Spatio-temporal variations of the land-use-related carbon budget in Southeast China: the evidence of Fujian province

The attainment of a regionally balanced carbon budget is fundamental for the realization of carbon neutrality. This study involved the quantification of the carbon budget related to land use across Southeast China from 2005 to 2020, which was achieved through the calculation of both carbon emissions and carbon sinks. Subsequently, we scrutinized the mechanisms driving the observed dynamic changes in the carbon budget, pinpointed the impact of land-use efficiency (LUE) on the carbon budget, and proposed sustainable spatial planning and management strategies for diverse functional areas at the county level. The core findings are as follows: The dynamics of the carbon budget were spatially heterogeneous, characterized by a gradual increase in carbon emissions over time, while carbon sinks remained relatively constant. The Gini coefficient (G) manifested a gradual increase throughout the study period, reflecting an imbalanced evolution between carbon sinks and emissions. There was also an observable imbalance in the distribution of the carbon ecological carrying coefficient between coastal and inland regions. Land-use-related carbon emissions demonstrated a substantial spatial spillover effect, whereas a weak spatial spillover effect was noted in land-use-related carbon sinks. The correlation between LUE and the carbon budget varied significantly across different functional areas, as the driving effects of LUE displayed remarkable spatial heterogeneity. A quantification of the spatio-temporal alterations and the driving mechanisms behind the carbon budget can aid in the advancement of urban sustainability and regional carbon neutrality.


Introduction
At present, 197 countries and over 800 cities globally have instituted policies pertaining to net-zero emission targets, aiming to combat climate change and accomplish regional sustainable development (Niklas et al 2020, Van Soest et al 2021).In September 2020, China pledged to reach peak CO 2 emissions before 2030 and achieve carbon neutrality before 2060 (Liu et al 2022, Yang et al 2022).Under the umbrella of carbon neutrality, the focus is not solely on achieving comprehensive carbon emission control within China but also on ensuring regional coordination and equitable reduction of carbon emissions.This indicates an urgent need for low-carbon transformations across various sectors, including industry, energy, land use, and transportation (Lamb et al 2. Methods and materials 2.1.Data sources Multisource datasets were used to spatially evaluate the land-use carbon budget and spatial determinants (table 1).The raster data in table 1, which have different spatial resolutions, were used to quantify the land-use carbon budget.The statistical data were derived mainly from the Statistical Yearbook of Fujian Province (2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015)(2016)(2017)(2018)(2019)(2020).Because of a lack of statistical data for Jinmen County, this county was excluded from the analysis in this study.

Study area
Fujian Province is located on the southeast coast of China, with a land area of 124,000 km 2 and a sea area of 136,000 km 2 (figure 1).It is adjacent to the Yangtze River Delta city cluster and the Guangdong-Hong Kong-Macao Greater Bay Area and is the starting point of the Maritime Silk Road.The terrain of Fujian Province is high in the northwest and low in the southeast.The area of mountains and hills within Fujian Province accounts for approximately 90% of the province.The climate is a subtropical marine monsoon climate.
In 2020, Fujian Province's gross domestic product (GDP) stood at 4,390.389 billion yuan, ranking seventh in China and displaying distinct imbalances between coastal and inland economic development.These disparities parallel the spatial pattern of China's broader economic trajectory, rendering the province a significant reference  point for the country's economic development.With an urbanization rate of 69.5%, urban growth is predominantly clustered in the two metropolitan areas of Fuzhou and Xiamen, marking some of the fastest urbanization trends on China's southeast coast.Moreover, in 2014, Fujian Province emerged as China's inaugural ecological civilization demonstration area, boasting an impressive ecological environment.It has the country's highest forest coverage rate at 66.8%, denoting a robust carbon sink.Therefore, to attain carbon neutrality, Fujian's developmental strategies must consider multiple objectives, making the province an excellent case study to understand the interplay between land-use and carbon budget in rapidly urbanizing Chinese regions.

Research framework for analyzing the carbon budget related to land use
In this study, we quantified the carbon emissions and carbon sinks related to land use to characterize the carbon budget capacity of the study area, which incorporates the carbon emissions and carbon sinks related to constant land use as well as changes in land-use type, plus the carbon emissions generated by human activities (figure 2 and table S3).
(1) The land-use carbon budget primarily comprises cropland planting, vegetation biomass changes, water responses, and total energy consumption resulting from human activities.Firstly, we employed the classification system from the Fujian Statistical Yearbook, encompassing 23 crop types and seven types of livestock and poultry, to determine the carbon emissions associated with cropland planting (table S4 and  S5).These calculations were based on the Guidelines for the Preparation of Provincial GHG Inventory (2011).Secondly, forests and grasslands play a significant role in carbon sequestration, which is closely linked to vegetation's carbon absorption capacity across various climatic regions (Ren and Fan 2021).The total amount of carbon sinks in unchanged forests and grasslands was estimated by overlaying vegetation types with forest and grassland data in Fujian Province from 2005 to 2020 (table S6).Thirdly, the carbon budget for water was calculated using carbon emission and carbon sink coefficients specific to different rivers, lakes, and wetlands (Tang et al 2021).
(2) The carbon budget, stemming from both natural and anthropogenic land-use shifts, captures variations in soil organic carbon and vegetation carbon density due to land-use modifications.Soil organic carbon alterations arise from transitions between distinct land-use categories (Table S1).We posited that impervious surfaces retain consistent soil organic carbon levels (Cao et al 2023).Changes in vegetation carbon density relate to carbon fluxes-both emissions and sinks-linked to forest and grassland transformations (table S2).Notably, our analysis excluded underground carbon reservoirs within vegetation.(3) The carbon emissions from total energy consumption resulting from human activities include industrial activities, energy consumption, transportation, and travel.We utilized the comprehensive results of the Carbon Emission Accounts and Datasets (https://www.ceads.net.cn/).Furthermore, we used corrected nighttime lights data (DMSP/OLS data from 2005 to 2020) to correlate with the carbon emission data from the Carbon Emission Accounts and Datasets, representing the distribution of carbon emissions.The average R 2 value was 0.92.

Carbon Gini coefficient
The Gini coefficient (G) is an indicator typically defined based on the Lorenz curve to quantitatively measure the equality difference in the income level of a population, which can better quantify the regional difference and equality degree of indicators (Shu and Xiong 2018).Here, we used G to measure the spatial difference in the land-use carbon budget of Fujian Province.The calculation formula is as follows: where x i and x j represent the carbon budget capacity of regions i and j, X represents the average carbon budget capacity of Fujian Province, and n represents the total number of regions.A higher value of G indicates a greater imbalance in the carbon budget, signifying elevated carbon emissions coupled with a lower proportion of carbon sink offset.

Ecological support coefficient
The ecological support coefficient (ESC) represents the quotient between the ratio of total carbon sinks in the whole region and the ratio of total carbon emissions in the whole region, reflecting the strength of the land-userelated carbon sink capacity in the region.The calculation formula is as follows: where CS and CE respectively represent land-use-related carbon sinks and carbon emissions, respectively, i represents a certain sub-region, and T represents the whole region.If ESC > 1, the ratio of land-use-related carbon sinks between the sub-region and the entire region is greater than that of land-use-related carbon emissions, and the carbon budget gap is relatively small, and vice versa.

Spatial Markov chain
A Markov chain was used to reveal the evolution mechanism of the spatio-temporal pattern of the land-use carbon budget in Fujian Province during 2005-2020.The Markov chain discretized the spatial pattern of landuse carbon emissions and carbon sinks in different time periods (Su et al 2018), and we used quartiles to divide the spatial pattern types of land-use carbon emissions and carbon sinks in Fujian Province into low, normal, high, and very high (table S6).By calculating the probability distribution and transfer of different types, the evolution process was approximated as a Markov process.The spatial Markov chain is used to construct the spatial weight matrix and decompose the traditional probability transfer matrix into the probability transfer matrix of multiple spatial lag types.Here, we used the spatial distance matrix for the analysis, as follows: where P ij (Y) is the transfer probability, X ij represents the number of types of spatial lag in the study period from type i to type j at t+1, and X i is the number of types i in the whole period.To verify whether the spatial evolution of the carbon budget in each region showed significant spatial correlation, we assumed that spatial evolution occurred independently of each other and independently of the type of neighborhood space.The model test was the chi-square test, as follows: where k is the spatial pattern type of carbon emissions, Y is the spatial lag type, and P ij (Y) and X ij (Y) are the probability and quantity of spatial Markov transfer with spatial lag type Y, respectively.Structural efficiency evaluates spatial composition among land types, encompassing cropland, forest, urbanrural construction, and industrial-mining-transportation land proportions (Yang et al 2020).The balance between these land types has direct implications for carbon storage and emissions (Li et al 2022, Kang et al 2023, Zhang et al 2023).Expanding cropland often results in forest clearance, diminishing carbon storage due to cropland's typically lower organic matter content.Conversely, increased forest coverage augments carbon sinks, with vegetation adeptly transforming carbon to organic matter via photosynthesis (Zhang et al 2023).Urbanrural construction land proportions correlate with residential energy consumption and building carbon footprints (Li et al 2022), while industrial, mining, and transportation lands, as primary functional construction lands, significantly influence industrial and transportation carbon emissions (Kang et al 2023).
Scale efficiency relates to population, economic, and environmental shifts across land uses (Liu et al 2021).It underscores the importance of balancing construction land with ecological conservation to maintain a sustainable carbon budget (Liu et al 2021, Yang et al 2022).Rising per capita construction land disrupts the carbon budget, inducing augmented carbon emissions from infrastructure development (Kang et al 2023).Ecological land protection, reduced land reclamation, and curtailed destructive practices can mitigate these emissions and bolster carbon sink capacities (Yang et al 2022).The GDP-construction land relationship offers insights into economic development and output value density beyond traditional GDP metrics.Elevated GDP in relation to construction land indicates denser economic activity and resource efficiency, curbing unnecessary land expansion and associated emissions.The GDP of new construction land units mirrors urban development stages, with initial phases driven by scale expansion and later stages pivoting to economic system enhancements due to urban space constraints (Liu et al 2021).
Intensive efficiency reflects the input-output ratio within a region, considering factors like investment intensity, labor intensity, and production levels (Peng et al 2017).Investment intensity in fixed assets, representing the economic input level per construction land unit, signifies the nexus between total fixed asset investments and constructed land area within a given region (Gao et al 2020).Labor intensity in secondary and tertiary sectors denotes land use labor intensity.Agricultural output reflects cropland productivity, with potential for higher yields and diminished carbon emissions.The output value from secondary and tertiary sectors epitomizes land's economic output efficiency (Peng et al 2017).
In this study, we integrated the distinct regional attributes of Fujian Province to formulate an evaluation index system for LUE.We standardized the LUE evaluation index using the range method and determined index weights through the entropy weight method (Liu et al 2021).Further details can be found in table 2 and figure S1.
We first explored the relationship between LUE and the carbon budget in different dimensions through the OLS regression method.Second, because of the complex relationship between LUE and the land-use carbon budget, the statistical analysis model may ignore the spatial non-stationarity and heterogeneity.The geographically weighted regression (GWR) model is a modified linear regression model that represents an extension of the general linear regression model and that incorporates the spatial location of the object into the linear regression model (Xu and Lin 2017).Independent equations are constructed for each research element in different units, resulting in strong local analysis ability.Here, i is the research factor, (u i , v i ) is the coordinate of the research factor i, β k (u i , v i ) is the k regression coefficient of the i research factor, and ε i is the random error of the i research factor.We used the GWR model to study the impact of LUE on the spatial heterogeneity of the carbon budget from 2005 to 2020.The variation of LUE in different dimensions was taken as the independent variable of GWR, and changes in the land-use carbon budget were taken as the dependent variable of GWR.

Spatio-temporal distribution characteristics of land-use carbon emissions
From 2005 to 2020, land-use carbon emissions in Fujian Province gradually increased from 37.24 Tg C to 66.80 Tg C, showing a spatial distribution of high in coastal areas and low in inland areas (figure 3).Fuzhou and Xiazhangquan metropolitan areas, as the economic growth poles of the strategic land development in Fujian Province, were the agglomeration centers of high land-use carbon emissions.The total carbon emissions of all functional areas showed an obvious upward trend over time and varied in the following order: key development areas > agricultural development areas > ecological functional areas > optimized development areas.As the main population, economic, and industrial agglomeration areas in Fujian Province, key development areas have significantly higher carbon emissions than other functional areas, and their carbon emissions will continue to grow in the short term to meet the needs of economic development.Optimized development areas have a strong economic foundation and an economic development mode that has transformed from extensive LUE, high resource consumption, and high pollution emission to economic development.In 2015, these areas were the first to achieve total carbon emission reduction, which partly reflects the successful implementation of ecological civilization strategy during the 12th Five-Year Plan period and the promotion of regional carbon emission reduction.

Spatio-temporal distribution characteristics of land-use carbon sinks
From 2005 to 2020, the total amount of land-use related carbon sinks gradually decreased from 5.07 Tg C to 4.34 Tg C, showing a spatial distribution of low in coastal areas and high in inland areas (figure 3).The total carbon sinks of all functional areas showed a significant decreasing trend over time, and varied in the following order: agricultural development products > ecological functional areas > key development areas > optimized development areas.From 2005 to 2020, the rapid expansion of construction land in Fujian Province led to the conversion of large areas of forest land and grassland into construction land; however, urban and rural construction land had not reach the equilibrium state of increase and decrease, leading to a continuous reduction of carbon sinks in Fujian Province.Furthermore, continuous and stable forest cover leads to saturation of the carbon sequestration capacity.The growth of carbon sinks should not be dominated by an increase of forest coverage, but should focus on improvements to ecological environment quality and vegetation type structure.
In addition, the coastal areas were predominantly optimized development areas and key development areas.The carbon sinks in coastal areas were smaller than those in inland areas because of different locations of the main functional area.That is, low-carbon sink areas were distributed along the development axis of coastal towns, whereas inland areas were mainly agricultural development areas and ecological functional areas with a good ecological background.Thus, high carbon sink areas were mainly distributed along the ecological functional protection belt, with Wuyi Mountain and Dayun Mountain as the core.

Dynamic analysis of the land-use carbon budget
A comparative analysis of the land-use carbon budget in Fujian Province found that the carbon offset capacity of Fujian Province gradually decreased from 13.61% to 6.50% during 2005-2020 (figure 4), indicating that landuse carbon sinks in Fujian Province was far from sufficient to offset carbon emissions, and carbon emissions had not yet been peak.The G value of the carbon budget increased from 0.56 to 0.60 (figure 4(a)), indicating a transition from near imbalance to extremely imbalance, a gradual increase in the carbon budget difference between sub-regions, and development of carbon sinks and carbon emissions toward the direction of imbalance.
The land-use carbon budget across most functional areas exhibits a gradual shift towards imbalance, characterized by an increasing disparity between carbon emissions and carbon sinks, and a corresponding decline in the proportion of carbon sinks offset.Specifically, optimized and key development areas represent economically developed regions with high levels of industrialization, excessive energy consumption, and a substantial increase in carbon emissions that far surpasses the capacity of carbon sinks, thereby resulting in an imbalanced carbon budget.In contrast, ecological functional areas fall under the category of conservation zones with abundant carbon sinks, imposing restrictions on excessive industrial development and consequently leading to lower levels of economic development efficiency and resource utilization.As the process of economic development unfolds, more natural land with carbon sinks is converted into construction areas, causing a rise in land-use carbon emissions and a decrease in carbon sinks, thereby contributing to the progressive imbalance observed in the carbon budget.Furthermore, a comparative analysis of the dynamic changes in land-use-related carbon emissions and carbon sinks reveals a positive trend in carbon emissions in Fujian Province from 2005 to 2020, while carbon sinks predominantly experience a negative trend (figures 3(b) and 4(b)).
The ESC of the land-use carbon budget in Fujian Province showed a distribution pattern of low in coastal areas and high in inland areas (figure 5), which conforms to the strategic spatial development pattern of Fujian Province that consists of two poles, Fuzhou and Xia-Zhang-Quan metropolitan areas, two belts (coastal urban development belt and mountain green development belt), and three axes (mountain and sea development axes).The ESC of the two metropolitan areas continued to decrease throughout the study period, but gradually increased from coastal urban areas along the development axis of mountains and seas to inland mountainous areas; thus, the gap between coastal and inland ESC values progressively increased.
The ESC of the carbon budget differed substantially for each main functional area.The optimized development zone had the lowest ESC, which increased from 0.06 to 0.08; however, the relative carbon budget gap remained large.The ESC of key development areas was approximately 0.75, and the carbon budget gap was relatively large.For major agricultural development and ecological functional areas, the ESC was greater than 1, showing an obvious upward trend, and the relative gap between carbon budget gradually reduced over the study period.

Mechanism behind the spatio-temporal evolution of the land-use carbon budget
The probability of the spatial pattern of carbon emissions remaining unchanged was 73.13%, which was less than that of the spatial pattern of carbon sinks remaining unchanged (93.75%), indicating that the spatial pattern of carbon emissions is more likely to evolve (Table S6).The non-diagonal evolution probability was distributed on both sides of the diagonal, that is, the spatial pattern of land-use-related carbon emissions and carbon sinks showed a bidirectional evolution trend and was more likely to evolve to low carbon emissions than to high carbon convergence (6.56%>0), indicating that low carbon emissions are more important for generating a new pattern of land space development and protection.
However, the Markov chain without lag could not prove that the heterogeneous county-scale spatial evolution of the land-use carbon budget had an effect on the spatial evolution of the carbon budget in neighboring counties.Therefore, we constructed the spatial adjacency matrix to reveal the interactions and evolution mechanism of the carbon budget spatial pattern in neighboring counties.Without adjusting the degree of freedom, the chi-square test of land-use-related carbon emissions and carbon sinks were significant at the level of p<0.01, in which the chi-square value of carbon emissions was 58.50; larger than that of carbon sinks (49.89).This indicates that the spatial evolution of the land-use carbon budget in Fujian Province in each county was not independent from that of neighboring counties during 2005-2020, that is, spatial heterogeneity significantly affected the spatial pattern of the carbon budget in neighboring counties, with a significant spillover effect.
Specifically, when a neighboring county had high carbon emissions (Table S6), the probability of the county evolving to have high carbon emissions was relatively high (20%).Conversely, when a neighboring county had high carbon sinks, the probability of the county evolving to have high carbon sinks was relatively low (3.57%).Thus, a high-carbon development pattern in a county enhanced the spatial diffusion effect, that is, neighboring counties exhibited a more obvious clustering trend toward the spatial pattern of high carbon emissions, indicating that an increase of regional land-use carbon emissions promotes an increase of carbon emissions in neighboring counties (figure 6).The collaborative reduction and control of total and regional carbon emissions is crucial for achieving the strategic goal of carbon neutrality.

Impacts of land-use efficiency on the land-use carbon budget
There was little correlation between LUE and the carbon budget in optimized and key development areas, which differed appreciably in these respects (figure 7).Specifically, the optimized development zones had a higher LUE and a smaller carbon budget, whereas key development areas had a large carbon budget and low LUE.Moreover, agricultural development areas and ecological functional areas exhibited a high correlation between LUE and the carbon budget.
In addition, the correlation between scale efficiency and the carbon budget gradually increased from 2005 to 2020.Because improvements in the urbanization and socio-economic level increase the utilization of construction land, the continuous growth of vegetation cover rate leads to a continuous improvement of scale efficiency, and the impact on the carbon budget is gradually enhanced.However, the structural efficiency and intensive efficiency showed a trend of first decreasing then increasing over the study period.After the global financial crisis in 2008, Fujian Province accelerated the optimization of economic structure, followed a sustainable green development path, established development based on resources and the environment, promoted an increase of input and output levels, and improved the land structure and intensive efficiency.
We further analyzed the influence of LUE's spatial heterogeneity on the spatio-temporal evolution of the carbon budget, which resulted in an R 2 value of 0.41, indicating a reasonable fit (figure 8).The regression coefficients of LUE were positive in different dimensions, which indicates that the improvement of LUE in Fujian Province promoted improvement of the carbon budget at this stage.The impacts of LUE on the land-use carbon budget exhibited significant spatial heterogeneity, showing a ring-shaped pattern, as described below.
(1) The increase of LUE with distance from the two coastal metropolitan areas indicates that the impact of LUE on the carbon budget of metropolitan areas gradually increased from coastal areas to inland.This is mainly attributed to the high level of economy and urbanization in coastal areas, the high LUE in metropolitan areas, transformation of the urban development mode from urban expansion to urban internal renewal, and the low impact of LUE on the carbon budget.In contrast, the improvement of LUE in mountainous areas has greatly promoted growth of the carbon budget in the study area.(2) The impact of structural efficiency on the carbon budget gradually increased from the southeast coast to the northwest inland area, which shows that improvement of the structural efficiency had a greater impact on the carbon budget of inland areas.When promoting economic development, underdeveloped areas in inland counties will inevitably generate more intense human activities, which will have a significant impact on the land-use structure.In contrast, considering the advantages of a first-mover economy, coastal areas should pay more attention to adjusting the ecological land structure, improving carbon sinks, and promoting a balanced carbon budget.
(3) The effect of scale efficiency on the carbon budget decreased from the northeast coastal areas to the southwest inland areas.Improvement of the scale efficiency in Fujian Province stems from improvements in the population and land economic capacity, which have promoted growth of the land-use carbon budget in Fujian Province.In contrast, improvement of the scale efficiency in northeast coastal areas had a greater impact on the carbon budget because of rapid economic development and population agglomeration in these areas in recent years.
(4) The impact of intensive efficiency on the carbon budget decreased from the southeast coastal areas to the northwest inland areas.Compared with coastal areas, the level of economic development in inland areas is relatively low, and the level of resource utilization, economic benefits, and industrial structure is relatively poor.Thus, improving the intensive efficiency of inland areas will help improve productivity and increase the carbon budget.Furthermore, coastal areas are economically developed and densely distributed with industrial enterprises, so the intensive efficiency was relatively high.The intensive efficiency showed a relatively small impact on the carbon budget.

Discussion
4.1.Spatial effects of achieving regional carbon neutrality Achieving the goal of regional carbon neutrality requires recognition of the complex relationships impacting the land-use carbon budget and incorporation of the various spatial effects into the land-use carbon budget (Liu et al 2017, Wang et al 2021a, Yang et al 2022).For consistency, we resampled data with different spatial resolutions to 30 m, summarized the land-use carbon budget at the county level (Chen et al 2020, Long et al 2021).Because achieving regionally collaborative carbon reduction is critical to achieving regional carbon neutrality, we performed this research at the county level (the basic management unit of the Chinese government).
We examined the influence of spatial effects on land-use-related carbon emissions and sinks at the county scale.Regions with advanced economic development exhibit a pronounced spatial synergistic effect on carbon emissions, aligning with findings from Long et al (2021) and Nyamari and Cabral (2021).Conversely, regional carbon sinks remain relatively stable, presenting a subdued spatial effect at this scale .Areas with robust ecosystems predominantly host high carbon sinks, as corroborated by Tong et al (2020).Given these insights, strategies should prioritize curbing regional carbon emissions, factoring in spatial dynamics.A European analysis accentuates this, indicating that mitigating anthropogenic carbon emissions outweighs terrestrial ecosystem carbon sequestration benefits by a significant 17.4% (Pan et al 2023).Strengthening policy frameworks and implementing strategic interventions to augment regional carbon sinks are essential (Zhu et al 2021).The integration of urban impervious surfaces with streetscape vegetation offers promising avenues for enhancing carbon sequestration (Ye et al 2023).
Furthermore, the spatial effect was also observed in different functional areas (He et al 2021).Some of the optimized development areas have achieved a reduction of their total carbon emissions, mainly because these areas have good economic conditions and are predominantly urban centers, as found by Cheng et al (2020) and Shan et al (2018).However, the carbon emissions generated by agricultural functional areas and ecological reserves for economic development, which generally have poor economic conditions, should not be ignored (Wu et al 2021).Ensuring that the carbon budget is maintained in these areas should be a focus of future research.
4.2.Land-use efficiency and its correlation with the land-use carbon budget LUE directly influences the carbon budget associated with land-use activities (Yang et al 2020).We investigated the determinants of LUE at the county functional area scale.With the expansion of construction land and sustained growth in forest cover (figure S1), the scale efficiency of LUE increasingly correlated with the carbon budget from 2005 to 2020.However, the relationship between the carbon budget and both structural and intensive LUE efficiencies initially weakened before strengthening.Compact urban designs, incorporating mixed land-use, efficient transport, and infrastructure, curtail carbon emissions from urban expansion, deforestation, and land conversion (Hutyra et al 2011, Long et al 2021, Nyamari and Cabral 2021).Such designs enhance LUE scale efficiency, fostering a more balanced carbon budget.Refining spatial composition among land types and elevating regional input-output levels can further mitigate emissions and augment carbon sinks (Yang et al 2020, Kang et al 2023).Notably, LUE's influence on the carbon budget in Fujian Province intensifies from coastal to inland regions, underscoring the importance of understanding these dynamics for effective carbon budgeting and land-use optimization.Furthermore, we also transitioned from the traditional per capita GDP metric to GDP per unit of construction land.This shift aligns with the 'Standard for Evaluation of Saving and Intensive Use of Construction Land' (Ministry of Natural Resources of the People's Republic of China 2008), where this metric serves as a key indicator.It offers a nuanced perspective on land use efficiency, presenting a richer understanding of output value density and economic development than standard GDP measures (Liu et al 2021).This metric is also pivotal for pinpointing inefficient construction land, characterized by elevated carbon emissions and diminished economic returns (Kang et al 2023).

Implications for urban planning and management based on the land-use carbon budget
Subsequent to our examination of the land-use carbon budget in Fujian Province, we propose several spatial management and planning measures, as depicted in figure 9. Firstly, the development of a regional carbon budget policy necessitates accurate assessment of total regional carbon emissions and sinks (Friedlingstein et al 2020).Our study demonstrates a significant increase in carbon emissions for each county, accompanied by a gradual decrease in the proportion of carbon sink offset in Fujian Province.Consequently, we have partitioned the carbon budget based on functional areas for each county, encompassing payment, balance, and compensation areas.Dividing areas with a balanced carbon budget according to functional areas offers valuable guidance towards achieving regional carbon equity and fostering collaborative carbon emission reduction (Zhang et al 2018;Yang et al 2021a).
Secondly, carbon emissions exhibit spillover effects characterized by dynamic correlations and spatial interactions, rendering it challenging for local governance alone to address the emission problem (Zhu et al 2021).Instead, a holistic approach is required, considering the interconnections between economically developed core regions and their surrounding areas, and establishing a mechanism for regional collaborative carbon emission reduction (Wang et al 2021b).Based on these considerations, our study emphasizes the urgent need for coastal counties in Fujian Province to curtail carbon emissions through industry and economic development transformations, given their propensity for high carbon emission agglomerations.In contrast, the regional distribution of high carbon sinks is largely influenced by natural conditions.Consequently, efforts to enhance carbon sinks should focus on identifying significant functional areas for natural carbon sinks and artificial sink enhancements, integrating them into the layout of nationally important ecological reserves, ecological red line areas, and major ecological restoration project areas.Specifically for Fujian Province, it is crucial to prioritize the augmentation of natural carbon sinks in inland counties and expand urban greening spaces in coastal counties to bolster urban carbon sinks.
Third, we scientifically and rationally delineated the areas with a balanced carbon budget and, hence, propose corresponding optimization schemes to coordinate the regional, benefit-based relationships between carbon emission and carbon sink subjects to achieve regional carbon equity and low-carbon development (figure 9).Based on the analysis of the dynamic changes of the carbon budget in the four different functional areas, we used cluster analysis to divide the carbon payment, balance, and compensation areas from the perspective of regional comprehensiveness, and combined these with regional carbon budget, economic development condition (table S7), ESC, and LUE.Because of the different orientation and development directions of the different functional areas, certain economic or policy compensations could achieve the wholistic balance of the regional carbon budget.Scientific and reasonable compensation is a powerful measure to achieve balance in the regional carbon budget.

Limitations
Our study acknowledges several constraints.The land-use data classification adopted is broad, encompassing merely six primary categories and 25 secondary ones.For a nuanced understanding of regional carbon neutrality, high-resolution classifications in land-use carbon budget analyses are imperative.While parameters for calculating carbon emissions and sinks stem from relevant literature, we did not adjust for potential countyspecific variations, potentially impacting the accuracy of our carbon budget estimates.The significance of belowground carbon-pertaining to the carbon stored in plant roots, rhizomes, and tubers-cannot be understated, especially when assessing land-use carbon budgets in forthcoming studies.It's noteworthy that a recent global research indicated that underground plant components contribute to approximately 24% of the overall carbon storage (Ma et al 2021).Moreover, our analysis marginally addresses the carbon budget within urban interiors, especially concerning urban growth.Future research should scrutinize urban carbon budget shifts in the context of policy enactments, emphasizing China's spatial planning post-2020.

Conclusions
In this study, we explored the spatio-temporal distribution of the land-use carbon budget in Fujian Province, as well as the driving effects of LUE at the scale of county functional areas.Our findings can be used to support cross-scale spatial planning and management strategies.First, the distribution of the carbon budget was spatially heterogeneous but relatively stable throughout the study period.Second, high carbon emission areas were mainly distributed in metropolitan areas, whereas high carbon sink areas were mainly distributed in inland mountainous areas.Third, differences in the carbon budget gradually increased over time, and carbon sinks and carbon emissions developed in the direction of imbalance.In addition, LUE had different driving effects on the carbon budget for different spatial scales and functions, indicating that the priority of carbon budget management differs with scale.Finally, we proposed several targeted spatial planning and management strategies for different functional areas based on the calculated carbon budget and LUE drivers of the carbon budget to help achieve regional coordinated carbon emission reduction and carbon neutralization goals.

Figure 1 .
Figure 1.Map of the study area, showing (a) its location in China; (b) a detailed location map; (c) elevation; and (d) vegetation coverage.

Figure 3
Figure 3 Spatio-temporal distribution of land-use-related carbon emissions and carbon sinks from 2005 to 2020 in Fujian Province.(a)-(b) The temporal trend of carbon emissions and sinks; (c)-(d) the spatial distribution of carbon emissions and sinks.

Figure 4 .
Figure 4. Dynamic changes in the land-use carbon budget from 2005 to 2020 in different functional areas of Fujian Province: (a) evolution of G; (b) evolution of land-use-related carbon emissions and carbon sinks.

Figure 5 .
Figure 5. Dynamic changes in the ecological support coefficient from 2005 to 2020.

Figure 6 .
Figure 6.Spatio-temporal changes in the land-use carbon budget: (a) changes in carbon emission type; (b) changes in carbon emission type affected by neighboring counties; (c) changes in carbon sink type; (d) changes in carbon sink type affected by neighboring counties.

Figure 7 .
Figure 7. Correlation between LUE and the land use-carbon budget from 2005 to 2020 in different functional areas of Fujian Province.(a) Correlation between LUE and carbon budget; (b) correlation between structural efficiency and carbon budget; (c) correlation between scale efficiency and carbon budget; (d) correlation between intensive efficiency and carbon budget.

Figure 8 .
Figure 8. Driving effects of LUE on the land-use carbon budget in Fujian Province from 2005 to 2020.(a) Driving effect of LUE on the carbon budget; (b) driving effect of structural efficiency on the carbon budget; (c) driving effect of scale efficiency on the carbon budget; (d) driving effect of intensive efficiency on the carbon budget.

Figure 9 .
Figure 9. Spatial management and planning strategies based on the land-use carbon budget.The payment areas need to remunerate the compensation areas through economic or policy measures; the balance areas could not need to remunerate or receive compensation; and the compensation areas need economic or policy compensation from payment areas.

Table 1 .
Summary of the primary data.

Table S1 (
Lai et al 2016) / Soil organic carbon data Text Table S2 (Lai et al 2016) / Crop yield Text Statistical Yearbook of Fujian Province / Livestock and poultry Text Statistical Yearbook of Fujian Province /

Table 2 .
Evaluation index system of LUE in Fujian Province.