Exploring the Pathways of Achieving Carbon Peaking and Carbon Neutrality Targets in the Provinces of the Yellow River Basin of China

Abstract

Achieving carbon peaking and carbon neutrality is an intrinsic requirement for sustainable development. The industrial structure primarily characterized by the chemical and energy industries poses a hindrance to the attainment of carbon peaking and carbon neutrality goals in the provinces of the Yellow River Basin of China. Predicting the time of carbon peaking and carbon neutrality and exploring the pathways of carbon peaking and carbon neutrality is an urgent issue for the government to address. The STIRPAT and InVEST models were used for the carbon emissions and carbon sequestration estimation in the nine provinces and regions of the Yellow River Basin from 2010 to 2060. The results show that the study area will realize carbon peaking in 2030 under the baseline scenario, with the carbon emission of 4146 million tons. Under the high-emission scenario, the study area will realize carbon peaking in 2035, with the carbon emission of 4372 million tons. Under the low-carbon energy-saving scenario, the study area will realize carbon peaking in 2025, with the carbon emission of 3909 million tons. The entire study area cannot achieve carbon neutrality in 2060 under the three scenarios. Under the baseline and high-emission scenarios, only Qinghai and Sichuan can realize carbon neutrality by 2060, and under the low-carbon energy-saving scenario, Sichuan, Qinghai, Shaanxi, and Gansu will achieve carbon neutrality on time. This research indicates that attaining carbon peaking and carbon neutrality can be accomplished by implementing strategies such as encouraging the growth of clean energy, managing energy usage, refining the industrial structure, and strengthening the ecosystem’s carbon sink.

1. Introduction

The continual increase in atmospheric CO₂ is a significant contributor to extreme weather events, climate warming and acid rain, which can have detrimental effects on crop growth and human health [1,2,3]. To mitigate ecosystem degradation caused by CO₂ emissions, many nations have set carbon neutrality targets and actively promoted the development of green, low-carbon initiatives. The European Commission published the draft European Climate Law proposing legislation to achieve carbon neutrality by 2050 [4]. Similarly, countries like the USA [5], Canada [6], Argentina [7], Japan [8] and other countries have also made commitments to reduce greenhouse gas emissions in the future. In order to actively address environmental issues related to climate change, China set out a “dual carbon” target [9,10,11]. Until 2016, China achieved the 2020 carbon emissions reduction target promised by the government. If the current rate is maintained, China will be able to fulfill the carbon peaking target by 2030. However, there are still several regions with low-carbon emission reduction potential, such as Shanxi, Ningxia, Gansu, Inner Mongolia, Shaanxi, distributed along the Yellow River [12]. Therefore, determining the timing of carbon peaking and carbon neutrality, along with assessing the spatial–temporal dynamics of carbon emissions and sequestration in the nine provinces of the Yellow River basin (NPYRs), are hot issues.

Regarding the study on carbon emissions, the mainstream methods include the Environmental Kuzentz Curve (EKC) method, the Logarithmic Mean Divisia Index (LMDI) model, the BP neural network model, the gray prediction method and the STIRPAT model, etc. Compared with other models, the STIRPAT model is expanded from the IPAT model, with more flexible model parameters. It uses elasticity coefficients to evaluate how wealth, population, and technological advancements affect environmental pressure [13,14,15]. The STIRPAT model facilitates the estimation of coefficients in the model, provides a simple analytical framework to appropriately decompose and evaluate the influencing factors on environment, and it is able to establish accurate and reasonable regression relationships between explanatory variables and explained variables for the identification and prediction of carbon-emission-influencing factors based on the elimination of covariance between the explanatory variables [16]. Liao et al. (2022) used the STIRPAT model to evaluate the factors impacting carbon emissions in Sichuan province [17]. Ofori et al. (2023) investigated the factors influencing carbon pollution in the emerging economy with the STIRPAT model [18]. Roy et al. (2017) explored the factors influencing environmental change in India and other nations using the STIRPAT model [19]. Therefore, this paper chooses the STIRPAT model to analyze the factors influencing carbon emissions along the NPYRs and to forecast the carbon emission to study the carbon peak.

However, the technological level is not a single component; rather, it is a combination of many different independent factors that affect the environment [20,21]. Vélez-Henao et al. (2019) suggested that the technological level should be expressed as a group of variables to capture the various dimensions of technology, with industrialization, energy intensity, and energy structure being seen as the most comprehensive and suitable approach to define the technological level [22]. Numerous studies have demonstrated that the rate of urbanization significantly affects carbon emissions [23,24]. Therefore, this paper divides the technology level into urbanization rate, industrial structure, energy intensity, and energy structure to extend the STIRPAT model, thereby improving the accuracy and convenience of prediction. Then, OLS regression is utilized to screen out the key factors affecting CO2 emissions. The extended STIRPAT model, incorporating key influencing factors, can effectively capture nonlinear relationships and interactions among variables and thus more accurately predict carbon dioxide emissions in the NPYRs.

The goal of carbon neutrality can be defined as minimizing carbon emissions to the lowest possible level and then offsetting or sequestering any remaining emissions through a certifiable process [25]. Carbon neutrality therefore encompasses both the goal of offsetting or sequestering remaining carbon emissions and the goal of reducing carbon emissions. Carbon sequestration means transferring atmospheric carbon dioxide into long-lived global reservoirs such as the oceans, soils, and biological and geological formations in order to mitigate the net increase in carbon dioxide in the atmosphere [26]. In achieving the carbon neutrality goal, carbon sequestration plays an important role by capturing and storing atmospheric carbon [27], as well as acting as a negative emission technology to reduce carbon emissions [28], and ultimately reducing the atmospheric carbon concentration. In this study, since it is difficult to predict the “minimum emissions”, we only compare the carbon sequestration capacity and carbon emissions in the current year to determine whether the carbon neutrality goal can be achieved.

Conventional approaches for estimating carbon stocks, like the box technique, biomass method, and accumulation method, are straightforward, user-friendly, and widely used. But they frequently fall short of capturing the intricate interactions between long-term trends and large swings in the carbon stock caused by both human activity and natural processes. With the advancement of information technology, the InVEST model, used for estimating carbon stock through model simulations, has emerged. Compared to other research methods, the InVEST model is characterized by low data requirements, fast operation speed, high accuracy and easy operation, and it makes it possible to map the distribution of carbon stocks spatially and dynamically, demonstrating the connection between changes in land use and carbon stocks, and realizing dynamic quantification of the ecosystem service functions’ value, which is different from the monotonous and static assessment of the value of the services in the past [29], and it has been utilized globally to estimate carbon sequestration [30,31,32,33]. He et al. (2016) evaluated the prospective impacts of urban sprawl on regional storage of carbon by applying the InVEST model [34]. Aitali et al. (2022) examined how changes in land use in North Africa’s coastal wetlands affected the storage of carbon with the InVEST model [35]. Huang et al. (2022) analyzed the impact of the rubber forest expansion on regional carbon storage using the InVEST model [36]. Kohestani et al. (2023) evaluated the impacts of landscape changes on carbon stock in river basins in mountainous areas by the InVEST model [37]. Therefore, this study opted for the InVEST model to forecast carbon sequestration in the NPYRs.

Since the reform and opening up, China’s economy has been rapidly developing along the path of “high energy consumption and high greenhouse gas emissions” [38], and China’s carbon dioxide emissions have risen sharply, making it the largest carbon emitting country since 2006 [39]. By 2020, China’s carbon emissions reached 11,895.5 million tons, accounting for about 30 per cent of global carbon emissions. The NPYRs, as an important hub for the country’s energy, chemical, raw materials and basic industries, are responsible for more than one-third of the nation’s carbon emissions in 2022. The dominance of the energy and chemical sectors in the industrial structure poses a challenge to achieving peak carbon and neutrality in the NPYRs. In recent years, several studies have been conducted on the factors driving carbon emissions and the projected trends in carbon sequestration capacity within the Yellow River Basin. The impact of each factor on the change in carbon dioxide emissions and carbon sequestration in the Yellow River basin varies widely. Wu et al. (2023) analyzed the spatial and temporal distribution characteristics of carbon emissions in the Yellow River basin and concluded that population density is the primary determinant of carbon emissions in the region, and predicted that the Yellow River basin will realize carbon emissions between 2030 and 2044 [40]. Wang et al. (2023) identified that energy intensity, economic activity, and population density have positive correlations with carbon emissions in the Yellow River Basin, of which the population effect has the most significant influence, and predicted that the region will realize carbon peaks in 2030 at the earliest [41]. Liu et al. (2022) found that the factors influencing carbon emissions in the Yellow River Basin include economic growth, energy consumption structure, and the structure of the industry [42]. Fang et al. (2021) investigated the factors influencing carbon sequestration capacity in the Yellow River Basin, which is negatively and linearly correlated with the median centrality of forest nodes and positively correlated with the clustering coefficients of grassland nodes [43]. Zhang et al. (2023) examined the spatiotemporal fluctuations and dynamic persistence of carbon source and sink ecosystems in the Yellow River Basin as well as changes in net ecosystem productivity (NEP) [44]. Bu et al. (2020) investigated how soil carbon sequestration capability was affected by wetland conservation and restoration initiatives in the Yellow River Basin. The study showed that wetland ecological restoration and protection initiatives can significantly increase the carbon sequestration capacity of the ecosystems in the Ningxia Basin of the Yellow River, and human activities also have an impact on soil carbon sequestration capacity [45]. However, to date, few studies integrated carbon emissions and carbon sequestration to investigate realization pathways of carbon peaking and carbon neutrality. This study focused on the following three key questions: (1) When will the NPYRs reach their goal of peaking carbon emissions? (2) Can the NPYRs reach their goal of carbon neutrality by 2060? (3) How can the goals of carbon peaking and carbon neutrality be achieved?

2. Materials and Methods

2.1. Study Area

The Yellow River originates from Bayan Har Mountains in Qinghai Province, China, and flows in a specific shape through nine provinces, including Qinghai, Sichuan, Gansu, Ningxia, Inner Mongolia, Shaanxi, Shanxi, Henan and Shandong, and finally injects into the Bohai Sea in Kenli District, Dongying City, Shandong Province (Figure 1). The NPYRs cover an area of approximately 1.3 million km², have a population of 420 million and a GDP of RMB 28.7 trillion. The region is short of water resources, and most of the upper and middle reaches of the Yellow River are located west of the 400-milliliter precipitation line, with a dry climate and little rain. The natural ecology is fragile and economic development is relatively underdeveloped.

2.2. STIRPAT Model

The STIRPAT model was developed from the IPAT model and imIPAT model [46]:

$$I=P\times A\times T$$

(1)

where $I$ is environmental pressure; $P$ is population size; $A$ is affluence; and $T$ is technology level.

The IPAT model and imIPAT model’s applications are extremely limited because they cannot reflect the difference of influencing factors due to simply setting the elasticity factors for the size of the population, affluence and level of technology to 1, and contradicting the environmental Kuznets curve assumption. The STIRPAT model can overcome the deficiencies of the above assumptions in the IPAT model to modify and extend the IPAT model. The equation is as follows:

$$I_i=a\times P_i^b\times A_i^c\times T_i^d\times {\epsilon }_i$$

(2)

where $a$ is the model coefficient; $b$, $c$, and $d$ denote the elasticity coefficients of the factors of population size, affluence, and technological level, respectively. $\epsilon$ is an error term., and the subscript $i$ represents time.

Numerous studies indicate that the population factor exhibits an elasticity close to 1 concerning its relationship with total carbon emissions [46,47]. Therefore, this study directly sets the population factor as unit elasticity and modifies the model to the following form:

$$c{e}_i=a\times A_i^c\times T_i^d\times {\epsilon }_i$$

(3)

where $ce$ represents per capita carbon emissions. After taking the logarithm of the above equation, the model is as follows:

$$lnc{e}_i=lna+cln{A}_i+dln{T}_i+{\epsilon }_i$$

(4)

York et al. (2003) noted that sociological or other controls can be added to the equation as long as they are conceptually consistent with the multiplicative specification of the model (4) [48]. This allows for the inclusion of new variables in the model to decompose influences based on relevant characteristics. Vélez-Henao et al. (2019) suggested that the T should be expressed as a group of variables to capture the various dimensions of technology, with industrialization, energy intensity, and energy structure being seen as the more comprehensive and suitable approach to define the technological level. Industrialization provides information on economic productivity and energy consumption, energy intensity is a direct measure of the efficiency of the industry, and energy structure indicates the permitted capture of resource productivity [22]. Considering that urbanization is a process of industrial agglomeration and increased energy consumption, which affects the changes in carbon emissions [49], this study decomposes $T$ into $IS2$, $IS3$, $ES$, $EI$, $U$. Meanwhile, considering that there is a correlation between per capita GDP ($A$) and population size, the influence factor $P$ is introduced. The equation is as follows:

$$lnc{e}_i=lna+bln{P}_i+cln{A}_i+dln{IS2}_i+eln{IS3}_i+fln{EI}_i+gln{ES}_i+hln{U}_i+{\epsilon }_i$$

(5)

Considering the existence of a large number of non-linear relationships between economic variables, the squared terms of affluence and energy structure are added to the model to determine whether the environmental impact has an inverted U-shaped Kuznets curve with economic development and energy consumption, which ultimately constitutes the following quadratic model:

$$\begin{array}{c}lnc{e}_i=lna+bln{P}_i+cln{A}_i+dln{IS2}_i+eln{IS3}_i+fln{EI}_i\\ +gln{ES}_i+hln{U}_i+{k\left(ln{A}_i\right)}^2+j{\left(ln{ES}_i\right)}^2+{\epsilon }_i\end{array}$$

(6)

where $ce$ denotes the per capita carbon emission. $P$, $A$, $IS2$, $IS3$, $EI$,$ES$, and $U$ denote the size of population, gross domestic product per capita at constant prices, the proportion of secondary industry, the proportion of tertiary industry, the energy intensity, the energy structure, and the rate of urbanization, respectively. $a$ is constant term. $b$, $c$, $d$, $e$, $f$, $g$ and $h$ denote the elasticity coefficients after logarithmic operation of the variables such as the size of population, gross domestic product per capita at constant prices, the proportion of secondary industry, the proportion of tertiary industry, the energy intensity, the energy structure, the rate of urbanization, and other variables. $k$ and $j$ denote the coefficients of elasticity of GDP per capita at constant prices and of the energy structure after logarithmization and their terms in the square. $\epsilon$ is the error term, and the subscript $i$ denotes time. The main variables and their descriptions are shown in

Table 1. Main variables and descriptions.

Variable name	Explanation	Unit
Carbon emissions per capita ($ce$)	Carbon emissions/year-end resident population	millions of tons per 10,000 people
Population size ($P$)	Year-end resident population	ten thousand people
Constant price GDP per capita ($A$)	Real GDP with 2010 as base year/year-end resident population	RMB 10,000
Urbanization rate ($U$)	Urban population/total population	%
Percentage of secondary industry ($IS2$)	Share of secondary industry value added to GDP	%
Percentage of tertiary sector ($IS3$)	Share of tertiary industry value added to GDP	%
Energy intensity ($EI$)	Energy consumption per unit of GDP	tons of standard coal/RMB 10,000
Energy structure (ES)	Clean energy as a share of total energy consumption	%

.

2.3. InVEST Model

The “Carbon Storage and Sequestration” module of the InVEST model provides a good assessment of carbon stocks [50]. The module categorizes ecosystem carbon sequestration into four primary carbon reservoirs: above-ground carbon pool, below-ground carbon pool, soil carbon pool, and dead organic matter carbon pool. The above-ground carbon pool denotes the quantity of carbon sequestered within all living organisms found on the Earth’s surface; the below-ground carbon pool denotes the quantity of carbon sequestered in all living organisms beneath the Earth’s surface; the soil carbon pool denotes the quantity of carbon sequestered in mineral and organic soils; and the dead organic matter carbon pool denotes the quantity of carbon sequestered in all dead organic matter. The specific calculation formula is as follows:

$$C_i=C_{i,above}+C_{i,below}+C_{i,soil}+C_{i,dead}$$

(7)

$$C_{total}=\sum _{i=1}^n{C}_i\times S_i$$

(8)

where $C_i$ is the total carbon density of the type of land use $i$; $C_{i,above}$ is the above-ground carbon density of the type of land use $i$; $C_{i,below}$ is the below-ground carbon density of the type of land use $i$; $C_{i,dead}$ is the dead organic matter carbon density of the type of land use $i$; $C_{i,soil}$ is the soil carbon density of the type of land use $i$; $S_i$ is the area of the type of land use $i$; and $C_{total}$ is the total terrestrial ecosystem carbon stock. Specific values of $C_{i,above}$, $C_{i,below}$, $C_{i,dead}$ and $C_{i,soil}$ for each province in the NPYRs can be found in the Supplemental Materials (Supplemental Materials S1).

2.4. Data Sources

The data of carbon emissions were from the China Emission Accounts and Datasets (https://ceads.net.cn/data/province/by_sectoral_accounting/ (accessed on 6 January 2023)). The data of the size of population, GDP per capita, the rate of urbanization, and the proportion of secondary and tertiary industries were from the Statistical Yearbook issued by the Bureau of Statistics. The energy intensity and energy structure were calculated based on the basis of real GDP, total energy consumption and total clean energy consumption of each province.

Table 2. Mean, median, standard deviation, maximum, minimum of data related to carbon emission variables.

	Mean	Median	Standard Deviation	Minimum	Maximum
P	41,345.4859	41,307.6272	592.2464	40,523.7150	42,147.0419
A	0.0402	0.0392	0.0111	0.0233	0.0577
U	49.5049	49.8934	5.0240	41.9504	56.4624
IS2	45.5187	44.2849	4.8382	38.2334	51.7527
IS3	44.9529	46.0613	5.5693	37.4134	52.0789
EI	0.8051	0.8043	0.1511	0.6319	1.1111
ES	18.3752	17.4259	3.0962	13.7912	23.4613
I	3401.6749	3410.8843	232.3903	2931.2000	3798.6019
ce	0.0824	0.0827	0.0047	0.0723	0.0903

shows the relevant descriptive information about the data. The land use data in 2010–2020 were obtained from the resource and environmental science data registration and publication system (http://www.resdc.cn/DOI (accessed on 19 March 2023)). The land use simulation products for the three scenarios of SSP1-RCP19, SSP5-RCP85, and SSP2-RCP45 in 2025–2060 were obtained from the Geographical Simulation and Optimization System (https://geosimulation.cn/index.html (accessed on 19 March 2023)). According to the land-use classification standard of Ministry of Natural Resoures of China, the land-use types were classified into six types: cultivated lands, forests, grasslands, water bodies, artificial surfaces and bare lands.

3. Results

3.1. Changes in Carbon Emissions

Figure 2 shows that the carbon emissions in the NPYRs rose in volatility from 2010 to 2020. The lowest carbon emission occurred in 2010, with the value of 2931 million tons, while the highest was recorded in 2019, with the value of 3799 million tons. By the end of 2020, regional carbon emissions had reached 3638 million tons, accounting for approximately 36.77% of China’s total carbon emissions. In 2010–2020, notable increases in carbon emissions were observed in 2011, 2012, 2018, and 2019, with a significant increase exceeding 4%. Conversely, reductions in the carbon emissions greater than 2% were noted in 2013 and 2020.

Figure 3 illustrates that the carbon emissions in Shandong and Inner Mongolia were higher than in other provinces, with the average carbon emissions of 829 and 643 million tons, respectively. In contrast, the carbon emissions in Qinghai, Gansu, and Ningxia showed lower emissions, and the average values of their carbon emissions were 48, 155, and 169 million tons, respectively. In the past 11 years, several provinces experienced an overall increase in carbon emissions. Gansu, Ningxia, Inner Mongolia, Shaanxi, Shanxi, Qinghai and Shandong all displayed upward trends. Inner Mongolia showed the highest average annual growth rate, while Qinghai showed the lowest. Sichuan notably experienced a larger decrease in carbon emissions in 2014, and Henan showed a gradual decrease in emissions after 2014 and an increase in 2020.

3.2. Changes in Carbon Sequestration

During the period of 2010–2020, the overall amount of carbon sequestered in the NPYRs increased gradually, rising from 1166 million tons in 2010 to 1204 million tons in 2020. Sichuan and Inner Mongolia sequestered more carbon, with average values of 294 and 259 million tons, while Ningxia had the lowest average value of only 9 million tons (Figure 4). In the past 11 years, Ningxia and Shandong had the highest growth rates in carbon sequestration, with average annual growth rates of 15% and 5.54%, while Gansu and Inner Mongolia had the lowest growth rates, with average annual growth rates of 0.55% and 1.05%. The carbon sequestration in Sichuan exhibited a slight downward trend, with a reduction of 0.05%.

Compared with the carbon emission data (Figure 4), it is evident that except for Qinghai province, the other eight provinces had higher carbon emissions than carbon sequestration. In general, the disparity between carbon emissions and carbon sequestration in the upper and middle reaches of the Yellow River Basin was less pronounced compared to those in the lower reaches. The average annual carbon sequestration in Qinghai was 38.6 million tons higher than the carbon emissions, suggesting that Qinghai had successfully achieved carbon neutrality. In contrast, Sichuan had a relatively small difference between carbon emission and carbon sequestration. In 2020, the carbon emission in Sichuan was 308 million tons, while the carbon sequestration was 294 million tons. Thus, the pressure of achieving the carbon neutrality target was relatively low. On the other hand, Shandong, Inner Mongolia and Henan had a large difference between carbon emissions and carbon sequestration, with a difference of 630, 560, and 427 million tons, respectively, in 2020. These substantial disparities exacerbate the pressure associated with achieving the goal of carbon neutrality in these provinces.

3.3. OLS Stepwise Regression

In this study, a step-by-step regression of per capita carbon emissions of the NPYRs from 2010 to 2019 was conducted (n = 90), systematically evaluating each influencing factor to select relevant variables for inclusion in the final SRIRPAT model. The regression results are shown in

Table 3. Stepwise regression results.

	(1)	(2)	(3)	(4)
lnA	0.926 **	0.696 *
	(3.089)	(2.502)
(lnA)²	0.093 **	0.070 *
	(3.025)	(2.443)
lnIS2	0.875 ***	0.759 **	0.272 *
	(3.923)	(3.484)	(2.423)
lnU	2.317 ***	2.490 ***	3.000 ***	2.944 ***
	(12.035)	(14.479)	(29.4)	(28.7910
lnEI	1.39 ***	1.331 ***	1.225 ***	1.262 ***
	(21.399)	(22.966)	(22.307)	(23.289)
lnES	0.08	−0.176***	−0.210***	−0.214 ***
	(0.581)	(−6.928)	(−12.461)	(−12.423)
(lnES)²	−0.043
	(−1.887)
lnIS3	0.682 *	0.556 *
	(2.457)	(2.032)
lnP	0.373 ***	0.363***	0.339 ***	0.364 ***
	(10.615)	(10.285)	(10.636)	(11.758)
_cons	−18.276 ***	−18.104 ***	−17.378 ***	−16.322 ***
	(−8.048)	(−7.854)	(−25.216)	(−29.733)
R2	0.971	0.97	0.963	0.961
adj.R2	0.968	0.966	0.961	0.959
F	292.193	318.089	436.178	513.949

Note: t-statistics are in parentheses; *, **, and *** indicate the 5%, 1%, and 0.1% significance levels.

. As this study uses panel data, unit root tests were carried out following parameter estimation to prevent “pseudo-regression” issues. The ce, P, U, ES, and EI were tested by the ADF test and the KPSS test. The results show that the original sequence of the ES was stable. P, EI, and ce were integrated in order 1, while the U was integrated in order 2. The final model is as follows:

$$lnce=1.262\times lnEI+2.944\times lnU-0.214\times lnES+0.364\times lnP-16.322$$

(9)

Equation (8) shows that EI, U and P all had positive effects on the ce, while ES had an opposite effect on the ce. Among them, U had the greatest impact.

3.4. Carbon Peaking Prediction

This paper set the change rate of P, U, EI, and ES under the three scenarios (

Table 4. Scenario setting.

Scenario	Specific Settings
	P	U	EI	ES
High-emission scenario	high	high	high	low
Baseline scenario	medium	medium	medium	medium
Low-carbon energy-saving scenario	low	low	low	high

). Under the baseline scenario, the change rates of P, U, EI, and ES were all set as medium speed based on the historical change patterns, the Outline of the 14th Five-Year Plan (2021–2025) for National Economic and Social Development and Vision 2035 of the People’s Republic of China, the Action Plan for Carbon Dioxide Peaking Before 2030, the Provincial Carbon Peaking Implementation Programs, and the results reported by other researchers [51,52].

Under the high-emission scenario, the change rates of P, U, EI, and ES were set as high, high, high and low speed, indicating, respectively, the high population growth or low population decline, high economic growth and rapid promotion of urbanization, highlighting the economic value of the energy-intensive industry scene, and ignoring the advancement of clean energy and technological upgrades.

Under the low-carbon energy-saving scenario, the change rates of the P, U, EI, and ES were set as low, low, low, and high, indicating, respectively, the slowing down of the rate of population growth, sacrificing a certain level of economic growth and pursuing high-quality urbanization, reducing energy intensity, shifting energy structure, and promoting clean energy. The specific inter-annual variation rates of the four parameters under the three scenarios for each province in the NPYRs can be found in the Supplemental Materials (Supplemental Materials S2).

Based on the STIRPAT model, this paper simulated the carbon emissions of the NPYRs from 2010 to 2060 under the three scenarios (Figure 5). The results show that under the high-emission scenario, the NPYRs will meet their carbon peaking target by 2035 with a peak value of 4372 million tons. This means that a relatively high rate of economic growth, focusing only on population growth, urbanization, and the development of high energy-consuming industries, but neglecting energy structure adjustment, will put a lot of pressure on the NPYRs to reduce carbon emissions, and the goal of peak carbon emissions may not be met on time, in addition to the peak value being considerably higher.

Under the baseline scenario, the NPYRs will meet their peak carbon peaking target by 2030, and the peak value will be 4146 million tons. In other words, if the NPYRs maintain the current developmental rate, essentially realize the targets outlined regarding energy structure and energy intensity in the Peak Carbon Implementation Plan, as well as the targets of urbanization rate and population growth in the Five-Year Plan, the NPYRs will achieve carbon peaking target as scheduled.

Under the low-carbon energy-saving scenario, the NPYRs will achieve carbon peaking target by 2025, with a peak value of 3907 million tons. The low-carbon energy-saving scenario appropriately reduces the growth rate of urbanization and population, further reduces energy intensity, actively adjusts the energy structure, and facilitates the advancement and integration of clean energy, progressively augmenting the proportion of clean energy within the overall energy consumption, which will make the carbon peaking time come earlier and effectively reduce the peak carbon emission. Figure 6 illustrates the carbon peaking time and the peak carbon emission of each province in the NPYRs under the three scenarios.

Under the baseline scenario, Ningxia and Inner Mongolia will achieve the carbon peaking target by 2025, with peak values of 252 and 842 million tons, respectively. The remaining seven provinces will reach peak carbon emissions by 2030. Under the high-emission scenario, Ningxia will achieve its carbon peaking target by 2030, five years later than the baseline scenario, with a peak emission of 2.69 billion tons. The rest of the provinces will realize their carbon peaking targets by 2035; the carbon peaking time of Inner Mongolia will be ten years later than the baseline scenario, with a peak emission of 881 million tons. Under the low-carbon energy-saving scenario, Ningxia had already achieved the goal of carbon peaking in 2020, five years earlier than projected in the baseline scenario, with a peak emission of 240 million tons. Gansu, Shanxi, and Shandong will meet the carbon peaking targets in 2030, consistent with the baseline scenario, with peak emissions of 226, 478, and 946 million tons, respectively. Sichuan, Qinghai, Inner Mongolia, Henan, and Shaanxi will reach peak carbon emissions by 2025. Among them, Inner Mongolia will achieve its carbon peaking target by 2025, as in the baseline scenario, while the other four provinces will achieve carbon peaking targets five years earlier than projected in the baseline scenario. The peak carbon emissions from large to small are as follows: high-emission scenario > baseline scenario > low-carbon energy-saving scenario. Shandong and Inner Mongolia have higher carbon emissions, while Qinghai and Ningxia are the provinces with lower carbon emissions.

3.5. Carbon Neutrality Prediction

The InVEST model was used to predict the carbon sequestration in the NPYRs from 2020 to 2060 under the three scenarios. The results show that the entire study area is unlikely to reach the goal of carbon neutrality by 2060 (Figure 7).

Table 5. The years in which provinces in the NPYRs achieve carbon neutrality.

Carbon Neutrality Time	Low-Carbon Energy-Saving Scenario	Baseline Scenario	High-Emission scenario
Sichuan	2030–2035	2035–2040	2040–2045
Qinghai	2010	2010	2010
Shandong	-	-	-
Shaanxi	2055–2060	-	-
Shanxi	-	-	-
Henan	-	-	-
Inner Mongolia	-	-	-
Gansu	2055–2060	-	-
Ningxia	-	-	-

shows the years in which provinces in the NPYRs achieve carbon neutrality under the three scenarios. The pressure to achieve the carbon neutrality target under the low-carbon energy-saving scenarios is low, with 515 million tons of carbon not being neutralized in 2060. However, under the baseline scenario, this number increases to 767 million tons, indicating higher pressure. The highest pressure is observed under the high-emission scenario, with 1023 million tons of carbon not being neutralized by 2060.

At the provincial scale, under the baseline scenario, only Sichuan and Qinghai can achieve carbon neutrality target before 2060. Qinghai had achieved carbon neutrality in 2010, with 32 million tons of carbon emission and 83 million tons of carbon sequestration, while Sichuan will achieve the carbon neutrality target between 2035 and 2040, with 301 million tons of carbon emission and 311 million tons of carbon sequestration. In 2060, the carbon sequestration in Shandong will be far less than the carbon emission, with a difference of 382 million tons, posing considerable challenges in achieving carbon neutralization. In Shaanxi and Gansu, the differences between carbon sequestration and carbon emissions will be smaller, but still need to be neutralized by 8 and 17 million tons, respectively.

Under the high-emission scenario, Sichuan and Qinghai can achieve the carbon neutrality target. Sichuan will achieve the carbon neutrality target between 2040 and 2045, five years later than the baseline scenario, with 290 million tons of carbon emission and 314 million tons of carbon sequestration.

Under the low-carbon energy-saving scenario, Sichuan, Qinghai, Shaanxi, and Gansu will achieve carbon neutrality on schedule. Sichuan will achieve the carbon neutrality target between 2025 and 2030, about a decade earlier than the baseline scenario, with carbon emissions and carbon sequestration of 304 and 305 million tons, respectively. Gansu and Shaanxi are both anticipated to meet their carbon neutrality goals by 2055 to 2060. In 2060, the predicted carbon sequestration in Gansu Province and Shaanxi Province will be 2 million tons and 11 million tons higher than the carbon emissions, respectively. However, several provinces, including Ningxia, Inner Mongolia, Shanxi, Henan, and Shandong, will still be unable to achieve the carbon neutrality targets by 2060. Shandong, Shanxi, and Henan have significant disparities between carbon emissions and carbon sequestration, requiring neutralization by 314, 176, and 131 million tons, respectively, in 2060.

4. Discussion

4.1. Comparison with Previous Studies

In this study, the carbon peaking results were analyzed in comparison to other relevant references for time to peak and peak value (

Table 6. Comparison of carbon peak time and peak value with the previous studies.

Author	Area	Peaking Time	Peaking Value (MT)	Model
Wang (2024) [53]	Qinghai	2030	66.64	LEAP Model
This study		2030	66.00	STIRPAT
Liao et al. (2023) [17]	Sichuan	2024	334.38	STIRPAT
This study		2030	316.11	STIRPAT
Xie et al. (2022) [54]	Gansu	2029	182.00	IMED|CGE
This study		2030	256.01	STIRPAT
Duan et al. (2023) [55]	Inner mongolia	2030	-	EPS Model
This study		2025	841.74	STIRPAT
Wang (2024) [53]	Ningxia	2030	253.73	STIRPAT
This study		2025	251.72	STIRPAT
Hu et al. (2022) [56]	Shanxi	2025	-	Grey relational analysis
This study		2030	495.55	STIRPAT
Cai et al. (2022) [57]	Shaanxi	2031–2032	-	NIP-GGM(1,1)
This study		2030	325.78	STIRPAT
Wei et al. (2023) [58]	Henan	2030	-	STIRPAT
This study		2030	608.55	STIRPAT
Tian et al. (2022) [59]	Shandong	2028	799.45–822.68	STIRPAT
This study		2030	977.54	STIRPAT

). Based on various studies with different scenario settings, the carbon peak timings and peak values under the baseline scenario were chosen for comparison. In terms of peak attainment values, Ningxia, Sichuan and Qinghai are more consistent with the values in the relevant literature, and the projected values in Shandong and Gansu provinces are higher than those in the relevant studies. In terms of time to peak, the predicted time to peak in Qinghai, Gansu, Henan, Shandong, and Shaanxi is similar to that in related studies, and there is a gap between the predicted time to peak in Inner Mongolia, Sichuan, Shanxi, and Ningxia and that in related studies. The reasons for the gap between the predicted times to peak may be caused by the differences in model selection, variable selection, and rate of change.

4.2. Regional Differences in Carbon Neutrality Degree

The examination of the attainment of carbon neutrality can be made easier by calculating the carbon neutrality degree, which is the difference between carbon emissions and carbon sequestration. Figure 8 shows that under the three scenarios, the carbon neutrality degree is consistently greater than 0. This means that the NPYRs cannot achieve the carbon neutrality target by 2060. Under the baseline scenario and the high-emission scenario, the carbon neutrality degree will first rise, and then gradually decrease after the carbon peaking time, while under the low-carbon and energy-saving scenarios, the carbon neutrality degree will continue to decrease. This indicates that using and developing green energy, adjusting the energy structure, controlling population growth, and developing high-quality urbanization under low-carbon and energy-saving scenarios are effective ways to reduce carbon emissions, while increasing efforts are still required.

Regionally, the middle and lower reaches of the Yellow River basin exhibit a higher degree of carbon neutrality compared to the upper reaches, which show a lower degree. Among them, Shandong, Inner Mongolia and Henan have larger carbon neutrality degrees, while Sichuan and Qinghai have smaller carbon neutrality degrees, which will be negative in 2060, meaning that Sichuan and Qinghai can achieve the carbon neutrality target. Qinghai has a small permanent population, with clean energy utilization ranking second only to Shanxi among the nine provinces. Its carbon emissions are the lowest among the nine provinces, amounting to less than 10% of Shandong’s in 2019. As for Sichuan, by 2020, the province’s forest area accounted for 40.26% and arable land for 10%, boasting a good ecological environment and having the highest carbon sequestration capacity among all nine provinces, far exceeding its carbon emissions. Shandong, Inner Mongolia and Henan are located in the middle and lower reaches of the Yellow River Basin. Shandong has a large population, and the clean energy portion in its energy structure is much smaller than the national average. Inner Mongolia possesses abundant coal resources, and its heavy reliance on coal energy leads to high energy intensity and a low proportion of clean energy. Henan has a high population density and a low share of clean energy in its energy mix, which contributes to high carbon emissions. Although the carbon sequestration capacity of Henan and Inner Mongolia is relatively high among the NPYRs, there is still a big gap between them and their carbon emissions.

4.3. Realization Pathways

Based on the aforementioned findings, the following recommendations are proposed to effectively realize the carbon peaking target by 2030 and achieve the carbon neutrality target by 2060.

(1) Promoting new urbanization construction and pursuing a path of high-quality urban development.

The NPYRs must rationally modify the scale of cities, formulate the construction scale scientifically, coordinate and optimize city structure, encourage the growth of new urbanization, and assist in the creation of urban and rural planning, building, and management mechanisms that give priority to low-carbon and green initiatives. Qinghai, Shaanxi, Shanxi, Ningxia and Inner Mongolia have the high urbanization rates, which were 60%, 63%, 63%, 64%, and 68%, respectively, in 2020, and the urbanization growth rate was too fast, with an average annual growth rate of 2%. Therefore, it is necessary to concentrate on controlling the speed of urbanization, developing urban clusters, accelerating the construction of livable, low-carbon communities and villages, and promoting new urbanization construction.

(2) Optimizing industrial structure, encouraging the growth of sustainable energy, controlling carbon emissions, and pursuing a path of emission reduction and low-carbon development.

At present, the biggest flaws of the NPYRs are that the industry mainly relies on energy and heavy industry, and the industrial structure is single and heavy [60,61]. Ningxia, Inner Mongolia and Qinghai have high energy intensity, and in 2020, their energy intensity was 218, 157, and 137 tons of standard coal per ten thousand yuan, respectively. As a result, there is more pressure to modify their energy structure. Therefore, the NPYRs should actively optimize industrial structures, accelerate the withdrawal of underdeveloped production capacity, reduce carbon emissions, promote the development and utilization of clean energy, and develop a sustainable, low-emission, and high-efficiency energy system.

Ningxia should be in favor of the optimization of its energy system and the progressive utilization of resources, encourage the construction of distributed energy projects in development zones, proactively establish low-carbon industrial parks, and advocate for mandatory cleaner production audits and transformations within key industries. Additionally, Ningxia can rely on desert, Gobi, desert and other areas to build photovoltaic bases, develop solar thermal energy, and actively promote the application of hydrogen energy.

Inner Mongolia ought to aggressively encourage the modernization of its current coal-fired power plants to use less energy and undertake exemplar projects demonstrating pivotal technologies for energy conservation and carbon reduction. Meanwhile, Inner Mongolia can also promote the construction mode of “new energy + energy storage”, accelerate the construction and development of new power systems, strategically plan the construction of pumped storage power plants, and expand the multi-scenario application of energy storage.

Benefiting from the rich light resources, Qinghai has the potential to attract a multitude of new energy and new material enterprises, thereby establishing a comprehensive photovoltaic manufacturing industry chain. This has resulted in an accelerated advancement in the clean energy equipment manufacturing sector.

Moreover, Shaanxi can take advantage of its geographical location to comprehensively promote large-scale harnessing of wind and photovoltaic energy, as well as facilitate the development of existing hydropower projects. Furthermore, Shanxi possesses the potential to establish a national natural gas base, fully exploit its resources and industries, and systematically develop renewables like geothermal and methanol.

(3) Emphasizing ecological environmental protection and pursuing a path of prioritizing ecological development.

The NPYRs play a crucial role in maintaining China’s ecological security due to their strategic location and unique ecosystem [62,63]. To decrease carbon emissions and enhance carbon sequestration capacity, the NPYRs should promote the integrated protection and restoration of the ecosystems and improve ecosystem stability so as to enhance ecosystems’ capacity of carbon sequestration. In conjunction with the formulation and implementation of land use planning, a framework for land development and protection conducive to achieving carbon neutrality should be established. The nine provinces also need to strictly abide by the ecological red line, strictly prohibit any encroachment upon ecological spaces and stabilize the role of carbon sequestration in existing ecosystems. Governments also need to implement ecological protection and restoration projects at different levels, tailored to the specific ecological conditions of each region, vigorously advocate fir initiatives such as returning farmland to lakes, forest and grassland projects, enhancing the conservation of forest resources, strengthening the restoration and management of degraded land in the middle and upper reaches of the Yellow River Basin, and actively carrying out the comprehensive soil and water erosion management. Simultaneously, the natural resource survey and testing system should be utilized to provide feedback on the changes in the ecosystems so as to better protect and repair the ecological environment.

4.4. Limitations of This Study and Future Work

In this study, the model used to predict carbon peaking is constructed using panel data from the NPYRs, and the change settings of the parameters in the model only rely on the existing policies and short-term changes, which could potentially result in errors in the outcomes. The follow-up work can consider introducing the time series analysis method to accurately predict the future changes in the driving factors and improve the reasonableness of the scenarios.

5. Conclusions

This paper used the STIRPAT model and the InVEST model to predict the time of carbon peaking and carbon neutrality and explored the realization pathways of carbon peaking and carbon neutrality. The results show the following:

(1)

The carbon emissions in the NPYRs exhibited an overall increasing trend in 2010–2020. The NPYRs will achieve carbon peaking target by 2025 under the low-carbon energy-saving scenarios, by 2030 under the baseline scenario, and by 2035 under the high-emission scenario.

(2)

Despite efforts, the entire study region is projected to fall short of achieving the carbon neutrality target by 2060 under all three scenarios. However, Qinghai notably achieved carbon neutrality as early as 2010. Sichuan demonstrates potential to achieve carbon neutrality under all three scenarios, while Shaanxi and Gansu could do so under the low-carbon energy-saving scenario. Moreover, Shandong, Inner Mongolia and Henan face greater challenges in meeting the carbon neutrality target.

(3)

In order to expeditiously meet the targets for carbon peaking and carbon neutrality, it is recommended that the governments promote new urbanization construction, optimize industrial structure, foster the advancement of clean energy, control carbon emissions, and protect the environment and ecosystems.