest land is mainly concentrated in the Qionglai Mountains and Min Mountains in the west, the Daliang Mountains and Dalou Mountains in the south, the Wushan Mountains in the east, and the parallel ridge valley area in eastern Sichuan. The distribution of construction land is relatively concentrated, being mainly located in the two core cities of Chengdu in the northwest and Chongqing in the southeast (Figure 4). The area of unused land is small, indicating that the regional land utilization rate is high and that the development intensity is high. From the trend of land use dynamics between 2000 and 2020, agricultural land remained the dominant land use type, although exhibiting a gradual decline over time. Forest cover increased prior to 2010 but declined thereafter, while construction land expanded rapidly, particularly during 2010–2020, with a growth rate of 4.57%. The area of grassland and water areas changed relatively little, and the area of unused land changed significantly between 2000 and 2010. The dynamic degree of single land use was 8.42%, but it tended to be stable after 2010. The overall dynamics of comprehensive land use in the study area are relatively low: 0.15% from 2000 to 2010 and 0.14% from 2010 to 2020 and from 2000 to 2020 (

Table 5. The proportion of land use area and land use dynamics in the Chengdu–Chongqing Economic Circle from 2000 to 2020.

Land Use Type	2000			2010			2020
	Area (km2)	Proportion%	2000–2010 Dynamics	Area (km2)	Proportion%	2010–2020 Dynamics	Area (km2)	Proportion%	2000–2020 Dynamics
Agriculture	118,483.51	63.48%	−0.13%	116,946.02	62.66%	−0.17%	114,904.55	61.57%	−0.30%
Forest	50,029.48	26.81%	0.10%	50,552.81	27.09%	−0.03%	50,400.09	27.00%	0.07%
Grass	12,187.24	6.53%	−1.09%	10,862.57	5.82%	−0.33%	10,500.62	5.63%	−1.38%
Water	2795.35	1.50%	0.83%	3028.10	1.62%	0.81%	3274.49	1.75%	1.71%
Construction	3041.36	1.63%	6.64%	5062.16	2.71%	4.57%	7375.53	3.95%	14.25%
Unused land	101.27	0.05%	8.42%	186.55	0.10%	−0.19%	182.93	0.10%	8.06%
Comprehensive dynamics and total	186,638.21	100%	0.15%	186,638.21	100%	0.14%	186,638.21	100%	0.14%

). Specifically, they are as follows:

2000–2010: The change in unused land is the most significant, with single land use dynamics of 8.42%. The area of grassland and agricultural land decreases, and the change rate is slow. The area of forest land, water areas, and construction land increases.

2010–2020: The change in construction land is the most significant, with single land use dynamics of 4.57%. The area of agricultural land and forest land decreases, with dynamics of −0.17% and −0.03%, respectively.

Land use dynamics under different scenarios are mainly affected by the following driving factors:

Natural development scenario (NDS): The expansion of construction land is mainly driven by population growth and economic development, and the reduction of agricultural land and forest land exhibits a strong correlation with the urbanization process.

Urban development scenario (UDS): High-intensity urbanization leads to the rapid expansion of construction land, a significant reduction in agricultural land and forest land, and a significant decline in carbon storage.

Carbon peak scenario (CPS): Under policy intervention, the expansion of construction land is restricted, the area of forest land and water areas is protected, and the downward trend of carbon storage slows down.

Carbon neutral scenario (CNS): By optimizing the land use structure and increasing the area of forest land and water areas, carbon storage shows a rebound trend.

3.1.2. Dynamics in Land Use from 2000 to 2020

From 2000 to 2020, agricultural land accounted for the largest proportion in the study area, followed by forest land, while grassland, water areas, construction land, and unused land accounted for a relatively small proportion. In the past 20 years, agricultural land was the main land use type in the study area, accounting for more than 60%; the agricultural land area continued to decrease. The construction land area increased slightly, but the total proportion was small. The proportion of construction land increased to 3.95%, the grassland area decreased, and the proportion decreased by nearly 1% in the past 20 years. The water area increased slightly; the unused land area increased significantly before 2010 but stabilized after 2010 (Table 5). The Chengdu–Chongqing Economic Circle experienced significant land use dynamics from 2000 to 2020. The agricultural land area continued to decrease, from 63.48% in 2000 to 61.57% in 2020, which reflects the pressure of urbanization and industrialization on agricultural land. At the same time, the area of forest land remains relatively stable, which is an important part of land use in the Chengdu–Chongqing Twin Cities Economic Circle and is of great significance for maintaining regional ecological balance and carbon storage. The rapid growth of construction land, from a low proportion in 2000 to a high level in 2020, indicates a clear trend of urban expansion.

The current land use situation in the Chengdu–Chongqing Twin Cities Economic Circle from 2000 to 2020 shows that agricultural land is dominant, forest land is supplemented, construction land is growing rapidly, grassland and water areas change differently, and the area of unused land increases first and then stabilizes. This trend of change reflects both the needs of regional economic development and the importance of ecological and environmental protection.

During the 20 years from 2000 to 2020, the areas of the six types of land have changed significantly, as follows:

Agricultural land occupies a dominant position: In the study area, the agricultural land area has always occupied the largest proportion, maintaining above 60%, and is the main type of land use in the Chengdu–Chongqing Twin Cities Economic Circle. However, despite the large base of agricultural land area, the agricultural land area has shown a continuous decreasing trend in the past 20 years, reflecting that agricultural land may be affected by factors such as urbanization and industrialization.

Forest land is an important component: Forest land is the second largest type of land use in the Chengdu–Chongqing Twin Cities Economic Circle, and its area is relatively stable, which plays an important role in maintaining regional ecological balance and carbon storage.

There is rapid growth of construction land: Although the construction land area accounts for a relatively low proportion overall, its increase is very significant. From 1.63% in 2000 to 3.95% in 2020, the proportion has doubled (Table 5). This shows that in the process of urbanization, the demand for construction land in the Chengdu–Chongqing Economic Circle has increased sharply, which has promoted the rapid development of the regional economy.

The dynamics in grassland and water areas are different: The grassland area has decreased in 20 years, and the proportion has dropped by nearly 1 percentage point, which may be squeezed by factors such as agricultural expansion and urbanization. The water area has increased slightly, which may be related to the implementation of regional water resource protection and water ecological restoration measures.

The area of unused land increased first and then stabilized: the area of unused land increased significantly before 2010 but then flattened. This may reflect that the region was relatively active in the development and utilization of unused land in the initial stage, but as time went on, the difficulty of development increased, and with the improvement of policy regulation and ecological protection awareness, the development speed of unused land gradually slowed down.

The current land use situation in the Chengdu–Chongqing Economic Circle from 2000 to 2020 shows that agricultural land is the main land type, forest land is the secondary land type, construction land is growing rapidly, grassland and water areas are changing differently, and the area of unused land increases first and then stabilizes (

Table 6. Land use transfer matrix of Chengdu–Chongqing economic circle from 2000 to 2020 (km²).

Land Type	Agriculture	Forest	Grass	Water	Construction	Unused Land	Total	Transfer Out
Agriculture	107,681.65	4801.64	950.52	629.56	4427.26	13.71	118,504.34	10,822.69
Forest	4850.33	44,010.78	667.83	121.11	289.51	88.68	50,028.25	6017.47
Grass	1686.89	1477.87	8836.43	75.45	85.31	10.69	12,172.63	3336.20
Water	295.52	38.31	10.33	2384.43	63.27	4.48	2796.35	411.91
Construction	425.84	42.01	21.13	45.02	2494.11	0.23	3028.33	534.22
Unused land	7.16	7.45	2.29	18.02	1.39	64.60	100.90	36.30
Total	114,947.38	50,378.05	10,488.53	3273.59	7360.84	182.40	186,630.80	21,158.80
Transfer in	7265.73	6367.27	1652.10	889.15	4866.73	117.80	21,158.80

). This trend of change reflects both the needs of regional economic development and the importance of ecological and environmental protection.

3.1.3. Multi-Scenario Prediction of Future Land Use

The FLUS model was used to simulate the land use types of the Chengdu–Chongqing Economic Circle under four scenarios—NDS, UDS, CPS, and CNS—in 2030 and 2060. The simulation results are as follows (Figure 5):

Compared with 2020, in 2030, under the NDS, the areas of agricultural land, forest land, grassland, and unused land have successively decreased, while the areas of construction land and water areas have increased. Among them, the increase in construction land area is the most significant. The construction land area has increased by 2017 km², which is three times that of 2000 and nearly twice the construction area in 2010, indicating that urbanization has expanded rapidly under the NDS in 2030; under the UDS in 2030, the construction land area has increased by 2813 km² compared with 7376 km² in 2020, increasing to 10,189 km² in 2030.

Under the CPS in 2030, the areas of agricultural land and unused land have decreased, while the areas of forest land, grassland, water areas, and construction land have increased under the background of carbon peak. Finally, under the background of carbon neutrality, the dynamics in the areas of various types are not as drastic as in other scenarios. The areas of agricultural land, forest land, and grassland have all decreased, while the water areas and construction land areas have increased. Specifically, the agricultural land area decreased by 116 km², the forest area decreased by 84 km², and the grassland area decreased by 302 km². The water area increased by 255 km², and the construction land area increased by 247 km² (Figure 6).

Under the CNS in 2060, compared with the CPS in 2030, the agricultural land, forest land, and grassland areas decreased in all four scenarios, except for the increase in forest land under the CPS. Compared with the CPS in 2030, the agricultural land area under the CNS in 2060 decreased the most, from 111,410 km² to 105,635 km², a decrease of 5775 km² (Figure 6). This change may reflect the trend of agricultural land shifting to more efficient and environmentally friendly utilization methods driven by the carbon neutrality goal. At the same time, it may also be related to the continued advancement of urbanization, the reconfiguration of land resources, and the implementation of ecological protection policies. It is worth noting that the forest area increased under the CPS, while the forest area decreased slightly under the CNS. Similar to the agricultural land area, the grassland area decreased in all four scenarios, which may be related to the reallocation of land resources and the implementation of ecological protection policies. Under the CNS, the reduced grassland may be partially converted into forest land or other ecological land to further enhance the carbon sink capacity of the region. This reflects the importance and protection of forest resources in the Chengdu–Chongqing Twin Cities Economic Circle under the goal of carbon neutrality because forests are an important part of carbon sinks and play an important role in absorbing and storing carbon dioxide.

The water area increased in all four scenarios, which may be related to the strengthening of regional water resources management and ecological protection policies. As an important part of the ecosystem, water areas are of great significance for maintaining ecological balance and biodiversity. At the same time, water areas are also one of the carbon sinks, and the increase in their area will help improve the carbon absorption capacity of the region. Under the UDS in 2060, the construction land area increased by 6015 km² compared with 2030, reaching 16,204 km² (Figure 6). This increase is the largest among the four scenarios, reflecting the continuous advancement of urbanization and the demand for land resources for economic development. However, under the CNS, although the construction land area will also increase, it may pay more attention to green and low-carbon construction methods to reduce the negative impact on the environment. The land use dynamics in the Chengdu–Chongqing Twin Cities Economic Circle under the CNS in 2060 reflect the reconfiguration and optimal use of land resources driven by the carbon neutrality goal. Measures such as reducing the area of agricultural land, increasing the area of forest land and water areas, and reasonably controlling the area of construction land are aimed at improving the carbon sink capacity and ecological environment quality of the region and achieving sustainable economic and social development.

3.2. Dynamics in Carbon Storage from 2000 to 2060

3.2.1. Spatial and Temporal Characteristics of Carbon Storage

From 2000 to 2020, carbon storage in the Chengdu–Chongqing Twin Cities Economic Circle has undergone significant dynamics, showing an overall trend of “first increasing and then decreasing”. Carbon storage in the Chengdu–Chongqing Twin Cities Economic Circle presents a spatial distribution pattern of “high around and low in the middle”.

Initial growth stage (2000–2010): During this stage, the land use carbon storage in the Chengdu–Chongqing Economic Circle increased steadily from 4970.55 × 10⁶ t in 2000 to 4979.61 × 10⁶ t in 2010, an increase of 9.06 × 10⁶ t (Figure 7 and Figure 8). This shows that during this decade, land use dynamics in the region (such as increased forest cover and improved agricultural land management) had a positive impact on the increase in carbon storage.

Late adjustment stage (2010–2020): Entering the second decade of the 21st century, carbon storage in the Chengdu–Chongqing Economic Circle showed a downward trend, from 4979.61 × 10⁶ t in 2010 to 4974.55 × 10⁶ t in 2020, a decrease of 5.06 × 10⁶ t. This change may reflect the negative impact of dynamics in land use patterns (such as expansion of construction land, deforestation, etc.) on carbon storage in the context of economic development and accelerated urbanization in the region.

Overall dynamics: Although there was a decrease in carbon storage between 2010 and 2020, carbon storage in the Chengdu–Chongqing Twin Cities Economic Circle still achieved positive growth in the 20-year period, from 4970.55 × 10⁶ t in 2000 to 4974.55 × 10⁶ t in 2020, a total increase of 4.00 × 10⁶ t. This shows that despite fluctuations, the carbon storage capacity of the region has increased overall.

The current status of carbon storage in the Chengdu–Chongqing Twin Cities Economic Circle from 2000 to 2020 showed a dynamic process, which experienced both initial growth and later adjustments, but generally maintained positive growth in carbon storage. This change reflects the region’s efforts and challenges in seeking a balance between economic development and ecological environmental protection.

3.2.2. Dynamics in Carbon Storage Under Different Scenarios

Using the carbon storage module of the InVEST model, combined with the carbon density table and land use data, it is calculated that the overall trend of carbon storage in the Chengdu–Chongqing Twin Cities Economic Circle in 2030 is a slow increase, and this change is significantly synchronized with the change in forest area. This is mainly due to the high carbon density and large carbon storage of forest land in the region and its high proportion in the land use structure.

In 2030, the Chengdu–Chongqing Economic Circle reached peak carbon storage under the CPS, but did not reach peak carbon storage under the NDS, UDS, and CNS.

2020 to 2030 (different scenarios):

NDS: Carbon storage is expected to decrease from 4974.55 × 10⁶ t in 2020 to 4969.22 × 10⁶ t in 2030, a slight decrease (about 5.33 × 10⁶ t). UDS: Carbon storage further decreased to 4967.44 × 10⁶ t, and the decrease was greater than that in 2020 (about 7.11 × 10⁶ t). CPS: Carbon storage increased to 4996.12 × 10⁶ t, the highest among all scenarios, reflecting the enhancement measures for carbon sinks under this scenario. CNS: Carbon storage recovered to a level close to 2020, which was 4972.44 × 10⁶ t, indicating that this scenario is more effective in maintaining stable carbon storage (Figure 7 and Figure 9).

2030 to 2060 (different scenarios):

NDS: Carbon storage is expected to continue to decline to 4966.60 × 10⁶ t. UDS: Carbon storage further declines to 4955.29 × 10⁶ t, showing the continued impact of urbanization on carbon storage. CPS: Carbon storage increases significantly to 5015.18 × 10⁶ t, the highest among all forecast scenarios, indicating that this scenario has a significant effect on enhancing carbon sinks. CNS: Carbon storage remains at a high level of 4982.72 × 10⁶ t, indicating that this scenario performs well in balancing carbon emissions and carbon absorption. CPS: Under the CPS, the land use carbon storage of the Chengdu–Chongqing Twin Cities Economic Circle reaches 4996.12 × 10⁶ t, an increase of 21.60 × 10⁶ t compared to 2020 (Figure 9 and Figure 10). This increase shows that under specific policies and management measures (i.e., CPS), the carbon storage of the Chengdu–Chongqing Twin Cities Economic Circle can be effectively controlled and reach its peak. This is a positive signal that through reasonable planning and management, regions can find a balance between economic development and carbon emission reduction.

In summary, only under the CPS can the Chengdu–Chongqing Economic Circle reach peak carbon storage in 2030. This result emphasizes the important role of policies and management measures in controlling carbon emissions and achieving the carbon peak target. At the same time, it also suggests that in the future, in the process of promoting the sustainable development of the Chengdu–Chongqing Economic Circle, it is necessary to continue to strengthen ecological protection, optimize land use structure, and increase forest coverage, so as to further increase carbon storage and respond to the challenges brought by climate change.

The change in carbon storage in the Chengdu–Chongqing Economic Circle exhibits a strong correlation with the change in land use. Under different scenarios, the trend of carbon storage is different. Under the NDS, carbon storage may decrease due to economic activities, urbanization, and other factors, and the carbon peak has not been reached. Under the CPS, through specific policies and management measures, the carbon storage in the Chengdu–Chongqing Economic Circle is effectively controlled and reaches the peak, showing that reasonable planning and management can find a balance between economic development and carbon emission reduction.

In the future, changes in land use and carbon storage in the Chengdu–Chongqing Economic Circle will be affected by many factors. According to forecasts, by 2030, land use dynamics under different scenarios will further affect the change of carbon storage. Under the CPS, carbon storage reaches a peak, showing a positive carbon emission reduction effect. Under the CNS, although the change in carbon storage is not as drastic as that of the CPS, it still shows a certain trend of carbon emission reduction as a whole.

4. Discussion

4.1. Similarities and Differences with Existing Studies

(1)

Policy-Driven Scenario Design for Carbon Neutrality Pathways

Unlike Wu Dan et al.’s two generic scenarios (“natural development” and “ecological protection”), this research introduces the “carbon peak scenario (CPS)” and “carbon neutrality scenario (CNS)”, which operationalize China’s “dual carbon” goals with granular policy levers. For example, the CPS restricts construction land expansion by reducing forest/cultivated land conversion by 50% and promotes ecological restoration (e.g., a 20% increase in unused land-to-forest conversion), simulating policies like intensive land use and ecological red line enforcement. This scenario achieves a carbon storage peak of 5015.18 × 10⁶ t in 2030, 3.6% higher than the highest value in Wu Dan et al.’s study (2022) [8]. The CNS further integrates carbon taxation and sink protection mechanisms, reducing cultivated land-to-construction conversion by 60% and introducing carbon trading to maintain carbon storage at 4982.72 × 10⁶ t in 2060, 1.5% above the natural development scenario. These scenarios bridge the gap between macro policy goals and on-the-ground land use decisions, offering actionable roadmaps for regional planners.

(2)

Refined Modeling and Carbon Pool Disaggregation

The study employs a 15-factor driver system (e.g., GDP, traffic accessibility, and DEM) in the PLUS model, achieving a Kappa coefficient of 0.84, significantly higher than the 0.82 reported in Wu Dan et al.’s study (2022) [8]. Critically, it distinguishes four carbon pool components (aboveground biomass, belowground biomass, soil carbon, and dead organic carbon; Table 4), revealing that soil carbon alone constitutes 41% of forest land carbon storage (206.45 t/ha). This addresses a major oversight in prior studies, which often used aggregated carbon density values and overlooked soil carbon dynamics. By extending the simulation to 2060 (10 years beyond [8]) and refining our analysis to the county level, the research uncovers a “peripheral high, central low” carbon storage pattern tied to urban expansion in the Chengdu–Chongqing plain and carbon-rich mountain ecosystems (e.g., Qionglai Mountains), providing granular insights beyond macro-scale analyses.

(3)

Integration of Policy Tools and Systemic Limitations

Missing cultural dimensions: Unlike Wu Dan et al.’s (2021) [8] integration of heritage protection preferences, this study lacks data on culturally significant landscapes (e.g., Chuanxi Linpan settlements and Bayu terraces). Such factors require participatory modeling or anthropological fieldwork to quantify, which are beyond the current macro-scale scope.

Incomplete policy constraints: Compared to Wang Fang et al.’s study (2024) [32], which incorporated ecological protection red lines, this study’s dataset (up to 2020) does not reflect post-2021 policy updates (e.g., 2023 Chengdu–Chongqing land space planning). This omission may lead to a 12–18% overestimation of construction land expansion, as noted by Wang Fang et al. (2024) [32], but is necessary due to unavailable data.

4.2. Generalizability and Limitations of the Research Results

With the potential for extrapolation under the following conditions, this method may be partially extended to other regions:

Geographical analogy: Regions with “basin–mountain” terrain (such as the Guanzhong Plain and the middle reaches of the Yangtze River urban agglomeration) can adopt PLUS model parameters such as a slope limit (>25°) and forest carbon density values. Arid or karst areas need to be adjusted, such as giving priority to grassland protection or modifying vegetation conversion rules.

Policy adaptation: Carbon peak policy tools (such as a construction land cap) can be transferred to the “dual carbon” pilot areas, but they need to be adapted to local conditions. For example, water-scarce regions (such as the Yellow River Basin) may need to increase water conservation weights from 0.7 to 1.0, while forest-rich regions (such as Northeast China) can relax forest conversion restrictions but strengthen carbon sink compensation. The lack of data on the “Three Regions and Three Lines” plan (also not covered by Wu Dan et al. (2022) [8]) may introduce bias in areas with strict policies, and future calibration will be needed based on public land use planning.

Data requirements: High-resolution land use data (100 m) and socioeconomic grid data (such as the GDP and population density) are essential. In data-scarce regions (such as Southeast Asia), remote sensing inversion (such as Sentinel-2) or statistical interpolation may introduce errors of 5–8%, limiting accuracy.

4.3. Omissions of Cultural Dimensions and Policy Rigidities

Cultural landscape complexity: The impact of cultural factors on land use, such as the Linpan settlements in western Sichuan and the Bayu terraces, is highly local, but existing datasets (such as GlobeLand30) lack sufficient granularity to capture traditional agricultural systems. Quantifying their impacts requires participatory modeling or anthropological fieldwork, which is beyond the scope of this study. Future research could integrate “culturally sensitive” scenarios through interdisciplinary approaches.

Policy data timeliness: Rigid constraints such as ecological protection red lines are refined after 2021, but the dataset of this study (as of 2020) cannot reflect the 2023 Chengdu–Chongqing Twin Cities Economic Circle Land Spatial Planning (such as the Yangtze River Ecological Barrier Boundary). Wu Dan et al. (2022) [8] also omitted such constraints, indicating a common limitation in dynamic policy integration. This study prioritizes macro policy goals (such as carbon peak) rather than undetermined spatial boundaries to maintain model consistency and plans to update parameters after official data are released.

To address current limitations and enhance policy relevance, three priority areas emerge:

Short term (2025–2030): develop dynamic carbon density models using GEDI lidar (30 m resolution) and ground surveys to quantify climate–vegetation feedbacks, reducing estimation errors by 8–12%;

Medium term (2030–2035): build a policy-responsive scenario framework via Delphi-AHP methods, integrating cultural ecosystem values (e.g., heritage site protection) and updated “three zones and three lines” data to refine land use transition matrices;

Long term (2035+): create a “digital twin” system merging night light data, social media geography, and real-time policy databases to simulate abrupt policy shifts (e.g., “non-grainization” controls) and industrial transformations (e.g., new energy bases), ensuring adaptive management for carbon-neutral development.

This study did not include cultural factors (such as heritage protection preferences and the spatial distribution of industrial output value) and rigid policy constraints such as ecological protection red lines, mainly due to limited data availability: cultural-related data are highly local and lack public quantitative spatial datasets, requiring field surveys to obtain, which exceeds the research’s macro data support capabilities; policy boundaries such as ecological protection red lines have been gradually refined after the research data cutoff (2020), and the latest planning data for 2023 has not yet been fully disclosed. Forcible inclusion may lead to simulation bias, so quantifiable natural and socioeconomic driving factors are preferred, and the model will be further calibrated after the policy data are improved.

Based on the coupling framework of the PLUS and InVEST models, this study conducted a multi-scenario land use and carbon stock evolution analysis for the Chengdu–Chongqing Economic Circle. The results showed that forest land and cultivated land were the main land use types, and the expansion of construction land caused carbon storage to show an “increase first and then decrease” trend. The carbon peak (CPS) and carbon neutrality (CNS) scenarios significantly improved the carbon sequestration capacity through policy intervention. The model method provided a scientific path for regional carbon management and cross-regional low-carbon development, but the study did not incorporate cultural landscapes and the latest policy constraints. In the future, it is necessary to combine dynamic data with regional characteristics to optimize the model to improve its applicability.

5. Conclusions

Based on the PLUS-INVEST coupling model, this study simulated the dynamics in land use and carbon storage in the Chengdu–Chongqing Twin Cities Economic Circle in multiple scenarios. The main conclusions are as follows:

(1)

Model application effect: Model verification shows that the PLUS model has high accuracy (Kappa > 0.8) in regional scale simulations, and the coupling application with the InVEST model can effectively predict the impact of land use dynamics on carbon storage. This result is consistent with the complex geographical environment characteristics of the region, verifying the applicability of the model in the study of large urban agglomerations.

(2)

Land use shows significant dynamics: Forest and cultivated land constitute over 90% of the land use in the Chengdu–Chongqing Twin Cities Economic Circle. Multi-scenario projections indicate that construction land will expand, while ecological land declines under all modeled conditions. Between 2000 and 2020, the area of cultivated land and grassland will decrease, while the area of forest land, water area, construction land, and unused land will increase. The rapid expansion of construction land is the main reason for the area change, increasing by 4400 km² in 20 years, and the proportion has increased from 1.63% to 3.95%. The amount of cultivated land decreased the most, reaching 3600 km². At the same time, the land use transfer matrix from 2000 to 2020 shows that the largest area of cultivated land was transferred out, which was 12,414.13 km², and the largest area of grassland was transferred in, which was 9107.21 km². The conversion between cultivated land and forest land was the main type of land use conversion.

(3)

The spatiotemporal evolution characteristics of carbon storage are obvious: the carbon storage in the Chengdu–Chongqing Economic Circle first increased and then decreased during 2000–2020, with an overall slight increase. Future forecasts show that carbon storage under different scenarios shows a trend of first decreasing, then increasing, and then decreasing. In the CPS scenario, carbon storage reaches the maximum value, and the carbon emission reduction effect is significant; in the CNS scenario, carbon storage remains at a high level, which helps to achieve the regional carbon neutrality goal. Spatially, carbon storage shows a distribution characteristic of “high around and low in the middle”, which is exhibit a strong correlation with factors such as the topography of the Chengdu–Chongqing Twin Cities Economic Circle (mostly mountains and forests around, with a plain urban dense area in the middle), land use type (mostly forests around, with most construction land in the middle), and human activity intensity (frequent human activities in the middle city and large carbon emissions).