Climate Change Through Urbanization: The Coupling Effects of Urbanization, Water Resources and Forests on Carbon Emissions

Abstract

The purpose of this paper is to quantitatively study the impact mechanism of urbanization, water resources, and forestry system coupling on carbon emissions, and explore new ways to reduce carbon emissions, as complex relationships exist among urbanization, water resources, and forestry systems. Based on the data of provincial regions in mainland China from 2015 to 2024, this paper analyzes the impact of urbanization, water resources, and forestry system coupling on carbon emissions by constructing the STIRPAT model. The findings reveal significant heterogeneity in the impact of the coupling degree among urbanization, water resources, and forestry systems on carbon emissions across Chinese provinces. Most regions exhibit insufficient carbon reduction effects. Enhancing the carbon mitigation effect through improving the coupling coordination of urbanization, water resources, and forestry systems presents a novel pathway toward achieving carbon neutrality during urbanization processes. Heterogeneity analysis further indicates that disparities in economic aggregate alter the mechanisms through which the STIRPAT model influences carbon emissions. The main contribution of this paper is to establish the evaluation index system of urbanization, water resources, and forestry development, analyze the mechanism of urbanization, water resources, and forestry coupling system affecting carbon emissions with the STIRPAT model, and explore new pathways for achieving carbon neutrality within urbanizing systems.

1. Introduction

Urbanization, as a critical pathway in the industrialization and modernization process of countries worldwide, serves as one of the key indicators for evaluating a country’s socioeconomic development level. Urban population growth promotes the increase in energy demand, leading to the increase in carbon dioxide emissions, which in turn leads to environmental degradation and related population pressure [1]. The role of socioeconomic factors, technological progress, and policy intervention in the relationship between urbanization and carbon emissions remains a subject of ongoing research [2]. Existing studies on the impact of urbanization on carbon emissions generally classify the effects into three categories: positive, negative, and nonlinear. Some scholars believe that urbanization has increased carbon emissions [3]. The consumption demand of urban residents leads to the growth of energy consumption and carbon emissions [4]. The extensive expansion of cities has brought about large-scale population aggregation and more intensive urban economic activities, resulting in an increase in energy demand and carbon emissions in the fields of housing, transportation, entertainment, and so on [5]. On the contrary, some scholars have pointed out that there is a negative correlation between urbanization and carbon emissions. Urbanization improves environmental quality by reducing carbon emissions [6]. Infrastructure construction in line with green policies will reduce carbon emissions [7]. Some literature focuses on the nonlinear impact of urbanization on carbon emissions [8]. Carbon emissions in the early and middle stages of urbanization increase rapidly, but the speed and intensity of carbon emissions in the late stage of urbanization show a downward trend [9], and there may be an inverted U-shaped relationship between urbanization and carbon emissions [10]. How, then, can we explain the intricate and diverse relationship between urbanization and carbon emissions? A macro-level, long-term analytical framework is required. Urbanization, water resources, and forestry systems should be integrated and analyzed in the study.

Many regions worldwide are undergoing varying stages of urbanization and industrialization, particularly in developing countries, exerting diverse—and often substantial—impacts on water quality. While China has achieved remarkable economic development through industrialization, this progress has been paralleled by escalating industrial pollution, triggering severe environmental degradation [11]. Despite the significant increase in water use efficiency, the water pollution problem in most parts of China still needs improvement, and the water resources in China’s coastal areas appear to be more fragile than those in the rest of the country [12,13]. China’s rapid urbanization and industrialization have propelled the nation to become the world’s second-largest economy; however, this accelerated urban development exerts substantial pressure on ecosystem integrity and environmental carrying capacity. Urban growth remains fundamentally dependent on natural ecosystem services, making the balance between urbanization processes and ecological sustainability a critical concern for contemporary human development.

Rapid population growth and industrialization have intensified the global water crisis through supply–demand imbalances and widespread contamination. As an important part of the theory of sustainable forestry, the carrying capacity of forest ecology is an important indicator for evaluating regional forest sustainability and social sustainability. Forests are the main body of terrestrial ecosystems, with a series of ecological functions including water conservation and soil conservation. China’s forestry sector, encompassing primary, secondary, and tertiary industries, represents both a fundamental economic sector and pillar industry that prioritizes ecological construction while enhancing forest resources, ensuring ecological security, and supporting economic development. This integrated system delivers substantial ecological and economic benefits, particularly through carbon sequestration and emission reduction mechanisms [14,15,16].

Given the substantial contribution of secondary production to China’s forestry sector, carbon emissions from forestry activities warrant significant attention. Studies have shown that although the forest products sector is generally cleaner compared to other sectors, carbon emissions remain considerable in developing nations, particularly China [17,18]. China’s urbanization rate represents one of the most rapid urbanization transitions in modern history [19]. Urbanization stimulates forestry’s secondary industry primarily through consumer market expansion, manifested in two key areas: first, the scale of new housing demand in cities and towns has led to the scale of the housing renovation and decorative wood market; second, emerging urban industries—including paper manufacturing, food processing, pharmaceuticals, and forestry chemicals—create diverse demand for forest-derived materials [20,21,22,23]. In order to alleviate the severe environmental problems, the National New Urbanization Plan (2014–2020), published by China in 2014, revealed a new urbanization path with Chinese characteristics, pursuing compatibility with the carrying capacity of resources and the environment [24] and focusing on green and low-carbon development. This policy framework establishes green urbanization as the fundamental paradigm for sustainable urban development.

Contemporary scholarship increasingly recognizes the critical importance of water and forestry resources in urbanization–carbon emission dynamics, with growing focus on how these natural resources influence the carbon footprint of urban development. While research in urbanization–carbon emission relationships continues to evolve with diverse methodological approaches, definitive conclusions remain elusive, necessitating further investigation. In order to clarify the mechanism and emission reduction path of the impact of urbanization, water resources, and forestry on carbon emissions, this study adopts a three-phase research framework. First, a comprehensive literature review identifies current research progress and methodological gaps, establishing the foundation for research objectives and design specifications. Second, empirical analysis examines the carbon emission reduction mechanisms operating under the integrated influence of urbanization, water resources, and forestry systems. Finally, policy recommendations demonstrate the practical applications and societal value of the research findings.

2. Literature Review

The literature review examines three interconnected research domains: the relationship between urbanization and water resources, the relationship between urbanization and forestry, and the relationship between water resources and forestry. Figure 1 shows the logical relationship between the topics in the literature. The literature review focuses on several pairs of variable relationships among urbanization, water sources, and forestry as shown in Figure 1. Existing scholarship has employed diverse theoretical frameworks and methodological approaches to investigate the complex relationships among urbanization, water resources, and forestry systems, yielding substantial empirical findings. However, the heterogeneous nature of these research outcomes has generated considerable debate and identified several unresolved theoretical and methodological questions requiring further investigation.

2.1. Relationship Between Urbanization and Water Resources

Urbanization, a primary driver of economic growth and social development, is often limited by the total amount of water resources [25,26]. Water resource availability is subject to climatic conditions, resulting in varying carrying capacities for urban populations. It is therefore imperative to investigate the supportive capacity of water resources for urban development through region-specific analyses that account for distinct climatic environments [27,28,29]. Research has identified the effects of “Urban Stream Syndrome”, whereby human activities that damage the natural environment cause river structure degradation, directly affecting the quantity and quality of surface water resources, thereby rendering them unable to support normal human life and production activities, and ultimately impacting humanity itself [30].

Unplanned urbanization poses significant threats to water resource sustainability, with particularly pronounced impacts in small and medium-sized cities characterized by limited technological capacity and lower developmental indices [31]. Urban expansion necessitates the conversion of substantial agricultural and rural land areas, fundamentally altering regional land use patterns. The transformation of cultivated land and forested areas into urban infrastructure significantly diminishes critical ecosystem services, including water conservation and environmental regulation functions [32]. This degradation correspondingly reduces the landscape’s capacity to mitigate water and air pollution [33], thereby intensifying urban heat island effects and compromising overall environmental quality [34].

Urban expansion typically involves substantial increases in impervious surface area, with the consequent loss of agricultural and forested land generating well-documented environmental impacts, including water scarcity [32], degraded water and air quality [33], and intensified urban heat island effects [34]. In particular, urbanization affects watershed microclimate, surface water dynamics, groundwater recharge, fluvial geomorphology, and river ecology [33]. Despite these recognized impacts, existing research lacks knowledge on how urbanization affects ecosystem structure and function, as well as its social and cultural dimensions, and how forest management can mitigate urbanization’s negative impacts.

2.2. Relationship Between Urbanization and Forestry

While urbanization drives economic and social development, it simultaneously generates substantial negative environmental consequences. Urban expansion consumes approximately 60% of converted land from agricultural areas [35], with this land use transformation representing a significant source of carbon emissions. Furthermore, the urban heat island effect elevates urban temperatures, which not only deteriorates the thermal environment of urban human settlements but also causes huge energy consumption, thereby contributing to global warming trends. In addition, in the severe form of global climate change, urban forest resources are affected to varying degrees by the land use and land cover change (LULC) caused by urbanization, which has a great impact on urban forest resources. Additionally, urbanization processes increase the frequency and severity of extreme weather events, creating cascading environmental risks.

Land use and land cover change (LULCC) research and its underlying driving mechanisms have emerged as critical frontiers in contemporary science on global climate change. Extensive scholarly investigation has examined the spatiotemporal characteristics of urbanization processes, their driving mechanisms, predictive modeling approaches, and associated environmental impacts [36,37].

2.3. Relationship Between Water Resources and Forestry

Forest management practices, including deforestation and reforestation activities, significantly influence land use patterns and subsequently affect water resource availability in agroforestry landscapes. Interdisciplinary water–forestry research has identified critical biophysical mechanisms governing forest–hydrological interactions, demonstrating that forest ecosystems serve essential functions in water resource conservation [38]. However, despite this recognized significance, comprehensive quantitative validation of these forest–water relationships remains lacking at the global scale [39]. The complexity of these challenges stems from their multi-scalar nature and the intricate interplay of direct and indirect biophysical watershed responses with socioeconomic dynamics and feedback mechanisms [40]. Ecological forestry practices significantly influence on watershed hydrology through multiple mechanisms, including peak flow reduction, delayed peak flow timing, and enhanced dry season base flows. For instance, afforestation initiatives in Danish river basins have effectively reduced river salinity and salt discharge, with water quality potentially returning to pre-deforestation conditions [41,42,43].

Water quality demonstrate significant correlations with forest coverage, while being concurrently influenced by climatic factors and forest management practices. Quantitatively, the stream increases with reduced forest cover but decreases following reforestation or diminished precipitation. Similarly, stream salinity rises with deforestation and declines with reforestation efforts [44]. These hydrological responses are particularly observable where forest cover modifications have been systematically implemented and maintained at watershed scales [45]. Comprehensive observations of water–forest linkages were increasingly tested in the twentieth century. An interdisciplinary field of research emerged in which foresters and hydrologists empirically examined water–forest interactions in complex models of influencing factors. The U.S. Forest Service conducted the first pairwise comparisons of watersheds, comparing the effects of interventions on water composition in one watershed to a reference watershed [46,47].

In summary, despite extensive research on pairwise systems (urbanization–water, urbanization–forestry, and water–forestry), the emergent properties of their triadic symbiotic relationship remain underexplored. Existing studies have mainly analyzed the relationship between urbanization, water resources, and forestry from the perspective of ecology and environmental science, yet few studies have focused on the relationship between urbanization, water resources, and forestry from the perspective of socioeconomic development, and have additionally overlooked the carbon emission impacts within the urbanization–water–forestry symbiotic system [35,36,37,38,39,40,41,42,43,44,45,46,47]. A coordinated development of urban, water, and forestry systems can effectively achieve carbon emission reduction effects.

Narrow analytical frameworks focus on bilateral relationships rather than comprehensive system interactions. For instance, Although recent scholarship has begun acknowledging these analytical constraints and exploring multifactorial approaches, substantial knowledge gaps persist regarding the carbon emission dynamics within integrated urbanization–water–forestry systems [35,36,37,38,39,40,41,42,43,44,45,46,47]. To address these limitations and better elucidate the complex interdependencies among urbanization processes, water resources, forestry management, and carbon emissions, this study adopts an ecological symbiosis framework that conceptualizes these components as interconnected elements within a unified system.

Research examining urbanization’s impact on carbon emissions exhibits significant methodological limitations, predominantly characterized by pairwise relationships. For instance, studies investigating urbanization–water resource dynamics typically neglect the carbon emission implications of forest ecosystems, while research examining water–forestry relationships frequently overlooks urbanization’s broader effects on natural resource allocation, technological innovation, and socioeconomic development trajectories. Recently, scholars have begun to investigate these limitations and explore multiple factors. In order to better explore the complex relationship between urbanization, water resources, forestry, and carbon emissions, this study adopts an ecological framework that conceptualizes these components as interconnected elements within a unified system.

To address the gaps in existing research, this study aims to propose a systematic methodology for analyzing the mechanism through which the coupling of urbanization, water resources, and forestry influences carbon emissions. Based on the results of this systematic analysis, strategies for improving the symbiotic system will be proposed. Accordingly, using provincial-level regions in China as a case study, this research applies the STIRPAT model to investigate the relationships among urbanization, forestry, water resources, and carbon emissions. It seeks to precisely quantify the carbon reduction effects of the urban–water–forest symbiotic system, thereby providing a scientific basis for advancing urbanization, promoting forestry development, and supporting the sustainable utilization of water resources. Furthermore, it offers novel insights for facilitating the coordinated and sustainable development of urban–water–forest systems at the provincial level.

Based on the results of this systematic analysis, it puts forward countermeasures to improve the symbiotic system and further explores new pathways for achieving carbon neutrality through forest ecosystem development. Using the STIRPAT model and empirical data on urbanization, water resources, and forestry at the provincial level in China, this study conducts an empirical analysis and provides recommendations for enhancement based on the measurement of symbiotic efficiency. The research achieves its intended goals, with the following key contributions: (1) It examines the interactions among urbanization, water resources, and forestry from a coupling perspective. (2) It constructs an extended STIRPAT model to investigate carbon emission mechanisms, thereby expanding upon existing research frameworks. (3) It measures the carbon emission effects of the symbiotic coupling among urbanization, water resources, and forestry, and proposes a pathway to carbon neutrality by enhancing the carbon reduction efficiency of the forestry symbiotic system.

3. Materials and Methods

The main hypotheses of this study are as follows: (1) Forestry, water resources, and urbanization form a symbiotic system, the mechanisms of which can be explained through a coupling degree model; (2) The coupled system of forestry, water resources, and urbanization exhibits heterogeneity, with different levels of urbanization corresponding to distinct coupling degrees; (3) A harmoniously coupled system of forestry, water resources, and urbanization is conducive to achieving carbon neutrality.

This study investigates the integrated application of analytical methods, including the construction of an evaluation indicator system, the entropy weight method, coupling degree analysis, and the STIRPAT model, in the analysis of carbon neutrality pathways. The research is conducted from the perspective of the complex symbiotic relationships among urbanization, water resources, and forestry development. The research process is illustrated in Figure 2.

As shown in Figure 2, this study mainly consists of four steps. First, a theoretical model is constructed through the selection of factors affecting carbon emissions. Then, based on STIRPAT, the benchmark model regression is established, followed by an examination of coupling mechanisms and an investigation into factor heterogeneity.

3.1. Construction of Basic Regression Model

To comprehensively examine the impact of various higher education factors on carbon dioxide emissions, it is essential to adopt appropriate and scientifically robust research methods. Numerous methodologies are widely employed in studying the influencing factors of CO₂ emissions, including the Environmental Kuznets Curve (EKC) hypothesis, the Kaya identity, the Logarithmic Mean Divisia Index (LMDI) decomposition method, and the STIRPAT model, among others. Research on the relationship between environmental quality and economic growth primarily relies on the Environmental Kuznets Curve (EKC) hypothesis, which posits an inverted U-shaped relationship between environmental pollution intensity and economic growth [48,49].

The Kaya identity is an equation that expresses CO₂ emissions as the product of four factors [50,51]. A large body of literature commonly employs time-series technology to evaluate the relationship between CO₂ emissions and GDP [52,53,54]. When studying the complex relationship between variables by controlling energy use, GDP, and carbon intensity based on the Kaya identity, the study found the long-term coupling characteristics of multiple factors [55]. However, the Kaya identity has certain limitations as an accounting equation and is not allowed to assess the hidden causal relationship between factors [56].

The Logarithmic Mean Divisia Index (LMDI) method is widely used in the research of carbon emissions, urban energy consumption, construction industry, and other fields, providing substantial benefits for theoretical investigation and evaluation [57,58,59,60]. LMDI is highly reliable and effective in energy use and carbon emission decomposition analysis. It decomposes the total energy use into various factors, so as to identify and quantify the contribution of each factor to the change in energy use [61]. In addition, the LMDI method offers several advantages over traditional Divisia decomposition approaches [62]. First, the decomposition results are independent of the decomposition path, which ensures consistency and reliability. Second, the LMDI method effectively processes zero values in the data and avoids the decomposition deviation caused by them.

The STIRPAT model has evolved through long-term development based on related research. After Ehrlich proposed the population–environment pressure model [63] in the 1970s, Commoner [64] proposed the classic IPAT model, an environmental impact equation, to explain the relationship between population, wealth, and environmental pressure. On this basis, different scholars reconstructed and extended the IPAT model according to research needs, among them the IGT model for studying the relationship between economic growth and ecological environment [65], and the ImPACT model for evaluating the impact of potential behavioral factors on the environment [66]. Although the above models are simple and intuitive, they lack the ability to explain the impact of a single element in the model on the environment. On the basis of the above models, Dietz et al. proposed the Stochastic Impacts by Regression on Population, Affluence, and Technology (STIRPAT), which overcomes the shortcomings of the IPAT model, and can show the degree of effect of each independent variable on the dependent variable in the model [67]. In recent years, the STIRPAT model has been widely employed to analyze the influencing factors of China’s greenhouse gas emissions [68,69,70,71].

A review of the literature in the field of research methodologies reveals that each method has its own advantages and disadvantages. The STIRPAT model plays an important role in revealing the influencing factors and change rules of carbon emissions, and has important value in further understanding carbon emissions. The STIRPAT model has a simple and clear structure and strong operability. The model has strong universality and practicability. The STIRPAT model can quantify the influencing factors of carbon emissions through common statistical analysis methods such as data validation and parameter estimation, yielding highly reliable results.

As mentioned above, based on the relevant literature [67], this paper directly gives the logarithmic form of the STIRPAT model of carbon emissions (1),

lnC = α₀ + α₁lnP + α₂lnA + α₃lnT + e

(1)

where C is the environmental pressure, expressed in terms of the total regional carbon emissions, P is the regional population, A is the degree of economic development and prosperity, expressed in terms of per capita GDP, and T is the technical level of carbon emissions, expressed in terms of carbon emissions per unit of GDP. e is the random disturbance term and α₀ is the constant term. In order to discuss the impact of urbanization, water resources, and forestry on carbon emissions, this study extends the STIRPAT model to obtain model (2),

lnC = α₀ + α₁lnP + α₂lnA + α₃lnT + α₄U + α₅lnW + α₆lnF + α₇ C_uwf + e

(2)

where U denotes the urbanization index, W represents the total regional water resources, and F indicates the development and conservation activities of regional forestry resources, measured by the total investment in forestry. In light of the complex interrelationships among urbanization, water resources, and forestry discussed earlier, this paper defines the interdependent and co-evolutionary interaction among these three systems as a coupling relationship. By incorporating the coupling level of these three dimensions as a control variable, Equation (3) is derived as follows:

lnC = α₀ + α₁lnP + α₂lnA + α₃lnT + α₄U + α₅lnW + α₆lnF + α₇ C_uwf + e

(3)

where C_uwf represents the coupling dependency level of the coordinated development of urbanization, water resources, and forestry. Coupling dependency is a measure of the interdependence of two or more entities.

3.2. Evaluation Index (For Urbanization, Water Resources, and Forestry) and Data Sources

Urbanization represents a comprehensive development process encompassing economic, social, and cultural dimensions, which fundamentally manifests as an interactive system dominated by endogenous agglomeration effects and supplemented by exogenous expansion mechanisms. It not only manifests itself in the transfer and gathering of population from rural to urban areas and the gradual increase in the urban population, but also includes the urbanization of land and of lifestyle, etc. During the urbanization process, individuals migrate to urban areas in pursuit of improved living standards and better quality of life, progressively transitioning from rural lifestyles to adopting urban modes of living and working. The massive concentration of population in urban areas drives the expansion of city scales, leading to both spatial enlargement of existing cities and the emergence of new urban centers. Urbanization constitutes a complex process of social progress in which modern civilization replaces traditional civilization by transforming the population and employment structure and upgrading the industrial structure.

Evaluation of urbanization requires the integration of multidimensional indicators spanning economic, social, and environmental domains. These factors not only include industrial structure, economic volume, and the structure of the working population, but also social development factors such as education, science and technology, medical care, and environmental protection. Based on the relevant studies [72,73,74], the evaluation index system is shown in

Table 1. Urbanization evaluation system.

Primary Indicator	Secondary Index
Demographic indicator	Urban population proportion
Economic indicators	GDP per capita
	Share of output value in secondary and tertiary industries in GDP
Social development indicators	Health technicians per 1000 people
	Expenditure on education
Science and technology	Number of patents granted
Urban construction	Investment in real estate enterprises
Pollution control	Investment in industrial pollution control

.

Table 2. Descriptive statistics of sample data.

Statistical Features	lnC	lnP	lnA	lnT	U	lnW	lnF	Cuwf
Mean	5.343	7.804	1.739	4.549	0.306	5.768	4.320	0.534
Standard Error	0.040	0.041	0.023	0.035	0.007	0.079	0.050	0.005
Median	5.348	7.869	1.682	4.450	0.286	6.054	4.416	0.527
Standard Deviation	0.732	0.747	0.427	0.645	0.118	1.449	0.933	0.083
Sample Variance	0.536	0.559	0.183	0.416	0.014	2.099	0.870	0.007
Kurtosis	0.434	0.188	−0.160	−0.153	−0.001	0.516	0.629	0.113
Skewness	−0.618	−0.694	0.450	0.048	0.766	−1.035	−0.346	0.694
Range	3.436	3.237	2.137	3.406	0.568	6.291	5.582	0.414
Minimum	3.329	5.857	0.885	2.720	0.105	1.925	1.540	0.379
Maximum	6.890	9.450	2.997	6.202	0.655	8.082	6.99	0.807
Count	510	510	510	510	510	510	510	510

shows the statistical characteristics of the observed data. Data on urbanization, water, and forests in provincial-level regions of China from 2008 to 2024 are sourced from the China Statistical Yearbook. Based on the urbanization evaluation index system, this study uses the entropy weight method to evaluate the new urbanization level in different regions. This paper preprocesses the evaluation index data when calculating the urbanization index. GDP and per capita GDP data were deflated with the GDP deflator index. After logarithm calculation of relevant indicators, the urbanization index is evaluated by the entropy weight method. The entropy weights of the different indicators are then measured, thereby enhancing analytical objectivity [75].

3.3. Coupling Strength

The reciprocal and symbiotic relationship among urbanization, forestry, and water resources development is primarily reflected in the following two aspects. The reciprocal and symbiotic relationship between urbanization, forestry, and water resource development is mainly reflected in the following two aspects. On the one hand, forestry and water resources can promote the development of new urbanization from multiple dimensions including natural resource attributes, economic value, and social development value. On the other hand, new urbanization is environmentally friendly urbanization, which minimizes the damage to natural resources during the process. At the same time, better urbanization can provide better technology, funding, and other resources for the protection and sustainable development of natural resources. This paper uses the coupling index to reflect the development level of urbanization, forestry, and water resources [76,77], where the coupling strength coefficient is [78,79]:

$$C_{uwf}=3\times {\left[{\displaystyle \frac{lnU\cdot lnW\cdot lnF}{{\left(lnU+lnW+lnF\right)}^3}}\right]}^{{\displaystyle \frac{1}{3}}}$$

(4)

4. Results

In panel data analysis, this study tested the multiple correlation of variables and selected the variance expansion factor (VIF) method for testing (as shown in

Table 3. Variance inflation factor test.

Variable	VIF	1/VIF
lnP	2.851	0.351
lnA	5.641	0.177
lnT	2.733	0.366
U	5.862	0.171
lnW	5.650	0.177
lnF	2.522	0.397
Cuwf	6.891	0.145
Mean VIF	4.593

). The average value of the variance expansion factor of variables in the model is 4.593, meeting the data requirements of panel regression. At the same time, after conducting Hausman test fixed-effects and random-effects models, it is found that the fixed-effects model is suitable for this study.

4.1. Baseline Regression and Model Robustness Test

In order to ensure the accuracy of the research results, this study changes the sample size by changing the sample data selection time period, and tests the stability of the model on the basis of changing the sample size. The data in

Table 4. Baseline regression and model robustness test results.

Time Interval	lnP	LnA	LnT	U	LnW	LnF	Adj.R2
2007–2023	0.532 ***	−0.597 ***	0.375 ***	3.671 ***	−0.057 **	0.053 *	0.992
2017–2023	0.510 ***	−0.821 ***	0.383 ***	4.039 ***	−0.025 *	0.005 *	0.990
2016–2022	0.518 ***	−0.846 ***	0.354 ***	4.204 ***	−0.032 *	0.020 *	0.991
2015–2021	0.554 ***	−0.681 ***	0.297 ***	3.245 ***	−0.055 **	0.033 *	0.990
2014–2020	0.559 ***	−0.634 ***	0.269 ***	3.048 ***	−0.071 ***	0.065 *	0.990

* p value < 0.1, ** p value < 0.05, *** p value < 0.01.

shows that the analysis results of panel data with different sample sizes show good model stability. The sign direction of the variable does not change, and the variation range of the regression coefficient is small. The Appendix B presents the model robustness test of the transformation variables, and tests the adaptability of forestry investment variable in the model through the transformation of forestry indicators.

According to the national panel data analysis, the factors influencing total carbon emissions, ranked in descending order of the impact of each factor on the total carbon emissions, are as follows: urbanization index (U), total population (lnP), technology level (lnT), forestry resources (lnF), water resources (lnW), economic development, and wealth level (lnA).

The urbanization index (U) has the highest impact coefficient on carbon emissions, indicating that urbanization is the main factor for the increase in carbon emissions in China. In this study, the new urbanization index used in the calculation of urbanization index is used to examine the urbanization level of different regions from multiple dimensions such as economic growth and social development. This shows that the new urbanization index will still significantly increase the increase in carbon emissions in the short term.

The coefficient of the variable total population (lnP) is positive, indicating that the indicator changes in the same direction as carbon emissions. In general, there is a positive correlation between the total population and carbon emissions. The findings presented here must be interpreted within a specific historical context. In recent years, China has experienced a precipitous decline in births, leading to a slowdown in population growth across many regions. Despite this demographic shift, total carbon emissions have continued to rise. Nevertheless, the deceleration in population growth is expected to contribute to the long-term control of aggregate carbon emissions.

The impact of technology level of carbon emissions (lnT) on carbon emissions is also positive, which shows that China’s economic growth still depends on high-intensity energy input, and a large number of carbon emissions are generated. In this study, forestry investment is used to represent the increase, protection, and development of forestry resources.

The impact of forestry resources (lnF) on carbon emissions is also positive, which shows that the impact of people on forestry resources promotes the increase in carbon emissions. The carbon emission benefit of forestry resource investment needs to be improved. This is an interesting discovery, which has a specific impact mechanism behind it. The mechanism of the positive impact of forestry investment on carbon emissions may be as follows: (1) Forestry investment has short-term effects on carbon emissions. For example, the use of logging machinery, forestry related infrastructure construction, and reforestation in forestry investment causes large short-term carbon emissions. (2) Forestry investment has not reached the state of return to scale, or has entered the range of no return to scale. (3) Low investment efficiency caused by low level of forestry technology. The economic and ecological benefits of forestry investment need to be improved.

Water resources (lnW) and economic development and wealth level (lnA) have a negative impact on carbon emissions, indicating that both factors contribute to emission reduction through specific mechanisms. Areas rich in water resources are more suitable for the development of hydropower and water conservancy projects and the application of clean energy such as hydropower to achieve carbon emission reduction. This study uses per capita GDP to represent the level of economic development and wealth. Regions with high per capita GDP have generally more advanced and well-established industrial structures, scientific and technological innovation capabilities, educational level, infrastructure systems, etc. Economic development and wealth level reflect the quality of economic and social development to a certain extent, and high-quality development is conducive to reducing carbon emissions.

Table 5. Regional differences in carbon emission mechanisms.

Area	lnP	lnA	lnT	U	lnW	lnF	Adj.R2
North China	0.458 ***	0.268 ***	0.420 ***	−0.245 ***	0.160 ***	−0.031 ***	0.976
Northeast	0.993 ***	0.175	−0.333	0.782	−0.116	−0.050	0.957
East China	1.170 ***	−0.513 ***	−0.583 ***	0.949 **	−0.212 ***	0.132 ***	0.983
Central South	1.131 ***	−0.206 *	−0.619 ***	−0.838	−0.126 ***	0.040 *	0.980
Southwest	0.322 ***	−0.182	0.319 ***	1.774 *	0.060	0.158 **	0.970
Northwest	0.962 ***	2.409 ***	−0.153 **	−11.153 ***	−0.226 ***	−0.189	0.974
Yangtze River	0.652 ***	−0.616 ***	0.042 *	3.510 ***	−0.168 ***	0.253 **	0.988
Yellow River	0.372 ***	0.424 *	0.137 ***	1.291	−0.240 ***	0.515 **	0.985

* p value < 0.1, ** p value < 0.05, *** p value < 0.01.

shows the regional differences in carbon emission mechanisms. The panel data analysis of sub regions shows that there is significant heterogeneity between regions. The population factor is the mechanism to promote the growth of carbon emissions in different regions. Wealth level, technical factors of carbon emission intensity, urbanization, water resources, and forestry factors all have the characteristics of regional carbon emission impact differentiation. In regions with higher levels of economic development, such as East China, South Central China and Southwestern China, the impact of wealth indicators on carbon emissions is negative. There are four factors that have a negative impact on carbon emissions in the northwest and central and southern regions with better carbon emission reduction mechanisms. The North China region exhibits the fewest negative influencing factors on carbon emissions, indicating that its carbon reduction mechanisms require further improvement.

A common feature in both the Yangtze River Basin and the Yellow River Basin is the negative impact of water resources on carbon emissions. Abundant water resources facilitate the development of clean energy sources such as hydropower and support the substitution of traditional thermal power generation. Figure 3 shows the regional differences in the carbon emission mechanism. If the Pentagon of a region converges to the center, the carbon emission mechanism tends to be eco-friendly. The figure shows that the regional difference in carbon emission mechanism mainly comes from the difference in urbanization index (U), followed by the difference caused by lnA.

4.2. Impact Mechanism of Urbanization, Water Resources and Forestry Coupling on Carbon Emissions

The difference in carbon emissions in different regions may be mainly caused by the different coupling degrees of urbanization, water resources, and forestry. By introducing coupling degree as a control variable, the underlying mechanisms influencing regional carbon emissions can be evaluated more accurately. To ensure the accuracy of the research results, this analysis changes the sample size by changing the sample data selection time period, and tests the stability of the model on the basis of changing the sample size. The data in

Table 6. The impact of urbanization, water resources, and forestry coupling on carbon emissions.

Time Interval	lnP	lnA	lnT	U	lnW	lnF	Cuwf	Adj.R2
2007–2023	0.952 ***	0.941 ***	0.929 ***	6.765 ***	−0.373 ***	−0.159 ***	−13.653 ***	0.995
2017–2023	1.000 ***	0.893 ***	0.945 ***	6.168 ***	−0.339 ***	−0.185 ***	−13.821 ***	0.991
2016–2022	0.930 ***	0.818 ***	0.913 ***	6.746 ***	−0.320 ***	−0.147 ***	−13.230 ***	0.993
2015–2021	0.937 ***	0.863 ***	0.912 ***	6.770 ***	−0.329 ***	−0.145 ***	−13.405 ***	0.993
2014–2020	0.928 ***	0.860 ***	0.904 ***	7.173 ***	−0.334 ***	−0.134 ***	−13.480 ***	0.993

*** p value < 0.01.

shows that the results of panel data analysis with different sample sizes demonstrate relatively high model stability. The sign direction of the variable does not change, and the variation range of the regression coefficient is small. The model with coupling control variable has better stability and higher coefficient regression significance.

From the national panel data analysis (Table 6), it is found that the factors affecting total carbon emissions are ranked in descending order of impact as follows: urbanization index (U), total population (lnP), wealth level (lnA), and technology level (lnT), forestry resources (lnF), water resources (lnW), and the coupling degree of urbanization, water resources and forestry development. The urbanization index (U) has the highest impact coefficient on carbon emissions, indicating that urbanization is the main factor for the increase in carbon emissions in China. The coefficient of the variable total population (lnP) is positive, which indicates that the indicator changes in the same direction as carbon emissions. The impact of technology level of carbon emissions (lnT) on carbon emissions is also positive, which shows that China’s economic growth still depends on high-intensity energy input, and a large number of carbon emissions are generated. The impact of wealth level (lnA) on carbon emissions is also positive.

Forestry resources (lnF) and water resources (lnW) have a negative impact on carbon emissions. Water resources and forestry development activities have a mechanism to reduce carbon emissions. Areas rich in water resources are more suitable for the development of hydropower and water conservancy projects and the application of clean energy such as hydropower to achieve carbon emission reduction. The coupling degree has a negative impact on carbon emissions, which fully illustrates the importance of coordinated development of urbanization, water resources, and forestry resources. The coefficient of C_uwf (−13.653) means that the 0.01 unit change in the coupling index has a relatively negative impact on 13.653% of the emission coefficient. In view of China’s vast territory and unbalanced regional development, it is imperative to analyze the regional differences in carbon emission mechanisms. The analysis results are shown in

Table 7. Regional differences in carbon emission mechanisms considering coupling degree variables.

Area	lnP	lnA	lnT	U	lnW	lnF	Cuwf	Adj.R2
North China	0.802 ***	0.948 ***	0.973 ***	3.679 ***	−0.193 ***	−0.279 ***	−10.337 ***	0.930
Northeast	0.988 ***	0.949 ***	0.933 **	8.858 ***	−0.338 ***	−0.163 ***	−15.106 ***	0.912
East China	0.788 ***	1.059 ***	1.016 ***	5.823 ***	−0.324 ***	−0.054 **	−12.440 ***	0.990
Central South	1.312 ***	0.725 ***	−0.103 *	2.085 ***	−0.289 ***	−0.018 *	−9.308 ***	0.983
Southwest	0.982 ***	0.847 ***	0.702 ***	7.157 **	−0.370 ***	0.031 *	−13.137 ***	0.973
Northwest	1.031 ***	1.384 ***	1.057 ***	9.434 ***	−0.444 ***	−0.249 **	−16.746 ***	0.892
Yangtze River	0.920 ***	0.748 ***	0.638 ***	5.290 ***	−0.340 ***	0.027 *	−10.544 ***	0.910
Yellow River	0.971 ***	1.353 ***	0.864 **	6.458 **	−0.380 ***	−0.149 ***	−13.884 ***	0.902

* p value < 0.1, ** p value < 0.05, *** p value < 0.01.

. The regression results show that the impact mechanism of carbon emission in most regions is similar to the overall situation of the country. Population, wealth, technology, and urbanization all show a positive impact on carbon emission, which is consistent with the reality of China’s social and economic development.

The data in Table 7 shows that the heterogeneity of carbon emission mechanisms among different regions has decreased after the coupling control variable is incorporated. Different regions and watersheds show similar carbon emission impact mechanisms, and the total population, wealth level, carbon emission intensity, and urbanization all show positive effects. This suggests that various influencing factors of carbon emissions in the process of economic growth and urbanization have an impact on different regions. There is no fundamental difference in the carbon emission mechanism in different regions of China, and the difference is mainly reflected in the degree of influence of the influencing factors. Water resources and forestry have a negative impact on carbon emissions, especially when the coupling degree of urbanization, water resources, and forestry is taken into account.

Regions with high coupling degree of urbanization, water resources, and forestry can better promote the coordinated development of economy and environment. For example, Inner Mongolia’s afforestation initiatives demonstrate synergistic interactions by simultaneously advancing ecological conservation and economic development, thereby fostering complementary relationships across all three subsystems. Regional analysis reveals distinct patterns of system coordination. Collaborative dynamics characterize Kunming, Yuxi, and Honghe in Yunnan Province, areas concentrated in central and southern regions where elevated economic development levels, optimized industrial structures, and integrated water resource management during urban construction processes yield superior overall coordination. The regional difference in carbon emission mechanism under the adjustment of coupling degree is shown in Figure 4. The figure shows that the coupling degree of urbanization and urbanization, water resources, and forestry are the main factors causing the difference in carbon emission mechanism.

4.3. Heterogeneity Analysis

With the continuous promotion of urbanization, a large number of rural populations are transferred to cities, and the number of urban populations is increasing, leading to the increase in the demand for infrastructure and the serious problem of carbon emissions while driving the regional economic development. However, the agglomeration of population can bring scale effect, as well as knowledge and technology spillover effect, thereby reducing regional carbon emissions. This study argues that regional carbon emissions are closely related to the level of economic development, population, and human capital. It mainly analyzes the heterogeneity of regional carbon emissions from the perspective of urban economic scale, population scale, and human capital level.

The economic development level has different effects on regional carbon emissions. In the early stage of economic development, when the level of social production is low, the industrial structure is dominated by the secondary industry. In order to meet the needs of people’s lives, enterprises increase material production, increase infrastructure construction, etc., and the consumption of resources and energy increases, aggravating the problem of carbon emissions. When the economy develops to a relatively developed level, the public’s awareness of environmental protection gradually increases, the government’s environmental regulation efforts are improved, the industrial structure is transformed and upgraded, and the technical level and energy utilization efficiency is improved, contributing to the reduction in regional carbon emissions.

From the perspective of urban population scale, the increase in regional population is conducive to the formation of scale effect, the promotion of industrial agglomeration, the encouragement of enterprises’ specialized and cleaner production, the improvement in energy efficiency, and the reduction in carbon emissions. On the other hand, the improvement in agglomeration improves the level of regional economic development, and the government and enterprises have a positive awareness of environmental protection, which is conducive to energy conservation and emission reduction. However, the population agglomeration will also have a crowding effect, increasing the production and living costs.

According to the endogenous economic growth theory, the improvement in human capital will stimulate the spillover effect of knowledge and technology, thereby improving energy efficiency and promoting the reduction in regional carbon emissions. When the regional human capital level is low, most of the labor force is limited to agriculture or traditional industrial production, which is not conducive to the optimization and upgrading of the industrial structure, resulting in high energy consumption and rising carbon emissions. However, in areas with high human capital, the scientific and technological level is more advanced, which is more conducive to the research, development, promotion, and application of cleaner regional production technology, and addressing the problem of carbon emissions from the source.

In order to test the difference in the impact of economic development, urban population, and human capital scale heterogeneity on carbon dioxide emissions, this paper divides the research samples into high and low groups according to the total economic output, the number of urban population, and the number of people with regional higher education degrees, and carries out panel data regression, respectively. The regression results are shown in

Table 8. Heterogeneity analysis.

Heterogeneity of Scale	lnP	lnA	lnT	U	lnW	lnF	Cuwf
High economic aggregate	0.453 ***	−0.414 **	0.105 **	1.774 *	0.015 *	−0.001 *	2.919 *
Low economic aggregate	0.630 ***	0.070 *	0.237 *	7.089 ***	−0.096 *	0.090 *	−4.582 *
High urban population	0.783 ***	0.896 ***	0.831 ***	4.708 ***	−0.273 ***	−0.044 ***	−10.323 ***
Low urban population	0.931 ***	1.083 ***	1.260 ***	8.708 ***	−0.406 ***	−0.269 ***	−16.518 ***
High human capital	0.921 ***	1.105 ***	1.045 ***	5.479 ***	−0.254 ***	−0.218 ***	−14.073 ***
Low human capital	0.935 ***	0.834 ***	0.992 ***	10.607 ***	−0.394 ***	−0.082 ***	−15.583 ***

* p value < 0.1, ** p value < 0.05, *** p value < 0.01.

.

Economic development has a significant heterogeneous impact on carbon emissions. Wealth indicators and forestry variables in regions with high economic output have a negative effect on carbon emissions, while water resources and coupling variables in regions with low economic output have a negative effect on carbon emissions. The impact coefficient of the urbanization index on carbon emissions in regions with large economic scale (1.774) is considerably smaller than that in regions with low economic scale (7.089). A possible explanation is that economically advanced regions benefit from relatively abundant fiscal revenue, higher public demand for environmental quality, more advanced scientific and technological capabilities, and greater efficiency in pollution management. With the continuous improvement in economic level, the positive externalities of urbanization gradually emerge, which is conducive to reducing regional carbon emissions, while regions and cities with low development may still be in the stage of industrialization and undertake more pollution-intensive industries relocated from developed regions.

Regions with large urban population and those with small urban population have similar carbon emission mechanisms, with various factors influencing carbon emissions in the same direction. The impact of heterogeneity of urban population size on carbon emissions is mainly reflected in the numerical difference in regression coefficient. Regions with larger urban population have smaller regression coefficients, indicating lower carbon emission intensity. A possible explanation is that the emission reduction effect of economies of scale and agglomeration effect brought by population agglomeration in large and medium-sized cities is greater than that of the increase in people’s demand for infrastructure. For small cities, the advancement of urbanization leads to substantial energy consumption, thereby diminishing the positive externalities of urbanization on carbon reduction.

The heterogeneity in how human capital affects carbon emissions exhibits a mechanism similar to that of urban population size. Regions with high levels of human capital demonstrate lower carbon emission coefficients, particularly reflected in the urbanization indicator. A potential explanation is that in areas with high human capital, during the early stages of urbanization, population agglomeration led to congestion effects, and increased demand for infrastructure prompted greater production and construction of related facilities, thereby exacerbating carbon emissions. However, after urbanization reached a critical threshold, further urban development facilitated more efficient resource sharing and widespread knowledge spillovers, which accelerated technological progress and promoted the development and application of clean production technologies. Consequently, the growth rate of regional carbon emissions slowed. In contrast, regions with lower human capital may lack the capacity for clean technology innovation due to relatively lower educational attainment, resulting in less advanced technology and lower energy efficiency. Figure 5 shows the difference in carbon emission mechanism in different regions under the condition of heterogeneity of environmental factors. Observing the figure, we can see that the source of the difference mainly lies in urbanization and its coupling with water resources and forestry.

5. Discussion and Comparison of Existing Literature

The non-equilibrium characteristics and regional differences within China’s urbanization–water resources–forestry symbiotic system primarily stem from variability in urbanization processes. While urbanization facilitates coordinated development between urbanization and water resources [78], disparities in urbanization development levels generate uncoordinated interactions between urbanization and both water resources and forestry [79]. Industrial structure upgrading during urbanization processes demonstrates significant negative impacts on coordination levels in the lower reaches of the Yangtze River [80]. Urbanization levels exhibit significant diversification effects on the coordination between urbanization and natural resources, consistent with the environmental Kuznets curve framework proposed by Grossman and Krueger [81,82].

The environmental Kuznets curve describes the relationship between pollution and per capita income, wherein pollution increases with GDP per capita at low-income levels and decreases with GDP growth at high-income levels. Applying this framework, this study’s findings demonstrate that symbiotic synergy among urbanization, water resources, and forestry increases with urbanization at low urbanization levels and decreases with urbanization at high urbanization levels. This pattern indicates that diverse landscape planning scenarios tailored to different urbanization intensities are critical for maintaining or improving species diversity in urban forests. These findings underscore the importance of landscape planning in conserving urban species diversity, ensuring urban ecosystem stability, and promoting sustainable development [83]. Furthermore, the role of human capital in advancing water ecological civilization during new urbanization processes requires consideration [84], particularly given the significant disparities in human capital quantity and quality across China’s provincial regions.

Enhancing symbiosis among urbanization, water resources, and forestry fundamentally depends on technological innovation and management innovation, necessitating increased scientific and technological investment and improved management systems. Current scientific research funding intensity for urbanization–water resources–forestry symbiosis remains insufficient to effectively enhance the symbiotic level and scale efficiency. Furthermore, achieving comprehensive improvement in symbiotic ecology, industry, science, and technology innovation requires coordinated measures encompassing government leadership, core enterprise and industry initiatives, and resource support mechanisms. Water resources and forest resources constitute the material foundation for synergistic development within the urbanization–water resources–forestry symbiotic system.

Given insufficient per capita forest area and water resources, enhancing total natural resource availability represents a fundamental strategy. Forest quality can be improved through cultivating superior seeds and seedlings for afforestation and strengthening forest production base construction. Scientific forest resource management requires enhancement, including upgrading existing forest resource quality, accelerating the cultivation of middle-aged and young forests, and reforming low-yield and low-efficiency forests. Strengthening ecological supervision and regional environmental early warning systems, improving environmental law enforcement levels, constructing ecological function protection zones, and establishing long-term construction and management mechanisms for these protection zones will collectively enhance total forestry resource availability.

Simultaneously, developing ecological and cultural industries that integrate urbanization, water resources, and forestry should establish a resource-saving and ecologically friendly industrial policy system. This requires constructing a development framework with rich local characteristics that encompasses ecological and cultural dimensions within the integrated urbanization–water resources–forestry system. Ecologically friendly industries should be developed by maximizing environmental resource advantages from scenic areas, nature reserves, and forest parks. The urbanization of integrated ecological and cultural industries must be enhanced to strengthen system cohesion. Publicity and education regarding ecological culture in integrated water resources and forestry development should be intensified, with ecological cultural knowledge disseminated to increase nationwide participation in ecological cultural construction. Sustainable development of the urbanization–water resources–forestry symbiotic system will be promoted through coordinated forestry resource development, enhanced water resource availability, and urban ecological cultural construction.

Ecological protection challenges during urbanization processes are widespread across developing countries. This comparative analysis examines research findings from similar studies in India, Brazil, and other regions to provide broader contextual understanding. Rapid urbanization near India’s Kerwa Forest Area (KFA) has negatively impacted ecosystem services including water filtration and supply [85]. Similarly, research on Delhi’s urban forests examined how urbanization alters underground fungal biodiversity, subsequently affecting water resource regulation [86]. Land use changes during urbanization and industrialization in Malaysia generate sustained negative impacts on water quality across various water systems [87]. In northern Thailand, primary land use changes involve increased rubber plantation and building areas, resulting in reduced forest coverage and negative impacts on basin ecosystem support functions [88]. Brazilian land use impact studies demonstrate that urbanization constitutes the primary cause of water quality degradation, while simultaneously revealing forests’ protective effects on water quality. Research on Rio de Janeiro’s informal settlements emphasizes urban forestry’s importance in improving living conditions, particularly within densely populated low-income areas [89]. These examples illustrate the interconnected nature of urban, water, and forestry systems across diverse environments, with empirical results predominantly reflecting urbanization’s negative effects. This study demonstrates the feasibility of exploring coordinated development pathways for urbanization, water resources, and forestry.

Based on the perspective of urbanization, water resources, and forestry coupling, measures to promote the construction of sponge city can be taken. The construction of sponge city can solve the problems of rainwater resource utilization and runoff pollution, and take into account the prevention and control of urban waterlogging and the development of urban forestry resources. As a part of urban elements, green space should not only be balanced with other space, but also establish cooperation with other elements to jointly realize rainwater and flood management. Using the regulation and storage function of green space, the city can freely absorb excess rainwater and store and release water resources like a “sponge”.

6. Conclusions

This study achieves its research objectives by examining symbiotic mechanisms among urbanization, forestry, and water resources, and exploring novel pathways for carbon neutrality in forestry ecosystems through symbiotic mechanism analysis. Findings reveal significant heterogeneity within urbanization–water resources–forestry symbiotic systems across China’s provincial regions. The symbiotic system has not yet achieved comprehensive collaborative evolution, and sustainable development levels within the symbiotic system require enhancement. Carbon reduction effects of urbanization–water resources–forestry symbiotic systems remain insufficient. Enhancing these carbon reduction effects represents a novel approach to achieving carbon neutrality in forestry ecosystems. The study’s main conclusions align with existing research perspectives on the coordinated development of urbanization, water resources, and forestry [78,79,80], environmental curve theory [81,82], and sustainable development frameworks [83,84].