Abstract
Balancing economic growth with emission reductions presents a pivotal challenge under China’s "dual-carbon" strategy. Hunan Province, characterized by its diverse industrial and ecological contexts, offers a quintessential setting for county-level analysis. This study investigates 122 counties within Hunan from 2000 to 2022, employing the Tapio decoupling model, spatial autocorrelation, ordinary least squares (OLS) regression, an augmented STIRPAT framework, and scenario simulations. It analyzes the spatiotemporal dynamics between carbon emissions and GDP, elucidates the principal driving mechanisms, and forecasts peaking trajectories under scenarios of low, medium, and high carbon emissions. The findings reveal a continued rise in emissions, albeit at a reduced pace post-2010, with the proportion of counties achieving strong decoupling escalating from 6% during 2006–2010 to 53% in 2016–2022. Counties dominated by eco-cultural (80%) and agricultural activities (67%) demonstrated the most significant transitions. Regression analyses, based on standardized coefficients, underscore the positive influence of carbon intensity (0.917), economic development (0.792), and population (0.447) on emissions, whereas industrial upgrading (-0.014), energy efficiency (-0.046), and urbanization (-0.010) contribute to mitigating impacts. GeoDetector analysis corroborates strong interaction effects. Scenario simulations indicate that only the low-carbon pathway might enable all four economic regions to achieve emission peaks between 2030 and 2035, while medium- and high-carbon trajectories fail to achieve effective peaking before 2040. These results highlight the imperative for tailored low-carbon strategies at the regional level, incorporating industrial restructuring, enhancement of energy efficiency, promotion of ecological-based industries, and implementation of county-level monitoring and market incentives.
Keywords: Carbon decoupling, Carbon neutrality, Low-carbon development, Hunan province
Introduction
The contemporary global climate scenario is deteriorating, characterized by frequent and severe extreme weather events that pose substantial threats to both ecosystems and human societies [1]. In response to this escalating climate crisis, numerous countries have set forth objectives for achieving carbon peak and carbon neutrality to facilitate transitions to low-carbon energy and foster sustainable development [2–5]. Nevertheless, the synchronization of carbon reduction with economic growth continues to be a significant challenge [6]. Major obstacles include elevated costs associated with mitigation [7], uncertainties in technological pathways [8], and the dual burden that developing nations face in aligning economic development with effective emission reductions [9]. As a crucial marker for evaluating the quality of low-carbon development, the notion of carbon decoupling has begun to be recognized in some developed nations [10]. However, in many developing regions, carbon emissions persistently rise in conjunction with economic growth, thereby underscoring the complexity of decarbonization strategies and the marked regional disparities [11]. As the largest developing country in the world, China officially declared its "dual-carbon" objectives in 2020, which aim to cap carbon dioxide emissions by 2030 and achieve carbon neutrality by 2060 [12].
Carbon decoupling serves as a critical indicator for assessing the alignment between economic growth and environmental sustainability [13]. Over the period from 2000 to 2020, China’s carbon emissions increased at an average annual rate of approximately 5.4%, while its gross domestic product (GDP) grew by 9.5% annually, signifying an initial transition towards a more sustainable growth model. Some regions have progressively achieved weak decoupling, and instances of short-term strong decoupling have been observed in several economically advanced cities [14]. Within this context, the Tapio model based on elasticity and the OECD classifications of strong and weak decoupling have gained widespread acceptance for their simplicity and robust comparability [15]. Wang et al. [16] utilized the Tapio approach to scrutinize the relationship between municipal solid waste generation and ED across 285 Chinese cities. Xi et al. [17] investigated the decoupling dynamics between industrial growth and pollution within China’s Circum-Bohai Sea Region. Recent studies have further expanded the application of decoupling analysis across diverse sectors and scales. Luo et al. [18] developed a supply-side analytical framework to examine digitalization’s role in decoupling China’s tourism economy from carbon emissions, demonstrating that digital-enabled tourism value-added increased approximately 18 times from 2002 to 2017 while emissions only grew about 11 times, highlighting the importance of examining decoupling patterns at finer spatial scales. The STIRPAT model is predominantly employed to explore the drivers of carbon emissions at a regional level, providing a quantitative foundation for devising targeted mitigation strategies [19, 20].
China’s administrative hierarchy encompasses national, provincial, prefectural, and county levels. Existing scholarship has predominantly concentrated on the national, provincial, or prefectural scales [21, 22]. However, as the foundational governance unit within China’s administrative system, counties play an indispensable role in implementing regional development strategies, resource allocation, and policy execution. Achieving national carbon neutrality goals critically depends on coordinated county-level actions, where localized emission patterns, industrial compositions, and development stages vary substantially. Unlike provincial or prefecture-level studies that often mask internal heterogeneity, county-level analysis enables the identification of specific areas where effective emission reductions have not yet been achieved and facilitates the design of targeted mitigation strategies. Counties represent the primary arena where national policies are translated into concrete actions, and where the trade-offs between economic development and environmental protection are most directly experienced by local governments and residents. In Hunan Province specifically, county-level units exhibit remarkable diversity ranging from resource-based economies reliant on coal mining in Western Hunan to eco-cultural regions dominated by tourism, and from heavy-industry-based counties in the Chang-Zhu-Tan urban agglomeration to agriculture-dominated areas in the Dongting Lake region. This diversity makes Hunan an ideal setting for examining how different development pathways influence decoupling trajectories. Recent studies have begun to investigate county-level carbon emissions, for instance linking carbon-emission intensity to construction-land-use intensity at the county scale [23]. Moreover, recent studies have shown that land-use change and spatial carbon balance play a critical role in shaping regional carbon emission dynamics and carbon sink-source interactions, highlighting the need to integrate land use considerations into subnational decarbonization research. For example, research on land-use carbon emissions and carbon balance zoning in China’s economic regions has demonstrated pronounced spatial heterogeneity arising from land transformation processes and network interactions between socioeconomic and environmental factors [24, 25]. Liu et al. [26] employed Exploratory Spatiotemporal Data Analysis to delineate the spatiotemporal dynamics of carbon emissions across Chinese counties. Liu et al. [27] calculated land-use carbon emissions of Chinese counties and employed GeoDetector to analyze spatial–temporal patterns and driving factors. Huang et al. [28] further investigated the spatial network structure and carbon balance zoning of land-use carbon emissions using land-use and socio-economic data, revealing pronounced spatial heterogeneity, core–periphery structures, and interregional spillover effects at the subnational scale. Beyond land-use composition, urban spatial structure has likewise been shown to shape the spatiotemporal pattern of carbon emissions [29], reinforcing the need to account for spatial configuration when analyzing emissions at subnational scales. Scenario simulation methods have been increasingly utilized to forecast carbon peak trajectories at the provincial scale [30–32]. At the city scale, related studies have further decomposed the driving factors of emissions and predicted their peak trajectories [33]. Despite these advances, the application of such models at the county level remains limited.
Although land-use carbon emissions and carbon balance zoning have been increasingly studied at regional and provincial scales [24, 25] few studies have systematically examined how county-level decoupling patterns are shaped by land transformation and carbon flux dynamics, especially in central and western China. This gap underscores the need for finer-scale analyses that integrate land use, socioeconomic drivers, and decoupling mechanisms. First, a systematic examination of carbon decoupling at the county level, particularly focusing on regional heterogeneity and stage-based evolution, is conspicuously absent. This deficiency is particularly pronounced in Hunan Province, where a detailed identification of areas with ineffective emission reductions is still missing. Second, previous studies have predominantly concentrated on single-factor drivers or static regression analyses, paying scant attention to the interactive effects and cumulative impacts of multiple factors on county-level carbon emission dynamics. Third, there is a notable dearth of scenario-based simulation studies that integrate regional strategies and localized parameter adjustments for differentiated carbon peaking pathways.
This study focuses on 122 county-level administrative units in Hunan Province, aiming to systematically assess the decoupling relationship between carbon emissions and economic growth, analyze its spatiotemporal evolution patterns and principal driving mechanisms, and construct three developmental scenarios (low-carbon, medium-carbon, and high-carbon) to simulate future regional carbon peaking trajectories. In doing so, it makes four contributions that distinguish it from existing work: (1) Whereas carbon-decoupling research has been concentrated at the national, provincial, or prefectural scale, where aggregate indicators average over the internal heterogeneity that determines where mitigation actually stalls, this study conducts a continuous county-scale diagnosis of 122 units over a 22-year period, resolving the specific locations and stages at which growth remains coupled to emissions; (2) In contrast to studies that apply decoupling diagnosis, driver attribution, or scenario forecasting in isolation, it integrates the Tapio model, spatial autocorrelation, an augmented STIRPAT/OLS specification, GeoDetector, and scenario simulation into a single diagnosis-to-projection framework, linking the explanation of why decoupling occurs with the projection of when peaking can be achieved, and using GeoDetector to capture the interactive rather than single-factor nature of emission drivers; (3) Unlike studies that treat counties as a homogeneous group, it classifies them into six functional types and four economic regions, enabling differentiated decoupling trajectories and their drivers to be identified by type of place rather than by administrative unit alone; and (4) Beyond a single-province case, its scenario design is anchored in Hunan’s official planning documents to ensure policy relevance, while the framework itself is transferable to other provinces and developing economies marked by strong sub-national heterogeneity, where aggregate national targets risk masking local constraints.
Method and data
This study examines 122 county-level administrative units in Hunan Province, employing exploratory spatial analysis to investigate the decoupling dynamics between ED and carbon emissions within the context of China’s dual-carbon strategy. The objective is to identify the key factors influencing emission changes and to model potential pathways toward carbon peaking and neutrality across various developmental scenarios. The research adopts an integrated methodological framework that includes the Tapio decoupling model, spatial autocorrelation techniques, Ordinary Least Squares (OLS) regression, and an expanded STIRPAT model. The Tapio model, an elasticity-based decoupling metric, is widely used to assess the coordination between environmental pressures and economic growth due to its simplicity and cross-regional comparability [17]. The STIRPAT framework, a flexible stochastic extension of the IPAT identity, facilitates the quantification of the contributions of socioeconomic and technological factors to carbon emissions [20]. The GeoDetector tool is employed to detect spatial stratified heterogeneity and factor interactions, which is suitable for analysing nonlinear and spatially clustered emission systems [34] (Fig. 1).
Fig. 1.
Research framework
Method
Calculation of influential factor indicators
(1) Carbon emission intensity (CEI)
CEI serves as a crucial metric for evaluating the carbon efficiency of ED, reflecting the volume of carbon dioxide emitted per unit of economic output. In this study, CEI is computed using the ratio of total carbon emissions to GDP, directly assessing the relationship between environmental pressure and economic performance at the county level [35]. The formula for CEI is given by:
 |
1 |
where 
represents the CEI of county i in year t, 
denotes the total carbon emissions from county i in year t, and 
represents the GDP of county i in year t. All GDP data utilised in this research are adjusted to constant 2000 prices to counteract the effects of inflation and ensure comparability over time. Carbon emission data are sourced from the EDGAR dataset and aligned with county-level administrative units.
(2) Industrial development level
The level of industrial development in this analysis is quantified using the Industrial Structure Upgrading Index (ISU), which reflects the transition from primary and secondary industries to more advanced, service-oriented sectors. This index not only captures the qualitative shift but also quantifies the optimization of the IS, serving as a marker of the region’s economic modernization. The formula used is:
 |
2 |
Here, 
represents the ISU for region i in year t, and 
denotes the proportion of the added value of the j-th industry to total GDP, where j=1 for primary, j=2 for secondary, and j=3 for tertiary industries.
(3) Energy consumption intensity (ECI)
Owing to constraints in data availability at the county level, especially the scarcity of comprehensive statistics on total energy consumption quantified in standard coal equivalent (SCE), this study employs electricity consumption per unit of GDP as a proxy for ECI. This approach reflects the efficiency of regional energy use, particularly in areas where electricity is the predominant source of energy for both industrial and residential activities. The calculation is conducted as follows:
 |
3 |
where 
represents the ECI of county i in year t, 
denotes the total electricity consumption of county i in year t, and 
is the GDP of county i in year t.
Spatial autocorrelation
To investigate the spatial agglomeration and heterogeneity characteristics of carbon emissions, the study applies spatial autocorrelation analysis to quantitatively assess the spatial distribution patterns of county-level carbon emissions. This analysis includes two principal metrics: Global Moran’s I and Local Moran’s I (LISA), which together reveal both overarching and localised spatial dependencies. Global Moran’s I determines whether carbon emissions exhibit significant spatial clustering at a regional scale [36]. The formula is presented as follows:
 |
4 |
In this equation, n represents the number of samples, 
and 
are the carbon emission values of counties i and j, respectively, and 
is the mean of the sample values. The element 
in the spatial weight matrix, with the study utilising an adjacency matrix to define spatial relationships, is elucidated in:
 |
5 |
Carbon emission decoupling index
This research employs the Tapio decoupling model to explore the relationship between carbon emissions and economic growth in the counties of Hunan Province [3, 37]. The carbon emission decoupling index determines whether the increase in carbon emissions is proportional to economic growth or if there is evidence of decoupling, indicative of progress towards low-carbon development. The formula for this index is as follows:
 |
6 |
where 
indicates the carbon emission decoupling index for county i in period t, 
denotes the growth rate of carbon emissions between periods t0 and t1, and 
indicates the growth rate of GDP in the same timeframe. The detailed computation of ΔC and ΔGDP is provided in:
 |
7 |
 |
8 |
Based on the criteria outlined in Table 1, the study categorises the various types of CO2 decoupling statuses.
Table 1.
CO2 decoupling statuses
| ΔC | ΔGDP | Decoupling index (D) | Status | Implication |
|---|---|---|---|---|
| < 0 | > 0 | D < 0 | Strong decoupling | GDP increases while carbon emissions decrease |
| > 0 | > 0 | 0 < D < 0.8 | Weak decoupling | GDP increases, but carbon emissions grow at a slower rate |
| > 0 | > 0 | 0.8 ≤ D ≤ 1.2 | Expansive coupling | GDP and carbon emissions grow at comparable rates |
| > 0 | > 0 | D > 1.2 | Expansive negative decoupling | Carbon emissions grow significantly faster than GDP |
| > 0 | < 0 | D < 0 | Strong negative decoupling | GDP declines while carbon emissions increase |
| < 0 | < 0 | 0 < D < 0.8 | Weak negative decoupling | Both GDP and carbon emissions decline, but emissions decrease less than GDP |
| < 0 | < 0 | 0.8 ≤ D ≤ 1.2 | Recessive coupling | GDP and carbon emissions decline at comparable rates |
| < 0 | < 0 | D > 1.2 | Recessive Decoupling | GDP declines, and carbon emissions decrease even more rapidly |
STIRPAT model
To conduct a rigorous investigation into the potential pathways towards carbon neutrality for county-level administrative units in Hunan Province, and to scientifically ascertain the principal determinants and progressive trends of carbon emissions within the framework of the “dual-carbon” targets, this study utilises an augmented IPAT/STIRPAT model for regression analysis and scenario forecasting. The IPAT model, which is extensively employed to examine the relationship between environmental pressures, such as carbon emissions, energy consumption, and ecological footprints, and variables including P, ED, and technological advancement, plays a pivotal role in elucidating the mechanisms underlying environmental transformations [38]. By decomposing environmental impacts into the multiplicative interactions of P, A, and T, the IPAT model provides a cogent analytical framework and has been widely adopted in research concerning the drivers of carbon emissions, attribution analyses of shifts in energy consumption, assessments of sustainability, and evaluations of environmental policies. The classical expression of the IPAT model is depicted as follows:
 |
9 |
Here, I denotes environmental impact, exemplified by carbon emissions; P indicates the P size; A represents affluence, typically quantified by per capita GDP; and T denotes technological influences, commonly measured in terms of energy intensity or the CEI. Given the identity nature of the IPAT equation, which precludes direct statistical estimation, this study utilises its stochastic regression-based extension, the STIRPAT model [39], articulated as follows:
 |
10 |
After logarithmic transformation, this model is employed for regression analysis:
 |
11 |
To enhance its applicability to county-level analysis and drawing upon pertinent literature, this study develops a multivariate extended STIRPAT model that integrates essential socioeconomic factors, including CEI, P, ED, IS, ECI, as well as UR. The model is formalised as follows:
 |
12 |
Variable selection is anchored in the STIRPAT framework, where population, affluence, and technology are recognised as fundamental determinants of environmental impacts [40–42]. In addition, variables such as industrial structure, urbanization, and energy intensity have been frequently documented to influence carbon emissions in regional studies [43–45]. Given the county-specific development context in Hunan, these variables also reflect critical policy levers under the “dual-carbon” strategy. Table 2 provides detailed descriptions of the six key socioeconomic factors identified as primary determinants of county-level carbon emissions in this study. The results from multicollinearity tests indicate that all variables display variance inflation factors (VIFs) below 10 and tolerance levels above 0.1, confirming that multicollinearity is not a significant concern in the model. Although the VIFs for ED and ECI are relatively high at 7.125 and 6.268 respectively, they remain within acceptable thresholds and are unlikely to significantly affect the reliability of the regression estimates. Consequently, this study employs OLS regression to estimate the extended STIRPAT model, thus ensuring both the robustness and interpretability of the results. OLS, a classical method for parameter estimation, seeks to minimise the sum of squared residuals between observed and predicted values, thereby providing the best linear unbiased estimates of the regression coefficients. Due to its favourable statistical properties, OLS is capable of yielding consistent and efficient results in the absence of severe multicollinearity. Since the selected variables adhere to the fundamental assumptions of linear regression and exhibit minimal multicollinearity, OLS is considered an appropriate methodology for elucidating the elastic effects of various factors on county-level carbon emissions. The general form of the OLS regression is presented below:
 |
13 |
where 
represents the total carbon emissions of the i-th county in the t-th year; 
denotes the k-th independent variable; 
signifies the parameter to be estimated; α is a constant term; 
is the random error term.
Table 2.
Variable description
| Variable | Description | Unit | VIF | Tolerance |
|---|---|---|---|---|
| Carbon emission intensity (CEI) | Per unit GDP Carbon emission | 10,000 tonness / 10,000 CNY | 1.373 | 0.729 |
| Population (P) | Total population at the end of the year | 10,000 people | 2.429 | 0.412 |
| Economic development (ED) | Per capita GDP | 10,000 CNY | 7.125 | 0.14 |
| Industrial Structure (IS) | Advanced Industrial Structure index | — | 1.663 | 0.601 |
| Energy consumption intensity (ECI) | The overall electricity consumption intensity of the entire society | kilowatt-hours / 10,000 CNY | 6.268 | 0.16 |
| Urbanization rate (UR) | Urbanization ratio | % | 3.013 | 0.332 |
Geo detector
To elucidate the predominant factors influencing county-level CEI and their interactive effects, this study employs the GeoDetector method, focusing particularly on the Interaction Detector module. Operating under the premise of spatial stratified heterogeneity, the GeoDetector method evaluates the explanatory power of both individual and combined factors on the spatial distribution of CEI. This is achieved by comparing the spatial variance of the dependent variable across different stratifications of the independent variables. This approach aids in identifying primary driving mechanisms and their synergistic impacts. The principal statistical measure utilised in the GeoDetector is the q-value, defined as follows:
 |
14 |
where h represents the h-th subregion or county category as determined by the stratification of an influential factor; Nh and 
signify the number of samples and the variance of the dependent variable within subregion h, respectively, while N and 
denote the total number of samples and the overall variance across the entire study area. The q-value, ranging from 0 to 1, indicates the strength of explanatory power.
Data sources, processing and limitations
This study employs county-level datasets comprising socioeconomic indicators, carbon emissions, and administrative boundaries for 122 counties in Hunan Province. The socioeconomic indicators include population size (P), gross domestic product (GDP), industrial structure (IS), electricity consumption, and urbanization rate (UR). These were sourced from the Hunan Statistical Yearbook (https://tjj.hunan.gov.cn/hntj/tjfx/hntjnj/hntjnjwlb/index.html), the China County Statistical Yearbook (https://www.stats.gov.cn/), and official statistical bulletins issued by relevant government departments. The data on carbon emissions were obtained from the Emission Database for Global Atmospheric Research (EDGAR) and then downscaled to the county level using spatial matching and weighted allocation procedures. Administrative boundary data were acquired from the most recent county-level vector shapefiles provided by the Geographic Information System (GIS) platform of Hunan Province. Additionally, to assist in inter-year calibration and ensure continuous temporal coverage, night-time light products from DMSP/OLS and NPP/VIIRS were utilised.
To maintain cross-year comparability, all datasets were harmonised to the 2020 administrative boundaries, and earlier socioeconomic records were re-aggregated accordingly. Annual GDP values were adjusted to constant 2000 prices using official deflators to mitigate the impacts of inflation. The night-time light products from DMSP/OLS and NPP/VIIRS underwent an inter-calibration process based on established methodologies to create a consistent long-term series. Procedures such as unit harmonisation, missing-value imputation, and cross-source reconciliation were implemented to enhance data integrity and consistency. From these processed datasets, six key variables were devised to characterise county-level development and emission patterns [26, 27, 46]: Carbon emission intensity, demographic scale, economic level, industrial upgrading, energy efficiency, and urban development. These variables were integrated into a unified spatial database to facilitate subsequent empirical analysis. This comprehensive dataset enables a detailed examination of regional disparities in emission trajectories and provides robust empirical support for formulating differentiated carbon peaking and neutrality strategies at the county level.
Several limitations of this study warrant acknowledgment. Firstly, the estimation of county-level carbon emissions is indirectly derived using night-time light proxies, potentially introducing uncertainties due to sensor saturation or inter-calibration discrepancies. Secondly, socioeconomic statistics at the county scale are occasionally incomplete, necessitating interpolation or cross-source integration. Thirdly, changes in administrative boundaries from 2000 to 2020 may impact temporal consistency, although this issue is mitigated by the harmonisation to the 2020 boundary. While these limitations do not compromise the identification of overall emission patterns, they may affect the accuracy of absolute estimates and the analysis of spatial variability.
Results
County-level carbon emission and GDP spatial–temporal evolution patterns in Hunan
Spatial–temporal evolution pattern of carbon emissions
Figure 2 illustrates the trends in total carbon emissions and their respective growth rates across the four principal economic regions of Hunan Province from 2000 to 2022. Throughout this period, a consistent upward trajectory in carbon emissions was noted in all regions, with particularly marked increases occurring between 2005 and 2012. The Chang-Zhu-Tan Region and the Dongting Lake Ecological Economic Region have been identified as the primary contributors, collectively accounting for a significant portion of the province’s total emissions. Conversely, the Southern Hunan and Western Hunan regions, whilst also exhibiting an upward trend, have registered comparatively lower levels of emissions. In terms of the growth rates of emissions, substantial fluctuations were observed in 2003 and 2006, with Western Hunan experiencing a peak growth rate in excess of 50% in 2006, subsequently followed by a gradual decline. After 2010, the growth rates across all regions have tended to stabilise, signifying a deceleration in growth and, in some instances, a transition towards emission stabilisation or reduction. The advent of the COVID-19 pandemic in 2020 precipitated a general reduction in growth rates across the regions; however, the Dongting Lake region experienced a temporary resurgence in 2021. In summary, while total carbon emissions in Hunan Province have continued to rise, the rate of increase has progressively moderated, reflecting a shift towards sustainable, high-quality, green development.
Fig. 2.
Trends in total carbon emissions and their growth rates at the county level in Hunan Province
The analysis of Moran’s I statistics for total carbon emissions across 122 county-level administrative units in Hunan Province, covering the period from 2000 to 2022, is presented in Table 3. These results reveal that carbon emissions at the county level exhibit a consistently positive spatial autocorrelation throughout the study period, though the overall magnitude is relatively modest, suggesting a lack of pronounced spatial clustering. Specifically, the annual Moran’s I values uniformly indicate positive autocorrelation, ranging from 0.067 to 0.126. This range indicates a slight tendency towards spatial clustering, albeit with limited intensity. In terms of statistical significance, only the years 2004 (Moran’s I = 0.123, Z = 2.064, P = 0.039) and 2005 (Moran’s I = 0.126, Z = 2.099, P = 0.036) recorded values below the 5% significance threshold, thereby indicating statistically significant spatial clustering during these years. For the remaining years within the study period, however, the P-values exceeded 0.05, suggesting that the spatial autocorrelation of carbon emissions did not achieve statistical significance. In conclusion, the spatial distribution of county-level carbon emissions in Hunan Province appears relatively dispersed, characterised by significant inter-regional variability and an absence of stable high- or low-value clusters. This observed distribution pattern is likely influenced by the province’s complex topography, the uneven pace of regional development, and diverse energy consumption structures.
Table 3.
Global Moran’s I
| Year | Moran’s I | Z-score | P-value | Year | Moran’s I | Z-score | P-value |
|---|---|---|---|---|---|---|---|
| 2000 | 0.073 | 1.321 | 0.186 | 2012 | 0.076 | 1.342 | 0.180 |
| 2001 | 0.067 | 1.205 | 0.228 | 2013 | 0.077 | 1.348 | 0.178 |
| 2002 | 0.070 | 1.285 | 0.199 | 2014 | 0.077 | 1.351 | 0.177 |
| 2003 | 0.080 | 1.408 | 0.159 | 2015 | 0.086 | 1.494 | 0.135 |
| 2004 | 0.123 | 2.064 | 0.039 | 2016 | 0.084 | 1.451 | 0.147 |
| 2005 | 0.126 | 2.099 | 0.036 | 2017 | 0.084 | 1.461 | 0.144 |
| 2006 | 0.075 | 1.327 | 0.185 | 2018 | 0.081 | 1.425 | 0.154 |
| 2007 | 0.083 | 1.447 | 0.148 | 2019 | 0.082 | 1.435 | 0.151 |
| 2008 | 0.091 | 1.556 | 0.120 | 2020 | 0.086 | 1.496 | 0.135 |
| 2009 | 0.091 | 1.558 | 0.119 | 2021 | 0.074 | 1.296 | 0.195 |
| 2010 | 0.092 | 1.574 | 0.116 | 2022 | 0.071 | 1.260 | 0.208 |
| 2011 | 0.083 | 1.444 | 0.149 |
Figure 3 presents the results of the Anselin Local Moran’s I analysis on total carbon emissions across county-level administrative units in Hunan Province at five distinct time points: 2000, 2005, 2010, 2015, and 2022. In 2000, the spatial distribution of carbon emissions showed pronounced clustering. Low-Low clusters predominantly appeared in seven counties in Western Hunan and three counties in Southern Hunan, indicating zones with consistently low emissions in close proximity. Conversely, Low–High clusters were identified in three counties within the Chang-Zhu-Tan Region and two counties in the Dongting Lake Ecological Economic Region, which denote areas of low emissions adjacent to regions with higher emissions. A single High-Low cluster was observed in Zhijiang Dong Autonomous County in Western Hunan, representing an isolated area with elevated emissions. High-High clusters were primarily located in four counties within the Chang-Zhu-Tan Region and Yueyang County, highlighting significant concentrations of high emissions. By 2005, the spatial clustering pattern experienced minor changes. A new Low–High cluster emerged in Yongxing County in Southern Hunan, alongside additional Low–High clusters in Shifeng District and Liling City within the Chang-Zhu-Tan Region. Simultaneously, Lusong District and Xiangtan County developed into new High-High clusters, while most other regions maintained relative stability. From 2010 onwards, the overall spatial clustering trend began to diminish. Zhijiang County in Western Hunan transitioned from a High-Low to a non-significant status, a change also noted in Shifeng and Liling within the Chang-Zhu-Tan Region. Ningxiang City emerged as a new Low–High cluster, while Lengshuijiang and Lianyuan evolved into new High-High clusters, and Xiangxiang shifted from a High-High to a non-significant status. By 2015, Anxiang County in the Dongting Lake region and Lanshan County in Southern Hunan had transitioned from Low-Low clusters to non-significant statuses, indicating a trend towards spatial stabilisation. In 2022, Yongshun County in Western Hunan emerged as a new Low-Low cluster, with Lanshan County also reverting to this category. Concurrently, Ningxiang City and Yuetang District in the Chang-Zhu-Tan Region transitioned from Low–High and High-High clusters, respectively, to non-significant statuses, signalling a weakening of localised high-emission clustering. These findings suggest that the spatial distribution of county-level carbon emissions in Hunan Province from 2000 to 2022 followed a general pattern of dispersion, clustering, and redistribution. Initially, emissions exhibited distinct spatial heterogeneity and local clustering. However, due to economic restructuring, energy optimisation, and the implementation of regional coordination strategies, the clustering effects in high-emission zones have progressively diminished. Inter-county disparities in carbon emissions have begun to narrow, and the overall spatial configuration has evolved into a more balanced and diversified state. This trend reflects a shift towards regional coordination and green transformation in the governance of carbon emissions in Hunan Province.
Fig. 3.
Total carbon emissions of county-level administrative units in Hunan Province Anselin Local Moran’s I results
Figure 4 depicts the centroid positions of total county-level carbon emissions in Hunan Province for the years 2000 and 2022, as delineated by standard deviation ellipses. The analysis indicates that the centroid of emissions consistently remained within Xiangxiang City, which is part of Xiangtan Municipality. Over this period, however, a clear southward migration of the centroid was observed. Specifically, the centroid transitioned from the northern to the southern parts of Xiangxiang, covering a total distance of 11.132 km, yet it remained within the city’s administrative boundaries. Concurrently, the morphology and extent of the standard deviation ellipse underwent considerable alterations: in 2000, the ellipse was relatively concentrated and oriented towards the northern area, but by 2022, it had expanded and rotated along its major axis, indicating a more pronounced north–south spatial extension. This shift suggests that the spatial distribution of carbon emissions in Hunan has progressively widened, evolving from a concentrated state in the north to encompass the central and southern regions of the province. Moreover, a more rapid increase in emissions has been observed in the southern areas. This trend highlights the heterogeneous development of industry, energy consumption, and urbanisation across different regions in Hunan. Consequently, this implies that a heightened focus on managing the escalating emissions burden in the central and southern areas is imperative when formulating carbon governance and reduction strategies.
Fig. 4.
Standard deviation ellipse and centroid migration of Hunan county-level carbon emissions in 2000 and 2022
Spatial–temporal evolution of GDP patterns
Figure 5 illustrates the trends in total county-level GDP and its corresponding growth rates across the four principal economic regions of Hunan Province from 2000 to 2022. During this period, all regions demonstrated a consistent upward trajectory in GDP, with particularly pronounced increases following 2010. Notably, the Chang-Zhu-Tan Region, as the economic nucleus of the province, consistently recorded the highest GDP figures and exhibited strong growth. This region was succeeded by the Dongting Lake Ecological Economic Region and the Southern Hunan Region, while the Western Hunan Region, despite lower GDP figures, also showed steady growth. In terms of growth rates, significant fluctuations were observed between 2000 and 2011. The Dongting Lake Ecological Economic Region reached a peak growth rate of nearly 50% in 2011, suggesting substantial economic expansion or possible adjustments in statistical methodologies. Before 2008, the Southern Hunan Region experienced considerable volatility in GDP growth, with rates approaching 30% in some years, whereas the Western Hunan Region displayed a pattern of rapid growth followed by a gradual slowdown. After 2012, growth rates across the regions stabilised, transitioning to a phase of moderate annual growth of approximately 10%, with narrower disparities among the regions. Despite the general slowdown triggered by the COVID-19 pandemic in 2020, none of the regions recorded negative GDP growth, and signs of gradual economic recovery were evident from 2021 onwards. These observations indicate a spatial gradient in ED across Hunan Province, with the Chang-Zhu-Tan Region consistently acting as the primary engine of growth. Concurrently, the other regions have shown trends of developmental catch-up and convergence. In recent years, Hunan’s economic trajectory has shifted towards high-quality growth, fostering a more balanced regional economic structure.
Fig. 5.
Trends in GDP and growth rates at the county level in Hunan Province
Hunan county-level carbon emissions decoupling index
To enhance the understanding of how carbon emissions and ED decouple across varied regional contexts, this study categorises 122 counties and districts in Hunan Province into six distinct groups. These categorisations are based on criteria including the proportion of primary industry in the local economy, ecological assets, and the prevalent types of energy consumption. The designated categories are as follows: agriculture-dominated, ecological-cultural, energy-resource-based, heavy-industry-based, light-manufacturing-based, and comprehensive service-oriented regions. Detailed criteria for these classifications are provided in Table 4.
Table 4.
County types in Hunan province
| Category | Classification criteria |
|---|---|
| Agriculture-dominated (AR) | Primary industry accounts for more than 30% of GDP; low per capita GDP |
| Ecological-cultural (ECR) | Located in key ecological function zones; dominated by tourism and cultural resources |
| Energy-resource-based (ERR) | High proportion of energy-intensive industries such as coal mining, electricity generation, and smelting |
| Heavy-industry-based (HIR) | High share of secondary industry, primarily focused on heavy industry |
| Light-manufacturing-based (LIR) | Dominated by light industries such as food processing and textiles |
| Comprehensive service-oriented (CSR) | Dominated by the tertiary industry; high urbanization rate |
Figure 6 illustrates the Tapio decoupling status between carbon emissions and economic growth across counties in Hunan Province during four distinct periods: 2000–2005, 2006–2010, 2011–2015, and 2016–2022. Additionally, subfigures 6(b), 6(e), 6(h), and 6(k), along with 6(c), 6(f), 6(i), and 6(l), respectively, delineate the proportion of different decoupling types across regional and county-level typologies for each timeframe.
Fig. 6.
Tapio decoupling effect of county-level carbon emissions and economic growth in Hunan Province over various periods
Between 2000 and 2005, the majority of counties in Hunan Province predominantly demonstrated a weak decoupling status (57%), whereby economic output and carbon emissions grew concurrently, albeit with relatively low carbon elasticity. Additionally, a considerable proportion of counties experienced expansive negative decoupling (22%) and expansive coupling (12%), indicating that in some areas, carbon emissions either increased at a rate surpassing GDP growth or progressed in complete synchrony with economic expansion. The prevalence of strong negative decoupling was notably minimal (9%), suggesting that only a few counties achieved significant reductions in emissions during this time frame. From a regional analysis, 57% of the counties in the Chang-Zhu-Tan area were characterized by weak decoupling. Similarly, both Western Hunan and Southern Hunan displayed predominantly weak decoupling statuses, each at 53%. However, Western Hunan was distinct in exhibiting a significantly higher incidence of expansive negative decoupling (36%), indicative of a pattern where economic growth was accompanied by rapid increases in carbon emissions. In contrast, in the Dongting Lake region, 67% of counties achieved weak decoupling, which may reflect relatively effective carbon management practices in this area. Examining the situation from the perspective of functional typologies, counties classified as eco-cultural and agriculture-dominated primarily displayed weak decoupling, at 90% and 60% respectively. This suggests that while their ED levels were modest, carbon emissions remained comparatively controlled. Conversely, 50% of counties that are reliant on energy resources were marked by expansive negative decoupling, illustrating a development trajectory heavily burdened by carbon emissions. Notably, counties with a focus on light industry demonstrated the highest incidence of strong negative decoupling, at 40%. This may be attributable to technological advancements and energy efficiency improvements within these regions.
Between 2006 and 2010, with the ongoing advancement of ED, there was a significant improvement in the overall decoupling level in Hunan Province. The predominant pattern was weak decoupling (84%), indicating that in most counties, carbon emissions increased at a slower rate than GDP. This reflects an initial movement towards low-carbon development. The proportion of strong decoupling (6%) also rose compared to the earlier period, while the instances of expansive negative decoupling (5%) and expansive coupling (2%) decreased markedly, suggesting a diminishing reliance of carbon emissions on economic growth. From a regional perspective, as many as 96% of the counties in the Dongting Lake region were classified as weakly decoupled, representing the most favorable decoupling status among the four regions. Both Southern Hunan (79%) and Western Hunan (89%) exhibited high proportions of weak decoupling, indicating that most areas within the province were beginning to move away from carbon-intensive development patterns. In contrast, while the Chang-Zhu-Tan region primarily remained weakly decoupled (75%), 14% of its counties were still categorized under expansive negative decoupling, highlighting the increasing pressures of carbon emissions faced by this urban-industrial core amid rapid industrialization and urbanization. In terms of county typology, counties dominated by agriculture (91%) and those identified as eco-cultural (90%) demonstrated the strongest trends in weak decoupling, suggesting that these areas had distinct advantages in managing the growth of carbon emissions. However, 22% of counties identified as heavy-industry-oriented continued to experience expansive negative decoupling, reflecting the challenges posed by their energy-intensive ISs. Notably, half of the energy-resource-based counties achieved strong decoupling, possibly benefiting from structural adjustments or technological improvements in resource-intensive sectors. Light-industry-oriented counties displayed significant internal variation, with 20% falling under expansive negative decoupling and another 20% achieving strong decoupling, indicating uneven development trajectories within this group.
During the period from 2011 to 2015, the overall decoupling level in Hunan Province improved further. Approximately 68% of the counties achieved a state of weak decoupling, which continued to dominate the overall pattern. However, the proportion of counties experiencing strong decoupling rose significantly to 29%, indicating that some counties had achieved a notable separation between economic growth and carbon emissions, thereby marking substantial progress in the region’s transition to a low-carbon economy. Meanwhile, the incidences of expansive negative decoupling and expansive coupling both declined to just 2%, suggesting that scenarios involving high-carbon development had been significantly reduced. Regionally, the Chang-Zhu-Tan area was particularly impressive: 36% of its counties achieved strong decoupling status, with the remainder exhibiting weak decoupling, and none fell into any expansive decoupling categories. This performance underscores the region’s role as the province’s economic powerhouse and highlights its achievements in green development and technological upgrading. Strong decoupling was also observed in 21%, 29%, and 28% of the counties in the Dongting Lake, Southern Hunan, and Western Hunan regions, respectively, indicating that the entire province was accelerating its shift toward low-carbon development during this period. In terms of county typologies, as many as 75% of the energy-resource-dependent counties achieved strong decoupling, the best performance among all types. This likely resulted from structural adjustments and improvements in energy efficiency within resource-intensive industries. Furthermore, 60% of light-industry-oriented counties also achieved strong decoupling, suggesting substantial progress in clean production and industrial upgrading. While agricultural-dominated, eco-cultural, and heavy-industry-oriented counties primarily remained in weak decoupling (all exceeding 70%), the proportion of counties achieving strong decoupling was steadily increasing. Notably, 25% of agricultural counties and 20% of eco-cultural counties achieved strong decoupling, signaling their growing potential for low-carbon transformation.
Between 2016 and 2022, under the enhanced focus on ecological and environmental protection mandated by China’s 13th and 14th Five-Year Plans, the decoupling of carbon emissions from economic growth at the county level in Hunan Province was significantly improved. The proportion of counties achieving strong decoupling increased sharply to 53%, establishing it as the predominant category. This trend indicates that a majority of counties successfully reduced carbon emissions while sustaining economic growth, signaling a robust trajectory towards low-carbon development. Concurrently, weak decoupling, which was observed in 43% of the counties, suggests that some counties remained in a transitional phase characterized by a moderate increase in emissions, albeit at a rate slower than that of economic growth. The proportions of counties experiencing expansive negative decoupling and strong negative decoupling decreased markedly to 2% and 1%, respectively, indicating a decline in high-carbon growth patterns. The emergence of recessive decoupling in 1% of the counties likely reflects the economic instability or adjustment challenges encountered by certain areas. From a regional perspective, Southern Hunan and Western Hunan exhibited the most impressive results, with 68% and 64% of counties, respectively, achieving strong decoupling. These traditionally less developed regions have realized a notable balance between reducing emissions and fostering development through industrial restructuring and enhancements in energy efficiency. In contrast, the Dongting Lake and Chang-Zhu-Tan regions reported slightly lower levels of strong decoupling, with 38% and 36% of counties, respectively. However, a majority of counties in both regions demonstrated weak decoupling (63% and 61%, respectively), suggesting that while economic momentum remained robust, emissions were effectively managed. In terms of functional county types, eco-cultural counties exhibited the most notable success, with 80% achieving strong decoupling and 10% displaying recessive decoupling. This underscores the potential of tourism- and culture-based low-carbon industries. Agricultural-dominated counties (67% strong decoupling) and light-industry-oriented counties (40% strong decoupling) also showed favorable trends towards a green transition. In contrast, comprehensive service-oriented counties predominantly experienced weak decoupling but demonstrated consistent improvements in carbon control capabilities. Notably, energy-resource-dependent counties presented the most varied decoupling patterns, with 25% categorized under expansive negative decoupling and another 25% under strong negative decoupling, highlighting ongoing challenges in carbon reduction within these energy-intensive industrial sectors.
County-level carbon emissions influential factors and carbon peak path prediction in Hunan
Influential factors of carbon emissions at the county level in Hunan
The outcomes of the OLS regression are presented in Table 5. The model exhibits a robust fit, with R2 and adjusted R2 values of 0.994, affirming that the selected explanatory variables adequately account for the variance in county-level carbon emissions (ln CE). The overall significance of the model is substantiated by an F-test (F = 76,212.612, p < 0.01). Although the Durbin–Watson statistic (D-W = 0.923) reveals a degree of autocorrelation, its impact is minimal given the extensive sample size.
Table 5.
OLS regression results
| Variable | Non-standardized coefficient | Standardized coefficient | Std.Err | t | p | 95% CI |
|---|---|---|---|---|---|---|
| constant | -4.444 | - | 0.025 | -176.697 | 0.000** | -4.493 ~ -4.394 |
| ln CI | 0.978 | 0.917 | 0.002 | 531.016 | 0.000** | 0.974 ~ 0.981 |
| ln P | 0.948 | 0.447 | 0.005 | 194.748 | 0.000** | 0.939 ~ 0.958 |
| ln ED | 0.984 | 0.792 | 0.005 | 201.159 | 0.000** | 0.975 ~ 0.994 |
| ln IS | -0.154 | -0.014 | 0.021 | -7.289 | 0.000** | -0.195 ~ -0.112 |
| ln ECI | -0.051 | -0.046 | 0.004 | -12.543 | 0.000** | -0.059 ~ -0.043 |
| ln UR | -0.023 | -0.01 | 0.006 | -3.746 | 0.000** | -0.036 ~ -0.011 |
| R2 | 0.994 | |||||
| Adj R2 | 0.994 | |||||
| F | F (6,2747) = 76,212.612,p = 0.000 | |||||
| D-W value | 0.923 | |||||
* p < 0.05 ** p < 0.01
Among the explanatory variables, carbon emission intensity, population, and economic development significantly increase county-level carbon emissions. Carbon emission intensity is the predominant factor, with a standardized coefficient of 0.917 (p < 0.01); a 1% rise in emission intensity corresponds to an approximate 0.978% increase in total carbon emissions. Population has a standardized coefficient of 0.447 (p < 0.01), with a 1% increase in population associated with about a 0.948% rise in emissions, while economic development has a standardized coefficient of 0.792 (p < 0.01), each 1% increase in per-capita output corresponding to roughly a 0.984% increase in emissions.
Conversely, industrial structure, energy-consumption intensity, and urbanization are negatively associated with carbon emissions. Industrial structure has a standardized coefficient of -0.014 (p < 0.01), with a 1% increase in the industrial-structure upgrading index corresponding to an approximate 0.154% decline in emissions, indicating that the shift from heavy industry toward service and high-tech sectors reduces emissions. Energy-consumption intensity has a standardized coefficient of -0.046 (p < 0.01), a 1% improvement in energy efficiency being associated with about a 0.051% reduction in emissions. Urbanization has a standardized coefficient of -0.010 (p < 0.01), a 1% rise in the urbanization rate corresponding to roughly a 0.023% decrease in emissions, potentially reflecting improved energy management and infrastructure optimization in urban areas. In summary, the principal drivers of county-level carbon emissions in Hunan Province are emission intensity, population, and economic development, whereas industrial restructuring, improved energy efficiency, and urbanization exert mitigating effects.
The development of a predictive model based on the STIRPAT framework for county-level carbon peaking in Hunan Province, following the results of the OLS regression, is outlined as:
 |
15 |
This study extends the existing model by analysing the principal developmental factors that ought to be prioritised across different regions to optimise carbon peaking pathways. The research employs the GeoDetector method to scrutinise the driving forces and coupling effects of various developmental components across four regions, utilising a predictive model. The findings indicate that both the explanatory power of individual variables for CEI and their interaction effects demonstrate significant temporal variations (Fig. 7). In 2000, the interaction values between CEI and variables such as P, ED, and IS were relatively low, indicating weak independent driving effects and a balanced stage of regional development at that time. Conversely, ECI and UR exhibited stronger individual explanatory power, highlighting their critical roles in shaping carbon emission patterns due to the prevailing energy structure and early-stage urbanisation. By 2005, the interaction q-values for all factors had increased markedly, particularly between CEI and ECI (0.939), CEI and UR (0.828), and CEI and ED (0.737). This trend suggests that energy-intensive growth and urban expansion had significantly augmented their influence on carbon emissions, with initial signs of regional heterogeneity becoming evident. The year 2010 was characterised by the zenith of factor interaction effects, with q-values for CEI and ED (0.895), CEI and IS (0.862), CEI and ECI (0.916), and CEI and UR (0.917) all exceeding 0.85. During this period, ED, industrial transformation, and energy consumption collectively drove changes in carbon emissions, leading to more pronounced spatial differentiation in carbon peaking pathways. Post-2015, some interaction effects began to diminish, such as those between CEI and ED (0.771) and CEI and IS (0.724), while the interaction between CEI and ECI (0.83) remained substantial, emphasising the continued importance of energy efficiency in emission control. The interaction q-values of CEI with P and UR stabilised around 0.77, indicating that both P concentration and UR continued to be significant drivers of CEI. By 2022, the interaction effects among all variables had stabilised, with q-values around 0.7, exemplified by CEI and ECI at 0.748 and CEI and UR at 0.776, reflecting a shift towards a more diversified and synergistic mechanism of carbon emission management.
Fig. 7.
Geo detector two-factor interaction results
Overall, ECI, UR, and ED have emerged as critical long-term determinants of changes in CEI. Notably, the interaction effects between these variables and CEI have demonstrated significant influence during the mid-to-late periods under review. Consequently, in the formulation of carbon peaking pathways, it is imperative for various regions to consider their unique characteristics and devise tailored policy packages. These packages should focus on enhancing energy efficiency, optimising ISs, and controlling urban expansion rates. Such a strategic approach will facilitate the development of region-specific and diversified strategies essential for achieving carbon peaking objectives.
Carbon peak prediction
To align the carbon peaking scenario simulations for Hunan Province with existing strategic plans, this study establishes core parameters for three developmental scenarios, low-carbon, medium-carbon, and high-carbon, covering the period from 2022 to 2040. These parameters were determined based on a comprehensive review of key policy documents, including the 14th Five-Year Plan and the 2035 Long-Term Vision Outline for Hunan Province, the Carbon Peaking Implementation Plan for Hunan Province, the New Urbanisation Plan for Hunan Province (2021–2035), and the 14th Five-Year Development Plan for the Service Sector in Hunan Province. The selection of parameter values adheres strictly to the actual developmental conditions and planning objectives of the province, as well as to the national-level targets.
The parameter settings under consideration primarily reflect both the historical development trajectory and the future planning guidelines of Hunan Province (Table 6).
Table 6.
Parameter assignment
| Development level | Year | P | ED | IS | CEI | UR | ECI |
|---|---|---|---|---|---|---|---|
| Low speed (LS) | 2022–2025 | -0.90% | 5.50% | 0.30% | -3.50% | 1.20% | 2.00% |
| 2026–2030 | -1.20% | 5.00% | 0.50% | -3.00% | 1.10% | 1.50% | |
| 2031–2035 | -1.50% | 4.50% | 0.70% | -2.50% | 1.00% | 1.00% | |
| 2036–2040 | -1.80% | 4.00% | 1.00% | -2.00% | 0.90% | 0.50% | |
| Middle speed (MS) | 2022–2025 | -0.60% | 7.00% | 0.80% | -5.00% | 1.50% | -1.00% |
| 2026–2030 | -0.90% | 6.50% | 1.20% | -4.50% | 1.40% | -1.50% | |
| 2031–2035 | -1.20% | 6.00% | 1.50% | -4.00% | 1.30% | -2.00% | |
| 2036–2040 | -1.50% | 5.50% | 1.80% | -3.50% | 1.20% | -2.50% | |
| High speed (HS) | 2022–2025 | -0.30% | 8.50% | 1.50% | -6.50% | 1.80% | -3.00% |
| 2026–2030 | -0.60% | 8.00% | 2.00% | -5.50% | 1.70% | -3.50% | |
| 2031–2035 | -0.90% | 7.50% | 2.50% | -5.00% | 1.50% | -4.00% | |
| 2036–2040 | -1.20% | 7.00% | 3.00% | -4.50% | 1.00% | -4.50% |
Specifically, 1) P Level: Prior to 2021, Hunan Province experienced sluggish P growth. According to the Seventh National Population Census Bulletin for Hunan Province, the average annual growth rate from 2010 to 2021 was only 0.11%. After 2021, a decline in P was observed, with the numbers decreasing from 66.22 million in 2021 to 65.68 million in 2023, representing an average annual decrease of 0.8%. These figures indicate that the province may have entered a period of negative P growth in recent years. Consequently, under the medium-growth scenario, the projected annual average growth rate of the year-end permanent P from 2022 to 2025 is set at -0.6%, which seeks to balance the observed decline with potential policy interventions aimed at reversing this trend.
2) ED: In 2022, Hunan’s per capita GDP reached RMB 73,598. Between 2001 and 2022, the average annual growth rate was 9.6%, although this rate decreased to 7.2% from 2017 to 2022. By comparison, the national average per capita GDP in 2022 was RMB 81,400. Considering the potential for economic advancement, in alignment with the objectives specified in the 14th Five-Year Plan and the 2035 Long-Term Vision Outline for Hunan Province, and the resolutions of the 12th Congress of the CPC Hunan Provincial Committee—which established a target of achieving approximately RMB 100,000 per capita GDP by 2026—the medium-growth scenario adopts a per capita GDP growth rate of 7.0% for the period 2022–2025. This adjustment indicates a transition to a more moderate growth pace, with an anticipated gradual deceleration thereafter.
3) IS: From 2017 to 2022, the proportion of the tertiary sector within Hunan’s GDP increased from 52.68% to 53.46%, with the growth rate decelerating post-2020. The five-year average growth rate was approximately 0.5%. In light of the aims outlined in the Development Plan for Modern Service Industries of Hunan Province, which sets a target for the service sector’s contribution to GDP at 53% by 2025, the medium-growth scenario forecasts an annual increase of 0.8% in the industrial upgrading index. This rate is expected to gradually accelerate thereafter, reflecting the enhanced policy support for this sector.
4) ECI: According to the 14th Five-Year Plan for Renewable Energy Development in Hunan Province, by 2025, the installed capacity of renewable energy is projected to reach 44.5 million kilowatts, and the consumption of non-fossil energy is expected to account for over 22% of the total primary energy consumption. Consequently, the consumption of standard coal in Hunan is projected to decrease annually. In this analysis, total electricity consumption is used as a proxy indicator for energy consumption. Under the medium-growth scenario, the annual reduction rate in ECI is initially set at –1.00%, with a projected gradual decline in subsequent years.
5) Social Development: From 2017 to 2021, the UR in Hunan increased from 54.62% to 59.71%, with an average annual growth rate of 1.71%. However, this growth rate decelerated from 2.68% to 1.62% over the same period. According to the 14th Five-Year New Urbanization Plan of Hunan Province, the target is to elevate the UR of permanent residents to approximately 63% by 2025 and further to 72% by 2035. The medium-growth scenario anticipates an annual increase of 1.5% in the UR from 2022 to 2025, followed by a gradual slowdown thereafter. Consistent with the Plan’s emphasis on “balancing speed and quality,” the period from 2022 to 2025 focuses on acceleration, while the period from 2031 to 2035 concentrates on enhancing quality. Accordingly, a gradual reduction in the annual growth rate from 1.50% to 1.30% is anticipated, reflecting a natural deceleration trend.
6) CEI: In 2021, Hunan’s CEI was recorded at 0.63 tonnes per 10,000 RMB of GDP. Between 2017 and 2021, the average annual change rate in CEI was –6.2%, although a rebound of 1.1% was observed in 2021, indicating mounting challenges in emission reduction efforts. The 14th Five-Year Plan of Hunan Province, in conjunction with the Carbon Peak Implementation Plan of Hunan Province, establishes a mandatory target of reducing CO₂ emissions per unit of GDP by 18% from 2020 to 2025. Additionally, the 2024–2025 Energy Conservation and Carbon Reduction Action Plan specifically mandates a reduction in CO₂ emission intensity of 3.9% for 2024. By 2030, it is projected that China’s emission intensity will have decreased by more than 65% relative to 2005 levels. In this framework, the medium-growth scenario for Hunan prescribes an annual reduction rate in CEI of –5.0% for the period 2022–2025, with expectations of a continued gradual decrease thereafter.
The medium-speed scenario is established as the benchmark, while the low-speed and high-speed scenarios are defined by adjusting the rate of development such that their key indicators reach a comparable extent of progress approximately five years later or earlier, respectively, than that of the medium-speed scenario. Consequently, three representative development scenarios are outlined: low-carbon, medium-carbon, and high-carbon (refer to Table 7).
Table 7.
Scenario combinations
| Scenarios | Scenarios setting | ||||||
|---|---|---|---|---|---|---|---|
| TP | RG | SG | TG | EI | UR | FI | |
| Low-carbon development | MS | MS | HS | HS | HS | MS | HS |
| Medium-carbon development | MS | MS | MS | MS | MS | MS | MS |
| High-carbon development | MS | HS | LS | LS | LS | MS | LS |
1) Medium-Carbon Development Scenario: This scenario mandates that all significant variables evolve at a moderate pace. It is based on the historical development trends of Hunan and projects potential future trajectories of carbon emissions under the objectives delineated in the 14th Five-Year Plan and the 2035 Long-Term Vision. It forecasts the province’s carbon emissions peak in line with existing policies.
2) Low-Carbon Development Scenario: In this scenario, the UR, per capita GDP, and year-end resident P follow medium-speed trajectories, while the proportions of secondary and tertiary industries, the CEI, and the rates of change in fiscal balance proceed at an accelerated pace. Based on the assumption of stable socioeconomic development in Hunan, this scenario emphasises expedited industrial advancement, a rapid decrease in the proportion of secondary industries, a swift transition towards green and low-carbon energy sources, enhanced adoption of clean energy, and an expansion of the service sector. It also involves the improvement of fiscal expenditure structures and the management of urban carbon emissions.
3) High-Carbon Development Scenario: In this scenario, both the UR and the year-end resident P maintain a medium pace, while IS, CEI, and fiscal balance evolve at a slower rate; per capita GDP grows rapidly. This model is predicated on a development strategy that prioritises economic expansion, supported by a strong secondary sector and energy supply. Compared to the medium-carbon scenario, it delays industrial restructuring and efforts to reduce emissions. Simultaneously, it promotes rapid development in transportation, foreign trade, and urban infrastructure to sustain economic dynamism. Under this scenario, the carbon peak outlook for Hunan assumes minimal additional mitigation measures.
Figure 8 illustrates the projected trajectories of carbon emissions and the anticipated timings for carbon peaking across four major economic zones in Hunan Province: the Chang-Zhu-Tan Region, the Dongting Lake Ecological Economic Region, the Southern Hunan Region, and the Western Hunan Region, under three distinct carbon development scenarios (low-carbon, medium-carbon, and high-carbon). The analysis indicates that, within the low-carbon development scenario, all four regions are likely to reach their peak carbon emissions around 2030, with emission growth rates declining significantly and stabilising by 2035. Specifically, the Chang-Zhu-Tan Region and the Dongting Lake Region are expected to achieve their emission peaks around 2030, subsequently maintaining stable emission levels, which suggests effective emission reduction strategies and policy responses.
Fig. 8.
Trends in carbon emissions and expectations for carbon peaking in economic regions of Hunan Province
In contrast, the Southern and Western Hunan Regions are projected to reach their emission peaks closer to 2035. Despite these slightly delayed peaks, the emission trajectories align with China’s “dual carbon” goals. Under the medium-carbon scenario, while carbon emissions in all regions are expected to continue to increase, the rate of increase is slower than that predicted under the high-carbon scenario. However, it is projected that none of the regions will achieve an effective carbon peak by 2040, highlighting the challenges that these development paths face in achieving carbon neutrality without more stringent regulatory measures. In the high-carbon development scenario, all regions experience a significant and continuous increase in carbon emissions, reaching unprecedented levels. The Chang-Zhu-Tan and Dongting Lake Regions, in particular, are subject to accelerated emission increases due to intensive industrialisation and urbanisation, making a medium- to long-term carbon balance untenable under this scenario.
Significant disparities characterize the trajectories of carbon emissions and the capacities to achieve carbon peaks across various economic regions of Hunan Province. These variations arise from differences in IS, energy consumption patterns, technological advancement levels, and policy implementation capacities. As the economic hub of the province, the Chang-Zhu-Tan region, which boasts advanced ED levels and technological infrastructure, leads the provincial agenda in green industrial transformation and improvements in energy efficiency. Consequently, this region is expected to reach its carbon peak earliest under a low-carbon scenario, achieving the most substantial emission reductions. In contrast, the Western Hunan region, with its strong dependence on natural resources, limited industrial base, and a predominantly coal-based energy framework, exhibits significant inflexibility in its carbon emissions, thereby hindering the achievement of peak emission levels. Meanwhile, the Dongting Lake Ecological Economic Region, despite its substantial ecological protection responsibilities, is also influenced by agricultural and basic industrial activities, positioning its emission control challenges at an intermediate level among the regions. The Southern Hunan region, possessing potential for industrial growth, may see a continued increase in emissions in the absence of strong policy interventions. These regional disparities underscore the necessity for tailored strategies to achieve a province-wide carbon peak. Differentiated emission reduction pathways and policy packages, designed to suit specific regional conditions, are crucial for enhancing coordination and achieving balanced regional development. The simulation results underscore the critical role of developmental pathway choices in meeting carbon peak targets, indicating that, alongside sustained economic growth, an accelerated transition to low-carbon practices is essential. This is particularly vital in less developed areas such as Southern and Western Hunan, where achieving carbon peaks will depend significantly on policy support and green investments.
Discussion
Regional energy conservation and low-carbon transition strategies
Following a comprehensive decoupling analysis of county-level carbon emissions and ED in Hunan Province from 2000 to 2022, and multiple simulations of carbon peak scenarios, it has become evident that the province’s four principal economic regions exhibit significant disparities in their carbon emission trajectories, peak potentials, and mitigation capacities. These differences are primarily attributed to regional disparities in IS, energy consumption patterns, technological infrastructure, and policy responsiveness. Therefore, to support Hunan’s overarching goals of achieving a carbon peak and fostering a green transition, it is crucial to develop energy conservation and low-carbon strategies that are tailored, targeted, and phased, reflecting the unique attributes of each region.
Chang-Zhu-Tan Region: Consolidate Leading Advantages and Build a Green Development Engine
The Chang-Zhu-Tan region, recognized as the most industrialized and urbanized area within Hunan Province, accommodates a significant proportion of energy-intensive manufacturing industries and is the primary hub for technological innovation in the province. Simulations suggest that, under a low-carbon scenario, the region is expected to reach its carbon emissions peak before 2030, followed by a stabilization of emissions. Future strategies should focus on enhancing the green transformation of the manufacturing sector, accelerating the intelligent and low-carbon modernization of traditional industries, and strictly enforcing entry criteria for projects categorized as “high energy-consuming, high-emission, and overcapacity.” Industrial parks within the region are encouraged to promote the cascading utilization of energy and the substitution of clean energy, as well as to implement carbon footprint management systems and green supply chain frameworks. Leveraging its strong innovation capabilities, the region should also advance cutting-edge technologies, including carbon capture, utilization, and storage (CCUS), digital energy management, and building energy efficiency. The objective is to develop a low-carbon urban agglomeration and a sustainable industrial ecosystem, thereby positioning the Chang-Zhu-Tan region as a paradigm of low-carbon leadership in Hunan and across central China.
(2) Dongting Lake Ecological Economic Region: Leveraging Ecological Potential to Enhance Synergistic Emission Reductions in Agriculture and Industry
The Dongting Lake Ecological Economic Region, characterized by its extensive plains, farmlands, and crucial wetlands, plays a pivotal role in ecological conservation and food security within Hunan Province. Moreover, it serves as a foundational hub for various industries. This study demonstrates that although the region’s carbon emissions exceed those of Southern and Western Hunan, substantial potential exists for achieving carbon peaking with considerable flexibility. It is crucial to prioritize the adoption of low-carbon agricultural practices such as straw return, conservation tillage, and the resource recovery of agricultural waste to facilitate the agricultural sector’s green transformation. Simultaneously, energy-intensive sectors, including traditional energy and building materials industries, should be subject to more stringent energy efficiency regulations and guided towards the adoption of clean energy solutions and the optimization of energy structures. In counties with abundant wetland resources, the initiation of pilot carbon sink trading programs, coupled with the expansion of forest carbon sequestration and cropland carbon storage projects, could produce significant co-benefits for both ecological protection and carbon reduction.
-
(3)
Southern Hunan: Capitalizing on the Critical Period of Industrial Transformation to Mitigate “Carbon Lock-in” Risks
Southern Hunan has recently emerged as one of the most dynamically industrializing regions in the province, undergoing a rapid expansion phase. According to this study, under medium- and high-carbon scenarios, the region is projected to face sustained upward pressure on emissions, underscoring the critical necessity to prevent long-term high-carbon developmental lock-in. It is advisable for the region to implement stringent energy efficiency assessments and environmental impact controls on new industrial projects to curtail the unchecked growth of energy-intensive and high-emission industries. Enhanced support should be extended to high-tech industries and advanced manufacturing sectors, including the development of green factories and eco-industrial parks. Given the trends of urban expansion in the region, the early implementation of green infrastructure, such as sustainable transport systems, energy-efficient buildings, and waste-to-resource initiatives, is pivotal for the structural optimization of urban carbon emissions.
(4) Western Hunan: Accelerating Energy Substitution and Industrial Diversification to Overcome Carbon Rigidity in Resource-based Regions
Western Hunan, characterized by its complex terrain and a coal-dominated energy structure, faces significant challenges due to its strong reliance on high-carbon energy, a weak industrial base, and relatively underdeveloped ED, all contributing to low carbon flexibility. In scenarios of high carbon emissions, this region not only peaks later but also maintains elevated carbon intensity over time. To overcome the dual constraints imposed by its existing energy structure and developmental trajectory, it is imperative for the region to prioritize the expansion of clean energy sources. These include distributed photovoltaics, small-scale hydropower suited to its mountainous terrains, and biomass energy, which would collectively reduce its dependence on coal. Simultaneously, there should be a strategic shift from resource-dependent industries to those with lower carbon footprints, such as cultural tourism, healthcare services, and specialty agriculture. This transition could be accelerated through cross-regional industrial coordination with areas such as Chang-Zhu-Tan, which would facilitate the relocation and upgrading of industries through mechanisms of “industrial cage replacement” and the transfer of green gradients. From a policy and financial perspective, it is crucial to increase investment in low-carbon infrastructure within Western Hunan, coupled with the exploration of integrated mechanisms of “ecological compensation plus carbon trading” to enhance the region’s capacity for carbon reduction.
Policy recommendations for low-carbon development at the county level in Hunan province
The regression and spatial analyses conducted in this study demonstrate that, at the county level in Hunan Province, carbon emissions are significantly influenced by CI, ED, and P size. In contrast, factors such as IS, ECI, and UR appear to exert a restraining influence on emissions. To facilitate the achievement of carbon peaking and carbon neutrality at the county level, amidst the challenges posed by developmental disparities, this study proposes three distinct policy directions:
Coordinated Industrial Restructuring and Energy Efficiency Enhancement
The high carbon intensity and substantial reliance on the secondary sector significantly contribute to regional emissions. Regression analyses reveal that counties with more advanced ISs exhibit lower emissions. Furthermore, the GeoDetector identifies a robust correlation between CI and ECI. It is therefore essential for counties to prioritize the transformation of their industrial bases, gradually transitioning from energy-intensive manufacturing to the promotion of high-tech and low-carbon service industries. This shift must include the integration of clean energy technologies, particularly in counties known for resource-based or heavy industries. In these areas, restructuring energy use and enhancing efficiency can lead to considerable marginal reductions in emissions. Practical steps include the development of green industrial parks, energy-efficient upgrades to existing facilities, and improved integration of renewable energy sources into the local energy infrastructure. Additionally, the implementation of performance-based energy management systems in major industrial enterprises, coupled with expanded access to clean electricity in rural areas, will be pivotal in this transformation.
(2) Spatially Differentiated Pathways Based on County Functions and Development Stage
The considerable heterogeneity among the counties of Hunan, which range from agriculture-dominated to industrial hubs, renders the implementation of a uniform decarbonization strategy both unrealistic and inefficient. The research indicates that counties in western or southern Hunan, primarily characterized by ecological and agricultural activities, have experienced varying degrees of decoupling, limited by their modest industrial bases. These regions should leverage their ecological advantages by enhancing forest carbon sinks, developing eco-tourism, and promoting sustainable agricultural practices. In contrast, industrialized counties within areas such as the Chang-Zhu-Tan urban agglomeration, which are experiencing persistent expansionary decoupling or strong coupling, must implement stringent carbon control measures and green their industrial processes. Additionally, emerging urban centers undergoing rapid population increases should focus on transit-oriented development and the implementation of green infrastructure to avoid future lock-in effects. This strategy of spatially differentiated governance not only aligns with the principle of “local peaking before provincial peaking” but also provides a resilient and regionally adaptive framework for achieving carbon neutrality.
(3) Institutional Capacity Building and Market-Based Incentives for Long-Term Transition
The success of a low-carbon transition fundamentally relies on the governance capacity and the strength of institutional frameworks. Currently, the lack of county-level emissions monitoring systems severely hampers informed decision-making. It is imperative to establish a hierarchical carbon inventory and a real-time monitoring platform at the county level, which should be interconnected with the provincial carbon data hub. This system ought to include major emitters, significant energy consumers, and variables pertinent to land-use changes. Such an infrastructure will facilitate effective carbon budget management and performance tracking. Additionally, the transition requires sustained financial investment. It is advisable for counties to create low-carbon development funds, which could be financed through mechanisms like green bonds, environmental taxes, or fiscal transfers. Where feasible, pilot initiatives for county-level carbon markets or offset mechanisms should be considered, especially in counties with robust industrial sectors or significant ecological resources. These mechanisms aim to internalize environmental costs and encourage private sector investment in decarbonization efforts.
Comparison with existing literature and case illustrations
The empirical findings of this study both corroborate and refine the existing research on carbon decoupling in China by illustrating how decoupling mechanisms are more pronounced at the county level. Aligning with studies at the national and provincial scales, which report a progression from weak to strong decoupling in economically advanced regions [14, 26], our results demonstrate a significant increase in the proportion of counties achieving strong decoupling in Hunan Province over time. However, the heterogeneity of decoupling trajectories at the county level is markedly greater than at higher administrative levels, emphasizing the need for detailed analysis to identify differentiated transition pathways. When set against provincial-scale evidence, this contrast is especially informative: whereas Liu et al. [32], coupling decoupling analysis with scenario simulation, reported that China’s overall decoupling status first improved and then deteriorated at the provincial scale, our county-level results instead reveal a sustained rise in the share of strongly decoupled counties that nonetheless conceals sharply divergent within-province trajectories, underscoring the value of resolving decoupling dynamics below the provincial scale.
In terms of driving mechanisms, the predominant influence of carbon emission intensity in our regression analysis is consistent with findings from provincial-level studies that highlight intensity reduction as a key driver of emission mitigation [22]. Notably, the negative impact of industrial structure upgrading in this study is more pronounced than that reported in analyses at the prefecture or provincial levels [20]. This indicates that industrial restructuring has a more immediate and observable impact on emission reduction at the county level, where changes in dominant industries directly alter local energy consumption patterns. Furthermore, the GeoDetector results reveal that the interaction between carbon intensity and energy consumption intensity possesses particularly strong explanatory power, underscoring that emission reduction outcomes are seldom driven by single factors but rather result from synergistic interactions among energy efficiency, industrial transformation, and economic development. These intensity- and structure-driven patterns are broadly consistent with city-scale driver decompositions and peak projections [33] and with county-scale evidence from Chongqing linking carbon-emission intensity to construction-land-use intensity [23], yet our analysis extends them by showing that, at the county scale, such drivers operate through pronounced functional heterogeneity rather than uniformly across space.
These mechanisms can be exemplified through contrasting county-level cases within Hunan Province. Changsha County, situated within the Chang-Zhu-Tan urban agglomeration, represents a transition-driven decoupling pathway. According to our empirical findings, counties experiencing rapid industrial upgrading and declining energy intensity are more likely to achieve strong decoupling. In Changsha County, the expansion of the tertiary sector, coupled with enhancements in energy efficiency, facilitated a shift from weak to strong decoupling post-2015. This transformation occurred not merely through emission control policies but through the gradual substitution of energy-intensive manufacturing with industries focused on technology and services. This is consistent with the negative coefficients for industrial structure and energy intensity observed in our regression analysis.
Conversely, Longshan County in Western Hunan represents a resource-dependent pathway with limited decoupling potential. Despite moderate economic growth, the county continued to exhibit weak decoupling throughout most of the study period. This outcome reflects the strong positive correlation between carbon intensity and emissions identified in the empirical analysis. The ongoing prevalence of coal-based energy and the lack of industrial diversity undermined the effectiveness of policy interventions, demonstrating that financial transfers alone are inadequate for achieving strong decoupling in resource-dependent counties without concurrent investments in alternative industries and energy infrastructure. Collectively, these case studies illustrate that county-level decoupling outcomes are influenced by structurally embedded development pathways. Counties engaged in industrial upgrading and efficiency improvements typically experience more pronounced decoupling effects, whereas those dependent on resources encounter path-dependent limitations. This insight enhances the existing body of decoupling literature by emphasizing that the efficacy of carbon mitigation strategies depends on local structural conditions. This finding underscores the importance of differentiated and place-based policy design in addressing the challenges faced by China’s counties.
Limitation and future expectations
Despite conducting a comprehensive assessment of county-level carbon decoupling patterns, driving mechanisms, and scenario-based peaking pathways in Hunan Province, it is important to acknowledge several limitations. These limitations also highlight promising avenues for future research.
Firstly, data-related limitations present a significant challenge at the county level. Carbon emissions and various socioeconomic indicators are compiled from multiple datasets, each differing in spatial resolution, statistical definitions, and reporting frequencies. Despite the application of extensive preprocessing, harmonization, and interpolation, residual inconsistencies remain. These may impact the precision of coefficient estimates and the robustness of scenario projections, particularly in counties with sparse or incomplete data records. Future studies would benefit from access to improved county-level energy and emission statistics, higher-resolution remote sensing products, and the integration of firm- or sector-specific data to enhance measurement accuracy.
Secondly, methodological limitations emerge from the reliance on linear econometric models and static interaction detection methods. Although the STIRPAT-based regression and GeoDetector analysis effectively identify dominant drivers and interaction effects, they do not account for dynamic feedbacks, nonlinear responses, or time-varying causal relationships among variables such as economic growth, industrial structure, energy use, and emissions. Future research should explore nonlinear modeling approaches, dynamic panel models, and causal inference frameworks, including difference-in-differences, instrumental variables, or machine learning-based causal methods. These approaches could provide deeper insights into the mechanisms underlying county-level decoupling.
Thirdly, scenario-related limitations arise from the simplifying assumptions used in our projection framework. The scenario simulations rely on stylized development pathways and do not account for potential abrupt technological breakthroughs, policy shifts, or external disturbances. Future work should aim to refine the scenario design by incorporating adaptive policy scenarios, stochastic elements, or system dynamics models. This would allow for a more realistic representation of the uncertainties and complexities associated with long-term low-carbon transitions. Addressing these limitations will not only enhance the robustness of county-level decoupling assessments but also improve the policy relevance of future studies, especially in supporting dynamic and evidence-based low-carbon governance at sub-provincial scales.
Conclusion
This study offers a county-level analysis of the decoupling relationship between carbon emissions and economic growth in Hunan Province from 2000 to 2022. Integrating methods such as decoupling diagnosis, econometric identification of key drivers, and forward-looking scenario simulation, the results demonstrate significant spatial heterogeneity in emission trajectories and decoupling transitions. Over the study period, an increasing number of counties demonstrated strong decoupling, suggesting that achieving a low-carbon transition at the county level is feasible but heavily dependent on specific local development conditions, industrial foundations, and patterns of energy use.
The findings of this study have several policy implications that are immediately applicable at the county level. Firstly, the pronounced impact of carbon intensity and its significant interactions with economic growth, industrial structure, and energy-use efficiency indicate that counties should avoid single-instrument approaches and should, instead, implement comprehensive policy packages. In counties with robust industrial bases and high urbanization rates (e.g., within major urban agglomerations), it is crucial to prioritize the acceleration of industrial upgrading towards activities that are high in value-added and low in emissions. These counties should also consider tightening entry standards for energy-intensive projects and enhancing energy efficiency through process optimization and increased electrification. In contrast, for counties that are resource-dependent or less developed, a different approach is required for decarbonization that includes stabilizing livelihoods and fiscal capacity, promoting industrial diversification, expanding distributed renewable energy sources and clean heating options where feasible, and employing targeted fiscal instruments (e.g., earmarked transfers, green credit, and performance-based subsidies) to mitigate the risk of carbon lock-in. Secondly, the scenario analysis suggests that peaking timelines are highly sensitive to chosen development pathways. Consequently, counties should incorporate carbon-peaking objectives into their annual and medium-term strategic planning, complete with measurable sector-specific roadmaps. These efforts should be supported by a robust county-level monitoring system that tracks carbon intensity and energy-use metrics. Thirdly, to address cross-county spillovers and interactive effects, county governments need to enhance regional coordination mechanisms. This could include collaborative planning for industrial layouts, shared energy infrastructure, and coordinated regulatory measures to prevent the displacement of high-emission activities from core to peripheral counties.
The empirical analysis conducted in a single province has implications that extend beyond Hunan, both in methodological and policy dimensions. This study presents a transferable county-level analytical framework to other provinces in China for identifying decoupling patterns, diagnosing dominant drivers, and testing alternative peaking pathways under various development strategies. This approach is particularly pertinent for provinces characterized by significant internal heterogeneity—where industrialized, agricultural, and resource-based counties coexist. This is because aggregated provincial targets may obscure local challenges and result in the implementation of inefficient “one-size-fits-all” policies. The typology-based interpretation emphasized in this study offers a logical framework for provinces to allocate responsibilities among counties: advanced counties might achieve earlier emission peaks through industrial upgrading and efficiency improvements; transitional counties could concentrate on structural adjustments and clean energy substitution; and resource-dependent counties may need coordinated support packages that include transformation finance, infrastructure investment, and employment-oriented diversification. Such differentiated governance is likely to enhance the effectiveness and equity of policies while mitigating the risk of carbon leakage across sub-provincial entities.
More broadly, this study provides insights applicable to other countries, particularly developing economies, where subnational heterogeneity and uneven governance capacities are prevalent. Many countries implement climate and energy policies at the local level, such as districts, counties, or municipalities, yet national-level indicators often fail to reflect place-based constraints like industrial specialization, energy access, and fiscal capacity. The county-level decoupling perspective advanced in this research demonstrates how detailed evidence can inform differentiated low-carbon transition strategies: (i) identifying locations where economic growth is still closely linked with emissions; (ii) prioritizing mechanisms such as intensity reduction, structural upgrading, and energy efficiency, tailored to local conditions; and (iii) designing scenario-based pathways that synchronize mitigation efforts with realistic development trajectories. Thus, although specific parameter values may vary according to context, the fundamental policy logic, “diagnose local coupling status, target key drivers, and sequence transition measures according to local development pathways”, is universally relevant to countries striving to balance growth with decarbonization amid limited resources and regional disparities.
Acknowledgements
We are gratefully acknowledge the support of the university and affiliated college throughout the completion of this research.
Author contributions
Conceptualization, Chaohui Zheng; methodology, Chaohui Zheng and Heng Wei; software, Chaohui Zheng; validation, Chaohui Zheng and Heng Wei; formal analysis, Chaohui Zheng; investigation, Chaohui Zheng; resources, Chaohui Zheng; data curation, Chaohui Zheng; writing original draft preparation, Chaohui Zheng; writing review and editing, Chaohui Zheng and Heng Wei; visualization, Chaohui Zheng; supervision, Chaohui Zheng; project administration, Chaohui Zheng; funding acquisition, not applicable. Chaohui Zheng and Heng Wei contributed equally to this work and share first authorship. All authors have read and agreed to the published version of the manuscript.
Funding
The author(s) declare that no financial support was received for the research, authorship, and/or publication of this article.
Data availability
All datasets used in this study are publicly available from open-access sources. County-level socioeconomic data for Hunan Province can be obtained from the Hunan Statistical Yearbook (https://tjj.hunan.gov.cn/hntj/tjfx/hntjnj/hntjnjwlb/index.html), the China County Statistical Yearbook (https://www.stats.gov.cn/), and official government statistical bulletins. Carbon emission data are available from the EDGAR Global Emissions Database (https://edgar.jrc.ec.europa.eu/). All data supporting the findings of this study are thus accessible through these open repositories.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Mirzabaev A, Bezner Kerr R, Hasegawa T, Pradhan P, Wreford A, Cristina Tirado Von Der Pahlen M, et al. Severe climate change risks to food security and nutrition. Clim Risk Manag. 2023;39:100473.
- 2.Herrador M, De Jong W, Nasu K, Granrath L. Circular economy and zero-carbon strategies between Japan and South Korea: a comparative study. Sci Total Environ. 2022;820:153274. [DOI] [PubMed] [Google Scholar]
- 3.Owen A, Barrett J. Reducing inequality resulting from UK low-carbon policy. Clim Policy. 2020;20:1193–208. [Google Scholar]
- 4.Szasz J. Which approaches to climate policy decrease carbon dioxide emissions? evidence from US states, 1997–2017. Energy Res Soc Sci. 2023;97:102969. [Google Scholar]
- 5.Wang Y, Guo C-H, Du C, Chen X-J, Jia L-Q, Guo X-N, et al. Carbon peak and carbon neutrality in China: goals, implementation path, and prospects. China Geol. 2021;4:1–27. [Google Scholar]
- 6.Huang C, Ren W, Fatima N, Zhu J. Carbon intensity constraint, economic growth pressure and China’s low-carbon development. J Environ Manage. 2023;348:119282. [DOI] [PubMed] [Google Scholar]
- 7.Benhelal E, Shamsaei E, Rashid MI. Challenges against CO2 abatement strategies in cement industry: a review. J Environ Sci. 2021;104:84–101. [DOI] [PubMed] [Google Scholar]
- 8.Kim H, McJeon H, Jung D, Lee H, Bergero C, Eom J. Integrated assessment modeling of Korea’s 2050 carbon neutrality technology pathways. Energy Clim Change. 2022;3:100075. [Google Scholar]
- 9.Ahmed J, Alvi AA. Balancing economic growth and environmental sustainability in developing countries: the role of financial innovation. J Energy Environ Policy Options. 2024;7:9–19. [Google Scholar]
- 10.Gbadeyan OJ, Muthivhi J, Linganiso LZ, Deenadayalu N. Decoupling economic growth from carbon emissions: a transition toward low-carbon energy systems—a critical review. Clean Technol. 2024;6:1076–113. [Google Scholar]
- 11.Leal Filho W, Ben Hassen T, Matandirotya N, Ng A. Empty promises: some requirements for a successful implementation of decarbonisation strategies in developing countries. Sci Total Environ. 2025;977:179409. [DOI] [PubMed] [Google Scholar]
- 12.Jia X, Zhang Y, Tan RR, Li Z, Wang S, Wang F, et al. Multi-objective energy planning for China’s dual carbon goals. Sustain Prod Consum. 2022;34:552–64. [Google Scholar]
- 13.Bianco V, Cascetta F, Nardini S. Analysis of the carbon emissions trend in European union: a decomposition and decoupling approach. Sci Total Environ. 2024;909:168528. [DOI] [PubMed] [Google Scholar]
- 14.Wang YQ, Huang HZ. Research on the decoupling effect, spatio-temporal characteristics, and driving factors of carbon emissions in China. Acta Ecol Sin. 2024;44:2228–43. [Google Scholar]
- 15.Watson D, Lander Svendsen N, Kaarsberg S, Värttö H. Is economic growth compatible with a sustainable Nordic future. Nordic Council of Ministers; 2021. [Google Scholar]
- 16.Wang K, Zhu Y, Zhang J. Decoupling economic development from municipal solid waste generation in China’s cities: assessment and prediction based on Tapio method and EKC models. Waste Manag. 2021;133:37–48. [DOI] [PubMed] [Google Scholar]
- 17.Xi Y, Yan D, Zhang J, Fu X. Decoupling analysis of the industrial growth and environmental pollution in the Circum-Bohai-Sea region in China. Environ Sci Pollut Res. 2021;28:19079–93. [DOI] [PubMed] [Google Scholar]
- 18.Luo X, Yin X, Li Y, Zhou X. Decoupling China’s tourism economy from carbon emissions through digitalization: a supply-side analytical framework. Sustainability. 2025;17:5183. [Google Scholar]
- 19.Xing L, Khan YA, Arshed N, Iqbal M. Investigating the impact of economic growth on environment degradation in developing economies through STIRPAT model approach. Renew Sustain Energy Rev. 2023;182:113365. [Google Scholar]
- 20.Wu R, Wang J, Wang S, Feng K. The drivers of declining CO2 emissions trends in developed nations using an extended STIRPAT model: a historical and prospective analysis. Renew Sustain Energy Rev. 2021;149:111328. [Google Scholar]
- 21.Liu XJ, Jin XB, Luo XL, Zhou YK. Multi-scale variations and impact factors of carbon emission intensity in China. Sci Total Environ. 2023;857:159403. [DOI] [PubMed] [Google Scholar]
- 22.He Y, Xing Y, Zeng X, Ji Y, Hou H, Zhang Y, et al. Factors influencing carbon emissions from China’s electricity industry: analysis using the combination of LMDI and K-means clustering. Environ Impact Assess Rev. 2022;93:106724. [Google Scholar]
- 23.Cao W, Liu Z, Zhou M, Gao R. Spatiotemporal correlation analysis between carbon emission intensity and intensive use level of construction land at county scale in Chongqing of China. Carbon Balance Manag. 2025;21:27. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Liu S, Jia J, Huang H, Chen D, Zhong Y, Zhou Y. China’s CO2 emissions: a thorough analysis of spatiotemporal characteristics and sustainable policy from the agricultural land-use perspective during 1995–2020. Land. 2023;12:1220. [Google Scholar]
- 25.Jia J, Xin L, Lu C, Wu B, Zhong Y. China’s CO2 emissions: a systematical decomposition concurrently from multi-sectors and multi-stages since 1980 by an extended logarithmic mean divisia index. Energy Strategy Rev. 2023;49:101141. [Google Scholar]
- 26.Liu X, Jin X, Luo X, Zhou Y. Quantifying the spatiotemporal dynamics and impact factors of China’s county-level carbon emissions using ESTDA and spatial econometric models. J Clean Prod. 2023;410:137203. [Google Scholar]
- 27.Liu C, Hu S, Wu S, Song J, Li H. County-level land use carbon emissions in China: spatiotemporal patterns and impact factors. Sustain Cities Soc. 2024;105:105304. [Google Scholar]
- 28.Huang H, Jia J, Chen D, Liu S. Evolution of spatial network structure for land-use carbon emissions and carbon balance zoning in Jiangxi province: a social network analysis perspective. Ecol Indic. 2024;158:111508. [Google Scholar]
- 29.An Y, Wen G, Fan M, Zhao P, Sun J, He M, et al. Towards sustainable urban development: decoding the spatiotemporal relationship between urban spatial structure and carbon emissions. Carbon Balance Manag. 2025;20:17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Huo T, Xu L, Feng W, Cai W, Liu B. Dynamic scenario simulations of carbon emission peak in China’s city-scale urban residential building sector through 2050. Energy Policy. 2021;159:112612. [Google Scholar]
- 31.Zhan J, Wang C, Wang H, Zhang F, Li Z. Pathways to achieve carbon emission peak and carbon neutrality by 2060: a CASE study in the Beijing-Tianjin-Hebei region. China Renew Sustain Energy Rev. 2024;189:113955. [Google Scholar]
- 32.Liu Z, Zhang J, Zhang P, Jiang L, Yang D, Rong T. Spatial heterogeneity and scenario simulation of carbon budget on provincial scale in China. Carbon Balance Manag. 2023;18:20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Zhang Y, Zhang Y, Chen W, Zhang Y, Quan J. Decomposition of driving factors and peak prediction of carbon emissions in key cities in China. Carbon Balance Manag. 2025;20:20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Dang H, Deng Y, Hai Y, Chen H, Wang W, Zhang M, et al. Integrating Geodetector and GTWR to unveil spatiotemporal heterogeneity in China’s agricultural carbon emissions under the dual carbon goals. Agriculture. 2025;15:1302. [Google Scholar]
- 35.Liu K, Dong S, Han M. Exploring the impact of green innovation on carbon emission intensity in Chinese metropolitan areas. Ecol Indic. 2023;156:111115. [Google Scholar]
- 36.Chen Y. Spatial autocorrelation equation based on Moran’s index. Sci Rep. 2023;13:19296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Wu Y, Yuan C, Liu Z, Wu H, Wei X. Decoupling relationship between the non-grain production and intensification of cultivated land in China based on Tapio decoupling model. J Clean Prod. 2023;424:138800. [Google Scholar]
- 38.Chertow MR. The IPAT equation and its variants. J Ind Ecol. 2000;4:13–29. [Google Scholar]
- 39.Li H, Mu H, Zhang M, Li N. Analysis on influence factors of China’s CO2 emissions based on path–STIRPAT model. Energy Policy. 2011;39:6906–11. [Google Scholar]
- 40.Pham NM, Huynh TLD, Nasir MA. Environmental consequences of population, affluence and technological progress for European countries: a Malthusian view. J Environ Manage. 2020;260:110143. [DOI] [PubMed] [Google Scholar]
- 41.Liddle B. Population, affluence, and environmental impact across development: evidence from panel cointegration modeling. Environ Model Softw. 2013;40:255–66. [Google Scholar]
- 42.Dietz T, Rosa EA. Rethinking the environmental impacts of population, affluence and technology. Hum Ecol Rev. 1994;1:277–300. [Google Scholar]
- 43.Ding Y, Li F. Examining the effects of urbanization and industrialization on carbon dioxide emission: evidence from China’s provincial regions. Energy. 2017;125:533–42. [Google Scholar]
- 44.Zhang YJ, Liu Z, Zhang H, Tan TD. The impact of economic growth, industrial structure and urbanization on carbon emission intensity in China. Nat Hazards. 2014;73:579–95. [Google Scholar]
- 45.Wang Y, Zhao T. Impacts of urbanization-related factors on CO2 emissions: evidence from China’s three regions with varied urbanization levels. Atmospheric Pollut Res. 2018;9:15–26. [Google Scholar]
- 46.Xing H, Gao D, Zhu L, Qiu P. Exploring spatiotemporal characteristics of land cover carbon emissions and their driving mechanisms in the Yellow River Basin: A multi-scale analysis from provincial-city-county level. Environ Dev Sustain. 2024:1–29.
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All datasets used in this study are publicly available from open-access sources. County-level socioeconomic data for Hunan Province can be obtained from the Hunan Statistical Yearbook (https://tjj.hunan.gov.cn/hntj/tjfx/hntjnj/hntjnjwlb/index.html), the China County Statistical Yearbook (https://www.stats.gov.cn/), and official government statistical bulletins. Carbon emission data are available from the EDGAR Global Emissions Database (https://edgar.jrc.ec.europa.eu/). All data supporting the findings of this study are thus accessible through these open repositories.