Decoupling and decomposition analysis on the CO 2 emissions of tourism industry: A case study of Hainan

. Currently, little attention was paid to the tourism’s CO 2 emission (CE) at province level. Thus, taking Hainan as a case, we computed this province’s CE, and analyzed the relationship between Hainan’s tourism economy and its CE, and the drivers of the CE. The results showed that Hainan’s tourism CE increased rapidly from 99.88 ×10 4 t in 2001 to 475.07 ×10 4 t in 2015. Particularly, Tourism transport always accounted for the largest proportion of tourism CE (more than 74%). Moreover, Hainan presented a holistic weak decoupling (0.68) during 2001-2015. But the decoupling rate was only 57.14%. Thus, Hainan still has much potential to improve the energy-use efficiency of tourism industry for accelerating the decoupling process. In addition, the effect of population was the dominant driver to promote Hainan’s tourism CE followed by expenditure size effect with the contribution rates of 132.52% and 11.78%, respectively. Whereas energy intensity effect played the most primary role in inhibiting CE followed by industrial structure effect, and their contribution rates were -38.65% and -5.58%, respectively. Last, based on these results above, some reasonable countermeasures and suggestions are proposed.


Introduction
Currently, global warming or climate change has deeply exerted obvious influence on the environment [1]. Thus, more efforts should be taken to mitigate the corresponding greenhouse gas, such as the carbon dioxide (CO 2 ) emissions. The CO 2 emissions (CE) are mainly produced by burning fossil fuels (e.g. coal, oil, gas). In particular, tourism has become the industry with the largest number of participants among many developed and developing countries [2]. Such a huge human activity will inevitably consume quantities of energies and produce the corresponding emissions which, in turn, go on impairing environment and aggravating global warming [3]. So, it is necessary and imperative to comprehensively study the related issues of tourism's CE.
In reality, China's economic sectors including the tourism industry could not further develop without energy resources. But energy consumption (EC) will inevitably lead to further environmental pollution and rise of carbon emissions. Therefore, how to coordinate the economic growth and environmental protection should be scientifically made clearer, especially in tourism industry. Wu and Shi (2011) used a bottom-up method to systematically estimate the whole China's tourism CE [4]. Then, using the same method, Tang et al. (2018) estimated the tourism-related CO 2 emissions in China from 2000-2015, the results showed the highest tourism-related CO 2 existed in eastern coastal China, while the least was in the west of China [5]. Moreover, Kuo and Chen (2009) adopted a life cycle assessment approach to study the energy use, the corresponding CE, as well as some other environment influences caused by tourism in Taiwan's island [6]. In addition,  introduced the Tapio decoupling index and Logarithmic Mean Divisa index (LMDI) to reveal the relationship between tourism economic growth and tourism CE in Yangtze River Delta [7].
Although people have performed such fruitful studies, these studies most focused on national perspective. In other words, the studies at the province's level are little. Moreover, few cases have analysed the nexus between tourism economy and its CE, meanwhile, studied the drivers of tourism CE as well. Thus, the two works have been simultaneously studied in this paper, which has a certain innovation. First, taking Hainan as the case study, we assess the nexus and decoupling process between the tourism economy and its tourism CE; second, we use the LMDI model to further analysis the underlying drivers of the tourism's CE.

Methodology
The bottom-up method used to computer Hainan's tourism EC and CE is from the article [4]. The Tapio decoupling indicator is from the article [8], and the LMDI is from the article [1].  Therefore, it could be concluded that Hainan should pay more attention to optimizing the tourism transport section and formulating more reasonable policies or measures to reduce the related transport's energy consumption. Similarly, the trajectory of the tourism CE in Hainan was basically consistent with the tourism EC. The tourism CE in Hainan increased from 99.88×10 4 tonnes (10 4 t) in 2001 to 475.07×10 4 t in 2015 respectively, with an annual growth amount of 26.80×10 4 t and rate of 11.78%. The obviously negative growth was -4.91% in 2003 and -7.96% in 2009, which was consistent with the EC's change. From three sub-sectors' percentage of tourism CE (Fig. 2), we could easily see that tourism transport emitted the largest proportion of the total tourism CE in Hainan. Overall, tourism transport accounted for no less than 74%, a very high share, of total tourism CE. Besides, it should be noteworthy that the percentage of tourism transport's CE increased constantly during 2006-2015 (11 th -12 th FYP), then reached to the summit (94.24%) in 2015. In contrast, the percentage of tourism accommodation's CE experienced an opposite change: the proportion of tourism accommodation's CE was from 17.87% in 2001 to 21.62% in 2005, then encountered a continuous decline to 4.58% in 2015. As for the tourism activities' CE, it usually only occupied few shares of the tourism CE (<1.2%).

Decoupling analysis
As shown in Table 1 During the 11 th FYP (2006-2010), the nexus between tourism revenues and its CE turned to weak decoupling (0.54). At this stage, Hainan's tourism revenues and its CE increased with an annual average growth rate of 16.18% and 9.69%, respectively. Undoubtedly, Hainan's tourism was deeply affected by the financial crisis in 2008. Therefore, the tourism was in expansive negative decoupling during 2007-2008. In addition, China's government took some powerful measures to stimulate economy, so that strong decoupling and weak decoupling reoccurred in sequence during 2008-2010.
During the 12 th FYP (2011-2015), Hainan's tourism revenues rose with an annual average rate of 12.06%, meanwhile, Hainan's government paid more attention to environmental protection. Thus, Hainan's tourism still maintained the weak decoupling (0.61). Although the decoupling value was higher than that in the 11 th FYP (0.54), its tourism decoupling status was more stable. Apart from expansive coupling in 2012-2013, all the remaining years were weak decoupling. This phenomenon was echoed with whole China's economy.

Drivers of CE's Changes at three stages
Based on the LMDI model mentioned above, the additive decomposition results of drivers were expressed in  From 2001 to 2015, the contribution rates of these four drivers were 132.52% (population growth effect), 11.78% (expenditure size effect), -38.65% (energy intensity effect) and -5.58% (industrial structure effect), respectively. The total promotion effect (147.30% = 135.52%+11.78%) was more than that total reduction effect (-44.23% = -38.65% -5.58%). So, overall, these effects led to a 103.07% increase in the total CE over the studied time.
Population effect always had a positive influence on the tourism CE growth, and its contributions for increasing CE were 36.07×10 4 t, 115.61×10 4 t and 226.68×10 4 t in the 10, 11 and 12 th FYP, respectively. The growth speed was fast. For instance, the number of tourists and the tourism revenues were 53.36 million and 57.25 billion in 2015, which was 4.74 times and 6.51 times than that in 2001, respectively. These implied the scale of Hainan's tourism industry continuously expanded, and it inevitably caused the increase of tourism CE. Obviously, tourism transport occupied the biggest share of the tourism CE. Hence, with the development of Hainan's tourism industry, the effect of population triggered 338.53×10 4 t CE, nearly 89.5% of the total CE (378.36×10 4 t). In addition, the tourism accommodation and activities totally led to 36.77×10 4 t and 3.05×10 4 t, but their growth rates were slow. This also verified the above analysis in Figure 2.
Expenditure size effect was another important driver to promote CE's growth. Its contribution for increasing CE was totally 33. However, energy intensity effect was the key factor to inhibit the tourism CE. Its contribution for reducing tourism CE was 110.35×10 4 t CO 2 during 2001-2015. Especially, the CE's decrease speed due to energy intensity effect was faster in the later time. This was because most of China's districts (including Hainan) improved their energy-use efficiency to mitigate carbon emissions during 2006-2015. Robust evidence could also be found in the tourism's three sub-sectors. The energy intensity effects of tourism transport, accommodation and activities resulted in -38.76×10 4 t, -5.97×10 4 t, and -0.26×10 4 t, respectively. Thus, the energy intensity effect of tourism transport played a significant role in declining CE. Similarly, the contributions of industrial structure effect were -15.94×10 4 t in 2001-2015, indicating that industrial structure was also a factor of reducing the tourism CE. Specifically, the changes of the CE caused by tourism transport, tourism accommodation and tourism activities were 10.88×10 4 t, -28.77×10 4 t and 1.96×10 4 t over 2001-2015.

Conclusions
Based on the data of 2001-2015 of Hainan province, we applied the bottom-up method to computer the total EC and CE of Hainan's tourism. Then, using the Tapio index (LMDI model), we analyzed the decoupling nexus between tourism economy and the matching CE (the drivers of CE But the decoupling rate was only 57.14%. Thus, Hainan still has much potential to improve the energy-use efficiency of tourism industry for accelerating the decoupling process. The contributions of the population effect, expenditure size effect, energy intensity effect and industrial structure effect were 378.36×10 4 t, 33.63×10 4 t, -110.35×10 4 t and -15.94×10 4 t, respectively. The results indicated that the effect of population was the most significant driver of CE growth followed by expenditure size effect. However, energy intensity effect played the most primary role in habiting CE followed by industrial structure effect.
According to these results above, the following countermeasures could be put forward to help the lowcarbon tourism development in Hainan. First, to achieve energy savings and emission reduction in the long run, Hainan should lower the tourism transport CE, especially the airplanes' CE. Thus, Hainan should take some effective policies or countermeasures to optimize the structure of tourism transport. Second, the government should make efforts to provide some low-carbon travel routes and advocate tourists to strengthen the awareness of environmental protection. Third, the energy intensity had an obvious negative effect on tourism CO 2 emissions. Therefore, Hainan should invest more in related R&D business and accelerate the pace of innovation to improve energy efficiency to cut down the corresponding tourism CE.