Analysis of spatial and temporal evolution and driving factors of carbon emission in Shandong Province: based on the perspective of land use

Abstract

Land use/cover change is the second major contributor to carbon emissions, following energy emissions. Studying provincial land-use carbon emissions is crucial for achieving the “double carbon” goal. This study selects 16 prefecture-level cities in Shandong Province as the research object. It analyzes the spatial and temporal distribution pattern of carbon emissions in Shandong Province based on land-use data and energy consumption. In terms of net carbon emissions, this study utilizes the standard deviation ellipse and kernel density estimation to analyze net carbon emissions change from the municipal and regional perspectives. In terms of carbon ecological carrying capacity, not only the carbon ecological carrying capacity of forest and grassland was considered, but also the carbon ecological carrying capacity of crops in Shandong Province, which is a large grain province. Using the geographic detector to explore the drivers. Research findings indicate that carbon sources and sinks show a clear spatial and temporal distribution pattern, with the center of gravity of net carbon emissions extending to the northeast. Areas with high carbon ecological carrying capacity have high forest coverage, grassland coverage, and crop yields. Regarding driving factors, the urbanization rate, economic aggregate, and technological progress demonstrate significant explanatory power through single and interaction tests, suggesting that these factors are critical drivers of land-use carbon emissions within Shandong Province. Based on the spatiotemporal pattern analysis of land-use carbon emissions in Shandong Province, each city's growth rate and spatial distribution characteristics can be clarified, providing a scientific basis for the local government to formulate regional and differentiated emission-reduction policies. In addition, by exploring the driving factors of land-use carbon emissions in Shandong Province, the influence level of factors on carbon emissions can be revealed to provide references for formulating regional sustainable development strategies.

Introduction

Global warming, melting glaciers, rising sea levels, and the emergence of haze all reveal that climate change, triggered by the greenhouse effect, has a profound impact on human life [48]. According to the Sixth Assessment Report issued by the International Panel on Climate Change (IPCC), global temperatures are now 1.1 °C higher than before industrialization [27], which has led to an enhanced greenhouse effect and an intensification of the global warming problem. A critical factor in climate change is human activities' production of greenhouse gases, especially carbon dioxide [56]. Land use change is the second largest source of greenhouse gas emissions after fossil fuel combustion, accounting for one-third of global emissions since 1850 [2]. Because of anthropogenic land use and cover change, there has been a significant shift in the carbon exchange between the atmosphere and the surface, significantly impacting the response of the land surface to climate [20]. As an important carrier of economic and social development and production activities, a large part of carbon emissions from human activities is closely related to land use. The study of carbon emissions from land use significantly mitigates climate warming and protects the ecological environment.

China, the world's largest developing country [23], has experienced rapid economic growth since implementing the reform and opening-up policy. Nonetheless, the toll of economic expansion has been substantial energy consumption, leading to a notable surge in carbon emissions. Concurrently, urbanization has accelerated, with a marked increase in construction land, thereby hastening the rise of carbon emissions. China overtook the US as the world's biggest carbon dioxide emitter [28]. As China's economy and carbon emissions are projected to grow [32, 39], China has introduced a 'dual carbon target' to promote green and low-carbon transformation and actively implement carbon reduction initiatives. China has a broad geographic scope, and there are noticeable differences among provinces regarding economic development, industrial layout, energy consumption, and land use. A comprehensive analysis of carbon emissions resulting from provincial land use can offer a more precise insight into the carbon emissions of each region and provide robust scientific backing for formulating targeted emission reduction strategies. Moreover, provincial administrative units represent the highest level of local administrative institutions under direct government management, possessing substantial regulatory capacity over economic activities and land use within their respective regions. Analyzing carbon emissions at the provincial level can guide local governments to adopt emission reduction measures and effectively implement relevant policies.

In this context, researchers have conducted many studies on carbon emissions from land use, mainly focusing on the measurement, spatial and temporal distribution characteristics of carbon emissions, and analyzing influencing factors. For the measurement of land-use carbon emissions, the researchers mainly used the sample point inventory method [4, 9], bookkeeping model [5, 54], and carbon emission factor method [19, 38]. In the study on spatial and temporal patterns of carbon emissions from land use, the researchers measured the carbon emissions from different land-use types and analyzed the spatial and temporal evolution using carbon sources and sinks as criteria. The study of land-use carbon emissions covers the global [40, 44] and national levels [8, 50]. However, in China, research on land-use carbon emissions mainly focuses on the national [36], provincial [13, 21], and city cluster scales [29, 65]. To analyze the drivers of land-use carbon emissions, the researchers used a variety of models for their assessment, such as the logarithmic mean zoning index (LMDI) [59], the geographically weighted regression model (GWR) [24], the extended Kaya identity [34], and the stochastic effects of population, affluence, and technology regression model (STIRPAT) [3]. The driving factors of land-use carbon emissions mainly include population size, economic development level, urbanization rate, and industrial structure. Among them, population and economic growth and the increase in urbanization rate have contributed to the growth of land-use carbon emissions [47, 69]. The optimization and adjustment of industrial structure, especially the decline of the proportion of the secondary industry, has an inhibiting effect on the growth of land-use carbon emissions [40, 62].

Most studies on carbon ecological carrying capacity are combined with the study of carbon footprint to conduct a dynamic analysis of a specific area. Taking Yunnan Province as the research object, the carbon footprint model was constructed to evaluate the carbon security status of Yunnan Province [15]. Taking Shanxi Province as a case study, the spatial and temporal evolution law of carbon footprint and carrying capacity was discussed, and the carbon ecological pressure index was used to evaluate and predict the carbon ecological security in this region [64]. Taking Shandong Province as the research object, the new concepts and calculation models of carbon-carrying capacity and net carbon footprint are introduced, and the dynamic analysis of carbon footprint and carbon-carrying capacity of Shandong Province is carried out by using this optimized carbon footprint method [61].

In summary, domestic and foreign scholars have analyzed land-use carbon emissions at different geographical scales from different perspectives, but there are still some limitations. Firstly, research into the spatial and temporal distribution of carbon emissions due to land use has predominantly concentrated on national and provincial scales and urban clusters, such as the Yellow River Delta, with earlier and fewer studies on individual provinces. The spatiotemporal analysis of carbon emissions from land use at the provincial and municipal levels is limited to comparing year-by-year results, which does not reveal the intrinsic pattern of such changes. Due to the differences in economy, society, and population size, carbon emissions also show different characteristics at different spatial scales. The dynamic characteristics of urban carbon emissions' spatial and temporal patterns can be more clearly explored from the provincial and municipal spatial scales. Secondly, the traditional approach to studying carbon emission drivers is mainly based on panel data, which cannot fully elucidate the drivers behind spatial differences and lacks consideration of the spatial correlation of the drivers. In contrast, as an efficient means of spatial analysis, the geographical detector model provides a more powerful tool to identify and quantify various factors' different degrees of influence on the spatial differences of land-use carbon emissions and their interactions [33]. Finally, previous studies on carbon ecological carrying capacity mainly focused on calculating regional carbon-carrying capacity and analyzing spatial and temporal dynamic characteristics. The pressure index was established to classify the regional carbon ecological security, and the future security was predicted using the relevant model, and the corresponding measures were put forward. Nonetheless, this research merges the analysis of carbon's ecological carrying capacity with research on carbon sources and sinks. After assessing the region's carbon ecological carrying capacity, its potential for reducing carbon emissions was examined from a carbon ecological carrying capacity standpoint.

In recent years, Shandong Province has made world-renowned achievements in economic development. However, Shandong's energy consumption and carbon emissions have been increasing, causing severe environmental hazards in the province. As Shandong Province is one of the high-carbon emitting provinces in the world major carbon emitters, the study of land-use carbon emissions is not only of great significance to the carbon emission reduction in Shandong Province and China, but also can provide a reference value for the world's carbon emission reduction. Previous studies on carbon emissions in Shandong Province have primarily focused on assessing [71], predicting peaks [72], and driving factor analysis of carbon emissions [10, 58]. The research findings of the scholars above serve as a foundation for the analysis of spatiotemporal evolution and driving factors in this paper. However, there are significant disparities in population conditions, economic status, and natural resources among cities in Shandong Province. Given that Shandong Province is a typical area with high carbon emissions, researching city-level land-use carbon emissions holds great significance.

Thus, to explore the spatial and temporal characteristics of carbon emissions and their drivers in Shandong Province, this study takes 16 prefecture-level cities as the research object, measures the carbon emissions of each land-use type, and analyzes the spatial and temporal patterns of carbon emissions in Shandong Province. Then, based on the carrying capacity of forest, grassland, and crops, analyze the effect of carbon-carrying capacity on the environment. Further, the migration trajectory of the center of gravity and the distribution trend of carbon emissions in each region of Shandong Province will be analyzed using the standard deviation ellipse (SDE) and kernel density estimation (KDE). Finally, to identify the driving factors affecting carbon emissions by utilizing geo-detectors. As a robust economic development province in China, the study results will help cities in Shandong Province implement emission reduction programs suitable for their local environments and provide a unique theoretical framework for long-term low-carbon growth in Shandong Province. It will also provide reference measures for other provinces and cities in China, promoting the realization of China's “dual-carbon” goal.

Materials and methods

Study area

Shandong Province (34° 22.9′–38° 24.01′ N, 114° 47.5′–122° 42.3′ E), also known as “Lu”, is a provincial-level administrative region with Jinan as its capital and 16 prefectural-level cities under its jurisdiction. It is located on the eastern coast of China, bordering the Bohai Sea, the Yellow Sea, and the provinces of Hebei, Henan, Anhui, and Jiangsu from north to south (Fig. 1).

Fig. 1

Overview of the study area

Shandong is one of the leading economic provinces in China, consistently holding the third position in national economic rankings since 2007. As of the end of 2022, the resident population of Shandong Province has an urbanization rate of 64.54%. Shandong Province's GDP reach 874,351 billion yuan by 2022. As urbanization speeds up, there is a significant shift in land resources, leading to a growing conflict between economic growth, social progress, and environmental sustainability. This triggers various environmental issues, including the deterioration of land quality and a decline in biodiversity. Thus, it is of great significance to explore the space for carbon emission reduction in the process of land use in order to realize the goal of “double carbon” by practicing green and low-carbon development.

Data sources and processing

Data sources

Given the data availability and statistical caliber issue, land data for Shandong Province, featuring a 30-m resolution, were sourced from the Chinese Academy of Sciences Cloud Platform for Resource and Environmental Data (https://www.resdc.cn) for this research. Additionally, calculating carbon emissions from construction land relies on energy consumption data obtained from the Statistical Yearbook of Local Municipalities and the Statistical Yearbook of Shandong Province (2010–2020). Socio-economic data for crop yields used to measure ecological carrying capacity and analyze carbon emission drivers were obtained from the statistical yearbooks of Shandong Province and prefecture-level cities.

Construction of driving factor index system

Population size, economic development, urbanization rate, and industrial structure influence carbon emissions. This paper also considers land-use structure as an independent variable to analyze the factors affecting land-use carbon emissions comprehensively. Based on existing research findings and considering the regional context and data availability [41, 70], this study selects five influencing factors: population, economy, land, technology, and science. The specific drivers selected are shown in Table 1.

Table 1 Index system of driving factors of land-use carbon emissions

Population: As land users, all human productive activities impact carbon emissions from land. First, the demand for energy, such as oil and electricity, increases as the population grows. Second, as the population increases, construction activities increase accordingly, further triggering a shift in land-use patterns, thus indirectly affecting the carbon cycle. Therefore, population size is selected in this study to analyze its effects on carbon emissions.

Urbanization rate: As urbanization continues to accelerate, people are flocking to cities, leading to a continuous growth in the size of cities, while the demand for land and energy is also rising rapidly. In addition, in the process of urbanization, energy consumption has also risen significantly to better meet the needs of urban construction and the daily life of residents. Considering this, this study chooses urbanization rate as a driver to explore its specific impact on carbon emissions.

Level of economic development and economic aggregate: The Gross Regional Product represents the aggregate economic operations in a given area. With the expansion of economic activities, energy use has a corresponding rise, leading to higher carbon emissions. Concurrently, as the per capita GDP incrementally escalates, there is a corresponding rise in the buying power of individuals. Alterations in consumption habits, like the increased need for energy-demanding items such as cars and home appliances, will additionally lead to elevated energy usage and carbon emissions. Consequently, this study identifies regional and per capita GDP as influencers in examining their impact on carbon emissions.

Industrial structure: There are significant differences in carbon emissions by industry type. Secondary industries tend to consume much energy and produce high emissions. However, secondary industry occupies an essential position in China's economic system. Thus, this paper selects the proportion of the secondary industry in the GDP to gauge its impact on carbon emissions.

Technological progress: As economic growth attains a specific level, the conventional catalysts for economic expansion dwindle. More rigorous environmental regulations exacerbate the situation, forcing companies to overhaul their economic growth and enhance production techniques. As a result, advancements in technology significantly reduce carbon emissions. Consequently, this research opts to analyze the aggregate of patents issued in China to evaluate the effect of technological advancements on carbon emissions.

Land-use structure: Given its role as the main contributor to carbon emissions in land utilization, expanding construction land area will lead to more construction activities, leading to the rise of energy consumption and carbon emissions. In this paper, the proportion of construction land in the total land area is selected as the measurement index of land-use structure, which can intuitively observe the change of land-use type and intensity to understand its specific impact on carbon emissions.

R&D investment (Research and Experimental Development investment): Breakthroughs in environmental protection and clean energy technologies have the potential to significantly mitigate carbon emissions during the production process. Furthermore, increased government investment in science and technology can provide guidance and support for enterprises, scientific research institutions, and other entities to develop and implement low-carbon technologies. In light of this, this study incorporates the ratio of science and technology expenditure to public budget expenditure as one of the factors driving land-use carbon emissions.

Methods

This paper takes Shandong Province as the research object and analyzes the spatial and temporal evolution of carbon sources, sinks, and net carbon emissions. The center of gravity migration and distribution trends of net carbon emissions in Shandong Province were analyzed more profoundly using standard deviation ellipses and kernel density estimation. The carbon ecological carrying capacity was estimated based on forest, grassland, and crops, and the carbon emission reduction potential of Shandong Province was analyzed from the perspective of the carbon ecological carrying capacity. Finally, the drivers of land-use carbon emissions were analyzed. The detailed analysis flow is shown in Fig. 2.

Fig. 2

Flowchart and research method

Carbon emission measurement methodologies

Regional land-use change significantly influences the carbon cycle in terrestrial ecosystems and is a crucial element in the rise of regional carbon emissions [31]. Estimates of carbon emissions from land use can be categorized in two ways. Direct land-use carbon emissions are related to land-use changes [51]. Indirect land-use carbon emissions refer to certain factors associated with anthropogenic emissions from construction land, such as energy consumption [1].

Since cultivated land is both a carbon source and a carbon sink, the coefficient of greenhouse gases and the coefficient of CO₂ absorption by crops in the growth process must be taken into account at the same time when calculating the carbon emission of cultivated land. The difference between the two is cultivated land's net carbon emission coefficient. Based on the previous research results, the carbon emission coefficient of crops is 0.0504 kg/(m²·a) [6, 7], and the carbon sequestration coefficient of crops is 0.00074 kg/(m²·a) [18], which leads to the net carbon emission coefficient of cultivated land number is 0.0497 kg/(m²·a). The carbon sequestration coefficient of − 0.581t/(hm²·a) for forest [14, 60]. According to researchers, the average carbon sink coefficients of China's water area and the Great Lakes are 0.0248 kg/(m²·a) and 0.0253 kg/(m²·a) [11, 26], respectively. In this paper, the average of the two is used to calculate the carbon sequestration of the water area. The carbon sequestration coefficient reaches 0.253t/(hm² a). The carbon absorption coefficients for the unutilized land and grassland used in this article are shown in Table 2, according to the findings of earlier researchers [14, 43, 52, 66]. The specific formulas used are as follows:

Table 2 Carbon emission/absorption factor for land-use type (t/(hm² a))

$${C}_{i}={A}_{i}\times {L}_{i}.$$

(1)

In the given formula, ${C}_{i}$ represents the carbon emission (t) produced by the ith land type. When ${C}_{i}$ is positive, it indicates carbon emission; when it is negative, it represents carbon sequestration. The term ${A}_{i}$ denotes the size of the ith type of land, and ${L}_{i}$ symbolizes the carbon emission (or sequestration) coefficient.

The carbon emission of energy consumption of construction land serves as an indirect representation of the carbon emission of construction land. Carbon emissions are quantified by end-use energy consumption. In this paper, eight types of energy sources are chosen as representatives, and their standard coal coefficient and carbon emission coefficient are presented in Table 3. This formula can be applied to determine the carbon emissions emanating from construction land:

Table 3 Standard coal discount factor for CO₂ emission factors for energy sources

$${B}_{i}={\sum }_{i=1}^{n}{K}_{i}\times {E}_{i}\times {F}_{i}.$$

(2)

In the given formula, ${B}_{i}$ represents the energy carbon emission; ${K}_{i}$ denotes the consumption of the ith energy source; ${E}_{i}$ represents the coefficient of the ith standard coal; and ${F}_{i}$ represents the carbon emission coefficient of the ith energy source.

The total regional carbon emissions may be calculated by including both direct and indirect carbon emissions using the following formula:

$$C={A}_{i}+{B}_{i}.$$

(3)

Standard deviation ellipse and kernel density estimation

Standard deviation ellipse is a spatial statistical method used to reveal research objects' spatial distribution and multi-directional characteristics [12]. We employ the standard deviation ellipse analysis method to depict the spatial intensity and overall distribution of net carbon emissions from land use in Shandong Province. The center of gravity of the standard deviation ellipse represents the average distribution center of net carbon emissions. The center of gravity coordinate formula was employed to ascertain the study elements' center of gravity, while its movement direction signifies the spatial trajectory of distribution change. The azimuth angle indicates the primary trend direction of net carbon emissions distribution. Furthermore, the long and short semi-axes reflect the degree of dispersion and concentration of net carbon emissions in their spatial distribution. These formulas are as follows [16]:

• (a) Center of gravity coordinates:

  $$X ={\sum }_{i=1}^{n}{T}_{i}{X}_{i}/{\sum }_{i=1}^{n}{T}_{i}, Y={\sum }_{i=1}^{n}{T}_{i}{Y}_{i}/{\sum }_{i=1}^{n}{T}_{i},$$

  (4)

• (b) Azimuth angle:

  $${\theta }_{\left(m+1\right)-m}=\frac{n\pi }{2}+\text{arctg}\frac{\left({y}_{m+1-}{y}_{m}\right)}{\left({x}_{m+1-}{x}_{m}\right)},$$

  (5)

• (c) The center of gravity’s spatial shift distance:

  $$D=K\times \sqrt{\left(\sqrt{{\left({y}_{m+1}-{y}_{m}\right)}^{2}+{\left({x}_{m+1}-{x}_{m}\right)}^{2}}\right),}$$

  (6)

  where $X$ and $Y$ represent the coordinates of the center of gravity, $\theta$ represents the azimuthal angle, and $D$ represents the spatial displacement distance of the center of gravity.

Kernel density analysis is a non-parametric estimation method that can describe the temporal distribution characteristics of carbon emission intensity with a continuous kernel density curve. The horizontal position of the curve can represent the level of carbon emission intensity, and the height and width of the wave crest can reflect the concentration degree of carbon emission intensity within the interval. In addition, the greater the number of wave peaks, the higher the polarization of the sample data, and the degree of curve extension can describe the degree of difference within the region. The longitudinal comparison of the nuclear density curves can be used to analyze the dynamic evolution of carbon emission intensity distribution in this region. The kernel density for the density function $f\left(c\right)$ related to carbon emissions is shown in equation:

$$f\left(c\right)=\frac{1}{Nh}\sum_{i=1}^{i}k\left(\frac{{c}_{i}-c}{h}\right),$$

(7)

$$K\left(c\right) =\frac{1}{\sqrt{2\pi }}\text{exp}\left(-\frac{{c}^{2}}{2}\right),$$

(8)

where $f (C)$ represents the kernel function, $h$ signifies the bandwidth. $N$ indicates the count of observed, and $K [ ({C}_{i}-C)/h)]$ denotes the kernel function, serving as a type of weighting or smoothing transformation function. ${C}_{i}$ represents the count of independent and identically distributed observations, and $C$ denotes the average.

Carbon ecological carrying capacity

In recent years, the idea of carbon-carrying capacity has garnered increasing interest due to the ongoing advancement of a low-carbon economy. Existing researchers mainly define carbon-carrying capacity in terms of three dimensions: specific areas, different vegetation, and the amount of CO₂ fixed [35, 68], but there has yet to be a universally applicable standard. Therefore, this study defines carbon-carrying capacity as the sum of carbon dioxide fixed by forest, grassland, and crops in a specific period and region [61], including the seven main crops. Through the analysis of carbon ecological carrying capacity, the carbon emission reduction potential of Shandong Province was discussed in three aspects: forest, grassland, and crops.

Forest and grassland's carbon ecological carrying capacity can be measured by multiplying their area by their average net ecological productivity, NEP, and then by the molecular weight ratio of carbon dioxide to carbon. The formulas are as follows:

$${C}_{\text{e}}={T}_{\text{e}}\times {\text{NEP}}_{\text{e}}\times \frac{44}{12},$$

(9)

$${C}_{\text{f}}={T}_{\text{f}}\times {\text{NEP}}_{\text{f}}\times \frac{44}{12}.$$

(10)

Within the equation, ${C}_{\text{e}}$ and ${C}_{\text{f}}$ are the carbon ecological carrying capacity of forest and grassland, respectively. ${T}_{\text{e}}$ and ${T}_{\text{f}}$ are their area, and ${\text{NEP}}_{\text{e}}$ and ${\text{NEP}}_{\text{f}}$ are their average net ecological productivity [63], both of which have a CO₂ to carbon molecular weight ratio of $44/12$.

A crop's carrying capacity results from five key factors. Table 4 displays the estimated carbon storage parameters for crops [67]. This research chose seven crops to determine their carbon-bearing capacity using a specific formula:

Table 4 Estimation parameters of vegetation carbon storage for different crops

$${C}_{\text{g}}=\sum_{i=1}^{n}{C}_{i}\times (1-{M}_{i})\times \left(1+{N}_{i}\right)\times {P}_{i}\times \frac{44}{12}.$$

(11)

Within the equation, ${C}_{\text{g}}$ denotes the crop's carbon capacity, ${C}_{i}$ r represents the carbon content rate per biomass unit, ${M}_{i}$ is the water coefficient for crop type I, ${N}_{i}$ is the root–crown ratio coefficient for crop type, and ${P}_{i}$ is the crop type I is economic yield.

The geographical detector model

The Geographical Detector Model is a model that detects the combined influence of multiple factors on the variability of geographic phenomena [57]. Factor probing refers to the magnitude of the explanatory power of the probing factor to the observed variable. In order to analyze the driving influence of various factors on land-use carbon emissions in Shandong Province, this study employs single-factor detection and interactive detection to conduct a driving analysis of land-use carbon emissions in the region. The formula for calculation reads:

$$q=1-\frac{{\sum }_{h=1}^{L}{N}_{h}{\sigma }_{h}^{2}}{{{N}_{\sigma }}^{2}}.$$

(12)

In this equation: $q$ represents the magnitude of influence, where higher figures signify increased impact; $L$ denotes the division of the independent variable; ${N}_{h}$ signifies the sample size for class $h$; ${\sigma }_{h}^{2}$ indicates the variance of class h's independent variable; $N$ denotes the sample size for each city; ${\sigma }^{2}$ is the variance.

Interaction detection recognizes the interplay among various influencing elements, that is, evaluating if the explanatory strength is enhanced or diminished when these factors collaborate (Table 5). The approach involves initially determining the q-values for ${x}_{1}$ and ${x}_{2}$ on $y$, namely $q({x}_{1})$ and $q({x}_{2})$, respectively, and then computing these q-values during their interaction, namely $q({x}_{1}\cap {x}_{2})$, followed by a comparison.

Table 5 Basis and type of interaction judgment

Results

Land use changes in Shandong Province

Figure 3 shows the land use map of Shandong Province during the study period. The size of each land use category was derived, as depicted in Table 6. Shandong Province's land utilization predominantly centers on cultivated and construction land, a smaller share of other land categories. Generally, there was a decline in cultivated land, grassland, and unutilized land, evidenced by a reduction of 173,016.63 hm² in cultivated land, 3348.45 hm² in grassland, and 7108.38 hm² in unutilized land. Construction land and forest increased by 32,882.85 hm² and 453.87 hm², respectively. Notably, from 2015 to 2018, there was a notable expansion of the water area by 190,142.46 hm², with the most significant growth at 22.96%. From 2010 to 2015, Shandong Province's construction land area saw a notable surge, attributed to its swift economic growth, evidenced by a 9.42% average GDP increase and rapid urbanization. After 2015, the local government recognized that the uncontrolled growth of construction land area had caused serious ecological problems. The “Management Measures for the Pilot Increase and Reduction of Urban and Rural Construction Land in Shandong Province” to revitalize cultivated land and control the blind growth of construction land through land reclamation and reuse adjustments is consistent with the actual statistics on the increase in the size of cultivated land and the decrease in construction land during this period.

Fig. 3

Land use map of Shandong Province for 2010, 2015, 2018 and 2020

Table 6 Variations in the size of different types of land utilization (hm²)

Changes in carbon emissions in Shandong

Carbon emissions for all land-use types were calculated based on land use and energy consumption data, as shown in Table 7. It reveals that construction land contributes the most carbon emissions in the carbon source lands, accounting for ~ 98%, while cultivated land's carbon emissions are the least, hovering around 2%. Of the carbon-absorbed land, a significant portion, ~  65–70%, is absorbed by forest and 27–32% by water area, in contrast to grassland and unutilized land, which make up a minor fraction. Between 2010 and 2020, the carbon emissions from construction land exceeded the original amount by 11,673.55 × 104 tons, nearly 1.4 times, leading to a notable increase in carbon emissions in Shandong Province. As a result, managing the growth of construction land and enlarging forest and water areas can diminish carbon emissions and enhance carbon absorption, thereby effectively lowering overall carbon output in Shandong Province. The remaining land types in the city have less variation in carbon emissions.

Table 7 Carbon emissions by in Shandong Province in 2010, 2015, 2018 and 2020 (tons)

Given the unique circumstances of each city, the carbon emissions differ among cities (Fig. 4). From 2010 to 2015, every city in Shandong Province, except Heze, Taian, Zaozhuang, and Zibo City, experienced increased carbon emissions. From 2015 to 2020, there was a deceleration in the increase of carbon emissions in various cities. Among them, Binzhou and Dongying City have the most significant changes in carbon emissions, especially in the period of 2010–2015, from 1200.83 × 10⁴ tons and 1513.92 × 10⁴ tons to 4724.25 × 10⁴ tons and 2955.65 × 10⁴ tons, respectively. This is because, as a traditional industrial city, Binzhou City's industrial structure is dominated by heavy industry. Developing these industries requires a large amount of energy consumption, resulting in a rapid increase in carbon emissions in Binzhou City. Statistics show that Binzhou City's electricity consumption ranks third in China. Dongying is rich in oil resources, relying on the rich oil resources to form a petrochemical industry and other pillar industries, resulting in a rapid increase in carbon emissions. In addition, Jining's carbon emissions have shown a downward trend, which may be due to Jining's active adjustment of industrial structure and optimization of energy consumption structure. As local governments realized the importance of low-carbon green development and eliminated backward production capacity, the growth trend of carbon emissions in each city slowed down after 2015. Yantai, Linyi, and Jinan ranked among the top three cities in Shandong Province regarding carbon sequestration by urban land. Compared with carbon emissions, changes in land carbon sequestration were relatively stable during the study period.

Fig. 4

Carbon emissions by land-use type in cities of Shandong Province (tons)

Spatial and temporal distribution of carbon emissions from land use

Analysis of spatial and temporal patterns of carbon sources

This paper utilized cross-sectional data from four periods in Shandong Province to calculate carbon sources in each city, using ArcGIS software and the natural discontinuity method to classify carbon sources into five levels: level 1 is the lowest, and level 5 is the highest, as shown in Fig. 5.

Fig. 5

Carbon source pattern in Shandong Province (× 10⁴ tons)

In particular, Binzhou and Jining City are the primary regions for carbon sources. Heze, Weihai, and Rizhao City exhibit comparatively mild effects from carbon sources. The pattern of carbon source distribution in Shandong Province exhibits significant variations over time and across different regions. In 2010, areas with high carbon sources predominantly resided in the southwestern and central parts. In 2015, regions with high carbon sources predominantly resided in the northwestern part of Shandong province and the Jiao Lai area. During 2018 and 2020, areas with high carbon sources started expanding into the central region, with a rise in cities with high carbon sources. During the study, Binzhou City witnessed the most notable shift in carbon source quantities, transitioning from a level 2 carbon source zone in 2010 to a level 5 carbon source zone in 2015, maintaining a high level. Heze and Weihai City sustain a steadier condition and are situated in regions with low carbon sources.

Analysis of spatial and temporal patterns of carbon sinks

Based on the natural discontinuity method, this paper classifies carbon sinks into five levels, with level 1 being the most robust and level 5 being the weakest (Fig. 6).

Fig. 6

Carbon sink patterns in Shandong Province (× 10⁴ tons)

Both high and low-carbon sink areas remained relatively stable during the study. Yantai City has always been a first-class carbon sink city, with carbon sinks between 11.66 × 10⁴ tons and 12.20 × 10⁴ tons, contributing to roughly 15% of Shandong Province's total carbon absorption. Liaocheng City records the least carbon sinks. During 2010 and 2015, the primary distribution of level 5 carbon sinks occurred in Dezhou, Liaocheng, and Heze City, making up a mere 1.81% to 1.85% of Shandong Province's total carbon sinks; between 2018 and 2020, these carbon sinks occurred in Dezhou, Liaocheng, Heze, and Zaozhuang City, representing roughly 4.12% to 4.71% of Shandong Province's total carbon sinks.

Analysis of spatial and temporal patterns and trajectory analysis of net carbon emissions

This study classifies net carbon emissions into five levels based on the natural discontinuity approach, where level 1 represents the minimum and level 5 the maximum (Fig. 7).

Fig. 7

Net carbon emission patterns in Shandong Province (× 10⁴ tons)

The disparities in net carbon emissions across regions are apparent, showing varying levels of variation in each city, with Binzhou City experiencing the most notable increase in net carbon emissions. In 2010, the southwestern part of Shandong Province was the primary area for high net carbon emissions. From 2015 to 2020, except for Jining City, which consistently remained in the level 5 net carbon emission zone, the central and JiaoLai regions of Shandong Province predominantly had high net carbon emissions. During the study period, the remaining cities' net carbon emissions transitioned to more advanced zones, except Tai'an City and Zaozhuang City, which shifted from a level 4 to a level 2 net carbon emission zone. Heze, Dezhou, Weihai, and Rizhao city consistently fall within the zone of low carbon emissions.

The trajectory of the center of gravity movement (Fig. 8) and the evolution trend (Table 8) can further reveal the spatial distribution pattern of net carbon emissions. The center of gravity shifted path from Yiyuan County to Zichuan District in 2010–2015 and then shifted slightly to the northwest in Zichuan District in 2015–2020, with an offset distance of about 34.3298 km. Zibo is located in the center of Shandong Province. As an old industrial city, it is the only city in China that covers three types of resource-exhausted cities: independent industrial and mining zones and old industrial bases. It has formed an industrial structure dominated by traditional industries such as chemicals and textiles, and the relevant statistical data show that the heavy chemical industry of Zibo accounts for nearly half of the industrial structure, making the city of Zibo a city with high net carbon emissions.

Fig. 8

Standard deviation ellipse and center of gravity change of net carbon emissions

Table 8 Standard deviation elliptic parameters of carbon emissions in Shandong Province

Regarding spatial differentiation patterns, carbon emissions in Shandong Province as a whole are dominated by the northeast–southwest direction, with a gradual shift to the northeast, suggesting that the influence of the northeastern region on the spatial pattern of carbon emissions in the province has been strengthened. The cities in northeastern Shandong Province have seen their carbon emissions grow faster than average, and the region with the fastest growth in total carbon emissions continues to expand to the northeast. The main reason for this is the significant increase in urbanization, the expansion of land for construction, and the large-scale mechanized production of agriculture, which is accompanied by a large amount of energy consumption. Specifically, net carbon emissions in Dongying and Binzhou have soared, much higher than in other cities in northeastern Shandong Province. This is because Dongying, which has gradually taken heavy industry and manufacturing as its leading industries, is the city with the largest capacity of ground refining in the country, and the city's petroleum equipment manufacturing industry accounts for one-third of the output value of the same industry in the country. Binzhou City is mainly developing the manufacturing industry, which has a higher demand for raw materials.

Regarding the variables of the standard deviation ellipse for carbon emission alterations, the variation in the ellipse's area during 2010–2015 exceeds that in 2015–2020, demonstrating that the shift in carbon emissions' center of gravity in the initial phase is more widespread than in the subsequent phase. Regarding migration distance, the movement of the center of gravity of carbon emissions is divided into two stages: from 2010 to 2015, the center of gravity migrated by 32.0854 km, and from 2015 to 2020, the center of gravity migrated by 2.2444 km. This indicates that the growth of carbon emissions in the cities of northeastern Shandong Province in the former stage is more significant and substantially impacts the overall carbon emission pattern. Regarding the divergence shape, the proportion between the short and long axes of the standard deviation ellipse declines annually. This suggests the shape of the distribution of carbon emissions in the province is flattening in the province's carbon emission distribution pattern, with a gradual evolution towards the long axis (northeast–southwest) and a deceleration towards the short axis (northwest–southeast). In the direction of divergence, the standard deviation ellipse azimuth of the spatial distribution of carbon emissions in Shandong Province increased continuously from 64.0096° to 69.0016°, indicating that the spatial dispersion of carbon emission pattern in Shandong Province tended to expand and showed a shrinking trend in the main direction, and the direction was stable from northeast to southwest.

Figure 9 illustrates that the patterns of net carbon emissions in both northwestern and central Shandong Province exhibit a consistent leftward skew, with regional variances displaying a rising pattern. In contrast, the patterns in Southern Shandong Province and JiaoLai regions exhibit intermittent phases. A steady rightward shift in the distribution curve of Northwest Shandong Province and Central Shandong Province, signifying an upward trajectory in net carbon emissions throughout the research period. The peak count in the JiaoLai region shifted from a singular peak in 2010 to a dual peak in 2020, with a continuous rightward shift in peak, suggesting a more polarized net carbon emission. Additionally, there is a left–right trailing trend in the Southern Shandong Province region's net carbon emissions, signifying an increase in both cities with low and high net carbon emissions. In 2015, the JiaoLai region's net carbon emission waveform showed a leftward shift relative to 2010. In 2020, there was a rightward shift in the waveform relative to 2015. This suggests that the net carbon emission polarization in the JiaoLai area fluctuates between positive and negative. A decline follows a pattern of growth in the regional disparity.

Fig. 9

Evolution of regional net carbon emissions dynamics

Spatial and temporal analysis of the carbon ecological carrying capacity

This study classifies the carbon ecological carrying capacity into five levels based on the natural discontinuity approach, where level 1 is the minimal and level 5 the maximal (Fig. 10). For an in-depth examination of each city's carbon ecological capacity, create a chart to depict (Fig. 11).

Fig. 10

Spatial and temporal distribution of carbon ecological carrying capacity (× 10⁴ tons)

Fig. 11

Carbon carrying capacity of forest, grassland, and crops

Cities with high carbon ecological carrying capacity include Weifang, Linyi, Dezhou, Heze, and Yantai city, among others, while those with low carbon capacity are chiefly Dongying, Weihai, Rizhao, and Zaozhuang. Within this context, Binzhou City's carbon ecological carrying capacity exhibits a fluctuating pattern, transitioning from a level 2 carbon ecological carrying area in 2010 to a level 3 one between 2015 and 2018 before shifting to level 2 in 2020. Post-2018, Weifang City's carbon ecological capacity exhibited a downward trajectory, evolving into a level 4 carbon ecological carrying area by 2020. During the research phase, Zibo, Dongying, Tai’an, and Qingdao maintained a steadier condition.

As shown in Fig. 11, the carrying capacity of grassland is relatively stable in all cities, with Linyi and Yantai having a higher carrying capacity than the other cities. The difference in the carrying capacity of forests is more evident than that of grasslands. Jinan, Linyi, and Yantai have significantly higher woodland carrying capacities than the others, while Binzhou, Dongying, Heze, and Dezhou have the lowest. Differences in crop carrying capacity were most pronounced among prefecture-level cities, especially in Dezhou, Heze, Liaocheng, and Weifang, where crop carrying capacity was significantly higher than in other cities. This aligns with the actual situation of more developed local agriculture and larger grain-sowing areas. Among them, the most obvious change in the carrying capacity of crops in Heze is due to the vegetable industry revitalization plan introduced by the Heze municipal government in 2018, the strengthening of infrastructure construction, the introduction and promotion of advanced vegetable cultivation technology and management mode, and a significant increase in vegetable production in Heze.

Analysis of factors affecting carbon emissions

Detection of single factor

By employing geo-detectors to examine the factors influencing land-use carbon emissions in Shandong Province, the q-value for each was determined, revealing varying levels of contribution from each factor to land-use carbon emissions (Table 9). PS, UR, LED, EA, IS, TP, LS, and R&D are population size, urbanization rate, level of economic development, economic aggregate, industrial structure, technological progress, land-use structure, and R&D investment.

Table 9 The q-value of the explanatory force of driving factors on carbon emissions

In 2010, the influence of various factors on carbon emissions due to land use ranked in a decreasing manner: economic aggregate > urbanization rate > technological progress > level of economic development > land-use structure > industrial structure > R&D investment > population size. The q-value for economic aggregate, urbanization rate, technological advancement, and economic development level exceeded 0.7, representing 63.66% of the overall explanatory power impacting carbon emissions. Thus, these elements primarily influence the carbon emissions from land utilization. In 2015, except for population size and R&D investment, whose explanatory power increased, the explanatory power of the rest of the factors showed a decreasing trend. Only urbanization rate, economic aggregate, technological progress, and R&D investment had a q-value greater than 0.5, and the influence of a single factor showed a decreasing trend. During the 13th Five-Year Plan period, Shandong Province has seen rapid economic development, expanding cities and increasing urbanization rates, while cities have focused on optimizing and adjusting industrial structure, accelerating the conversion of old and new kinetic energy, and slow technological progress and low R&D investment, which have become the main drivers of carbon emissions. The trend of decreasing explanatory power of the single factors becomes more pronounced after 2018, with only the 2018 economic aggregation factor and the 2020 economic aggregation and technological progress factors having q-values greater than 0.5.

Interaction test

Interaction between the indicators was further explored using a two-way interactive detection method with a geographical detector (Fig. 12).

Fig. 12

Two-factor interaction heat map

After the interaction of various influencing factors, their influence was significantly enhanced, and the enhancement relationship is a predominantly bilinear enhancement, i.e., the variations in carbon emissions from land use within the research zone stem not from a solitary element but from creating a complex interplay of multiple factors. After the interaction between population size and level of economic development, the effect was significantly increased. In 2018 and 2020, the interaction between population size and level of economic development was measured at 0.999 and 0.991, respectively, indicating that population size growth and level of economic development significantly impact carbon emissions. After the interaction of industrial structure, land-use structure, and R&D investment, the influence of the interaction was significantly more potent than that of the single factor, with the influence of industrial structure and R&D investment interaction reaching over 0.9 in 3 years. Industrial structure and land-use structure are essential components of regional development. The impact of these two actions on regional carbon emissions is not apparent, but their interaction will increase the impact on land carbon emissions after increasing R&D investment. In 2020, the single factors of economic aggregate and R&D investment enormously influence Shandong's carbon emissions. After interacting with R&D investment, they are incredibly influential compared to the single factors, with the interaction detection result being as high as 0.98 or more. In total, urbanization rate, economic aggregate, and technological progress show high explanatory power in single-factor and interaction tests, strongly suggesting that these factors are the most critical drivers of land-use carbon emissions in Shandong Province.

Discussion

Discussion of the findings

In terms of land use, Shandong Province is dominated by cultivated land and construction land, and there was a trend of decreasing the cultivated land area and expanding the construction land area. A possible explanation for this is the increasing level of urbanization and industrialization in Shandong Province. According to the Shandong Provincial Urbanization Development Program (2012–2020), the government supports the expansion of cities within the province to augment construction land. During the study period, the urbanization rate in Shandong province increased from 40.25% to 50.38%. The accelerated urbanization has led to more and more people coming to the region to work, and the number of urban residents increased by 12,863,600 between 2010 and 2020. Land for factory construction and workers' residential land increased as part of the source of construction land expansion. This is also consistent with the findings of other researchers [25]. Although there was a general trend of expansion in the area of construction land during the study period, there was a decrease between 2015–2018 and 2018–2020. One possible explanation is the different categorization of construction land in the statistical yearbook and remote sensing data. The construction land in the statistical yearbook of Shandong Province includes residential land, industrial and mining land, transportation land, and water-use land. The construction land in the remotely sensed data includes urban land, rural settlement land, and other construction land. Other construction land includes factories, mines, large industrial areas, oil fields, salt fields, quarries, transportation roads, airports, and special land. The difference in statistical caliber may lead to the difference between the area of construction land extracted from remote sensing data and the area recorded by statistical caliber. Other scholars have also concluded that the construction land area has decreased, which the author explains [30]. The author explains that the decrease in the construction land area in the Yellow River Delta and coastal area of Binzhou may be due to the impact of the implementation of aquaculture wetland restoration projects in the area. Since 2015, the water area has increased significantly. In 2009, China's State Council officially approved the Development Plan for the Yellow River Delta High-Efficiency Eco-Economic Zone, which has a near-term target until 2015 and a long-term outlook until 2020. The plan proposes mandatory protection of water sources and coastline nature protection zones, strict restrictions on all development and construction activities, acceleration of comprehensive water area management, rationalization of town construction, and strict control of the population scale. This is also in line with the actual situation, which is that the water area in the northern part of Dongying and Binzhou City has increased significantly since 2015 (Fig. 3). The unutilized land in Shandong Province is mainly distributed in Dongying, Binzhou, and the northern part of Weifang City in the Yellow River Delta. Dongying City accounts for 17.87% of the province's unutilized land due to the high degree of soil salinization. The General Land Use Plan of Shandong Province (2006–2020) states that the Yellow River Delta should be guided to reasonably increase the amount of land for construction by developing of unutilized land. The “One Body, Two Wings” development strategy is also appropriately tilted towards the Yellow River Delta in allocating targets, focusing on supporting the development of unutilized land into construction land. The suitability of unutilized land was evaluated, and development was carried out on alkaline land with concentrated distribution and conditions for development, consistent with the results of the reduction in the area of unutilized land during the study period.

Regarding the analysis of carbon emissions, the total carbon emissions showed an increasing trend during the study period, and the carbon emissions of each city also showed different degrees of increase. However, the growth trend of carbon emissions slowed after 2015, which may be related to the energy-saving and emission-reduction policies implemented by the local government to recognize the importance of the ecological environment. In 2011, it was proposed in the 12th Five-Year Plan that new energy and renewable energy would be vigorously developed, and the proportion of coal in energy consumption would be reduced. As a result, the trend of increasing carbon emissions in Shandong Province will slow down after 2015. At the city scale, the most significant increase in carbon emissions was observed in Binzhou during the study period (Fig. 4). This phenomenon occurred because Binzhou City has developed a secondary industry, a developed chemical and aluminum industry, and high energy consumption, which generates a large amount of carbon emissions. The high inertia of energy carbon emissions in economic development makes it difficult for Binzhou City to realize the low-carbon development of its economy in the short term [55]. Analyzed from the aspect of land use, construction land is the main source of carbon emissions, and the total amount of carbon sources is much larger than the total amount of carbon sinks, which is also consistent with the conclusions of previous studies [22, 42]. Therefore, controlling carbon emissions should not only start from land use, but also control the scale of construction land. It is also particularly crucial to adjust the industrial structure and improve energy utilization efficiency.

In the analysis of carbon sinks and carbon ecological carrying capacity, forest's carbon sequestration accounts for a more significant proportion of total carbon sequestration, about 65–70%, and the proportion of water area is 27–32%. Forest carbon sinks are one of the economical and effective ways to cope with climate change [17]. By analyzing the carbon ecological carrying capacity, the study found that crops contributed more to the carbon ecological carrying capacity of the region than forest and grassland. A reasonable farming system and improved cultivated land quality can increase carbon uptake and thus enhance the carbon sequestration capacity of cultivated land [46]. To this end, the Shandong Provincial Government has introduced a series of cultivated land protection policies. The Shandong Province Land Rectification Plan (2016–2020) and Shandong Province Land Use General Plan (2006–2020) state that the strictest cultivated land protection system will be adhered to, the red line of cultivated land protection will be strictly adhered to, and the occupation of cultivated land, especially basic cultivated land, for non-agricultural construction will be strictly controlled. The construction of high-standard cultivated land has been strengthened, giving full play to the characteristics of the region's large cultivated land area and centralized layout to achieve a combination of centralized and decentralized protection and curb the phenomenon of “non-agriculturalization” of cultivated land. Promoting the comprehensive utilization of saline and alkaline land has not only increased the cultivated land, but also improved the region's ecological environment. Shandong Province is a significant center for agricultural output and a crucial zone for safeguarding cultivated lands in China, holding the third position in total grain production. Its cultivated land protection policies are much stronger than those of provinces with a higher degree of urbanization or a lower agricultural focus. Provinces with a higher degree of urbanization are likely to focus more on urban expansion and intensive land use, while provinces with a lower agricultural focus are likely to place relatively less emphasis on arable land protection. In contrast, Shandong Province has achieved a win–win situation between cultivated land protection and agricultural development by protecting agricultural land while promoting agricultural modernization and rural revitalization. Based on this, more significant efforts should be made to protect the region's forest, water area, and grassland resources to maximize carbon sequestration. In addition, as a food-intensive province, the carbon sink role of crops should be fully taken into account, and the occupation of cultivated land should be strictly prohibited to ensure that cultivated land is occupied and replenished in a balanced manner. The constraining effect of the cultivated land protection policy in curbing the expansion of construction land should be maximized [49].

The study analyzed drivers of land-use carbon emissions and found that urbanization rate, economic aggregate, and technological progress are the main drivers. As urbanization continues to accelerate, the area of construction land is expanding. The main contributors to carbon emissions are industrial production, energy use, and human endeavors in construction land [37], reflecting how the urbanization rate and economic aggregate affect carbon emissions. In addition, technological progress is one of the critical factors that cannot be ignored, and it plays an inhibiting role in carbon emissions. It is especially crucial to improve energy utilization efficiency and realize green production through technological progress, which is also in line with the results of previous studies [45].

Limitations and prospects of the study

Although this paper analyzed the carbon emission characteristics of Shandong Province from the aspects of land-use type, city level, center of gravity migration and distribution trend, and analyzed the emission reduction potential based on carbon ecological carrying capacity, revealing the impact of the interaction of different driving factors on carbon emission in Shandong Province, bridging the shortcomings of existing research literature and providing references for realizing coordinated development among regions. However, there are still some shortcomings: (1) uncertainties in the carbon emission factor. First, the method of calculating the carbon emission factor is based on an empirical model [53]. There are significant differences in the geographic conditions (e.g., climate, soil, vegetation, etc.) and socio-economic conditions (e.g., the level of economic development, the industrial structure, the structure of energy consumption, etc.) of different regions, which will lead to differences in the carbon emission factor of the same land-use type in different regions. Therefore, if the same carbon emission coefficient is directly used for estimation, such regional differences may possibly be neglected. Even if the average value is calculated based on the results of several researchers, it is difficult to avoid errors completely. Furthermore, local policies (e.g., carbon emission reduction and land-use strategies) are undergoing changes as technology advances and economic development strategies shift. Changes in these policies may not be reflected in the carbon emission factors on time, thus potentially generating errors. (2) Selecting energy sources and crop types is not comprehensive. Based on data availability, only eight energy sources and seven crop types were selected to estimate the carbon emissions from construction land and the carbon-carrying capacity of crops in Shandong Province. Although the main energy sources and crop types in Shandong Province were selected, they still cannot fully reflect the carbon emissions from energy consumption and the actual crop ecological carrying capacity in Shandong Province. (3) Due to data availability, the land-use data in this paper are derived exclusively from remote sensing data. Given the extensive study area, the resolution of our base data may not be optimal, potentially leading to inconsistencies in the extracted land-use data compared to actual conditions. At the same time, the land-use data are discontinuous, which may make the time cross-section data unable to truly reflect the development trend and characteristics of the study period.

Thus, future research will concentrate on constructing a precise system for carbon emission factors and calculating carbon emissions across various geographical regions, timeframes, and scales. Efforts should be made to establish a sound local data structure, improve the accessibility of city- and county-level data, and provide a solid database for research from smaller administrative units. In addition, the accuracy and quality of remote sensing data should be improved, and complete time series data should be established to provide continuous data support for the research.

Policy recommendations

Optimize the land-use structure and promote intensive land use

Carbon emissions mainly come from construction land, and the government should encourage all land-use entities to use construction land more intensively, economically, and in a low-carbon manner, strengthening the regulation of construction land's disorderly expansion. Forest and water area are the main contributors to carbon sequestration in Shandong Province, so it is necessary to implement the strategic planning of territorial space, strengthen the cultivation of grassland resources, strengthen the ecological protection and restoration of water area, and give full play to its role in carbon sequestration and emission reduction. In addition, the reduction of cultivated land is mainly converted into construction land. Therefore, the encroachment of cultivated land should be strictly prohibited, and illegal acts should be investigated and punished in time to ensure the “balance of occupation and compensation” of cultivated land to give full play to crops' carbon emission reduction potential.

Optimizing energy structure and developing a green economy

Economic aggregates mainly drive carbon emissions in Shandong Province, making it vital to modify the industrial structure for sustainable and low-carbon growth. On the one hand, we need to accelerate the transformation and upgrading of traditional industries and reduce the proportion of conventional high-energy-consuming sectors in the industry. At the same time, we also need to increase the development of new service industries and gradually optimize the secondary industry-center industrial layout in many regions through industrial alternation. The principle of “who pollutes, who pays” is adhered to, and the standards for the payment of sewage charges and environmental taxes are strictly enforced. Use economic means to incentivize and constrain polluters' behavior and take the green and low-carbon development path.

Emphasizing regional differences and tailoring policies to local conditions

The research uncovers notable regional variances in the amount and elements influencing carbon emissions due to land utilization in Shandong Province. Consequently, in developing plans to lower carbon emissions, we should consider the uniqueness of the local area and choose methods that are compatible with the region and in harmony with the overall development strategy so that we can gradually realize the goal of “dual-carbon”. At the same time, combining the characteristics of spatial aggregation, we should formulate differentiated carbon emission reduction targets for each region and reasonably allocate carbon emission reduction tasks. For cities with vast forests, grasslands, and high-yield crops, it is necessary to develop solid ecological protection strategies to ensure the long-term sustainability of their ecosystems and stabilize the carbon ecological carrying capacity.

Conclusions

After analyzing the spatial and temporal distribution characteristic of land-use carbon emissions and the driver factors in Shandong Province, the following conclusions were derived:

• (1) Land use is dominated by construction and cultivated land, with a smaller share of other land types. Land use shows an increasing trend in construction land, water area and forest, and a decreasing trend in cultivated land, grassland and unutilized land;

• (2) Construction land's carbon emissions were the highest, constituting 98% of the overall carbon source, while cultivated land's carbon emission remained around 2%. In forest and water area, carbon uptake is more significant, in contrast to grassland and unutilized lands, where it is less significant. Between 2010 and 2020, carbon emissions from construction land increased more rapidly, hitting 41,042.44 × 10⁴ tons;

• (3) The distribution patterns of carbon source and sink in Shandong Province exhibited distinct spatial and temporal features. Shandong Province's high-carbon sink area was primarily concentrated in Yantai, while the primary locations for low-carbon sinks were concentrated in Dezhou, Liaocheng, Heze, and Zaozhuang City, and the gravity of net carbon emissions stretched towards the northeast;

• (4) Crop carrying capacity varies more across municipalities than forest and grassland. Weifang, Linyi, Dezhou, Heze, and Yantai have high carbon ecological carrying capacity. Dongying, Weihai, Rizhao, and Zaozhuang, in the low carbon ecological carrying capacity zone. Urbanization rate, economic aggregate and technological progress are the main drivers of carbon emissions in Shandong Province.

This study provides theoretical and data support for various land management policies and local government's carbon emission reduction measures from the land use perspective. In order to achieve the goal of “double carbon”, the land-use structure should be optimized, and the intensive use of land should be promoted. At the same time, the energy structure should be optimized, and a green economy should be actively developed. Considering the differences between regions, formulate corresponding emission reduction measures according to local conditions. In addition, climate change is a global issue that requires the joint efforts of all countries to cope with it and advance the global climate governance process. This study is a reference for other regions in China to conduct carbon emission studies. It can provide valuable insights for other regions in Europe and the world to address climate issues by optimizing land-use structures. Specifically, the study shows that construction land is the main source of carbon emissions in Shandong Province's land use. Rapid urbanization has led to the conversion of arable land mainly into construction land. This finding suggests that in the process of urbanization in European countries, the unnecessary expansion of construction land and the occupation of other land-use types should be given high priority, and appropriate land-use policies should be formulated to mitigate this adverse effect. Through an in-depth analysis of the drivers of Shandong Province, we find that the urbanization rate, the overall economy, and technological progress are the dominant factors. Among them, urbanization rate and economic aggregate play a driving role in the increase of carbon emissions, while technological advancement can effectively inhibit the rise of carbon emissions. This result reveals that in pursuing economic growth, European countries should focus more on choosing a low-carbon development path and reducing carbon emissions through technological innovation. In addition, from the perspective of policy insights, the Shandong government has adopted economic strategies such as the polluter pays principle in environmental policymaking. Europe can further explore and improve the application of economic instruments in environmental policies, for example, through environmental taxes and emissions trading, with the help of market mechanisms to encourage enterprises and individuals to reduce carbon emissions.

However, only the carbon emissions from land use from 2010, 2015, 2018, and 2020 have been studied. In future research, land-use data over a long period and energy data of smaller administrative units should be combined to study the carbon emission status of land use at the county level to provide solid theoretical support for land use planning.