Social Science Perspectives on Hazards and Vulnerability Science

Part of the International Year of Planet Earth book series (IYPE)


What makes people and places vulnerable to natural hazards? What technologies and methods are required to assess this vulnerability? These questions are used to illustrate the circumstances that place people and localities at risk, and those circumstances that enhance or reduce the ability of people and places to respond to environmental threats. Vulnerability science is an emerging interdisciplinary perspective that builds on the integrated tradition of risk, hazards, and disasters research. It incorporates qualitative and quantitative approaches, local to global geography, historic to future temporal domains, and best practices. It utilizes technological sophistication and analytical capabilities, especially in the realm of the geo-spatial and computation sciences (making extensive use of GPS, GIS, remote sensing, and spatial decision support systems), and integrates these with perspectives from the natural, social, health, and engineering sciences.

Vulnerability research focuses on the intersection of natural systems, social systems, and the built environment. These three component areas intersect with the spatial social sciences to play a critical role in advancing vulnerability science through improvements in geospatial data, basic science, and application. The environment, individuals, and societies have varying levels of vulnerability that directly influence their ability to cope, rebound, and adapt to environmental threats. At present, we lack some of the basic operational understanding of the fundamental concepts of vulnerability, as well as models and methods for analyzing them. The focus on place-based applications and the differential susceptibility of populations to hazards is a key contribution of vulnerability science. Using examples derived from recent disasters, the role of the spatial social sciences in advancing vulnerability science are reviewed.


Vulnerability science EM-DAT SHELDUS 


We know geohazards are distributed unevenly across the earth’s surface – some regions are seismically active, while others are not; certain regions experience hydro-meteorological hazards with regularity, while others endure water deficits for most of the year. Understanding physical processes and their variability across the landscape provides the fundamental science of hazards – where, when, and why they occur, and the risks posed to society (Beer et al. 2004). While we possess some level of understanding on the distribution of natural hazards and their historical frequency, we know less about the risk (probability of an event occurring) or its likely impact on society.

Just like geohazards, the Earth’s population also is unevenly distributed on the natural landscape, often clustered near coastlines, along rivers, or in seismically active zones. Population size and location coupled with the socio-economic and demographic characteristics of that population are the drivers of societal impacts, and help explain why the same geophysical event produces quite different impacts at the local level. Societal factors intervene between nature (and the natural processes) and the built environment to redistribute the risk prior to an event, and to amplify or attenuate the losses after an event. The interaction of society, nature, and the built environment creates dangerous places (Reisner 2004), and often elevates routine hazards into disasters.

Vulnerability Science

Simply stated, vulnerability is the potential for harm or loss. It examines those circumstances that place people and localities at risk as well as those characteristics that enhance or reduce the ability of society to respond to environmental threats. While there is considerable theoretical and conceptual development within the field of vulnerability science in many different literatures (engineering, social sciences, natural sciences), most agree on the underlying construct – vulnerability is the susceptibility to harm. Understanding vulnerability to hazards, especially geohazards, requires three separate, but intersecting knowledge domains: natural systems, social systems (and the built environment), and local places (Fig. 1).
Fig. 1

Vulnerability science incorporates knowledge from the intersection of physical systems, human systems, and place-based research

Natural processes, independent of human agency, do not produce hazards, it is only when these geophysical processes interact with human populations that hazards arise (Burton, Kates and White 1993).1 Humans create hazards by altering the natural landscape and affecting natural system processes, such as locating in floodplains and altering flow regimes and runoff through landscape modification. Interactions between social systems and the built environment also contribute to vulnerability. The pressing demand for shelter in many of the world’s megacities can result in shoddy construction and poor placement of housing, much of it in high-risk areas such as floodplains or steep, unstable slopes. Once a landslide, earthquake, or flood occurs it destroys homes and human lives, yet this unsustainable pattern continues with increasing demand for new housing by the influx of new residents. The third knowledge component is the understanding of the local places – landscape, history, economics, culture, demographics, politics – in other words, the local geography. For example, in New Orleans the topography of the city reflects its social geography: wealthier residents live in higher elevations and have historically done so; while the poor and minority residents occupy the low-lying areas, the latter often subjected to the most intense flooding (Colten 2005; Kates et al. 2006).

The integration of all three knowledge domains (Fig. 1) provides the intellectual basis for vulnerability science, an interdisciplinary field that entails the development of methods and metrics for analyzing societal vulnerability and resilience to environmental hazards and extreme events (Cutter 2003). One of the primary goals of vulnerability science is to provide the scientific basis for disaster and hazard reduction policies.

The Recipe for Disaster

Prior to the implementation of any risk or vulnerability reduction policy, we must first understand the geographic variability in the hazards and their impacts; the geographic variability in the populations at risk; the vulnerability of populations; and the context in which these three interact at the local or regional scale. There are rich sources of general information on the spatial distribution of hazards and disaster risks (Dilley et al. 2005; International Strategy for Disaster Reduction 2004), as well as annual overviews of disasters, embodied in the yearly World Disasters Report (International Federation of Red Cross and Red Crescent Societies 2008). The development of robust monitoring and surveillance networks have advanced our understanding of the earth sciences and provide a rich data source for hazard event parameters such as location, magnitude, intensity, duration. Global seismic and tsunami monitoring systems, and the international suite of weather-related satellites produce enormous data streams on precipitation patterns (that can lead to floods and droughts), tropical cyclones and other wind events, and severe weather (heat and cold). The US Geological Survey, the National Climatic Data Center (NCDC), the National Geophysical Data Center (NGDC), and the Smithsonian’s Global Volcanism program routinely catalog data describing the physical attributes of natural hazards in the U.S. However, the societal impact of such events (lives lost or economic losses) is not consistently included in such databases (Gall et al. 2009).

Hazard Losses

At present, there is no systematic information on the societal losses to geohazards. These losses include direct monetary losses such as the value of destroyed homes, businesses, or infrastructure (e.g. roads and bridges). Yet societal losses also include human losses – loss of life, injuries, loss of livelihoods, and displacements from one area to another. If we are to reduce vulnerability to natural hazards, a good starting point is to determine the annual losses associated with natural hazards for each nation. Are such losses dispersed throughout the country or are they concentrated in one or more specific areas? Are the losses a consequence of a singular, infrequent catastrophic event, or are they the result of periodic smaller scale (and impact) events that overtime add up to significant losses for the country? In other words, what are the hazard loss profiles of nations?

Proprietary data collected by reinsurance companies such as Munich Re and Swiss Re computes insured losses from large-scale disaster events. These insurance databases provide only a limited picture of losses since low density and low capitalized countries are often excluded, especially countries in Asia, Africa, and Latin America. Within the more developed world regions, there is better coverage of urban versus rural areas, but the monetary impact of slow onset hazards such as drought are not included because of the difficulty in assessing monetary damages. At present, there is no systematic open-access inventory or accounting for individual nations or aggregated to the global scale of hazard events and losses by location or by hazard agent. How can we reduce the societal impact of natural hazards when we lack fundamental information on how large the losses are and where they occur?

There are two on-going efforts to produce such an accounting – the EM-DAT disaster database managed by CRED for nations with global coverage; and SHELDUS, a US-centric database for hazard events and losses at the county scale. These are described next.


EM-DAT is a global emergency events database developed and maintained by The Centre for Research on the Epidemiology of Disasters (CRED) at the Université Catholique de Louvain (Brussels, Belgium) ( This country-level database covers the period 1900–2007, but the majority of entries are from 1975 to present. The database includes information on people killed, people affected, direct and indirect economic damages. It can be sorted by date and by hazard category using three main groupings: natural hazards, technological hazards, and complex emergencies. Under the natural hazards category, the specific hazards include earthquakes, tsunamis, epidemics (viral and parasitic), floods (flash, general, storm surge/coastal), mass movements (landslides), storms (local, tropical cyclones), volcanic, and wildfire (forest fires and shrub/grassland fires). Beginning in 2003, each disaster event was issued a GLIDE (GLobal IDEntifer) number. GLIDE numbers provide a consistent identifier so scholars and practitioners can link information across databases on specific disasters and their impacts ( In EM-DAT, where the disaster affected more than one country, the same GLIDE number is used, but the losses are attributed to the country where they occurred. EM-DAT employs a threshold for inclusion into the database that each event must meet. The criteria include more than 10 people killed, more than 100 affected, a declared state of emergency, or a call for international assistance.

EM-DAT is a searchable database that can create either country profiles or hazard specific profiles. The database is user friendly, although there are limitations in accessing the raw data. Instead, country profiles contain summary statistics by event type for a specified time period (1900 to present; 1975 to present). For example, as shown in Fig. 2, there is an upward trend in disaster damages, globally, and a downward trend in hazard-related fatalities. While the data are the best available, globally, it is often difficult to download individual year data by country, but this functionality should improve with new versions and updates on the web site.
Fig. 2

International trends in disaster losses. Source: EM-DAT: The OFDA/CRED International Disaster Database, Université Catholique de Louvain, Brussels, Belgium (

The CRED EM-DAT database is the best available for country-level comparisons of disaster impacts, both temporally and spatially. The data are routinely used by international aid and disaster relief organizations. In fact, EM-DAT has been part of the World Disasters Report published annually (IFRCRC 2008), and provided the statistical background for the World Bank’s Natural Disaster Hotspots report (Dilley et al. 2005).


In the U.S. many different federal and state agencies collect hazard data such as the US Geological Survey (geophysical and hydrological data), and NOAA (atmospheric and hydrometerological data). As noted above, only some event parameters are monitored (magnitude, frequency, location in long/lat), while others (deaths, injuries, or economic losses) languish. At present, the US lacks any baseline information on such loss patterns. Despite repeated calls beginning in 1999 for a national inventory of hazard-losses (Mileti 1999; National Research Council 1999), and despite its experience with the costliest single disaster ever (Hurricane Katrina estimated at more than $100b in losses), the U.S. still does not have standardized loss inventory.

An independent research effort has tried to remedy this situation through the development of the Spatial Hazard Events and Losses Database, U.S. (SHELDUS) ( This is a geo-referenced (to the county level) database of natural hazard events and losses for the U.S. from 1960 to present. It includes 18 different natural hazards (Table 1), location (state and county), deaths, injuries, property losses, crop losses, and beginning/end dates. The database is searchable by hazard type, location (state and county), date, major event names (e.g. Hurricane Katrina), Presidential Disaster Declaration number, and GLIDE number. The latter is an important feature, as it links SHELDUS into international databases using a globally common identifier for specific disaster events. In SHELDUS, the output from the user-initiated query includes beginning date, hazard type, state, county, injuries, fatalities, property damage, and crop damage. The economic damages reported are in period dollars, but there is an inflation adjustment, which computes the losses to current dollars (e.g. $2007). The records (which now exceed 450,000) are downloadable in multiple formats for ease of use in statistical programs. Lastly, the SHELDUS website ( provides metadata (compliant with the U.S. Federal Geographic Data Committee standards), frequently asked questions (FAQs) to aid navigation through the site, and an annual year in disasters report, called the SHELDUS Clock.
Table 1

Hazards included in SHELDUS database






Severe Storm/Thunderstorm











Hurricane/Tropical Storm

Winter Weather

SHELDUS was constructed from U.S. federal government data sources, notably from USGS and National Climatic Data Center records. It includes any event that resulted in more than $50,000 in economic losses or any death during the time period (1960 to present). When a single event affected a number of different counties, the losses were attributed equally across the counties when no other specific spatial information was available on where the loss occurred. This results in fractional deaths, injuries, and dollar losses. Given changing data collection procedures over the time period by the federal agencies, SHELDUS represents a very conservative estimate of the losses (Gall et al. 2009), especially when compared to other databases. Nevertheless, SHELDUS represents the first centralized approximation of a national inventory for U.S. natural hazard losses.

Hazard losses (economic) in the U.S. are escalating (Fig. 3a) and weather-related losses account for most of them. It is equally clear from the timeline, that Hurricane Katrina was (and remains) an unprecedented event in terms of dollars lost. The large spike in 1994 represents the losses from the Northridge earthquake. The pattern for fatalities (Fig. 3b) shows the opposite trend, an overall declining pattern, but one with some periodic spikes, most notably the Chicago heat wave (1995). Other fatality peaks include the 1972 heat wave in Baltimore and flooding fatalities in the Rapid City flash flood event in South Dakota. The 1980 peak has no singular event, but rather represents a multitude of events. The leading cause of property and crop damage in the U.S. is tropical storms and hurricanes, followed by severe weather, flooding, and geophysical events (earthquakes) (Fig. 4a). For mortality, the leading hazard causes are severe summer weather (including lightning strikes), heat, and winter weather (Fig. 4b).
Fig. 3

Temporal patterns of losses in the United States, 1960–2007 for (a) property and crops damages (in billions of $2007), and (b) human fatalities.Source: Spatial Hazards Event Loss Database for the United States (

Fig. 4

Loss-causing hazards in the United States, 1960–2007 by type. The proportion of losses attributed to each hazard type for (a) $2007 reported damage of property and crops, and (b) human fatalities.Source: Spatial Hazards Event Loss Database for the United States (

Spatial trends in economic losses clearly illustrate the impact of tropical storms and hurricanes along the U.S. Gulf and Atlantic Coasts (Fig. 5). The greatest losses appear in coastal counties and in selected regions in the west – notably southern California (earthquakes and wildfires); Northern California (earthquakes); Washington state (earthquakes, flooding, and volcanic eruptions), and in Idaho (wildfire). When taking the economic losses and aggregating by causal agent for each state a hazard profile can be developed. This state-level profile (Fig. 6) graphically illustrates the leading cause of losses. As can be seen, geophysical sources dominant in the western US, while severe weather dominates in the Great Plains and Midwest. Flooding is a ubiquitous hazard, and dominates in many eastern states. Finally, the US hurricane-prone states in the Southeast and along the Gulf of Mexico are clearly visible on the map. The geographic representation of such a hazard profile (based on actual losses) is a first step in the development of targeted mitigation strategies, designed to address the most costly hazard for that place, be it a county or a state.
Fig. 5

Geographic variability in the pattern of hazard losses (property and crop) in the United States, 1960–2007.Source: Spatial Hazards Event Loss Database for the United States (

Fig. 6

Individual states hazard profile. Source: Spatial Hazards Event Loss Database for the United States (

Understanding the distribution of deaths and economic damages, globally and locally, is the first step towards building disaster resilient communities. The need for baselines against which to evaluate risk reduction options is the first step. Yet we need more than just event-driven baseline data for the increasing trend and pattern of losses (economic and human life) is likely a function of two different phenomena. The first is the increasing value and density of property in harm’s way (or exposure), which is driving the escalating losses. The second is the increasing vulnerability of the population that could partially explain injuries and deaths.

Populations at Risk

To assess the populations at risk from natural hazards requires not only estimates of the number of people potentially affected, but also those characteristics of the population that contribute to the social burdens of risk. The latter is often termed “social vulnerability”.

For many world regions, the timely response to and delivery of disaster relief is challenging, and complicated by the lack of data about people in need of assistance. It is not uncommon for relief teams to be deployed to a disaster area without full knowledge of how many people will need aid or where they’re located relative to the impact area, let alone have information on age and gender – characteristics that are vital in the delivery of food, water, shelter, and other assistance needs. As noted by a US National Research Council report, “How many people, their characteristics, and where they are constitute critical information needed by agencies and organizations charged with disaster response. Inaccurate numbers and locations for populations can slow the relief effort and literally mean the difference between life and death (NRC 2007: 16)”. National census of population data provide the backbone for estimating populations at risk and in planning for risk management and risk reduction policies. To be useful, such data must be collected at sub-national levels and include demographic characteristics such as age, gender, race or ethnicity, and economic well-being. Such geo-referenced data must be collected at periodic intervals to insure the most accurate and timely data are available. More than 85% of the world’s population has been enumerated within a national census since 2000 (NRC 2007). Unfortunately, population growth and migration can render such censuses obsolete in just a few years. Furthermore, censuses are normally conducted for where people live (at night), not necessarily, where they work or go to school (daytime). If the disaster occurs during the day, the census counts may severely underestimate the likely affected population.

A range of methods is employed to improve such estimations of populations at risk. Remote sensing imagery can assist in determining settlement patterns and housing (Lo 2006), but is ineffective in determining how many people actually live in each structure. Global population databases such as the Gridded Population of the World (GPW) product from the CIESIN ( (Balk et al. 2006), and Oak Ridge National Laboratory’s Landscan (Dobson 2007) help to model the distribution and density of populations, thus providing an approximation of the sub-national at risk population.

Social Vulnerability

Social vulnerability describes those pre-existing characteristics of groups or conditions within communities that make them more susceptible to the impacts of hazards, in other words, those factors that shape the social burdens of risk. Social vulnerability also influences the uneven capacity of individuals, groups, or communities to prepare for, respond to, and recover from disasters. There is a wealth of empirical data (much of it based on post-event field studies of affected populations conducted over the past half century) (National Research Council 2006), which help us understand which segments of society or sections of the community are more susceptible to disaster impacts than others. For example, age is one such characteristic. Age not only affects the mobility to get out of harm’s way without some additional assistance, but it also entails the need for special care for both children and elderly, especially if they are infirmed or frail. Thus, both ends of the age spectrum, the very old and the very young, tend to increase social vulnerability in communities where they constitute a significant proportion of the population (Heinz Center 2002). Another example is socioeconomic status. Wealth influences the ability to absorb losses and recover from disasters. Wealthier groups and communities have greater assets but they also have insurance and other financial reserves to absorb the loss, enabling faster recovery after the disaster. Poorer subpopulations and communities have less material goods to lose, but the impacts of any loss are disproportionally greater, and the ability to recover compromised. Thus, poorer communities are more socially vulnerable than wealthier places. The impact of Hurricane Katrina on the poor residents of New Orleans illustrates the differential social vulnerability and the role of inequality in disasters (Laska and Morrow 2006; Brunsma et al. 2007). A third example is gender, long recognized as influencing vulnerability. Gender-specific employment, lower wages, lower status, and women’s roles as caregivers combine to create disproportionate impacts from hazards and disasters on women (Enarson and Morrow 1998).

There is considerable interest in the development of robust metrics to measure the multidimensional concept of social vulnerability. The development of social vulnerability indicators has progressed at the international level with the work of Birkmann (2006) and King and MacGregor (2000). At a national scale, the social vulnerability index (SoVI) (Cutter et al. 2003) provides a comparative assessment of social vulnerability at the county level for the U.S. The statistically determined metric consists of 42 socio-economic and demographic variables reduced through factor analysis into a series of dimensions that are then summed to create the overall index score. Nearly 75% of the variance in the data is explained by the index, which has been replicated for five different census dates (1960–2000) with the same level of explained variance (Cutter and Finch 2008). Numerous sensitivity analyses at other spatial scales (census block group, census tract) suggest that the SoVI is a robust algorithm for assessing the comparative level of social vulnerability of places (Schmidtlein et al. 2008). The geographic depiction of social vulnerability highlights the clustering of high vulnerability counties where the driving factors that produce such vulnerability include low socioeconomic status, age extremes (children and elderly), and higher levels of density in the built environment. It is worth noting that Orleans parish (where New Orleans is located) was among the most socially vulnerable counties in the US and thus, it came as no surprise the enormity of the impact on that population (Cutter and Emrich 2006). The SoVI has wide applicability ranging from research and replication, to applied emergency management practice as a component of disaster mitigation plans (

Place-Based Science

People and the communities where they live and work are integral parts of the natural hazards system as are the physical systems (Haque and Etkin 2007). The mechanism for integration of physical processes and human systems is through the suite of geospatial tools to describe such places. There are many considerations in understanding the nexus of physical and social vulnerability – some are unique to the geosciences, others are limited to the social sciences. In both perspectives the role and impact of scale is critical. Physical features are generally point or line patterns, while some census enumeration unit (or polygon) collects social information. Reconciling the overlapping geometries is an important function, thus the need for and use of Geographic Information Systems (GIS). There are also problems with aggregation and disaggregation biases that ultimately may mask many of the subtle differences in the spatial impacts between social and physical systems. Depending on scale, these differences may completely disappear at one scale, but when the scale is local, become readily apparent. There are also limitations on social data availability. Governments collect the most consistent social and economic data sets at a resolution or scale useful for place-based studies normally on a decadal basis through national census. Unfortunately, the data collected are not always comparable between census years, and because they are collected at various levels of geography (census blocks, counties, postal codes) due to the need to follow stringent privacy act protections. The tradeoff between finer resolution data and less frequency versus more coarse grained data collected more frequently dictates how far we can push the place-based science and understanding of the impacts of hazards.

The spatial representation of place vulnerability is a powerful heuristic for assessing the likely impact of hazards on society. For example, a study examining coastal erosion vulnerability in the U.S. integrated physical indicators (sea level rise, slope, mean wave height, erosion/accretion rate) and social vulnerability and found that the physical parameters explained most of the variability in vulnerability (Boruff et al. 2005). However, when examining the regional trends (Atlantic Coast, Gulf Coast, and Pacific Coast) different stories emerge for each geographic area. For the Atlantic and Pacific coasts, physical parameters explained more of the geography of vulnerability, while for the Gulf Coast social vulnerability was a more significant driver, especially age (elderly), and high birth rates. The physical hazards can be delineated using composite indices as noted above, in-situ measurements (1% chance flood zone), or modeled output such as hurricane storm surge inundation zones or peak ground acceleration. When overlain with social vulnerability indicators, the geography of vulnerability becomes readily apparent (Fig. 7).
Fig. 7

Intersection of social vulnerability and modeled peak ground acceleration for the U.S.

There is now a solid body of research on how the vulnerabilities based in physical systems interact with social conditions to produce hazard vulnerability. Much of this research uses a single threat source such as drought (Polsky 2004); earthquakes and tsunamis (Rashed and Weeks 2003; Wood and Good 2004; Wood et al. 2009); sea level rise (Wu et al. 2002), hurricanes (Chakraborty et al. 2005), and levee failures (Burton and Cutter 2008).

The development of multi-hazard vulnerability assessments poses a different set of challenges in both the representation of the hazards as well as the social conditions. The different hazard zones (and geometries) must be represented as accurately as possible, which means there are often methodological and scientific questions regarding such generalizations. At the same time, there are issues in the social and demographic data in terms of quality and availability as noted earlier. For example, Fig. 8 provides a place-base assessment for Richland County in South Carolina integrating both social and physical vulnerability indicators. The circular features are protective action distances from chemical facilities, while the linear features are rail and highways that potentially carry hazardous materials. The social vulnerability is greatest in the center of the county, which has a significant transient population (a major military base), and an institutionalized population (prison). Originally developed at the county level of geography (Cutter et al. 2000), this GIS-based approach to hazard vulnerability assessment has also been implemented in small island nations (Boruff and Cutter 2007).
Fig. 8

Hazards of place spatial model of vulnerability illustrating the spatial integration of all hazards physical vulnerability and social vulnerability for Richland County, South Carolina

Vulnerability Indices: Strengths and Weaknesses

Vulnerability indices are useful constructs for developing a rough assessment of the distribution and likely impact of hazards and disasters. Their utility is exploratory and diagnostic in nature, enabling the researcher or policy maker to understand the underlying drivers of vulnerability and differences between places or among social groups (Smit and Wandel 2006). Out of necessity, indices are simplifications, in the same way that models are simplifications of real-world processes and interactions. While there is no such thing as a perfect vulnerability index (or a sustainability index for that matter), they are useful in benchmarking or establishing baseline conditions, and tracking changes over time and across space.

At present, most of the vulnerability metrics are descriptive, not predictive indices and are mainly used as a representation of multi-dimensional phenomena, such as those pre-existing conditions in communities that make them susceptible to harm. The problem with validation still hampers the utility of vulnerability indices for there is still no good outcome measure. Dollar losses or mortality from natural hazards are inadequate, but temporal trends in population (including out and in-migration), employment, or households in need of assistance may provide more relevancy for linking pre-existing vulnerability to outcomes.

The second major issue with vulnerability indices is their application and the unit of analysis. Considerable attention focused on country-level indicators of environmental vulnerability as a companion to the UN’s Human Development Index (equally fraught with many of the same of scale, adequacy of measurement, validation). This top-down approach (Barnett et al. 2008; Birkmann 2006) does not capture the intra-country variability in vulnerability or the nature of the local impacts, and as Barnett et al. correctly state. The bottom-up approach (Pelling 2003; O’Brien et al. 2004) provides the local case study (normally qualitative) but often fails to situate the local within a meso-scale analysis thus limiting comparisons across different spatial units using the same metrics. Some combination of the two is perhaps the best solution – specific enough to impart knowledge of local impacts and processes; yet measured with standardized variables that are readily transportable and comparable across different spatial units.

Applications to Policy at Home and Abroad

The differential exposure of places and people to geohazards results in uneven impacts and clearly influences disparities in recovery. There are many examples of such disparities in the societal impacts, not only in the U.S. but globally as well. To achieve substantial reductions in disaster losses (lives lost, social, economic and environmental assets) the World Conference on Disaster Reduction and its Hyogo Framework for Action (ISDR 2005), determined five priority actions. These include:
  • Making risk reduction a national and local priority

  • Identify, assess, and monitor risks

  • Build a culture of safety and resilience

  • Reduce the underlying risk factors

  • Strengthen disaster preparedness

We cannot develop risk reduction strategies or policies in the absence of baseline geospatial and economic data on hazard events and losses. The enhancement of existing global and national databases such as EM-DAT or SHELDUS are critical to our understanding of the distribution of losses and the selective targeting of regions or places for immediate risk reduction strategies. Further, there is a critical need to consider and incorporate spatial and social inequities in risk and vulnerability and include efforts to address such disparities in risk reduction strategies. Before we can reduce the underlying risk factors and strengthen disaster preparedness, we need to have consistent information on the populations at risk and their social vulnerability. Finally, it is clear that the drivers of community vulnerability to hazards are a product of the interactions between physical systems and human systems. The development of social vulnerability metrics described in this chapter and the application of place-based hazard and social vulnerability analyses is the step in the right direction. However, more work is needed to enhance the integration between physical models and social processes – research and application that can only be achieved through multi-disciplinary research and practice where the social sciences have a key role.


  1. 1.

    There is a difference in usage of the terms risk and hazards between the geophysical and social science community that studies disasters, hazards and risk. Rather than going into the nuances of these linguistic differences, I chose to adhere to the definitions from the social science community. From the social science viewpoint, risk is the probability of an event occurring, while hazards include the probability of an event happening as well as the impact of that event on society. In other words, natural hazards are threats to people and the things they value and arise from the interaction between human systems and the natural processes. We follow the terminology developed by Gilbert F. White and his colleagues.


  1. Balk, D. L., U. Deichmann, G. Yetman, F. Pozzi, S. I. Hay, and A. Nelson, 2006. Determining the global population distribution: methods, applications and data. Advances in Parasitology 62: 119–156.CrossRefGoogle Scholar
  2. Barnett, J., S. Lambert, and I. Fry, 2008. The hazards of indicators: insights from the environmental vulnerability index. Annals of the Association of American Geographers 98 (1): 102–119.CrossRefGoogle Scholar
  3. Beer, T., P. Bobrowsky, P. Canuti, S. Cutter, and S. Marsh, 2004. Hazards – minimising risk, maximizing awareness. Leiden, The Netherlands: Earth Sciences for Society Foundation, International Year of Planet Earth. Online at Google Scholar
  4. Birkmann, J. (ed.), 2006. Measuring vulnerability to natural hazards: towards disaster resilient societies. Tokyo: United Nations University Press.Google Scholar
  5. Boruff, B. J. and S. L. Cutter, 2007. The environmental vulnerability of Caribbean island nations. Geographical Review 97 (1): 24–45.CrossRefGoogle Scholar
  6. Boruff, B. J., C. Emrich, and S. L. Cutter, 2005. Hazard vulnerability of U.S. coastal counties. Journal of Coastal Research 21 (5): 932–942.CrossRefGoogle Scholar
  7. Brunsma, D. L., D. Overfelt, and J. S. Picou, 2007. The sociology of Katrina: perspectives on a modern catastrophe. Latham, MD: Rowman & Littlefield.Google Scholar
  8. Burton, C. and S. L. Cutter, 2008. Levee failures and social vulnerability in the Sacramento-San Joaquin Delta area, California. Natural Hazards Review 9 (3): 136–149.CrossRefGoogle Scholar
  9. Burton, I., R. W. Kates, and G. F. White, 1993. The environment as hazard (2nd Edition). New York: Guildford Press.Google Scholar
  10. Chakraborty, J., G. A. Tobin, and B. E. Montz, 2005. Population evacuation: assessing spatial variability in geophysical risk and social vulnerability to natural hazards. Natural Hazards Review 6 (1): 23–33.CrossRefGoogle Scholar
  11. Colten, C. E., 2005. An unnatural metropolis: wresting New Orleans from nature. Baton Rouge: Louisiana State University.Google Scholar
  12. Cutter, S. L., 2003. The science of vulnerability and the vulnerability of science. Annals of the Association of American Geographers 93 (1): 1–12.CrossRefGoogle Scholar
  13. Cutter, S. L., B. J. Boruff, and W. L. Shirley, 2003. Social vulnerability to environmental hazards. Social Science Quarterly 84 (1): 242–261.CrossRefGoogle Scholar
  14. Cutter, S. L. and C. T. Emrich, 2006. Moral hazard, social catastrophe: the changing face of vulnerability along the hurricane coasts. Annals of the American Academy of Political and Social Science 604 (1): 102–112.CrossRefGoogle Scholar
  15. Cutter, S. L. and C. Finch, 2008. Temporal and spatial changes in social vulnerability to natural hazards. Proceedings of the National Academy of Sciences: 105(7): 2301–2306.Google Scholar
  16. Cutter, S. L., J. T. Mitchell, and M. S. Scott, 2000. Revealing the vulnerability of people and places: a case study of Georgetown County, South Carolina. Annals of the Association of American Geographers 90 (4): 713–737.CrossRefGoogle Scholar
  17. Dilley, M., R. S. Chen, U. Deichmann, A. L. Lerner-Lam, M. Arnold with J. Agwe, P. Buys, O. Kjekstad, B. Lyon, G. Yetman, 2005. Natural disaster hotspots: a global risk analysis. Washington D.C.: Hazard Management Unit, World Bank.CrossRefGoogle Scholar
  18. Dobson, J. E., 2007. In harm’s way: estimating populations at risk. In National Research Council, Tools and methods for estimating populations at risk from natural disasters and complex humanitarian crises (pp. 183–191). Washington D.C.: National Academies Press.Google Scholar
  19. Enarson, E. and B. H. Morrow, 1998. The gendered terrain of disaster: through women’s eyes. Westport, CT: Praeger.Google Scholar
  20. Gall, M., K. A. Borden, and S. L. Cutter, 2009. When do losses count? Six fallacies of natural hazard loss data. Bulletin of the American Meteorological Society 90 (6): 799–809.CrossRefGoogle Scholar
  21. Haque, C. and D. Etkin, 2007. People and community as constituent parts of hazards: the significance of societal dimensions in hazards analysis. Natural Hazards 41: 271–282.CrossRefGoogle Scholar
  22. Heinz Center, 2002. Human links to coastal disasters. Washington D.C.: The H. John Heinz III Center for Science, Economics, and the Environment.Google Scholar
  23. International Federation of Red Cross and Red Crescent Societies, 2008. World disasters report 2008. Available online:, Accessed 29 October 2008.
  24. International Strategy for Disaster Reduction, 2004. Living with risk. New York and Geneva: The United Nations.Google Scholar
  25. International Strategy for Disaster Reduction, 2005. World Conference on Disaster Reduction, Hyogo Framework for Action 2005–2015. Available online: Accessed 30 October 2008.
  26. Kates, R. W., C. E. Colten, S. Laska, and S. P. Leatherman, 2006. Reconstruction of New Orleans after Hurricane Katrina: a research perspective. Proceedings of the National Academy of Sciences 103 (40): 14653–14660.Google Scholar
  27. King, D. and C. MacGregor, 2000. Using social indicators to measure community vulnerability to natural hazards. Australian Journal of Emergency Management 15 (3): 52–57.Google Scholar
  28. Laska, S. and B. H. Morrow, 2006. Social vulnerabilities and Hurricane Katrina: an unnatural disaster in New Orleans. Marine Technology Society Journal 40 (4): 16–26.Google Scholar
  29. Lo, C. P. 2006. Estimating population and census data. In M. K. Ridd and J. Hipple (eds.) Remote sensing of human settlements (pp. 337–378). Bethesda, MD: American Society for Photogrammetry and Remote Sensing.Google Scholar
  30. Mileti, D. S., 1999. Disasters by design: a reassessment of natural hazards in the United States. Washington D.C.: Joseph Henry Press.Google Scholar
  31. National Research Council, 1999. The impacts of natural disasters: a framework for loss estimation. Washington D.C.: National Academies Press.Google Scholar
  32. National Research Council, 2006. Facing hazards and disasters: understanding human dimensions. Washington D.C.: National Academies Press.Google Scholar
  33. National Research Council (NRC), 2007. Tools and methods for estimating populations at risk from natural disasters and complex humanitarian crises. Washington D.C.: National Academies Press.Google Scholar
  34. O’Brien, K., R. Leichenko, U. Kelkar, H. Venema, G. Aandahl, H. Tompkins, A. Javed, S. Bhadwal, S. Barg, L. Nygaard, J. West, 2004. Mapping vulnerability to multiple stressors: climate change and globalization in India. Global Environmental Change 14 (4): 303–313.CrossRefGoogle Scholar
  35. Pelling, M., 2003. The vulnerability of cities: natural disasters and social resilience. London and Sterling, VA: Earthscan Publications.Google Scholar
  36. Polsky, C., 2004. Putting space and time in Ricardian climate change impact studies: the case of agriculture in the U.S. Great Plains. Annals of the Association of American Geographers 94 (3): 549–564.CrossRefGoogle Scholar
  37. Rashed, T. and J. Weeks, 2003. Assessing vulnerability to earthquake hazrds through spatial multicriteria analysis of urban areas. International Journal of Geographic Information Science 17 (6): 547–576.CrossRefGoogle Scholar
  38. Reisner, M., 2004. A dangerous place: California’s unsettling fate. New York: Penguin.Google Scholar
  39. Schmidtlein, M. C., R. Deutsch, W. W. Piegorsch, and S. L. Cutter, 2008. A sensitivity analysis of the social vulnerability index. Risk Analysis 28 (4): 1099–1114.CrossRefGoogle Scholar
  40. Smit, B. and J. Wandel, 2006. Adaptation, adaptive capacity and vulnerability. Global Environmental Change 16 (3): 282–292.CrossRefGoogle Scholar
  41. Wood, N. J., C. G. Burton, and S. L. Cutter, 2009. Community variations in social vulnerability to Cascadia-related tsunamis in the U.S. Pacific Northwest. Natural Hazards, DOI 10.1007/s11069-009-9376-1. Published online: 26 March 2009.Google Scholar
  42. Wood, N. J. and J. W. Good, 2004. Vulnerability of port and harbor communities to earthquake and tsunami hazards: the use of GIS in community hazard planning. Coastal Management 32 (3): 243–269.CrossRefGoogle Scholar
  43. Wu, S. Y., B. Yarnal, and A. Fisher, 2002. Vulnerability of coastal communities to sea level rise: a case study of Cape May, New Jersey, USA. Climate Research 22: 255–270.CrossRefGoogle Scholar

Copyright information

© Springer Science + Business Media B.V. 2009

Authors and Affiliations

  1. 1.Department of Geography, Hazards and Vulnerability Research InstituteUniversity of South CarolinaColumbiaUSA

Personalised recommendations