Natural multi-hazard risk perception and educational insights among Geography and Tourism students and graduates amid the Covid-19 pandemic

Multi-hazard risk perception is an emerging research topic that has been gaining more and more interest since the beginning of the Covid-19 pandemic, due to the unprecedented interactions between the management practices of the pandemic and of other co-occurring natural hazards, and to the compounded impact of these multi-hazards. This paper aims to investigate the specialization and study level-dependent differences concerning multi-hazard risk perception and hazard-related education insights of future potential specialists in natural hazard-induced risk management and tourism reconstruction. These differences were explored by applying a multi-layer questionnaire on 547 Geography and Tourism students and graduates from two universities in Iași City (Romania) and Novi Sad (Serbia), and performing descriptive and differential statistical analyses. The statistically significant differences that emerged refer to estimations of the impact level of the Covid-19 pandemic and of other natural hazards at different spatial scales and on different socio-economic activities, on the training/career of the respondents, and to the hazard-related education improvements. Up to date, none of the papers in the existing literature integrate both the Covid-19 pandemic and the co-occurrent natural hazards as objects of students’ perception, which makes the present study a starting point for such research endeavours.


Introduction
The beginning of the third decade of this century has brought significant global scale challenges that affected both the environment and society in different ways and to various degrees, also hindering the advancement towards the 17 Sustainable Development Goals [1,2]. While the world confronted the Covid-19 pandemic, natural destructive events continued to occur [3][4][5], leading to more severe human toll and damage [6,7]. This overlapping of a biological hazard with a global outbreak and other natural hazards has given way to new research that focuses on multi-hazard risks and specific coping and management strategies [8][9][10]. In this regard, the negative impact of Covid-19 pandemic has been studied in conjunction with many natural hazards (i.e., earthquakes, floods, droughts, hurricanes, wildfires) [11][12][13][14][15][16][17][18][19][20][21].
To date, there are no papers that approach multi-hazard risk perception in a comparative way, focusing on young professional or students who follow education programmes that aim to construct knowledge and skills related to the very subject of perception (i.e., natural hazards and risks). Therefore, this study is a trailblazing, valuable one, which stands out among the ones that focus on the risk perception of students. It is worth mentioning that, while there are numerous papers that deal with the effects of the Covid-19 pandemic on the educational/early career pursuits of Tourism students around the world [74][75][76][77][78][79][80][81], or on the opinions of academics on the pandemic-induced changes of the Tourism sector [82,83], the scientific literature on multi-hazard risk perception in the times of Covid-19 is rather thin and does not account for the perspective of students and early-career specialists [84][85][86]. Also, there is only one study that examines the perspective of Tourism students from two countries (i.e., the Czech Republic and Slovakia) on the present-day changes in the Tourism sector, and the implications for their future career in this field [77], and only one study that briefly explores the impact of the pandemic on the training of Geography students [87]. However, none of the papers in the existing literature integrates both the Covid-19 pandemic and the co-occurrent natural hazards as objects of students' perception.

Regional synopsis of natural hazards
The two universities from which participants were selected for this study are located in Iași City in Northeastern Romania, and Novi Sad in Northern Serbia. These are both large size (over 350,000 inhabitants), administrative urban centres of regional importance, with complex economic, social and cultural profiles [88,89]. Both cities have well-defined areas from which students are attracted, that largely overlap the aforementioned regions. As a direct consequence of the geographical position of Northeast Romania and Northern Serbia in the Central-South-Eastern continental part of Europe, the most representative natural hazards are hydrological and climate-related hazards (droughts, floods, soil erosion and landslides), and earthquakes [43,90], while many other hazards, such as sea-related, and volcanic ones are missing. A particular note of the natural hazard spectrum in the regions of interest concerns wildfires, that have a relatively larger incidence in Northern Serbia compared to Northeast Romania [91][92][93].
Droughts affect both regions, especially the rural communities and agricultural sector, forcing people to find various adaptive solutions to cope with water scarcity [94]. In the Northeast of Romania, droughts occur particularly frequent in August [95], with a historical trend of increased frequency in 1983-2014 [96] a tendency that is foreseen to continue [97,98]. A similar pattern characterizes Northern Serbia [99,100], with a maximum intensity of droughts in July-August, and an increased incidence of droughts in the 1920-2014 period [101]. The impact of floods has varied from moderately to severe in the last century in Northeastern Romania [102], and the recent flood events in 2005, 2008, 2010 and 2020 along the Siret and the Prut Rivers [103,104] had the power to affect many people. The most recent and impactful floods in Northern Serbia took place along the Danube and Tisa floodplains in 2013 [105], and their tributaries in 2014 [106]. Soil erosion and landslides are widespread in both regions due to a large spectrum of natural settings such as geological structure and lithology [107][108][109], relief fragmentation, soil properties, climate and hydrographic network particularities, but also due to the influence of anthropogenic factors [110][111][112].
Seismic risk has medium values for Northern Serbia and high values for Northeast Romania [91,[113][114][115][116]. The territory of Romaniaand even parts of the neighbouring countries, including the Northern part of Serbia are subject to significant seismic hazard, originating in the Vrancea Seismic Zone, which is one of the most active in Europe [117]. It was estimated it can be the source of 2-3 major earthquake events/century [118]. Recent important earthquakes were recorded on the November 3, 2010 in central Serbia (5.4 M W ), when two people died, many others were injured, and the damage was assessed at more than 100 million € [119,120]. The impact of the last major event that occurred in Romania on the March 4, 1977 (7.4 M W , 109 km seismic focus depth) was much greater: damage exceeded 2.8 billion $ at that time, claiming the lives of 1578 people and injuring 11,321 people [121]. Since then, minor damage was reported from earthquakes of over 6 M W in August 1986 and May 1990 [121].
The Covid-19 pandemic strongly affected both countries, causing a high number of deaths, prolonged health problems, and the overburdening of the medical systems [122][123][124][125][126]. The WHO [127] reports that over 3 million Covid-19 infection cases, and almost 66,000 deaths were recorded in Romania from the beginning of the pandemics until the beginning of August 2022. For the same reference period, the numbers specific to Serbia offer a similar overview: 2.1 million infection cases and over 16,250 deaths [128]. Both countries were affected by the last pandemic waves in September-October 2021, and December 2021-February 2022 [127,128]. Following the international trend, the educational systems in Romania and Serbia were also affected by the pandemic crisis. Iași and Novi Sad Universities suspended physical courses, switching to the online system between mid-March 2020 and March-April 2022.

Geography and Tourism higher education at Iași and Novi Sad universities
At the "Alexandru Ioan Cuza" University of Iași, Geography education has a centennial tradition (since 1904). Currently, it is organized following the European Bologna system, which comprises three levels (i.e., Bachelor, Master and PhD studies) and different specializations that focus on Geography or Tourism. Regardless of specialization, at the Bachelor level, there is a dedicated subject on hazards and the associated risks, and there are certain thematic subjects that also refer to the natural and anthropogenic elements and phenomena that contribute to the manifestation of hazardous events. At the Master level, the number of subjects dealing with hazards and risks increases. Moreover, there is a dedicated Master programme called Natural Risks and Territorial Planning. During the PhD studies, PhD students may choose to conduct research on hazard and risk-related topics. At the Department of Geography, Tourism and Hotel management from the University of Novi Sad, which was founded in 1962, the studies are organized following the same study level structure. Students have elective subjects related to natural hazards and risk perception at all levels of studies. However, there is no specialized Master or PhD programme related to Natural hazards or risk management.

Participants
Participants included in this study were Romanian and Serbian students of Geography and Tourism. The sample was collected in Romania ("Alexandru Ioan Cuza" University of Iași) and Serbia (University of Novi Sad). The study sample also included those who graduated in the last five years, and are no longer enrolled in any educational programmes related to their Geography/Tourism background. Both universities are recognized for the tradition of geographical education, having well-defined areas of students' origin, with a regional coverage: the Northeast part of Romania in the case of the "Alexandru Ioan Cuza" University of Iași and Northern Serbia for the Novi Sad University.
The total number of respondents was 547: 318 respondents from Romania and 229 respondents from Serbia. There was a larger number of female respondents (66.2%), and the average age of the sample was 21.7 years old ( Table 1). The sample includes 55.9% Geography students and graduates (GSG), and 44.1% Tourism students and graduates (TSG).

Survey procedure
The research was conducted in February-April 2022 at the two aforementioned universities, when the fifth wave of the Covid-19 pandemic was almost over [127,128]. This was the largest wave in both countries at least in terms of infection numbers [129]. The Serbian respondents were invited to participate in this research by their professors. Both pen-and-paper survey and online survey (distributed through email and students' Facebook groups) options were available. A specialized questionnaire platform [130] helped the survey process in Romania. All students were informed that participation in the survey was voluntarily and anonymous, and that the collected data will be used only for scientific purposes.

Survey structure
The survey instrument used to explore the perception of the GSG and TSG of the Covid-19 pandemic and multi-hazards consisted of a five-layer questionnaire. This was elaborated based on expert insights provided by the scientific literature, and the educational and scientific experience of the authors in the field of risk perception analysis [131]. The first part included demographic information about respondents (gender, age), and their educational background (level of studies, area of specialization). The second part referred to the self-reported knowledge of the respondents concerning risks and natural hazards, and to their knowledge sources. The risks of interest were classified according to a classification inspired by Ref. [29]: (i) Geophysical hazards (earthquakes, volcanic eruptions, landslides, tsunami); (ii) Hydrological hazards (floods, droughts); (iii) Atmospheric hazards (tornado, tropical cyclone, storm, hail, snowstorms, heat waves, climatic changes); (iv) Biophysical hazards (wildfires); (v) Epidemiologic hazards (Table 2). Afterwards, the respondents were asked to estimate on a five-point Likert scale (1-Very low, 2-Low, 3-medium, 4-High, 5-Very High) the dimension of the impact specific to geophysical, hydrological, atmospheric, biophysical and epidemiologic hazards at global, national and continental, local and regional scale ( Table 4). The last question concerned the impact level of the aforementioned natural hazards on tourism activities.  The third layer of the questionnaire included questions related to the relationships between the Covid-19 pandemic and other risks and hazards. Respondents were asked to assess on a five-point Likert scale (1-Very low; 2-Low; 3-Medium; 4-High; 5-Very high) the level to which the Covid-19 crisis has amplified different hazards. Also, participants were asked to express their agreement level related to the fact that the current pandemic crisis resulted in an increase of the vulnerability of human communities to certain types of natural hazards and risks ( Table 7). The fourth part of the questionnaire inquired about the impact of the Covid-19 pandemic and natural hazards on social and economic activities (i.e., Industry, Commerce, Tourism, Education and Social life at the national scale). Also, the participants were asked about the impact of the Covid-19 pandemic crisis, and about the impact of future potential natural hazards with comparable severity on their training/career. The fifth layer of the questionnaire referred to the need for educational improvement concerning natural hazards and risks programmes. The participants were asked to state their agreement level regarding several claims that are detailed in the Results section. All claims were assessed on a five-point Likert scale (1-I strongly disagree, to 5-I strongly agree).

Data analysis
The statistical analysis was performed both in a descriptive and in a differential way. The specialization-dependent differences between participants were analysed via the Independent sample t-test, while ANOVA was performed in order to investigate the study level-related differences. The independent t-test (also called the two sample t-test, independent-samples t-test or Student's t-test) is an inferential statistical test that determines whether there is a statistically significant difference between the means in two unrelated groups. On the other hand, ANOVA test (Analysis of Variance) is a statistical test used to analyze the difference between the means of more than two groups [132]. The next section presents the statistically significant results described by a p-value <.05. The analysis has been performed in SPSS (Statistical Package for Social Sciences).

Results
The presentation of the results is divided according to the main themes of the study: (i) the knowledge on hazards and associated risks; (ii) the links between the Covid-19 pandemic and other risks and hazards, and their impact on economic and social activities, (iii) the hazard-related education and possible improvements.

Knowledge about different risk categories
The analysis shows that a large part of the respondents self-reported a medium level of knowledge about different risk categories ( Table 2, Fig. 1). The participants are most knowledgeable about the risks associated to present days Covid-19 pandemic and natural hazard-induced risks (yet, the mean value is not above 3.5). This is followed by environmental risks (3.2 mean value). Technological and economic risks are the topics about which the respondents are least knowledgeable, according to self-evaluations. Such results indicate that there is still a lot of room for improvement regarding students' general knowledge about risks. The analysis also shows that the main sources of information about risks are education in university (3.5 mean value), as well as social media and mass media (3.4 mean value).
Furthermore, the study intended to compare GSG and TSG regarding their knowledge about risk categories (Fig. 1). The results show that there is a statistically significant difference between GSG and TSG regarding the knowledge about technological risks (industry, transports), and the risk associated to the Covid-19 pandemic. GSG are more knowledgeable about technological risks, while TSG estimate that they have more knowledge about the risks associated to the COVID-19 pandemic (Fig. 1, Table 3). When it comes to different levels of studies, statistically significant differences have been found in the case of natural hazard-induced risks and technological (industry, transports) risks. Specifically, the post hoc LSD test of ANOVA showed that Bachelor participants have less knowledge about natural hazard associated risks than Masters respondents while the latter estimated that they have a higher knowledge level on technological risks than the participants from lower and higher study levels.
Regarding the sources on which the hazard-related knowledge levels are based, there are also some differences between the two groups of students concerning education in university (18+ age), mass media, social media, and friends and family. More precisely, TSG build their knowledge about natural hazards through mass-media, social media, and friends and family more than GSG, who have the university education as their primarily source of knowledge about risks and hazards.
There are statistically significant differences among the study levels concerning the sources of knowledge: education in undergraduate cycles (5-18 age); education in university (18+ age); mass-media, social-media; family, friends. The educational cycles are well highlighted in the responses (Fig. 2). Specifically, the ANOVA showed that those with Bachelor degree, more than those of Master level, learned about natural hazards via education in undergraduate cycles (5-18 age). On the other hand, the participants from higher study levels rely on education in university (18+ age) for hazard-related knowledge. Compared to Master level respondents, Bachelor ones learned more about natural hazards via mass-media, social-media, family and friends.

Perception of the impact level of natural hazards
In the next set of questions, the participants were asked to estimate the dimension of the impact specific to certain hazard categories at (i) local and regional, (ii) national and continental, and (iii) global scale. The respondents were also asked to estimate the impact level of the natural hazards on tourism activities ( Table 4).
The analysis shows that epidemiologic hazards, followed by atmospheric hazards, and geophysical hazards were considered to have the largest impact at the global scale. At the national and continental scale, there is a slight difference in the way that the impact of natural hazards is perceived; but the epidemiologic hazards seem to be considered the most impactful, while the places of the atmospheric and geophysical hazards swap. This perception trend was probably shaped by the Covid-19 pandemic, also showing a lower-level awareness of climate change. At the local and regional level, hydrological hazards (floods, droughts), and geophysical hazards were considered to be the most impactful, followed by the epidemiological ones. Moreover, the respondents consider that all types of natural hazards have quite significant effects on tourism activities, especially in the case of epidemiological and geophysical hazards.
The t-test pointed out statistically significant differences between GSG and TSG regarding the impact of all types of natural hazards on tourism activities: geophysical, hydrological, atmospheric, biophysical hazards; except for the epidemiologic hazards (Table  5). In all cases, TSG consider the impact of the mentioned hazards on Tourism to be greater than the one estimated by GSG. When it comes to different levels of study, there is a statistically significant difference only in the case of the atmospheric hazards: Bachelor participants, more than PhD ones, estimated a higher level impact of atmospheric hazards on Tourism. Table 3 The results of t-test (differences in perception between GSG and TSG) and ANOVA test (level of study) -knowledge about risk categories and sources of knowledge about risks.  Table 4 The mean values of the dimension of the impact specific to natural hazards at different scales, and particularly on tourism activities.

Table 5
The results of t-test (differences in perception between GSG and TSG) and ANOVA test (level of studies) regarding the dimension of the impact specific to natural hazards at different scales, and particularly on tourism activities.  Regarding the impact of natural hazards at local and regional scale, there is no statistically significant difference between GSG and TSG. However, TSG considered that geophysical hazards and hydrological hazards have a high-level impact at national and continental scale, more than GSG. Similarly, TSG considered that geophysical hazards, hydrological hazards, atmospheric hazards and biophysical hazards have greater impact at the global scale, compared to GSG.
The ANOVA showed statistically significant differences in the case of biophysical hazards (wildfires) at local and regional scale, and at global scale (Table 5). Specifically, those PhD level participants, less than Bachelor and Master ones, considered that biophysical hazards (wildfires) have a higher impact at local and regional scale. On the other hand, Bachelor respondents, more than Master and PhD ones, considered that biophysical hazards (wildfires) are more impactful at global scale.

The links between the Covid-19 pandemic and other risks and hazards, and their impact on economic and social activities
The participants considered that the Covid-19 pandemic has mostly amplified economic risks, and that the natural hazard-induced risk were the least amplified ( Table 6, Fig. 3). Also, they largely disagree that the vulnerability to all the proposed hazard types was increased due to the pandemic crisis (Table 7). This proves that both GSG an TSG have yet to acquire a deeper understanding of the interactions behind multi-hazard risks in times of the Covid-19 pandemic. However, there is one particular statistically significant difference (F = 3.453, p = .032) that is most interesting: Bachelor participants, more than PhD ones, considered that the Covid-19 pandemic has amplified natural hazard-induced risks to a larger extent.
GSG and TSG differed in their opinion regarding the influence of the Covid-19 crisis on the vulnerability of human communities to economic risks, geophysical, and biophysical hazards. GSG, more than TSG, considered that human communities became more vul-M.C. Mărgărint et al.

Table 7
The results of t-test (differences in perception between GSG and TSG) and ANOVA test (level of studies) on the influence of the Covid-19 crisis on vulnerability of human communities to different risks and hazard.
The influence of the Covid-19 crisis on the vulnerability of human communities to: The results of independent sample t-test (differences in perception between GSG and TSG) ANOVA testdifferences in level of study nerable to economic risks, while TSG believed that collective vulnerability will increase if geophysical and biophysical hazards are involved (Table 7). When it comes to different levels of study, there is a statistically significant difference between geophysical hazards and biophysical hazards. Bachelor participants, more than doctoral respondents, consider that the Covid-19 pandemic increased the vulnerability of communities to geophysical hazards. The latter category of participants considered that the current pandemic crisis has increased the vulnerability of communities to biophysical hazards to a larger extent than the one estimated by the participants from lower levels of study.
In terms of training/career impact, the mean value of 3.62 (Std. 1.181), shows that the respondents perceive the effects of the Covid-19 pandemic on their training/career as moderate to high. Moreover, there is no statistically significant difference between GSG and TSG regarding this matter. Significant differences appear at the level of studies: Bachelor and Master participants considered that their training/career was highly impacted by the pandemic more than doctoral respondents (F = 4.441, p = .012). More than 60% of the Bachelor and Master respondents indicated high and very high impact levels, while the proportion dropped to 50% in the case of PhD students and recent PhDs (Fig. 4).
The participants were also asked to estimate if different types of natural hazards may affect their training/career in the future to a comparable extent as the Covid-19 crisis did ( Table 8). The analysis shows that more than half of the students appreciate that all types of natural hazards have the potential to affect their training/career, especially the hydrological and atmospheric ones. However, there is a study level-dependent difference regarding the perception of the biophysical hazards' impact (F = 5.671, p = .004): Bachelor participants, more than Masters ones, considered that this type of hazards might be more impactful on their career/training.
The questionnaire required an estimation of the level to which the Covid-19 crisis has affected Industry, Commerce, Tourism, Education and Social Life at national scale, as well as the potential of future natural hazards to affect the same socio-economic sectors    5, Table 9). The respondents considered the Covid-19 pandemic has highly affected tourism, social life and education at national scale, and that future natural hazards have the potential to affect tourism activities in significant ways (Table 9). In addition, the Independent sample t-test indicated that GSG, more than TSG, considered that the current pandemic has highly affected education. On the other hand, TSG considered that future natural hazards may affect tourism and social life to a larger extent than GSG estimated.

Hazard-related education and its improvement perspectives
In this part of the survey, the respondents were asked to express their agreement level to the statement that a high level of natural hazard-related education can help to better the preparedness of society to hazardous events. The mean value of 4.20 (Std. 0.907) suggests that the participants agree. There was no statistically significant difference between GSG and TSG regarding this matter. Fur-

Table 9
The level to which the Covid-19 crisis and natural hazards have affected/may affect the main socio-economic activities. thermore, the respondents were asked to express their agreement level to the statement that the Curriculum from different levels of study should be improved in terms of knowledge about natural hazards (Fig. 6, Table 10). The respondents considered that natural hazards-related education should be improved mainly in higher education (18+ age) and high-school (14-18 age). The ANOVA showed statistically significant differences in case of the primary cycle (5-10 age), the secondary cycle (10-14 age), the college cycle (14-18 age) and the university cycle (18+). Bachelor level respondents, more than PhD ones, considered that improvements are necessary in the primary cycle (5-10 age), while Master and PhD participants believed, more than Bachelor students and graduates, that improvements are needed in the secondary cycle (10-14 age). Master participants displayed quite particular opinions compared to Bachelor level respondents (that improvements are needed in the high-school Curriculum), and to Bachelor and doctoral participants (that improvements are needed in higher education) (Table 10).
Further on, the questionnaire inquired about the agreement extent to the listed claims related to different aspects concerning the improvement of knowledge about natural hazards (Table 11, Fig. 7). According to the results, respondents appreciated that a higher level of climate change awareness may influence society's view of natural hazards (Table 11). It should be highlighted that GSG were supporting this claim (t = 2.003, p = .046) more than TSG. Both GSG and TSG believed that there should be more subjects dealing with risks and hazards in the Curriculum of Geography/Tourism students. Moreover, ANOVA (F = 9.995, p = .000) indicated that Master participants, more than Bachelor and PhD ones, considered that a higher level of climate change awareness may influence society's view of natural hazards.
After the current pandemic crisis, Bachelor and Master students and graduates appreciated that they be more interested in natural hazards, but still there is a statistically significant difference between them (F = 6.137, p = .002) (53.3% of Bachelor and 64.3% of Fig. 6. The Likert plot of the respondents' answers (divided by study levels) regarding the need for an improved Curriculum on natural hazards at different education cycles (A. Primary school (5-10 age); B. Gymnasium (10-14 age); C. High-school (14-18 age); D. Higher education (18+ age)).

Table 10
The agreement level on the necessity to improve the natural hazard-related Curriculum at different study cycles.  Master level respondents indicated high and very high agreement levels to Q16, Fig. 7). PhD level participants, more than Bachelor and Master ones, reported that they will be more interested in natural hazards. The next question explored the opinion of the respondents concerning the statement that decision-makers in Education will change their perspective regarding the modification of the Curriculum specific to three main topics: anthropogenic and natural risks (Mean = 3.87, Std. 1.018), natural hazards (Mean = 4.059, Std. 0.998), and climate change (Mean = 3.99, Std. 1.018), in the coming years. These mean values suggest that the students and graduates agree to this statement, and t-test confirmed statistically significant differences (t = 4.575, p = .000), as the claim is more supported by GSG than by TSG. Regarding this question, there was no statistically significant study level-dependent difference.

Discussion
This section addresses the research questions proposed in the Introduction, starting with an overview of the survey's results that deal with RQ1-3, and proceeding with a larger perspective of the answers specific to RQ4 (i.e., Are there differences between the participants of various education levels and geographic specializations in terms of knowledge about and attitudes towards risks?) and RQ5 (i.e., Are there differences in the perception of young geographers/Tourism students and graduates about the impact of different natural hazards on social and economic activities?). These findings complement the existing scientific literature, contributing to the robustness of the emerging research that focuses on the links between the Covid-19 pandemic and other hazards, as well as on their impact on students and human communities in general.
This study is the first to address the multi-hazard perception of students enrolled in/recent graduates of hazard-related education programs. Therefore, linking the presented findings with previous ones of similar nature is limited to the comparison regarding the impact of the Covid-19 pandemic on the training/career of Tourism students [74][75][76][77][78][79][80][81], and the multi-hazard perception of rural communities in West Sumatra, Indonesia [86], of the general population in Sweden and Italy [84]. Due to the dynamic nature of the Covid-19 pandemic and its local scale particularities, caution is advised when comparing the current results with the ones of the aforementioned studies, as the research specific to each of them was performed during different stages of the Covid-19 outbreak.

Overview on the survey's results
Keeping in mind the first research question of this study (i.e., RQ1: Do the educational systems in Romania and Serbia provide sufficient knowledge about the manifestation, assessment, and management of different risks?), the results point out a relatively high awareness regarding the natural hazards and the associated risks among the participants, especially in the case of the hazards that are more frequent in Romania and Serbia. This is in line with the results of other related studies of risk perception among young people [133,134]. The answers gathered for Q1, Q7, Q8, and Q11 (Figs. 1, 3 and 5) point out that the participants possess a medium level of knowledge and awareness concerning the impact of the Covid-19 crisis and of the other natural hazards that co-occurred in the last two years, both on socio-economic activities, and the participants' training/career. Nevertheless, the interpretation of these results must take into account the fact that the questionnaire was applied in February-April 2022, when many uncertainties related to the Covid-19 pandemic had already been removed (compared to the year 2020). The moderate to a high level of risk perception is specific not only to the types of hazards that can directly affect these territories (e.g., earthquakes, floods), but also in the case of other categories of hazards that, due to their slow manifestation process may appear less significant, and mainly affect rural communities (e.g., droughts), from which a significant part of the participants come from. The sharper perception of the hazards which are more frequent in the area of origin of the respondents is confirmed by (i) [84] who found out that people in Italy considered that droughts and floods are more likely to occur and more impactful, particularly because these hazards are frequent in Italy, and by (ii) [86] who state that rural communities in West Sumatra selected droughts, the Covid-19 pandemic and earthquakes as the top three hazards in terms of worry, perceived impact and the likelihood of occurrence.
When it comes to the student's perception regarding the extent to which the Covid-19 pandemic has amplified the impact of other risks (i.e., RQ2), economic risks are the ones who are believed to be the most impacted. These are followed by technological risks and environmental risks. The participants considered that natural hazard-induced risks were the least amplified by the Covid-19 pandemic, which shows a poor understanding of the interactions between the pandemic crisis and other natural hazards, which may come from the novel character (therefore less known) of multi-hazard research. The scientific literature includes numerous examples that illustrate that the strategies that helped to reduce the SARS-Cov-2 infection rate made the management of natural hazardous events (or even disasters) more demanding in terms of human and material resources [3][4][5]135]. The reality that stands behind the multi-hazard risk term is cogently expressed by Ref. [3]: "COVID-19 has the potential to significantly impact the exposure, vulnerability and response elements associated with natural disasters and vice-versa", but escapes the understanding of many respondents in this study; perhaps due to the fact that their cognition focuses on the different origin of the concurring hazards at hand (i.e., the Covid-19 pandemic and other natural hazards). Moreover, the fact that the participants estimated that the vulnerability of human communities to different types of hazards has been amplified by the pandemic crisis to a low to moderate extent proves that their understanding of the interactions between concurrent hazards needs to be deepened.
Addressing RQ3 (i.e., Do the students and graduates consider that the Curriculum on natural hazards should be improved in various stages of the educational process?), the results provide optimistic perspectives. Most of the participants in this study acknowledged the need to upgrade the hazard and risk-related information provided by educational programs, especially in the higher tiers of the educational system (i.e., higher education and high-school). Similar findings emphasize the importance of learning about other global themes, such as climate change or sustainable development, especially in university [136][137][138][139][140].
The concentration of hazard-related education programs at higher levels and in specialized clusters of faculties is detrimental to the overall understanding of the concepts of risk, hazard, vulnerability, resilience, and also to the awareness of the importance of risk preparedness and mitigation [141]. [142] show that students who have participated in hazard-related education programs are more likely to implement household preparedness measures and to be knowledgeable about safe behaviours. This is also supported by the findings concerning the enhanced knowledge about earthquakes of the people who participated in organized evacuation exercises, and the better survival skills of the people who were provided with official information about earthquakes [42]. Thus, educating the population about the manifestation of natural hazards and about the preventive measures that ensure safety and minimise damage is a prerequisite of the safety culture, and it is needed at all levels of education. Generally speaking, this type of education is not restricted to institutionalized forms, but is also transferred from students to parents and other members of their social networks [58,143]. A sharp risk perception and adequate knowledge about natural hazards and risks are of primary importance, especially since the participants of this study are future specialists in risk management, natural hazard mitigation, tourism reconstruction or even tomorrow's Geography teachers; in other words, the professionals who can help to reach the Sustainable Development Goals (i.e., 4. Quality education, 10. Reduced inequalities, 11. Sustainable cities and communities, 17. Partnership for the goals [73]).

The role of specialization (Tourism vs. Geography) in shaping the multi-hazard risk perception and the educational insights
The participants in this study are young people ages between 18 and 34 years old, oriented towards understanding and inquiring about the spatial and temporal social and natural realities the world we live in, and interested in travelling experiences. The analysis confirms that the Curriculum influences the students' and graduates' level of knowledge on various natural hazards and the associated risks (Fig. 1). On the other hand, their multi-hazard risk perception is not necessarily concordant to their self-reported knowledge levels.
Generally, GSG direct their attention to the spatial distribution of natural and social phenomena, while TSG focus on economic and entrepreneurial aspects, which is expected. Fig. 1 show these differences, which become relevant in the case of technological risks, about which Geographers are more knowledgeable, and of the Covid-19-induced risks, which are better known by TSG. More-over, the fact that TSG estimated higher impact levels of natural hazards on Tourism than GSG show that these perspectives are correlated with the potential threats to the future activity of participants of different specializations, and to the current activity of the graduates. Similar anxious perspectives on future Tourism careers were signalled by Refs. [74,80,81].
Certain statistically significant differences were found concerning the impact of various natural hazards (Table 4) at different spatial scales between the two categories of participants, and also regarding the potential impact of all categories of natural hazards on Tourism. It is worth mentioning that in the case of epidemiological hazards, the participants indicated very high impact levels ( Table  4), regardless of their specialization. It appears that TSG tend to estimate higher impacts of geophysical and hydrological hazards at the national and continental scale, and of geophysical, hydrological, atmospheric and biophysical hazards at the global scale, than GSG. This shows that their understanding of the multi-scalar character of natural hazards needs to be improved.
Counterintuitive results emerge from Q7 and Q8, as GSG overlooked the fact that the vulnerability of human communities to natural hazard-induced risks has been significantly amplified by the Covid-19 pandemic crisis; a fact which was not missed by Tourism participants in the case of geophysical and biophysical hazards. On the other hand, TSG underestimated the increase of vulnerability to economic risk as a result of the pandemic, which was not missed by GSG. These findings contradict the educational background of the two categories of respondents: Geography participants should have a deeper understanding of the interactions between concurring or cascading hazards, and TSG should be more focused on the economic impact of the multi-hazard events. This outcome is not explained by the dominance of Bachelor students (79.9%) in the sample, who still have to acquire hazard-related knowledge, because ANOVA showed that Bachelor respondents, more than PhD ones, considered that the pandemic has amplified natural hazard-induced risks.
When it comes to the improvement of hazard-related education, some statistically significant differences appeared between respondents with different specializations. GSG agreed to a larger extent that climate change will influence society's view on natural hazards, and that the Education decision-makers will change their perspective regarding the upgrading the Curriculum to better integrate knowledge on hazards and risks. In addition, they considered that the pandemic has a higher impact on education than TSG did. These findings are concordant with their supposed higher awareness of these topics, which is based on more hazard-focused educational programs, as well as with the hopes specific to potential Geography teachers; a career some of these students are known to be interested in.

The role of the study level (Bachelor, Master and PhD) in shaping multi-hazard perception and the educational insights
The constant accumulation of knowledge, the sharpening of awareness and the ever-increasing probability of accumulating personal experiences are highlighted by differences found at the three study levels. Certain topics integrated in the questionnaire were differently perceived by Bachelor, Master and PhD students and graduates, as shown by several statistically significant differences found via ANOVA. These differences become even more representative as the average age of the participants, which correspond to the three levels of study, ranges between 20 and 30 years old (Bachelor: 20.6 years; Master: 25.1 years; PhD: 28.3 years, Table 1).
Master and PhD students or graduates tend to base their knowledge about natural hazards on university education, while the younger Bachelor participants rely more on the information gathered during pre-university education stages (Fig. 2). In addition, the horizon of personal experiences grows larger in time, which means that doctoral students approach the concepts of hazard and risk more analytically, and with a higher awareness of the hazardous events they encounter or study about. A decreasing trend among the three study cycles is also noticeable in the case of knowledge accumulation from mass-media and social-media, as a direct consequence of professional emancipation and increasing interest in and access to the scientific literature. It is the period in which natural hazards and their effects are studied and thought about more and more often, following phenomenology and cartography-based approaches, with less and less interest in media outlets that focus more on the aspects of novelty, emotional, and momentary impact [144][145][146][147], or even present erroneous information [148].
These configurations of the knowledge sources may also motivate the study level-dependent differences concerning the estimation of impact levels specific to certain natural hazards. Bachelor students considered that biophysical hazards are more impactful at global scale, and that their training/career may be endangered by future biophysical hazard to a comparable extent, which may be a consequence of higher mass media and social media exposure. On the contrary, PhD level participants adequately assessed that usually these hazards are more impactful at local and regional scale. Nonetheless, Bachelor students correctly estimated that the Covid-19 pandemic has significantly amplified the risks associated with natural hazards; an interaction that was overlooked by the respondents from superior education levels.
The impact of the Covid-19 pandemic on the training/career of the participants was perceived differently from one study cycle to the other: Bachelor and Master level participants were more affected by the pandemic than respondents from the highest tire of university education. This is directly related to the transition from physical classes to the online system, which was more impactful at the Bachelor and Master levels that base their activity on many mandatory courses, laboratory work and practical activities. The better part of the studies that focus on the Covid-19 pandemic on the training/career of Tourism/Geography students [87] involve only Bachelor level participants, which means that the comparison of the current findings to previous ones is not possible. However, the TSG from the two universities in Romania and Serbia estimate that the impact of the pandemic crisis on their training/career was of moderate to high level (mean value 3.62), which is concordant with the balanced perspectives of the students following Tourism specializations in the Czech Republic and Slovakia [77], and India [76]. In contrast, the literature includes several examples of more pessimistic opinions about the impact of the Covid-19 pandemic on Tourism education [80,81].
Regarding the necessity of improving the Curriculum on natural hazards and risks, statistically significant differences were observed among the students and graduates from the three levels of study. Master level participants seem to have the stronger opinions on the cycles of study that should benefit from the upgrading of the hazard and risk-related Curriculum, perhaps due to the fact that they are the middle link between the fresh, inexperienced perspective of Bachelor students and the more mature vision of doctoral students.

Limitations of the study and further research directions
Although the present study provides new insights related to GSG and TSG perception of multi-hazards risks in a period when society faces one of the most impactful pandemics, some limitations arise. Firstly, this study does not account for the influence of classic socio-demographic and personality predictors of risk perception, such as gender or age, neither does it discuss the differences that may originate from cultural and nationality-related factors [40,41,85]. A design-related objective limitation is the dominance of Bachelor students and graduates (79.9%) in the sample. This was an unavoidable outcome, as the largest number of students in the two selected universities follows first-tier higher education programmes. Proceeding with the design weaknesses of the survey, in the endeavour to obtain a larger picture of the multi-hazard risk perception, some questions were formulated in a general, less focused manner. This may have created some problems during the stage of filling out the questionnaire (e.g., grouping all types of hazards into five classes, without providing details on the classification criteria). Also, special attention could have been dedicated to the factors that influenced the impact level of the Covid-19 pandemic on the training/career of the respondents.
Nevertheless, the presented findings have great social and scientific value, especially due to the fact that the participants' knowledge on risk perception does not only echo to their personal lives, but has greater significance since these students and graduates may have future decision-making, hazard-related research or teaching jobs. Therefore, understanding their multi-hazard risk perception during their training years and at the start of their careers offers important cues about tomorrow's hazard management, and tourism reconstruction practices.
Further research endeavours may be tailored focusing on the perception of students of a large spectrum of environment and hazard-related specializations (e.g., Ecology, Biology, Environmental engineering, Economy, Medicine, Water engineering, Agriculture, Silviculture or Civil protection), among which climate change associated ones are most acknowledged. Looking at a research landscape where interdisciplinarity and transdisciplinarity are more and more appreciated, we consider that a comparative approach that confronts the opinions of participants with different, but complementary educational backgrounds and experiences, perhaps also from different countries, offer further-reaching results than single scientific field-focused approaches.

Conclusions
This study explores the perception of GSG and TSG of "Alexandru Ioan Cuza" University of Iasi, Romania and Novi Sad University, Serbia, on natural multi-hazard risks in the times of the Covid-19 pandemic, as well as their impact on different socio-economic activities, using a multi-scalar approach. The main objective of this research work was to identify the specialization and study leveldependent differences in multi-hazard perception and hazard-related education improvements. Secondarily, the study aimed to examine the differences concerning the impact of the Covid-19 on the respondents' training/career, also inquiring about their opinion if future natural hazards may affect these areas of life to a comparable extent.
The descriptive analysis indicated that the respondents self-reported medium to high general hazard and risk-related knowledge levels, as also displaying high awareness levels of the necessity to improve the hazard-related education programs, especially in higher education and the high-school cycle. Subsequently, the differential analysis performed via the t-test indicated several statistically significant differences between the participants of different specializations that may be summarized as follows.
• TSG proved to have a sharper perception of certain natural hazards at national and continental scale, as well as at global scale; but GSG seemed to provide more on point scalar estimations regarding these matters.
• GSG failed to see the amplification effect of the Covid-19 pandemic crisis on the vulnerability to geophysical and biophysical hazards; which was not missed by TSG. On the other hand, the latter category considered that the vulnerability of human communities to economic risks was amplified by the pandemic to a lesser extent than the one estimated by GSG; which is counter-intuitive to their more economic-oriented educational background.
• TSG considered that Tourism was negatively affected by different hazard types to a larger extent than the one estimated by GSG.
• The participants that are following or recently graduated from Geography programs seem to have stronger opinions concerning the positive effects of hazard-related education than TSG.
There were no statistically significant differences between the impact of the Covid-19 pandemic on the training/career of GSG and TSG. However, study level-dependent differences were identified: the participants from lower study levels reported that they were affected by this epidemiological hazard to a larger extent than doctoral students and recent PhDs. Other study level-dependent statistically significant differences that were pointed out by ANOVA concern the scale of the impact specific to biophysical hazards, the amplification effect of the pandemic on different hazard and vulnerability types, and the different education cycles that the Curriculum upgrade should be performed at.
These findings may be considered starting points for future research projects dealing with multi-hazard risk perception amid the Covid-19 pandemic, which are particularly valuable considering that the respondents of the survey are potential specialists in hazard management, risk mitigation and tourism activities, who can significantly contribute to the race towards Sustainable Development Goals. The counter-intuitive results provided by the differential analysis may highlight important challenges for future risk management practices and tourism reconstruction endeavours in Romania and Serbia; which have to be taken with a grain of salt as the survey is limited only to one higher education institution in each country. Nonetheless, it is only by understanding the perception of to-day's students and early career specialists on multi-hazard risks and risk-related education, that we can hope for the improvement of multi-hazard management and for resilience-oriented socio-economic development paths.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
The data that has been used is confidential.