Driving Forces of Tourism Carbon Decoupling: A Case Study of the Yangtze River Economic Belt, China

Abstract

Although decoupling tourism growth from carbon emissions is vital for sustainable tourism development, the driving forces of tourism carbon decoupling in the Yangtze River Economic Belt (YREB) are little known. Herein, our study applies the geo-detector model and the Tapio decoupling index to investigate the decoupling trend and driving mechanism of the tourism economy in the YREB from carbon emissions from 2009 to 2019. Our results show that (1) the tourism carbon decoupling status has gradually evolved from connection to decoupling, and the average decoupling index was optimized from 1.36 in 2011 to 0.34 in 2019; (2) the dominant factors promoting the evolution of decoupling are the industrial structure (with an average q of 0.64 (2009–2019)) and the urbanization index (with an average q of 0.61 (2009–2019)), with government policy, technological innovation capability and consumption, and regional GDP also being important drivers; and (3) the double and nonlinear enhancement between the driving factors imply that regions in poor decoupling areas, such as Shanghai and Chongqing, can promote the evolution of decoupling through multi-factor interactions to realize the sustainability of the tourism industry. Finally, an integrative and proactive policy framework that has important theoretical, methodological, and management implications for the construction of green demonstration areas in the YREB is proposed.

1. Introduction

The world has witnessed the production of excessive carbon emissions through economic growth driven by energy consumption. The reduction of carbon emissions has become a global challenge, and many countries have taken legislative measures, policy oaths, and other actions to reduce emissions [1]. As the world’s largest developing country and one of the leading emitters of CO₂, China promotes green transformation in economic development and is achieving remarkable development in this direction, forcing China’s industrial transformation and upgrading and promoting high-quality economic growth. Meanwhile, China has promised to reach a carbon peak by 2030 and carbon neutrality by 2060, which indicates the direction for China’s low-carbon transformation to deal with the international problem of greenhouse gas emissions. However, China’s “double carbon” target, as a national action with the largest and shortest emission reduction in the world, faces great challenges. To achieve the goal of double carbon while maintaining stable and orderly economic and social development, it is necessary to complete the decoupling of economic growth from carbon emissions as soon as possible [2].

Tourism is a pivotal engine of economic growth in China. Decoupling tourism-related carbon emissions from economic growth is vital to providing practical solutions for China’s low-carbon development [3]. The tourism economy has been heavily affected by COVID-19 [4]; the global tourism market is recovering slowly and remains unstable. However, there are no lasting structural changes in the tourism systems due to COVID-19 [3].Tourism is still an integral part of China’s extremely large-scale market advantages and domestic demand potential. In addition, studies have shown that tourism-related carbon emissions account for about 8% of total carbon emissions and are gradually increasing at an annual rate of 3.2% [5]; thus, there is great potential for energy conservation and emission reduction in tourism. However, some scholars argue that policies that focus too much on reducing carbon emissions may also increase the risk of tourism-related economic loss [6]. It is also impractical to achieve decoupling by curbing the scale of the growth of tourism or its level of expenditure [7]. Meanwhile, sustainable tourism is one of the targets set in goals 8, 12, and 14 of the Sustainable Development Goals (SDGs) of the European Union [8]. In light of this, addressing the conundrum of the completely durable and robust decoupling of tourism development from carbon emission is vital to realizing sustainable tourism development.

At present, the research on the relationship between the tourism economy and environmental cost mainly focuses on three aspects. One is using the environmental Kuznets curve (EKC) to describe the linkage between tourism-related economic development and ecological environment quality [9]. The second is testing the two-way causal relationship between tourism development and carbon emissions through measurement [3]. The third is using the decoupling model to analyze the decoupling relationship between tourism-related economic growth and carbon emissions [10,11]. The decoupling model is widely used in the research on the relationship between tourism growth and carbon emissions because it can quantitatively analyze the dependence on carbon emissions in the process of tourism-related economic development and effectively identify the evolution stage [12]. However, most existing studies on decoupling have concentrated only on the descriptive analysis of the decoupling status [7,13,14,15,16,17,18]. The potential driving mechanism of the decoupling between tourism growth and environmental cost remains poorly understood.

In China, the tourism development level, decoupling effect, and influencing factors of tourism-related carbon emissions are always accompanied by high spatial heterogeneity. The geographical detector model is mainly used to explore the stratified spatial heterogeneity and the affecting drivers of variables [19]. Currently, the geographical detector model has been applied in many research fields, such as human health [20], urbanization [21], enterprises [22,23], and the environment [24,25,26]. However, few studies applied this spatial statistical model in tourism. Therefore, we combined the geo-detector technology and Tapio decoupling model to detect the driving factors of the decoupling relationship between the tourism economy and tourism-related carbon emissions, and the spatial interaction knowledge between tourism growth and carbon emissions can provide valuable theoretical support for the sustainable development of tourism in the Yangtze River Economic Belt (YREB).

2. Materials and Methods

2.1. Research Area

As the first demonstration zone for constructing an ecological civilization and the central economic belt of China (Figure 1a), the YREB has natural ecological advantages and a solid economic foundation; it covers 11 provinces and cities, occupies 21.4% of China’s area, and accounts for over 40% of its GDP and population. However, the total carbon emissions of the YREB also account for about half of those in China [27]. Thus, in the 14th Five-Year Period, it is pivotal for the YREB to take the lead in achieving regional “carbon peaking” and “carbon neutrality” by building a green and low-carbon development demonstration zone. The YREB has rich and unique tourism resources, attracting quite a few tourists from home and abroad (Figure 1b). In 2019, the total tourism revenue was CNY 10,598.1 billion, accounting for 23.2% of the regional GDP. In total, 8.389 billion tourists were received, including 8.331 billion domestic tourists. Even in 2020, which was seriously affected by the pandemic, tourism revenue was as high as CNY 6228.5 billion. Therefore, given its popularity, typicality, and representativeness, the YREB is selected as a case to estimate the applicability of our analytic framework.

2.2. Methods

2.2.1. Carbon Emission Estimation Model of Tourism

Based on the “bottom-up” method and previous studies [28,29,30] (

Table A1. Carbon emissions from various transportation modes.

Transportation Modes	Railways	Highways	Waterways	Civil Aviation	Researchers
$W_i$(%)	32.7	27.9	10.6	37.6	2011 [45], 2012 [46], 2014 [10]
$a_i$(g/p km)	27	133	106	137	2012 [46], 2014 [10], 2016 [47]

), this study selects the tourism transportation, accommodation, and activities as the sources of carbon emissions from tourism in the research area (

Table A2. Carbon emissions from various tourism activities.

Tourism Activities	Sightseeing	Leisure Vacations	Business Conferences	Visiting Relatives and Friends	Others	Note
${\beta }_k$(g/visitor)	417	1670	786	591	172	2006 [48], 2021 [3]

).

The estimation formula of the CO₂ emission of tourism is as follows:

$$C_{total}=C_{tran}+C_{act}+C_{accom}$$

(1)

where $C_{total}$ is the CO₂ emissions of the tourism industry; and $C_{tran}$, $C_{act}$, and $C_{accom}$ are the CO₂ emissions of tourism transport, tourism activities, and tourism accommodation activities, respectively.

1.

The estimation of tourism transportation is as follows:

$$C_{tran}={\displaystyle \sum }_{i=1}^n(Q_i\ast W_i\ast a_i)$$

(2)

where $Q_i$,$W_i$, and $a_i$ are the passenger turnover of the $i$ mode, the proportion of tourists in passenger turnover, and the carbon emission factor, respectively.

2.

The estimation of tourism activities is as follows:

$$C_{act}={\displaystyle \sum }_{k=1}^n\left(m\ast {\beta }_k\ast P_k\right)$$

(3)

where $m$,${\beta }_k$, and $P_k$ are the tourist reception, carbon emission factor, and proportion of visitors to the k-activity, respectively.

3.

The estimation of tourism accommodation is as follows:

$$C_{accom}=b\ast s\ast T\ast \beta \ast \epsilon$$

(4)

where $b$,$s$,$T ,\epsilon ,$ and $\beta$ are the number of beds, room occupancy rate, days in the year (365), energy consumption per bed per night (MJ/per bed night), and carbon content per unit of calorific value (g C/MJ), respectively. We set $\beta$ as 130 MJ/per bed night [3,31] and $\epsilon$ as 43.2 g C/MJ [3,32].

2.2.2. The Tapio Decoupling Model

First, the OECD (Organisation for Economic Co-operation and Development) first proposed the concept of the decoupling indicator [33]; then, Tapio (2005) redefined decoupling indicators and distinguished eight decoupling states, proposing the Tapio decoupling model [34]:

$${\theta }_i=\frac{\Delta C_i/C_i}{\Delta T_i/T_i}=\frac{\left(C_i^t-C_i^0\right)/C_i^0}{\left(T_i^t-T_i^0\right)/T_i^0}$$

(5)

where ${\theta }_i \mathrm{is}$ the decoupling index $;$ and $C_i$,$T_i,\Delta C_i, \Delta T_i,C_i^0, T_i^0,C_i^t, \mathrm{and} T_i^t$ are the carbon emissions, tourism income, changes in carbon emissions, changes in tourism income, base-period carbon emissions, base-period tourism income, carbon emissions in the t period, and tourism income in the t period, respectively. The classification of ${\theta }_i$ is listed in Figure 2.

2.2.3. The Geographic Detector Model

To explore the effects of different factors on spatial heterogeneity, Wang et al. (2010) [35] proposed the geographical detector model, which consists of four different detectors: the factor detector, the interaction detector, the risk detector, and the ecological detector. This study mainly uses factor and interaction detectors. The geo-detector is a statistical method to detect spatial differentiation and reveal the driving factors behind it.

• Factor detector
  The association can be examined as follows:

  $$q=1-\frac{{{\displaystyle \sum }}_{h=1}^L{{\displaystyle \sum }}_{i=1}^{N_h}{\left(y_{h1}-\overline{y_h}\right)}^2}{{{\displaystyle \sum }}_{i=1}^N{\left(y_i-\overline{y}\right)}^2}=1-\frac{{{\displaystyle \sum }}_{h=1}^L{N}_h{{\sigma }_h}^2}{N{\sigma }^2}$$

  (6)

In this expression, q denotes the degree to which the influencing factors explain the spatial heterogeneity of dependent variables (tourism carbon decoupling index); the value interval of q is (0, 1) and the larger the value, the stronger the explanatory ability of this factor to change the tourism carbon-decoupling index.$L$ denotes the classification number of factors.$N_h$ denotes the number of samples in type h, while $N$ is the number of all sample areas under study. The variables ${\sigma }_h{}^2,{\sigma }^2$ denote the variance of the independent variable of layer h and the whole region, respectively.

2.

Interaction detector

The evaluation method of the interaction detector is to first calculate the q of any two influencing factors on the dependent variables, q(X 1) and q(X 2). Then, we calculate q when the two influencing factors interact, q(X 1∩X 2). Finally, we compare q(X 1), q(X 2) and q(X 1∩X 2) to judge their interaction (

Table A3. The evaluation criteria of interactive relationships.

Evaluation Criteria	Interaction	Graphical Representation
$q\left(X_i{\displaystyle \cap }{X}_j\right)<Min(q\left(X_i{\displaystyle \cap }{X}_j\right)$	Weak, nonlinear
$Min(q\left(X_i{\displaystyle \cap }{X}_j\right)<q\left(X_i{\displaystyle \cap }{X}_j\right)<Max(q\left(X_i{\displaystyle \cap }{X}_j\right)$	Weak, univariate, nonlinear
$q\left(X_i{\displaystyle \cap }{X}_j\right)>Max(q\left(X_i{\displaystyle \cap }{X}_j\right)$	Enhanced, linear, bi-near
$q\left(X_i{\displaystyle \cap }{X}_j\right)=q\left(X_i\right)+q\left(X_j\right)$	Independent
$q\left(X_i{\displaystyle \cap }{X}_j\right)>q\left(X_i\right)+q\left(X_j\right)$	Enhanced, nonlinear

Note: Sort according to Wang et al. (2010) [35].

).

2.3. Influencing Factor Indicators

The determinants of tourism carbon decoupling are attributed to the influence and disturbance outside the tourism eco-economic system. Thus, by analyzing the socioeconomic environment that is closely related to the tourism eco-economic system and relevant literature [36,37], including the regional economic strength, the degree of industrial structure, urbanization, residents’ consumption ability, tourist scale, consumption level, government policy, and technological development, we obtained the eight drivers listed in

Table 1. The influencing factor indicator system.

Type	Detection Factor	Indicator	Unit of Indicator
Industrial Structure (IS)	$\mathrm{X}1$	The proportion of tertiary industry	%
Consumer Spending Power (CSP)	$\mathrm{X}2$	Per capita disposable income	Yuan
Urbanization Index (UI)	$\mathrm{X}3$	The proportion of the urban population	%
Regional GDP (RGDP)	$\mathrm{X}4$	The total number of domestic and foreign tourists	100 million
Technological Innovation Capability (TIC)	$\mathrm{X}5$	Patent application authorization	10,000 pieces
Government Policy (GP)	$\mathrm{X}6$	The government expenditures on energy protection and environmental conservation	100 million
Tourist Arrival (TA)	$\mathrm{X}7$	The total number of domestic and foreign tourists	10,000 person
Consumption Level (CL)	$\mathrm{X}8$	The total tourism revenue/tourism arrival	None

.

2.4. Data Sources

Since the tourism industry was severely impacted by COVID-19 beginning in 2020, the research period in this study is from 2009 to 2019. The data on total tourism revenue $T_{ total}$ were calculated by multiplying the tourism foreign exchange income by the average exchange rate of the current year plus the domestic tourism income, and the passenger turnover $Q_i$ was obtained from the statistical yearbooks of provinces and cities (https://ip.cn/huilv.html, accessed on 14 July 2022).

The number of beds $q$, the room occupancy rate $s$, and the number of visitors $m$ were extracted from the Yearbook of China Tourism Statistics. The proportion of tourists each year was collected from the Tourism Sample Survey Information. Detection factors X1, X2, X5, X6, X7, and X8 were obtained from the statistical yearbooks of provinces and cities, while detection factor X3 was calculated by $T_{ total}$. Moreover, detection factor X4 was also extracted from the Yearbook of China Tourism Statistics, and the data for a few regions or years were obtained from the local Statistical Bulletin.

3. Results

3.1. Decoupling Situation of Tourism-Related Carbon Emissions

3.1.1. Analysis of Tourism Carbon Emission Trend

The growth curve of tourism-related carbon emissions in the YREB can be divided into three periods: a rapid growth period from 2009 to 2012, a descent phase from 2012 to 2014, and a relatively slow growth stage from 2014 to 2019 (Figure 3a). Meanwhile, for the sectoral emission structure, the proportions of tourism transportation, tourism activities, and accommodation decreased in turn. With the increase in living standards and the improvement in leisure time, tourist travel demand has been stimulated, increasing the proportion of carbon emissions from tourism activities, while the proportions of carbon emissions from tourism transportation and tourism accommodation show a downward trend. Furthermore, concerning the provinces, the changing trend of carbon emissions in the tourism industry is consistent with that in the YREB, while the carbon emissions from tourism in Chongqing and Sichuan are still high (Figure 3b).

3.1.2. Decoupling Tourism-Related Carbon Emissions

Based on the Tapio decoupling model, this study examines the decoupling relationship between tourism growth and carbon emissions in the YREB (Figure 4). Only three states of decoupling appeared in the tourism economy system of the YREB during the study period: weak, expansive negative, and strong; the economy finally stabilized in a weak decoupling state, with an average decoupling index of 0.28. In general, the decoupling index over 2010–2019 is deemed optimistic concerning the green tourism economy, except in 2011 due to the expansive negative decoupling (Figure 4a).

Specifically, in the provinces, the decoupling index exists as a volatile trend over the research period, and the spatial differentiation among provinces is obvious (Figure 4b). In particular, the decoupling index fluctuated wildly in Yunnan, Jiangsu, and Anhui provinces, while the fluctuations in other provinces are not obvious.

3.2. Driving Factors of Decoupling Tourism-Related Carbon Emissions

Before the analysis of the geo-detector, it is necessary to discretize the influencing factors [38]. The factors are divided into five levels in this study using the natural breakpoint method in ArcGIS10.8 software (Figure 5). Then, the eight discretized influencing factors are used as detection factors, and their effects on the spatial differentiation of the tourism carbon emission decoupling index are analyzed through geographic detectors.

3.2.1. Driver Analysis

As shown by the factor detector (

Table 2. Geo-detector q statistics of factors affecting the carbon decoupling index of the tourism industry.

Year	Factor Detector	IS (X1)	CSP (X2)	UI (X3)	RGDP (X4)	TIC (X5)	GP (X6)	TA (X7)	CL (X8)
2010	q statistic	0.78 ***	0.34	0.58 ***	0.22	0.17 ***	0.87 ***	0.72	0.68 ***
	p value	0.001	0.94	0.006	0.14	0.008	0.005	0.21	0.005
2011	q statistic	0.34 ***	0.19	0.56 ***	0.20 *	0.19 *	0.29 ***	0.22	0.27 *
	p value	0.008	0.96	0.007	0.09	0.09	0.008	0.89	0.09
2012	q statistic	0.53 ***	0.51	0.51 **	0.32 ***	0.27 *	0.60 ***	0.15	0.53 *
	p value	0.003	0.68	0.044	0.008	0.09	0.004	0.93	0.06
2013	q statistic	0.46 *	0.54	0.73 **	0.61 **	0.66 **	0.51 ***	0.27	0.28 *
	p value	0.054	0.49	0.019	0.04	0.02	0.006	0.82	0.08
2014	q statistic	0.99 ***	0.99	0.99 ***	0.10 *	0.25 ***	0.45 ***	0.45	0.28 *
	p value	0.001	0.38	0.0004	0.09	0.007	0.005	0.58	0.07
2015	q statistic	0.96 ***	0.94 **	0.96 ***	0.14	0.17 *	0.28 ***	0.45	0.48
	p value	0.004	0.01	0.002	0.90	0.08	0.007	0.58	0.54
2016	q statistic	0.68 ***	0.64	0.40 *	0.38 *	0.25 *	0.41 ***	0.41	0.37 *
	p value	0.007	0.70	0.085	0.08	0.08	0.006	0.83	0.07
2017	q statistic	0.46 ***	0.38	0.44 ***	0.06 *	0.43 ***	0.30 ***	0.44	0.11 *
	p value	0.008	0.85	0.007	0.09	0.005	0.007	0.75	0.09
2018	q statistic	0.60 ***	0.51	0.29 ***	0.32 ***	0.30 *	0.12 ***	0.51	0.33 *
	p value	0.007	0.82	0.009	0.009	0.07	0.009	0.79	0.07
2019	q statistic	0.73	0.84	0.78	0.04 *	0.73 **	0.41 ***	0.88 *	0.51 **
	p value	0.375	0.13	0.27	0.09	0.03	0.006	0.07	0.04

Note: ***, **, and * indicate significance at 1%, 5%, and 10%, respectively.

), most of the eight driving factors affecting the decoupling of tourism-related carbon emissions in the YREB pass the significance test and are statistically significant. The factor detector also presents the explanatory power of these significant drivers with regard to the spatial differentiation of the tourism carbon decoupling index.

The results show that (1) the impact of IS (X1) showed a trend of increasing and then decreasing. Except for in 2019, the IS was significant in all years. It had the greatest influence in 2014 and showed a decreasing trend in the later years. (2) CSP (X2) was not a driving factor. The impact of CSP on the decoupling index only passed the significance test of 5% in 2015. It was not significant in any of the other years. (3) The influence of UI (X3) increased and then decreased. The impact of UI on the decoupling index only failed to pass the significance test in 2019, and it was significant in other years. Its influence was the largest in 2014, while in 2018, its impact was the smallest. (4) RGDP (X4) was not significant in 2010 and 2015 but was significant in the other years. In 2013, it exerted the greatest influence on the decoupling index, while in 2019, it had the least impact. In general, its q value fluctuated between 0.0426 and 0.6182. (5) The impact of TIC (X5) was significant, and its influence shows a trend of strengthening, then weakening, and then strengthening. (6) The impact of GP (X6) remained significant during the research period, and the impacts tended to increase later. (7) TA (X7) was significant in 2019 but not in any other year. (8) CL (X8) was not significant in 2015, but it was statistically significant in other years. In 2010, it exerted the greatest influence on the decoupling index. Consequently, CL (X8) had a reduced impact on the decoupling index. In summary, its influence also fluctuated.

3.2.2. Interaction Analysis

The explanatory power of the interactions between the different influencing factors is mainly manifested in two-factor enhancement and nonlinear enhancement, among which two-factor enhancement is the more common (Figure 6). Therefore, the spatial differentiation of the tourism carbon decoupling index for the YREB is the result of the comprehensive action of multiple influencing factors. The interaction and superposition of multiple factors will enhance the evolution of tourism carbon decoupling.

Take 2018 as an example: After IS (X1), UI (X3), and RGDP (X4) interact with other influencing factors, the explanatory power of these factors is significantly greater than the total of two factors, characterized by the enhancement of two factors. Meanwhile, after TIC (X5), GP (X6), and CL (X8) interact with other influencing factors, the explanatory power of these factors is significantly greater than the sum of two factors, which is characterized by nonlinear enhancement. This shows that the interaction of IS (X1), UI (X3), RGDP (X4), TIC (X5), GP (X6), and CL (X8) with other factors will increase the correlation with the spatial differentiation of carbon decoupling in the tourism industry.

4. Discussion

This study innovatively combines the geo-detector model with the Tapio decoupling model, which provides a comprehensive analysis framework for studying the decoupling effect and driving factors between tourism development and carbon emissions. We take the YREB as a case study to examine the applicability of the proposed framework and offer insights into the sustainable development of tourism. We analyze the (1) dynamic changes in carbon emissions in the tourism industry during the observed period; (2) the decoupling nexus between the tourism economy and carbon emissions; (3) the driving factors affecting the decoupling evolution; and (4) the interactive relationships among factors.

4.1. Tourism Carbon Decoupling

First, the total carbon emissions of tourism showed a trend of rapid rise–sharp decline–slow rise during the observation period (Figure 3a). From 2012 to 2014, carbon emissions from tourism showed a strong downward trend related to China’s implementation of the three major emission reduction measures. In addition, the tourism-related economic growth rate showed a downward trend at this stage (Figure 3), indicating that achieving carbon emission reduction in tourism by limiting the scale of growth of tourism does not constitute healthy and sustainable development, which is consistent with previous studies [7]. It is worth noting that the uneven terrain characteristics of Sichuan and Chongqing lead to more energy consumption for tourism transportation than in other regions. Meanwhile, as popular tourist destinations, Sichuan and Chongqing attract a large number of tourists every year, which has led to the high carbon emissions caused by tourism. Similarly, influenced by COVID-19, the tourism-related economic growth rate of the YREB in 2020 was −38.13%, which surely led to a decline in carbon emissions from tourism. In particular, in those areas where tourism is the main economic sector, the impact of COVID-19 has triggered a heated discussion [39,40]. Some scholars believe that COVID-19 has not caused permanent structural changes in the tourism system and that the downward trend of carbon emissions from tourism cannot be maintained for a long time [3]. In this study, due to the lack of data, we did not quantitatively analyze the impact of COVID-19 on tourism-related carbon emissions. Some scholars also deem that cutting carbon emissions cannot and should not be the single goal of sustainable tourism [41]. However, there is no doubt about the impact of COVID-19 on tourism. The lack of any quantitative evaluation of the impacts of COVID-19 on tourism-related carbon emissions is indeed a deficiency of this study, and it will be an important research direction in the future.

Second, with the exception of 2011 due to the expansive negative decoupling, the decoupling relationship between tourism-related economic growth and carbon emissions was confirmed and achieved important results. This is consistent with the research conclusions of many scholars [3,18,42]. Tourism is still an ideal area for emission reduction [43], and it will become an essential breakthrough area in low-carbon development and ecological civilization construction. Meanwhile, the stable weak decoupling state that has existed for many years has proved that it is necessary to investigate the driving factors of tourism carbon decoupling to adopt more effective technology and further promote the evolution of the decoupling effect to strong decoupling, in turn realizing the healthy and sustainable development of tourism.

4.2. Driving Forces of Tourism Carbon Decoupling

The factors that affect the decoupling evolution of tourism-related carbon emissions in the YREB are the industrial structure, urbanization index, regional GDP, government policy, technological innovation capability, and consumption level; however, their impact is gradually weakening.

First, the industrial structure and urbanization index are the leading factors driving the evolution of decoupling, and the average interpretation strength during the study period was more than 50% (Table 2). The impact of industrial structure fluctuated, which may be related to the continuous adjustment of the industrial structure and the positive response of provinces and cities to the policies of industrial structure optimization and upgrading. Moreover, the impact of the urbanization index shows an increasing trend year on year. Compared with tourism destinations far away from cities, urban tourism produces higher carbon emissions [44]. The increase in rural tourism has enhanced the role of new urbanization levels in promoting the evolution of tourism carbon decoupling. In addition, the impact of government policies and consumption levels has also increased.

Regional GDP shows a downward trend that is consistent with the current situation of China’s development. The decoupling effect in the eastern region, with better economic development, is poor, whereas the decoupling levels of the less economically developed central and western regions are good [7]. Compared with the western region, the eastern region’s natural resources and landscape style are slightly insufficient, and they rely more on urban tourism, resulting in higher energy consumption and environmental pollution.

Second, residents’ consumption capacity and passenger-reception capacity are not the driving factors, which is consistent with the research of Zheng [44]. Although the improvement of residents’ consumption capacity and passenger-reception capacity will increase tourism-related carbon emissions, it will also boost tourism income, which will have a small impact on promoting the evolution of tourism carbon decoupling.

Third, the interaction of each driving factor is characterized by superposition and enhancement (Figure 6). The evolution of tourism carbon decoupling in the YREB results from the comprehensive action of multiple influencing factors. The interaction and superposition of multiple factors will be more conducive to the evolution of tourism carbon decoupling.

4.3. Suggestions and Implications

From the above discussion, the future development of tourism in the YREB should continue to promote the evolution of tourism carbon decoupling and give full play to the comparative advantage of tourism as an ideal area for carbon reduction. The following are the recommendations of this study.

First, adjust the structure of the tourism industry and realize transformation and upgrading. Implement the concept of green and low-carbon development [2,3]. Adjust the economic structure of the tourism industry, realize industrial transformation and upgrading, and evolve to a green and low-carbon economic growth model.

Second, encourage technological innovation and improve energy efficiency. Technological innovation, energy conservation, and emission reduction are the main paths to reducing carbon emissions in tourism. Technological progress is conducive to improving energy efficiency and promoting the evolution of decoupling [13,18]. Phase-out low-end and high-energy-consumption tourism products and speed up the upgrading of energy-saving and environmental protection technologies. Focus on high carbon emission departments such as tourism and transportation; develop intelligent transportation; introduce networks, cloud computing, and other new generation technologies into the tourism transportation system; reasonably arrange tourism routes; and improve the flow efficiency of tourists.

Third, promote the construction of new urbanization and reasonably transfer the main battlefield of tourism. In the new urbanization process, pay attention to improving the public service system of rural tourism and strive to shift the principal locations of tourism from cities to villages [27,30]. Based on an area’s tourism resource endowment, reasonably develop low-carbon hotels and scenic spots, build a more regional and unique rural tourism development model, and build and maintain an ecologically pleasant rural tourism resort.

Fourth, strengthen carbon emission monitoring and improve the reward and punishment mechanism. For regions with poor decoupling in the YREB, strengthen the carbon emission monitoring of tourism, clarify the carbon emission responsibilities of relevant actors, establish reward and punishment mechanisms, and focus on institutional and environmental protection policy innovation to realize low-carbon tourism growth.

Fifth, strengthen the leading role of radiation and realize the decoupling and agglomeration effects. Remove administrative restrictions and give full play to the radiation and driving roles of excellent decoupling areas. Promote green and low-carbon cooperation among provinces and cities in the region, promote the coordinated development of the low-carbon integration of tourism, and form a spatial pattern of carbon reduction in tourism development with a matching development orientation and regional synergy and complementarity. Finally, achieve the green demonstration construction of the YREB under the goal of “double carbon”.

5. Conclusions

This study creatively integrates the geo-detector model into the decoupling model, explores the dynamic decoupling relationship between tourism growth and carbon emission, and detects the driving factors of decoupling evolution and the interactions between various factors. It vigorously promotes the evolution of carbon decoupling in tourism and further provides decision-making guidance for the sustainable development of tourism. It is found that the carbon decoupling of tourism in the YREB exhibits a healthy trend. The driving factors of decoupling evolution include the industrial structure, the urbanization index, regional GDP, government policies, the technological innovation capability, and the consumption level. The interaction between the driving factors indicates two-factor and nonlinear enhancement. Finally, we put forward countermeasures and suggestions from five dimensions: the transformation and upgrading of the tourism industry, the technological innovation of tourism energy consumption, the rational transfer of tourism’s main battlefield, the monitoring of tourism carbon emissions, and the agglomeration of the tourism carbon decoupling effect. We seek to further the evolution of tourism carbon decoupling and the construction of green demonstration areas in the YREB.