Existing studies hold that risk can be considered as a measure of uncertain results. Disaster risk refers to the level of potential impact and its potential occurrence probability in the future. Risk perception, which was first discussed by Bauer (1960) in the consumer behavior field, is defined as the expectation of loss caused by uncertainty about the outcome that is likely to occur given the consumer’s choice (Ahyoung, 2021). A representative study was Slovic Paul’s theory of “personality profile” in 1987 which measured risk perception using the psychometric paradigm of risk perception (Slovic, 1987). Many studies have also measured risk perception from multiple dimensions, such as functional, physical, economic, social, and psychological risks, rather than a single component (Hong, 2015; Ahyoung, 2021). In the context of disaster tourism issues, such as natural disasters and landslides, risk perception is a comprehensive embodiment of disaster risk, exposure, vulnerability, and disaster prevention and reduction ability (Pan et al., 2016). Risk perception refers to people’s feelings and understanding of risks taken in the surrounding world, including personal experience based on intuitive judgment and subjective feelings (Meng et al., 2010). It is also people’s beliefs, attitudes, judgments, and emotions about hazards and its potential benefits, as well as cultural and social tendencies in the broader sense (Wu and Zhang, 2012).

In previous studies, risk perception has been widely taken as a key variable that affects people’s travel intentions or as barriers to leisure travel. Pahl et al. (2005) explored that if a person thinks that he is at risk and feels vulnerable because of natural disasters, he is more likely to take pro-environmental action. Thus, risk perception is positively related to tourists’ behavioral intention.

Knowledge is “a fact or condition of knowing something with a considerable degree of familiarity through experience, association, and contact” (Zhou et al., 2022). Hoffmann and Muttarak (2017) revealed that those who had been influenced by disaster in the past 3 years show nearly three times more preparedness behavior. Meanwhile, other literature studies demonstrated that disaster could improve people’s experiences, especially for those who perceived the severity of the hazard (Wiwik, 2020). According to these studies, tourists’ disaster knowledge refers to tourists’ information about natural disasters that happened in tourism attractions.

Some studies found insignificant correlations between disaster knowledge and people’s experiences (Sim et al., 2018). A person who can obtain disaster knowledge through different channels will show an overall lower impact on his/her risk perception. A person who has experienced natural disasters has a significant impact on his/her risk perception. Persons with past flood experience and flood knowledge education are quite sensible to flood risk perception (Okaka and Odhiambo, 2019). The fear of experiencing an earthquake positively and indirectly affects the risk perception of earthquake survivors (Bronfman et al., 2020). Chinese residents’ perception level of flood disaster risk is relatively low (Zhou and Liu, 2019), while residents’ perception level of geological disaster risk in the mining area is relatively high (Feng and Ning, 2020). Li et al. (2011) found that disaster knowledge, as the most critical variable, has a significant impact on people’s psychological and behavioral processes when sudden disasters such as earthquakes occur. Gao (2019) believed that the perceived use of new media and tourism information will affect the tourism behavior of youth groups. Therefore, there is a positive correlation in the relationship between disaster knowledge and tourists’ intentions because of the disaster experience.

Drawing upon the theory of attachment, place attachment refers to a person’s complex emotional and affective response to a place or location, value, and attribution (Liu et al., 2021; Morais and Lin, 2010). It has been applied to investigate tourists’ emotional, functional, affective, and social attachment to tourism destinations (Bowlby, 1969; Hidalgo and Hernandez, 2001; Kyle et al., 2004; Xu and Zhang, 2015). Place attachment has been conceptualized as a complex construct with multiple dimensions (Woosnam, et al., 2018). Researchers have proposed two (Hwang, 2005), three (Yuksel et al., 2010), and even four dimensions (Ramkinssoon et al., 2012) for place attachment.

Tourists’ emotional attachment significantly affects their risk perception and travel behavioral attitude. A person’s dependence on the place of disaster positively affects the perception of disaster risk. Participants from flood prone areas believe that the perception of flood risk is higher (Ullah et al., 2020). Most participants with a higher socio-economic status have higher risk perception of tsunami due to their residence being near the coastline (Gotham et al., 2018). By measuring risk perception, Shen et al. (2020) revealed that the presence of children or older people in the respondent’s family induces a higher risk perception. There are significant differences in risk perception within the non-expert population and between the expert and non-expert populations (Pelling, 2011). Leiter (2011) found that risk and specific characteristics of individuals are important determinants of participants’ perception of risk. Kim et al. (2016) found that an experience from past destinations only plays a regulatory role in the relationship between age and tourism risk perception (Jang et al., 2020). Therefore, there is a significant effect on emotional attachment for tourist behavioral intentions.

Some studies found that women who had experienced geological disasters have higher risk perception of geological disasters and pay more attention to disaster knowledge (Yang et al., 2018). Women with high levels of risk perception tend to have negative attitudes toward disaster reduction (Yang et al., 2018). Other studies revealed that educated youth have significant intention to travel to places where they have been in their youth time for work, and the yacht tourists’ behavior and attitude have a significant positive impact on their travel behavioral intention (Liu and Jia, 2018; Gou et al., 2020). Young people are usually the main demographic of tourists in mountain tourism destinations. Hu et al. (2014) analyzed the characteristics of mountain tourism recreation behavior in Wugong Mountain in China and found that there are mainly young students and tourists who are satisfied and willing to revisit the mountain tourism area. Zheng et al. (2016) also found that vacationers in the Nankun Mountain in Guangdong Province of China are mostly young and middle-aged tourists. Varasteh et al. (2015) studied the travel behavior of international students studying in Malaysia and found that most international students tend to travel during semester holidays for the purpose of entertainment and exploring new places. Youth group tourists pay attention to the tourism safety measures applied in mountain areas, but they have low safety awareness and lack of knowledge on mountain tourism safety (Cen and Huang, 2011).

Thus, different groups, such as male and female, and high education level and low education level, had different impacts on the relationship between tourists’ risk perception, disaster knowledge, emotional attachment, and their travel behaviors (Gao et al., 2017).

Tourists’ behavioral intention is defined as “subjective judgments about how we will behave in the future” (Ahyoung, 2021). In regard to disaster tourism matters, tourists’ behavioral intention can be described as the extent to which groups will take actions to continue their travel in various ways, such as overcoming all kinds of disaster impacts to visit tourist attractions in the future (Ramdas and Mohamed, 2014).

Based on the above literature review, we consider tourists’ emotional attachment, disaster knowledge, and disaster risk perception as the main driving factors, and analyze the impacts on tourists’ behavioral intentions in this study. Therefore, the following research framework and conceptual model of this study are developed (Figure 1).

The Qinling Mountain Range (32°28′53″-34°32′23″N, 105°29′18″-111°01′54″E) in the hinterland of Shaanxi Province in China stretches from the Jialingjiang River in the west to the Bahe River in the east. This range is connected with the Sichuan Basin in the south and the Loess Plateau in the north, spanning six cities of Shaanxi Province—Xi’an, Baoji, Weinan, Hanzhong, Ankang, and Shangluo (Figure 2).

The Qinling Mountain and Huaihe River are the watersheds between northern and southern China. The Qinling Mountain Range is about 1,600 km long from east to west and 300 km wide on average from north to south.

The average elevation is about 1,500 m with the highest peak, Taibaishan Peak (3,771.2 m above sea level), located in Baoji City. Situated between the subtropical zone and the temperate zone with a mountainous continental climate, the Qinling Mountain has an obvious vertical zonality. The annual precipitation can reach 1,000–1,300 mm, creating climatic conditions very conductive to the formation and development of landslide and debris flow in Qinling Mountain.

Known as the “Central National Park of China,” rich natural resources and historical culture in Qinling Mountain feed numerous tourism attractions, such as the Huashan National Park, Zhongnanshan Global Geopark, and Nanwutai Cultural site, and their exotically beautiful landscapes attract tourist attention worldwide.

In fact, quite a few tourist attractions in Qinling Mountain frequently suffer heavy geological hazards (Ma et al., 2021). According to the investigation results of Shaanxi Geological Survey Department, landslide and debris flow that existed in the Qinling Mountain caused an average of 55 people missing or dead each year with a direct economic loss of 57 million RMB.

To measure the constructs proposed in the conceptual model, a questionnaire was first developed based on previous studies and then tested on 30 college students. The final questionnaire was composed of 24 measurement items, which are described in Table 1. A 5-point Likert-type scale was used to measure each statement of risk perception, disaster knowledge, emotional attachment, and behavioral intentions (from 1 = strongly disagree to 5 = strongly agree).

1) The measurement of disaster risk perception is based on the research items published in Fuchs and Reichel’s (2006) tourism disaster risk perception, including six items in terms of the risk of disaster-causing factors and the vulnerability of the carrier.

This analysis investigates the perception and influencing factors of tourism disaster risk in Qinling Mountains using youth group tourists. The data are obtained from a questionnaire survey. The first survey was conducted during winter season in Qinling Mountain in January 2021, and the second survey was conducted before the summer season in May 2021 to collect a comprehensive dataset. Two surveys were conducted in the form of electronic questionnaires, and a total of 542 questionnaires were received. Then, a range between 15 and 44 was defined as the age of youth group tourists. After screening, 511 questionnaires were considered valid, with an effective rate of 94.3%. The survey results were imported into SPSS 25.0 software for sample descriptive statistical analysis, and the results are shown in Table 2.

In terms of gender distribution, males accounted for 42.1% and females accounted for 57.9%. Concerning education level, most of them have a good education level: 279 tourists have a master’s degree or above, accounting for 54.6%; 219 tourists have a college or bachelor’s degree, accounting for 42.9%; and 13 tourists have a diploma of high school, vocational school, or technical secondary school and junior middle school or below, accounting for 2.6%. From the perspective of personal monthly income, 172 tourists earned below 1,000 RMB, accounting for 33.7%, followed by 5,000–10,000 RMB, 1,001–3,000 RMB, 3,001–5,000 RMB, and more than 10,000 RMB, accounting for 22.1, 20, 14.1, and 10.25%, respectively. In terms of current residence, 248 tourists currently live in Xi’an city which is the capital city of Shaanxi Province, accounting for 48.5%, while 20.2% live in other areas of Shaanxi Province and 31.3% live in other provinces. The occupations of the survey participants are mainly students; enterprise managers; personnel of public institutions; and professional, cultural, educational, scientific, and technological personnel, accounting for 53.6, 11.4, 11.2, 9, and 6.5%, respectively. There are a few state agency personnel, workers, service personnel, soldiers, farmers, and retirees, accounting for less than 10% of the total participants. Based on the above analysis, the research object and data have been considered with enough quality to meet the research objective.

In order to test the adaptability between the item and the variable factor, a single-factor test was applied. The results show that a total of 4 factors and 18 items can be extracted and their eigenvalues are all greater than 1, and the interpretation rate of cumulative variance after rotation is 78.2%, indicating that 18 items in this study have good adaptability to variable factors.

In this study, SPSS software was used to analyze the reliability and validity of the questionnaire. In the reliability test, it is generally considered that if the Cronbach’s alpha coefficient is greater than 0.7, the date would have high reliability. The Cronbach’s alpha coefficients of the four variables in this study are shown in Table 3, as follows: 0.951, 0.940, 0.946, and 0.919, respectively, and the overall reliability is 0.936, indicating that the data are highly reliable and can be considered for further analyses. In addition to the reliability test, the content and structure validity were also examined. Since the variable measurement items draw on the mature scales of existing research, they have good content validity. A KMO test was conducted to test for structural validity. The KMO value of this questionnaire is 0.920, and the significance in the Bartlett sphere test is 0.000, less than the significance level of 0.05, indicating that the research data have good discriminant validity.

Previous tourism and hospitality research published in the literature shows that tourism emotional attachment of a visitor will be beneficial to his/her travel attitudes and behavioral intentions (Kim et al., 2017). Yuksel et al. (2010) found a strong link between place attachment and visitors’ satisfaction, a key determining factor of tourists’ attitude (Prayag & Ryan, 2012). Tourism emotional attachment can be a good predictor of visitors’ emotional experience of satisfaction and might have a direct effect on destination loyalty. It has been found that tourists’ attachment has an indirect significant positive effect on their behavioral intentions (e.g., revisit intentions and travel purchase intentions) (Cheng et al., 2017).

According to the analysis results, young tourists’ emotional attachment to Qinling Mountains positively and significantly affects their travel behavioral intentions. The higher the emotional attachment level of the youth group to Qinling, such as emotional experience and cognitive level, the stronger their behavioral willingness to travel to this area will be.

It is reasonable to assume that changing travel environment also impacts travelers’ preference of travel services. The sociological concept of “cocooning” is widely used in social science to explain the human behavior when a person prefers to stay in his/her room instead of going out (Wut, et al., 2021).

According to the results of our empirical analysis, among youth group tourists who visited Qinling Mountains, the quality of disaster knowledge has a positive and significant impact on their travel behavioral intentions. When youth group tourists got access to the information of natural disasters in Qinling Mountains as comprehensively and accurately as possible, such as the time, frequency, and damage degree of disaster, their travel intentions and travel behavioral intentions became more obvious.

Existing research studies in the literature found that threat severity and susceptibility can induce “travel fear,” which leads to protection motivation and protective travel behaviors after the outbreak of the COVID-19 pandemic (Zheng, et al., 2021). Findings also revealed that “travel fear” can trigger different coping strategies, which increases people’s psychological resilience and adoption of cautious travel behaviors. Second, disease has an impact on tourists’ behaviors; one possible explanation is that travelers might develop adaptive behaviors (personal health interventions) which may prevent them from contracting the disease (Cahyanto, et al., 2016). The pandemic is also likely to change tourists’ lifestyles, travel behaviors, and patterns (Guan, et al., 2020).

The analysis of the results concludes that the tourists’ perception of disaster risk has a negative impact on the tourists’ behavioral intentions of youth group tourists in Qinling Mountains. If the youth group has a greater perception of the risk of various natural disasters of Qinling Mountains, they may cancel their travel plans or change to another destination. On the contrary, if they positively perceive the potential risk of Qinling Mountain, they may keep their travel plans.

According to the analysis results of the moderating effect of gender and education level, the moderating effect of gender is non-significant, but the moderation role of education level is significant which has a positive moderating effect on disaster knowledge for tourists’ travel intentions. In other words, there is no significant difference between male tourists and female tourists with regard to their behavioral tendencies. Tourists with a high education level can comprehensively and accurately assimilate the information of various natural disasters about tourism destinations, which strengthens the relationship between disaster knowledge and their travel behavioral intentions.

This analysis expands the theoretical study on risk perception of tourism disaster. First, on the basis of explaining the concepts of disaster risk, risk perception, and tourism perception, combined with the current research literature, we constructed a research framework for analyzing the influencing factors of tourists’ travel behavioral intentions in disaster-prone tourism destinations. Second, risk perception theory and TPB were applied jointly in this study, and risk perception and tourism emotional attachment were both taken as influencing factors to expand the category of influencing factors of tourists’ behavioral intentions in TPB. Third, we analyzed the moderating role of tourists’ gender and their education level, which enriches the research on the moderating effects in tourism disaster management studies.

The conclusion has great importance for policymakers dealing with the risk management in disaster-prone tourist destinations. On the one hand, tourism destinations can strengthen tourists’ emotional experience of destinations through well-provided services, good tourist images, and unforgettable tourist experiences, and thereby promoting their willingness to travel again. On the other hand, tourism destinations need to release natural disaster knowledge comprehensively, such as the time and damage degree of disasters, through various channels via Internet, TV, and social media, to help tourists establish a positive risk perception. And, well-educated tourists should be regarded as the main source of tourists in the Qinling Mountains destination, take them as the target tourists for the Qinling Mountains, and promote and serve them well.