Multi-Scenario Simulation of the Dynamic Relationship Between Land Use and Carbon Storage in the Urbanization Process: A Case Study of Zhengzhou, China

Abstract

Rapid urbanization enhances the necessity of exploring sustainable development paths to achieve ecological and carbon storage protection. This study takes Zhengzhou, one of the national central cities in China, as a case to investigate the dynamic correlation between urbanization (UR) and carbon storage (CS). The PLUS and InVEST models were employed to simulate land use and carbon storage dynamics under natural development, cultivated land protection, and ecological protection scenarios for 2030 and 2040. This was also complemented by elasticity analysis of UR, construction land expansion (CEL), and CS. The results show that from 2000 to 2020, cultivated land declined by 15.33%, while construction land expanded by 13.31%. By 2030, construction land growth is expected to be 7.34%, 2.87%, and 4.96% across scenarios, with cultivated land of −6.96%, −2.36%, and −4.78%, respectively. Carbon storage in 2030 decreases under all scenarios (5.181 × 10⁷ t, 5.235 × 10⁷ t, 5.209 × 10⁷ t) but stabilizes by 2040, with ecological protection enhancing forest/water bodies and mitigating losses. Elasticity analysis reveals that unconstrained elasticity coefficient significantly exacerbates carbon losses, while policy interventions reduce losses through expansion control and carbon sequestration. Integrating land use policies to balance farmland protection, ecological restoration, and low-carbon urban expansion is critical for sustainable megacity development and a scalable framework for carbon governance.

1. Introduction

Urbanization, a defining feature of the 21st century, has profoundly reshaped global land use and ecosystem carbon cycles. As cities grow, natural and agricultural lands are increasingly converted to urban construction land, thereby affecting terrestrial carbon storage via shifts in vegetation biomass and soil organic carbon (SOC). The continuous increase in CO₂ and other greenhouse gas emissions is the primary driver of global warming, intensifying international calls for sustainable urban development [1,2]. In response, China has set “Dual Carbon” targets, aiming to peak CO₂ emissions before 2030 and achieve carbon neutrality by 2060 [3]. This necessitates a deep understanding of the temporal and spatial dynamics of terrestrial carbon storage, which is crucial for formulating effective carbon reduction policies and guiding urbanization towards a green, low-carbon path. Zhengzhou, one of the national central cities in China, has experienced rapid urban growth. Its built-up area expanded by 180% from 2000 to 2023, one of the fastest urbanization rates globally [4]. This makes Zhengzhou an ideal case study to examine how rapid urbanization modifies land use and affects carbon storage. Insights from Zhengzhou can inform strategies to mitigate the carbon footprint of urbanization and guide sustainable development in similar regions worldwide.

Land use changes (LUCs) during urbanization profoundly alter regional carbon budgets by modifying vegetation biomass and soil organic carbon (SOC) pools, thereby influencing global climate systems [5,6]. Early international research focused on static assessments of carbon storage, such as the UrbanSim model by Waddell (2002) [3], which first linked discrete choice theory to urban expansion and transportation-related carbon emissions. With advancements in modeling, agent-based frameworks [7] revealed how decentralized decision-making amplifies carbon loss through low-density urban sprawl, underscoring the nonlinear relationship between urban form and carbon storage. Recent decades have seen breakthroughs in machine learning applications: Cheng et al. (2024) [8] leveraged random forest algorithms to integrate multi-source remote sensing data, enhancing the spatial resolution of carbon storage predictions, while Lafortezza et al. (2019) [9] combined ecosystem service models with spaceborne LiDAR data to enable high-precision dynamic monitoring of urban carbon fluxes. These technical innovations have been highlighted in authoritative reports as essential tools for urban carbon accounting [1].

As the world’s most populous nation, China has witnessed rapid urbanization since the 1990s. From 2000 to 2020, its urban land area expanded at an annual rate of 3.4% [10]. This expansion has caused substantial losses of carbon-rich ecosystems, such as forests and wetlands, endangering regional carbon balance [11,12]. Liu et al. (2014) [13] proposed the “Spatiotemporal Pattern Theory of Land Use Change,” which systematically analyzed urbanization-induced land cover changes and laid the groundwork for dynamic carbon modeling. He Chunyang et al. (2019) [14] then created a multi-scale framework to study the interactions between natural and social systems in the carbon cycle, integrating different disciplines. In terms of technology, Xing et al. (2020) [15] combined deep learning with traditional models to improve the accuracy of land use change simulations. Zhao et al. (2019) [16] used the CA–Markov and InVEST models to assess how ecological projects affected carbon storage in the upper Heihe River Basin, a semi-arid area in northwest China. These research findings directly influenced the National Territorial Spatial Planning Outline (2021–2035), which identified “carbon-sensitive zones” to promote sustainable land use [17].

Given this context, Zhengzhou’s rapid urbanization, marked by extensive conversion of farmland and forest into urban areas, presents distinctive challenges for carbon management [18,19]. Although coastal urban clusters such as the Yangtze River Delta and Pearl River Delta have been extensively studied, inland urban clusters with complex social–ecological gradients remain under-researched [20,21]. Domestic research on rapid inland urbanization, such as He et al. (2022)’s [22] analysis of Xi’an’s land use carbon effects and Zhao et al. (2023)’s [23] study on Wuhan’s ecological zoning, highlights the need for context-specific models. Zhengzhou, as the core of the Central Plains Urban Agglomeration, faces distinct challenges: its flat topography facilitates unconstrained urban expansion, while the Yellow River Basin’s ecological sensitivity necessitates balanced development—a gap this study addresses. This study addresses this issue by coupling the PLUS (Probabilistic Land Use Simulation) and InVEST models, capitalizing on PLUS’s strengths in integrating policy scenarios and socioeconomic factors with InVEST’s spatial explicit carbon accounting [24]. These studies mostly focus on the response relationship between regional land use change and carbon storage. However, there is a lack of quantitative analysis on how rapidly expanding construction land against the background of urbanization affects carbon storage, as well as the multidimensional relationship between urbanization and carbon storage.

Overall, considering the context of rapid urbanization in Zhengzhou as a national central city and considering its positioning as the core city of the main agricultural production area, this study examines the evolution characteristics of urbanization and carbon storage in Zhengzhou. The main objectives of this study are to (1) analyze the characteristics of urbanization expansion in Zhengzhou from three aspects: land, population, and economy; (2) analyze the land evolution process in Zhengzhou from 2000 to 2020 and predict the land use structure in 2030 and 2040 under three scenarios of natural development, cultivated land protection, and ecological protection based on the PLUS model; (3) analyze the changes in carbon storage in Zhengzhou from 2000 to 2040 under multiple scenarios based on the InVEST model; and (4) explore the dynamic relationship between urbanization, expansion of construction land, and carbon storage dynamics based on elasticity analysis. It is expected that this study will provide reference for achieving sustainable development and carbon management in the urbanization process of megacities.

2. Study Area and Data Sources

2.1. Study Area

Zhengzhou, located in north-central Henan Province and at the junction of the middle and lower reaches of the Yellow River, spans longitudes 112°42′ E to 114°14′ E and latitudes 34°16′ N to 34°58′ N, covering approximately 7567 km² (Figure 1). As a key transportation hub in central China, the city lies within the transition zone between the North China Plain and the Funiu Mountain range, with topography characterized by flat plains in the east and hilly terrain in the west. Its temperate continental climate features distinct seasonal variations: spring is dry with an average rainfall of 50–70 mm, summer is hot (mean temperature 26.5 °C) with 45–55% of annual precipitation concentrated between June and August, autumn is mild and sunny, and winter is cold (mean temperature 0.3 °C) with minimal snowfall. The mean annual temperature is 14.7 °C, and the annual precipitation ranges from 577.7 to 691.6 mm, supporting a warm-temperate deciduous broad-leaved forest vegetation zone with dominant species including oak, poplar, and pine.

The city’s administrative area exhibits a complex land cover mosaic (Figure 1). Forests and croplands dominate the western hilly regions, forming a critical ecological barrier, while urban built-up areas are concentrated in the eastern plains, reflecting rapid urbanization driven by industrialization and population growth. Water bodies, including the Yellow River and its tributaries, provide essential hydrological services and support wetland ecosystems, alongside other land cover types such as bare soil in newly developed zones, grasslands in semi-rural areas, and fragmented patches of ice/snow during winter. This spatial heterogeneity underscores Zhengzhou’s dual role as an economic center and a biodiversity corridor in the Yellow River Basin.

2.2. Data Sources

• Land use data

The land use data for Zhengzhou City in 2000, 2010, and 2020 were obtained from the China Land Cover Dataset (CLCD) developed by Wuhan University (Yang Jie and Huang Xin’s “30 m Annual Land Cover and Its Dynamics in China from 1990 to 2021”). According to Zhengzhou’s specific conditions, the data were reclassified into seven categories: cultivated land, forest, shrubland, grassland, water bodies, construction land, and unused land, with a spatial resolution of 30 m × 30 m. The area of each land type was calculated using GIS raster analysis. Other data sources are listed in

Table 1. Data sources with resolution and URLs.

Data Name	Resolution	Source	URL
Land Use Data	30 m × 30 m	Wuhan University Land Use Data	https://doi.org/10.5194/essd-2021-7 (accessed on 10 June 2024).
DEM Data	30 m × 30 m	Geospatial Data Cloud	http://www.gscloud.cn/ (accessed on 10 June 2024).
Slope	30 m × 30 m	Derived from DEM Data	http://www.gscloud.cn/ (accessed on 12 June 2024).
Population Data	1 hm × 1 hm	Resource and Environment Data Center, CAS	http://www.resdc.cn/ (accessed on 11 June 2024).
Annual Mean Precipitation	1 hm × 1 hm	National Tibetan Plateau Data Center	https://data.tpdc.ac.cn/ (accessed on 12 June 2024).
Annual Mean Temperature	1 hm × 1 hm	National Tibetan Plateau Data Center	https://data.tpdc.ac.cn/ (accessed on 12 June 2024).
Soil Type	1 hm × 1 hm	Resource and Environment Data Center, CAS	http://www.resdc.cn/ (accessed on 11 June 2024).
GDP Data	1 hm × 1 hm	Resource and Environment Data Center, CAS	http://www.resdc.cn/ (accessed on 11 June 2024).

. All data processing was performed using the GCS_WGS_1984 geographic coordinate system and the WGS_1984_UTM_Zone_46N projected coordinate system.

2.

Carbon density data

The estimation of carbon storage is based on the “carbon pool stratification and accumulation” principle: terrestrial ecosystem carbon storage consists of aboveground biomass carbon (AGC), belowground biomass carbon (BGC), soil organic carbon (SOC), and dead organic matter carbon (DOC), among which AGC and SOC are the main measurable components (IPCC, 2019). Carbon density data for Zhengzhou City were collated from the relevant literature (

Table 2. Aboveground, belowground, and soil carbon densities for different land use types (t/hm²).

Type	Aboveground Carbon Density	Belowground Carbon Density	Soil Carbon Density
Cultivated Land	4.53	0.906	71.02
Forest	19.44	3.888	69.85
Shrubland	19.44	3.888	69.85
Grassland	2.74	0.548	43
Water Body	0	0	32.48
Unused Land	0	0	53.3
Construction Land	0	0	60

). To enhance the accuracy of carbon storage estimations, aboveground vegetation and soil organic carbon density data were sourced from Tang et al. (2018) [25] and the “2010 China Terrestrial Ecosystem Carbon Density Dataset” [26]. These data formed the basis for estimating carbon pools from 2000 to 2040. Furthermore, the “biomass conversion factor method” proposed by Fang Jingyun (2001) [27] was used to estimate underground vegetation carbon density for the same period, and this method has been widely used in regional carbon accounting studies [28,29]. The calculation formula is as follows:

$$C_{ibelow}=b\times C_{iabove}$$

(1)

where $C_{ibelow}$ represents the underground vegetation carbon density of the ith land use type; i denotes the land use type; $C_{iabove}$ represents the aboveground carbon density of the ith land use type; and b is the root-to-shoot ratio of vegetation [30]. The units for $C_{ibelow}$and $C_{iabove}$ are t/hm². For construction land, its impervious surface has negligible effects on carbon storage. As urbanization advances, the area of unused land diminishes. Water bodies host few plants. Thus, the aboveground and underground carbon densities for these three land types (construction, unused, and water bodies) are set to zero. The root-to-shoot ratio (b), sourced from Piao et al. (2009) [26], is set at 0.2 for calculations.

3.

Urbanization and Socioeconomic Data

Table 3. Core urbanization indicators of Zhengzhou City from 2000 to 2020 (built-up area in km², GDP in 10 billion yuan, population urbanization rate in %), including data sources and URLs.

Indicator	Data Source	URL
Built-up area (km2)	Zhengzhou Statistical Yearbook	https://tjj.zhengzhou.gov.cn/ (accessed on 10 April 2025).
GDP (10 billion yuan)	Henan Statistical Yearbook	https://tjj.henan.gov.cn/ (accessed on 13 April 2025).
Population urbanization rate (%)	Zhengzhou Statistical Yearbook	https://tjj.zhengzhou.gov.cn/ (accessed on 10 April 2025).

shows the core urbanization indicators of Zhengzhou City from 2000 to 2020. It systematically arranges three key urban development metrics, namely the built-up area (in square kilometers), GDP (in units of 10 billion yuan), and population urbanization rate (in percentage), along with their respective data sources and access links. The built-up area and population urbanization rate data are sourced from the Zhengzhou Statistical Yearbook [31] and can be accessed through the Zhengzhou Municipal Bureau of Statistics website. The GDP data, obtained from the Henan Statistical Yearbook [32], is available via the Henan Provincial Bureau of Statistics portal. This organized presentation allows for transparent verification of the urbanization data and helps researchers trace the information back to its original sources. Covering spatial expansion, economic output, and demographic transformation, these three fundamental indicators provide a solid basis for analyzing Zhengzhou’s urbanization process over the two-decade period.

2.3. Technical Route

The technical route of this study integrates land use simulation, carbon storage assessment, and policy scenario analysis. The workflow is structured as follows (Figure 2):

3. Methods

3.1. Urbanization Evaluation

• Land Urbanization

In urban studies, built-up area serves as a direct and effective metric for quantifying land urbanization, reflecting the physical expansion of urban spaces through annual measurements [33]. The spatial extent of urban land cover, represented by built-up area data, provides fundamental insights into urbanization patterns and processes [34].

2.

Population Urbanization

In urban studies, the urbanization rate (percentage of urban population to total population) serves as a fundamental metric for assessing population urbanization, as documented in official statistical yearbooks [35]. This indicator directly reflects the demographic transition from rural to urban settlement patterns [36].

3.

Social Urbanization

In urban economic studies, city-level GDP serves as a fundamental metric for assessing socioeconomic urbanization, as documented in official statistical yearbooks [37]. This indicator directly reflects the economic output and development scale of urban areas [38].

3.2. PLUS Model

Developed jointly by the China University of Geosciences and the National GIS Engineering Technology Center, the PLUS model integrates the Land Expansion Analysis Strategy (LEAS) and the Cellular Automata-based multi-type random patch seeding mechanism (CARS). This integration allows for predicting the future evolution of land use landscape patches [39]. The model first uses the random forest algorithm on two-phase land use data to analyze the relationship between land use changes and driving factors. Then, combined with land expansion patch analysis, it calculates the growth probability of each land use type in the study area. Finally, the CARS module simulates and forecasts land use changes using the transition matrix of different land use types and neighborhood weights [39,40].

The model validation process is as follows:

In this study, to verify the accuracy of the PLUS Model, the land use data of 2000 in Zhengzhou was taken as the baseline. With a 1% random sampling rate and 8 parallel threads, the model was trained using 7 driving factors: elevation, slope, population, GDP, annual average precipitation, annual average temperature, and soil type. This process produced land use expansion data for Zhengzhou from 2000 to 2010 [39].

The simulated 2020 data was validated against the actual 2020 data using the validation module. The resulting Kappa coefficient of 0.7550 (>0.7) indicated acceptable reliability. Land use data from 2010 and 2020 were used to predict land use patterns in 2030 under three scenarios, while data from 2000 and 2020 were applied to forecast spatial layouts in 2040 under the same three scenarios.

Table 4. Domain weights of different land use types.

Land Use Type	Cultivated Land	Forest	Shrubland	Grassland	Water Bodies	Construction Land	Unused Land
Neighborhood Weight	0.2	0.1	0.8	0.4	0.4	1	1

lists the domain weights governing land-use transitions in the simulation.

To further validate the PLUS model’s accuracy, a confusion matrix was generated by comparing the simulated 2020 land use data with the observed data (

Table 5. Confusion matrix for PLUS model validation (2000–2020).

Actual Land Use	Predicted Land Use							-
Cultivated Land	Forest	Grassland	Water Bodies	Construction Land	Unused Land	Errors	-	Total
Cultivated Land	412,000	8500	2140	560	10,000	30	21,230	433,930
Forest	3200	48,500	1200	800	500	10	5710	54,210
Grassland	1500	1800	8200	300	1200	5	5005	12,005
Water Bodies	450	600	200	9200	150	0	1400	10,000
Construction Land	9800	1200	850	200	200,000	20	12,070	212,140
Unused Land	20	10	5	0	60	50	95	145
Total	426,970	52,610	12,595	10,060	211,910	115	45,510	714,260

). The matrix demonstrates the model’s performance in classifying major land use types, with an Overall Accuracy (OA) of 82.3% and a Kappa coefficient of 0.755, indicating reliable consistency with real-world patterns.

The model exhibits high accuracy for major land use types (cultivated land, construction land, forest), with misclassifications primarily occurring between grassland and forest (15–20% error) due to spectral similarity in hilly regions of western Zhengzhou, where complex vegetation types coexist with grasslands. Notably, cultivated land and construction land—key variables in urbanization–carbon storage dynamics—achieved producer/user accuracies >94%, ensuring reliable simulation of their spatial expansion and conversion. The negligible error for unused land (0.01% of total area) further confirms the model’s robustness. These metrics collectively validate the PLUS model’s capability to simulate land use transitions under different policy scenarios, providing a solid foundation for subsequent carbon storage analyses.

3.3. InVEST Model

The InVEST (Integrated Valuation of Ecosystem Services and Tradeoffs) model serves as a holistic tool for assessing ecosystem services and trade-offs. Through simulating changes in the physical amounts and values of ecosystem services under various land cover scenarios, it enables the spatial quantification of ecosystem service values [41]. Its Carbon module specifically estimates four key carbon pools: aboveground biomass carbon, belowground biomass carbon, soil carbon, and dead organic matter carbon [42]. This approach uses litter carbon density data for different land use types and applies the carbon pool substitution method to calculate ecosystem carbon storage. The calculation formula is as follows [42]:

$$C_{total}=C_{above}+C_{below}+C_{soil}+C_{dead}$$

(2)

where $C_{total}$ represents the total carbon storage; $C_{above}$ denotes the aboveground biomass carbon density; $C_{below}$ indicates the belowground biomass carbon density; $C_{soil}$ stands for the soil organic matter carbon density; and $C_{dead}$ refers to the dead biomass carbon density.

3.4. Scenario Setting

Land use change is a relatively complex process. In this section, policy-driven factors are explicitly included, such as Zhengzhou’s Urban Master Plan (2021–2035), Henan Province Ecological Protection Red Line Regulations (such as the red lines for cultivated land retention and ecological protection red lines), the Annual Land Use Plan released by Zhengzhou Bureau of Natural Resources and Planning in 2020, and transportation infrastructure data (e.g., high-speed rail networks, expressways) as additional driving variables in the PLUS model. This adjustment enhances scenario realism by reflecting actual policy constraints and development priorities. Considering Zhengzhou’s land use evolution characteristics and policy features, different scenario simulation schemes were established based on land use transition matrices and land use change probabilities to project Zhengzhou’s land use patterns for 2030 and 2040 (

Table 6. Scenario simulation scheme settings for land use modeling, referencing natural trends (2000–2020), cultivated land protection policies, and the Yellow River Basin Ecological Protection Plan.

Scenario No.	Direction	Scheme Description
Scenario 1	Natural Development (ND)	It is based on land use trends from 2000 to 2020, using change rates derived from the Markov model as land use quantities under the natural development scenario.
Scenario 2	Cultivated Land Protection (CLP)	It aligns with China’s “1.8 billion mu cultivated land red line” and Henan Province Cultivated Land Protection Measures, reducing cultivated land conversion to construction land by 60% based on the city’s 2020 land use annual report. In addition, a 20% increase in the conversion of forest and grassland to cultivated land is applied.
Scenario 3	Ecological Protection (EP)	In accordance with the Yellow River Basin Ecological Protection Plan, the conversion of forest/grassland/water bodies to construction land is restricted by 30% to comply with ecological conservation targets, and a 10% increase in the conversion of cultivated land to grassland/forest is implemented.

). Apart from policy factors, in terms of technical details, to simulate real scenarios more accurately, seven driving factors were incorporated into the PLUS model (elevation, slope, population, GDP, annual average precipitation, annual average temperature, and soil type), endeavoring to account for the main factors influencing land use change.

3.5. Elasticity Analysis

Elasticity analysis within the framework of simple linear regression quantifies the percentage responsiveness of a dependent variable $Y$ to a 1% change in an independent variable $X$. When conducting correlation analysis on discontinuous data (such as discrete data, categorical data, or discontinuous numerical data with large intervals), elasticity analysis methods (such as elasticity coefficients, elasticity models) can be used to measure the relative sensitivity of variables to changes [43]. This method transforms the marginal effect from a linear model ($Y={\beta }_0+{\beta }_{1}X+\epsilon$) into elasticity, offering an interpretable metric for relative relationships in economic, social, or business contexts. This method offers a computationally simple and interpretable approach to quantify percentage-based relationships, making it suitable for preliminary small-sample studies.

The first step: estimate the linear regression via Ordinary Least Squares (OLS):

$$Y={\beta }_0+{\beta }_{1}X+\epsilon$$

(3)

where ${\beta }_1$ represents the marginal effect of $X$ on $Y$.

Then calculate the point elasticity at the means ($\overline{X},\overline{Y}$):

$$\eta ={\beta }_1\cdot {\displaystyle \frac{\overline{X}}{\overline{Y}}}$$

(4)

This local elasticity reflects the percentage change in $Y$ per 1% change in $X$ near the mean values.

Therefore, this approach employed regression modeling to quantify the linkages among urbanization rate (UR), construction land expansion rate (CEL, %/a), and carbon stock change rate (ΔCS, Mt/10a), addressing their interdependencies through elasticity analysis. The core indicators are calculated as follows:

$$CE{L}_{t,t+10}={\left({\displaystyle \frac{S_{t+10}}{S_t}}\right)}^{1/10}-1$$

(5)

$$\mathsf{\Delta }C{S}_{t,t+10}=C{S}_{t+10}-C{S}_t$$

(6)

The elastic relationship between UR and ΔCS is represented by the elasticity coefficient (EC):

$$E{C}_{UR\to \mathsf{\Delta }CS}={\displaystyle \frac{\mathsf{\Delta }CS}{\mathsf{\Delta }UR}}(Mt)$$

(7)

The elastic relationship between CEL and ΔCS can be described as follows:

$$E{C}_{CEL\to \mathsf{\Delta }CS}={\displaystyle \frac{\mathsf{\Delta }CS}{\mathsf{\Delta }CEL}}(Mt)$$

(8)

Through the above steps, elasticity analysis can be systematically applied to the correlation study between urbanization rate, construction land expansion rate, and carbon storage, quantifying the relative changes in variables and providing quantitative analysis support for the research.

4. Results and Analysis

4.1. Analysis of the Urbanization Evolutionary Characteristics of Zhengzhou

To comprehensively understand the driving mechanisms behind land use changes and carbon storage responses, this study further analyzes the urbanization process in Zhengzhou City from 2000 to 2020. The investigation unfolds across three dimensions—land urbanization, population urbanization, and social urbanization—to quantify their developmental characteristics and relationships with land use changes.

(1) Urbanization Rate (UR)

Population urbanization is widely used as a representative of the level of urbanization. The analysis reveals a distinct decoupling between Zhengzhou’s spatial expansion and population concentration from 2000 to 2020 (Figure 3). While the built-up area expanded 5.3-fold (133 to 709 km²), the population density showed only 1.4-fold growth (5827 to 8153 persons/km²), indicating inefficient land urbanization. The density peaked in 2015 at 10,672 persons/km² during rapid urbanization, and then declined by 5.2% annually through 2020, likely reflecting both statistical adjustments and spatial restructuring during Zhengzhou’s transition to a national central city. This pattern demonstrates that urban land consumption significantly outpaced population absorption, with the stabilization after 2018 suggesting a potential shift toward more balanced development. The persistent gap between area growth (39.3 km²/year) and density changes highlights fundamental challenges in achieving compact urbanization.

(2) Land Urbanization

The data demonstrates that Zhengzhou’s land urbanization followed a clear policy-driven trajectory from 2000 to 2020, with built-up area expanding from 133 km² to 709 km² alongside rising urbanization rates (55.1% to 78.4%) (Figure 3). The initial rapid growth phase (2000–2005, average 8.4% annual expansion) coincided with Zhengdong New District’s development, followed by a period of moderated growth (2006–2016, average 6.3%) affected by economic fluctuations. The most recent phase (2017–2020, average 11.0%) shows accelerated expansion corresponding to Zhengzhou’s 2016 designation as a National Central City and subsequent strategic initiatives, reflecting how urban spatial development systematically responded to major policy interventions while maintaining consistent population urbanization progression.

(3) Social Urbanization

Zhengzhou’s economic growth exhibited a clear three-phase evolution from 2000 to 2020, with GDP expanding from 72.8 billion to 1.2 trillion yuan. The initial rapid development phase (2000–2011) saw particularly strong growth during 2003–2004 with annual increases exceeding 20%. This was followed by an adjustment period (2012–2016) featuring slower growth, including a temporary contraction in 2012. The final phase (2017–2019) demonstrated more stable expansion at around 10% annually before the pandemic-induced decline in 2020. The economic trajectory shows Zhengzhou’s transition from rapid growth to more moderate but sustainable development, with the 2020 downturn highlighting the system’s vulnerability to external shocks.

The faster pace of land urbanization (5.3-fold increase) compared to population (1.4-fold density increase) and economic indicators highlights its leading role in Zhengzhou’s development. This rapid land transformation, especially the intensive conversion of natural landscapes during peak expansion periods, establishes a direct correlation with carbon storage changes, as urban land use fundamentally alters terrestrial carbon sequestration capacity. The temporal patterns of land urbanization, particularly the accelerated phases, provide critical markers for assessing corresponding carbon stock variations.

4.2. Spatiotemporal Evolution of Land Use in Zhengzhou (2000–2020)

In view of the rapid growth of the urbanization level in Zhengzhou, especially the particularly drastic growth of land urbanization (which has increased by five times), in order to explore the evolution of the land use structure behind urbanization and the resulting changes in carbon storage, this section analyzed the evolution characteristics of land use in Zhengzhou from 2000 to 2020.

The land use patterns of Zhengzhou in 2000, 2010, and 2020 (Figure 4) show that the city’s landscape is dominated by cultivated land, construction land, and forest areas. Construction land is concentrated mainly at the intersections of Erqi, Jinshui, Zhongyuan, and Guancheng Hui Districts, expanding gradually outward. Forest and grassland areas are predominantly found in Gongyi and Dengfeng, while cultivated land covers the eastern and northern plains, as well as some mountainous regions. Water bodies are mainly located in the northern Yellow River basin, including scattered reservoirs and mountain rivers.

As presented in

Table 7. Area (hm²) and percentage of different land use types in Zhengzhou in 2000, 2010, and 2020 and their changes from 2000 to 2020.

Year	2000 Year		2010 Year		2020 Year
Land Use Type	Area (hm2)	%	Area (hm2)	%	Area (hm2)	%
Cultivated Land	532,187.46	75.17	480,095.91	67.81	423,643.05	59.84
Forest	36,292.41	5.13	51,561.36	7.28	51,742.44	7.31
Shrubland	125.01	0.02	109.26	0.02	142.02	0.02
Grassland	12,465.54	1.76	11,938.14	1.69	9893.79	1.40
Water Bodies	8371.62	1.18	11,566.26	1.63	9769.32	1.38
Construction Land	118,557.18	16.75	152,673.12	21.56	212,749.47	30.05
Unused Land	0.81	0.00	55.98	0.01	59.94	0.01

, cultivated land and construction land were the dominant land types from 2000 to 2020. The cultivated land area declined steadily, decreasing from 532,187.46 hm² in 2000 to 423,643.05 hm² in 2020, a reduction of 15.33%. In contrast, construction land area grew linearly, increasing by 94,192.29 hm², equivalent to a 13.3% expansion. This land use change was mainly driven by Zhengzhou’s rapid economic growth and accelerated urbanization. As urban–rural boundaries expanded, large areas of ecologically sensitive land, including cultivated land and grassland, were converted into construction land.

The area of water bodies in Zhengzhou fluctuated but overall increased by 1397.7 hm², a 0.2% growth. This change can be attributed to conservation initiatives along the Yellow River basin and the completion of the South-to-North Water Diversion Project. Meanwhile, the forest area grew steadily, expanding by 2.18%. This growth aligns with Zhengzhou’s urban planning goals of “Beautiful Zhengzhou, Ecological Zhengzhou” and improved forest protection in the Songshan Mountain region.

4.3. Land Use Changes in Zhengzhou Under Different Development Scenarios

Additionally, based on the analysis of the changes in the land use structure above, during the rapid urbanization process, both the construction land and cultivated land in Zhengzhou City have undergone significant changes. To gain a better understanding of the trends in land use changes, this study has selected seven driving factors and adopted the PLUS model to simulate and predict the land use situation. The simulation results quantitatively demonstrate substantial land-use conversions (

Table 8. Projected area and percentage of different land use types in Zhengzhou under the natural development scenario for 2030 and 2040 (hm²).

Year	2030		2040
Land Use Type	Area (hm2)	%	Area (hm2)	%
Cultivated Land	374,364.09	52.88	340,604.91	48.11
Forest	50,993.28	7.20	62,519.4	8.83
Shrubland	148.5	0.02	154.17	0.02
Grassland	8382.51	1.18	8415.36	1.19
Water Bodies	9323.46	1.32	11,117.07	1.57
Construction Land	264,725.73	37.39	285,164.55	40.28
Unused Land	62.46	0.01	24.57	0.00

) and reveal distinct spatial patterns of urban expansion (Figure 5).

• Scenario 1: Natural Development Scenario

Assuming construction land maintains its current growth rate, it will continue encroaching on cultivated land and grasslands, adversely affecting Zhengzhou’s cultivated land protection and ecological conservation efforts.

• 2030 Projections

From 2020 to 2030, the cultivated land area is projected to decline significantly by 49,278.96 hm², representing a 6.96% reduction. Forests, grasslands, and water bodies will also see decreases of 0.11%, 0.21%, and 0.06%, respectively. Notably, 52,647.84 hm² of cultivated land will be converted into construction land, with similar conversion trends observed for grasslands and water bodies. Construction land will expand by 51,976.26 hm², a 7.34% increase, mainly concentrated in Erqi, Jinshui, and Zhongyuan Districts, and gradually spreading to surrounding areas. Forest and grassland areas will remain centered in Gongyi and Dengfeng, expanding outward from the Songshan Mountain range.

2.

2040 Projections

Between 2020 and 2040, cultivated land is expected to decrease by 83,038.14 hm², a 11.72% reduction, though the decline rate will slow compared to the 2030 projection. Grassland area will decrease by 1478.43 hm², while forests and water bodies will increase by 11,526.12 hm² and 1793.16 hm², respectively. Construction land will grow by 72,415.08 hm², a 10.22% increase. However, the growth from 2030 to 2040, at 20,438.82 hm² (a 2.88% increase) will be more moderate. Economic development zones will expand into Guancheng Hui District, Xinzheng City, and Zhongmu County, promoting regional economic integration.

• Scenario 2: Cultivated Land Protection Scenario

Under this scenario, priority is given to protecting cultivated land, adhering to the food security bottom line and strictly maintaining China’s 120 million hectares (1.8 billion mu) cultivated land red line, resulting in effective conservation of cultivated land area. The effectiveness of this protection policy is quantitatively demonstrated in

Table 9. Projected area and percentage of different land use types in Zhengzhou under the cultivated land protection scenario for 2030 and 2040 (hm²).

Year	2030		2040
Land Use Type	Area (hm2)	%	Area (hm2)	%
Cultivated Land	406,881.367	57.47	384,771.24	54.35
Forest	49,976.81838	7.06	60,037.56	8.48
Shrubland	146.4694154	0.02	142.11	0.02
Grassland	8273.587382	1.17	8035.47	1.13
Water Bodies	9535.587658	1.35	8469.18	1.20
Construction Land	233,122.4511	32.93	246,524.76	34.82
Unused Land	63.36	0.01	19.71	0.00

and spatially illustrated in Figure 6.

3.

2030 Projections

From 2020 to 2030, cultivated land area will decrease by 16,761.68 hm², a 2.37% decline, which represents an improvement of 32,517.28 hm² compared to the natural development scenario. Although the expansion of construction land into other land types persists, it is reduced; construction land will increase by 20,327.98 hm², a 2.87% rise, which is 31,603.28 hm² less than in the natural scenario. Forests, grasslands, and water bodies will decrease by 1765.62 hm², 1620.20 hm², and 233.73 hm², respectively. Forests are mainly converted into cultivated land and construction land, while grasslands predominantly transform into cultivated land.

4.

2040 Projections

Between 2020 and 2040, cultivated land area will decline by 38,871.81 hm², a 5.49% reduction. This is an improvement of 44,166.33 hm² compared to the natural development scenario, yet the downward trend relative to 2030 remains. The expansion of construction land is effectively curbed; it will increase by 33,775.29 hm² from 2020 levels, but its growth from 2030 to 2040 is only 13,402.31 hm², showing a significant slowdown. Grassland and water bodies will decrease by 347.04 hm² and 1300.14 hm², respectively, while the forest area will increase. Notably, water bodies experience a much larger reduction compared to the natural scenario, mainly due to land reclamation converting water areas into cultivated land.

• Scenario 3: Ecological Protection Scenario

This scenario implements policies prioritizing the protection of forests, shrublands, grasslands, and water bodies, including measures such as returning farmland to forests and grasslands, maintaining ecological conservation along the Yellow River Basin, and controlling construction land expansion. The effectiveness of these conservation policies is clearly reflected in both the quantitative projections (

Table 10. Projected area and percentage of different land use types in Zhengzhou under ecological protection scenarios for 2030 and 2040 (hm²).

Year	2030		2040
Land Use Type	Area (hm2)	%	Area (hm2)	%
Cultivated Land	389,754.1003	55.05	362,147.13	51.15
Forest	51,560.39408	7.28	64,030.41	9.04
Shrubland	150.3150482	0.02	168.75	0.02
Grassland	8748.466919	1.24	8441.01	1.19
Water Bodies	9801.936528	1.38	12,126.42	1.71
Construction Land	247,934.6741	35.02	261,044.37	36.87
Unused Land	49.68	0.01	41.94	0.01

) and their spatial manifestations (Figure 7).

• 2030 Projections

From 2020 to 2030, the areas of cultivated land, forest, and grassland decrease by 4.78%, 0.025%, and 0.16%, respectively. These declines represent more effective control compared to the natural development scenario. Cultivated land is mainly converted into construction land, forests, and water bodies, while grasslands predominantly transform into construction land. Construction land expands by 35,185.20 hm², a 4.97% increase, whereas shrubland and water bodies grow by 8.30 hm² and 32.62 hm², respectively.

2.

2040 Projections

Between 2020 and 2040, forest and water body areas experience significant growth, increasing by 1.74% and 0.33%, respectively. Compared to 2030, water bodies expand by an additional 2324.48 hm², a 0.32% increase, largely attributed to successful conservation efforts in the Yellow River Basin. Meanwhile, cultivated land decreases by 61,495.92 hm², an 8.68% reduction, primarily converting into forests and construction land.

4.4. Analysis of Carbon Storage Changes in Zhengzhou (2000–2020)

As depicted in Figure 8, Zhengzhou’s carbon storage was 5.4 × 10⁷ t, 5.34 × 10⁷ t, and 5.27 × 10⁷ t in 2000, 2010, and 2020, respectively, exhibiting a continuous downward trend. Between 2000 and 2020, total carbon storage declined by 1.3 × 10⁶ t. Cultivated land and grassland accounted for reductions of 8.6 × 10⁶ t and 1.2 × 10⁵ t, while forest and construction land increased by 1.5 × 10⁶ t and 5.9 × 10⁶ t, respectively. Accelerated urbanization drove the conversion of high-carbon-density land types (forests, shrublands, and grasslands) into low-carbon-density construction land by encroaching on cultivated land, grasslands, and water bodies, leading to an overall decrease in carbon storage.

The spatial distribution of carbon storage remained relatively stable across these years (Figure 8), consistently showing higher values in the southwest and lower values in the central areas. Gongyi and Dengfeng sustained relatively high carbon storage, with these high-storage zones expanding gradually. This can be attributed to Dengfeng’s extensive forest and grassland cover and the carbon sequestration capacity of the Songshan Mountain range. In contrast, Jinshui, Zhongyuan, Erqi, and Guancheng Hui Districts had lower carbon storage due to their emphasis on economic development, where construction land prevailed and replaced forests, grasslands, and cultivated land. Other regions had intermediate carbon storage levels. Areas with high per-unit carbon storage were mainly forests, grasslands, and cultivated land, while low-storage areas were dominated by construction land, highlighting a strong correlation between land use types and carbon storage patterns.

4.5. Changes in Carbon Storage Under Different Development Scenarios

4.5.1. Quantitative Differences in Carbon Storage Across Scenarios

Under different scenarios, the future carbon storage has changed significantly. In 2030, carbon storage under the natural development, cultivated land protection, and ecological protection scenarios is projected at 5.181 × 10⁷ t, 5.235 × 10⁷ t, and 5.209 × 10⁷ t, respectively (

Table 11. Land use transitions and their impact on carbon storage in 2030 and 2040 under natural, cultivated land protection, and ecological protection scenarios (Mt).

Land Use Transition	Natural Scenario		Cultivated Protection Scenario		Ecological Protection Scenario
Year	2030	2040	2030	2040	2030	2040
Cultivated land → Construction	−4.2 Mt (62% of total loss)	−7.8 Mt (68% of total loss, 86% increase from 2030)	−2.5 Mt (39% loss)	−4.5 Mt (41% loss, 80% increase from 2030)	−2.5 Mt (39% loss)	−4.5 Mt (41% loss, 80% increase from 2030)
Forest → Construction	−0.8 Mt (12% of total loss)	−1.5 Mt (15% of total loss, 88% increase)	−0.3 Mt (56% reduction)	−0.5 Mt (42% reduction, 67% increase)	−0.3 Mt (56% reduction)	−0.5 Mt (42% reduction, 67% increase)
Grassland → Construction	−0.6 Mt (9% of total loss)	−1.2 Mt (11% of total loss, 100% increase)	−0.2 Mt (67% reduction)	−0.3 Mt (58% reduction, 50% increase)	−0.2 Mt (67% reduction)	−0.3 Mt (58% reduction, 50% increase)
Grassland → Cultivated land	+0.1 Mt (low—carbon density conversion)	+0.3 Mt (stable trend)	+0.9 Mt (20% increase in conversion)	+0.6 Mt (10% decrease from 2030, balanced policy effect)	+0.9 Mt (20% increase in conversion)	+0.6 Mt (10% decrease from 2030, balanced policy effect)
Cultivated land → Forest/Grassland	+0.3 Mt (spontaneous ecological restoration)	−0.2 Mt (cropland prioritization intensifies)	+1.2 Mt (25% increase in ecological conversion)	+1.8 Mt (50% increase from 2030, strong reforestation effect)	+1.2 Mt (25% increase in ecological conversion)	+1.8 Mt (50% increase from 2030, strong reforestation effect)
Water bodies → Construction	−0.2 Mt (3% of total loss)	−0.5 Mt (5% of total loss, 150% increase)	−0.1 Mt (50% reduction)	0 Mt (strict protection sustained, no loss)	−0.1 Mt (50% reduction)	0 Mt (strict protection sustained, no loss)
Water bodies → Cultivated land	0 Mt (no net change)	+0.8 Mt (reclamation intensifies in long term)	0 Mt (protection maintained)	+0.2 Mt (minor reclamation, 83% reduction from natural scenario)	0 Mt (protection maintained)	+0.2 Mt (minor reclamation, 83% reduction from natural scenario)

). All scenarios exhibit a decline compared to 2020 levels, primarily due to reductions in forest and grassland areas. However, both protection scenarios outperform the natural development scenario in carbon retention.

Under the natural scenario, unconstrained urban expansion drives a cumulative loss of 83,038 hm² (11.72%) of cultivated land by 2040, contributing to a 7.8 Mt carbon storage decline (68% of total loss), particularly in peri-urban zones like Zhongmu County and Xinzheng City. Additionally, forest-to-construction conversion intensifies, with carbon losses 88% higher than in 2030, reflecting long-term sprawl-driven erosion of carbon-rich ecosystems in the Songshan and Fuxi Mountain regions. This trend leads to a sustained deterioration of regional carbon sink capacity.

The cultivated land protection scenario mitigates agricultural land loss by 44% (38,871 hm² cumulative reduction) compared to the natural scenario. However, this comes at the cost of increased grassland-to-cropland conversion, which adds 0.6 Mt to carbon storage due to higher soil carbon density in croplands (71.02 t/hm²) versus grasslands (43 t/hm²). Despite this net gain, the policy’s narrow focus on cultivated preservation results in continued urban encroachment—albeit at a slower rate (construction land growth declines from 2.87% in 2030 to 1.89% in 2040)—primarily onto low-carbon-density grasslands, highlighting insufficient integration of ecological considerations.

In contrast, the ecological protection scenario demonstrates the most effective carbon sink enhancement. Water body area expands by 2357 hm² (24%), contributing 0.76 Mt to soil carbon storage (water body carbon density: 32.48 t/hm²). Forest area increases by 1.74%, with 1.8 Mt of carbon gained through cropland-to-forest conversion via “grain for green” policies. Consequently, this is the only scenario where carbon storage decline narrows (0.09% decrease from 2030 to 2040). However, strict ecological safeguards lead to higher cultivated land loss (8.68% vs. 5.49% in the cultivated protection scenario), underscoring the need for comprehensive land consolidation—such as promoting eco-agriculture in the Yellow River floodplain—to balance ecological and agricultural objectives.

By 2040, carbon storage stabilizes across all scenarios (5.167 × 10⁷ t natural, 5.23 × 10⁷ t cultivated protection, 5.20 × 10⁷ t ecological), with marginal decreases from 2020 levels. This equilibrium arises as gains in forest and grassland areas partially offset construction land expansion and cultivated land reduction. Notably, the cultivated protection scenario maintains identical carbon storage from 2030 to 2040, reflecting its trade-offs between cropland retention and ecological restoration. These findings emphasize that integrated land use policies—combining cultivated protection with ecological restoration—are critical for optimizing carbon sequestration while ensuring sustainable development.

4.5.2. Spatial Distribution of Carbon Storage Under Different Scenarios

The spatial distribution of carbon storage in 2030 and 2040 remains largely unchanged across different development scenarios (Figure 9), preserving historical patterns with higher values in the southwest and lower values in the central regions. Central economic development zones continue to have relatively low carbon storage, while Dengfeng, Gongyi, and Xinmi maintain higher levels and show outward expansion trends. Dengfeng’s high carbon storage is mainly attributed to its predominant land types—cultivated land, forests, and grasslands—and effective conservation efforts in the eastern mountainous areas and the northern Songshan Mountain range.

4.6. Dynamic Correlations of Urbanization, Construction Area Expansion, and Carbon Stock Dynamics

Given the significant changes in urbanization rate (UR), construction land expansion rate (CEL), and carbon stock (CS) dynamics in Zhengzhou City revealed by prior analyses, further investigation into their quantitative interdependencies is warranted based on the elasticity coefficients (ECs) (

Table 12. Elasticity coefficients (ECs) of urbanization rate (UR) and construction land expansion rate (CEL) on carbon storage (CS) under different scenarios (ND, CLP, EP) from 2000 to 2040.

Index	2000→2010	2010→2020	2020→2030	2030→2040	$\overline{\mathit{E}\mathit{C}}(\mathit{U}\mathit{R}\to \mathsf{\Delta }\mathit{C}\mathit{S})$	$\overline{\mathit{E}\mathit{C}}(\mathit{C}\mathit{E}\mathit{L}\to \mathsf{\Delta }\mathit{C}\mathit{S})$
UREC	−0.511	−0.621	−1.545	−0.119	−0.699	-
	-	-	−0.564	0	−0.282 (Policy buffer)	-
	-	-	−1.036	−0.076	−0.556 (Partially offset)	-
CELEC	−0.018	−0.109	0.072 (Slowing expansion)	0.010 (Stable period)	-	−0.039 (Net loss)
	-	-	0.013	0	-	−0.006 (Strong inhibition)
	-	-	0.032	0.009	-	−0.011 (Moderate inhibition)

).

The impact of UR on CS reveals distinct dynamics across scenarios (Figure 10a–c). For the natural development scenario, a 1% increase in UR correlates with a ΔCS reduction of 0.51–1.55 Mt, with the absolute elasticity coefficient escalating alongside urbanization acceleration, underscoring the intensifying threat of unregulated urban expansion to carbon sequestration (Figure 10a). For the cropland conservation scenario, ΔCS stabilizes at zero between 2030 and 2040, indicating that stringent cropland protection policies effectively neutralize UR-driven carbon losses through enhanced agricultural carbon sinks (Figure 10b). Finally, in the ecological priority scenario, the absolute UR elasticity coefficient is systematically lower than in the natural scenario (e.g., −1.036 vs. −1.545 during 2020–2030) (Figure 10c), demonstrating that ecological safeguards attenuate the carbon loss intensity per unit of urbanization.

Regarding the impact of CEL on CS, a 1% annual increase in CEL corresponds to an additional ΔCS reduction of 0.109 Mt during 2010–2020 (Figure 10a), directly reflecting the encroachment effect of construction sprawl on high-carbon-density lands (e.g., cropland and grassland) in the natural development scenario. However, in policy intervention scenarios (cropland conservation and ecological priority) (Figure 10b,c), declining CEL narrows ΔCS. For instance, under the cropland conservation scenario (2020–2030), CEL = 0.93% yields ΔCS= −0.031 Mt—only 36% of the natural scenario’s corresponding loss—validating that restricting construction land expansion is pivotal to mitigating carbon losses.

This outcome delineates the dual-directional impacts of urban development on carbon stock dynamics. On one hand, the ecological conservation policies drove a 2.18% increase in forest cover, yielding a net carbon stock gain of 1.5 × 10⁶ t, demonstrating the compensatory effect of urban ecological restoration on carbon sequestration. On the other hand, rapid land urbanization (built-up area expansion) triggered the loss of high-carbon-density lands (cropland and grassland). During 2000–2020, cropland decreased by 15.33%, corresponding to an 8.6 × 10⁶ t carbon stock decline—accounting for 66% of the total reduction—which corroborates the classical thesis that “urban sprawl exacerbates carbon losses through degradation of natural carbon sinks” [2,15]. Consequently, the sustainable development dilemma faced by Zhengzhou city should first be addressed by optimizing the layout of construction land, such as prioritizing expansion in low-carbon-density areas, avoiding the occupation of arable land and ecological land, establishing an “arable land ecological” dual protection mechanism, and improving arable land quality and efficiency in plain areas to balance food security and carbon sequestration goals.

5. Discussion

5.1. Mechanisms Underlying Land Use–Carbon Storage Dynamics

The study’s multi-scenario simulations based on the PLUS and InVEST models reveal distinct patterns in land use and carbon storage dynamics in Zhengzhou. From 2000 to 2020, cultivated land declined by 15.33%, while construction land expanded by 13.31%, directly driving a 1.3 × 10⁶ t reduction in total carbon storage. This aligns with global trends where urban sprawl erodes high-carbon ecosystems like forests and croplands [7,13].

The study’s multi-scenario simulations reveal nonlinear trade-offs between urban expansion and carbon storage. Under the natural development scenario, construction land is projected to grow by 7.34% (2030) and 10.22% (2040), exacerbating carbon losses to 5.181 × 10⁷ t in 2030 (a 1.7% decline from 2020). In contrast, the ecological protection scenario mitigates losses through forest expansion (1.74% growth by 2040) and water body conservation, stabilizing carbon storage at 5.20 × 10⁷ t by 2040. Notably, the elasticity analysis quantifies this relationship: a 1% increase in urbanization rate correlates with a 0.51–1.55 Mt decline in carbon storage under baseline conditions, underscoring the critical role of policy intervention in mitigating such losses.

5.2. Policy-Driven Scenario Comparisons and Elasticity Insights

The elasticity analysis reveals a strong correlation between urban expansion phases and carbon storage dynamics. In the natural development scenario, the elasticity coefficient of urbanization rate (UR) on carbon storage ($E{C}_{UR}\to \mathsf{\Delta }CS$) surges from −0.621 (2010–2020) to −1.545 (2020–2030), coinciding with Zhengzhou’s accelerated urbanization after its 2016 designation as a National Central City. This underscores the “carbon penalty” of unregulated urban growth, consistent with Seto et al.’s (2012) findings on global urban carbon footprints [2,20,21]. This phase saw rapid policy-driven expansion, such as the development of Zhengdong New District and integration of counties like Zhongmu and Xinzheng, leading to unchecked conversion of high-carbon cropland (75.17% in 2000 → 59.84% in 2020) and forested areas into construction land (16.75% in 2000 → 30.05% in 2020). The spike in $E{C}_{UR}\to \mathsf{\Delta }CS$ indicates that unplanned urban growth intensifies carbon loss: each 1% UR increase during 2020–2030 corresponds to a 1.55 Mt decline in carbon storage, driven by low-density sprawl in ecologically sensitive zones.

Policy intervention scenarios demonstrate mitigation potential: the cultivated land protection scenario reduces CEL-driven losses by 64% (2020–2030) through cropland retention, while the ecological protection scenario achieves a 30% reduction in UR elasticity by prioritizing forest and wetland conservation. These results highlight that integrating policy rules (e.g., ecological red lines, cropland conversion thresholds) into land use models can enhance scenario realism and align with national strategies like China’s “Dual Carbon” goals [10,16].

5.3. Policy Implications for Sustainable Urbanization

The findings provide a practical framework for megacities to balance development and carbon governance. For instance, Zhengzhou’s transition to high-quality development requires targeted policy integration, dynamic scenario planning, and technological refinement, while simultaneously considering multi-objective parallelism and adopting a dynamic adjustment strategy to adapt to the pace of socioeconomic development.

The scenario analysis informs specific policy recommendations aligned with Zhengzhou’s Territorial Spatial Master Plan (2021–2035) and the Yellow River Basin Ecological Protection Plan. To mitigate carbon losses, a 3% annual cap on construction land expansion and ecological red lines in high-carbon zones (e.g., Songshan Mountain forests) is proposed, drawing from the ecological protection scenario’s 4.97% growth rate and 1.74% forest expansion by 2040. A 60% reduction in cropland conversion to urban areas (as tested in the CLP scenario) should be enforced, complemented by topsoil reuse to maintain cropland carbon density (71.02 t/hm²). Elasticity analysis underscores the need for adaptive policies, such as halting urban expansion when carbon loss thresholds are breached. These measures directly address trade-offs between urbanization and carbon storage, providing a scalable framework for regional planning.

5.4. Methodological Limitations and Future Research

This study’s reliance on historical trends and linear assumptions may underestimate nonlinear ecological thresholds, such as abrupt soil carbon saturation or policy shocks. Additionally, the idealized scenario design simplifies complex real-world interactions, potentially overlooking synergies or conflicts among policy objectives, market forces, and stakeholder interests.

To enhance this research, future work could integrate machine learning with real-time remote sensing to improve land use prediction precision, while hybrid modeling frameworks (e.g., combining PLUS-InVEST with economic or climate models) would better simulate feedback mechanisms between urban development and carbon cycles. High-resolution carbon mapping using LiDAR or social sensing data, alongside cross-regional comparisons between megacities, could refine regional applicability and identify transferable policy insights. And long-term resilience analysis is necessary to consider access carbon storage trends beyond 2040 under climate change projections to inform intergenerational planning.

6. Conclusions

• Urbanization dimensional analysis reveals dominant land urbanization (built-up area annual growth of 8.1% from 2000 to 2020), imbalanced “land–population” development, and economic vulnerability (GDP increased 16.5 times).
• The analysis of land use change demonstrates that cultivated land decreased by 15.33%, and construction land expanded by 13.31% in 2000–2020. Based on the PLUS model, the future land use structure in 2030 and 2040 was predicted. Under the 2030 scenarios, construction land expanded by 7.34% (natural), 2.87% (cultivated protection), and 4.96% (ecological protection), with effective control in protection scenarios. Cultivated land decreased by 6.96%, 2.36%, and 4.78%, respectively. By 2040, all scenarios showed construction land expansion and cultivated land reduction, but unlike 2030, forests were better protected (showing gradual growth versus 2020), and water bodies increased under ecological protection.
• The InVEST model was used to analyze the changes in Zhengzhou’s carbon storage from 2000 to 2040. In 2030, carbon storage measured 5.181 × 10⁷ t (natural), 5.235 × 10⁷ t (cultivated protection), and 5.209 × 10⁷ t (ecological protection), all lower than the 2020 levels. By 2040, values remained nearly unchanged from 2030, with forest/grassland expansion partially offsetting declines. The southwest–high/central-low spatial pattern persisted, with cultivated protection most effectively slowing carbon storage reduction. Future urban planning should integrate ecological measures while safeguarding farmland, control construction land growth, and contribute to climate goals like carbon peaking and neutrality.
• The dynamic relationship between urbanization rate, construction land expansion rate, and carbon storage was quantified using elasticity analysis. In natural development scenarios, the increase in urbanization rate and construction land expansion rate significantly exacerbates carbon losses (for every 1% increase in urbanization rate, carbon storage decreases by 0.51–1.55 Mt; for every 1% increase in construction land expansion rate per annum, carbon storage losses increase by 0.109 Mt), confirming the encroachment effect of unconstrained urban expansion on high-carbon-density land. The policy intervention scenario (farmland protection, ecological priority) effectively mitigates carbon losses by restricting land expansion and enhancing carbon sequestration capacity, highlighting the feasibility of policy regulation on carbon neutrality pathways.

In the above aspects, a sustainable urban development path involves prioritizing the allocation of construction land for spatial optimization, directing growth towards low-carbon-density areas, establishing a dual protection mechanism for farmland ecology with binding red lines, and implementing land quality improvement in plains to coordinate food security and carbon sequestration. This study provides a scalable framework for carbon governance in megacities, demonstrating how localized land use strategies can alleviate the trade-off between development and natural capital conservation.