Exploring the dynamic impact of biomass energy, deforestation, fossil fuels and green technology on greenhouse gas emissions in China

China faces the dual challenge of sustaining economic growth while reducing greenhouse gas (GHG) emissions, yet the combined effects of energy structure, technological innovation, and land-use policies remain underexplored. This study examines the long- and short-run drivers of GHG emissions across Chinese provinces from 2000 to 2022 using second-generation panel estimators that address cross-sectional dependence and slope heterogeneity (CS-ARDL, AMG, CCEMG). A unique contribution of this work is the joint inclusion of biomass energy use, fossil energy consumption, energy intensity, green technology innovation, and deforestation dynamics within a unified empirical framework. The results show that fossil fuel use and energy intensity significantly accelerate emissions, while innovation and afforestation policies help mitigate environmental pressures. Biomass energy increases emissions due to continued reliance on traditional rural combustion practices. These findings provide new empirical evidence to guide China’s transition toward clean energy, innovation-led growth, and strengthened forest-based carbon sinks in support of long-term decarbonization strategies.

1 Introduction

China is the world’s largest emitter of greenhouse gases (GHGs), accounting for nearly 31% of global CO₂ emissions in 2022 (World Energy Outlook 2023 – Analysis - IEA, 2023). Over the past two decades, rapid industrialization and urbanization have fuelled economic growth but also intensified environmental pressures, with rising air pollution and climate-altering emissions threatening both national and global sustainability goals (Chen Y. et al., 2025; Meo and Anees, 2025; Zhao et al., 2025). According to the World Bank, China’s per capita carbon emissions increased from 2.7 metric tons in 2000 to 8.1 metric tons in 2022, highlighting the urgent need to design effective, evidence-based climate policies. These figures demonstrate the importance of examining the drivers of emissions within the context of China’s transition to a low-carbon economy.

To better understand these challenges, this study draws on two key theoretical frameworks. The Environmental Kuznets Curve (EKC) hypothesis suggests that environmental degradation initially worsens during early industrialization as economies prioritize growth, but eventually declines as societies adopt cleaner technologies and stronger environmental policies. In parallel, Ecological Modernization Theory (EMT) argues that technological innovation and institutional reforms can decouple economic growth from environmental harm by promoting cleaner production and sustainable practices. These frameworks provide a foundation for analyzing how China’s rapid development, coupled with its push for green innovation and afforestation, shapes its emissions trajectory.

While China has invested heavily in renewable energy, green technology, and afforestation programs, the effectiveness of these initiatives in reducing GHG emissions remains an open empirical question (Gupta et al., 2024; Yu et al., 2025; Zhang et al., 2023). Existing literature often focuses on fossil fuel dependency and industrialization but pays limited attention to the interactions between traditional biomass energy use, technological innovation, and land-use dynamics. In addition, regional disparities in energy practices and forest management policies create complexity that national-level studies frequently overlook. This highlights the need for a comprehensive analysis that integrates energy structure, ecological factors, and technological progress.

Recent studies have explored diverse strategies to address the intertwined challenges of energy use, greenhouse gas emissions, and sustainability. For instance, (Chen Z. et al., 2025) demonstrated that photocatalytic birnessite-nitrate amendments can simultaneously inhibit arsenic mobilization and greenhouse gas emissions in flooded paddy soils, offering an innovative approach to environmental management. Duan (2025) emphasized the need for a balanced pathway that harmonizes energy consumption, environmental protection, and sustainable development. Similarly, (Meng et al., 2023) highlighted the role of collaborative scheduling in integrated energy systems, showing that incorporating carbon restrictions into energy planning can significantly enhance the efficiency and sustainability of power distribution. Collectively, these studies underscore the importance of integrating technological, environmental, and policy measures to reduce emissions and promote long-term sustainability.

Figure 1 shows the trend of total greenhouse gas (GHG) emissions in China from 2000 to 2022. Emissions rose sharply from about 4,381 Mt CO₂-equivalent in 2000 to a peak of 13,835 Mt CO₂-equivalent in 2021, reflecting rapid industrialization, urbanization, and heavy reliance on fossil fuels, particularly coal. The steep increase between 2000 and 2012 was driven by export-led industrial growth and energy-intensive production. From 2013 onward, the growth rate moderated slightly as renewable energy adoption, energy efficiency measures, and environmental regulations began to take effect. A temporary slowdown occurred in 2020 due to the COVID-19 pandemic, followed by a sharp rebound in 2021 as industrial activity resumed. In 2022, a slight decline suggests early progress toward China’s dual carbon targets of peaking emissions before 2030 and achieving carbon neutrality by 2060. This trend underscores the urgent need for policies promoting clean energy, green innovation, and improved efficiency to curb emissions while sustaining economic growth.

Figure 1

Figure 1. Trends in provincial greenhouse gas (GHG) emissions in China (2000–2022). Data source: International Energy Agency (IEA, 2024). Emissions expressed in kilotonnes of CO₂-equivalent using the GWP-100 metric under IPCC AR5. Author’s processing.

A notable feature of China’s energy transition is its increasing reliance on renewable and biomass energy sources. While modern biomass energy is often positioned as a cleaner alternative to fossil fuels, in practice, much of China’s rural biomass use remains traditional and inefficient leading to emissions of particulate matter and greenhouse gases (Wang et al., 2023). Similarly, fossil fuel consumption, particularly coal, continues to dominate the national energy structure, undermining climate targets despite a surge in solar, wind, and hydro investments (Ibrahiem and Hanafy, 2020). Compounding this challenge is China’s relatively high energy intensity, which reflects inefficiencies in energy conversion and usage across industrial sectors.

In contrast, green technology innovation (GTI) has emerged as a promising driver of emissions reduction, supported by strong public R&D investment and a national commitment to clean energy patents and green manufacturing (Kashif et al., 2024; Liu et al., 2024). Simultaneously, land-use change, changes in forest cover, particularly increases in deforestation, weaken the carbon sink capacity and consequently raise GHG emissions. China’s ecological restoration programs—such as the Grain-for-Green initiative and Natural Forest Conservation Program—have significantly reshaped the country’s forest landscape, offering potential mitigation benefits (Wang et al., 2023).

The current study contributes to this area of research by examining the long-run and short-run impacts of biomass energy consumption (BIO), fossil fuel consumption (FOS), energy intensity (EI), green technology innovation (GTI), and deforestation trends (NFC) on GHG emissions in China from 2000 to 2022. Using advanced second-generation econometric techniques such as CS-ARDL, AMG, and CCEMG, the study accounts for cross-sectional dependence, slope heterogeneity, and non-stationarity to produce robust and reliable results. This approach provides a nuanced understanding of the relationships between these variables and offers both theoretical and practical implications for sustainable energy and environmental management.

The analysis focuses on 30 provinces in mainland China (excluding Hong Kong, Macao, and Taiwan) over 2000–2022, resulting in a balanced panel of 690 observations. This structure allows us to capture cross-provincial heterogeneity in economic structure, resource use, and environmental performance under unified national regulatory conditions.

The findings show that fossil fuel consumption and energy intensity are major contributors to emissions, while biomass energy also has a positive impact due to inefficient, traditional usage in rural areas. In contrast, green technology innovation and afforestation efforts play a significant role in mitigating emissions. By linking energy structure, innovation, and ecological restoration, this study not only broadens the literature on emission drivers but also offers practical guidance for policymakers. These insights align with China’s dual carbon goals of peaking emissions before 2030 and achieving carbon neutrality by 2060, while also providing lessons for other rapidly offering insights for China’s provinces and, by extension, for other industrializing economies seeking sustainable development pathways.

The study specifically addresses the following research questions.

1. What are the long-run and short-run effects of biomass energy, fossil fuels, energy intensity, green innovation, and deforestation on greenhouse gas emissions in China?

2. Does green technology innovation significantly mitigate GHG emissions in China?

3. Do biomass practices and deforestation policies undermine or support China’s emission-reduction goals?

To address these questions empirically, we construct a balanced panel dataset for Chinese provinces over the period 2000–2022, allowing for interprovincial comparison under a unified national policy framework.

This study contributes in three ways.

• It integrates energy structure, green technology innovation, and forestry policy variables into a unified empirical framework, providing a more comprehensive understanding of China’s emission drivers.

• It applies second-generation panel techniques (CS-ARDL, AMG, CCEMG) that explicitly address cross-sectional dependence and heterogeneous adjustment dynamics often ignored in earlier work.

• It offers new evidence on the effectiveness of China’s large-scale afforestation programs in strengthening carbon sink capacity, contributing to the growing literature on land-use–emissions interactions.

By answering these questions, the study contributes to the growing body of literature at the nexus of energy policy, ecological governance, and sustainable development. The results offer actionable insights for policymakers seeking to design regionally adaptive, low-carbon strategies that are both environmentally and economically viable.

2 Literature review

Greenhouse gas (GHG) emissions have become a critical concern in the global sustainability agenda, particularly for rapidly developing economies where industrialization and energy consumption patterns are evolving rapidly. China represents a unique case as it continues to pursue rapid growth while committing to peak carbon emissions by 2030 and carbon neutrality by 2060 (Miao et al., 2024). The literature highlights that emissions are shaped not only by conventional factors such as fossil fuel consumption and industrial energy demand but also by renewable energy adoption, innovation capacity, and land-use transformations (Raza and Lin, 2025). Therefore, a comprehensive review of these interlinked drivers is essential to frame the empirical setting of this study.

2.1 Fossil fuel consumption and energy intensity

A large body of research agrees that fossil fuel dependence remains the primary source of rising GHG emissions worldwide. Coal continues to dominate China’s energy system and remains a major obstruction to its decarbonization efforts (Mehmood et al., 2024; Martins et al., 2021; Meo and Adebayo, 2025). Studies consistently find that increased coal-based electricity generation, industrial combustion, and associated pollutant releases are responsible for escalating environmental degradation in emerging economies. In addition, energy intensity, representing the amount of energy consumed per unit of output, is widely recognized as a marker of production inefficiency. Higher energy intensity values imply lagging technology adoption and outdated industrial processes, leading to excessive emissions (Tian and Khan, 2025). Within China, energy efficiency levels vary significantly across provinces due to differing industrial structures and economic orientations, resulting in spatially uneven emission patterns.

Despite these insights, empirical studies rarely capture how fossil fuel use and energy intensity interact dynamically in China’s provinces within models that address cross-sectional dependence and heterogeneous responses across regions.

2.2 Biomass energy utilization

Biomass is often promoted as a renewable substitute for fossil fuels. However, contemporary research increasingly acknowledges that the climate benefits of biomass vary based on combustion technology, feedstock type, and regulatory oversight (IEA, 2021). In China, traditional biomass such as crop residues and firewood remains heavily used in rural and underserved areas, often burned in basic stoves that generate substantial carbon emissions and air pollutants (Chen J. et al., 2017). Evidence from South and Southeast Asia suggests that inefficient biomass utilization may produce more harmful outcomes than fossil alternatives if technological upgrading is absent (Balanay and Halog, 2024; Chowdhury et al., 2025). As China expands its renewable energy portfolio, the contribution of bioenergy to sustainability objectives remains uncertain and requires further investigation.

The environmental impact of biomass has not been adequately examined at the provincial panel level in China, especially when evaluated jointly with energy efficiency and fossil dominance (Bettarelli et al., 2023; Sun et al., 2025).

2.3 Green technology innovation (GTI)

Green innovation stands at the center of the ecological modernization vision, which proposes that technological progress can harmonize economic development with environmental improvements. Numerous studies show that GTI, measured through environmental patents, R&D expenditure, or clean technology diffusion, supports reductions in carbon intensity and pollutant emissions (Bergendahl et al., 2018; Chang et al., 2023; Chen Y. et al., 2022). China has strengthened its low-carbon transition through innovation-oriented strategies such as the Made in China 2025 initiative, carbon trading pilots, and green finance reforms. Even so, the impact of innovation is not uniform across regions; provinces equipped with research infrastructure and stronger governance experience greater environmental benefits compared to resource-dependent or under-developed areas (Kula, 2010).

Few studies incorporate GTI into a dynamic heterogeneous empirical framework that also considers traditional and modern energy structure features within China.

2.4 Deforestation and land-use dynamics

Land-use change plays a critical role in carbon cycling because forests serve as natural carbon sinks (Chiriacò et al., 2024). Conventional evidence from many countries shows that deforestation leads to higher GHG emissions by reducing sequestration capacity and releasing stored carbon from biomass and soil (Li Y. et al., 2022). However, China represents a contrasting case. Since the 2000s, the government has implemented unprecedented ecological restoration programs that have reversed decades-long deforestation trends. Projects such as the Grain-for-Green Program and the Natural Forest Conservation Program have substantially increased forest coverage, contributing to long-term emissions mitigation (He et al., 2023). The dual nature of land-use effects in China offers an empirical opportunity to understand the climate benefits of afforestation policy interventions.

Very limited research incorporates forest policy achievements and deforestation dynamics simultaneously with energy-related drivers when assessing GHG determinants.

2.5 Summary and overall gap

Taken together, prior studies improve understanding of how fossil reliance, energy inefficiency, biomass practices, innovation, and forest transitions affect environmental performance. However, three shortcomings remain clear.

• Fragmented perspective: Studies typically focus on only one or two drivers, overlooking the interconnected mechanisms that jointly influence emissions.

• Methodological limitations: Many neglects cross-sectional dependence and dynamic adjustments that are critical in provincial panel settings.

• Missing policy integration: The climate impact of China’s land-use restoration efforts has rarely been evaluated alongside its energy transition and innovation agenda.

2.6 Theoretical foundations and hypotheses

The Environmental Kuznets Curve suggests that early development intensifies emissions due to fossil fuel use and industrial scale. Traditional biomass combustion similarly increases emissions in rural China due to incomplete burning. In contrast, ecological modernization theory posits that green technology and environmental policies can progressively reduce emissions. Afforestation programs further enhance carbon sequestration.

Drawing on these theoretical foundations and empirical findings, the following hypotheses are proposed.

• H1: Biomass energy consumption has a positive effect on GHG emissions due to traditional and inefficient usage patterns.

• H2: Fossil fuel consumption positively influences GHG emissions by increasing carbon output and air pollutants.

• H3: Higher energy intensity is positively associated with GHG emissions as a result of inefficient production systems.

• H4: Green technology innovation has a negative impact on GHG emissions by promoting cleaner production and energy efficiency.

• H5: An increase in net forest change (i.e., net forest expansion through afforestation and reforestation) reduces greenhouse gas emissions by strengthening carbon sequestration capacity.

This hypothesis structure provides a clear and testable framework that links theoretical arguments with empirical evidence, ensuring alignment with the study’s objectives.

3 Data and methodology

3.1 Model specification

All estimations are performed on a provincial panel for China, where each cross-sectional unit represents a province, and the time dimension spans 2000–2022. This study uses a balanced provincial panel consisting of 30 provinces in China observed annually from 2000 to 2022, resulting in 690 province–year observations (30 × 23). This study examines the long- and short-run determinants of greenhouse gas (GHG) emissions in China by modeling emissions as a function of energy structure, technological progress, and ecological pressure. The general long-run model is specified as given in Equation 1.

${\text{GHG}}_{it}={\alpha }_i+{\beta }_1{\text{BIO}}_{it}+{\beta }_2{\text{FOS}}_{it}+{\beta }_3{\text{EI}}_{it}+{\beta }_4{\text{GTI}}_{it}+{\beta }_5{\text{NFC}}_{it}+{\epsilon }_{it}(1)$

Where.

• i and t denote province and time (year),

• β_j​ are the slope coefficients,

• ${\epsilon }_{it}$ is the error term.

This model allows the estimation of both long-run elasticities and short-run dynamic adjustments.

3.2 Data and variables

The variables in this study were carefully selected based on their theoretical relevance and empirical importance in explaining greenhouse gas (GHG) emissions. Table 1 provides the definition, proxy, and data source for each variable, along with supporting references from previous studies to justify their inclusion.

• Greenhouse Gas Emissions (GHG): Total GHG emissions, including CO₂, CH₄, and N₂O, are used as the dependent variable. This measure reflects the overall environmental burden caused by energy use and land-use change. Similar indicators have been used by (Shindell et al., 2012; Lynch et al., 2020) to analyze emission patterns across developing and developed economies.

• Biomass Energy Consumption (BIO): Biomass energy includes organic materials such as crop residues, wood, and animal waste. Although classified as renewable, traditional biomass use often causes incomplete combustion, leading to significant emissions of particulate matter and GHGs (Lanjekar et al., 2024).

• Fossil Fuel Consumption (FOS): The share of fossil fuels, including coal, oil, and natural gas, in total energy consumption represents the degree of carbon-intensive energy dependency. Several studies, such as (Hoa et al., 2024; Li W. et al., 2022), confirm fossil fuel use as a dominant driver of GHG emissions, particularly in emerging economies like China.

• Energy Intensity (EI): Energy intensity is measured as the amount of energy consumed per unit of GDP and serves as a proxy for energy efficiency. Higher energy intensity indicates inefficient production systems and greater environmental stress (Filippini and Hunt, 2015; Zanjani et al., 2022). This metric has been widely applied in previous studies to evaluate structural inefficiencies in industrial sectors.

• Green Technology Innovation (GTI): GTI is proxied through environmental patent applications, reflecting investments in cleaner production technologies and sustainable practices. (Cutcu et al., 2024; Chang et al., 2023) show that green innovation plays a critical role in mitigating emissions by promoting low-carbon industrial development.

• Net forest change (NFC): Net forest change captures the annual percentage change in provincial forest area, reflecting the combined effects of afforestation, reforestation, and deforestation. Positive values indicate net forest expansion, while negative values indicate net forest loss. This measure is particularly appropriate for China, where large-scale ecological restoration programs have substantially increased forest coverage since the early 2000s. Using net forest change rather than gross deforestation allows a more accurate assessment of how land-use dynamics influence greenhouse gas emissions through changes in carbon sink capacity.

Table 1

Table 1. Variable description and sources.

For each indicator, we specify its construction method as follows: CO₂ emissions are expressed as per capita CO₂-eq; renewable energy is measured as the share in total final provincial energy use; and technological innovation indicators represent the number of environmental patent applications. GDP and urbanization are measured per capita and as population share, respectively. These definitions ensure consistent interpretation of environmental and economic outcomes across provinces.

The dataset is organized as a balanced provincial panel for China over the period 2000–2022. For each province and year, greenhouse gas emissions (GHG) are measured as per capita CO₂-equivalent emissions, biomass energy (BIO) and fossil fuels (FOS) are expressed as shares of provincial final energy consumption, energy intensity (EI) is defined as provincial energy use per unit of provincial GDP, green technology innovation (GTI) is captured by the number of environmental patent applications, and Net Forest Change (NFC) is measured as the annual percentage change in provincial forest area. Since NFC measures forest loss, an increase in NFC indicates declining carbon sinks, which theoretically should lead to higher GHG emissions. This construction ensures that all indicators are comparable across provinces and over time. The dataset forms a balanced panel of N = 30 provinces and T = 23 years, yielding 690 province-year observations. This structure is well-suited for the CS-ARDL estimator, which accommodates cross-sectional dependence and heterogeneous dynamic adjustments across economic units. All variables are harmonized as per capita, share, or intensity indicators to ensure comparability across provinces and years.

All variables are converted into natural logarithms to interpret elasticities and to reduce heteroscedasticity.

A provincial-level panel is particularly appropriate for this study because China exhibits substantial interprovincial differences in industrial structure, population density, energy mix, and environmental regulation, while all provinces operate under a common national institutional framework. By focusing on provincial units, the analysis captures cross-provincial heterogeneity in the drivers of greenhouse gas emissions and enables more targeted policy insights for China’s internal low-carbon transition.

Figure 2 illustrates the hypothesized relationships between key factors and greenhouse gas (GHG) emissions in China. Biomass energy consumption (BIO), fossil fuel consumption (FOS), and energy intensity (EI) are expected to positively influence emissions, as traditional biomass use leads to incomplete combustion, fossil fuels release high levels of carbon dioxide, and high energy intensity reflects inefficient production systems. In contrast, green technology innovation (GTI) and deforestation policies (NFC) are anticipated to reduce emissions. GTI drives cleaner production and energy efficiency, while reforestation and sustainable land management enhance carbon sequestration. This framework highlights the combined role of energy practices, technological progress, and ecological policies in shaping China’s emission trajectory.

Figure 2

Figure 2. Conceptual framework showing hypothesized relationships.

3.3 Estimation strategy

Our goal is to recover credible long-run and short-run effects in a panel that exhibits cross-sectional dependence, slope heterogeneity, and non-stationarity. First-generation estimators perform poorly under these features, which are common in energy and environment panels. We therefore use second-generation procedures that explicitly model common shocks and allow parameters to vary across units. This design aligns with our research question, which requires both dynamic adjustment paths and heterogeneous long-run elasticities.

Step 1: Pre-estimation diagnostics.

We test for cross-sectional dependence using Pesaran’s CD test to detect common shocks and spillovers (Pesaran, 2004). We assess slope heterogeneity with the Δ and adjusted-Δ tests to verify that responses differ across units (Hashem Pesaran and Yamagata, 2008). These diagnostics guide the choice of estimators that remain valid when units are interlinked and heterogeneous.

Step 2: Integration order and cointegration.

We apply the CIPS panel unit root test, which is robust to cross-sectional dependence, to determine the order of integration (Pesaran, 2007). Given I(1) variables, we test for a long-run equilibrium using Westerlund’s panel cointegration tests that allow for dependence through common factors (Westerlund, 2007). Evidence of cointegration justifies models that separate long-run relations from short-run dynamics.

Step 3: Baseline estimator.

We estimate a Cross-Sectionally Augmented ARDL model to obtain unit-specific short-run coefficients and long-run elasticities while controlling for unobserved common factors via cross-sectional averages (Chudik and Pesaran, 2015). This framework accommodates endogenous dynamics, heterogeneous slopes, and weak exogeneity of regressors. Lag lengths are selected by information criteria. Long-run effects follow from the error-correction representation, and the error-correction term verifies adjustment toward equilibrium.

For the CS-ARDL specification, we set a maximum of two lags for the dependent variable and the regressors and selected the optimal lag length using the Akaike Information Criterion (AIC). To retain a parsimonious structure and avoid over-parameterization, a common lag order is imposed across provinces, while allowing the slope coefficients to differ across cross-sectional units in line with the heterogeneous panel framework. Robustness checks with alternative lag orders yielded very similar long-run elasticities.

Step 4: Robustness estimators.

We re-estimate long-run coefficients using AMG and CCEMG, which remain consistent under pervasive common shocks and heterogeneous responses. AMG captures a common dynamic process and then averages unit-specific coefficients. CCEMG augments each regression with cross-sectional averages and then takes the mean group estimate (Pesaran, 2006). Convergence of signs and magnitudes across CS-ARDL, AMG, and CCEMG supports robustness.

4 Results

The empirical analysis begins with an examination of the interrelationships among the variables through the correlation matrix presented in Table 2. The correlation results show that greenhouse gas emissions (GHG) are positively associated with fossil fuel energy consumption (FOS) (r = 0.621), energy intensity (EI) (r = 0.477), and Net Forest Change (NFC) (r = 0.332), suggesting that these variables contribute directly to increased emissions. Conversely, biomass energy consumption (BIO) (r = −0.512) and green technology innovation (GTI) (r = −0.385) exhibit negative correlations with GHG emissions, implying that these factors may play a mitigating role in environmental degradation. Additionally, GTI and NFC are inversely correlated (r = −0.451), supporting the idea that innovation contributes to conservation.

Table 2

Table 2. Matrix correlation.

Table 3 reports the cross-sectional dependence (CSD) test results using Pesaran’s method. All variables, including GHG, BIO, FOS, EI, GTI, and NFC, exhibit statistically significant dependence (p < 0.01), indicating the presence of strong interconnections among countries in the panel. This dependence could be attributed to the global integration of energy markets and coordinated environmental policies, necessitating panel data methods that account for such dependencies.

Table 3

Table 3. Cross-sectional dependence (CSD) test.

Table 4 provides evidence of slope heterogeneity using both the Δ̂ and Δ̃ tests. The highly significant test statistics (p < 0.01) confirm that slope coefficients differ across cross-sectional units, reinforcing the heterogeneity assumption and validating the use of flexible estimators such as CS-ARDL, AMG, and CCEMG.

Table 4

Table 4. Slope heterogeneity test.

The stationarity properties of the data are assessed using the CIPS unit root test, as presented in Table 5. All variables are found to be non-stationary at level but become stationary after first differencing, suggesting that the variables are integrated of order one, I(1). This finding justifies the subsequent application of cointegration techniques.

Table 5

Table 5. CIPS unit root test.

To evaluate the long-run relationship among the variables, the Westerlund (2007) cointegration test is employed. As shown in Table 6, all four test statistics (Gt, Ga, Pt, and Pa) reject the null hypothesis of no cointegration at the 1% significance level. This implies that a long-run equilibrium relationship exists between GHG emissions and its key determinants: BIO, FOS, EI, GTI, and NFC. Hence, it is appropriate to estimate both long-run and short-run dynamics using CS-ARDL and confirm robustness through AMG and CCEMG techniques.

Table 6

Table 6. Westerlund (2007) cointegration test results.

Together, the preliminary tests validate the suitability of second-generation panel estimators that accommodate heterogeneity, non-stationarity, cointegration, and cross-sectional dependence, ensuring that the empirical model yields robust and credible insights into the environmental implications of energy and innovation variables.

Table 7 presents the CS-ARDL results with greenhouse gas emissions (GHG) as the dependent variable. The estimated error-correction term ECT (−1) in Table 7 is negative and statistically significant, confirming the presence of a stable long-run equilibrium relationship between GHG and its determinants. Its magnitude implies that approximately X% of any short-run deviation from the long-run path is corrected in the following year.

Table 7

Table 7. CS-ARDL estimates (Dependent variable: GHG).

$\text{You\hspace{0.17em}compute\hspace{0.17em}}\mathrm{X}\text{\hspace{0.17em}as}:\mathrm{X}=-\text{ECT}-\text{coefficient}\times 100$

$\text{Example}:\text{if\hspace{0.17em}ECT}=-0.35\to \mathrm{X}\approx 35\%$

All variables were transformed into their natural logarithms, allowing the coefficients to be interpreted as elasticities, which quantify the percentage change in GHG emissions associated with a 1% change in each independent variable. This provides a direct measure of marginal effects, linking the statistical results to real-world policy implications.

In the long run, fossil fuel consumption (FOS) has the largest positive elasticity at 0.221 (p < 0.01), implying that a 1% increase in fossil fuel use leads to a 0.22% rise in GHG emissions. Similarly, energy intensity (EI) has a significant elasticity of 0.176 (p < 0.01), meaning that a 1% increase in energy inefficiency raises emissions by 0.18%, underscoring the urgent need for industrial modernization and efficiency policies. Biomass energy consumption (BIO) also exerts a positive effect with an elasticity of 0.084 (p < 0.05). While biomass is renewable, its traditional and inefficient use, especially in rural areas, results in incomplete combustion, releasing methane, carbon dioxide, and particulate matter (Al-qazzaz et al., 2024). This indicates that without modernizing biomass systems, such as through biogas or advanced pellet technologies, rural energy practices will continue to worsen emissions.

Conversely, green technology innovation (GTI) has a negative long-run elasticity of −0.143 (p < 0.01). This means that a 1% improvement in green innovation reduces GHG emissions by 0.14%, confirming that technological advancements promote cleaner production and greater energy efficiency. The negative and statistically significant coefficient for net forest change confirms that forest expansion contributes to lower greenhouse gas emissions in both the long run and short run. This result reflects the effectiveness of China’s large-scale afforestation and reforestation programs, which have enhanced provincial carbon sink capacity and partially offset emissions from energy-intensive activities. Importantly, the negative sign does not imply that deforestation reduces emissions, but rather that net forest gains play a mitigating role in China’s emission dynamics.

In the short run, the magnitudes are slightly smaller but remain consistent in direction. A 1% increase in fossil fuel consumption raises GHG emissions by 0.20%, while a 1% increase in energy intensity results in a 0.15% increase. Meanwhile, green technology innovation and reforestation efforts continue to have significant mitigating effects, reducing emissions by 0.12% and 0.08%, respectively. Biomass retains its positive impact with an elasticity of 0.06%, highlighting the immediate environmental burden of inefficient biomass practices.

These elasticity-based interpretations translate statistical outputs into practical policy insights. For instance, a 10% improvement in energy efficiency nationwide could lower GHG emissions by nearly 1.8%, while a 10% increase in green technology investment could reduce emissions by 1.4%. Conversely, a 10% rise in fossil fuel use would increase emissions by 2.2%, demonstrating the critical need to transition away from carbon-intensive energy sources.

The negative and highly significant ECT (−1) value indicates that approximately 34.5% of the previous year’s deviation from long-run equilibrium is corrected each year, confirming a stable and convergent long-run relationship among the variables.

Additional diagnostic statistics including R², Adjusted R², F-statistic significance, and LM serial correlation test results are reported in Appendix Table A1 to support model validity. Table A1 shows diagnostic statistics for CS-ARDL model.

Figure 3 displays the relationship between China’s energy mix and greenhouse gas (GHG) emissions from 2000 to 2022. The figure shows that as fossil fuel consumption remains dominant, GHG emissions have risen sharply, highlighting the country’s dependence on carbon-intensive energy sources. In contrast, the share of renewable energy has grown gradually, but its pace has not been sufficient to offset the effects of fossil fuels.

Figure 3

Figure 3. Energy mix vs. greenhouse gas (GHG) emissions in China (2000–2022).

These visualizations complement the statistical results in Table 7, providing a clear and intuitive understanding of the data trends. Trend lines illustrate how changes in the energy mix influence emissions over time, reinforcing the need for policy measures that accelerate the transition to clean energy while improving energy efficiency.

4.1 Robustness analysis

To ensure the robustness of these results, Table 8 reports estimate from the Augmented Mean Group (AMG) and Common Correlated Effects Mean Group (CCEMG) estimators. The coefficients across both methods are largely consistent in sign, magnitude, and significance with those from the CS-ARDL model. AMG and CCEMG are applied to validate the CS-ARDL long-run elasticities by addressing unobserved common factors and heterogeneous adjustment dynamics across Chinese provinces.

Table 8

Table 8. Robustness check via AMG and CCEMG estimators.

Specifically, fossil fuel consumption (AMG: 0.208, CCEMG: 0.195; both p < 0.01) and energy intensity (AMG: 0.162, CCEMG: 0.149; both p < 0.01) remain the strongest contributors to GHG emissions. Green technology innovation continues to exhibit a statistically significant negative effect (AMG: −0.135, CCEMG: −0.128), affirming its role as a strategic lever for environmental improvement. While the coefficients for biomass and Net Forest Change are slightly smaller than those in CS-ARDL, they retain statistical significance and directional consistency. Table 9 shows regional average of key variables.

Table 9

Table 9. Regional averages of key drivers (2000–2022).

Collectively, the robustness checks validate the core findings, confirming that structural and technological factors in the energy sector critically influence emission trajectories in the panel of Chinese provinces studied.

5 Discussion

This study explores the long- and short-run drivers of greenhouse gas (GHG) emissions in China, with emphasis on energy consumption patterns, green technology innovation, and land-use dynamics. The empirical results yield important insights into China’s dual challenge of maintaining economic growth while pursuing environmental sustainability.

The empirical findings provide clear evidence supporting the proposed hypotheses. H1 is confirmed, as biomass energy consumption shows a positive and significant effect on GHG emissions, indicating that traditional and inefficient biomass use remains a pollution source. H2 and H3 are also supported since fossil fuel consumption and higher energy intensity significantly increase emissions, validating the argument that China’s reliance on coal and energy-inefficient industrial processes intensifies environmental degradation. Conversely, H4 is validated by the negative and significant coefficient of green technology innovation, confirming that cleaner technologies and environmental R&D reduce emissions in line with ecological modernization theory. Finally, H5 is supported by the negative and statistically significant coefficient of net forest change, indicating that forest expansion driven by China’s large-scale afforestation and reforestation programs enhances carbon sequestration and contributes to lower greenhouse gas emissions. Taken together, these findings confirm that the proposed theoretical framework and hypothesized relationships are empirically validated.

The positive and significant effects of fossil fuel energy consumption (FOS) and energy intensity (EI) on GHG emissions are consistent with China’s current energy landscape. Despite rapid growth in renewable energy deployment, coal remains a dominant energy source, contributing significantly to the country’s carbon footprint. Additionally, high energy intensity, particularly in industrial sectors, reflects structural inefficiencies that exacerbate environmental degradation.

The apparent contradiction between the negative bivariate correlation between biomass energy consumption and GHG emissions and the positive long-run elasticity obtained from the CS-ARDL model reflects the difference between unconditional association and conditional causal effects. The correlation coefficient captures simple co-movement and is influenced by structural factors, such as higher biomass use in less industrialized provinces that also exhibit lower overall emission levels. In contrast, the CS-ARDL estimates isolate the partial effect of biomass energy while controlling for fossil fuel use, energy intensity, green innovation, and land-use dynamics. Once these confounding factors are accounted for, biomass energy—predominantly traditional biomass used in rural households—exerts a positive impact on emissions due to inefficient combustion and methane release. Quantitatively, the long-run elasticity implies that a 10% increase in traditional biomass consumption raises GHG emissions by approximately 0.8%, underscoring that without technological upgrading, biomass reliance can undermine decarbonization objectives.

The difference between the negative correlation between biomass energy consumption and greenhouse gas emissions and the positive coefficient obtained from the regression analysis reflects the distinction between unconditional association and conditional effects. The correlation matrix captures simple co-movement and is influenced by structural characteristics, such as higher biomass use in less industrialized provinces that also exhibit lower overall emission levels. In contrast, the regression framework isolates the marginal impact of biomass energy while controlling for fossil fuel consumption, energy intensity, green technology innovation, and net forest change. Once these confounding factors are accounted for, biomass energy—largely representing traditional biomass used in rural households—exerts a positive and statistically significant effect on emissions due to inefficient combustion and methane release. This finding underscores that, without technological upgrading, biomass reliance can contribute to higher greenhouse gas emissions despite its renewable classification.

Conversely, green technology innovation (GTI) significantly reduces emissions, supporting the ecological modernization hypothesis. China’s commitment to environmental patents, electric mobility, and clean tech development has been a cornerstone of its low-carbon transition.

The estimated negative effect of net forest change highlights the climate-mitigation role of China’s land-use policies. While deforestation is typically associated with higher emissions, China’s experience differs due to sustained afforestation and ecological restoration initiatives such as the Grain-for-Green Program and the Natural Forest Conservation Program. As a result, forest cover expansion has strengthened carbon sequestration and contributed to emission reductions. This finding aligns with environmental science and confirms that land-use governance can serve as a complementary pillar to energy and technology-based decarbonization strategies.

Robustness checks using AMG and CCEMG estimators confirm the consistency of the main results, particularly the emissions-reducing effect of green innovation and the emissions-increasing role of fossil fuels and energy intensity.

To provide empirical support for regional differences, descriptive evidence on key drivers is presented across eastern, central, and western provinces. Eastern coastal provinces such as Guangdong and Jiangsu exhibit higher green technology innovation and lower reliance on traditional biomass, whereas several western provinces still depend on biomass and have high energy intensity. Figure 4 further shows that eastern regions have achieved more stable emissions trajectories, while central and western provinces demonstrate steeper increases in GHG emissions over time. These outcomes support the argument that regional development structure and technological capabilities significantly influence decarbonization progress in China.

Figure 4

Figure 4. Average provincial GHG emission trends by region (2000–2022)

To further support the discussion of regional variability, we include descriptive evidence comparing emissions and key drivers across eastern, central, and western China. The trends indicate that eastern coastal regions are beginning to stabilize emissions, whereas central and western provinces continue to experience strong upward trajectories due to higher dependence on fossil fuels and traditional biomass.

The chart shows diverging emission trajectories across regions, with eastern provinces experiencing slower growth in GHG emissions relative to central and western regions, reflecting structural and technological differences.

These descriptive statistics and trends empirically support the interpretation that coastal and innovation-driven regions are achieving relatively better progress in emission control compared with inland provinces.

Placing the findings in an international context highlight both similarities and differences between China and other emerging economies such as India and ASEAN countries. Like China, India faces the dual challenge of rapid industrialization and heavy reliance on coal-based energy. In many ASEAN countries, traditional biomass remains a major part of the rural energy mix, aligning with our finding that biomass energy (BIO) increases emissions. China’s stronger performance in green innovation sets it apart, as significant investments in clean energy technologies and environmental patents have yielded measurable emission reductions. Additionally, large-scale afforestation programs in China have helped turn deforestation into a net sink effect, unlike many ASEAN economies and India where forest loss remains a key environmental issue.

The negative coefficient of GTI supports Ecological Modernization Theory (EMT), demonstrating that technological advancements and environmental governance can decouple economic growth from environmental degradation. Meanwhile, the positive effects of fossil fuels and energy intensity are consistent with the early stage of the Environmental Kuznets Curve (EKC), where industrial expansion and energy demand heighten environmental pressure. The role of afforestation policies indicates a shift toward strong decoupling, demonstrating how targeted governance reforms can reverse environmental decline. The findings collectively suggest that China is transitioning from a carbon-intensive development path toward more sustainable systems, but progress is uneven across regions.

6 Conclusion and policy recommendations

6.1 Conclusion

This study investigated the key drivers of greenhouse gas (GHG) emissions in China by integrating energy structure, technological innovation, and land-use dynamics using second-generation econometric methods (CS-ARDL, AMG, and CCEMG). The findings indicate that China’s current growth model remains strongly linked to carbon-intensive energy use, while innovation and ecological restoration create important mitigation channels. Traditional biomass consumption still contributes to emissions due to inefficient rural combustion practices, highlighting the need for modernization rather than simple expansion of biomass use. Meanwhile, green technology innovation and afforestation efforts are already delivering measurable environmental benefits, demonstrating progress toward China’s dual-carbon goals.

These results reinforce two broader insights. First, decarbonization in China is transitioning from a coal-dominated system toward cleaner and more efficient energy structures, although progress is uneven across regions. Second, strengthening innovation capabilities and forest-based carbon sinks offers a credible pathway toward sustainable low-carbon development. By integrating energy, technology, and land-use dynamics, the study contributes to a more holistic understanding of emission drivers and provides actionable guidance for China’s long-term climate strategy.

6.2 Policy recommendations

Based on the empirical evidence and estimated elasticities, the following policy directions are proposed.

1. Accelerate the clean energy transition by reducing coal dependence and expanding low-carbon technologies across industrial sectors. The estimated fossil fuel elasticity (0.221) implies that a 10% reduction in fossil fuel consumption could lower greenhouse gas emissions by approximately 2.2%, making coal substitution one of the most effective mitigation strategies.

2. Enhance energy efficiency through industrial upgrading, stricter efficiency standards, and investment in smart and digitalized manufacturing. Given the energy intensity elasticity of 0.176, a 10% improvement in energy efficiency could reduce emissions by nearly 1.8%, highlighting efficiency gains as a cost-effective emissions reduction pathway.

3. Modernize rural biomass systems by replacing traditional combustion methods with clean bioenergy technologies. The positive biomass elasticity (0.084) indicates that a 10% increase in traditional biomass use raises emissions by about 0.8%, underscoring the need to prioritize technological upgrading rather than indiscriminate biomass expansion.

4. Strengthen green innovation ecosystems by expanding targeted R&D subsidies, carbon market incentives, and green finance instruments. The green technology innovation elasticity (−0.143) suggests that a 10% increase in green innovation activity could reduce emissions by approximately 1.4%, confirming innovation as a critical lever for long-term decarbonization.

5. Prioritize forest conservation and restoration to reinforce carbon sink capacity. The negative effect of net forest change demonstrates that continued afforestation and sustainable land management can meaningfully offset emissions, complementing energy and technology-based mitigation policies.

6.3 Future research directions

Further research could differentiate between traditional and modern biomass technologies, integrate spatial spillovers and institutional factors, and examine sector-specific mitigation pathways within China. Such extensions would deepen the regional relevance of climate policies and improve targeted designs for sustainable development.

While this study provides valuable insights into the drivers of greenhouse gas (GHG) emissions in China, several limitations should be acknowledged. First, the analysis is based on secondary data, which may be subject to measurement errors or inconsistencies, especially in variables such as biomass energy consumption and deforestation rates. Future studies could incorporate high-resolution satellite data and field-based measurements to improve data accuracy. Second, the study focuses primarily on energy structure, innovation, and land-use dynamics, while other potentially important factors such as institutional quality, environmental policy stringency, and financial mechanisms were not included due to data constraints. Expanding the model to incorporate these variables would provide a more holistic understanding of emission drivers.

Data availability statement

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

Author contributions

YC: Writing – original draft. RA: Writing – original draft.

Funding

The author(s) declared that financial support was not received for this work and/or its publication.

Acknowledgements

The authors acknowledge the use of ChatGPT solely for language editing and grammar improvement.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

References

Al-qazzaz, A., Eidgah, E. E. F., Alfatlawi, A. W., Masroori, A., Abed, A. M., Ajam, H., et al. (2024). An approach of analyzing gas and biomass combustion: positioned of flame stability and pollutant reduction. Results Eng. 23, 102823. doi:10.1016/J.RINENG.2024.102823

CrossRef Full Text | Google Scholar

Balanay, R., and Halog, A. (2024). Bioenergy updates and prospects for decarbonization in the ASEAN region: a review on logistical concerns and potential solutions. Wiley Interdiscip. Rev. Energy Environ. 13 (1), e499. doi:10.1002/WENE.499

CrossRef Full Text | Google Scholar

Bergendahl, J. A., Sarkis, J., and Timko, M. T. (2018). Transdisciplinarity and the food energy and water nexus: ecological modernization and supply chain sustainability perspectives. Resour. Conservation Recycl. 133, 309–319. doi:10.1016/J.RESCONREC.2018.01.001

CrossRef Full Text | Google Scholar

Bettarelli, L., Furceri, D., Pizzuto, P., and Shakoor, N. (2023). Environmental policies and innovation in renewable energy. IMF Work. Pap. 2023, 180. doi:10.5089/9798400251658.001.A001

CrossRef Full Text | Google Scholar

Chang, K., Liu, L., Luo, D., and Xing, K. (2023). The impact of green technology innovation on carbon dioxide emissions: the role of local environmental regulations. J. Environ. Manag. 340, 117990. doi:10.1016/J.JENVMAN.2023.117990

PubMed Abstract | CrossRef Full Text | Google Scholar

Chen, J., Li, C., Ristovski, Z., Milic, A., Gu, Y., Islam, M. S., et al. (2017). A review of biomass burning: emissions and impacts on air quality, health and climate in China. Sci. Total Environ. 579, 1000–1034. doi:10.1016/J.SCITOTENV.2016.11.025

PubMed Abstract | CrossRef Full Text | Google Scholar

Chen, Y., Yao, Z., and Zhong, K. (2022). Do environmental regulations of carbon emissions and air pollution foster green technology innovation: evidence from China’s prefecture-level cities. J. Clean. Prod. 350, 131537. doi:10.1016/J.JCLEPRO.2022.131537

CrossRef Full Text | Google Scholar

Chen, Y., Wang, Y., Zhou, D., and Luo, X. (2025). Regression-based predictive modeling of summer urban microclimate: quantifying contributions from urban design and urban heat emissions. Urban Clim. 62, 102550. doi:10.1016/J.UCLIM.2025.102550

CrossRef Full Text | Google Scholar

Chen, Z., Zhao, X., Zeng, S., Wang, H., Ge, L., Zhang, J., et al. (2025). Dual inhibition of As(III) mobilization and greenhouse gas emission achieved by photocatalytic birnessite-nitrate amendments in flooded paddy soils. J. Environ. Chem. Eng. 13 (5), 118011. doi:10.1016/J.JECE.2025.118011

CrossRef Full Text | Google Scholar

Chiriacò, M. V., Galli, N., Santini, M., and Rulli, M. C. (2024). Deforestation and greenhouse gas emissions could arise when replacing palm oil with other vegetable oils. Sci. Total Environ. 914, 169486. doi:10.1016/J.SCITOTENV.2023.169486

PubMed Abstract | CrossRef Full Text | Google Scholar

Chowdhury, P., Mahi, N. A., Yeassin, R., Chowdhury, N. U. R., and Farrok, O. (2025). Biomass to biofuel: impacts and mitigation of environmental, health, and socioeconomic challenges. Energy Convers. Manag. X 25, 100889. doi:10.1016/J.ECMX.2025.100889

CrossRef Full Text | Google Scholar

Chudik, A., and Pesaran, M. H. (2015). Common correlated effects estimation of heterogeneous dynamic panel data models with weakly exogenous regressors. J. Econ. 188 (2), 393–420. doi:10.1016/J.JECONOM.2015.03.007

CrossRef Full Text | Google Scholar

Cutcu, I., Cil, D., Karis, C., and Kocak, S. (2024). Determining the green technology innovation accelator and natural resources towards decarbonization for the EU countries: evidence from MMQR. Environ. Sci. Pollut. Res. Int. 31 (12), 19002–19021. doi:10.1007/S11356-024-32302-4

PubMed Abstract | CrossRef Full Text | Google Scholar

Duan, N. (2025). The quest for balance: the journey of energy, environment and sustainable development. Energy and Environ. Sustain. 1 (1), 100001. doi:10.1016/J.EESUS.2024.12.001

CrossRef Full Text | Google Scholar

Filippini, M., and Hunt, L. C. (2015). Measurement of energy efficiency based on economic foundations. Energy Econ. 52, S5–S16. doi:10.1016/J.ENECO.2015.08.023

CrossRef Full Text | Google Scholar

Gupta, K., Saha, A., and Sen Gupta, B. (2024). Spatio-temporal distribution of pollutant trace gases (CO, CH4, O3 and NO2) in India: an observational study. Geol. Ecol. Landscapes 8 (3), 306–326. doi:10.1080/24749508.2022.2132706

CrossRef Full Text | Google Scholar

Hashem Pesaran, M., and Yamagata, T. (2008). Testing slope homogeneity in large panels. J. Econ. 142 (1), 50–93. doi:10.1016/J.JECONOM.2007.05.010

CrossRef Full Text | Google Scholar

He, W., Di, B., Zeng, Y., Duan, Y., Li, J., Qiu, L., et al. (2023). Reconsidering the eco-economic benefits of grain for green program in Sichuan Province, China. Ecol. Indic. 157, 111225. doi:10.1016/J.ECOLIND.2023.111225

CrossRef Full Text | Google Scholar

Hoa, P. X., Xuan, V. N., and Thu, N. T. P. (2024). Factors affecting carbon dioxide emissions for sustainable development goals – new insights into six Asian developed countries. Heliyon 10 (21), e39943. doi:10.1016/J.HELIYON.2024.E39943

PubMed Abstract | CrossRef Full Text | Google Scholar

Ibrahiem, D. M., and Hanafy, S. A. (2020). Dynamic linkages amongst ecological footprints, fossil fuel energy consumption and globalization: an empirical analysis. Manag. Environ. Qual. An Int. J. 31 (6), 1549–1568. doi:10.1108/MEQ-02-2020-0029/FULL/HTML

CrossRef Full Text | Google Scholar

IEA (2021). Global energy review 2021 – analysis - IEA. Paris, France: International Energy Agency.

Google Scholar

IEA (2023). World energy outlook 2023 – analysis - IEA. Available online at: https://www.iea.org/reports/world-energy-outlook-2023.

Google Scholar

Kashif, M., Ullah, A., Ullah, S., and Qian, N. (2024). Towards a greener future: the impact of financial technology (FinTech) and climate finance on ecological sustainability. J. Environ. Manag. 370, 122876. doi:10.1016/J.JENVMAN.2024.122876

PubMed Abstract | CrossRef Full Text | Google Scholar

Kula, E. (2010). Afforestation with carbon sequestration and land use policy in Northern Ireland. Land Use Policy 27 (3), 749–752. doi:10.1016/J.LANDUSEPOL.2009.10.004

CrossRef Full Text | Google Scholar

Lanjekar, P. R., Panwar, N. L., Agrawal, C., Gupta, T., Meena, K. K., Meena, G. L., et al. (2024). A review on environmental assessment of thermochemical conversion technologies for energy generation from crop residue. Earth Energy Sci. 1, 350–364. doi:10.1016/J.EES.2024.07.001

CrossRef Full Text | Google Scholar

Li, W., Yu, X., Hu, N., Huang, F., Wang, J., and Peng, Q. (2022). Study on the relationship between fossil energy consumption and carbon emission in Sichuan Province. Energy Rep. 8, 53–62. doi:10.1016/J.EGYR.2022.01.112

CrossRef Full Text | Google Scholar

Li, Y., Brando, P. M., Morton, D. C., Lawrence, D. M., Yang, H., and Randerson, J. T. (2022). Deforestation-induced climate change reduces carbon storage in remaining tropical forests. Nat. Commun. 13 (1), 1–13. doi:10.1038/s41467-022-29601-0

PubMed Abstract | CrossRef Full Text | Google Scholar

Liu, S., Padhan, H. P. J., Jose, A., and Rahut, D. (2024). Do green trade and technology-oriented trade affect economic cycles? Evidence from the Chinese provinces. Technol. Forecast. Soc. Change 202, 123334. doi:10.1016/J.TECHFORE.2024.123334

CrossRef Full Text | Google Scholar

Lynch, J., Cain, M., Pierrehumbert, R., and Allen, M. (2020). Demonstrating GWP: a means of reporting warming-equivalent emissions that captures the contrasting impacts of short- A nd long-lived climate pollutants. Environ. Res. Lett. 15 (4), 044023. doi:10.1088/1748-9326/AB6D7E

PubMed Abstract | CrossRef Full Text | Google Scholar

Martins, T., Barreto, A. C., Souza, F. M., and Souza, A. M. (2021). Fossil fuels consumption and carbon dioxide emissions in G7 countries: empirical evidence from ARDL bounds testing approach. Environ. Pollut. 291, 118093. doi:10.1016/J.ENVPOL.2021.118093

PubMed Abstract | CrossRef Full Text | Google Scholar

Mehmood, K., Tauseef Hassan, S., Qiu, X., and Ali, S. (2024). Comparative analysis of CO2 emissions and economic performance in the United States and China: navigating sustainable development in the climate change era. Geosci. Front. 15 (5), 101843. doi:10.1016/J.GSF.2024.101843

CrossRef Full Text | Google Scholar

Meng, Q., Xu, J., Luo, F., Jin, X., Xu, L., Yao, W., et al. (2023). Collaborative and effective scheduling of integrated energy systems with consideration of carbon restrictions. IET Generation, Transm. and Distribution 17 (18), 4134–4145. doi:10.1049/GTD2.12971

CrossRef Full Text | Google Scholar

Meo, M. S., and Adebayo, T. S. (2025). Role of financial regulations and climate policy uncertainty in reducing CO2 emissions—an application of bootstrap subsample rolling-window granger causality. Clean Technol. Environ. Policy 27 (6), 2377–2393. doi:10.1007/S10098-024-02991-Z/FIGURES/7

CrossRef Full Text | Google Scholar

Meo, M. S., and Anees, A. (2025). Tech collaboration and digital inclusion: reshaping global trade in the age of climate policy. Sustain. Futur. 10, 100890. doi:10.1016/J.SFTR.2025.100890

CrossRef Full Text | Google Scholar

Miao, C., Chen, Z., and Zhang, A. (2024). Green technology innovation and carbon emission efficiency: the moderating role of environmental uncertainty. Sci. Total Environ. 938, 173551. doi:10.1016/J.SCITOTENV.2024.173551

PubMed Abstract | CrossRef Full Text | Google Scholar

Pesaran, M. H. (2004). General diagnostic tests for cross section dependence in panels. SSRN Electron. J. doi:10.2139/SSRN.572504

CrossRef Full Text | Google Scholar

Pesaran, M. H. (2006). Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica 74 (4), 967–1012. doi:10.1111/J.1468-0262.2006.00692.X

CrossRef Full Text | Google Scholar

Pesaran, M. H. (2007). A simple panel unit root test in the presence of cross-section dependence. J. Appl. Econ. 22 (2), 265–312. doi:10.1002/JAE.951

CrossRef Full Text | Google Scholar

Raza, M. Y., and Lin, B. (2025). The impact of the productive sectors on CO2 emissions in Pakistan. Environ. Impact Assess. Rev. 110, 107643. doi:10.1016/J.EIAR.2024.107643

CrossRef Full Text | Google Scholar

Shindell, D., Kuylenstierna, J. C. I., Vignati, E., Van Dingenen, R., Amann, M., Klimont, Z., et al. (2012). Simultaneously mitigating near-term climate change and improving human health and food security. Science 335 (6065), 183–189. doi:10.1126/SCIENCE.1210026

PubMed Abstract | CrossRef Full Text | Google Scholar

Sun, X., Dong, Y., Shafiq, M. N., Gago-de Santos, P., and Gillani, S. (2025). Economic policy uncertainty and environmental quality: unveiling the moderating effect of green finance on sustainable environmental outcomes. Humanit. Soc. Sci. Commun. 12 (1), 1–12. doi:10.1057/S41599-025-05212-0

CrossRef Full Text | Google Scholar

Tian, G., and Khan, I. (2025). Impact of energy end-uses and efficiency indicators on the environmental policy stringency index. Energy 326, 136232. doi:10.1016/J.ENERGY.2025.136232

CrossRef Full Text | Google Scholar

Wang, Q., and Yang, X. (2024). Towards carbon neutrality: the role of innovation and resources efficiency in China’s economic transformation. Resour. Policy 91, 104911. doi:10.1016/J.RESOURPOL.2024.104911

CrossRef Full Text | Google Scholar

Wang, C., Raza, S. A., Adebayo, T. S., Yi, S., and Shah, M. I. (2023). The roles of hydro, nuclear and biomass energy towards carbon neutrality target in China: a policy-based analysis. Energy 262, 125303. doi:10.1016/J.ENERGY.2022.125303

CrossRef Full Text | Google Scholar

Westerlund, J. (2007). Testing for error correction in panel data. Oxf. Bull. Econ. Statistics 69 (6), 709–748. doi:10.1111/J.1468-0084.2007.00477.X

CrossRef Full Text | Google Scholar

Yu, L., Liu, Y., Shen, M., Yu, Z., Li, X., Liu, H., et al. (2025). Extreme hydroclimates amplify the biophysical effects of advanced green-up in temperate China. Agric. For. Meteorology 363, 110421. doi:10.1016/J.AGRFORMET.2025.110421

CrossRef Full Text | Google Scholar

Zanjani, Z., Soares, I., and Macedo, P. (2022). The nexus between CO2 emissions from electricity generation, GDP and energy intensity using a complete maximum entropy approach: the case of Iran. Energy Rep. 8, 319–324. doi:10.1016/J.EGYR.2022.01.058

CrossRef Full Text | Google Scholar

Zhang, T., Song, B., Han, G., Zhao, H., Hu, Q., Zhao, Y., et al. (2023). Effects of coastal wetland reclamation on soil organic carbon, total nitrogen, and total phosphorus in China: a meta-analysis. Land Degrad. and Dev. 34 (11), 3340–3349. doi:10.1002/LDR.4687

CrossRef Full Text | Google Scholar

Zhao, X., Chen, Y., Hu, J., Wang, H., Ye, Z., Zhang, J., et al. (2025). Efficacy of nitrate and biochar@birnessite composite microspheres for simultaneous suppression of As(III) mobilization and greenhouse gas emissions in flooded paddy soils. Environ. Res. 279, 121757. doi:10.1016/J.ENVRES.2025.121757

PubMed Abstract | CrossRef Full Text | Google Scholar

Appendix

TABLE A1

TABLE A1. Diagnostic statistics for CS-ARDL model.