1 Introduction

People around the world are vulnerable to environmental changes and disasters, with climate-related disasters such as windstorms, floods, and droughts becoming more frequent (Balica 2012; Kron 2014). Climate variability and change are major contributors to extreme rainfall events, which can lead to floods (Milly et al. 2002; Carter et al. 2009; Pörtner et al. 2022). Vulnerability assessment is an important aspect of flood risk management (Turner et al. 2003) and climate change adaptation strategies (Ledda et al. 2021). Previous research has emphasized the importance of incorporating climate change and vulnerability assessments into environmental impact assessments, as they are crucial for sustainable planning and future development (Burdge 2008; Jiricka et al. 2016; Climent-Gil et al. 2018; Fischer 2019; Rana et al. 2022).

Pakistan is highly vulnerable to floods and has low capacity to cope with them. Any variability in climate extremes makes it difficult for authorities to implement sustainable disaster risk management strategies (Kirsch et al. 2012). The 2010 floods in Pakistan received widespread attention and were considered the worst in the country’s history (Khan 2013). Approximately 14–20 million people were affected, and 1700 lost their lives (Shabir 2013). The floods also destroyed 1.1 million homes and around 400 healthcare facilities that cost about USD 9.7 billion (Houze et al. 2011). Some areas were flooded for over six months, and the rural economy was severely impacted. Additionally, 80% of the country’s forest reserves were lost (Polastro et al. 2011). The overall loss and damage reported under the Pakistan 2022 Flood Response Plan, particularly in the rural areas, was more than USD 40 billion (United Nations 2022). Several studies have highlighted vulnerability of rural areas in Pakistan (Shah et al. 2018; Jamshed et al. 2019; Hamidi et al. 2020; Shah et al. 2020; Khalid et al. 2021; Ullah et al. 2021). However, there has been a dearth of comprehensive studies that specifically address vulnerability from a multidimensional perspective, encompassing physical, social, institutional, environmental, economic, and attitudinal dimensions. Notably, the inclusion of attitudinal vulnerability in such multidimensional assessments is particularly scarce in existing research. Consequently, this study sought to address this gap by proposing a framework for evaluating multidimensional vulnerability in rural communities located in Muzaffargarh, Pakistan. Additionally, the research investigated how the incorporation of this framework into disaster risk management strategies can enhance their efficacy and effectiveness.

2 Conceptualization of Vulnerability

In the context of disaster risk reduction and climate change adaptation, vulnerability is often recognized as the root cause of the same disasters having different impacts on elements at risk (Adger 2006). There are numerous frameworks and definitions of vulnerability in the literature, as practitioners from different fields use the concept in different contexts (Cutter et al. 2003; O’Brien et al. 2004; Birkmann 2007). The United Nations International Strategy for Disaster Reduction (UNISDR) defined vulnerability as the predisposition of a system or community determined by social, physical, economic, and environmental factors that influence the impacts of hazards (UNISDR 2009). The United Nations Development Program (UNDP) stated that the likelihood of damage caused by a specific hazard is determined by anthropogenic conditions modified by economic, physical, social, and environmental components (Dao and Peduzzi 2004). The Intergovernmental Panel on Climate Change (IPCC) described vulnerability as the susceptibility or inability of a system to cope with the detrimental effects of climate change (Gallopín 2006). Climate change scientists often analyze vulnerability through the lens of sensitivity, exposure, and coping capacity. Exposure refers to the stressor that increases a system’s vulnerability and is location-dependent, while sensitivity refers to the extent to which a system is influenced by climate perturbations (Smit et al. 2000; Brenkert and Malone 2005; Ippolito et al. 2010). Coping capacity is defined as a system’s ability to mitigate the impacts of a hazard. In terms of climate change, it is the potential of a system to adjust or adapt to climate change and withstand adverse impacts (Roy and Blaschke 2015). While there are various explanations of vulnerability, all definitions agree that it is the inherent tendency of a system to be susceptible to external shock (Fekete 2010).

Previous research has reviewed several widely accepted methods for vulnerability assessment, including computer modeling, curve analysis, indicator-based approaches, and disaster loss analysis (Nasiri et al. 2016). The indicator-based approach is often used to quantify vulnerability because it provides a comprehensive, reliable, and quick evaluation of flood vulnerability in an area (Krellenberg and Welz 2017). Assessing the vulnerability of urban regions requires considering physical vulnerability, while past studies have highlighted the importance of economic and social factors in vulnerability analysis (Field et al. 2012). To improve the effectiveness of disaster risk reduction strategies, a comprehensive framework is needed for analyzing multidimensional vulnerability (Rana and Routray 2018). The flexibility of the framework can affect the comparability of results, so it is important to include widely accepted indicators (Carreño et al. 2007). This study focuses on various dimensions to determine multidimensional vulnerability by categorizing indicators into main dimensions (Fig. 1), standardizing the data, and calculating indices for each dimension. A composite index for multidimensional vulnerability is then determined by aggregating these indices (Adger and Agnew 2004).

Fig. 1
figure 1

Dimensions of vulnerability

Full size image

2.1 Physical Vulnerability

Physical vulnerability is defined as the degree to which vulnerable assets are exposed to natural hazards. Research has generally divided physical vulnerability into three main categories: buildings, occupancy, and infrastructure indicators (Mileti 1999; Douglas 2007). Structural factors, such as building type, age, floors, and construction material, fall under the category of buildings. Occupancy refers to the use of a building, such as industrial, residential, or commercial, which can affect the scale of damage from a specific hazard (Kircher et al. 2006; Lagomarsino and Giovinazzi 2006; Duzgun et al. 2011). Infrastructure, such as transportation infrastructure (railways, highways, and ports) and utility services (wastewater management, drinking water supply, electric power, and gas supply), can also impact vulnerability (Hahn et al. 2009; Ahsan and Warner 2014). The availability of essential services such as hospitals, schools, fire stations, and police, as well as proximity to these facilities, also play a significant role (Peng 2015). Physical vulnerability is also influenced by geographic proximity to the source of a disaster (Botzen et al. 2009a; Geneletti 2012).

2.2 Social Vulnerability

Social vulnerability refers to the inability of individuals, societies, and organizations to withstand the impacts of a hazard due to the intrinsic characteristics of institutions, social interactions, and cultural norms (Kelly and Adger 2000; de Loyola Hummell et al. 2016). The resilience of individuals, communities, and societies also affects the social aspect of vulnerability (Adger and Kelly 1999; Wisner et al. 2004; Lee 2014; de Sherbinin and Bardy 2015). Important indicators of social vulnerability include education level, religious beliefs, political accountability, traditional values, literacy rate, social equity, and community organization (Cutter et al. 2003; Bergstrand et al. 2015). Developing countries are often highly susceptible to natural hazards due to a lack of information and resources, resulting in unsustainable measures to reduce the impacts of hazards. Children, the elderly, women, and people with disabilities are typically considered highly vulnerable (Cutter et al. 2003; Turner et al. 2003; Khan 2012).

2.3 Institutional Vulnerability

The negative impacts of natural hazards on a community do not solely depend on visible factors such as poor infrastructure, geographic location, dependency ratio, and income level. There are also less obvious factors that can affect a community’s vulnerability to disasters (Smit and Wandel 2006). The efficiency of communication and coordination among responsible stakeholders, as well as the level of awareness, risk transfer mechanisms, and public participation, can all play a critical role in the institutional setup and disaster risk management of a society. The legislation and frameworks related to preparedness, reconstruction, response, and implementation also contribute to a community’s vulnerability (de Guzman and Unit 2003; Jongman et al. 2014; Scolobig et al. 2015). Institutional vulnerability refers to the way in which institutions approach and address vulnerability to disasters, rather than the vulnerability of the institutions themselves (Fuchs 2009; Papathoma-Köhle et al. 2021). It is important to consider these factors in order to improve the preparedness, responsiveness, and coping capacity of a community in all stages of a disaster (López-Martínez et al. 2017).

2.4 Environmental Vulnerability

Environmental vulnerability refers to the susceptibility of natural assets or capital to harm or damage as a result of any event or process (Jamshed et al. 2019). Previous research has demonstrated a cyclical relationship between disasters and the environment, with environmental changes often leading to the exploitation of natural resources and the degradation of natural barriers, reducing resilience and exacerbating the impact of disasters (Srinivas and Nakagawa 2008). Human activities that are not carefully planned and managed can undermine the integrity of ecosystems, leading to the loss of biodiversity and essential environmental services that support human well-being (Kaly et al. 1999). It is important to consider the environmental impacts of disasters in order to preserve and protect natural assets and reduce vulnerability to future events.

2.5 Economic Vulnerability

The economic vulnerability of an area is determined by several factors, including the source of income, the availability of natural resources, the ease of control and access to means of production, the availability of livelihood options, the dependency ratio, and the availability of loans and insurance (Browne and Hoyt 2000; Hahn et al. 2009; Birkmann et al. 2013; Nhuan et al. 2016). Low-income families are particularly vulnerable as they often lack resources and opportunities for stable employment, and may be forced to live in squatter settlements due to financial constraints. These families may be more susceptible to economic shocks and may have difficulty recovering from financial setbacks (Winsemius et al. 2018). It is important to address the economic vulnerabilities of low-income communities in order to promote stability and improve living conditions.

2.6 Attitudinal Vulnerability

The risk of disasters can also be influenced by the attitude of a community (Räsänen et al. 2020). A society with a negative attitude towards change and innovation may be more vulnerable to hazards (Rana and Routray 2018), as individuals may be less likely to take independent actions to prepare for and respond to disasters (Grothmann and Reusswig 2006). This lack of individual agency can lead to a greater reliance on external support and a lack of diversity in livelihood options, which can weaken the community’s overall resilience (Pomeroy et al. 2006; Appleby-Arnold et al. 2018). In addition, a lack of unity and collectivism can lead to conflicts and a reduced capacity to cope with disasters, further diminishing the community’s ability to withstand and recover from disasters (Norris et al. 2008; Cornia et al. 2016).

3 Methodology

This section explains the details regarding the study area and rationale behind selecting it. Muzaffargarh was selected as the case study area. It also explains the sampling plan and data collection procedure. Lastly, the indicators selected for the study and the technique adopted for data analysis are discussed in detail.

3.1 Study Area

Muzaffargarh is a district located in the southern part of Punjab, Pakistan. It has a long history and has experienced numerous floods in recent decades due to its location between the Indus and Chenab Rivers. The district is divided into four main administrative units, known as tehsils, which are further divided into 93 union councils. The annual average temperature in Muzaffargarh is 25.6 °C, and around 40% of the area is devoted to agriculture, with rice, sugarcane, wheat, and cotton being the primary crops. During the active period of the Chenab River (mid-July to September) and Indus River (August to October), an increased flow or movement of riverine water is observed (Qurratulain and Munazza 2014). During this period, for the purposes of this study, four union councils were selected based on their proximity to the rivers and past flood experience. These union councils are Shareef Chajra, Muradabad, Ghazi Ghat, and Rangpur. Shareef Chajra and Ghazi Ghat are located near the Indus River, while Rangpur and Muradabad are situated close to the Chenab River. Figure 2 illustrates the location of the study area within Muzaffargarh.

Fig. 2
figure 2

Study area map of Muzaffargarh in southern Punjab, Pakistan

Full size image

3.2 Data Collection

There are various sampling methods described in the literature, and this study employed the Yamane (1967) sampling technique to calculate the required sample size (Eq. 1). Random sampling was used in order to minimize biases, as the study area was not large and a confidence level of 95% and e = ± 10% was chosen. Based on these parameters, a sample size of 392 was determined. However, some of these questionnaires were deemed inadequate and were discarded, resulting in a total sample size of 365.

$$n=\frac{N}{1+N{\left(e\right)}^{2}}$$
(1)

where n denotes sample size, e is the precision level, and N refers to the size of the population.

Based on the computed sample size, data were collected from the local inhabitants of Shareef Chajra, Muradabad, Ghazi Ghat, and Rangpur. Respondents were selected using the simple random sampling technique in order to ensure a representative sample. The data collection process involved the utilization of a structured questionnaire survey, which was administered through face-to-face interviews conducted by the primary author. The questionnaire consisted of a comprehensive set of questions designed to assess each dimension of vulnerability. The respondents’ informed consent was duly obtained.

3.3 Indicator Selection

Data collection from the selected sample involved the use of a carefully constructed set of questions derived from a comprehensive set of vulnerability indicators. The identification and extraction of these indicators were performed through a meticulous review of pertinent literature. Each relevant study was thoroughly examined, allowing for the categorization of indicators across six distinct dimensions of vulnerability, namely physical, social, institutional, environmental, economic, and attitudinal. Subsequently, the selected indicators were accompanied by clear explanations and assigned response categories with corresponding weightages. The specific indicators, along with their respective explanations, are presented in Table 1. These indicators were organized into six main dimensions, which were used to create a composite index of multidimensional vulnerability. To maintain consistency among the indicators, ordinal scale was adopted for each indicator’s responses. In order to encourage widespread participation in the survey, the confidentiality of the responses was emphasized, and participants were assured that their responses would be analyzed as a group. This research examined six dimensions of vulnerability for each of the four union councils. A total of 10 indicators were selected to assess social vulnerability, 10 indicators were used to assess economic vulnerability, 12 indicators were used to measure physical vulnerability, 5 indicators were chosen to assess environmental vulnerability, 20 indicators were used to assess attitudinal vulnerability, and 8 indicators were used to evaluate institutional vulnerability. The selection of indicators aims to facilitate a simplified assessment and calculation of vulnerability, enabling a comprehensive understanding of the individual contributions of each indicator towards the overall vulnerability to floods in rural communities. These carefully chosen indicators are recognized for their significant influence on the vulnerability of individuals, particularly in the context of rural flood events. Notably, any observed increase or decrease in a specific indicator is expected to impact the corresponding dimension’s vulnerability level and subsequently influence the values of the multidimensional vulnerability index.

Table 1 Indicators and dimensions for assessing multidimensional vulnerability in the southern Punjab, Pakistan, study area
Full size table

3.4 Formulation of a Multidimensional Vulnerability Index

Numerous studies have examined vulnerability levels through diverse equations, models, and techniques. However, limited studies have developed composite index based on individual scores for each vulnerability dimension. In response to this gap, the present study formulated a multidimensional vulnerability index. This index possesses significant strengths as it provides a comprehensive understanding of community vulnerability by encompassing all relevant factors that influence household capacities and vulnerabilities. While the index is specifically designed for rural flood contexts, it can be adapted for measuring vulnerability to other hazards by replacing the pertinent variables accordingly. For formulation of multidimensional vulnerability index, the primary datasets were standardized according to the guidelines of previous studies. Classes were then assigned to each indicator using the subjective weighting technique to create composite indices for all six dimensions of vulnerability (Eq. 2). All transformed values fell between 0 and 1, with values closer to 1 indicating a higher level of vulnerability. Depending on the categories, each indicator was first divided into classes. For example, questions that could be answered with a “yes” or “no” were assigned 0 and 1 values, respectively. Indicators with five classes were assigned values of 0.2, 0.4, 0.6, 0.8, and 1, while indicators with four classes were assigned values of 0.25, 0.50, 0.75, and 1. Indicators with three classes were transformed into values of 0.33, 0.67, and 1. The composite indices for six vulnerability dimensions were then combined and their average was calculated to get the resultant multidimensional vulnerability index (Eq. 3).

$$Composite Index (CI)=\frac{{W}_{1}+{W}_{2}+{W}_{3}+\dots {W}_{n}}{n}$$
(2)

Social Vulnerability Index (SV) =\(\sum_{i=1}^{10}=SV/n\) (n = 10)

Economic Vulnerability Index (EV) = \(\sum_{i=1}^{10}=EV/n\) (n =10)

Physical Vulnerability Index (PV) =\(\sum_{i=1}^{12}=PV/n\) (n = 12)

Environmental Vulnerability Index (GV) = \(\sum_{i=1}^{5}=GV/n\) (n = 5)

Attitudinal Vulnerability Index (AV) = \(\sum_{i=1}^{20}=AV/n\) (n = 20)

Institutional Vulnerability Index (IV) = \(\sum_{i=1}^{8}=IV/n\) (n = 8)

$${\text{Multidimensional}}\,{\text{Vulnerability}}\,{\text{Index}} = \frac{SV + EV + PV + GV + AV + IV}{6}$$
(3)

4 Results and Discussion

Vulnerability indices were calculated using the methodology described in the previous section. Descriptive statistics were applied to each dimension in order to identify statistical differences among the communities. The results for each dimension are presented in detail, along with the calculation of the multidimensional index for each union council.

4.1 Social Vulnerability

The majority of the households (68%) in all four union councils had 5–10 members, while 38% had two children over the age of 18. Around 45.5% of the households had more than two elderly members (60 years and above), who may require special care and have limited mobility, making them more vulnerable. In this study, relatively few people in all four rural communities needed special care. A lack of education can also make individuals more vulnerable by preventing them from understanding early warning systems and emergency protocols. Approximately 26.8% of the households had no educated members. Most households (56.1%) had lived in flood-prone areas for over 40 years, making them more aware of evacuation routes and shelters in the event of an emergency. About 90% of participants had experienced a flood in the past, with the majority experiencing it more than twice. Muradabad and Ghazi Ghat had relatively high levels of social vulnerability (0.42), while the households in Shareef Chajra were comparatively less socially vulnerable (0.36). The ANOVA value was significant (F = 8.78, p = 0.00), indicating that there were significant differences among the four rural communities. Overall, 50% of the sampled households were found to be highly vulnerable (Fig. 3). The high social vulnerability observed in the selected rural communities can be attributed to key indicators such as the presence of elderly individuals, low educational attainment levels, and short length of residency.

Fig. 3
figure 3

Dimension-wise vulnerability in the southern Punjab, Pakistan, study area

Full size image

4.2 Economic Vulnerability

Most household heads (79.4%) were fully employed, resulting in a low dependency ratio in all four rural communities. Around 86% of the households had no members employed outside the district, indicating high vulnerability as those employed outside the flood-prone area are generally considered safer and can provide support to their families during a crisis. Additionally, the majority of participants belonged to low-income groups and struggled to meet their daily living expenses, leaving little or no money for savings. People who depend on agriculture for their livelihood are often highly vulnerable to disasters, as floods can cause significant damage to crops and seriously impact those involved in agriculture. In this study, most individuals were engaged in private business or services, with only a small percentage working in agriculture (7.9%). Approximately 30% of participants lived in social housing or rented privately and were unable to fortify their buildings, making them more vulnerable to floods. Significantly, 97% of the respondents did not have insurance and could face economic challenges during a flood. The majority of individuals were free of debt and owned private means of transportation. Shareef Chajra had the lowest level of economic vulnerability (0.49), while Ghazi Ghat was the most vulnerable (0.55). The ANOVA value was significant (F = 12.26, p = 0.00), indicating a clear difference among the communities. Overall, 46% of the sampled households were highly vulnerable (Fig. 3). The primary driver of high economic vulnerability in the studied rural communities was the prevalence of low-income and smallholder farmers who faced significant challenges in terms of obtaining insurance coverage and securing loans.

4.3 Physical Vulnerability

The majority of the sampled households were located within a buffer zone of 1–5 km from the nearest river and medical facility. Most houses were made of mud (katcha), were semi-detached, and were less than 10 years old. More than 90% of these houses were single-story with no basement, making them highly vulnerable to floods. In terms of utility services, almost all homes had access to electricity and means of communication. However, access to drinking water, sanitation, and gas supply was limited in all four rural communities. The ANOVA value was high (F = 12.04, p = 0.00), indicating a significant difference among the communities. Shareef Chajra was the most vulnerable in terms of physical vulnerability (0.66), followed by Muradabad (0.62). Around 62% of the sampled households were vulnerable in the infrastructural dimension (see Fig. 3). The predominance of single-story housing limited the households’ resilience to physical hazards, while the limited availability of basic necessities further exacerbated their vulnerability.

4.4 Environmental Vulnerability

The nature of past floods can impact people’s perceptions of flood risk and vulnerability. The majority of the respondents considered the severity of the previous floods to be high, with a height of less than 1 m. Around 26% of the households had family members who became ill as a result of the flood, indicating a higher level of vulnerability. About 90% of individuals were able to evacuate successfully during previous flood events, indicating that people were aware of evacuation routes in their areas. The duration of evacuation was typically 2–3 days. Among all four rural communities, those near the Indus River had relatively high levels of environmental vulnerability, including Ghazi Ghat (0.53) and Shareef Chajra (0.46). The ANOVA value was high (F = 30.69, p = 0.00), indicating a clear difference among the communities. Around 36% of the sampled households were vulnerable due to environmental characteristics (see Fig. 3). The geographical positioning of the selected communities in close proximity to both the Chenab and Indus Rivers renders them susceptible to recurrent rural flooding during the monsoon season.

4.5 Attitudinal Vulnerability

Around 88% of the households had never approached the government for information about floods. Most respondents had discussed flood preparedness with their family members and believed that they were capable of dealing with floods without external support. Around 56% of the households expressed an interest in seeking flood-related information in the future, indicating that people were taking floods seriously. Many respondents were highly inclined to take out insurance and floodproof their houses in the future. More than 40% of individuals considered the possibility of participating in flood training programs and taking flood preparedness measures in the future. However, understanding of early warning systems and emergency protocols was limited. In addition, people trusted the media more than the government. Most households were terrified of floods and believed that a flood would damage their houses and disrupt the food supply. Around 39% of the respondents believed that a flood could cause loss of life in their area. The perceived impact of a flood on personal health and the social environment was high. However, people were uncertain about the effect of a flood on utility services. More than 40% of the households believed that a flood was highly likely to occur in their area and would significantly impact their lifestyles. The attitudinal index was relatively high in Muradabad (0.60) and Rangpur (0.51), located near the Chenab River. A high index value indicates a negative attitude towards floods and a lower likelihood of preparation. The ANOVA value was very high (F = 86.58, p = 0.00), indicating that all four communities were statistically significant. The analysis found that 49% of the households were vulnerable due to their attitudes, beliefs, and perceptions (see Fig. 3).

4.6 Institutional Vulnerability

Institutional vulnerability included accessibility to reach institutions and the efficiency of early warning systems, emergency plans, and risk communication plans developed by institutions. Around 68% of the households had good knowledge of first aid. However, significantly 93% of the respondents had never attended any awareness drills that could help them cope with floods. Around 98% of the households considered the government responsible for protecting against floods. In addition, people continued to rely on traditional methods of obtaining information, resulting in higher vulnerability as these methods may not effectively deal with future floods. Government authorities were the least trusted source of information about floods or for seeking help during flood events, as they were unable to provide information and relief during previous floods. Institutional vulnerability was relatively high in communities located near the Chenab River, including Muradabad (0.71) and Rangpur (0.72). The ANOVA value was high (F = 5.88, p = 0.00), indicating statistical differences among all communities. In terms of the institutional dimension, 27% of the households were found to be vulnerable (see Fig. 3).

4.7 Multidimensional Vulnerability

In rural areas, the proximity of communities to the Indus and Chenab Rivers made them vulnerable to flooding. The communities located near the Chenab River had a higher level of attitudinal and institutional vulnerability compared to Muradabad and Rangpur. The respondents from these communities had a fatalistic attitude towards flood hazards, and government authorities were not trusted due to their poor management of previous floods. However, there was a slight variance in all dimensions except for environmental and attitudinal vulnerability. Shareef Chajra was more physically vulnerable, with a vulnerability score of 0.66. However, the participants from this community had the lowest vulnerability in all other aspects. Overall, Ghazi Ghat was the most vulnerable to floods compared to the other communities. Ghazi Ghat is located near the Indus River and has a high risk of flooding due to its location. A significant difference was found among all four union councils, with a p-value of 0.00.

This case study examined the local nature of vulnerability based on six dimensions that have been widely used in previous research (Ahsan and Warner 2014). In terms of social vulnerability, approximately 24.9% of the households were classified as highly vulnerable. The communities had large family sizes with a high number of children under 18 years of age, which increased their exposure to risk. In this study, the education level of household heads was mainly limited to middle school, resulting in a poor understanding of awareness and emergency protocols. Length of residency is also a factor that can impact vulnerability, as it determines people’s knowledge about evacuation routes in the area (Ullah et al. 2021). Most of the respondents in this study were long-term residents who had experienced a flood in the past, which likely influenced their perception of potential flood loss (Rana and Routray 2018). In terms of economic vulnerability, the surveyed households belonged to the middle-income group and had a low dependency ratio with few members employed outside the flood zone, increasing their susceptibility (Kirsch et al. 2012; Jamshed et al. 2019). However, the dominant livelihood source was not agriculture, which differs from other research findings (Jamshed et al. 2019). All communities were more economically challenged compared to their social vulnerability. In terms of infrastructure, most houses were single-story, made of mud, and semi-detached, which could be easily damaged by floods (Rana and Routray 2018). In addition, the limited access to drinking water and sanitation facilities exacerbated their vulnerability to water-borne diseases. Previous research has shown that past experience can distort flood risk perception and vulnerability (Siegrist and Gutscher 2006; Sullivan-Wiley and Gianotti 2017). In this study, most of the surveyed households were aware of the evacuation routes, but had limited knowledge about emergency protocols. Approximately 96% of the participants believed that the government was responsible for providing protection. However, they had more trust in the media and rarely approached local institutions for help, which is consistent with other research (Rana and Routray 2018). Additionally, about 93% of the surveyed households had never attended a mock drill and relied on traditional methods for obtaining flood information. The communities near the Chenab River had a higher attitudinal and institutional vulnerability compared to the others. Overall, 25% of the households had high multidimensional vulnerability.

5 Conclusion

Effective flood risk management requires an understanding of vulnerability from multiple perspectives. This study used an indicator-based approach to assess multidimensional vulnerability, including six main components: physical, social, institutional, environmental, economic, and attitudinal. The study also used a robust methodology to measure flood vulnerability and contributed to the literature by selecting the most suitable sub-indicators related to flood hazards. The methodology is comprehensive and can be used at different scales and in different contexts, with the sub-indicators adjusted according to the specific disaster being studied. The main findings of the study included a fatalistic attitude among the respondents and the inefficiency of local institutions, which contributed to the higher vulnerability of the selected rural communities. The respondents had little trust in government authorities due to the government’s poor management of past floods.

There are some limitations to this research that should be considered. The study area was limited to four communities in Muzaffargarh, and the respondents were only males due to cultural constraints. It is important to recognize and appreciate the role of women in developing and implementing flood risk reduction strategies. The methodology employed in this study can be enhanced by incorporating or eliminating sub-indicators as deemed necessary. Furthermore, a larger sample size would be beneficial in comprehensively understanding the intricacies of multidimensional vulnerability assessment. Despite these limitations, the study’s findings establish a benchmark for evaluating vulnerability levels in flood-prone rural communities, offering valuable insights into their capacity to withstand both direct and cascading impacts of floods. Local planning institutions can use this information to develop targeted capacity building programs, training initiatives, awareness workshops, early warning systems, contingency plans, and evacuation routes. Integrating social and institutional vulnerability dimensions can further assist these authorities in making well-informed decisions that align with existing institutional capacities and public preferences. It is crucial to emphasize that the development of integrated flood risk management strategies should be grounded in the meaningful representation of all community members, particularly women and other vulnerable groups. Future studies can also develop some tool or ranking methodology that may help the disaster risk management institutions in identifying which interventions are appropriate.