Disasters, whether unexpected or predictable, have the potential to inflict significant human and economic harm on affected communities.Reference Abolghasem Gorgi, Jafari and Shabanikiya 1 , Reference Hidalgo and Baez 2 In today’s world, we are witnessing an increase in the frequency and severity of disasters, with an average of 352 million people being exposed to various forms of disasters annually.Reference Räsänen, Lein and Bird 3 This trend has led to a rise in the magnitude of damages, casualties, and economic losses. 4 , Reference Jones, Guha-Sapir and Tubeuf 5 Among the regions most susceptible to natural disasters is South Asia, where Iran stands out due to its vast size, diverse geography, and climatic variations, making it one of the most hazard-prone countries in the area. 6 , Reference Ali, Xu and Henchirli 7 Iran has experienced 31 of 45 different types of disasters, with floods and earthquakes being the most prevalent and destructive.Reference Shabanikiya, Seyedin and Haghani 8 As a result, approximately 90% of Iran’s population is at risk of being impacted by these disasters. 9 Every year, natural disasters in Iran lead to a significant loss of life and inflict injuries upon many individuals. Additionally, a substantial portion of Iran’s Gross National Product is allocated towards compensating for these losses. 10 Therefore, effective disaster management plays a crucial role in mitigating these impacts.Reference Partridge, Proano and Marcozzi 11 By incorporating risk management strategies into the 4-stage process of disaster management—comprising prevention and mitigation, preparedness, response, and recovery—we can substantially reduce both the human and economic toll of disasters.Reference Koenig and Schultz 12 At the initial stage of prevention and mitigation, conducting thorough risk assessments is paramount. The cornerstone of disaster risk assessment lies in identifying areas that are vulnerable to disasters. 13 , Reference Abdalla 14 There are multiple methods for identifying vulnerable areas, with geospatial analysis being one of the most renowned and effective approaches, particularly when aiming to identify such areas across a large geographical region.Reference Gadelha Ferreira Rosa, Douglas Oliveira Silva and Elisa Fontana 15 , Reference Kanaroglou and Delmelle 16 One of the studies conducted in this field was led by Shi et al. in 1 of the provinces of China.Reference Shi, Cui and Tian 17 The study aimed to investigate the distribution and trends of changes in the occurrences of floods and droughts in Northeast China across 637 districts. Spatial analysis methods were employed to determine the trends and distribution of these hazards. The results revealed a rising trend in flood occurrences from 1984-2010, while the occurrence of droughts showed no significant trend in East China.Reference Shi, Cui and Tian 17 Another notable study in this domain is the research conducted by Abedini et al.Reference Abedini and Beheshti 18 The objective of this research was to identify the hydrological factors that contribute to flooding incidents, such as land slope, soil type, and precipitation rates in a specific county in Iran. Through spatial analysis, these variables were examined to pinpoint areas with elevated flood risk.Reference Abedini and Beheshti 18
While past and similar studies have primarily focused on the physical factors influencing hazard occurrence and severity, it is essential to recognize that a society’s vulnerability to disasters extends beyond the likelihood and impact of a hazard event. 19 This vulnerability encompasses a range of variables, including socio-economic factors. 19 Today, the concept of social vulnerability to disasters has garnered significant attention from managers and disaster planners at all levels.Reference Arvin and Ziyari 20 Measures for disaster risk management and reduction are now being tailored and strategized based on social vulnerability.Reference Arvin and Ziyari 20 Social vulnerability to disasters encompasses the characteristics of individuals or societies that influence their abilities to anticipate, withstand, and recover from disasters.Reference Flanagan, Hallisey and Adams 21 It is crucial to emphasize that social vulnerability is not solely relevant in the initial stages of the disaster management cycle, such as prevention/mitigation and preparedness. Rather, it also holds substantial importance in the post-disaster phases, including response and recovery. In past disasters, it has been evident that there exist glaring inequalities in the distribution and utilization of public assistance among affected communities. The populations with higher social vulnerability often receive fewer benefits from aid compared to other groups.Reference Flores, Collins and Grineski 22 , Reference Han, Jia and Xu 23 This disparity results in a prolonged and inadequate recovery for these vulnerable sections of society following disasters.Reference Flores, Collins and Grineski 22 , Reference Han, Jia and Xu 23
Given the significance of the topic and the identified knowledge gap in this domain, the present study aims to identify socially vulnerable areas to disasters in the regions of Razavi Khorasan Province (Iran) that fall within the jurisdiction of Mashhad University of Medical Sciences (MUMS).
Theoretical Framework: Social Vulnerability to Disasters
Social vulnerability to disasters is shaped by a complex interplay of socio-economic, demographic, and environmental factors.Reference Cutter, Boruff and Shirley 24 A geospatial approach enables the identification of high-risk areas by integrating multi-dimensional indicators such as income levels, age distribution, housing conditions, and access to resources. This framework draws on disaster risk reduction (DRR) theory, which emphasizes that vulnerability is not merely a function of hazard exposure but also of societal resilience and adaptive capacity. 25
Key Dimensions of Social Vulnerability
-
1. Socio-Economic Factors: Poverty, unemployment, and low education levels exacerbate vulnerability by limiting preparedness and recovery capabilities.Reference Flanagan, Gregory and Hallisey 26
-
2. Demographic Characteristics: Elderly populations, children, and marginalized groups often face higher risks due to mobility constraints and dependency ratios.Reference Wisner, Blaikie and Cannon 27
-
3. Infrastructure and Housing: Informal settlements and substandard housing increase susceptibility to disaster impacts.Reference Jha, Barenstein and Phelps 28
-
4. Institutional Capacity: Weak governance and inadequate disaster policies amplify vulnerability.Reference Birkmann 29
Comparative perspective on social vulnerability trends
Recent studies highlight divergent social vulnerability trends across different regions, influenced by socio-economic development and policy interventions. For instance:
-
- Global Patterns: The study by Cvetković, et al. reveals that while high-income nations experience significant economic losses from disasters, low- and middle-income countries suffer disproportionately higher human casualties due to pre-existing social vulnerabilities.Reference Cvetković, Renner and Lukić 30 This underscores the role of economic resilience in mitigating disaster impacts.
-
- Regional Disparities: Research on Serbia demonstrates how demographic shifts (e.g., aging populations and rural depopulation) influence disaster resilience. Compared to Western Europe, Serbia’s slower recovery from floods and earthquakes is attributed to weaker institutional frameworks and economic instability.Reference Cvetković and Šišović 31
Method
This study was cross-sectional and utilized a mixed method approach, carried out in 2 stages.
Stage One
The primary objective of this stage was to assess the social vulnerability level of the rural areas within the scope of MUMS.
Setting
Iran is in the Middle East and ranks as a developing country, boasting a population of 87 million. The Khorasan Razavi province, situated in the northeast of the country with its capital city of Mashhad, holds the second position in population and the fifth in area among Iran’s provinces. The Mashhad University of Medical Sciences (MUMS), serving as the provincial branch of the Ministry of Health, plays a pivotal role in safeguarding the health of Khorasan Razavi’s residents during emergencies and disasters.Reference Ghavami, Kokabi Saghi and Asghari 32 Encompassing 61 898 square kilometers, which make up 52.07% of the province’s total area of 118 851 square kilometers, MUMS oversees 16 counties, 43 districts, and 91 rural districts, catering to a population of 8 632 379 individuals, 7.19% of whom are Afghan immigrants. 33 , 34 Although the province is threatened by various types of disasters, earthquakes, floods, and droughts are the most important and common ones.Reference Farzanegan, Fischer and Noack 35
Social vulnerability measurement
For increased accuracy and improved regional differentiation in terms of vulnerability, this study focused on analyzing rural districts as the primary unit of analysis. The Social Vulnerability Index (SVI) developed by the Centers for Disease Control and Prevention (CDC) in the US was utilized to assess social vulnerability.Reference Flanagan, Hallisey and Adams 21 The original SVI consists of 15 sub-indicators, which were reduced to 8 in its adjusted version for Iran. This adjustment was based on the opinions of 5 experts in the field of disaster management. Experts considered 2 primary principles when evaluating the sub-indices: data availability and relevance to the study area. Specifically, the excluded indicators were those for which the necessary data for calculation was unavailable, such as the number of disabled household members over 5 years old or the number of vehicles owned by the household, as well as those that were not pertinent to the study area, such as language. The entire population of the study area spoke Persian, with no second language, although there were some regional accents. These indicators encompass:
-
1. The proportion of vulnerable populations (aged under 5 years and over 65 years)
-
2. The percentage of unemployed individuals
-
3. The proportion of illiterate individuals
-
4. The percentage of individuals with an educational attainment of high school diploma or below
-
5. The percentage of residential buildings with a floor area of less than 100 square meters
-
6. The proportion of dwellings constructed with brick and mud
-
7. The percentage of households with more than 5 members
-
8. The percentage of immigrant populations.
To calculate each sub-index, we obtained the necessary raw data from Iran’s statistical center and inputted it into predetermined forms. The value of each sub-indicator was then computed separately for every rural district. A higher value for each sub-index signifies a greater vulnerability of that district to disasters. We adopted the CDC’s SVI calculation method for calculation SVI in our study. In this method all sub-indices were assigned equal weight. Subsequently, the rural districts were ranked based on each sub-index, with the top rank signifying the lowest vulnerability among all districts. The average ranks of the sub-indices were then determined for each district. The district with the lowest average rank demonstrated the least vulnerability, while a higher average rank indicated increased susceptibility to disasters. This average rank was designated as the SVI for a rural district.
Stage Two
The objective of this stage was to identify rural districts that stand out significantly from others in terms of SVI calculated in the previous stage, through geo-spatial analysis. Simply put, it aimed to pinpoint clusters of rural districts with notably high or low vulnerability within the geographical region covered by MUMS. Anselin’s Local Moran’s I was employed as a tool for spatial statistics to recognize geographic disparities that can reveal such clusters. This stage was executed through 2 steps.
Step One: Data Collection and Preparation
In this initial step, the administrative divisions map of Iran was obtained in shapefile format, comprising 4 levels: province, county, district, and rural district. Subsequently, utilizing ArcMap v.10.8.2 software, basic layers essential for the study, covering the areas of interest for MUMS at province, county, and rural district levels, were developed. Following this, the Social Vulnerability Index (SVI) for each rural district, alongside other demographic attributes of the regions under investigation, were inputted into Excel software. Subsequently, the Excel dataset was linked to the fundamental map layers constructed within ArcMap software for further analysis.
Step Two: Data Analysis
In this stage, we utilized Anselin’s Local Moran’s I index for data analysis. Local Moran’s I is specifically designed to detect clusters of similar values (hotspots) or outliers in spatial data. This capability is crucial for disaster risk assessment, as it helps identify areas with significantly higher or lower risks compared to their neighbors. Unlike global measures of spatial correlation, Local Moran’s I facilitates the exploration of local patterns. This is particularly useful in rural districts where risks may vary significantly even within geographically proximate areas. Furthermore, the local measure is sensitive to the spatial scale of the analysis, which is important when dealing with heterogeneous data across rural districts with differing socio-economic contexts. These features align perfectly with the purpose and context of our study. Our aim was to identify high-risk areas across local regions (districts). In contrast, other geo-spatial analysis methods, such as Geographically Weighted Regression (GWR), were not suitable for our study. While GWR enables modeling of spatially varying relationships between dependent and independent variables, our aim was not to explore the relationship between these variables or make predictions. Overall, our study did not involve distinct concepts (variables) categorized as dependent and independent. Instead, all components were integrated into a primary concept known as social vulnerability.
Local Moran’s I method involves calculating 3 values: z score, P value, and the Local Moran’s I index for each of the features (i.e., rural districts in our study). The null hypothesis posits that Social Vulnerability Indices (SVIs) are distributed randomly across the rural districts covered by MUMS.
The formula for computing the Local Moran’s I index is as follows:
$$ {I}_i=\frac{Z_i-\overline{Z}}{\unicode{x03C3}^2}\sum \limits_{j=1,j\ne i}^n\left[{W}_{ij}\left({Z}_j-\overline{Z}\right)\right] $$
Ii = the value of the local Moran index for the location (rural district) i
Zi = the value of the SVI at the location i
Zj = the total SVI values at other locations
б2 = the variance of z
Wij = distance weighting between zi and zj
Z̅ = the mean of SVIs
Based on the calculated values of z score, P value, and Local Moran’s index for each rural district, they are categorized into 5 groups:
-
1. High-High cluster: P value ≤ 0.05, positive z score, and high positive Moran’s Index
-
2. Low-Low cluster: P value ≤ 0.05, positive z score, and low positive Moran’s Index
-
3. High-Low outlier: P value ≤ 0.05, negative z score, and high negative Moran’s Index
-
4. Low-High outlier: P value ≤ 0.05, negative z score, and low negative Moran’s Index
-
5. Not significant: P value ≥ 0.05.
A high-high cluster indicates that the rural district(s) has a high value of SVI and is surrounded by rural districts with high values of SVI. It occurs when a rural district has an SVI significantly higher than its neighbors while enjoying a high SVI compared to the overall average of the dataset. This indicates that certain rural districts exhibit significantly elevated concentrations of SVI. In simpler terms, it highlights hotspots of SVI, prompting further investigation or targeted interventions. Similarly, a low-low cluster indicates that the rural district(s) has a low value of SVI and is surrounded by rural districts with low values of SVI. In terms of outliers, a High-Low outlier indicates that the rural district(s) has a high value of SVI and is surrounded by rural districts with low values of SVI. Conversely, a Low-High outlier is a rural district(s) that is surrounded by rural districts with high values of SVI. Statistical significance was defined as a P value less than 0.05. The data analysis was conducted using ArcMap version 10.8.2 software.
The concept of disaster risk is a comprehensive one that encompasses both vulnerability and hazard. This can be represented by the schematic formula: Disaster Risk = Vulnerability × Hazard.Reference Frigerio and De Amicis 36 –Reference Ranke 38 While our study primarily focused on understanding vulnerable areas, we also explored the element of hazard in the study regions, particularly concerning 2 main natural hazards in the country and Khorana Rave province: floods and earthquakes. To achieve this, we assessed the areas covered by MUMS for exposure to these 2 hazards. To evaluate exposure to earthquakes, historical disaster records were utilized. We obtained a map of earthquake events in Iran spanning 118 years (from 1905-2023) and created a layer illustrating these features in the MUMS-covered areas. The total magnitude of all earthquakes that occurred within and up to a 15-kilometer radius (buffer) of each rural district was calculated. Fifteen kilometers is the maximum destructive distance from the epicenter for earthquakes with moderate magnitudes (6-6.5 on the Richter scale). The rural districts were then ranked based on their total magnitude, with the highest rank indicating the greatest exposure to earthquakes. This ranking was designated as the Earthquake Exposure Index (EEI).
Flood models could not be utilized to assess exposure to flooding because certain parameters, such as soil type and land use, were unavailable for each district or at the counties level. Consequently, only 2 parameters were employed to measure flood exposure: the area occupied by flood plains and average annual precipitation.
The first parameter was the percentage of each district’s area occupied by flood plains. To measure this, we obtained a map of flood plains in Iran and created a layer highlighting these features in the MUMS-covered areas. The total area of flood plains in each rural district was then calculated. This total was divided by the district’s total area, and the result was multiplied by 100. The rural districts were ranked based on the area occupied by flood plains, with the highest rank indicating the greatest exposure to flooding.
For the second parameter, the necessary data on average annual precipitation was sourced from the Khorasan Razavi Meteorological Department. 39 The rural districts were then ranked according to their average annual precipitation, with the highest rank indicating the greatest exposure to flooding. The mean of the ranks for these 2 parameters was considered the flood exposure index (FEI).
Finally, the layers for flood exposure and earthquake exposure (FEI & EEI) were created. These designed layers were overlaid on the layer of SVI to visualize both socially vulnerable and hazardous areas within a single frame.
Results
Table 1 displays the categorization of rural districts based on SVI values within 3 thresholds, along with EEI and FEI values.
The categorization of rural districts of MUMS based on SVI, along with EEI and FEI values
In Table 1, it is shown that the rural districts of Shandiz, Sheshtaraz, and Hume exhibited the lowest vulnerability to disasters, with Social Vulnerability Index (SVI) scores of 30.7, 31.5, and 33, respectively. Conversely, the rural districts of Marzdaran, Nishtifan, and Pol Khatun were identified as most vulnerable, with SVI scores of 60.875, 58.75, and 58, respectively. The confidence interval is reported to be between 39.65 and 42.34. Furthermore, the interquartile range (IQR) was calculated to be 8.8.
Figure 1 illustrates the regions and counties encompassed by MUMS. As depicted in this figure, the areas covered by MUMS consist of 3 separate regions: 1 larger zone and 2 smaller zones. The larger zone encompasses mostly the northern and central portions of Razavi Khorasan province, containing 9 counties. The 2 smaller zones, situated in the southwest and southeast of the province, encompass 3 and 4 counties, respectively.
Regions and counties covered by MUMS.
Figure 2 shows the population density of the rural districts encompassed by MUMS. According to Figure 2, the central areas of the large zone, primarily encompassing the rural districts of Mashhad County, exhibit the highest population density, with a gradual decrease in population density as one moves further away from the center.
The population density of the rural districts encompassed by MUMS.
Figure 3 illustrates SVIs across the rural districts of MUMS. This figure illustrates that within the large zone, the index value in the peripheral areas—particularly in the eastern and southern regions—is higher than that in the center. This disparity suggests that these areas exhibit higher levels of social vulnerability. Conversely, in the smaller western zone, the central regions display lower social vulnerability. Within the small eastern zone, the index is elevated in the northeast, indicating a heightened vulnerability of these areas to disasters.
SVIs across the rural districts of MUMS.
Figure 4 displays the outcomes of Anselin’s Local Moran’s I analysis on the SVI. In Figure 4, the Anselin-local Moran’s I analysis revealed 4 clusters and 4 outliers. Within the larger zone, a rural district (code 28) in the central eastern area was classified as a high-high cluster, while a low-low cluster was identified in the central western region comprising 11 rural districts. Additionally, 2 rural districts were noted as outliers in this zone: 1 low-high (code 46) in the southern part and 1 high-low (code 45) in the southwest. Moving to the eastern small zone, a high-high cluster consisting of 4 rural districts and a low-high outlier (code 60) were identified. In the smaller western zone, a low-low cluster containing 5 rural districts and 2 high-low outliers were observed.
Outcomes of Anselin’s Local Moran’s I analysis on the SVI.
Figure 5 illustrates the EEI and FEI of rural districts, along with their SVI. This map shows that nearly all rural districts covered by MUMS in all 3 zones are exposed to flood and earthquake hazards. This issue is especially pronounced in rural districts with higher social vulnerability. In these areas, the points are larger, indicating they are more exposed to natural hazards.
The EEI and FEI of rural districts, along with their SVI.
Discussion
The results of the present study indicate that there are regions within all 3 zones of MUMS exhibiting high social vulnerability to disasters. A comparison of these findings with similar studies highlights the pervasive nature of social vulnerability to natural disasters, posing a significant challenge to disaster management at any community and level. For instance, 1 study focused on assessing the social vulnerability in District Two of Tabriz city, Iran, concerning disasters.Reference Roustai and Maboudi 40 The study revealed that approximately 18.9% of the surveyed area, encompassing nearly 1/3 of the population, faced high vulnerability levels.Reference Roustai and Maboudi 40
In a separate study within this field, which encompassed 11 cities in China, researchers identified regions and populations vulnerable to geological disasters.Reference Xiao, Tang and Li 41 The study utilized a range of variables and indicators, including social structure, demographic details of the population, disaster perception, and risk perception, to evaluate vulnerability. The results revealed variations in vulnerability levels across the studied areas, with most cities displaying high to medium vulnerability levels.Reference Xiao, Tang and Li 41
The results of the current study revealed that numerous areas under investigation, classified as having high levels of social vulnerability, also faced significant exposure to natural hazards of earthquake and flood. This highlights the correlation between social vulnerability and susceptibility to hazards; specifically, regions with greater social vulnerability are at a higher risk of facing natural hazards. The findings of the Zarghami and Dumrak study support this assertion.Reference Zarghami and Dumrak 42 The purpose of their study was to provide a dynamic model for simulating and predicting socio-economic and demographic characteristics that affect vulnerability to disasters. In their study, they initially drew the relationship between different variables of social vulnerability using a Causal Loop Diagram (CLD). They then transformed the CLD into a Stock and Flow Diagram using the Analytic Hierarchy Process to create a social vulnerability index. Subsequently, a disaster risk assessment was conducted in an Australian city by considering both the social vulnerability index and the severity of the hazard.Reference Zarghami and Dumrak 42 This process resulted in the identification of 5 vulnerable clusters within the city through a cluster analysis. The findings of this study demonstrated a correlation between the likelihood of hazard exposure and social vulnerability.Reference Zarghami and Dumrak 42 It was revealed that the probability of hazard exposure in the identified vulnerable clusters was higher compared to other areas.Reference Zarghami and Dumrak 42
This issue was also highlighted in the study by Chen et al.Reference Chen, Liu and Ge 43 The study aimed to assess the risk of flood disasters by combining social vulnerability and likelihood of floodwater (flood hazard) in Nanjing city, China. Social vulnerability in different areas was determined based on various indicators and variables.Reference Chen, Liu and Ge 43 The likelihood of flood hazard was assessed using the Soil & Water Assessment Tool, an extension of ArcGIS software, for different areas of the city. The study findings revealed a central-periphery pattern in social vulnerability, with higher vulnerability in central areas decreasing gradually towards the suburbs.Reference Chen, Liu and Ge 43 This pattern was also observed in the risk of flood hazard occurrence, with central areas having the highest risk, exurban areas having a lower risk, and suburban areas having the lowest risk.Reference Chen, Liu and Ge 43
In general, areas with higher social vulnerability are more likely to be exposed to hazards because they often have unsuitable conditions for human habitation. For instance, these areas may include inaccessible mountainous regions or riverbanks, where people with higher socio-economic status tend to avoid residing. The results of a comprehensive survey conducted by the World Bank in 2017, titled “Disaster Poverty Household Survey” across 5 countries - Brazil, Ethiopia, Ghana, Haiti, and Tanzania - support this hypothesis.Reference Erman, Malgioglio and Yoshida 44 –Reference Touray and Perge 47 The primary objective of this study was to investigate why impoverished communities tend to reside in flood-prone regions. The research revealed that individuals living in poverty are often inclined to settle in flood-prone areas due to the affordability of housing and the availability of employment opportunities within these localities.Reference Erman, Malgioglio and Yoshida 44 –Reference Touray and Perge 47
This hypothesis was further validated in a separate study carried out by Kawasaki et al. in Bago, Myanmar.Reference Kawasaki, Kawamura and Zin 48 The study aimed to explore the correlation between flood exposure and poverty at the household level. With a sample size of 213 households, the research results demonstrated that impoverished households are more likely to reside in flood-prone regions.Reference Kawasaki, Kawamura and Zin 48 This can be attributed to the fact that their proximity to rivers and canals offers increased job prospects and consequently higher income opportunities.Reference Kawasaki, Kawamura and Zin 48
It is important to note that the issue at hand is not restricted solely to flooding; it also encompasses other hazards such as earthquakes.Reference Boustan, Kahn and Rhode 49 A research report from the National Bureau of Economic Research in the US analyzed data from over 100 diverse disasters, including earthquakes, spanning almost a century (1920-2010).Reference Boustan, Kahn and Rhode 49 The study identified a notable trend in terms of migration and population displacement in areas prone to hazards.Reference Boustan, Kahn and Rhode 49 The trend observed indicates that affluent individuals and households tend to gradually relocate from risky zones to safer regions, while those with lower incomes remain in hazardous areas and lack the financial resources to relocate.Reference Boustan, Kahn and Rhode 49
Below, successful disaster risk reduction strategies implemented in regions with comparable socio-economic profiles are examined. Examples include Southeast Asia (e.g., Bangladesh, Myanmar), Sub-Saharan Africa (e.g., Ethiopia, Somalia), and Latin America (e.g., Haiti, Nicaragua)—all regions that face pervasive poverty and vulnerability to natural hazards.
In Southeast Asia, the efficacy of community-based disaster risk management (CBDRM) and advanced early warning systems in enhancing community preparedness is evident. 50 , Reference Ahmed, Kelman and Fehr 51 Sub-Saharan Africa demonstrates the value of integrating disaster risk reduction into national development policies and promoting livelihood diversification programs. 52 –Reference Weldegebriel and Prowse 54 In Latin America, the adoption of robust building codes and urban planning initiatives, combined with effective emergency response networks, showcases strategies for mitigating earthquake and hurricane impacts. 55 , Reference Alvear, Campos and Ciancio 56
These diverse approaches collectively underscore the importance of tailored, context-specific interventions for building resilience in vulnerable populations.
Although it may appear that not adopting an all-hazard or multi-hazard approach was a limitation of the study, this research intentionally avoids a multi-hazard risk assessment strategy due to several practical constraints. The inherent diversity of hazards, both natural and man-made, presents significant challenges in establishing a unified metric for measuring magnitude, probability, and consequence. Discrepancies exist in the definition and measurement of different hazards, with some, like earthquakes, having established scales and monitoring systems, while others, such as droughts, lack standardized assessment. Furthermore, the geographical scale of impact varies considerably between hazards, rendering localized studies ineffective for widespread events like droughts. Beyond technical difficulties, the practical utility of a unified risk assessment is questionable, as distinct prevention and mitigation strategies are required for each hazard type, necessitating separate risk assessments. In conclusion, the all-hazards approach, while conceptually appealing, proves technically and practically cumbersome for effective disaster risk assessment and management in this specific study.
One limitation of the study was the lack of data for some sub-indices of the original version of SVI, as mentioned in the methods. However, the adjusted version of SVI used in the current study covers all main dimensions of the original version, ensuring that at least 1 sub-indicator from each main dimension of the original SVI is included in the adjusted version.
Another limitation of the current study was related to assessing exposure to flood hazards. Flood models could not be employed to evaluate exposure to flooding due to the unavailability of certain parameters, such as soil type and land use, as outlined in the methodology section. It should also be kept in mind that the hazard assessment was not the primary objective of the study, therefore we chose the simplest method based on the available data for hazard assessment. One strength of this study is its approach to simultaneously examining 2 key factors: social vulnerability and hazard across a relatively extensive geographical region.
Conclusion
In total, almost half of the studied areas exhibited a high level of social vulnerability and were prone to natural hazards like floods and earthquakes. The results of this study offer compelling evidence for disaster policymakers to prioritize disaster management efforts in these vulnerable regions. Many steps that can effectively mitigate social vulnerability to disasters align with initiatives that enhance the overall socioeconomic status of individuals and households, including improving literacy and education levels, and promoting economic empowerment through employment opportunities to increase income levels.
Additionally, taking specific actions, such as decreasing the reliance of vulnerable groups on residing in hazardous zones for their income while offering both financial and non-financial encouragements for households to relocate to safer areas, or constructing more resilient homes for those living in hazardous regions, can greatly contribute to mitigating risks of natural disasters.
In general, land-use planning is inevitable for reduction the risk of earthquake and flood in high risk districts. Some of the land-use planning measures include implementing stringent zoning regulations to prevent new developments in areas susceptible to floods or earthquakes, providing incentives for developers to build in safer zones, integrating green infrastructure in urban planning (such as parks and green roofs) to absorb excess rainfall and mitigate flooding risks, and retrofitting high-risk buildings and infrastructure to meet modern seismic and flood-resistant standards, especially in vulnerable neighborhoods.
Developing accessible and comprehensible early warning systems tailored to local contexts, ensuring that high-risk communities receive timely alerts about impending natural disasters, can also be beneficial. Infrastructure resilience is another critical issue in this regard. Prioritizing investments in high-risk areas for resilient infrastructure such as roads, bridges, and public transportation that can withstand natural disasters and provide safe evacuation routes is essential.
Furthermore, while we utilize static social vulnerability sub-indicators, developing an up-to-date database of these indicators could be advantageous for organizations engaged in disaster risk management.
Data availability statement
Data and the scale will be made available from the corresponding author upon reasonable request.
Acknowledgements
This research was financially supported by Mashhad University of Medical Sciences.
Author contribution
Maryam Salari: Formal analysis, Methodology, Writing – review & editing. Alam Abbasiyazdi: Data curation. Hamidreza Shabanikiya: Conceptualization, Formal analysis, Investigation, Methodology, Resources, Software, Writing – original draft, Writing – review & editing.
Competing interests
The authors declare that they have no competing interests.
