Quantification and sensitivity assessment of Chinese provincial ecological compensation in the perspective of carbon deficit redistribution

Abstract

The ecological compensation mechanism is a tool for managing regional development and promoting the green economy. This paper proposes a revised model for carbon emissions and absorption based on IPCC, then analyses the spatial and temporal variations of carbon emissions, absorption, and deficit in 31 provinces of China from 2001 to 2019. A model was developed to redistribute carbon deficits and compensate for ecological imbalances, with the aim of eliminating background differences between regions. The concept of ecological compensation sensitivity was proposed, and a redundancy analysis (RDA) was conducted to detect possible influencing factors. Results indicate that: (1) The carbon deficits were relatively substantial in the centre and developed coastal regions, which were the subject regions of China’s ecological compensation. Meanwhile, the northwest and southwest regions were the object regions of ecological compensation, and the compensated object regions tended to extend towards the southeast. The majority of compensated subject areas are generally less sensitive than compensated object regions. (2) The graph of carbon deficit volume over time shows that China’s two carbon targets are closely aligned. The distribution of China’s ecological compensation amounts follows the pattern of provincial economic development levels, with higher compensation amounts concentrated in the central and coastal regions. There is a positive correlation between GDP, year, and population size with carbon emission and carbon deficit. Additionally, there is a positive correlation between year, region, and carbon absorption. (3) This compensation model can strongly incentivise the compensation subject area to proactively adjust its economic development model to cope with the significant compensation pressure. Additionally, it can fully encourage the compensation recipient area to continue adhering to the green economic development model. The conclusions of the study hold significant reference value for promoting the development of a green, low-carbon economy.

Introduction

As the world’s second-largest economy, China’s rapid development has led to an escalation of environmental and ecological issues. In response, the Communist Party of China and the Chinese government have taken a proactive approach and accorded high priority to ecological conservation, actively promoting the improvement of ecological conditions¹. While traditional ecological compensation schemes focus mainly on the link between carbon emission units and the environment^2,3, achieving regional carbon emission reduction and fair and sustainable development among provinces requires horizontal ecological compensation among provinces⁴. Current research has identified significant challenges in China’s provincial inter-ecological compensation, including balancing ecological and economic interests⁵, appropriately penalizing those who damage the ecology and compensating those who protect it, and managing regional development coordination⁶. The tense relationship between regional ecological protection and economic development not only places a heavy burden on China’s ecological protection but also impacts the sustainable development of regional economies^7,8,9. To address this issue, it is imperative to establish a provincial ecological compensation mechanism to reconcile the ecological and economic interests of stakeholders in different regions, promote ecological environment protection, foster fair competition between regions, and facilitate overall coordinated social development⁷.

Ecological compensation, as an important means of protecting the ecological environment internationally, has been widely accepted and applied by the Chinese government and society. It focuses on using economic means to coordinate the interests of stakeholders, promote the implementation of compensation activities, and mobilize enthusiasm for ecological protection. Its basic guiding principle is: “Whoever develops, whoever protects; whoever destroys, whoever restores; whoever benefits, whoever compensates; whoever pollutes, whoever pays”⁸. At the same time, ecological compensation, as a new research field that has emerged in the context of global change and low-carbon economy, has also become a hot research direction for scholars in recent years⁹. Accurately assessing the value of ecosystems is the foundation and premise of ecological compensation¹⁰, but China’s ecological compensation practices are still in the exploratory stage, and there are no unified and widely applicable standards and models, which greatly hinders China’s efforts to promote regional green coordinated development¹¹.

To address this issue, numerous articles on the quantification of ecological compensation have emerged in recent years. The main accounting methods for ecological compensation quantification are currently the contingent valuation method^12,13, ecological footprint method^14,15, ecosystem service value method^16,17,18, opportunity cost method¹⁹, and Carbon Footprint Method²⁰. The contingent valuation method (CVM) primarily obtains residents’ willingness to pay or acceptance through questionnaire surveys, which are more susceptible to subjective factors^21,22. The ecological footprint method primarily determines the ecological compensation standard by determining the ecological deficit and surplus in the study area, but it ignores economic and social influences, resulting in a final result that is larger than the actual situation. The ecosystem service value method determines the final ecological compensation standard by measuring the value of ecosystem services, but the final result is also high, and the calculation process is often cumbersome. The opportunity cost method takes the opportunity cost loss caused by ecological compensation as the compensation standard, but it requires high data accuracy and is prone to result errors²³. The Carbon Footprint Method refers to determining the final ecological compensation standard by multiplying the carbon emission factor by the specific activity level to obtain the carbon account and then determining the final ecological compensation standard according to the regional carbon compensation model. Compared with other methods, this method is more convenient to calculate, more operational at the provincial level, and has been adopted by the Intergovernmental Panel on Climate Change (IPCC)²⁴ and widely used internationally. Therefore, this article chooses to use the Carbon Footprint Method for China’s provincial ecological compensation quantification analysis.

In the perspective of quantification, the current research mainly has three approaches: the first is to determine the ecological compensation standard from the carbon sink value angle, which is to calculate the value generated by the increase in carbon sink, and then use the estimated value to determine the carbon compensation standard^25,26. The second is to determine the ecological compensation standard from the compensation willingness angle, which is to ask the payers or beneficiaries to pay or receive the willingness amount through field questionnaire survey to determine the carbon compensation standard^3,22. The third is to determine the carbon compensation standard from the carbon balance angle, which is to calculate the difference between carbon emissions and carbon absorption, and then combine it with the carbon trading market to determine the ecological compensation standard, such as^13,27. Compared with determining the compensation standard from the carbon sink angle and the compensation willingness angle, determining the compensation standard from the carbon balance angle fully considers the calculation of ecological compensation standard based on the carbon emissions and carbon absorption, which is a more scientific and comprehensive calculation method. Therefore, this article chooses to calculate the ecological compensation standard from the carbon balance perspective.

The Carbon Footprint Method under the carbon balance perspective takes into account the calculation of carbon emissions and carbon absorption when calculating the ecological compensation standard, and completes the determination of the ecological compensation standard based on the IPCC calculation method combined with the regional ecological compensation model. Summarizing the existing literature, it is found that this method still has many problems. First, there are significant differences in the calculation of carbon emissions and carbon absorption using different methods, and the data or regional scope included are different. When considering factors, the focus is also different, and there is even the possibility of repetitive calculation and omission, which leads to insufficient accuracy of the calculation results²⁸. Moreover, many scholars adopt data from different years, and the provinces considered are not the same, which makes the regional ecological compensation data and standards lack intuitive comparability²⁹. Second, when establishing the compensation model, the focus is mainly on the determination of the compensation standard, and almost no one proposes a fair and promotional calculation model from the perspective of “carbon deficit redistribution”. For example, although Chen et al.³³ further processes the carbon deficit, it causes the carbon total value to be inadvertently amplified or reduced; finally, in the application of ecological compensation, few articles have appeared on the impact of ecological compensation quota on the local government’s financial situation. Some scholars have compared the compensation quota with GDP, but the proportion has certain differences. For example, Yu et al.³⁰ only mentioned the decoupling effect of carbon emissions and GDP in some provinces, but did not calculate the proportion of ecological compensation and GDP, and Gao et al.²⁰ used the compensation priority index to represent the size of the regional GDP, but this comparison is not very direct and cannot effectively reflect the essence of compensation and cannot effectively reflect the impact of compensation on the local government’s finances.

Given the current research problem, this paper aims to establish a modified carbon emission and carbon absorption calculation model based on available data, which can directly use the statistical data released by the National Bureau of Statistics of China and can eliminate redundant calculations and oversights to the greatest extent, resulting in more accurate calculation results. At the same time, in order to take into account the differences in regional background characteristics (population, geographical area, and GDP), an improved concept and model of carbon deficit redistribution are proposed, which has the advantages of “further rewarding the less and punishing the more”, “keeping the total carbon deficit unchanged”, and “balancing efficiency and fairness”. It plays an important role in promoting energy conservation and emission reduction. Finally, considering that the amount of regional compensation needs to be provided from the regional fiscal expenditure or included in the fiscal revenue, it is more realistic to compare the proportion between the compensation amount and fiscal revenue or expenditure. Finally, the concept of ecological compensation sensitivity is proposed.

Material and methods

The heterogeneity of China’s resources and economy has resulted in significant regional variability in CO₂ emissions³¹, with about half of the provinces’ regional carbon emissions not being fully absorbed internally, resulting in a carbon deficit. Because the data from Hong Kong, Macao and Taiwan are not available, this paper calculated and analyzed the spatial and temporal variation characteristics of carbon emission, carbon absorption, and carbon deficit of 31 provinces (municipalities and autonomous regions) in China from 2001 to 2019, a general view of the study area is shown in Fig. 1.

Fig. 1

Overview of the study area.

The various fossil energy consumption, electricity consumption data, and energy balance sheets used in this study to account for carbon emissions was obtained from data published on the website of the China Statistics Bureau, among which the carbon emissions of Tibet were estimated using the average carbon emissions per 10,000 Yuan GDP of Qinghai, Xinjiang, and Gansu provinces, which is at the same level of economic development with Tibet, due to the absence of statistics on fossil energy consumption. The forest area, grassland area, and crop yields used in the carbon sequestration calculations came from the China Statistical Yearbook, while the population, area, GDP, and fiscal revenue of other provinces (municipalities directly under the Central Government and autonomous regions) came from the China Statistics Bureau’s website.

Characteristics of the spatial distribution of ecological compensation were analysed using the Moran index, which is one of the commonly used indicators for measuring spatial autocorrelation, including the global Moran index and the local Moran index, and less literature has been published on the analysis of the spatial and temporal distribution patterns of carbon compensation using this method. Using the redundancy analysis (RDA) technique, the correlation analysis of carbon emissions, carbon sequestration, and carbon deficit by region with GDP, population size, regional area, and year revealed the overall correlation between the two types of factor indicators and the magnitude of their impact on each region.

Carbon emission accounting models

The IPCC’s emission factor method can calculate carbon emissions in a given region on the basis of fossil fuel consumption, which is a direct and convenient calculation technique. Most data on fossil fuel consumption can be obtained directly, but the national statistical data on coal consumption in China does not distinguish between the non-thermal power generation consumption ratio in each province. The electricity produced through coal consumption is transmitted to various provinces, so the part of coal consumption should not be counted in the province where the power plant is located. Ignoring this fact will inevitably result in the calculated value of regional carbon emissions being higher than the actual value. Therefore, we obtained the national average non-thermal power generation consumption ratio of coal based on the coal energy balance table in the national statistical data. Additionally, the consumption of electricity as a secondary energy source should also be included in the carbon emission calculation system. Since power generation is divided into thermal power, hydro-power, wind power, nuclear power, etc., it is particularly important to calculate the proportion of thermal power generation. Ignoring the fact that coal is consumed to generate electricity will cause the calculated value to be smaller than the actual value. We calculated the national thermal power generation proportion based on the power balance table. Finally, these two ratios are applied to the formula of the emission factor method to further refine the numerical value of carbon emissions.

Total carbon emissions are calculated as follows:

$$C_{{\text{e}}} = C_{{{\text{e}}1}} + C_{{{\text{e}}2}}$$

(1)

$$C_{{{\text{e}}1}} = \frac{44}{{12}}\sum\limits_{j}^{n} {\alpha_{j} \cdot {\text{A}}_{j} \cdot \beta }$$

(2)

$$C_{{{\text{e}}2}} = \frac{44}{{12}}\beta \left( {\alpha_{{\text{e}}} \cdot \lambda_{1} \cdot {\text{A}}_{e} + \alpha_{c} \cdot \lambda_{2} \cdot {\text{A}}_{c} } \right)$$

(3)

where, C_e is the total carbon emission, C_e1 is the carbon emission from fossil energy consumption (excluding coal), t/a; C_e2 is the coal and electricity consumption, t/a, α_j is the coal conversion factor of the jth energy source into standard coal (Table 1); A_j is the consumption of the jth energy source, t/a (α_j and A_j from Department of Energy Statistics, National Bureau of Statistics of China, 2001–2019); β is the carbon emission of 1t of standard coal, the value is 0.7326 t from the relevant literature (National Development and Reform Commission of China, 2001–2019); 44/12 is the molecular weight ratio of carbon dioxide and C molecular weight ratio; α_e is the coal conversion factor of electrical energy into standard coal; α_c is the coal conversion factor of coal (αe and αc from Department of Energy Statistics, National Bureau of Statistics of China, 2001–2019); λ₁ is the share of thermal power generation in electricity consumption; λ₂ is the share of non-thermal power generation in coal consumption. The parameters λ₁ and λ₂ are derived from the coal and electricity balance sheets issued by the Chinese National Bureau of Statistics.

Table 1 Standard coal conversion coefficient of various energy sources.

Carbon absorption accounting models

When accounting for carbon absorption, we have tried to consider the carbon sequestration of forests, grasslands, wetlands, and crops as comprehensively as possible. The carbon absorption calculation models for forests, grasslands, and wetlands are shown in formulas (5–7). When calculating the carbon absorption of crops, we considered both annual crops (rice, wheat, corn, millet, sorghum, legumes, tubers, cotton, peanuts, rapeseed, sesame, sunflower, sugar beet, tobacco, vegetables, fruits, and melons) and perennial crops (tea and garden fruits). Since perennial crops are only considered for their economic yield, the calculation model is slightly different, as shown in formulas (9–10).

$$C_{a} = C_{a}^{f} + C_{a}^{g} + C_{{\text{a}}}^{n} + C_{a}^{agr}$$

(4)

$$C_{{\text{a}}}^{f} = \frac{{{44}}}{{{12}}}{\text{M}} \cdot {\text{m}}$$

(5)

$$C_{{\text{a}}}^{g} = \frac{{{44}}}{{{12}}}{\text{N}} \cdot {\text{n}}$$

(6)

$$C_{{\text{a}}}^{n} = \frac{44}{{12}}P \cdot p$$

(7)

$$C_{a}^{agr} = C_{a1}^{agr} + C_{a2}^{agr}$$

(8)

$$C_{a1}^{agr} = \frac{44}{{12}}\lambda_{3} \cdot {\text{z}} \cdot \sum\limits_{{\text{t}}}^{n} {\frac{{p_{t} (1 - \omega_{t} )}}{{O_{t} }}}$$

(9)

$$C_{a2}^{agr} = \frac{44}{{12}}{\text{z}} \cdot \sum\limits_{{\text{t}}}^{n} {p_{t} (1 - \omega_{t} )}$$

(10)

where, C_a is the carbon carrying capacity, t/a; $C_{a}^{f}$ is the carbon sequestration capacity of forest, t/a; $C_{a}^{g}$ is the carbon sequestration capacity of grassland, t/a; $C_{a}^{n}$ is the carbon sequestration capacity of wetland, t/a; $C_{a}^{agr}$ is the carbon sequestration capacity of crops, t/a; M is the area of forest in the region, hm²; m is the average NEP of forest, which is the carbon sequestration capacity of 1 hectare of forest in one year, taking the value of 3.8096 t/(hm²·a); N is the total area of grassland in the region, hm²; n is the average NEP of grassland, which is the carbon sequestration capacity of 1 hectare of grassland in one year, taking the value of 0.9482 t/(hm²·a); P is the area of wetland in the region, hm²; p is the average NEP of wetland, which is the carbon sequestration capacity of 1 hectare of wetland in one year, and the value is 7 t/(hm²·a) (M, N and P from National Bureau of Statistics of China, 2001–2019; Xie et al.³⁶; Gao et al.²⁰); $C_{a1}^{agr}$ is the carbon sequestration capacity of annual crops, t/a; $C_{a2}^{agr}$ is the carbon sequestration capacity of perennial crops, t/a;λ₃ is the correction factor, which takes into consideration the amount of crop leftovers reused and burned and has a value of 0.07; z is the biomass and carbon conversion factor, which according to the literature is equal to 0.5 (Fang³⁷); pt is the economic yield of the tth crop, t/a; n is the number of crop species; ω_t is the water content of the economic yield of the tth crop; O_t is the ratio of the economic yield to biomass of the tth crop, also known as the economic coefficient in Table 2 (National Bureau of Statistics of China, 2001–2019; Li³⁸).

Table 2 Economic coefficient and water content of various crops.

Carbon deficit redistribution model

Using the above carbon emission and carbon absorption models, it is possible to calculate a regional carbon deficit, as indicated in E_q. below:

$${\text{C}}_{{\text{d}}} = {\text{C}}_{{\text{e}}} - {\text{C}}_{{\text{a}}}$$

(11)

where, $C_{d}$ is the carbon deficit before redistribution.

However, the current carbon deficit disregards the absolute statistics of regional population, area, and GDP, which lacks a certain degree of fairness; therefore, we propose the notion of carbon deficit redistribution, adhering to three redistribution rules.

• 1. The total carbon deficit in the study region stays unchanged prior to and during redistribution.

• 2. The redistribution is calculated using population, area, and GDP considerations taken into account.

• 3. The distribution reflects the incentive concept of “further rewards for less and penalties for more”.

The following formula was then developed for estimating the value of the carbon deficit redistribution:

$$C_{{{\text{di}}}}^{P} = C_{di} + P_{i} \times \left( {\frac{{C_{di} }}{{P_{i} }} - \frac{{C_{T} }}{{P_{T} }}} \right)$$

(12)

$$C_{di}^{A} = C_{di} + {\text{A}}_{i} \times \left( {\frac{{C_{di} }}{{A_{i} }} - \frac{{C_{T} }}{{A_{T} }}} \right)$$

(13)

$$C_{di}^{GDP} = C_{di} + {\text{GDP}}_{i} \times \left( {\frac{{C_{di} }}{{{\text{GDP}}_{i} }} - \frac{{C_{T} }}{{{\text{GDP}}_{T} }}} \right)$$

(14)

$$\overline{\overline{{C_{di} }}} = \frac{{C_{di}^{P} + C_{di}^{A} + C_{di}^{GDP} }}{3}$$

(15)

In the formula, $C_{di}^{P}$ represents the population correction for the i-th province or municipality; $C_{di}$ represents the carbon deficit for the i-th province; $P_{i}$ represents the population of the i-th province; $C_{T}$ represents the total carbon deficit for all regions in a certain year in China; $P_{T}$ represents the total population; $C_{di}^{A}$ represents the area correction for the i-th province; $A_{i}$ represents the area of the i-th province, 10 K (Km)²; $A_{T}$ represents the total area for all regions in a certain year in China, 10 K (Km)²; $C_{di}^{GDP}$ represents the GDP correction for the i-th province; ${\text{GDP}}_{i}$ represents the GDP of the i-th province, 100 M RMB; ${\text{GDP}}_{T}$ represents the total GDP for all regions in a certain year in China, 100 M RMB; $\overline{\overline{{C_{di} }}}$ represents the redistributed carbon deficit.

Model validity verification

Using data validation to evaluate the three settings of the model is reasonable. Considering the different contents considered in carbon emission, carbon absorption, and carbon price accounting in existing relevant studies, it would make the model comparison meaningless. Therefore, this paper fixed the data of carbon emission, carbon absorption, carbon deficit, and carbon price to the data of the Beijing–Tianjin–Hebei region in 2019. The relevant validation data and results are as follows (Table 3):

Table 3 Model validation data.

As the ecological compensation quota is a product of the modified carbon deficit and carbon price in each model, the values of each model’s ecological compensation quota are not presented in Table 4. Table 4 reveals that the modified carbon deficit in this paper has the advantage of maintaining the total carbon deficit unchanged before and after allocation, which meets the requirements of the first principle. In regards to the second principle, Miao Yang’s model only considers the impact of GDP through the improved Pearl growth curve, while Yang Chunguang takes into account the influence of population, GDP, and area simultaneously. Nevertheless, the final modified carbon deficit results are significantly different. Compared with the aforementioned models, the consideration of population, GDP, and area in this paper is more reasonable. When it comes to the third principle, by comparing the modified carbon deficit before and after modification, one can find that the carbon deficit in Beijing and Hebei Province decreases after modification, while the carbon deficit in Tianjin increases after modification. The shift in carbon deficit before and after modification reflects the principle of “Further reward less and punish more”.

Table 4 Verification result.

Further explaining the third principle from the formula itself, taking formula (12) as an example, according to the formula definition, we know that $\frac{{C_{T} }}{{P_{T} }}$ is the national per capita carbon deficit, and $\frac{{C_{di} }}{{P_{i} }}$ is the per capita carbon deficit of province i. If the per capita carbon deficit of province i is greater than the national per capita carbon deficit, the value of $\left( {\frac{{C_{di} }}{{P_{i} }} - \frac{{C_{T} }}{{P_{T} }}} \right)$ will be positive, and the carbon deficit of province i will increase, reflecting the principle of “further punishment for the more”. Conversely, if the per capita carbon deficit of province i is less than the national per capita carbon deficit, the value of $\left( {\frac{{C_{di} }}{{P_{i} }} - \frac{{C_{T} }}{{P_{T} }}} \right)$ will be negative, and the carbon deficit of province i will decrease, reflecting the principle of “further reward for the less”.

Ecological compensation models and sensitivity indices

The ecological compensation amount are calculated based on carbon trading prices and the allocated carbon deficit, where the carbon trading prices for 2013–2019 are the corresponding prices in China’s carbon trading market, and the carbon trading prices for 2001–2012 are the average price (23.26 RMB/t) in China’s carbon trading market from 2013 to 2019 due to lack of corresponding data. When the ecological compensation price is positive, it indicates that the region needs to pay the amount, and vice versa, it receives compensation. The compensation quota is calculated based on the carbon trading price for each year, as follows:

$${\text{CC}}_{{\text{i}}} = \overline{\overline{{C_{di} }}} \cdot {\text{V}}_{{{\text{CO}}_{2} }}$$

(16)

where, CC_i is the compensation amount for province i; ${\text{V}}_{{{\text{CO}}_{2} }}$ is the carbon trading price, taken as RMB 23.26/t.

Considering that future ecological compensation amounts will need to be allocated from regional revenues, we have developed the concept of regional ecological compensation sensitivity, which is the calculation of the offset amount as a percentage of a region’s annual fiscal revenues or expenditures, as shown in Equation.

$${\text{S}}en_{{\text{i}}} = \frac{{{\text{CC}}_{{\text{i}}} }}{{{\text{Re}}_{i} }}$$

(17)

where, $Sen_{i}$ is the regional ecological compensation sensitivity; ${\text{CC}}_{i}$ is the offset amount for province i; ${\text{Re}}_{i}$ is the fiscal revenue or expenditure for province i.

All data collected in this paper were ensured to be processed by missing value filling as well as outlier testing, missing data from provinces in 2001–2019 were ensured to be continuous and available by Scikit-learn method, and one-way ANOVA with Tukey post-hoc test was used to ensure that outliers were eliminated, All significant level was at P < 0.05.

Analysis of results

Characteristics of the spatial and temporal differentiation of carbon emissions

Figure 2 shows the trend of carbon emissions by Chinese provinces (municipalities and autonomous areas) from 2001 to 2019. Carbon emissions in each region increase year by year as the years pass. Shandong province’s carbon emissions increased by more than 1.1 billion tonnes over the course of 19 years, and from 2004 onwards, Shandong province’s carbon emission growth trend began to gradually surpass that of other provinces; Tibet province has the smallest growth, approximately 36.98 million tonnes from 2001 to 2019.

Fig. 2

2001–2019 Trend chart of China’s carbon emissions.

The provinces with the highest growth rates were Hainan, Inner Mongolia, Shandong, Guangxi, Fujian, Xinjiang, Shaanxi, Ningxia, and Qinghai. Between 2001 and 2019, Hainan Province’s carbon emissions multiplied by six, with relatively large growth rates from 2005 to 2007 and relatively flat growth rates in other years; Inner Mongolia’s carbon emissions multiplied by 6.8; Shandong, Guangdong, and Tibet Provinces’ carbon emissions multiplied by six; Carbon emissions increased to five times their 2001 level, as shown in the graph below, with Shandong Province showing the highest growth rate of approximately 28% from 2013 to 2016; Fujian Province increasing to four times its 2001 level; Xinjiang Province increasing to approximately 4.6 times its 2001 level; Jiangsu, Yunnan, Zhejiang, and Sichuan showing slightly larger growth in carbon emissions over the 19-year period. Other provinces’ carbon emissions (municipalities directly under the Central Government and autonomous areas) have changed relatively flat as year increases.

The carbon emissions of Liaoning and Shandong provinces decreased from 2012 to 2013, and the carbon emissions of Gansu, Jilin, and Yunnan provinces essentially overlapped during the 19-year period, with Fujian Province overlapping with the aforementioned three provinces from 2001 to 2009 and gradually increasing thereafter, removing the overlap area. In a manner of speaking, the trend of carbon emissions also exposes the economic development dynamics of each location.

Spatial and temporal partitioning characteristics of carbon absorption

Figure 3 shows the trajectory of carbon absorption by Chinese provinces (municipalities and autonomous regions) from 2001 to 2019. From 2001 to 2019, carbon intake in each region rose annually on average. The Xinjiang Uyghur Autonomous Region had the largest carbon uptake in all 19 years, with 4.7 billion tonnes in 2019. Tianjin had the lowest carbon uptake in all 19 years, with 12.62 million tonnes in 2019, followed by Shanxi and Hebei provinces, which also came last. Inner Mongolia ranks first in terms of growth rate, with an increase in carbon absorption of around 2.2 billion tonnes between 2001 and 2019; Shanghai province ranks last, with an increase of approximately 4.43 million tonnes. 2008 was a clearer turning point for most regions between 2001 and 2019; with the exception of Shanxi, Hebei, and Tianjin, which had minimal change in that year, most other provinces contrasted favourably to those three. In 2008, carbon absorption rose the maximum in Qinghai province, where it grew by around 115.82 million tonnes.

Fig. 3

2001–2019 Trend chart of carbon absorption in China.

Several provinces, including Shanxi, Chongqing, Guizhou, Tianjin, Hebei, Beijing and Ningxia, had higher growth rates from 2001 to 2019 overall, with Shanxi and Chongqing having the highest growth rates, increasing to approximately double their 2001 rates in 2019; followed by Guizhou with a growth rate of 95% over the 19-year period; Tianjin with a growth rate of 83% and Hebei, Beijing and Ningxia with a growth rate of 74%. In contrast, Tibet, Jilin, Hainan, and Heilongjiang provinces had lower carbon absorption growth rates over the 19-year period, with Tibet having the lowest growth rate of 8%, followed by Jilin with 15%, and Hainan and Heilongjiang with 18%.

Spatial and temporal variation characteristics of the carbon deficit

Comparison of forecasts

Based on the carbon deficit data obtained from the above calculation model and combined with China’s policy of double carbon target, this paper has fitted and projected the data. Assuming the carbon neutrality target is achieved in 2060, the following Fig. 4 shows the pattern of China’s national carbon deficit data.

Fig. 4

Fitted graph of carbon deficit data with year assuming the carbon neutrality target is achieved in 2060.

The fitted curve adheres to the law of quadratic distribution, where y = − 27.791x2 + 1769.923x− 6385.915 (R² = 0.955), and the fit is highly accurate. Where: axis of symmetry x = − b/2a = − 1769.923/27.791*2 = 31.84, i.e. The carbon deficit peaks in 2031.84 (which roughly coincides with China’s pledge to reach the carbon peak target), at which point the carbon deficit is approximately 220 million tonnes, and China has at least 30 million tonnes of carbon deficit credits remaining before reaching the carbon peak in 2030. To achieve carbon neutrality by 2060, the yearly average decrease in carbon deficit throughout the 30-year period from 2030 to 2060 would be around 7.85 million tonnes, with the rate of reduction increasing annually.

Assuming that the 2030 carbon peak target is achieved, the section outlines the pattern of China’s national carbon deficit data. The fitted curve also satisfies the quadratic distribution (Fig. 5), where y = − 30.436x2 + 1826.16x− 6616.835 (R² = 0.956), and the accuracy of the fit remains very high. The symmetry axis x = − b/2a = 30, it is the anticipated top of the carbon deficit in 2030, when the carbon deficit is around 20 million tonnes, and the intersection of the image with the x-axis indicates that 2056 is a carbon neutral year, four years before the “carbon neutral by 2060” goal.

Fig. 5

Fitted graph of carbon deficit with year assuming that the 2030 carbon peak target is achieved.

The above fitting demonstrates that the trend obtained from the data in this paper is essentially consistent with the trend and time of China’s projection, which is in line with China’s international commitment to “strive to reach the peak by 2030 and strive to achieve carbon neutrality by 2060,” which is a strong indication of the rigour and accuracy of China’s national policy formulation. Furthermore, it proves the scientific and coherence of the dual carbon targets.

Distribution pattern of carbon deficit before and after redistribution by province

To condense the expression, only the GIS graphs for the years 2001, 2005, 2010, 2015, and 2019 are presented because the study spans an excessively lengthy time period. Different colours of red and green represent the absolute extent of the carbon shortfall. It is essential to note that the intervals representing carbon deficits in the legends of all graphs in this section are consistent.

As shown in Fig. 6, as time progresses, the area with a positive carbon deficit increases and the area with a negative carbon deficit declines, the green area progressively gets lighter and the red area tends to become darker. From 2001 to 2019, the Inner Mongolia Autonomous Region, Heilongjiang, Jiangxi, Guangxi, Sichuan, Yunnan, Tibet, Gansu, and Qinghai all had negative carbon deficit readings, but the absolute value of the carbon deficit decreased each year. Prior to 2005, Jilin and Hainan had negative carbon deficits, whereas Shaanxi and Fujian had negative deficits until 2004 and 2005, respectively. Xinjiang Uyghur Autonomous Region only had a zero carbon deficit in 2017 and 2019, and had a carbon deficit in all other years. Notable is that Hunan Province’s carbon deficit fluctuates, reducing to a negative value in 2007, reverting to a carbon deficit in 2008, becoming negative again from 2009 to 2018 with increases and falls, and then reverting to a carbon deficit with unstable performance in 2019. Shandong province has demonstrated the most substantial improvement among regions that have always had a positive carbon deficit, with an increase of 660.26% over 19 years.

Fig. 6

Carbon deficit before redistribution in China 2001, 2005, 2010, 2015, 2019.

Following the redistribution of the model outlined above, the Fig. 7 shows the carbon deficit in each region. After the redistribution, both the red and green areas have become darker, indicating that the absolute magnitude of the carbon deficit has increased. Carbon deficits were recorded in all 19 years in Xinjiang Uyghur Autonomous Region and Hunan Province. In Jilin Province, the number of years with carbon deficits increased from negative values before to the redistribution to positive levels in 2007–2009 and 2015–2019. Other regions’ carbon deficits did not shift from positive to negative, except for an increase in absolute magnitude. In addition, unlike the negative range of carbon deficit before the redistribution, the negative region obtained after the correction gradually decreases.

Fig. 7

Carbon deficit after redistribution in China 2001, 2005, 2010, 2015, 2019.

Ecological compensation and spatial autocorrelation by region

The graph (Fig. 8) below shows the spatial and temporal distribution of ecological compensation amounts by province (municipality directly under the Central Government and autonomous regions), with the red areas representing the subjects of compensation and the green areas representing the objects of compensation. The graph demonstrates that the subject and object areas of compensation do not increase or decrease each year, but are more varied. However, the object area of compensation has decreased during the past 19 years, while the subject area has increased. Fujian, Hainan, Shaanxi, and Chongqing were compensated until 2009, 2007, 2006, and 2002, respectively, and have since become compensation subjects that have always been required to pay compensation. Jilin was a compensated object area from 2001 to 2009, shifted to a compensated subject area from 2010 to 2014, and is once again a compensated object area from 2015 to 2019; Guizhou was a compensated object area from 2001 to 2002, shifted to a subject compensated area from 2003 to 2014, and is once again a compensated object area from 2015 to 2019. From 2001 to 2019, provinces such as Shandong, Beijing, Tianjin, Hebei, and Zhejiang are required to pay compensation to other provinces as compensated subjects. In contrast, the territories of Inner Mongolia, Heilongjiang, Guangxi, Hunan, Sichuan, Yunnan, Tibet, Gansu, Qinghai, and Xinjiang have always been objects for compensation from other provinces.

Fig. 8

Ecological compensation in China 2001, 2005, 2010, 2015, 2019.

The figure (Fig. 9) below shows the findings of the spatial and temporal heterogeneity analysis of compensation amounts. The white portion represents the “insignificant area,” the pink portion represents the “high-value aggregation area,” the dark blue portion represents the “low–high-value aggregation area,” and the light blue portion represents the “low-low-value aggregation area.” The area in light blue represents the “low-low value aggregation area.” As the graph demonstrates, the central and coastal regions of China have greater compensation levels, as do their neighbouring areas. In contrast, the Inner Mongolia region has a lower absolute value but greater compensation in neighbouring regions. The ‘low-low value area’ tends to increase in a southeasterly direction with each passing year, whereas the ‘high-value aggregation area’ continues to shrink, indicating that the tendency for broad high-value aggregation is waning.

Fig. 9

Moran’s I of ecological compensation in China 2001, 2005, 2010, 2015, 2019.

RDA analysis

Figure 10 represents the redundancy analysis (RDA) for each province (municipality directly under the Central Government and autonomous region) in China from 2001 to 2019 before and after the redistribution of GDP, population size, year, regional area, carbon emissions, carbon absorption, and carbon deficit. The total of the eigenvalues of the first two axes is 64%, which can include the majority of the information; therefore, it is possible to use the first axis as the principal component axis, which is mostly decided by the first axis. For each region, the greater the similarity of the data, the closer the points of different colours in the graph are to one another.

Fig. 10

RDA analysis graph.

GDP, year, and population size are positively correlated with carbon emissions and carbon deficit before and after redistribution, but not with carbon absorption; year is positively correlated with carbon absorption, but the degree of correlation is not significant; regional area is not significantly correlated with carbon emissions, but negatively correlated with carbon deficit before and after redistribution. Carbon absorption is inversely connected with carbon deficit prior to and following redistribution, but not with carbon emissions; carbon emissions are positively correlated with carbon deficit prior to and following redistribution. GDP is highly and positively connected with population size. The degree of influence is proportional to the length of the ray, and it is evident from the graph below that carbon emissions, carbon absorption, and carbon deficit redistribution have a higher impact on each preceding and succeeding region. Comparing population size, GDP, and year, it can be determined from the length of the line that GDP has the greatest impact, followed by population size and then year.

Ecological compensation sensitivity

As shown in Table 5, the ecological compensation sensitivity of each region fluctuates from year to year. However, in terms of the overall trend, the absolute value of each region’s ecological compensation sensitivity has decreased gradually since 2001, increased sharply from 2013 to 2015, and then decreased again after 2016. Tibet, Qinghai, Inner Mongolia, Xinjiang, Heilongjiang, and Yunnan are relatively more sensitive, with Tibet receiving the largest share of compensated income and Shaanxi province receiving the smallest share, with an average annual compensation of − 0.35%; among the subject area of compensation, Shanxi province has the greatest ecological compensation sensitivity, i.e. the compensation payment has a relatively large impact on its economy. Except for the Tibet Autonomous Region and the Ningxia Autonomous Region, the average sensitivity of the object area of compensation is around − 18.54%, while the average sensitivity of the subject area of compensation is approximately 5.07%. In contrast, the sensitivity of the compensated object area is bigger in absolute terms, indicating that the financial expenditures arising from the compensation are less sensitive than the compensation’s income.

Table 5 Ecological compensation sensitivity in 2001, 2005, 2010, 2015 and 2019 (%).

With the exception of Tibet and Qinghai, the majority of regions have ecological compensation sensitivity values of approximately 10% in absolute terms over the majority of years. For the primary regions that pay ecological compensations, ecological compensation sensitivity values are typically between 0 and 10%, with 2 to 9 regions still able to reach 10%, or even 20% or 30% of the total before 2009, and after 2009, with the exception of five provinces with absolute values greater than 10% in 2013, all other years have values below this threshold. From 2001 to 2019, the number of regions with absolute values above 10% declined from 15 to 4, and the number of districts with absolute values exceeding 20% decreased from 11 to 2.

Discussion

About the carbon deficit

Carbon deficit before and after redistribution by province

In this paper, the carbon absorption accounting is more comprehensive, particularly the calculation of the main carbon absorption in forest and wetland areas, which increases the calculation of the absorption value, and the carbon absorption value increases or decreases after the triple correction of GDP, population, and area; consequently, the positive carbon deficit area differs from the results of other papers, and the negative carbon deficit range increases year by year b. The region of negative carbon deficit before redistribution increases year by year, whereas the negative area gained after redistribution diminishes.

Zhan et al.³² generated carbon absorption correction coefficients using the equivalent factor approach of ecosystem service function valuation and adjusted the agricultural carbon absorption of each region by assessing the ecosystem service capacity of the various regions. In addition, the modified pearl growth curve model was used to compute the carbon compensation coefficients for various economic development levels. As a result of the treatment of carbon absorption correction coefficients, the net carbon absorption of regions with poorer ecological backgrounds were amplified to a certain extent, resulting in a reduction in the carbon deficit of those regions. This paper argues that magnification is not conducive to the establishment of a policy mechanism that rewards less and penalises more.

Chen and Jiang³³ also obtained carbon compensation coefficients by adopting an enhanced growth curve, with the level of economic development and willingness to pay serving as a proxy for the compensation capacity. However, their calculations of carbon emission accounting is corrected based on the level of science and technology, so that regional carbon emissions are amplified, i.e. the amount of carbon deficit in areas with high technology level will increase, and this paper argues that this amplification effect is also not conducive to encouraging the development of regional technology level.

The carbon deficit redistribution model constructed in this paper avoids inappropriate scaling up and scaling down of the original data, so that the total amount of carbon deficit remains constant before and after redistribution, while redistribution is based on regional resource endowment and natural attributes, etc. The updated carbon deficit is conducive to promoting the continuation of the green development model or initiating proactive changes in the region, ultimately generating a development climate that “rewards less and punishes more.”

Carbon peaking and carbon neutrality targets

To actively reduce carbon emissions, Chinese President Xi Jinping announced at the 75th General Debate of the United Nations General Assembly in 2020, “China will increase its autonomous national contribution, adopt more strict policies and measures, strive to peak CO₂ emissions by 2030, and work toward carbon neutrality by 2060.” In the long run, achieving the “two-carbon” goal, reducing carbon emissions, and making joint global efforts will help mitigate the adverse effects of climate change, reduce the damage to the economy and society, and bring man and nature back to peace and tranquility³⁴. Achieving carbon peaking and carbon neutrality are an unavoidable decision to concentrate on solving the outstanding problems of resource and environmental constraints.

In this paper, we fit the carbon deficit data calculated by the constructed redistribution model, Chang³⁵ mentioned that China will reduce carbon dioxide emissions in the future, with GDP of less than 8.752%, and emissions will peak by 2030. Xie et al.³⁶ developed the STIRPAT model, discussed that all 30 provinces in China would reach the peak emissions by 2030, and strongly demonstrated the preconditions of this paper. And found that the curve is basically consistent with the national prediction of carbon peak and carbon neutral trend, and that the overall trend is similar to that of developed countries such as Belgium and Germany, with a stable peak after the carbon peak and no longer rebounding upwards, but on a downward trend after the peak, in accordance with the inverted “U” environment^37,38. This helps validate the reasonableness and scientificity of the national goal of “reaching the carbon peak in 30 years and carbon neutrality in 60 years” under the development of green, low-carbon, and high-quality development in parallel.

About ecological compensation amounts and spatial correlation patterns

Ecological compensation amounts by region

Shanxi (RMB 21 billion), Jiangsu (RMB 20.9 billion), Shandong (RMB 17.2 billion), Shaanxi (RMB 16.9 billion), and Tianjin (RMB 13 billion) are the top five provinces paying annual ecological compensation costs, according to Gao et al.²⁰ while this paper calculates that the regions with the highest average annual expenditure costs are Shandong (RMB 36.1 billion), Jiangsu (RMB 22.2 billion), Liaoning (RMB 21.6 billion), Hebei (20.7 billion yuan), and Shanxi (17.9 billion yuan), which are overall higher than Yang’s results. Tibet, Inner Mongolia, Qinghai, Yunnan, and Heilongjiang have the highest levels of ecological compensated objects according to this study, with southwesterly regions having greater amounts than Sichuan, Xinjiang, Guangdong, Heilongjiang, and Hunan. Due to the varied selection of the two for calculation, this tiny variance exists. Gao et al.²⁰ also suggest that the provinces paying ecological compensation costs are primarily in the central and eastern regions, whereas the provinces receiving compensation are primarily in the northwest, which is basically consistent with the conclusions of this paper: the provinces that should be compensated are concentrated in the western regions, whereas the provinces that should pay compensation costs are primarily in the central and developed coastal regions.

In calculating ecological compensation, Wan et al.³⁹ used carbon pressure (carbon emissions per unit area), carbon emissions per capita, and carbon intensity (carbon emissions per unit GDP) as measurement indicators, and calculated compensation amounts based on these indicators with carbon price, GDP, and population area respectively. Except for measurement indicators, the quota calculation methodology is similar to that of this paper. Chen and Jiang³³ included the level of science and technology and ecological background based on the consideration of population size and land area in different regions, and the provinces (cities) with the highest average compensation amounts were primarily Shanghai, Liaoning, Jiangsu, Tianjin, and Shandong, which differs slightly from the top-ranked provinces obtained in this study. We found there is a tendency for their papers to zoom in or out, making areas with high levels of science and technology more highly compensated. Yu et al.³⁰ applied to 47 major socio-economic industries in calculating carbon emissions, such as energy producing industries, light industries, heavy industries, etc., Which may include double counting. Carbon absorption, on the other hand, only takes into account carbon sequestration by forests and urban green vegetation, ignoring the significance of carbon sequestration by wetlands and grasslands, etc. Forests, grasslands, wetland, and crops are the primary carbon absorption data in this paper. The greater the area of the region, the greater the amount of carbon dioxide absorbed by these plants, which supports the conclusion that area is positively correlated with carbon absorption. Likewise, carbon deficit = carbon emission–carbon absorption, so when carbon absorption increases, carbon deficit decreases, which also confirms the negative correlation between carbon deficit before and after redistribution and area.

Meanwhile, Yu et al.³⁰ thought Shanxi is the region with the highest average yearly payment, while we think Shandong is the region with the highest average annual payout, with variances in the results. And its use of the modified pearl growth curve model in calculating ecological compensation to estimate the ecological compensation coefficient of each province merely magnifies the compensation amount for some locations without adjusting the payment area. This study does a better job of resolving the issue since it reflects less reward and more punishments.

Spatial autocorrelation analysis

In the central and coastal regions of China, where compensation levels are higher, the “low-low value zones” have a tendency to grow in a southeasterly direction over the years, indicating a rising trend of low compensation areas. In contrast, the “high value aggregation areas” tend to diminish with time, with a wide variety of high value aggregation locations, whilst the “low value aggregation areas” tend to travel southward. The overall decreasing compensation pattern is generally compatible with the findings of Chen and Jiang³³ and the compensation pattern is comparable to the structure of each region’s economic development level.

Drivers analysis

Regions with relatively high GDP are often accompanied by increases in carbon emissions. From 1995 to 2007, Zhu et al.⁴⁰ argued that the overall carbon emissions of all regions rose annually, which is consistent with this paper’s finding that year is positively connected with carbon emissions. The development of regional economy needs energy consumption and industrial expansion, so the carbon deficit is also positively related to it. Using the STIRPAT model and panel data methodologies, they also analysed the impact of population, economics, and technology on regional carbon emissions, concluding that rapid economic growth is the most significant factor in the increase of carbon emissions. The expansion of the population means increased demand for energy and resources, and increased carbon emissions from living consumption, which in turn affects the carbon deficit; Lu et al.³¹ also explored the causative relationship between GDP growth and carbon dioxide emissions in China using multivariate cointegration causality tests and concluded that there is a two-way causal relationship between the two variables, which is consistent with the findings of this paper. This is consistent with the premise of this study that there is a large positive correlation between GDP and CE (carbon emissions), which is also supported by the RDA analysis.

About ecological compensation sensitivity

There are few results that quantify pay as a percentage of fiscal revenue or expenditure, with the GDP serving as the predominant reference point. Wan et al.³⁹ compare the compensation amount to the regional GDP, whereas Li et al.⁴¹ calculate the compensation capacity as an indicator and present it as the regional GDP per capita ratio. This study claims that the amount of compensation will be factored into the local government’s finances and expenditures, hence it is more appropriate and reasonable to use these two variables for the computation. As GDP is defined in terms of final goods and services, the measurement of the impact of compensation costs on GDP lacks meaningfulness and comparability in terms of income and expenditures, a problem that this study can rectify.

We discovered that the fiscal expenditure sensitivity of the compensation subject regions is generally less than the fiscal revenue sensitivity of the compensation object regions, and that the few regions with the highest ecological compensation sensitivity are essentially identical to those with the highest compensation amount. Gao et al.²⁰ calculated the carbon compensation priority index and determined the order of compensation based on the percentage of GDP, such that the payment regions are Beijing (0.068%), Anhui (0.076%), Zhejiang (0.086%), Fujian (0.093%), and Jiangsu (0.127%), while the average annual sensitivity of the payment regions in this paper is relatively higher in absolute value, namely Beijing (3.223%), Anhui (2.766%), Zhejiang (3.989%), Fujian (0.666%) and Jiangsu (5.939%), the comparison of the data shows this approach further highlights the difference in the impact of provincial background factors on carbon compensation.

Ecological compensation sensitivity can play a significant role in the majority of provinces, according to an examination of the proportion of various industries in regional fiscal income vs ecological compensation sensitivity. The transport, storage, and postal industry in Sichuan Province, for instance, had a revenue share of 3.18% in 2019 and an ecological compensation sensitivity of 3.13% in the same year; the wholesale and retail industry in Heilongjiang Province had a revenue share of 7.36% in 2016 and an ecological compensation sensitivity of 7.84%; the construction industry in Yunnan Province had a revenue share of 9.7% in 2015 and an ecological compensation sensitivity of 9.67%; and the accommodation industry in 2013, the ecological compensation sensitivity of Hunan’s lodging and food service industries was 2.20% and 2.28%, respectively. It is evident that the ecological compensation sensitivity of the respective regions is comparable to the income touch of the major industries. By comparing them in this way, the ecological compensation amount is no longer an unclear number; it has the same economic impact as the region’s industries, which can lead to the emergence of new industries (related to ecological restoration, ecological remediation, etc.) and encourage regions to pay greater attention to the ecological compensation mechanism.

Comparing the proportion of specific items in regional fiscal expenditures to the ecological compensation sensitivity reveals that the ecological compensation sensitivity can have a considerable impact in the majority of provinces. For example, the share of public safety costs in Hebei province’s fiscal expenditure in 2012 was approximately 5.57%, which is very close to the ecological compensation sensitivity (6.0%) for the same period. The share of transport costs in Shanxi province’s fiscal expenditure in 2019 was approximately 6.81%, which is very close to the ecological compensation sensitivity (6.2%) for the same period. The aforementioned findings indicate that ecological compensation is roughly equivalent to the region’s spending on health care, public safety, or transportation in the same year, which will inevitably put a lot of pressure on regional finances and force governments to prioritise environmental protection and low-carbon economic development strategies, allowing the market to force governments to make corresponding changes.

Conclusions and suggestions

Conclusions

This paper presents a revised model for accounting carbon emission and carbon absorption based on publicly available data. The model was used it to calculate and analyse the spatial and temporal distribution characteristics of carbon deficits in 31 provinces (municipalities directly under the Central Government and autonomous regions) in China from 2001 to 2019.Additionally, a carbon deficit redistribution calculation model was proposed for the first time based on the basis of comprehensive data. This strategy promotes energy conservation and pollution reduction by implementing the concept to the idea of “further rewards for less and penalties for more” The spatial and temporal variation characteristics of ecological compensation amount for each province (municipality directly under the Central Government and autonomous region) was then derived by integrating the carbon trading price, taking into account the spatial and temporal variation characteristics. An RDA analysis was conducted on the total carbon emissions, total carbon absorption, carbon deficit prior to and after redistribution, and four factors: GDP, population size, regional area, and year for each Chinese region from 2001 to 2019. The study concluded with a new concept of regional ecological compensation sensitivity. This concept effectively measures the impact of the ecological compensation amount generated by the model on the fiscal revenue or expenditure of 31 provinces (including municipalities directly under the Central Government and autonomous regions).

The following are the specific conclusions:

• (1) Between 2001 and 2019, China’s carbon emissions and carbon absorption increased, With Shandong, Hebei, and Liaoning provinces having the highest emissions. Carbon absorption patterns were similar across most regions during the past 19 years, with the western, southwestern, and northeastern regions absorbing more carbon than the central and eastern regions. Changes in carbon indicators in each location strongly correlate with their respective demographic and economic growth.

• (2) The central and developed coastal regions of China have relatively larger carbon deficits and are the main regions of China that pay compensation, while other regions are the recipient regions that receive compensation. The compensated regions tend to expand towards the southeast, and the amount of compensation paid is less sensitive to the financial touch of the main regions, i.e. the regional ecological compensation, than the compensated guest regions. The fitted curve pattern of carbon deficit over time in this paper is consistent with China’s dual carbon ambitions (carbon peaking by 2030 and carbon neutrality by 2060), and there is close coordination among the aims.

• (3) High compensation aggregation areas are concentrated in the central-eastern and coastal regions, forming a decreasing pattern in all directions. Conversely low aggregation areas tend to move south. The overall spatial distribution characteristics are similar to the layout of economic development levels; GDP, year, population number and carbon emissions shows a positive correlation with carbon deficit redistribution before and after, while year and regional area show a negative correlation.

• (4) The sensitivity of ecological compensation is roughly proportional to the revenue share of the catering, storage, and postal industries among compensation clients and the expenditure share of health care, public safety, and environmental protection among compensation subjects. Therefore, it can have a significant impact on both compensation subjects and compensation clients, compelling regional governments to make necessary adjustments. This study’s findings provide a strong incentive for regional governments to make modifications towards ecological conservation and low-carbon economic growth.

The results will impact China’s quantification of inter-regional ecological compensation criteria and will help shape macroeconomic policies for green and low-carbon growth.

Suggestions

In the process of improving the ecological compensation mechanism, ensuring fairness between ecological benefits and ecological damages and mobilizing the enthusiasm of regional emission reduction and exchange enhancement have been the focus of research. Based on the research findings, this paper puts forward the following suggestions:

• (1) Determine the status of the main body of carbon revenue and expenditure of each region each year according to the positive or negative carbon deficit obtained in the previous year. As an important environmental tax in the context of dual-carbon, the strict setting of carbon offset costs will affect the attitude of regional emission reduction and sink enhancement, and the publicizing of regional offset sensitivity under the GDP ratio will effectively increase the activeness of Chinese provinces in participating in the mechanism and maintaining it.

• (2) For regions with faster GDP growth and larger carbon deficits, the design of eco-compensation should focus on incentivizing the optimization and upgrading of their industrial structure, promoting green technological innovation, and reducing high-energy-consuming and high-emission industries. When implementing eco-compensation, these regions can set stricter carbon emission targets and provide special funds for green development projects, such as support for the research and development and application of clean energy technologies; for densely populated regions with serious carbon deficits, eco-compensation can focus on improving energy efficiency, promoting green travel and popularizing low-carbon lifestyles. Specific measures may include subsidizing energy-saving home appliances and building public transportation facilities to reduce per capita carbon emissions, as well as strengthening environmental education to enhance residents’ awareness of environmental protection. At the same time, the eco-compensation policy should be dynamically adjusted and forward-looking, and the compensation standards and measures should be continuously updated over time to adapt to new emission reduction needs and technological progress.

• (3) In order to avoid the internalization of the externalities of ecological compensation, market forces should be introduced on the basis of touching the government’s finances, and the burden of the government’s “transfusion-type” support of compensation funds should be alleviated by promoting the realization of the value of ecological products, so as to allow synergy between the government and the market, expand diversified compensation channels, and increase the diversity of compensation types, compensation contents and compensation methods. The government and the market should work together to expand diversified compensation channels and increase the diversity of compensation types, contents and methods.

Although we have given some theoretical policy recommendations based on the empirical results, there are still parts of this paper that can be further investigated. In the future, when data are available, we can consider introducing more environmental factors as variables to clarify the impact of water resources, biodiversity, pollution load index and other variables on carbon deficit. In addition, the current calculations in the carbon field still rely on manual data processing and calculations, etc. The introduction of AI and the development of a complete set of algorithms are important to get rid of the complexity of data processing, to be able to explore the characteristics of the data of eco-compensation, and to improve the accuracy of the calculations. In the future, we can expand our research area, increase the number of years of research, and include more countries that implement environmental taxes in the scope of research.