Abstract
Under mounting global pressure for carbon emission mitigation, China—currently the world's largest carbon emission contributor—confronts the critical challenge of reconciling emission reduction targets with sustained economic growth and progressive enhancement of citizens' living standards. Adopting a land-intensive utilization framework, this paper systematically investigates the relationship between urban intensive land use (UILU) and carbon emissions (CEs) in China's three major eastern urban agglomerations—the Beijing-Tianjin-Hebei (BTH), the Yangtze River Delta (YRD), and the Pearl River Delta (PRD). Methodologically, standard deviation ellipse analysis, the HR coordination model, and the Environmental Kuznets Curve (EKC) were employed to identify viable emission reduction pathways. Key findings from the 2005–2021 study period reveal four principal patterns: (1) UILU levels across the three urban agglomerations demonstrate an overall upward trajectory; Inter-agglomeration disparities exhibit phased transitional characteristics, following a “contraction-expansion-contraction” sequence, whereas intra-agglomeration disparities manifest persistent widening trends. Spatial analysis through standard deviation ellipses demonstrates distinct orientation patterns: the BTH exhibits a predominant northeast-southwest alignment; the YRD displays northwest-southeast orientation; and the PRD maintains an approximate east-west axis. (2) CEs across the three urban agglomerations demonstrated an overall increasing trajectory, stabilizing during the final phase of the study period. Spatiotemporal analysis revealed distinct orientation patterns: the BTH's standard deviation ellipse maintained a northeast-southwest alignment, while the YRD and PRD exhibited northwest-southeast orientations. (3) The HR coordination degree displayed moderate fluctuation with an overall ascending trajectory, demonstrating hierarchical coordination levels: YRD > PRD > BTH. (4) EKC analysis delineates distinct morphological patterns: The EKC curves for UILU and CEs in the BTH and PRD demonstrated an inverse N-shaped pattern, with turning points at 0.14 and 0.49 for the BTH, 0.13 and 0.44 for the PRD, respectively. In contrast, the EKC relationship for the YRD follows a U-shaped curve with a turning point at 0.20.
1 Introduction
The synergistic effects of global population expansion (projected to reach 9.7 billion by 2050), accelerated economic globalization, and exponential growth in energy consumption have established anthropogenic CO2 emissions as the predominant driver of contemporary climate change, accounting for 76% of total greenhouse gas emissions (; ). China's dual status as the world's second-largest economy (GDP $17.7 trillion in 2021) and foremost developing nation positions it at the epicenter of this environmental challenge, with its energy consumption constituting 26.5% of global totals and CO2 emissions reaching 10.28 gigatons in 2021-representing 31% of worldwide emissions and maintaining its position as the largest carbon emitter since 2006 (; ). At the 75th UN General Assembly, China formally committed to achieving carbon peaking by 2030 and carbon neutrality before 2060 (the “Dual Carbon” targets) through nationally determined contributions (; ; ). Nevertheless, structural inertia persists as fossil fuels (coal: 56.20%, oil: 17.90% in 2022) continue to dominate the national energy mix (74.10% of total consumption), creating substantial path dependence that constrains low-carbon transition and green economic development. This energy configuration necessitates unprecedented mitigation efforts, requiring an estimated annual CO2 reduction rate of 8–10% to compress the typical 60-year decarbonization cycle into a 30-year timeframe-a commitment demanding cumulative emission cuts of 100–120 gigatons CO2 equivalent by 2060.
As the fundamental substrate for human subsistence and development, terrestrial systems exert profound impacts on carbon dynamics through distinct land-use modalities. Land-use and land-cover change (LUCC) not only induces ecosystem structural transformations but also directly modulates biogeochemical cycles, particularly through carbon flux alterations. Empirical evidence demonstrates that deforestation, agricultural extensification, and urban sprawl collectively compromise carbon sequestration capacity. This is unequivocally supported by foundational research, which has established land-use change as the second-largest source of anthropogenic carbon emissions after fossil fuels (). Consequently, strategic land-use optimization employing spatial zoning models and intensive utilization technologies emerges as a critical pathway for achieving emission abatement targets while maintaining ecosystem service provisioning.
Building upon this recognition, scholars, governmental agencies, and social organizations have conducted extensive research on the relationship between land use and CEs. For instance, the aforementioned foundational study by Houghton—investigating nine global regions—revealed that land-use changes contributed 124 PgC to atmospheric increases from 1850 to 1990, equivalent to approximately half of the emissions from fossil fuel combustion during this period (). advanced this macro-scale understanding by demonstrating that land management explains 30–45% of the interannual variability in the global net land carbon balance (1959–2015), underscoring its persistent and significant role. An official assessment by China's Ministry of Natural Resources indicates that land-use change accounted for 30% of anthropogenic CEs during the 55-year period from 1950 to 2005 (). took the intensive use of cultivated land and agricultural CEs in the five provinces and regions of northwest China as the research object, and found that the coupling coordination showed a trend of rising and then decreasing from 2000 to 2015, and had entered the stage of decoupling development. applied DEA and multi-objective evaluation methods to assess spatial correlations between intensive cropland utilization and carbon emission efficiency across 138 counties (municipalities) in Hebei Province, demonstrating a predominant negative correlation. At the provincial scale, 's mechanism analysis revealed that construction land intensive use indices—particularly R&D (research and development) intensity, industrial structure upgrading, and green space intensity-positively correlated with carbon efficiency enhancement, whereas labor and energy intensity exerted negative inhibitory effects. Sun Yixuan conducted panel regression analyses on industrial land intensive utilization (ILIU) and carbon emission efficiency across 15 sub-provincial cities in China, identifying statistically significant impacts of ILIU. Specifically, human capital, fixed-asset investment, and R&D expenditure positively influenced efficiency gains, whereas energy intensity exhibited negative suppression effects (). Xie Hualin employed the STIRPAT model to analyze provincial-scale industrial land intensive utilization impacts on carbon emissions in China. The findings demonstrate that technological R&D investments exerted more pronounced decarbonization effects compared to labor and capital inputs across the eastern, central, and western regions (). Wu Haiyan employed a dual-effect fixed effects model to analyze the impact of UILU on carbon emission reduction at the provincial level in China during the period of 2008–2020. The findings indicated that UILU not only reduces carbon emissions but also enhances carbon emission efficiency, achieving carbon emission reduction from both quantity and quality perspectives (). Xiao Ling examined the influence of UILU on low-carbon transformation from a spatial spillover perspective and found that it exhibits significant spatial spillover effects. Specifically, UILU promotes low-carbon transformation in neighboring cities through three dimensions: land use intensity, economic and social efficiency, and ecological benefits (). To complement these macroscopic findings, research has progressively shifted to unraveling the underlying microscopic mechanisms. revealed that intensive land use alters soil environment (e.g., pH) and microbial physiological traits, thereby directly influencing the accumulation and decomposition of soil organic carbon. These studies have laid a solid theoretical foundation for improving land use efficiency and advancing pollution and carbon reduction. However, significant research gaps remain in two key areas: first, there is a lack of systematic revelation and comparative analysis of the spatiotemporal evolution patterns, regional differences, and complex interaction mechanisms between intensive land use and carbon emissions; second, how to achieve precise regulation and optimization of carbon emissions through the management of intensive land use remains a critical scientific challenge requiring urgent resolution.
Addressing these research gaps, this study selects China's three major eastern coastal urban agglomerations as empirical cases. These regions are representative and unique: they serve as leading economic powerhouses, yet face an acute contradiction. They are core zones for both energy consumption and carbon emissions—accounting for as much as 30.2% of the national total ()—while also exhibiting the highest levels of land use intensity alongside internal development imbalances and land resource wastage in certain cities (). We aim to develop a comprehensive evaluation system for intensive land use and utilize nighttime light data to estimate energy-related carbon emissions, thereby accurately quantifying the spatiotemporal dynamics of both variables. Furthermore, this research will employ spatial analysis techniques and econometric models to delve into the spatial heterogeneity of their interactions and systematically reveal the underlying multidimensional driving mechanisms. The findings will not only provide a solid theoretical basis and practical guidance for optimizing regional land use structures and implementing targeted carbon reduction strategies but will also offer scientifically grounded support for achieving sustainable land resource utilization and contributing to the national “Dual Carbon” strategic goals.
2 Materials and methods
2.1 Study area
This paper focuses on three major coastal urban agglomerations in eastern China: the BTH, the YRD, and the PRD. Specifically, the BTH comprises two direct-controlled municipalities (Beijing and Tianjin) and 11 prefecture-level cities in Hebei Province. The YRD is composed of Shanghai (a direct-controlled municipality), three sub-provincial cities (Nanjing, Hangzhou, and Ningbo), and 12 prefecture-level cities in Jiangsu and Zhejiang provinces. The PRD comprises two sub-provincial cities in Guangdong Province (Guangzhou and Shenzhen) and seven prefecture-level cities. All selected cities are core members of their respective urban agglomerations. Their spatial distribution is presented in Figure 1.
Figure 1
2.2 Data sources
The data used in this study primarily consist of four components: administrative boundary vector data, nighttime light imagery data, energy consumption data, and socio-economic statistical data. (1) The administrative boundary vector data were obtained from the 1:4,000,000 National Fundamental Geographic Information System of China. These boundaries were used to extract and clip all other raster and statistical data, ensuring spatial consistency; (2) The nighttime light imagery data were sourced from the National Oceanic and Atmospheric Administration's National Centers for Environmental Information (https://ngdc.noaa.gov/eog/index.html). Specifically, the study utilized two sensor types to ensure temporal coverage: the Version 4 Annual Composite of the DMSP-OLS data for the earlier period, and the Annual Composite data from the NPP-VIIRS sensor for the later period. These datasets were integrated and calibrated to construct a consistent, long-term nighttime light time series; (3) The energy consumption data were derived from the China Energy Statistical Yearbook for the years 2006–2022; (4) The socio-economic statistical data, crucial for constructing the intensive land use evaluation system and analyzing driving factors, were compiled from the China City Statistical Yearbook, the China Urban Construction Statistical Yearbook, and the China Statistical Yearbook for Regional Economy for the corresponding years. The key variables included, but were not limited to, output value of secondary and tertiary industries per square kilometer of land, total retail sales of consumer goods per square, population density, financial expenditure of unit land, number of hospital and clinic beds per 10,000 population, and green coverage rate of built-up areas.
2.3 Methods
2.3.1 CEs calculation methods
Currently, the estimation of CEs based on fossil energy consumption is widely used for carbon accounting in urban systems. Based on the IPCC Guidelines for national greenhouse gas inventories and related research findings (; ; ), this paper calculates energy-related CEs at the provincial level for the three major urban agglomerations. The calculation covers 10 types of energy, including coal, coke, crude oil, gasoline, kerosene, diesel, fuel oil, natural gas, electricity, and heat, with the calculation method as shown in Equation 1.
Where, C represents the total CEs from fossil fuels, measured in 10,000 tons; Ei denotes the consumption of energy type i, converted into standard coal equivalent, measured in 10,000 tons; and Ki represents the carbon emission coefficient of energy type i. The carbon emission coefficients for various energy types are listed in Table 1.
Table 1
| Types | Conversion factor to standard coal equivalent (104 tce•104 t−1) | CE coefficient (104 t•104 tce−1) |
|---|---|---|
| Coal | 0.7143 | 0.7559 |
| Coke | 0.9714 | 0.8550 |
| Crude oil | 1.4286 | 0.5857 |
| Gasoline | 1.4714 | 0.5538 |
| Kerosene | 1.4714 | 0.5714 |
| Diesel | 1.4571 | 0.5921 |
| Fuel oil | 1.4286 | 0.6185 |
| Natural gas | 1.3300 | 0.4483 |
| Electricity | 0.3450 | 0.2720 |
| Heat | 34.1200 | 0.6700 |
Conversion factors to standard coal equivalent and carbon emission coefficients for energy types.
The conversion factors to standard coal equivalent for natural gas, electricity, and heat are measured in kgce·m−3, kgce·(kW·h)−1, and kgce/(106 kJ), respectively.
First, the calibrated nighttime light data from 2005 to 2021 were used as the foundational dataset. The total digital number (DN) values of nighttime lights within the municipalities and provincial-level regions of the three major urban agglomerations were calculated and subjected to regression analysis with the corresponding annual carbon emission data. Due to the characteristic of a large number of zero values in the nighttime light data, to ensure that there are no CEs data in zero-value areas after downscaling, non-intercept linear regression analysis was adopted, as referenced in related studies (; ). The results of the regression analysis are shown in Figures 2–4. It can be seen that there is a strong fitting relationship between the total nighttime light intensity and carbon emissions in the three major urban agglomerations, and all models passed the significance test at the 0.001 level.
Figure 2
Figure 3
Figure 4
In order to further improve the accuracy of carbon emission estimation, a correction factor Km(t) (Equation 2) was introduced into the carbon emission inversion model. The values calculated annually based on nighttime light data were then multiplied by the correction factor Km(t), ensuring that the fitted values for year t matched the CEs calculated from statistical data. This approach ultimately yielded relatively high-precision carbon emission estimates for built-up areas ().
Km(t) represents the ratio between the CEs calculated from the energy consumption statistical data of province or municipality m in year t and the CEs simulated based on nighttime light DN values; CEm(t) denotes the CEs of province or municipality m in year t, calculated from the statistical energy yearbook; CE′(mt)x represents the CEs of province or municipality m in year t, estimated through the fitting of nighttime light data; C(mt)x indicates the CEs distributed at pixel x in year t; and CE′(mt)x represents the fitted CEs at pixel x in year t.
2.3.2 Construction of the evaluation index system
Drawing upon relevant research findings, this study selected 16 indicators from five dimensions: land input intensity, land use intensity, land economic benefit, land social benefit, and land ecological benefit (; ; ) to construct an evaluation system for UILU in the three major urban agglomerations (Table 2). When calculating the weight of evaluation indicators, the CRITIC method and entropy weight method were adopted to take the average value to more objectively reflect their true weight ().
Table 2
| Target layer | Criteria layer | Indicator layer | Weight |
|---|---|---|---|
| UILU | Land input intensity | Financial expenditure of unit land | 0.07 |
| Number of employed persons per unit land area | 0.07 | ||
| Proportion of construction land area to built-up area | 0.03 | ||
| Land use intensity | Population density | 0.06 | |
| Per capita construction land area | 0.06 | ||
| Per capita road area | 0.05 | ||
| Land economic benefit | Financial revenue of unit land | 0.08 | |
| Total retail sales of consumer goods per square | 0.07 | ||
| Total commodity sales of designated-size enterprises in wholesale and retail sectors per unit land area | 0.09 | ||
| Output value of secondary and tertiary industries per square kilometer of land | 0.07 | ||
| Land social benefit | Number of public library books per unit area | 0.08 | |
| Number of hospital and clinic beds per 10,000 population | 0.06 | ||
| Number of schools per square kilometer | 0.06 | ||
| Land ecological benefit | Green coverage rate of built-up areas | 0.04 | |
| Per capita park green space area | 0.06 | ||
| Industrial sulfur dioxide emissions per unit area | 0.04 |
Evaluation index system for UILU.
2.3.3 HR coordination degree model
When measuring the coordination relationship between UILU and CEs, most scholars adopt the coupling coordination degree theory from physics. However, the calculated values tend to have insufficiently dispersed distribution intervals and poor hierarchical differentiation, making it difficult to categorize their coordination types. This study references methodologies proposed by , , employing the HR coordination degree model to evaluate the relationship between UILU and CEs.
Where, HR denotes the coordination degree between UILU and CEs; Q represents the standard deviation of UILU and CEs; indicates the arithmetic mean of UILU and CEs. Therefore, the coefficient of variation (, also known as the discrete coefficient) can be formulated as:
To facilitate the study, it is necessary to ensure that the HR value remains within the range [0,1], thus requiring the coefficient of variation Cv to also lie within [0,1]. However, relevant research findings indicate that Cv does not necessarily conform to this interval. Consequently, a transformation of Equation 4 is required, as demonstrated in Equation 6:
3 Results and analysis
3.1 Analysis of spatiotemporal characteristics of UILU in the three major urban agglomerations
3.1.1 Analysis of temporal evolution characteristics
As shown in Figure 5, over the 17-year period from 2005 to 2021, the average UILU in the BTH, the YRD, and the PRD showed a continuous upward trend. The disparity in UILU among these urban agglomerations evolved through a three-phase trajectory: initial narrowing, subsequent expansion, and final convergence. Meanwhile, during the final phase of the study period, the YRD and PRD urban agglomerations demonstrated convergence trends with narrowing gaps.
Figure 5
To further analyze the distribution of UILU in the three major urban agglomerations, kernel density curves were introduced for five representative years (2005, 2010, 2015, 2020, and 2021). As revealed in Figure 6: (1) In terms of position, the kernel density curves exhibit an overall rightward shift, indicating a steady improvement in UILU. This trend is characterized by an expansion of high-value zones and contraction of low-value zones, suggesting a consistent increase in the overall UILU levels across eastern China's coastal urban agglomerations. (2) In terms of peak characteristics, the curves transitioned from narrow to broad peaks, accompanied by a gradual decline in kurtosis. This pattern reflects the Matthew Effect-a divergence where high UILU values intensify and low values further diminish, indicating increasing intra-agglomeration disparities. (3) In terms of shape, the BTH exhibits a bi-modal pattern, while the YRD shows a tri-modal distribution. Both display reduced main peak magnitudes, declining secondary peaks, and widening inter-peak distances—evidence of spatial polarization. Conversely, the PRD transitioned to a uni-modal pattern in the final phase, reflecting increased homogeneity and reduced polarization.
Figure 6
As illustrated in Figure 5, a synchronous decline in the level of UILU was observed across China's three major coastal urban agglomerations during the 2013–2014 period. This phenomenon can be attributed to the stringent macro-regulatory policies implemented during this phase to fundamentally transform land use patterns. The intensive introduction of core policies such as the National New Urbanization Plan (2014–2020) imposed strict controls on the scale of new construction land, directly constraining the extensive development model reliant on land element inputs. During this intense window of policy transition, regional fixed-asset investments closely tied to new construction land consequently slowed significantly, thereby exerting downward pressure on the “input intensity” dimension within the UILU evaluation system in the short term. Concurrently, structural adjustments aimed at phasing out inefficient production capacity and fostering new growth drivers inevitably induced what may be termed “transition pains”—a temporary “window of vacancy” emerged between the clearance of obsolete industrial land and the delayed full operational capacity of new, high-efficiency projects. This transitional gap further contributed to the decline in intensive use indicators reflecting “economic output efficiency.” Consequently, this phase of retraction in UILU does not signify the failure of intensive use policies but rather represents an inevitable transitional manifestation as the prior extensive growth model was decisively interrupted and the shift toward a high-quality intensive development pattern was initiated, thereby underscoring the profound impact of macro-level policies on regional land use practices.
3.1.2 Analysis of spatial evolution characteristics
As shown in Figure 7 and Table 3, the standard deviational ellipse of UILU in the BTH exhibits a predominant northeast-southwest orientation; the YRD demonstrates a northwest-southeast oriented standard deviational ellipse; while the PRD presents a nearly east-west directional pattern in its standard deviational ellipse.
Figure 7
Table 3
| Region | Year | X (m) | Y (m) | Long axes (m) | Short axes (m) | Ratio of long and short axes | Azimuth (°) |
|---|---|---|---|---|---|---|---|
| BTH | 2005 | 539703.30 | 4519983.80 | 108435.42 | 247676.88 | 2.28 | 33.81 |
| 2010 | 537588.85 | 4522802.65 | 113857.84 | 247182.11 | 2.17 | 33.32 | |
| 2015 | 538627.71 | 4526022.63 | 112405.57 | 240692.72 | 2.14 | 34.29 | |
| 2020 | 536202.01 | 4530224.14 | 116660.76 | 236617.69 | 2.03 | 34.41 | |
| 2021 | 537235.93 | 4530462.82 | 116619.78 | 236102.74 | 2.02 | 34.46 | |
| YRD | 2005 | 1445224.69 | 3403451.97 | 185346.20 | 100967.73 | 1.84 | 146.83 |
| 2010 | 1442620.01 | 3403272.17 | 187677.89 | 99313.43 | 1.89 | 147.97 | |
| 2015 | 1442824.35 | 3404828.85 | 186065.71 | 100460.03 | 1.85 | 148.88 | |
| 2020 | 1444475.20 | 3403034.28 | 186352.24 | 100010.35 | 1.86 | 147.46 | |
| 2021 | 1445632.53 | 3403495.36 | 186049.03 | 99585.07 | 1.87 | 147.04 | |
| PRD | 2005 | 873178.69 | 2415139.60 | 92899.30 | 68631.78 | 1.35 | 85.32 |
| 2010 | 872569.81 | 2414864.15 | 93380.80 | 68132.00 | 1.37 | 89.42 | |
| 2015 | 876453.30 | 2414861.57 | 92831.74 | 66666.76 | 1.39 | 91.00 | |
| 2020 | 872211.53 | 2415633.15 | 94610.53 | 67073.31 | 1.41 | 92.28 | |
| 2021 | 873882.12 | 2414335.12 | 91562.80 | 66229.59 | 1.38 | 89.67 |
Standard deviation ellipse parameters of UILU in the BTH, the YRD, and the PRD.
Centroid Variation of Standard Deviational Ellipse
① The centroid migration of the standard deviational ellipse in the BTH demonstrates a northwest-northeast-northwest-northeast directional shift. The northward movement during 2005–2021 indicates that the UILU level in the northern region surpassed that of the southern area. ② The YRD exhibits a southwest-northeast-southeast-northeast trajectory in the centroid movement of its standard deviational ellipse. ③ The PRD manifests a southwest-southeast-northwest-southeast directional pattern in the centroid displacement of its standard deviational ellipse.
Shape Variation of Standard Deviational Ellipse
① The ratio of major to minor semi-axes in the standard deviational ellipse of the BTH demonstrates a progressively decreasing trend year by year, indicating that the ellipse configuration is gradually approximating a circular form with diminishing directional orientation. ② The YRD exhibits an initial increase followed by a decrease in the major-minor axis ratio of its standard deviational ellipse, while maintaining relatively stable directional characteristics overall. ③ Throughout the study period, the PRD manifests minimal variation in the major-minor axis ratio of its standard deviational ellipse.
Azimuthal Variation of Standard Deviational Ellipse
① The azimuth angle of the standard deviational ellipse for UILU in the BTH exhibited a minimal clockwise increase of merely 0.65° over the 17-year period, indicating relatively stable spatial distribution patterns oriented along the “northeast-southwest” axis. ② The YRD demonstrates an “initial increase followed by decrease” trend in the azimuth angle of its standard deviational ellipse for UILU, suggesting progressive improvements in land use efficiency within northwestern cities during later phases. ③ Similarly, the PRD manifests an “initial increase followed by decrease” pattern in the azimuth angle variation of its standard deviational ellipse for UILU.
3.2 Analysis of spatiotemporal characteristics of CEs in the three major urban agglomerations
3.2.1 Analysis of temporal evolution characteristics
As shown in Figure 8, the CEs of the three major urban agglomerations have demonstrated an overall upward trend from 2005 to 2021. Over the 17-year period, the total CEs reached 1,477.37 million tons in the BTH, 1,565.29 million tons in the YRD, and 812.76 million tons in the PRD. Regarding growth rates, the hierarchy followed the order: YRD > PRD > BTH. This indicates that the YRD exhibited both the highest cumulative emissions and the most rapid growth trajectory during the study period. This is primarily attributable to the YRD's status as China's largest and most energy-intensive manufacturing base, where its massive economic scale and industrial structure drive higher energy consumption demand.
Figure 8
From the temporal characteristics perspective, the evolution of CEs in the three major urban agglomerations can generally be divided into three phases: 2005–2010 was the rapid growth phase of CEs, particularly in 2009 when the YRD temporarily surpassed the BTH to become the largest emitter among the three. This phase of rapid growth is closely linked to China's deeper integration into the global industrial chain after its WTO accession, particularly the rapid development of heavy and chemical industries. 2011–2017 constituted the second phase characterized by fluctuating upward trends, during which total emissions continued to rise overall but exhibited intermittent declines at various time points across all three regions. These short-term declines and fluctuations are likely associated with the intervention of targeted emission reduction policies, such as the binding energy intensity targets implemented during the “12th Five-Year Plan” period. 2018–2021 marked the third phase of stable transition, where none of the three urban agglomerations experienced significant emission increases, maintaining relatively stable low-carbon emission levels. This suggests that under the top-level design of the “Ecological Civilization Construction” national strategy, a preliminary decoupling between economic development and carbon emissions has begun to emerge in the three regions.
3.2.2 Analysis of spatial evolution characteristics
As shown in Figure 9, from 2005 to 2021, the BTH exhibited an overall northeast-southwest spatial orientation, while the YRD and PRD demonstrated northwest-southeast spatial orientations.
Figure 9
In particular, the major semi-axis of the standard deviation ellipse for the BTH exhibited an “increase-decrease” evolutionary trend, indicating that CEs transitioned from a dispersed state to an aggregated state along the northeast-southwest principal axis. Similarly, the minor semi-axis demonstrated an analogous increase-decrease pattern, revealing that carbon emissions displayed a dispersing tendency along the northwest-southeast axis during 2005–2015, while gradually transitioning toward aggregation after 2015.
The major semi-axis of the standard deviation ellipse for the YRD exhibited a “decrease-increase” variation trend, suggesting that CEs shifted from an aggregated state to a dispersed state along the northwest-southeast principal axis. The minor semi-axis similarly demonstrated an identical decrease-increase pattern, revealing that CEs tended to cluster along the northeast-southwest axis during 2005–2020, while transitioning toward dispersion after 2021.
The PRD's standard deviation ellipse major semi-axis showed a decrease-increase variation trend, indicating the transition of carbon emissions from aggregation to dispersion along the northwest-southeast principal axis. The minor semi-axis displayed a progressive decreasing trend, demonstrating the persistent clustering tendency of CEs along the northeast-southwest axis.
3.3 Analysis of HR coordination degree between UILU and CEs
3.3.1 Temporal evolution characteristics
As evidenced by Figures 10–12, the HR coordination degree between UILU and CEs in the three major urban agglomerations demonstrated a fluctuating upward trend from 2005 to 2021. Specifically, the inter-city disparities in HR coordination exhibited a gradual narrowing trend over the study period. In terms of multi-year mean values, the YRD surpassed both the PRD and BTH in HR coordination levels. This spatial pattern indicates that the YRD has achieved a more balanced development between land intensification and carbon emission control. Its advanced stage of development likely enables the simultaneous application of stringent environmental regulations and efficient spatial planning, allowing economic agglomeration and decarbonization to progress in tandem.
Figure 10
Figure 11
Figure 12
To further analyze the spatial distribution patterns of HR coordination in the three major urban agglomerations, kernel density curves were constructed for five representative temporal nodes (2005, 2010, 2015, 2020, and 2021):
(1) Spatial positioning perspective: The kernel density curves of all three urban agglomerations exhibited rightward shifts, with expanding high-value zones and contracting low-value zones of HR coordination, indicating a steady enhancement trend in coordination levels.
(2) Peak characteristics: The original narrow peaks transitioned to broader peaks, demonstrating a “decline-rise-decline” pattern in kurtosis. Concurrently, the right-tailed elongation phenomenon intensified annually, revealing the emergence of Matthew effects in HR coordination development.
(3) Morphological features: All urban agglomerations maintained bimodal distributions, where the primary peak amplitude decreased significantly while the secondary peak showed moderate decline. The persistent spatial separation between primary and secondary peaks suggests potential polarization phenomena within each agglomeration's spatial configuration.
3.3.2 General trends and spatial heterogeneity characteristics
As illustrated in Figure 13, the HR coordination degree between UILU and CEs in the BTH exhibited a characteristic inverted U-shaped curve. The spatial differentiation pattern manifests distinct directional features: (1) East-west axis: A pronounced “east-high, west-low” spatial gradient emerges, where eastern regions encompassing Beijing, Tianjin, and Tangshan-characterized by advanced economic development and high urbanization levels-demonstrate significantly superior HR coordination compared to western areas like Zhangjiakou and Baoding. (2) North-south axis: A complementary “north-high, south-low” spatial configuration is observed, indicating systematically higher HR coordination levels in northern sectors of the agglomeration relative to southern counterparts.
Figure 13
As shown in Figure 14, the spatial distribution of HR coordination degree between UILU and CEs in the YRD can be mainly divided into two phases. The first phase was 2005–2015, which showed a “high in the west and low in the east” pattern in the east-west direction with a linear-like trend line; this was mainly because the western regions concentrated areas with relatively high HR coordination degree such as Nanjing and Hangzhou, while the eastern regions like Zhoushan and Ningbo had relatively low HR coordination degree during this phase. In the north-south direction, it presented a spatial differentiation pattern of “high in the north, low in the south, with central regions higher than both ends.” The second phase was 2020–2021, during which the HR coordination degree in the east-west direction exhibited a “high in the east and low in the west” spatial differentiation pattern, while maintaining the “north-high, south-low, central superiority” spatial configuration in the north-south direction.
Figure 14
As shown in Figure 15, the HR coordination degree between UILU and CEs in the PRD exhibits an inverted U-shaped curve relationship. The east-west spatial differentiation demonstrates a “higher in the east, lower in the west” pattern, where the eastern region (concentrating cities like Guangzhou and Shenzhen, early implementers of reform and opening-up policies) shows relatively higher UILU levels and lower CEs. The north-south spatial distribution displays a “central peak” configuration, with the northern area exceeding the southern sector. The central region clusters economically developed and highly urbanized areas-including Guangzhou, Shenzhen, Dongguan, and Foshan-where the HR coordination between UILU and CEs correspondingly reaches higher levels.
Figure 15
3.4 EKC curve relationship
3.4.1 Variable unit root and cointegration tests
In analyzing the EKC relationship between UILU and CEs, this study employs a bivariate model specification. This methodological choice is based on the following rationale: the UILU composite index constructed for this study is inherently a multidimensional concept, whose evaluation system has systematically incorporated core dimensions such as economic development level (e.g., output value of secondary and tertiary industries per unit area) and social input intensity (e.g., fiscal expenditure per unit area; see Table 2). Consequently, the UILU index serves as an integrated proxy that embodies multiple key drivers. Introducing control variables (e.g., industrial structure, economic development level) that are conceptually and empirically highly correlated with the constituents of the UILU index into the regression would lead to severe multicollinearity problems, thereby compromising the validity of the model estimates. Thus, the EKC analysis in this section aims to explore the macro-level statistical relationship between this comprehensive intensive land use level and carbon emissions, an approach that is logically coherent and methodologically robust.
As shown in Table 4, the statistical values of UILU and CEs are all below their critical values at the 1% significance level, passing the panel unit root test, indicating that they are stationary series. Building on this foundation, the Kao panel cointegration test was conducted. The test results presented in Table 5 demonstrate the existence of a cointegration relationship between UILU and CEs, providing a foundation for subsequent statistical analysis and model construction.
Table 4
| Variable | Test method | Three major urban agglomerations | BTH | YRD | PRD |
|---|---|---|---|---|---|
| UILU | LLC | −7.098*** | −3.819*** | −3.864*** | −4.346*** |
| IPS | −2.844*** | −1.343** | −1.115*** | −2.811*** | |
| CEs | LLC | −16.126*** | −6.378*** | −9.600*** | −7.861*** |
| IPS | −12.720*** | −4.282*** | −8.075*** | −7.169*** |
Results of panel unit root tests.
***, **, and * denote statistical significance at the 0.01, 0.05, and 0.1 levels, respectively.
Table 5
| Region | UILU and CEs | ||
|---|---|---|---|
| Test | t-statistic | p-value | |
| BTH | Dickey-Fuller | 1.898 | 0.029** |
| YRD | Dickey-Fuller | −1.512 | 0.065* |
| PRD | Dickey-Fuller | 1.683 | 0.046** |
Statistical results of the Kao test for panel data of the three major urban agglomerations.
The null hypothesis posits that no cointegration relationship exists between ILU and carbon emissions; ***, **, and * denote statistical significance at the 0.01, 0.05, and 0.1 significance levels, respectively.
3.4.2 EKC curve relationship analysis
This study employs the EKC framework by introducing a cubic function to empirically analyze the EKC relationship between UILU and CEs across three major urban agglomerations. With the cubic function as the a priori specified model, when regression results demonstrated poor goodness-of-fit, a quadratic function was re-estimated for comparative analysis (Table 6).
Table 6
| Specification | β1 | β2 | β3 | Curvature pattern |
|---|---|---|---|---|
| 1 | + | 0 | 0 | Positive correlation |
| – | 0 | 0 | Negative correlation | |
| 2 | – | + | 0 | U-shaped |
| + | – | 0 | Inverted U-shaped | |
| 3 | + | – | + | N-shaped |
| – | + | – | Inverted N-shaped |
Heterogeneous EKC relationship patterns.
In the equation, the symbols “+” and “–” denote coefficient signs (>0 and < 0, respectively).
Where, β1, β2, β3, are coefficients of the linear, quadratic, and cubic terms, respectively, indicates the constant term; ϵ signifies the error term.
Regression analyses were performed using Eviews 12.0 on 17-year panel data from the three major urban agglomerations. The models demonstrated satisfactory goodness-of-fit, with adjusted R2 values of 0.72, 0.78, and 0.73, respectively.
For the BTH, the EKC model between UILU and CEs is specified as:
According to Figure 16, the fitted curve diagram of UILU and carbon emissions in the BTH reveals an inverted N-shaped EKC relationship between these two variables. During the UILU intensity range of 0.14–0.49, the EKC manifests a positive growth phase, suggesting that CEs increase rapidly with the enhancement of land use intensification. When land use intensity falls below 0.14 or exceeds 0.49, the EKC demonstrates negative growth rates, indicating effective emission control through land intensification with CEs showing a gradual declining trend.
Figure 16
For the YRD, the EKC model between UILU and CEs is specified as:
As shown in Figure 17, the EKC relationship between UILU and CEs in the YRD exhibits a U-shaped pattern. When the UILU level is below 0.20, the growth rate of the Kuznets curve remains negative. However, once the UILU level exceeds 0.20, CEs demonstrate an increasing trend with the enhancement of land use intensity. This phenomenon can be attributed to the fact that although improved land use intensification has partially restrained the unrestrained expansion of construction land, the CEs generated by increased inputs of labor, capital, and energy in construction land utilization significantly outweigh the carbon sequestration benefits derived from reducing the conversion of carbon sink lands (including cultivated land, woodland, and grassland) to carbon source lands. Consequently, the overall carbon emissions manifest an upward trajectory during this phase.
Figure 17
The EKC Curve Model of UILU and CE in the PRD:
As shown in Figure 18, the fitted curve between UILU and CEs in the PRD demonstrates an inverted-N-shaped EKC relationship. Within the land use intensity range of 0.13–0.44, the EKC manifests a positive growth phase. Beyond the inflection point at 0.44, three synergistic drivers emerge: (1) accelerated urbanization, (2) efficiency gains from technological advancements, and (3) heightened environmental awareness. These factors collectively facilitate deliberate emission reduction strategies. Concurrently, constrained by the law of diminishing marginal returns, resource inputs (human capital and financial investments) reach saturation thresholds at this intensification stage. This dynamic effectively curtails agricultural land conversion to non-agricultural uses, while carbon sink land uses (cultivated fields, woodlands, and grasslands) dominate the land use structure, ultimately driving CEs into a deceleration phase.
Figure 18
4 Discussion
Research on the relationship between UILU and CEs has predominantly focused on economic perspectives. This paper, however, adopts a geographical lens to analyze the spatiotemporal evolution, HR coordination, and EKC relationship between UILU and CEs in three major coastal urban agglomerations of eastern China. This approach provides new insights into understanding the dynamic interplay between these two factors. First, UILU in the three major urban agglomerations showed an upward trend during the study period, while CEs followed an initial increase followed by a subsequent decline. This suggests that advancements in UILU technologies have enhanced output efficiency per unit land area, enabling the maintenance of equivalent economic output while reducing the demand for land development. Consequently, this has mitigated CEs associated with land use changes. During the study period, the standard deviation ellipses of UILU and CEs in the BTH urban agglomeration exhibited a predominant northeast-southwest orientation, while those in the YRD and the PRD urban agglomerations showed a northwest-southeast orientation. Additionally, the major and minor axes of these ellipses demonstrated distinct evolutionary trends. These spatial patterns are closely linked to the spatiotemporal heterogeneity in land use practices, economic development levels, and urbanization stages among cities within each urban agglomeration. Second, the HR coordination between UILU and CEs in the three major urban agglomerations has progressively improved. The YRD urban agglomeration exhibits a positive coordination relationship between UILU and CEs, attributable to its advanced economic development level, which provides a material foundation for technological innovation. Concurrently, innovative policy instruments such as intelligent spatial planning and optimization of UILU standards, combined with market-based mechanisms like ecological compensation systems, have systematically enhanced land use efficiency and constrained unplanned urban expansion. These efforts have ultimately achieved a dynamic equilibrium between land development and ecological conservation, offering spatial support for regional low-carbon transitions. Regarding the EKC relationship, the three major coastal urban agglomerations in eastern China exhibit distinct curve relationships between UILU and CEs. CEs display varying trends when the intensity of land use reaches different thresholds. Overall, in the early stages, improvements in UILU are often accompanied by more advanced industrial production methods and rational allocation of land resources, leading to reductions in energy consumption and carbon emissions per unit of GDP. However, during the mid-term stage of UILU, over exploitation of resources and unsustainable land use practices may emerge, leading to an increase in CEs. In the later stage of UILU, further technological advancements-such as the adoption of eco-friendly and efficient technologies, stricter carbon emission standards and taxation policies, and heightened environmental awareness among the public and enterprises-result in a renewed downward trend in carbon emissions. This pattern parallels the EKC observed in the relationship between economic development and environmental quality, demonstrating that intensive land use can, to a certain extent, facilitate carbon reduction.
The findings of this paper align with partial conclusions from existing literature but also reveal discrepancies. For example, , , and have identified an EKC relationship between UILU, urbanization, and CEs. Specifically, Jingru Zhou and Tong Wang focused on the Guanzhong urban agglomeration and the BTH urban agglomeration, respectively. Their results are consistent with the findings of this study for the BTH and PRD urban agglomerations, where the EKC relationship between UILU and CEs follows an inverted-N-shaped curve. In contrast, Miao Zhang analyzed 30 provinces across China and concluded that the EKC relationship exhibits an inverted-U-shaped curve. These differences may stem from variations in the study regions, time spans, or the intensity of policy interventions.
The findings from China's coastal agglomerations provide transferable insights for rapidly urbanizing regions globally, particularly in Southeast Asia, Latin America, and Africa. The core lesson is that the carbon reduction effect of UILU is not automatic; it is contingent on how the intensity is achieved. Simply increasing density without concurrent improvements in urban spatial structure and the efficiency of the economic activities housed on the land may fail to yield carbon benefits or could even backfire, as suggested by the initial upward trend in the EKC curve. Therefore, the policy implication extends beyond mere compactness: it necessitates integrated planning that links land use intensity directly with low-carbon industrial structures, energy efficiency standards, and sustainable transportation systems. This integrated approach, highlighted by the Chinese case, is crucial for any large urban agglomeration aiming to reconcile economic agglomeration with climate goals.
Therefore, future research could expand the geographical scope to test the generalizability of the findings across diverse regional contexts and incorporate additional variables—including dynamic policy frameworks, innovation-driven technological progress, and sectoral restructuring—into the analytical models to enable a more comprehensive understanding of the mechanisms linking UILU and CEs.
5 Conclusion
This study provides empirical evidence that the relationship between UILU and CEs varies significantly across different regional contexts. Our analysis of China's three major coastal urban agglomerations reveals two distinct patterns: while the BTH and PRD exhibit an inverted-N-shaped curve, the YRD follows a U-shaped trajectory.
These divergent pathways challenge the assumption that land intensification uniformly leads to carbon reduction. The findings indicate that the environmental outcome of compact urban development is mediated by region-specific factors, including industrial composition, energy infrastructure, and developmental priorities. The evidence suggests that the carbon reduction effect of land use intensification is conditional rather than automatic.
This understanding carries a clear and important implication for urban planning: effective low-carbon development requires moving beyond standardized intensification policies. Instead, planners and policymakers should develop tailored approaches that account for local economic and environmental conditions. For some regions, this may mean coordinating UILU with energy structure optimization; for others, it may require greater emphasis on industrial upgrading and technological innovation.
Ultimately, this research confirms that sustainable urban development demands contextualized strategies. The pursuit of compact cities must be coupled with a nuanced understanding of how regional characteristics influence the relationship between spatial organization and environmental performance.
Statements
Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.
Author contributions
QZ: Funding acquisition, Investigation, Methodology, Project administration, Writing – original draft, Writing – review & editing. ZH: Data curation, Software, Validation, Writing – original draft. XJ: Data curation, Formal analysis, Writing – original draft. WZ: Data curation, Methodology, Validation, Writing – original draft. AG: Data curation, Resources, Software, Writing – original draft. JG: Conceptualization, Funding acquisition, Supervision, Writing – original draft.
Funding
The author(s) declared that financial support was received for this work and/or its publication. Research Project Supported by the 2025 Annual Basic Scientific Research Operating Funds of Provincial Higher Education Institutions in Hebei GEO University (Funding Number: QN25024).
Acknowledgments
The authors are thankful to the editor and the reviewers for their valuable comments and suggestions for upgrading the quality of this paper.
Conflict of interest
The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
Generative AI statement
The author(s) declared that generative AI was not used in the creation of this manuscript.
Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.
Publisher’s note
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.
References
1
BatranceaL. M.NichitaA.TulaiH.RusM. I.MascaE. S. (2025). Fueling economies through credit and industrial activities. A way of financing sustainable economic development in Brazil. Green Finance7:24. doi: 10.3934/GF.2025002
2
CaiX.LiangY.HuangZ.GeJ. (2021). Spatiotemporal pattern and coordination relationship between urban residential land price and land use intensity in 31 provinces and cities in China. PLoS ONE16:e0254846. doi: 10.1371/journal.pone.0254846
3
ChenZ.GuoJ.LiW.JiaH.LiangX. (2023). Evaluating emission reduction potential at the “30-60 dual carbon targets” over China from a view of wind power under climate change. Sci. Total Environ.900:165782. doi: 10.1016/j.scitotenv.2023.165782
4
ChhabraV.SreethuS.KaurG.SinghA.KaurM.SiddiquiM. H.et al. (2025). Response of wheat crop to water-logged conditions under different land configurations and nutrient management. Sci. Rep.15:7168. doi: 10.1038/s41598-024-83752-2
5
China Internet Information Center (2022). Environmental Quality and Challenges of China's Urban Agglomerations: The Environment Dimension of the Comprehensive Development Index of Chinese Cities 2022. Available online at: http://www.china.com.cn/opinion/think/2023-12/22/content_116897480.shtml (accessed November 10, 2025).
6
DuH. B.WeiW.ZhangX. Y.JiX. P. (2021). Spatio-temporal evolution and influencing factors of energy-related carbon emissions in the Yellow River Basin: based on the DMSP/OLS and NPP/VIIRS nighttime light data. Geogr. Res. 40, 2051–2065.
7
GuoY.CaoX. Z.WeiW. D.ZengG. (2022). The impact of regional integration in the Yangtze River Delta on urban carbon emissions. Geogr. Res. 41, 181–192.
8
HouJ. C.WanM. H.SongM. L. (2024). Carbon emission inequality and fairness from energy consumption by prefecture-level cities in China. Ecol. Indic. 158:111364. doi: 10.1016/j.ecolind.2023.111364
9
HoughtonR. A. (1999). The annual net flux of carbon to the atmosphere from changes in land use 1850-1990*. Tellus B51, 298–313. doi: 10.3402/tellusb.v51i2.16288
10
JiangG.JiW.TianY.LuoY.MaW. (2025). Compact and livable? Identifying and managing the relationships between intensive land use and urban livability: evidence from 337 cities in China. Ecol. Econ.232:108562. doi: 10.1016/j.ecolecon.2025.108562
11
KöneA. Ç.BükeT. (2019). Factor analysis of projected carbon dioxide emissions according to the IPCC based sustainable emission scenario in Turkey. Renew. Energy133, 914–918. doi: 10.1016/j.renene.2018.10.099
12
LiS. Q.ShenJ. Q.WuR.HaX.FuhuaS. (2024). An integrated accounting system of the economic-social-ecological framework for assessing the value of intensive land use: a case study of the Taihu Lake governance project. Ecol. Indic. 158:111506. doi: 10.1016/j.ecolind.2023.111506
13
LingX.GaoY.WuG. (2023). How does intensive land use affect low-carbon transition in China? New evidence from the spatial econometric analysis. Land12:1578. doi: 10.3390/land12081578
14
LyuW.HuY.LiuJ.ChenK.LiuP.DengJ.et al. (2024). Impact of battery electric vehicle usage on air quality in three Chinese first-tier cities. Sci. Rep.14:21. doi: 10.1038/s41598-023-50745-6
15
MalikA. A.PuissantJ.BuckeridgeK. M.GoodallT.JehmlichN.ChowdhuryS.et al. (2018). Land use driven change in soil pH affects microbial carbon cycling processes. Nat. Commun.9:3591. doi: 10.1038/s41467-018-05980-1
16
NiuD.WangK.WuJ.SunL.LiangY.XuX.et al. (2020). Can China achieve its 2030 carbon emissions commitment? Scenario analysis based on an improved general regression neural network. J. Clean. Prod.243:118558. doi: 10.1016/j.jclepro.2019.118558
17
QuF. T.LuN.FengS. Y. (2011). Effects of land use change on carbon emissions. China Popul. Resour. Environ. 21, 76–83.
18
ShahW. U. H.HaoG.YanH.ZhuN.YasmeenR.DincăG. (2024). Role of renewable, non-renewable energy consumption and carbon emission in energy efficiency and productivity change: evidence from G20 economies. Geosci. Front.15:101631. doi: 10.1016/j.gsf.2023.101631
19
SongD. L.LuW. Y. (2023). Multi-scale analysis of spatial and temporal evolution of carbon emissions in urban agglomerations in the Yellow River Basin. J. N. China Univ. Water Resour. Electr. Power44, 1–12.
20
SunY. B.SuX. Z.XuG. (2016). Correlation between levels of cultivated land intensive use and carbon emission efficiency in Hebei Province. Trans. Chin. Soc. Agric. Eng. 32, 258–267.
21
SunY. X.ChengY.ZhangH. S. (2020). Study on the impact of intensive use of urban industrial land on carbon emission efficiency in China's 15 sub-provincial cities. Resour. Environ. Yangtze Basin29, 1703–1712.
22
WangJ.XueD. Q.MaB.SongY. (2019). Study on spatial-temporal coupling relationship of cultivated land intensive use and agricultural carbon emission in five provinces of Northwestern China. Environ. Sci. Technol. 10, 211–217.
23
WangR. Y.MoX.JiH.ZhuZ.WangY. S.BaoZ.et al. (2024). Comparison of the CASA and InVEST models' effects for estimating spatiotemporal differences in carbon storage of green spaces in megacities. Sci. Rep.14:5456. doi: 10.1038/s41598-024-55858-0
24
WilliamsR. G.RoussenovV.GoodwinP.ResplandyL.BoppL. (2017). Sensitivity of global warming to carbon emissions: effects of heat and carbon uptake in a suite of earth system models. J. Clim.30, 9343–9363. doi: 10.1175/JCLI-D-16-0468.1
25
WuH.QiuY.YinL.LiuS.ZhaoD.ZhangM. (2022). Effects of China's land-intensive use on carbon emission reduction: a new perspective of industrial structure upgrading. Front. Environ. Sci.10:1073565. doi: 10.3389/fenvs.2022.1073565
26
WuN.ShenL.ZhongS. (2019). Spatio-temporal pattern of carbon emissions based on nightlight data of the Shanxi-Shaanxi-Inner Mongolia Region of China. J. Geo-Inf. Sci. 21, 1040–1050.
27
WuT. (2020). A study on the relationship between intensive land use and carbon emissions: a case study of Beijing-Tianjin-Hebei urban agglomerat (Master's thesis). China University of Geosciences, Beijing, China.
28
XieH.ZhaiQ.WangW.YuJ. (2018). Does intensive land use promote a reduction in carbon emissions? Evidence from the Chinese industrial sector. Resour. Conserv. Recycl. 137, 167–176. doi: 10.1016/j.resconrec.2018.06.009
29
YuanK. H.MeiY.ChenY. R.LanM. T. (2017). Temporal and spatial evolution and influencing mechanism of construction land intensive utilization on carbon emissions efficiency in China. Resour. Sci. 39, 1882–1895.
30
YueC.CiaisP.HoughtonR. A.NassikasA. A. (2020). Contribution of land use to the interannual variability of the land carbon cycle. Nat. Commun. 11:3170. doi: 10.1038/s41467-020-16953-8
31
ZhangM.LiuX.PengS. (2023). Effects of urban land intensive use on carbon emissions in China: spatial interaction and multi-mediating effect perspective. Environ. Sci. Pollut. Res.30, 7270–7287. doi: 10.1007/s11356-022-22693-7
32
ZhangQ.GeJ.LiangY.ZhangM.DongL.ZhangJ. (2022). Does intensive land use decoupled from carbon emissions? an empirical study from the three grand economic zones of China. Front. Environ. Sci.10:941177. doi: 10.3389/fenvs.2022.941177
33
ZhouJ. R.ZhaoH. F.WuJ. H. (2017). Study on relationship and evolution characteristics of intensive land-use and carbon emission of guanzhong urban agglomeration. China Land Sci. 31, 55–61.
34
ZhouY.ZhangE.HeL.KeX.LinA.LaiX. (2024). The carbon emission reduction benefits of the transformation of the intensive use of cultivated land in China. J. Environ. Manage.370:122978. doi: 10.1016/j.jenvman.2024.122978
35
ZhouZ.KongX. Z. (2015). Pattern and influencing factors of coordinated implementation of “four tasks” in China: perspective of agricultural modernization. China Soft Sci. 30, 9–26.
Summary
Keywords
urban intensive land use, carbon emissions, coastal urban agglomerations, spatiotemporal dynamics, environmental Kuznets curve
Citation
Zhang Q, Huang Z, Jiao X, Zhang W, Guo A and Ge J (2026) Evolutionary characteristics of the relationship between urban intensive land use and carbon emissions in China's three major coastal urban agglomerations. Front. Sustain. Cities 8:1712186. doi: 10.3389/frsc.2026.1712186
Received
24 September 2025
Revised
11 November 2025
Accepted
14 January 2026
Published
06 February 2026
Volume
8 - 2026
Edited by
Peng Cheng, Wuhan Institute of Technology, China
Reviewed by
Shanzhong Qi, Shandong Normal University, China
Salahuddin M. Jaber, Hashemite University, Jordan
Updates
Copyright
© 2026 Zhang, Huang, Jiao, Zhang, Guo and Ge.
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
*Correspondence: Qingxi Zhang, qxz2022@163.com; Jingfeng Ge, gejingfeng@hebtu.edu.cn
Disclaimer
All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article or claim that may be made by its manufacturer is not guaranteed or endorsed by the publisher.