Research Article A Comparative Analysis of Climate Change Risk Response Perception Paths between Northern and Southern Shaanxi Siwen Xue1,3,*, , Zhou Qi1,2,3 1School of Geography and Environment, Baoji University of Arts and Sciences, Baoji 721013, China 2Shaanxi Key Laboratory of Disasters Monitoring and Mechanism Simulation, Baoji University of Arts and Sciences, Baoji 721013, China 3Shaan’xi Provincial Key Research Center for Socialism with Chinese Characteristics (Baoji Base), Baoji 721013, China 1. INTRODUCTION The fifth report of the IPCC pointed out that extreme weather and climate events have changed since 1950, and extreme climates have also occurred frequently [1]. In the context of global climate change [2], meteorological disasters occur frequently and the risks of cli- mate change are increasing. It is thus necessary to step up efforts to address the risks of climate change. Currently, there are three main ways to deal with climate change risks: mitigation, adaptation and avoidance [3]. The adaptation challenge grows with the magni- tude and the rate of climate change. Even the most effective climate change mitigation through reduction of Greenhouse Gas emissions or enhanced removal of these gases from the atmosphere (through carbon sinks) would not prevent further climate change impacts [4], making the need for adaptation unavoidable [5]. Climate change mitigation consists of actions to limit the magnitude or rate of global warming and its related effects [6]. The main challenge is move away from coal, oil and gas and replace these fossil fuels with clean energy sources [7]. As for avoiding dangerous climate change, a study published in 2018 points at a threshold at which tempera- tures could rise to 4° or 5° through self-reinforcing feedbacks in the climate system, suggesting it is below the 2° temperature target [8]. Therefore, different climate change risk response methods have certain challenges and shortcomings. And in the interaction between people and the environment, the perception of environ- ment is the main basis for human decision-making behavior [9]. Therefore, it is necessary to explore the formation mechanism of people’s climate change risk response perception, so as to overcome the difficulties and shortcomings in climate change risk response. Most scholars believed that behaviors influencing people’s response to climate change risks are diverse. In many instances, there are many factors that cam enhance people’s ability to cope with cli- mate change. These factors can include resources, education and information, gender, poverty, wealth, infrastructure, institutional efficiency as well as local indigenous practices, knowledge, and experiences [10,11]. Therefore, factors that influence climate change risk response are diverse. Owing to the interaction between behavior and perception. It is believed that the factors that impact climate change risk response perception are also diverse. This shows that structural equation model is suitable to deal with multiple fac- tors affecting climate change risk response perception simultane- ously. In this regard, some scholars have done extensive research. Momtaz et al. investigated the factors affecting perception and adaptation behavior of farmers in response to climatic changes in Hamedan. The findings indicated that knowledge, perception, and A RT I C L E I N F O Article History Received 23 December 2020 Accepted 11 March 2021 Keywords Northern Shaanxi southern Shaanxi climate change risk perception structural equation modeling A B S T R AC T The public’s awareness of climate change risks is the basis for their choice of adaptation action. A good understanding of the key factors that affect the public’s perception of climate change risk is critical to climate change risk management. In this paper, a path model was constructed to analyze the path of climate change risk response perception in northern Shaanxi based on 1660 public survey data in northern Shaanxi, which was compared with that of southern Shaanxi. The results showed that (1) there are three causal paths in northern Shaanxi, that is, the public’s awareness of climate change issues, awareness of ecological stability, and awareness of climate change causes, to affect response status; there are nine causal paths in southern Shaanxi. (2) There are four related routes in northern Shaanxi and 19 in southern Shaanxi. In short, compared with southern Shaanxi, there are fewer perception paths and simpler models for climate change risk response in northern Shaanxi. (3) The degree of concern for climate change issues and the perception of the causes of climate change influence the establishment of the causal path of climate change risk perception in northern Shaanxi. The major factors that influence climate change risk response perception in southern Shaanxi are climate change risk reason perception, industrial structure adjustment perception, and energy conservation, and emission reduction perception. (4) The response perception path in northern Shaanxi is simpler than that in southern Shaanxi, and there are fewer causal and related paths that impact climate change risk response perception. (5) Finally, through the comparative analysis of the path of climate change risk response perception in northern Shaanxi and southern Shaanxi, this paper provides a reference for coping with climate change risks in northern and southern Shaanxi. © 2021 The Authors. Published by Atlantis Press B.V. This is an open access article distributed under the CC BY-NC 4.0 license (http://creativecommons.org/licenses/by-nc/4.0/). *Corresponding author. Email: 1213268775@qq.com Journal of Risk Analysis and Crisis Response Vol. 11(1); April (2021), pp. 26–35 DOI: https://doi.org/10.2991/jracr.k.210312.001; ISSN 2210-8491; eISSN 2210-8505 https://www.atlantis-press.com/journals/jracr http://orcid.org/0000-0002-0809-5185 http://creativecommons.org/licenses/by-nc/4.0/ mailto:1213268775%40qq.com?subject= https://doi.org/10.2991/jracr.k.210312.001 https://www.atlantis-press.com/journals/jracr S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 27 belief had the maximum impact on the adaptation behavior, with path coefficients of, respectively, 0.53, 0.32, and 0.18, whereas belief and knowledge had the maximum impact on perception, with path coefficients of 0.56 and 0.35 respectively [12]. Xue et al. [13] pointed out In exploring new ecological paradigms and coping with climate change in China, highly educated respondents showed a signifi- cantly stronger path between risk perception and behavior than less educated respondents. Eriksson examined appraisals of threat (cognitive and emotional), personal resources (cost and self-effi- cacy), and strategies (response-efficacy) as predictors of proactive management responses (past behavior and future intention) among forest owners in Sweden by means of a questionnaire (n = 1482), and found that threat appraisals and response-efficacy are direct predictors of past risk management behavior and the intention to respond in the future [14]. Brown et al. studied the impact of Cyclone Evan in December 2012 on Fijian households’ risk attitudes and subjective expectations about the likelihood and severity of nat- ural disasters over the next 20 years, and pointed out the main fac- tors that influence the perception of climate change risk response. Their results showed that extreme event substantially changes indi- viduals’ risk perceptions as well as their beliefs about the frequency and magnitude of future shocks [15]. In summary, most scholars believed that education, knowledge, experience and concepts are important in the perception path of climate change risk response. However, few have incorporated environmental and experiences factors into the climate change risk response perception path model at the same time. In fact, some have conducted research on the fac- tors that influence climate change risk response perception from the perspectives of environment or experience. Marlon et al. ana- lyzed a representative statewide survey of Floridians and compared their risk perceptions of 5-year trends in climate change with local weather station data from the 5 years preceding the survey. Their research compared to local experience, risk perceptions of climate change were more strongly predicted by subjective experiences of environmental change, personal beliefs about climate change, and political ideology [16]. Retchless used an interactive map of sea level rise in Sarasota, Florida and an accompanying online survey, it con- siders how college students from nearby and far away from Sarasota, and with different views about climate change, vary in their risk per- ceptions. The results showed that, consistent with spatial optimism bias, risk perceptions increased more from pre- to post-map for respondents far away from Sarasota than for those nearby [17]. Nowadays, although domestic and foreign studies have achieved certain progress in the public’s climate change risk perception and its influencing factors, there are still the following shortcom- ings. First, most of the research subjects focused on investigating the single relationship between environment or experience and response perception, but failed to combine the two to systemati- cally reflect the interaction between various factors and the impact mechanism of climate change risk response perception. Moreover, the research was mostly conducted based on the opinions of peas- ants. Northern Shaanxi and southern Shaanxi are important geo- graphic regions in China, with relatively frequent meteorological disasters. Comparing the research results of northern Shaanxi with southern Shaanxi can further highlight the perception path of cli- mate change risk response in northern Shaanxi, and provide a typ- ical reference case for risk management and response. Therefore, based on the structural equation model, this paper explored the path of climate change risk response perception in northern Shaanxi and conducted a comparative analysis with southern Shaanxi. This paper attempted to find the answers to the following questions: (1) How many paths are there to respond to climate change risk perception in northern Shaanxi and southern Shaanxi? How does it impact on people’s climate change risk response perceptions? (2) What reference can this regular pattern provide for people in northern and southern Shaanxi to deal with the risk of climate change? 2. MATERIALS AND METHODS 2.1. Study Area Northern Shaanxi is located in the northern part of Shaanxi Province, between 107°28¢ and 111°15¢ east longitude, and between 35°21¢ and 39°34¢ north latitude (Figure 1). The loess hilly and gully area of northern Shaanxi is in the middle reaches of the Yellow River and the northern part of the Loess Plateau [18]. It borders Gansu Province and Ningxia Hui Autonomous Region in the west. It is adjacent to the Inner Mongolia Autonomous Region in the north and Fu county, Luochuan and Yichuan counties in Yan’an City in the south, covering 12 counties (districts) including Yuyang District and Dingbian County in Yulin City, and Pagoda District, Ansai County, Zichang County, Yanchuan County, Yanchang County, Ganquan County, Zhidan County, and Wuqi County in Yan’an [19]. Northern Shaanxi consists of two regions, Yan’an and Yulin. The former is a typical dry farming area, and the latter belongs to the agro-pastoral zone in the northern area of China. There are many meteorological disasters in the whole northern Shaanxi region. Drought, frost, rainstorm, gale, hail of varying degrees occur almost every year, among which drought, hail and frost are particularly serious [20]. The south of Shaanxi is close to the Qinling Mountains in the north, and the Bashan Mountain in the south, with Han River flowing from its west to east. The natural conditions of Hanzhong and Ankang in southern Shaanxi have typical characteristics of the southern region. They are located at 105°30¢–110°01¢E and between 31°42¢ and 34°24¢N, as shown in Figure 1. They have a humid climate in the northern subtropical zone, and most of mountains have a warm temperate humid climate. The shallow valleys in southern Shaanxi are the warmest areas in the province, with temperatures mostly Figure 1 | Topography and geomorphology of the study area. 28 S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 ranging from 14 to 15°C. The average temperature in January, the coldest month, is 0–3°C, and the average temperature of July, the hottest month, is 24–27.5°C. The annual precipitation is 700–900 mm. There are many flood disasters in southern Shaanxi, and the rainy season is in the autumn, which generally lasts from early and late to mid-early September. The main meteorological disasters there are summer drought, heavy rain, continuous rain, hail, frost, strong wind, cold wave etc. [20]. 2.2. Data Sources The questionnaire data came from a random sampling of public in northern Shaanxi. A total of 1660 valid questionnaires were received, and the response rate was 80%. Among the respondents, 829 were male, accounting for 49% of the total, and 831 were female, accounting for 51% 412 were at the age of 20 or below, accounting for 24.8% of the total; 629 aged 21–30, accounting for 37.9%, and 248 aged 31–40, accounting for 14.9% [21]. There were 186 respondents aged 41–50, accounting for 11.2% [21], 168 aged 51–70, accounting for 10.1%, and 24 aged over 70, accounting for 1.4%. In this survey, the data of the Shaanxi Provincial. Statistical Yearbook (2018) were used in the design of the popu- lation structure of the respondents, and appropriate adjustments were made based on the status of Yan’an and Yulin and large sample requirements. It is for us to consider the representativeness and validity of the sample as much as possible. Table 1 shows other basic characteristics of the surveyed public [22]. The correlation coefficients between the perception of environ- ment beauty and the living environment and risk concepts in northern Shaanxi are 0.435 and 0.238 respectively, which are both significant at the level of 0.01. The correlation coefficient between risk perception and perception of environmental stability is 0.174. The correlation coefficients of the degree of concern for climate change issues with the perception of response situations and the perception of climate change causes are 0.245 and 0.149 respec- tively (Table 2), both significant at the level of 0.01. Therefore, the questionnaire indicators selected in northern Shaanxi have a rela- tively significant correlation, which indicates that the public’s per- ception of climate change risks and response paths, and the content validity is high [24]. 2.3. Research Methods 2.3.1. Construction of structural equation model Based on the field survey in northern Shaanxi and the analysis of the validity of the questionnaire [25], this paper proposes the following hypotheses, and constructs a path model of the role of risk concepts, living environment, and climate change information mastery on public climate change risk perception (Figure 2). Hypothesis H1: The public’s perception of climate change issues, perception of environmental stability, and perception of the causes of climate change affect the response status [26]. Hypothesis H2: Risk perceptions are positively correlated with the perception of living environment and environment beauty perception. Hypothesis H3: The living environment and the perception of environment beauty perception are positively correlated. Hypothesis H4: The degree of concern for climate change issues is positively correlated with the perception of the causes of climate change [27] (Table 3). Table 1 | Basic characteristics of the surveyed public [22] Survey item Category Frequency Ratio (%) Education Elementary school or below [23] 408 29.30 Junior high school [23] 284 17.10 High school [23]/Technical secondary school 60 3.60 Undergraduate/Junior college [23] 278 16.70 Postgraduate and above 553 33.30 Monthly income 500 and below 870 52.40 500–1000 292 17.50 1001–2000 205 12.30 2001–3000 179 10.70 3001–5000 114 6.90 Profession Agriculture, forestry, animal husbandry and fishery 229 13.70 Production and transportation 53 3.1 Business services 177 10.1 Government institutions 28 1.6 Expertise 173 10.4 Doctors 135 8.5 Teachers 557 33.5 Soldiers 132 7.9 Self-employed people 99 5.9 Students 33 1.9 Table 2 | KMO value and Bartlett test in northern Shaanxi Kaiser– Meyer–Olkin measures sampling suitability KMO value and Bartlett test in northern Shaanxi Kaiser–Meyer–Olkin measures sampling suitability Bartlett’s sphere test 0.678 Approximately chi-square 1272.646 df 406 Significance 0.000 Figure 2 | The impact mechanism model of public climate change risk perception in northern Shaanxi (hypothetical model). S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 29 2.3.2. Variable selection and descriptive statistics Table 4 is the descriptive statistics of independent variables and dependent variables of structural equation model. In addition, it also includes specific questionnaire items corresponding to different indicators. 2.3.3. Path analysis In order to identify the key factors that affect the perception of cli- mate change risk in southern Shaanxi and the path of these factors, this paper uses path analysis to construct a path map and calculates the effect value (including overall effect, direct effect and indirect effect) in the AMOS26.0 environment [28]. In the structural equa- tion model, the structural model between latent variables with only one observation variable is called path analysis. It is used to test the accuracy and reliability of the hypothetical causal model, the strength of the causal relationship between the measured variables, and it can accommodate the multi-link causal structure and use a path diagram to express it [29]. The basic expression is: h h x z= + +B Γ where x is the exogenous variable matrix [30], h is the endoge- nous variable matrix [30], B is the structural coefficient matrix that represents the influence between the constituent factors of the endogenous variable matrix h, Γ is the structural coefficient matrix [31], which represents the influence of the exogenous variable matrix x on the endogenous variable matrix h [31], and z is the residual matrix which represents the unexplained part [31]. 3. RESULT ANALYSIS 3.1. Model Fit Test In AMOS 26.0 environment, path model framework is established and calculated, original path is debugged according to model correction prompts, and the final model of northern Shaanxi is determined (Figure 3). When response path model freedom degree in northern Shaanxi is 9, its Chi-square value is about 9.312. The corresponding significance Figure 3 | The impact mechanism model of public climate change risk perception in northern Shaanxi (standard model). Table 3 | Correlation coefficient matrix of climate change risk perception in northern Shaanxi Index Understanding the reasons of climate change Coping situation Scenic beauty perception Environmental stability awareness Living environment −0.035 0.07 0.435** 0.062 Risk concept 0.01 0.075 0.238** 0.174** Concern about climate change 0.149** 0.245** 0.069 0.076 **represents significant at the 0.01 level. Table 4 | Description of explanatory variables in northern Shaanxi Variables Measurement standard Assignment Mean Standard deviation Living environment Regional climate comfort C72 Strongly agree = 1; agree = 2; uncertain = 3; disagree = 4 2.98 1.078 Severe surrounding pollution C73 strongly disagree = 5; strongly agree = 1; agree = 2; uncertain = 3; disagree = 4; strongly disagree = 5 2.694 1.16 Regional environmental livability C74 Strongly agree = 1; agree = 2; uncertain = 3; disagree = 4; strongly disagree = 5 2.665 0.983 Risk concept Risk perception B1 Strongly agree = 1; agree = 2; uncertain = 3; disagree = 4; strongly disagree = 5 3.2 1.86 Risk option B3 There is 80% chance of getting 4000 yuan, 20% chance of getting nothing = 1, 100% chance of getting 3000 yuan = 2 1.611 0.569 Understanding the causes of climate change Evaluation of causes of climate change C91 Natural reasons humanistic reasons = 1–7 4.925 2.073 Understanding the causes of climate change C81, C82, C83 Very well understanding = 1; relatively understanding = 2; general = 3; not very understanding = 4; not at all = 6 2.291 0.861 Concern about climate change issues Degree of concern for climate change issues D1 Very concerned = 1; more concerned = 2; general = 3; not very concerned = 4; very unconcerned = 5 2.134 0.876 Coping situation awareness Climate change event participation status D6 Very willing = 1; more willing = 2; unclear = 3; reluctant = 4; very unwilling = 5 1.958 0.967 Daily coping behavior D7 Always = 1; sometimes = 2; not sure = 3; rarely = 4; never = 5 1.99 1.042 Scenic beauty perception Scenic beauty recognition C71 Strongly agree = 1; agree = 2; unsure = 3; disagree = 4; strongly disagree = 6 3.112 1.142 Environmental stability awareness Environmental stability awareness B2 The natural world is fragile, even a small change can cause catastrophic consequences = 2 2.982 0.919 The natural world is very stable, even if it is greatly disturbed, it can be restored to its original state = 4 30 S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 probability p = 0.811 > 0.05, which does not reach the significance level of 0.05. In addition, the ratio of chi-square freedom degree (CMIN/DF) is 0.665 < 2; RMSEA value is 0.000 < 0.050; the GFI, AGFI, IFI, TLI, CFI values are 0.993, 0.985, 1.050, 1.081, and 1.000 respectively, all of which are over 0.900, complying with the stan- dard. The preset model’s AIC, BCC, BIC, CAIC, ECVI values are all smaller than those of independent model and saturation model, indi- cating that the hypothetical model fits well with actual data (Table 5). 3.2. Analysis of Results in Northern Shaanxi Test results show that the overall effect of climate change reason perception, climate change problems concern degree, public’s envi- ronmental stability perception and public response status percep- tion is 0.217, 0.200, and −0.18 respectively. Furthermore, the direct effects are 0.217, 0.200, and −0.18, respectively. The direct effects of the degree of concern for climate issues and the perception of the causes of climate change on the situation are significant at the 0.01 level (Figure 3). This shows that the environmental stability per- ception, climate change reasons perception, and concern degree for climate change issues have a significant positive impact on climate change response perception [32]. It is believed that hypothesis of H1 is valid. In contrast, climate change issues concern degree has a greater impact than the above two (Table 6) [33]. As for correlation path in northern Shaanxi, risk concern is posi- tively correlated with living environment and environment beauty perception, with covariances of 0.137 and 0.203, respectively, assuming H2 holds. Among them, the covariance of risk concepts and living environment, beautiful scenery perception is significant at the level of 0.01, which is inferred to be related to the fragile geographical environment in northern Shaanxi. Further covari- ance analysis of living environment and scenic beauty perception is 0.362, among which relationship with scenic beauty perception is significant at the level of 0.01, assuming H3 holds. Moreover, the covariance between concern degree of climate change issues and perception of climate change reasons is 0.121, significant at the level of 0.05. Therefore, H4 is confirmed. This shows that the better the living environment in northern Shaanxi, the stronger risk con- cept and environmental beauty perception. The higher the concern degree of climate change issues, the better the perception of climate change reasons [34] (Table 7). 3.3. Comparative Analysis The path model of public climate change risk response perception in northern Shaanxi was constructed based on risk concepts, living environment, and concern for climate change issues. The climate change risk response path model in southern Shaanxi was con- structed based on risk concepts, human and land concepts, cultural level, living environment, and concern degree for climate change issues, and they have all passed test. It is inferred that in north- ern Shaanxi region, due to the relatively harsh environment, con- servative ideological concepts, serious soil erosion, and frequent disasters, education degree has a smaller impact on climate change risk response perception [35]. Instead, concern degree for climate change issues and climate change reason perception influence the causal path of climate change risk perception [36]. In south- ern Shaanxi, the mountains and rivers are beautiful, so it is less hit by natural disasters. Therefore, climate change result perception, human and land concepts, risk concepts, educational level, and concern degree for climate change issues impact the establishment of climate change risk perception’s causal path in southern Shaanxi. In addition, northern Shaanxi is dominated by the secondary industry, whereas southern Shaanxi is dominated by the primary and tertiary industries (Figure 4). According to research by relevant scholars, the tertiary industry can break through Hu Huanyong line [37], so industrial structure adjust- ment perception in southern Shaanxi has a significant impact on climate change response perception [26]. From Figures 3 and 4, it can Table 5 | Index parameters of model adaptation in northern Shaanxi Evaluation index Preset model Saturation model Independent model CMIN/DF (Relative chi-square) 0.665 5.12 RMSEA 0 0.107 GFI 0.993 1 0.915 AGFI 0.985 0.887 IFI 1.05 1 0 CFI 1 1 0 TLI 1.081 0 AIC 37.312 56 121.511 BCC 37.947 57.269 121.829 BIC 91.795 164.966 148.753 CAIC 105.795 192.966 155.753 ECVI 0.103 0.155 0.337 Table 6 | Overall effect, direct effect, and indirect effect among variables Reason variable Result variable Overall effect Direct effect Indirect effect Climate change reason perception Coping situation perception 0.217 0.217 0 Concern degree for climate change problems 0.200 0.200 0 Environmental stability perception −0.108 −0.108 0 Table 7 | Climate change risk perception covariance matrix Variables Index Estimate SE CR p Living environment← → Risk concept 0.137 0.032 4.319 *** Living environment← → Scenic beauty perception 0.363 0.049 7.424 *** Scenic beauty perception← → Cimate change reason perception 0.121 0.043 2.787 0.005 Risk concept← → Scenic beauty perception 0.203 0.050 4.096 * *, ***represents significant at the 0.05 and 0.001 level. S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 31 Figure 4 | The impact mechanism model of public climate change risk perception in southern Shaanxi (standard model). Table 8 | Linear regression analysis results (n = 24) Constant Non- standardized coefficient Standard error Normalized coefficient T p VIF R2 Adjusted R2 F –0.243 – 0.104 – –2.334 0.030* – 0.864 0.843 (3,20) = 42.217, p = 0.000 Longitude 0.28 0.173 0.171 1.615 0.122 1.643 Latitude 0.813 0.203 0.613 4.002 0.001** 3.444 Altitude 0.42 0.188 0.314 2.24 0.037* 2.887 Dependent variable: MMS_ganzhi. D-W (Durbin-Watsonstatistic) value: 1.248. *p < 0.05, **p < 0.01. be seen that there are three causal paths in northern Shaanxi: pub- lic’s of climate change issues concern degree, environmental stability perception, and climate change reason perception influence climate change response perception. There are nine causal paths in southern Shaanxi, namely, climate change consequences perception, human and land concept [38], cultural level for climate change issues con- cern degree and industrial structure adjustment perception impact on climate change response status perception; Public human-land and risk concept influence climate change response perception via impact on of climate change reasons perception; human-land and risk concept influence climate change reason perception. As for related routes, there are four in northern Shaanxi and 19 in southern [39]. In short, compared with southern Shaanxi, there are fewer per- ception paths and simpler models of climate change risk response perception in northern Shaanxi. 3.4. Analysis of Influencing Factors In order to explore the factors influencing climate change risk response perception in different counties and regions, and to reveal the mechanism of differences in climate change risk response perception path in northern and southern Shaanxi, lat- itude, longitude and average altitude of each county were taken as independent variables, and climate change risk response per- ception intensity of each county as dependent variable for linear regression analysis. As shown in Table 8, linear regression model R squared is 0.864, indicating a high fitting degree of model. Further analysis of data in Table 8 shows that latitude and altitude are the most influential factors on climate change risk response perception, with regression coefficients of 0.203 and 0.188 respec- tively, significant at the levels of 0.01 and 0.05 respectively. This result demonstrates that the greater the differences in terrain and latitude, the greater the difference of climate change risk response perception intensity, which probably leads to difference in paths (Table 8). 4. DISCUSSION Domestic and foreign studies have also confirmed that the environ- ment and people’s experience will influence perception of climate change risk response [40]. For example, Bradley et al. believed that antecedent psychological and socio-demographic variables predict climate change risk perceptions, which lead to enhancing levels of response efficacy and psychological adaptation in relation to climate change, and ultimately to environmentally-relevant behav- iors [41]. The study found that: Risk perception (hot), response (both hot and direct) and psychological adaptation (directly) pre- dicted behavior [41]. Smith provided some ground-breaking work on human behavior as it relates to perception and response to risks associated with climate change and climatic variability in the rural communities of Sandy Bay and Fancy. The study examined house- holds’ knowledge and perception of the climate change phenom- enon and their responses to climate-related events. The results showed that an investigation of responses or the decision to respond to some of the impacts that they have experienced as a result of climate change and climatic variability leads to the development of different types of perceptions, including religious, ill informed, experienced-based, and knowledge-based perceptions. It is argued here that these forms of perception may result in non-adaptive, pro- active or reactive adaptive behavior [42]. After studying farmers’ response to and perception of climate change risks, Wang et al. [43] believed that extreme climate changes such as rising temperature, decreased precipitation and increased frequency of drought would affect farmers’ perception and response to climate change. In the hutt valley, New Zealand et al., through a family survey, as well as seminar and interviews with local government officials, found that flood experience can influence flood risk perceptions, and that flood experience can stimulate increased risk reduction and adaptation actions where climate change risks are likely to occur. It is argued here that these forms of perception may result in non- adaptive or reactive adaptive behavior. These studies have confirmed the rationality of using the two major variables of environment and concept to design the climate change risk response pathway model in northern Shaanxi and southern Shaanxi [44]. To verify the reliability of results of this paper, the climate change risk response perception path model of various cities in northern Shaanxi (Figure 5) and southern Shaanxi (ensure RMSE = 0) is calculated. It is found that climate change risk response percep- tion path of Yulin and Yan’an in northern Shaanxi is much simpler than that in southern Shaanxi (Figure 6). The climate change risk 32 S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 response path in northern Shaanxi has four factors included in the model, while southern Shaanxi has at least five. Finally, geographic detectors are utilized to investigate the factors affecting the per- ception of climate change risk response in northern Shaanxi and southern Shaanxi respectively. It is found that in northern Shaanxi, the explanatory power of each factor on the perception of climate change risk response is: education level > risk concepts > climate change reason percep- tion > living environment > environmental stability perception > industrial structure adjustment perception > scenic beauty percep- tion = climate change problems perception = energy saving and emission reduction perception, as shown in Figure 7). This shows that impact of industrial structure adjustment, energy conserva- tion and emission reduction perception on climate change risk response perception is not much different from that of environ- mental stability and grace perception. Thus, the above factors can be substituted for each other, but they cannot be incorporated into the model of climate change risk response perception in northern Shaanxi. This shows that climate change risk response percep- tion path in northern Shaanxi is not a complete mediation model, which is more consistent with the conclusions drawn by Song and Shi [45]. As for the climate change risk response perception paths in southern Shaanxi, most of them are fully intermediary or partial intermediary models, and there is no non-intermediary model (Figure 6). Figure 7 shows the explanatory power of climate change risk response perception factors from small to large. It can be found that energy conservation, emission reduction percep- tion, and industrial structure adjustment in southern Shaanxi have greater explanatory power to climate change risk response percep- tion than environmental stability or beautiful scenery perception. Therefore, it is believed that energy conservation and emission reduction in southern Shaanxi, climate change reasons, and indus- trial structure adjustment perception are three important interme- diary variables that influence their perception of climate change risk response. The view that climate change risk perception path model in southern Shaanxi is more complicated can be empirically proved by Zhou, who demonstrated that public in Hanzhong area influences their perception and response to climate change risks through their perceptions of reasons, knowledge, facts and conse- quences, which in turn influence their behavior and willingness to climate change risks response [46]. The above discussions indicate that the path of climate change risk response perception in northern Shaanxi is simpler than that in southern Shaanxi, and corresponding influencing factors are also less. The following conclusions can be drawn from the above dis- cussions. First, the main influencing factors of climate change risk response perception in northern Shaanxi [47] are climate change reason perception and climate change issues concern degree. Second, Figure 5 | A perceived path model for climate change risk response of all cities in northern Shaanxi. Figure 6 | A perceived path model for climate change risk response of all cities in southern Shaanxi. S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 33 northern Shaanxi region should increase basic network platforms construction to strengthen publicity of climate change risk informa- tion. Third, regarding complex perception path in southern Shaanxi to deal with climate change risks [48], people’s path of climate change risks response perception is diverse. For this reason, a well understanding of the intermediary variables in climate change risk response perception path model is necessary. Fourth, because indus- trial structure adjustment, climate change reason perception and cli- mate change problems concern degree are important variables in the climate change risk response perception path model, it is necessary to vigorously promote development of tertiary industry in southern Shaanxi and understanding of climate change risk information and reasons. In terms of demographic factors, education level, monthly income and age in northern Shaanxi have greater explanatory power for climate change risk response perception, and can be considered for inclusion in model in the future. In addition to education level in southern Shaanxi, age also has a greater influence on climate change risk response perception. Therefore, it is necessary to strengthen exploration of age on climate change risk response perception to reduce systematic errors induced by the model. The reasons for the difference in climate change risk response perception path in north- ern and southern Shaanxi have been well explained in Subsection 3.4. The facts that vertical difference in topography in southern Shaanxi is more significant than in northern Shaanxi, and that they are located in the southern and northern parts of the Qinling Mountains respec- tively, further confirm that climate change risk response perception in northern Shaanxi is simpler than southern. The contribution of this paper is mainly reflected in the flowing aspects. First, this paper combines the environment and public expe- rience to explore factors influencing the risk perception of climate change. In addition, the public’s perception and experience of risk is divided into two measuring dimensions, which is more innovative than the previous psychology measurement paradigm. Second, this paper, by comparing the two regions of northern Shaanxi and south- ern Shaanxi, provides a more typical case for public climate change risk management. Finally, most scholars tend to study on people’s climate change risk response behavior, whereas this paper directly investigates the path and factors of climate change risk response per- ception [49], with a better design of the research plan. Nevertheless, it should be pointed out that this research has a small problem in the selection of indicators for the perception of climate change risk response. That is, the indicator of the living environment needs fur- ther improvement although it can replace the objective environment where people live. For example, temperature and precipitation can be used to replace the indicator of the living environment. There are less paths in northern Shaanxi than in southern Shaanxi. Previous studies have shown that in the Hanzhong City in southern Shaanxi, age, occupation, education level, income level and public percep- tions of climate change knowledge, facts, and reasons perception, perception of consequences, willingness to respond, and response behavior have varying degrees of influence [50]. Therefore, the paths that affect the perception of climate change risk in southern Shaanxi are diverse. Some scholars analyzed the adaptation behaviors and influencing factors of peasants in the hilly loess regions of north- ern Shaanxi and concluded that peasants’ adaptation behaviors are affected by the perception of climate change (Figure 7). In addition [51], family socioeconomic attributes have a significant impact on the probability of peasants’ adaptation behaviors, while other attri- butes such as age and education level are independent of the proba- bility of farmers adopting adaptive behaviors [52] (Figure 8). Figure 8 | The explanatory power of demographic factors in Shaanxi. Figure 7 | Detection of impact factors in Shaanxi. 34 S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 5. CONCLUSION Based on the research purpose proposed in the introduction part and the results of the discussion part, the following conclusions can be drawn. Firstly, there are three causal paths in northern Shaanxi, that is, the public’s awareness of climate change issues, awareness of ecological stability, and awareness of climate change causes, to affect response status. There are nine causal paths in southern Shaanxi. Secondly, there are four related routes in northern Shaanxi and 19 in southern Shaanxi. In short, compared with southern Shaanxi, there are fewer perception paths and simpler models for climate change risk response in northern Shaanxi. Thirdly, the degree of concern to climate change issues and the perception of climate change causes affect the establishment of the causal path of climate change risk perception in northern Shaanxi. Fourthly, the related paths of climate change risk perception in northern Shaanxi can be summarized into the following two: the better the living environment, the stronger the risk perception of places; the higher the degree of concern for climate change issues, the better the perception of the causes of climate change. Finally, according to the above conclusions, we put forward the following suggestions for northern and southern Shaanxi to deal with the risks of climate change. Northern Shaanxi and southern Shaanxi should be dif- ferent in managing climate change risk. Northern Shaanxi should strengthen advocacy on the causes of climate change, and southern Shaanxi should strengthen publicity on the effects of climate change risk response and increase adjustment of industrial structure, and at the same time, actively carry out energy conservation and emission reduction activities to promote climate change risk response. CONFLICTS OF INTEREST The authors declare they have no conflicts of interest. ACKNOWLEDGMENTS This work is supported by the National Natural Science Foundation of China “Regional Climate Change Risk Perception and Response” (41771215). The authors appreciate the time and effort of the edi- tors and reviewers in providing constructive comments which have helped to improve the manuscript. REFERENCES [1] Shen YP, Wang GY. Key findings and assessment results of IPCC WGI fifth assessment report. J Glaciol Geocryol 2013;35:1068–76. [2] Solomon S, Qin D, Manning M, Marquis M, Averyt KB, Tignor M, et al. Climate change 2007: the physical science basis. Working group I contribution to the fourth assessment report of the IPCC. Cambridge, UK: Cambridge University Press; 2007, pp. 3–53. [3] Grüneis H, Penker M, Höferl KM. The full spectrum of climate change adaptation: testing an analytical framework in Tyrolean mountain agriculture (Austria). Springerplus 2016;5:1848. [4] Fu S, Zou J, Zhang X, Qi Y. Review on the latest conclusions of working group III contribution to the fifth assessment report of the intergovernmental panel on climate change. Chin J Urban Environ Stud 2015;3:1550005. [5] Mackay A. Climate change 2007: impacts, adaptation and vul- nerability. Contribution of working group II to the fourth assess- ment report of the intergovernmental panel on climate change. J Environ Qual 2008;37:2407. [6] Somorin OA, Visseren-Hamakers IJ, Arts B, Tiani AM, Sonwa DJ. Integration through interaction? Synergy between adaptation and mitigation (REDD+) in Cameroon. Environ Plan C: Politics Space 2016;34:415–32. [7] Salvia M, Reckien D, Pietrapertosa F, Eckersley P, Spyridaki NA, Krook-Riekkola A, et al. Will climate mitigation ambitions lead to carbon neutrality? An analysis of the local-level plans of 327 cities in the EU. Renew Sustain Energy Rev 2021; 135:110253. [8] Feng Q, Yang L, Deo RC, AghaKouchak A, Adamowski JF, Stone R, et al. Domino effect of climate change over two millennia in ancient China’s Hexi Corridor. Nat Sustain 2019;2:957–61. [9] Walmsley DJ, Lewis GJ, et al. Introduction to behavioral geogra- phy [Xingzhong W, Guoqiang Z, Guicai L, Trans]. Xi’an: Shaanxi People’s Publishing House; 1988. [10] Nkya K, Mbowe A, Makoi JHJR. Low - cost irrigation technology, in the context of sustainable land management and adaptation to climate change in the Kilimanjaro region. J Environ Earth Sci 2015;5:45–56. [11] Onofri L, Nunes PALD. Economic valuation for policy support in the context of ecosystem-based adaptation to climate change: an indicator, integrated based approach. Heliyon 2020;6:e04650. [12] Mahmoodi Momtaz A, Choobchian S, Farhadian H. Factors affecting farmers’ perception and adaptation behavior in response to climate change in Hamedan province, Iran. J Agric Sci Technol 2020;22:905–17. [13] Xue W, Marks ADG, Hine DW, Phillips WJ, Zhao S. The new ecological paradigm and responses to climate change in China. J Risk Res 2018;21:323–39. [14] Eriksson L. The importance of threat, strategy, and resource appraisals for long-term proactive risk management among forest owners in Sweden. J Risk Res 2016;20:868–86. [15] Brown P, Daigneault AJ, Tjernström E, Zou W. Natural disasters, social protection, and risk perceptions. World Dev 2018;104:310–25. [16] Marlon JR, van der Linden S, Howe PD, Leiserowitz A, Lucia Woo SH, Broad K. Detecting local environmental change: the role of experience in shaping risk judgments about global warming. J Risk Res 2019;22:936–50. [17] Retchless DP. Understanding local sea level rise risk perceptions and the power of maps to change them: the effects of distance and doubt. Environ Behav 2018;50:483–511. [18] Yang L, Xie GD, Zhen L, Leng YF, Guo GM. Spatio-temporal changes of land use in jinghe watershed. Resour Sci 2005;27: 26–32 (in Chinese). [19] Zhao J, Li C, Yang T, Tang Y, Yin Y, Luan X, et al. Estimation of high spatiotemporal resolution actual evapotranspiration by combining the SWH model with the METRIC model. J Hydrol 2020;586:124883. [20] Changyan W, Yaochuang Y. Research on the temporal and spa- tial changes of climate change perception by people in southern Shaanxi. Chin Agric Sci Bull 2015;31:244–51. [21] Orczyk M, Firlik B. The assessment of vibroacoustic comfort in trams on the basis of experimental studies and surveys. In: Suchanek M, editor. New research trends in transport sustainability and innovation. Springer: Springer Proceedings in Business and Economics; 2018, pp. 46–55. https://doi.org/10.1186/s40064-016-3542-1 https://doi.org/10.1186/s40064-016-3542-1 https://doi.org/10.1186/s40064-016-3542-1 https://doi.org/10.1142/S2345748115500050 https://doi.org/10.1142/S2345748115500050 https://doi.org/10.1142/S2345748115500050 https://doi.org/10.1142/S2345748115500050 https://doi.org/10.2134/jeq2008.0015br https://doi.org/10.2134/jeq2008.0015br https://doi.org/10.2134/jeq2008.0015br https://doi.org/10.2134/jeq2008.0015br https://doi.org/10.1177/0263774X16645341 https://doi.org/10.1177/0263774X16645341 https://doi.org/10.1177/0263774X16645341 https://doi.org/10.1177/0263774X16645341 https://doi.org/10.1016/j.rser.2020.110253 https://doi.org/10.1016/j.rser.2020.110253 https://doi.org/10.1016/j.rser.2020.110253 https://doi.org/10.1016/j.rser.2020.110253 https://doi.org/10.1016/j.rser.2020.110253 https://doi.org/10.1038/s41893-019-0397-9 https://doi.org/10.1038/s41893-019-0397-9 https://doi.org/10.1038/s41893-019-0397-9 https://doi.org/10.1016/j.heliyon.2020.e04650 https://doi.org/10.1016/j.heliyon.2020.e04650 https://doi.org/10.1016/j.heliyon.2020.e04650 http://jast.modares.ac.ir/article-23-30786-en.html http://jast.modares.ac.ir/article-23-30786-en.html http://jast.modares.ac.ir/article-23-30786-en.html http://jast.modares.ac.ir/article-23-30786-en.html https://doi.org/10.1080/13669877.2016.1200655 https://doi.org/10.1080/13669877.2016.1200655 https://doi.org/10.1080/13669877.2016.1200655 https://doi.org/10.1080/13669877.2015.1121905 https://doi.org/10.1080/13669877.2015.1121905 https://doi.org/10.1080/13669877.2015.1121905 https://doi.org/10.1016/j.worlddev.2017.12.002 https://doi.org/10.1016/j.worlddev.2017.12.002 https://doi.org/10.1080/13669877.2018.1430051 https://doi.org/10.1080/13669877.2018.1430051 https://doi.org/10.1080/13669877.2018.1430051 https://doi.org/10.1080/13669877.2018.1430051 https://doi.org/10.1177/0013916517709043 https://doi.org/10.1177/0013916517709043 https://doi.org/10.1177/0013916517709043 https://doi.org/10.1016/j.jhydrol.2020.124883 https://doi.org/10.1016/j.jhydrol.2020.124883 https://doi.org/10.1016/j.jhydrol.2020.124883 https://doi.org/10.1016/j.jhydrol.2020.124883 S. Xue and Z. Qi / Journal of Risk Analysis and Crisis Response 11(1) 26–35 35 [22] Temelová J, Novák J, Jíchová J. Safe life in the suburbs? Crime and perceptions of safety in new residential developments in Prague’s hinterland, Czech Republic. Eur Urban Region Stud 2016;23:677–96. [23] Ke X, Zhang L, Li Z, Tang W. Inequality in health service utiliza- tion among migrant and local children: a cross-sectional survey of children aged 0–14 years in Shenzhen, China. BMC Pub Health 2020;20:1668. [24] Damert M, Feng Y, Zhu Q, Baumgartner RJ. Motivating low- carbon initiatives among suppliers: the role of risk and opportu- nity perception. Resour Conserv Recycl 2018;136:276–86. [25] Feng S, Yin J. The study on villa price volatility of Shanghai with structural equation. 2012 International Conference on Management Science & Engineering 19th Annual Conference Proceedings. Dallas, TX, USA: IEEE; 2012, pp. 1819–25. [26] Azócar G, Billi M, Calvo R, Huneeus N, Lagos M, Sapiains R, et al. Climate change perception, vulnerability, and readiness: inter-country variability and emerging patterns in Latin America. J Environ Stud Sci 2021;11:23–36. [27] Joyce RE, Hunt CL. Philippine nurses and the brain drain. Soc Sci Med 1982;16:1223–33. [28] Kim S, Lee JE, Kim D. Searching for the next new energy in energy transition: comparing the impacts of economic incen- tives on local acceptance of fossil fuels, renewable, and nuclear energies. Sustainability 2019;11:2037. [29] Minglong WU. Structural equation model-AMOS operation and application. Chongqing: Chongqing University Press; 2009. [30] Li W, Wei XP, Zhu RX, Guo KQ. Study on factors affecting the agricultural mechanization level in china based on structural equation modeling. Sustainability 2019;11:1–16. [31] Sun H, Fan Z, Zhou Y, Shi Y. Empirical research on competitive- ness factors: analysis of real estate industry of Beijing and Tianjin. Eng Constr Architect Manage 2010;17:240–51. [32] Kwon SA, Kim S, Lee JE. Analyzing the determinants of individ- ual action on climate change by specifying the roles of six values in South Korea. Sustainability 2019;11:1834. [33] Zahran S, Brody SD, Grover H, Vedlitz A. Climate change vulner- ability and policy support. Soc Nat Resour 2006;19:771–89. [34] Lata S, Nunn P. Misperceptions of climate-change risk as barriers to climate-change adaptation: a case study from the Rewa Delta, Fiji. Climatic Change 2012;110:169–86. [35] You M, Ju Y. Interaction of individual framing and political ori- entation in guiding climate change risk perception. J Risk Res 2019;22:865–77. [36] Cutler MJ, Marlon JR, Howe PD, Leiserowitz A. The influence of political ideology and socioeconomic vulnerability on perceived health risks of heat waves in the context of climate change. Weather Climate Soc 2018;10:731–46. [37] Li L, Xueyan Z, Yaru W, Qin Z, Bing X. Research on farmers’ per- ception of climate change in ecologically fragile Alpine region based on structural equation model-a case study of Gannan plateau. Acta Ecol Sin 2017;37:3274–85. [38] Chengchao W, Yusheng Y, Wen Z, Jing H, Jianbin X. Research on the perception and adaptation of domestic and foreign farmers to climate change/variation. Geogr Sci 2017;37:938–43. [39] Yudong Z, Qi Z, Rong H, Kepeng G. Research on the relationship between public perception and response to climate change based on structural equation—a case study of Hanzhong City, Shaanxi Province. Henan Sci 2019;37:429–38. [40] Liu Y. Chapter 6 logistics development in Shaanxi Province. Springer Science and Business Media LLC; 2014. [41] Bradley GL, Babutsidze Z, Chai A, Reser JP. The role of climate change risk perception, response efficacy, and psychological adaptation in pro-environmental behavior: a two nation study. J Environ Psychol 2020;68:101410. [42] Smith RA. Risk perception and adaptive responses to climate change and climatic variability in northeastern St. Vincent. J Environ Stud Sci 2018;8:73–85. [43] Wang J, Kim S. Analysis of the impact of values and perception on climate change skepticism and its implication for public policy. Climate 2018;6:99. [44] Lawrence J, Quade D, Becker J. Integrating the effects of flood experience on risk perception with responses to changing climate risk. Nat Hazards 2014;74:1773–94. [45] Zhen S, Xingmin S. Farmers’ adaptation behaviors to climate change and path analysis of influencing factors in rain-fed agri- cultural areas. Adv Geogr Sci 2020;39:461–73. [46] Yudong Z, Qi Z, Rong H, Kepeng G. Research on the relationship between public perception and response to climate change based on structural equations: taking Hanzhong City, Shaanxi Province as an example. Henan Sci 2019;37:429–38. [47] Becerra MJ, Pimentel MA, De Souza EB, Tovar GI. Geospatiality of climate change perceptions on coastal regions: a systematic bibliometric analysis. Geogr Sustain 2020;1:209–19. [48] Theory and practice of climate adaptation. Springer Science and Business Media LLC; 2018. [49] Elshirbiny H, Abrahamse W. Public risk perception of climate change in Egypt: a mixed methods study of predictors and impli- cations. J Environ Stud Sci 2020;10:242–54. [50] Echavarren JM, Balžekienė A, Telešienė A. Multilevel analysis of climate change risk perception in Europe: natural hazards, political contexts and mediating individual effects. Saf Sci 2019; 120:813–23. [51] Terorotua H, Duvat VKE, Maspataud A, Ouriqua J. Assessing perception of climate change by representatives of public author- ities and designing coastal climate services: lessons learnt from French Polynesia. Front Mar Sci 2020;7:160. [52] Lu W, Xingmin S, Lifan S. The adaptation behavior and influenc- ing factors of farm households in the loess hilly and gully area of northern Shaanxi. Progr Climate Change Res 2017;13:61–8. https://doi.org/10.1177/0969776414534254 https://doi.org/10.1177/0969776414534254 https://doi.org/10.1177/0969776414534254 https://doi.org/10.1177/0969776414534254 https://doi.org/10.1186/s12889-020-09781-4 https://doi.org/10.1186/s12889-020-09781-4 https://doi.org/10.1186/s12889-020-09781-4 https://doi.org/10.1186/s12889-020-09781-4 https://doi.org/10.1016/j.resconrec.2018.05.002 https://doi.org/10.1016/j.resconrec.2018.05.002 https://doi.org/10.1016/j.resconrec.2018.05.002 https://doi.org/10.1109/ICMSE.2012.6414419 https://doi.org/10.1109/ICMSE.2012.6414419 https://doi.org/10.1109/ICMSE.2012.6414419 https://doi.org/10.1109/ICMSE.2012.6414419 https://doi.org/10.1007/s13412-020-00639-0 https://doi.org/10.1007/s13412-020-00639-0 https://doi.org/10.1007/s13412-020-00639-0 https://doi.org/10.1007/s13412-020-00639-0 https://doi.org/10.1016/0277-9536(82)90147-2 https://doi.org/10.1016/0277-9536(82)90147-2 https://doi.org/10.3390/su11072037 https://doi.org/10.3390/su11072037 https://doi.org/10.3390/su11072037 https://doi.org/10.3390/su11072037 https://doi.org/10.3390/su11010051 https://doi.org/10.3390/su11010051 https://doi.org/10.3390/su11010051 https://doi.org/10.1108/09699981011038042 https://doi.org/10.1108/09699981011038042 https://doi.org/10.1108/09699981011038042 https://doi.org/10.3390/su11071834 https://doi.org/10.3390/su11071834 https://doi.org/10.3390/su11071834 https://doi.org/10.1080/08941920600835528 https://doi.org/10.1080/08941920600835528 https://doi.org/10.1007/s10584-011-0062-4 https://doi.org/10.1007/s10584-011-0062-4 https://doi.org/10.1007/s10584-011-0062-4 https://doi.org/10.1080/13669877.2017.1422785 https://doi.org/10.1080/13669877.2017.1422785 https://doi.org/10.1080/13669877.2017.1422785 https://doi.org/10.1175/WCAS-D-17-0105.1 https://doi.org/10.1175/WCAS-D-17-0105.1 https://doi.org/10.1175/WCAS-D-17-0105.1 https://doi.org/10.1175/WCAS-D-17-0105.1 https://doi.org/10.5846/stxb201603090415 https://doi.org/10.5846/stxb201603090415 https://doi.org/10.5846/stxb201603090415 https://doi.org/10.5846/stxb201603090415 https://doi.org/10.1016/j.jenvp.2020.101410 https://doi.org/10.1016/j.jenvp.2020.101410 https://doi.org/10.1016/j.jenvp.2020.101410 https://doi.org/10.1016/j.jenvp.2020.101410 https://doi.org/10.1007/s13412-017-0456-3 https://doi.org/10.1007/s13412-017-0456-3 https://doi.org/10.1007/s13412-017-0456-3 https://doi.org/10.3390/cli6040099 https://doi.org/10.3390/cli6040099 https://doi.org/10.3390/cli6040099 https://doi.org/10.1007/s11069-014-1288-z https://doi.org/10.1007/s11069-014-1288-z https://doi.org/10.1007/s11069-014-1288-z https://doi.org/10.1016/j.geosus.2020.09.002 https://doi.org/10.1016/j.geosus.2020.09.002 https://doi.org/10.1016/j.geosus.2020.09.002 https://doi.org/10.1007/s13412-020-00617-6 https://doi.org/10.1007/s13412-020-00617-6 https://doi.org/10.1007/s13412-020-00617-6 https://doi.org/10.1016/j.ssci.2019.08.024 https://doi.org/10.1016/j.ssci.2019.08.024 https://doi.org/10.1016/j.ssci.2019.08.024 https://doi.org/10.1016/j.ssci.2019.08.024 https://doi.org/10.3389/fmars.2020.00160 https://doi.org/10.3389/fmars.2020.00160 https://doi.org/10.3389/fmars.2020.00160 https://doi.org/10.3389/fmars.2020.00160 https://doi.org/10.12006/j.issn.1673-1719.2016.075 https://doi.org/10.12006/j.issn.1673-1719.2016.075 https://doi.org/10.12006/j.issn.1673-1719.2016.075