European Union
European Union's Flag

Faroe Islands
Faroe Islands' Flag

Iceland's Flag

New Zealand
New Zealand / Aotearoa's Flag

Overseas Territories
United Kingdom
United Kingdom's Flag



Island Vulnerability Background

Island Vulnerability explores the challenges which isolated geographies face when dealing with risk and disasters by examining the processes which create, maintain, and could be used to reduce their vulnerability. To understand Island Vulnerability, the following information is provided:

See also a November 2001 paper by James Lewis:
"Island characteristics and vulnerability: some perspectives"
(25 kb in Rich Text Format).

The text on this page is by Ilan Kelman unless otherwise noted. I have plenty yet to learn, so please contact me with feedback and suggestions, including omissions or errors. The views here are my own and do not necessarily represent others quoted.

Barbados Sunset.

Barbados Sunset.
(Copyright Ilan Kelman 1998.)

What are islands, isolated geographies, and small states?

A set definition for terms such as "small", "island", "small island", or "isolated" does not exist in a geographical context. Many authors (e.g., Briguglio et al., 1996; Crowards, 2002; King, 1993; Ratter and Sandner, 1996; Streeten, 1993) have discussed definitions of these terms using such criteria as population size; land area; arable land area; gross national product; environmental influence, for example defining an island to be a landmass which does not create its own climate or ecology due to its volume; and characteristics of social geography, such as the presence of a unique people or culture. Royle (2001) goes through some practical problems encountered when defining an "island" according to the dictionary.

Scoglietto, Italy.

Scoglietto, Italy: Island or rock?
(Copyright Ilan Kelman 2003.)

One suggestion would be to consider islands and isolated geographies as lacking an adequate land transportation network connected to a larger land mass. Land transportation networks for products and people tend to be cheaper than air and quicker than water transportation networks while wire-based communication and energy lifelines are easier to construct and maintain across land. For sustainability, such lifelines should be minimised along with the need for transport of products. The lack of a land transportation unit thus implies an island or isolated geography as a self-contained human system or a self-contained society. This approach to the definition may be appealing, but then encounters the quagmire of defining the key phrases "adequate", "larger land mass", and "self-contained". After all, no geographic location on the planet is unaffected by events and actions elsewhere. The only legitimate island would be the universe!

Two further provisos emerge. First, the construction of a land transportation route, normally a bridge or tunnel, may create ambiguities and "pseudo-islands". Characteristics may lie in the transition zone between those of islands and those of non-islands. Holyhead, Prince Edward Island, Skye, and Singapore each have a "fixed link" with the "mainland" and all are certainly islands. In many cases, fears were expressed regarding the loss of island characteristics due to the fixed link. Research would assist in determining how much "islandness" has been lost due to the fixed link, although "islandness" would need to be defined first. An interesting case is Orkney where many outlying islands such as Burray and South Ronaldsay, are connected to the "Mainland", the main island of Orkney, by causeways with a road. Have the outlying islands lost any of their "islandness" due to the connection with the island Mainland? Has the Mainland been affected? Could contrasts be made with Shetland where few outlying islands have land connections to the Mainland?

Canvey Island, U.K. on the Thames Estuary downstream from London.

Canvey Island, U.K. on the Thames Estuary downstream from London.
(Copyright Ilan Kelman 1999.)

Another interesting U.K. example is Canvey Island on the Thames Estuary downstream from London. Most of Canvey Island, with a population of over 36,000, is below the mean high water mark, so the island is surrounded by a concrete wall. Two roads connect the island to the mainland, but they both pass through the same intersection (roundabout) just before going off-island. The sense of island community and island spirit are strong, making it difficult to claim that the fixed link makes Canvey a non-island.

The second proviso is that some islands belong to more than one state, such as Haiti and the Dominican Republic forming Hispaniola and New Guinea split between the country of Papua New Guinea and the Indonesian region (for the moment) of Irian Jaya. Irrespective of the land borders between states, Hispaniola and New Guinea could easily be considered islands. Whether or not "an adequate land transportation network" across each international border exists may be irrelevant since it would not be "connecting with a larger land mass".

In the end, it is perhaps best to answer the question "What is an island?" with an intuitive concept of a comparatively small landmass, generally without an adequate land transportation network connecting with a larger land mass. This statement is not a definition but an attempt to describe an intuitive concept. Obvious examples of islands are the Chatham Islands (New Zealand), Kiribati, Lakshadweep (India), the Philippines' constituents, St. Lucia, Sardinia (Italy), Tasmania (Australia), Tristan da Cunha (U.K.), and Vancouver Island (Canada). Africa, Brazil, Galicia, Kamchatka, Nebraska, and Tanzania are clearly not islands, at least in the sense of physical geography. Brazil and Galicia, for example, are arguably cultural and linguistic islands.

Plenty ambiguities emerge, such as Andorra, Antarctica, Australia, Ceuta (Spain), Greenland, Sri Lanka, Madagascar, Malaysia, Vatican City, and Venice. Monaco is certainly a small state, but could it really be considered to be an island or an isolated geography, other than one of decadence and ugly development? Despite the questions which may not have lucid answers, none of these locations detracts from the intuitive approach, the list of obvious islands, and the list of obvious non-islands. The fairest way of dealing with these ambiguities is to suggest that if a location or people feel part of the global island community or wish to become involved in the global island community for a specific issue, then they should not be excluded. The onus is not on the island community either to legalistically and irrefutably define itself or to include and exclude members at their whim. Instead, the onus is on any physical or human geographical entity to decide to be part of the island community. Anyone who asks would be accepted as part of the standard and essential exchange process which must pervade sustainability, development, environmental management, and vulnerability management activities.


Monaco is certainly a small state, but is it an isolated geography?
(Copyright Ilan Kelman 2002.)

Commentary by
James Lewis on 6 January 2002:

A common misconception is that islands are single entities. Island countries include both island groups and archipelagos under one national identity (though not necessarily with common ethnicity). For example, it is often put that Fiji or Tonga are each an island, whereas they both contain several hundred islands.

The experience of smallness (as distinct from its theoretical analysis) should have a part to play alongside other criteria such as population size, land area, and income. For example, the 122 km2 of St. Helena or the 400 km2 of Barbados are substantial "smallnesses" when compared to that of Tuvalu's 26 km2; but much more significant is that Tuvalu's minimal land area is further separated between its nine atolls (which are further separated again) over oceanic distances of 650 km. Although all are "small" (islands or states), the experience of their land formations, whether by visitors or indigenous inhabitants, is vastly different--a difference which it is not (yet) possible to discern from the present categories. The same issue would apply to any island group or archipelago; the prevailing differences that exist between dispersive land forms and contiguous continental countries are surely significant.

Would an additional category of "islands" be useful, with an indication of (e.g.) "dispersive" or "entire" to begin to convey the fragility and the remoteness that is a part of a special experience of smallness? "Remoteness from what" (e.g. continental land mass or capital/other islands) would need to be considered. A more precise grouping may be required than by seemingly arbitrary cut off points--or by self-appointed memberships or political affiliation, such as that of the Pacific Islands Forum Secretariat being inclusive of both Tuvalu and Papua New Guinea. Smallness may be more a complex matter than simply size.

Commentary by Stefan Paetow on 4 February 2005:

My definition is that if natural processes or events isolate a piece of land from the rest of the land mass, it becomes an island. When human intervention returns enough land mass that renders the gap between the island and the land mass invisible, then it no longer should be considered an island.

Canvey Island is an island. Although it is connected by bridges, it remains an island as the bridges do not constitute enough of a land mass to render the gap between Canvey Island and the rest of Essex insignificant. If the creeks between Canvey and Essex were to be filled in, then Canvey Island could not any longer be considered an island per se.

Mumbai, India faces the same dilemma. Mumbai is an island city (it is physically located on the island of Salcette), and the landscaping and land reclamation by India over the years combined the seven islands off the coast into one big island. If the two creeks that separate Salcette from the mainland were to be reclaimed and landscaped, Salcette would cease to exist as an island.

The same goes for the Florida Keys which are connected to the mainland by the island highway, but would still be discrete islands if the highway were to be destroyed.

Monaco is not an island. It is geographically isolated, but it still forms part of the European land mass. Ditto for Ceuta and Gibraltar for example. They may not have any road connection to the mainland (in Gibraltar's case it was more a political gesture than a physical non-connection), but they are physically part of the land mass they are based on (Ceuta is part of Africa geographically).

Fiji, Tonga, and others are archipelagos of islands. New Guinea and Hispaniola are islands, as are Sumatra, Madagascar, Sri Lanka, and Barbados. Australia is a contradiction. It is an island continent, as is Antarctica. Greenland is an island also.

Response by Ilan Kelman on 4 February 2005:

Thank you for your comments which yield useful insights and, in particular, simplicity. If I could ask the following questions about your definition:

1. Your definition implies a starting point with one big land mass, parts of which split off due to a natural process or event. Islands, such as Surtsey off the coast of Iceland, are born from the ocean due to tectonic activity and were never part of an original land mass (continental shelf excepted). Your definition does not specifically exclude such islands, but nor does it encompass such islands. Is that an issue?

2. How much "separation" or "split" would be required? If the creek separating the island from the land mass is one centimetre wide, is there an island? And if that creek dries up in a drought? If the creek is tidal, ranging from dry to inundated, would we have a "tidal island"? What about islands which are entirely inundated at high tide?

3. How does comparative size of the island and land mass enter the definition? Presumably, the island's area is less than the land mass area. In considering islands of islands (and islands of islands of islands), many cases exist where both the parent and the child are similar size. Which is the island and which is the land mass?

4. If New Guinea and Hispaniola are islands, what are their respective land masses? What is the land mass of Tristan da Cunha, Pitcairn Island, and New Zealand's North and South Islands? Presumably, Australia and Antarctica as island continents do not have respective land masses.

These questions are, of course, nitpicking and pedantic. They simply illustrate the challenges inherent in defining "island" and how particular one must be in order to have a consistent, fully-embracing definition, which in the end will be rarely succinct. Thank you again for your comments and interest.

Response by Stefan Paetow on 4 February 2005:

In answer to your questions:

1. No issue there, because natural processes are involved. If an island is born by volcanic process (as the Canary Islands or Surtsey were, for example), then it is an island. When a volcanic flow (as an example) bridges the separation between the landmass and the island, then the island ceases to be an island.

2. It should be discernible and identifiable as being a proper separation between landmass and island, nominally involving a discrete body of water. Does a canal count? Probably not as it is artificial. As for the question about a tidal creek, yes, I believe a 'tidal island' would be an acceptable term. I believe that Mont-Saint-Michel in France is such a case (see At high tide, it is an island; at low tide, people can walk across the sands from mainland France. The causeway built in the late 1800's blurs the definition somewhat (Mont-Saint-Michel is also described as a 'quasi-island') as it has caused some natural changes which have turned the island into a quasi part of the mainland (described at That URL indicates that France is attempting to reverse the altered current flows that, over time, have changed Mont-Saint-Michel into what it is today. The goal is to restore it to what it used to be, a tidal island.

3. I believe a land mass by default (if it exists) would be assumed to be of larger size than the island (or else the island would become the land mass, the smaller body would become the island). If there are islands of islands, all of similar size, then they become a group of islands (such as the changed Keys on the west coast of Florida after hurricane Charlie).

4. Islands that are not associated with a land mass in close proximity most likely constitute their own land mass. Australia and Antarctica are their own land mass as they are island continents. New Zealand, Hispaniola and New Guinea are probably remnants of natural processes (tectonic activity?). Ditto for Pitcairn, Hawai'i (the group of islands), the Canaries, the Azores etc.

A succinct definition of "island" probably would be "a land mass surrounded by a body of water". However, the definition of "land mass" and "body of water" are up for debate, again. Which is why this subject is incredibly fascinating. Where does one eventually draw the line?

Île St. Louis, in the Seine in Paris.

The eastern end of Île St. Louis, in the Seine in Paris. A small island in a narrow river with three bridges across it.
(Copyright Ilan Kelman 2004.)

What is vulnerability?

Discussion that consolidates definitions and key references are provided in the document Understanding Vulnerability to Understand Disasters (92 kb in PDF) or as an RTF file that should open in Word (136 kb).

Discussion on inadequate science regarding vulnerability and resilience is provided in the document Critique of Some Vulnerability and Resilience Papers (84 kb in PDF) or as an RTF file that should open in Word (149 kb).

Some case studies supporting this text appear as the anthology from James Lewis titled The Creation of Cultures of Risk: Political and commercial decisions as causes of vulnerability for others (79 kb in PDF).

Whereas IPCC's definition of "vulnerability" focuses almost exclusively on climate change as defined by IPCC and requires definitions of other phrases such as "adaptive capacity", UNISDR's definition is more generic, practical, communicable, and straightforward. IPCC's definition of "resilience" is even worse, full of jargon, whereas UNISDR uses almost the same concepts but with language which is easier to understand.

The most important difference, though, is IPCC implying that a specific measure can be taken of vulnerability ("the degree to which") in contrast to UNISDR including "factors and processes". That latter point suggests that "vulnerability" is more than a snapshot in space and time. This focus on the process, while noting that climate change is considered to be a "hazard" within a disaster risk reduction context, connects better with the detailed and comprehensive literature on vulnerability (Hewitt, 1983; Lewis, 1999; Mileti et al., 1999; Oliver-Smith, 1986; Wisner et al., 2004), the full depth of which IPCC does not always consider.

Another example of a relevant definition, although it is a rather superficial approach that does not factor in past literature, is from coastal management (after Capobianco et al., 1999 and Klein and Nicholls, 1999) that vulnerability is, mathematically, a function of and, non-mathematically, is composed of:
1. Resistance, the ability to withstand change due to a hazard.
2. Resilience, the ability to return to the original state following a hazard event.
3. Susceptibility, the current physical state, without taking into account temporal changes.
Capobianco et al. (1999) and Klein and Nicholls (1999) define vulnerability for only the natural environment, but their definition may be interpreted for individuals, societies, structures, infrastructure and social networks, communities, cultures, and the built environment.

An example of a coastal settlement in a tropical cyclone illustrates this definition's applicability. The people have resistance in their ability to live through the tropical cyclone without having nightmares for the next year and without being fearful each time the wind blows more strongly than usual. The houses have resistance in their ability to withstand structural collapse due to wind speed, debris, and external water pressure from rain or a storm surge. The people's resilience appears through their ability to carry on with their normal routines soon after the tropical cyclone has passed. The houses also have resilience in its material's physical properties relating to the ability to dry without damage following inundation. Finally, people have susceptibility due to gender and age distributions, prior disabilities, familiarity with survival and coping strategies, and mental and physical states when the tropical cyclone strikes. The houses have susceptibility due to their value, state of maintenance, and location.

Tyres on a Samoan House's Roof to Prevent Wind Damage.

Reducing Vulnerability to Cyclones: Tyres on a Samoan House's Roof to Prevent Wind Damage. Will the Corrugated Metal Roof Remain Undamaged?
(Copyright Ilan Kelman 2004.)

The vulnerabilities of the houses and people influence each other. People can affect their house's vulnerability by boarding up windows and securing or removing loose objects in the garden. Houses can affect the people's vulnerability by its ability to withstand damage. The difference is that houses do not make decisions and choices, whereas people do.

The potential for damage or harm to occur, though, goes beyond a specific, extreme environmental event with conditions that still overwhelm a community's ability to cope, yet which manifest every day. Examples are a low-quality water supply, energy overuse with dependence on non-renewable supplies, and inadequate waste management. Sometimes, population numbers, population densities (including urbanization), and population inequalities (determined through discrimination, livelihoods, poverty, or entitlements) are highlighted as being the chronic vulnerability or chronic disaster faced. The longer-term vulnerability processes could be interpreted as leading to "disaster conditions", or "disastrous conditions", in contrast to "disaster events".

An example of an informal, perhaps illegal, settlement illustrates the applicability of this understanding of "vulnerability". The people have resistance in their ability to raise the next generation without succumbing to the cycle of poverty, poor individual and social health, and potentially violence which often pervades such settlements. The people's resilience appears through their ability to build up a sense of community and to create opportunities for the future, despite being subjected to the continual disasters of inequity, lack of resources, and injustice. Finally, people have susceptibility due to gender and age distributions, prior disabilities, familiarity with survival and coping strategies, and their mental and physical states shaped by their living conditions.

Informal, though not illegal, settlement of displaced Afghans:  overview.

Informal, though not illegal, settlement of displaced Afghans:  inside.

An informal, though not illegal, settlement of displaced Afghans. An island in the desert? Aside from the catastrophic, immediate disasters of drought and war, the population faces chronic disasters including poverty, repression, and lack of education.
(Copyright Joseph Ashmore 2002.)

Variations on the definition of vulnerability exist. For example, in considering property, susceptibility may be interpreted as "exposure", the accumulated value and proximity to hazard of buildings (after Crichton, 1999). Other authors define vulnerability as being susceptibility. Resistance and resilience are important and must be considered but are not part of vulnerability. In contrast, UN DHA (1992) does not worry about the constituents of vulnerability, instead defining vulnerability quantitatively as "Degree of loss (from 0% to 100%) resulting from a potentially damaging phenomenon", a form of proportional vulnerability buildling on earlier work.

SOPAC (2002) takes a different view:

  • Vulnerability: Vulnerability is the tendency for an entity to be damaged.

  • Resilience: Resilience is the opposite of vulnerability and refers to the ability of an entity to resist or recover from damage.

  • Entities: Entities can be physical (people, ecosystems, coastlines etc) or abstract concepts (societies, communities, economies, countries etc) that can be damaged (responders).

  • Vulnerability and resilience: Vulnerability and resilience are two sides of the same coin. Something is vulnerable to the extent that it is not resilient.

  • Overall vulnerability (OV): Overall vulnerability (OV) is the result of many vulnerability factors working together. For example, we might be concerned with the OV of a country. It includes information on the risk of hazards, natural resilience and acquired vulnerability.

Lewis (1999) avoids definitional games by providing an insightful summary into, and expansion of, not only what vulnerability means but also how vulnerability should be interpreted and used:

Victims are often said to have been 'in the wrong place at the wrong time'. Vulnerability, as the degree of susceptibility, may be interpreted as simply a matter of location or place--and some places are more vulnerable than others. There is much more, however, to the understanding of vulnerable conditions than their physical recognition and identification or, for that matter, than in physical resistance to natural forces in constructional and infrastructural technology. Social and political issues may have had a greater part to play.

Vulnerability, similarly, is often interpreted as a physical state of exposure related to location and quality of construction. Vulnerability becomes manifest when locations and constructions are seriously affected by storm, flooding or landslide for example, and when some poorer areas of construction sustain more damage than others. The external pressures over which victims may have no control, which prescribe such locations, and the lack of options as to where they live or their quality of life, as well as construction, are examples of how vulnerability is related to overall policies and activities in the control of others. To seek modifications through development programmes and projects to avoid or reduce such pressures, is one effective objective of vulnerability reduction in development.

The recognition and identification of locationally or socially vulnerable sectors of populations is itself only an indicator of the processes that have brought about those conditions. They are the visible and tangible manifestations on the surface, so to speak, of invisible and intangible social, economic, and political undercurrents; and they will have been active remotely and indigenously, contemporarily and historically.


Vulnerability as the condition of exposure to the initial impact and its immediate effects is only a part of the overall, pervasive and negative condition of vulnerability.

Vulnerability is not only about the present state, but also about what we have done to ourselves and to others over the long-term, why and how we have done that in order to reach the present state, and how we may change the present state to improve in the future.

Taal, Philippines, September 1965.

Taal, Philippines, September 1965. The volcano is now dormant, but small farming and fishing communities surround the crater (Pyle, 1998). What vulnerabilities exist? What future impacts may occur? How might the situation be changed?
(Public domain photo courtesy of the National Geophysical Data Center's photo library, National Oceanic and Atmospheric Administration, Department of Commerce, U.S.A Government, photographer J.G. Moore USGS.)

What significance does island vulnerability have?

Given the previous discussions of "island" and "vulnerability", "island vulnerability" straightforwardly refers to the potential for harm to come to locations deemed to be islands. More than the present state of potential harm, however, must be covered. "Island vulnerability" also examines why that state of potential harm exists in such a relatively small environment, how it arose, the advantages and disadvantages of its existence, how it should be changed, and how that change might be effected within the challenges which islands face. As per Lewis (1999), the processes which created that vulnerability and the processes which that vulnerability creates are examined in the island context

Pelling and Uitto (2001) quantify island vulnerability by using socioeconomic data to develop a vulnerability index or indicator number by which islands may be ranked. They conclude that "The larger, and least globally connected island states are those most severely affected by disaster (Haiti, PNG, Jamaica). Although it is the smaller islands that are most at risk from 'knock-out' by a single event." The vulnerability experienced by islands is aptly illustrated.

UN DHA (1992) also provided a quantitative definition of vulnerability: "Degree of loss (from 0% to 100%) resulting from a potentially damaging phenomenon". When considering vulnerability as a percent loss, the importance of islands becomes apparent, as Lewis (1999) and Mossler (1996) note in discussing "proportional impact" which also refers to "potential proportional impact" or "proportional vulnerability". Proportional impact and proportional vulnerability report percentages; for example, the percent of the population killed or affected or the percent of the infrastructure damaged. In contrast, disaster magnitude reports numbers; for example, the number of people killed or affected and numbers of buildings damaged or destroyed.

The total population of Montserrat is approximately half the death toll from the 26 January 2001 earthquake in Gujarat, India. Superficially, it appears that the tragedy in India far surpasses any event which could possibly afflict Montserrat. Proportional impact, however, shows the territory of Montserrat to have far greater vulnerability than the state of India. Since Montserrat's volcano started erupting in 1995, every single person on Montserrat, 100% of the population has been directly affected. More than 50% of the population has migrated away from the island. The 25 June 1997 pyroclastic flow killed at least 19 people, which proportionally would be equivalent to more than one million people dying in an earthquake in India. Similarly, close to 100% of Montserrat's main infrastructure has been destroyed.

Bethel, Montserrat after the 25 June 1997 pyroclastic flow and surge.

Bethel, Montserrat after the 25 June 1997 pyroclastic flow and surge
(Public domain photo courtesy of the
United States Geological Survey, Volcano Hazards Program, photographer R.P. Hoblitt USGS.)

Lewis (1999) provides more examples:

The tropical cyclone which swept into the Indian state of Andhra Pradesh in 1974 rendered two million people homeless, 4 per cent of the state population and less than 1 per cent of India's national population. When, in 1972, 120,000 people were made homeless by hurricane Bebe in Fiji, they represented more than one-fifth of the national population. Eighty per cent of Dominica's housing stock was destroyed by hurricane Allen in 1980. Twenty-two per cent of Tonga's housing stock was destroyed and 50 per cent of the national population were made homeless by hurricane Isaac in 1982".

The issue is not to downplay the disasters in India, but to give the island disasters the prominence and importance they deserve. The events in India were horrific, requiring far more mitigation and response activities than were experienced. We should be outraged at the vulnerabilities in India and we should be working towards avoiding the recurrence of such catastrophes. Similarly, the events on the islands were horrific, requiring far more mitigation and response activities than were experienced. We should be outraged at the vulnerabilities on the islands and we should be working towards avoiding the recurrence of such catastrophes. Examining proportional impact and proportional vulnerability reveals the true importance of events on islands.

Therefore, island vulnerability encompasses "proportional impact" and "potential proportional impact" / "proportional vulnerability" as indicators of the importance of a disaster or potential disaster. Island vulnerability investigates the processes which produce the proportional impact and proportional vulnerability, of which the small size of islands is an obvious element, and, more importantly, looks at ways in which vulnerability could be reduced by lessening proportional impact and proportional vulnerability in such a small environment.

Luxembourg Flood Levels.

Luxembourg Flood Level, 9 February 1756.

Luxembourg Flood Level, 4 March 1806.

Floods in the small state of Luxembourg. The top sign indicates the flood level in 1756. The bottom sign indicates the flood level in 1806. (Copyright Ilan Kelman 1999.)

Why islands?

The Island Anthology (Lewis, 2003) along with Kelman (2003b) illustrate the fascination and importance of islands from many perspectives. Furthermore, the physical and psychological isolation of islands tends to prioritise them disproportionately low in comparison to their importance. Reasons for neglecting islands include small size, lack of land-based resources, and relative inaccessibility, yet these same characteristics make islands more unique and more vulnerable. Pelling and Uitto (2001) also note "the literature on SIDS [Small Island Developing States] and natural disaster vulnerability has an under-developed critical voice" and ask why islands are vulnerable with the answer:

Small size
Limited natural resource base, high competition between land use, intensity of land-use, immediacy of interdependence in human-environment systems, spatial concentration of productive assets
Insularity and remoteness
High external transport costs, time delays and high costs in accessing external goods, delays and reduced quality in information flows, geopolitically weakened Environmental factors
Small exposed interiors, large coastal zones
Disaster mitigation capability
Limited hazard forecasting ability, complacency, little insurance cover
Demographic factors
Limited human resource base, small population, rapid population changes, single urban centre, population concentrated on coastal zone, dis-economies of scale leading to high per capita costs for infrastructure and services
Economic factors
Small economies, dependence on external finance, small internal market, dependence on natural resources, highly specialis[ed] production

When the risk in a vulnerable situation manifests, such as in the form of a disaster, an island's insularity tends to preclude a timely response with the needed resources. The result is worse consequences than would occur at another location experiencing a similar situation.

Islands are often shaped by their remoteness, or insularity, which develops ecologies and cultures that are usually unique to their location. Islands tend to have fragile environments, fragile economies, and are highly vulnerable to some of the most devastating hydrometeorological and geological disasters. One of the most lethals tsunami of the 20th century struck Papua New Guinea on 17 July 1998 killing more than 2,000 people. The most lethal volcanic eruption of the 20th century killed approximately 28,000 people on Martinique when Mount Pelée erupted on 8 May 1902.

Islands are not immune to some of the most devastating disasters not involving environmental phenomena:

  • The world's worst peacetime sea disaster to that date occurred in 1987 in the Philippines when the overloaded ferry Dona Paz collided with an oil tanker, killing more than 3,000 people.

  • A Korean Air Lines Boeing 747-3B5 crashed in Guam on 6 August 1997 killing 228 people. 26 survivors meant a challenging mass casualty / major incident scenario.

  • Terrorist bombs on Bali on 12 October 2002 killed 202 people.

Moreover, islands often experience longer-term, more chronic vulnerabilities such as maintaining adequate water and energy supplies, preventing emigration which depletes the population and removes a needed skill base, maintaining self-sufficient economies, and preserving their culture.

Such characteristics of island geography yield challenges which have common themes throughout all islands yet which produce an incredible diversity of environments and societies. Thus, a strong international island community may be built and islands readily lend themselves to specific studies of and solutions for their problems.

These difficulties and challenges may lead to advantageous outcomes. For example, the vulnerability of isolation implies an impetus towards developing local, small-scale, renewable energy sources which reduce vulnerability wherever they are used. The same argument applies to other resources because the potential exists for focusing on self-sufficiency and locally sustainable resource management. Water management, for example, would be greatly assisted by sustainable desalination technology and many islands have a clear impetus towards developing such solutions. Such solutions may then be applied to non-island geographies.

A further advantage of islands is that they provide the compactness and isolation needed to fully analyse and communicate the inter-relationships amongst vulnerability, sustainability, and risk. The ability to demonstrate and to articulate the processes which create, maintain, and could be used to reduce vulnerability can be learned through island study and then apply to non-island geographies. The transferability of lessons from islands to other locations, particularly given the innovative solutions islands may develop, is a vital outcome from island studies.

Some specific examples are coastal zones and volcanoes. Due to their proximity to the sea and the affiliated vulnerabilities, islands and coastal zones display similar characteristics. Thus, islands yield lessons for all coastal zones, and vice versa. Similarly, since many islands are volcanoes, the lessons learned from island volcano management may be applied to non-island volcanoes.

Islands are examined due to their increased importance, their increased vulnerability, and the lessons which can be gained.

Suva, Fiji Bus Station.

Suva, Fiji Bus Station.
(Copyright Ilan Kelman 2004.)

Why vulnerability?

The words of Lewis (1999) and SOPAC (2002) elegantly answer this question.

Lewis (1999):

Vulnerability is the product of sets of prevailing conditions within which disasters occur. Vulnerability has to be addressed therefore, not only by post-disaster concern and response, but as a part of the day-to-day management of change--whether or not that change is called development. The pervasive conditions of vulnerability cannot be allocated as the responsibility of one desk or department; they are the prerogative of all desks and departments for all kinds of business and activity, both policy and practice. In these policies and activities may otherwise ferment the causative conditions for disaster, which so often go unrecognized and unattended until disaster has happened.


Consideration of vulnerability...looks at the processes at work between the two factors of hazard and risk. It reverses the conventional approach, and focuses upon the location and condition of the element at risk and reasons for that location and condition of infrastructure, construction, community, dwelling, population or person. Focus upon source or origin of hazards maintains a difference between them; whereas focus upon vulnerability maintains a similarity of sustained effects as a result of hazards experience By attending to vulnerability, the effects of all potential hazards can be accommodated to some degree--from the point of view of the victim's potential to survive and to recover. Measures to decrease vulnerability are partially integrative with normal collective conditions, small-scale, and individually are less costly and more achievable.


Risk is static and hypothetical, (though reassessable from period to period of time) but vulnerability is accretive, morphological and has a reality applicable to any hazard. It is not dependent upon, or applicable to, only specified sources of hazard.

Therefore, it is not only risk that needs to be considered; it is also vulnerability. Vulnerability is actual; risk is actuarial. Whether or not risk is appropriately assessed, vulnerability will accrue or change. Conversely, changes to susceptibility, either planned or fortuitous, may cause vulnerability to increase, diminish or to be held in check--independently of assessments of risk.

SOPAC (2002):

Vulnerability management is emerging as a critical part of any sustainable development strategy. It focuses not only on conditions now, but also on likely conditions in the future. It examines risks of hazards, natural and acquired abilities to resist damage (natural resilience and acquired vulnerability), giving us the opportunity to balance strengths and weaknesses.


The vulnerability of our environmental, social and economic systems is made up of more than just the risk of disasters and good or bad management. It is not just about climate change, or globalisation, or trade agreements. It must also include an understanding of how well any system (environmental, social and economic) can cope with any hazards that may come its way and that might harm it. It would be impossible to work towards good quality of life and growth for countries under a sustainable development model if no account were made of the damage that can occur from internal and outside influences.

For development to be sustainable, we clearly need to learn to manage our vulnerabilities. We need to be able to understand and/or manage hazards, natural resilience and acquired resilience. This understanding for the first time opens up opportunities for improving our overall vulnerability because it forces us to examine the problem from all angles, instead of just focusing on the risk of disasters. Vulnerability management is emerging as a critical part of any sustainable development strategy.

The interesting thing about vulnerability is that it can be examined at different levels for different issues. That is, it can be used to look at a single issue, or to assess a complex entity such as a country.


Vulnerability is a new way of looking at an age-old problem. Instead of focusing just on what has been going wrong in the past and the effects of hazards, vulnerability gives us the opportunity to focus on getting things right for the future. As a future-focused approach, vulnerability is a way of using strengths and strategically improving weaknesses.

An tree uprooted by Cyclone Heta in January 2004 in Apia, Samoa.

An tree uprooted by Cyclone Heta in January 2004 in Apia, Samoa.
(Copyright Ilan Kelman 2004.)

Why island vulnerability?

Much of the justification for examining island vulnerability has been provided in previous answers. Lewis (1999) reiterates and summarises the arguments:

Island countries and countries of islands have, in their relative smallness, an extraordinary vulnerability. Tropical cyclones in their destructive power can engulf entire island groups and cause devastation on a proportional scale unknown in larger and sub-continental countries. At the same time, their diverse and scattered smallness, in archipelagos for example, has special implications for administration and management which provide both constraints and opportunities for development strategy.


The characteristics of place have a significant bearing on the identification of development strategies anywhere--economically, socially and culturally. The place will have had its influence upon local culture and vice versa. Local and regional, popular and official perceptions and analysis of them, are thus necessary for understanding local and regional influences upon vulnerability and strategies to counter them--whether in the aftermath of one disaster, or contexts of vulnerability to the next.

In the shifting interplay of contributing factors of vulnerability, what happens in islands will be much the same as what could happen at local level anywhere. The difference is that islands are immediately 'local level' and appropriately identified strategies are implementable there, being manageable and small scale. Islands offshore of a mainland or continental country especially require processes of development designed on their behalf, and with their participation, rather than to share only in prevailing national programmes.


Large disasters in large countries and in urban areas have nevertheless become established as the basis for public and official opinions and action concerning generalized response. It is these disasters that have become either the emotive persuader, on the one hand or, on the other hand, the equally persuasive medium for despair with regard to the apparent impossibility of doing anything about disasters of such destructive and disruptive power.

In either event the small disasters, which recur much more frequently than the large ones, and affect similarly large numbers of people in total, escape attention and escape international action. It is these disasters in the islands of the archipelagos, in remote mountain villages and in the fishing communities on the maritime coastlines that are losing out to the international global programmes. Islands could inform the continents, were they given the chance, by a reversal of this unfortunate global norm.

Volcanic landform islands viewed from Oahu, Hawai'i

Volcanic landform islands viewed from Oahu, Hawai'i
(Copyright Ilan Kelman 2005.)

Alternatives to Vulnerability

Some people dislike the term "vulnerability". Sometimes definitions, connotations, and understandings of "vulnerability" vary but at other times the term is misunderstood. In particular, vulnerability is often seen as just the current state, in a sense referring to what society is at the moment regarding characteristics such as its fragilities, weaknesses, and susceptibilities. In contrast, this website considers vulnerability not only as the current state but also the process by which that current state was reached and the direction in which the current state is heading. The "vulnerability process" refers to the actions, behaviours, values, ideas, and systems which have led to characteristics such as fragilities, weaknesses, and susceptibilities and which can perpetuate or absolve these issues. To absolve these issues, aspects including resistance, resilience, capacity, capability, strength, power, empowerment, and sustainability are necessarily addressed by vulnerability.

Nonetheless, some people still contend that the word "vulnerability" is too technocratic, negative, or otherwise inappropriate, especially if it would be frequently misunderstood. Other possible phrases or ideas which refer to, encompass, or complement Island Vulnerability as defined by this website are Island Affairs, Island Capability, Island Capacity, Island Empowerment, Island Power, Island Resilience, Island Resiliency, Island Risk, Island Strength, and Island Sustainability.

Language is powerful and terminology is important. "Vulnerability" through the vulnerability process has been deliberately chosen for this website for the reasons described on this page. This choice does not imply that "vulnerability" is always the superior term irrespective of the circumstances, that other people are wrong or misguided, or that all inadequacies have been addressed. Instead, it suggests that for the perspective and interests displayed by this website, "vulnerability" is the best term, particularly with the definition provided. The future might bring changes.

Simultaneously accepting Island Vulnerability, SOPAC's (2002) focus on resilience, and others' choice of other terms is not contradictory. Instead, it is complementary, indicating the richness of island issues. Debating definitions, vocabularies, and context is important (see, for example Kelman, 2003a and, but the activities beyond word discussions are important too. Rather than becoming mired in the inadequacies of language--particularly English--or rejecting anyone's work due to word choice, continued thinking, discussion and exchange on the terminology would help to ensure that our "doing" continually helps islands.

The Main Street of Nuku'Alofa, Capital of Tonga.

The Main Street of Nuku'Alofa, Capital of Tonga.
(Copyright Ilan Kelman 2004.)


Briguglio, L., R. Butler, D. Harrison, and W.L. Filho. 1996. Sustainable Tourism in Islands and Small States: Case Studies. Pinter, London/New York.

Capobianco, M., H.J. DeVriend, R.J. Nicholls, and M.J.F. Stive. 1999. "Coastal Area Impact and Vulnerability Assessment: The Point of View of a Morphodynamic Modeller". Journal of Coastal Research, vol. 15, no. 3, pp. 701-716.

Crichton, D. 1999. "The Risk Triangle", pp. 102-103 in Ingleton, J. (ed.), Natural Disaster Management, Tudor Rose, London.

Crowards, T. 2002. "Defining the Category of 'Small' States". Journal of International Development, vol. 14, issue 2 (March), pp. 143-179.

Hewitt, K. (ed.). 1983. Interpretations of Calamity from the Viewpoint of Human Ecology, Allen & Unwin, London, U.K.

Kelman, I. 2003a. "Defining Risk". FloodRiskNet Newsletter, issue 2, Winter 2003, pp. 6-8, full text (166 kb in PDF).

Kelman, I. 2003b. "Intrigue of Islands – What are They?". Views (U.K. National Trust), issue 38, Summer, pp. 70-72, full text (353 kb in PDF).

Klein, R.J.T. and R.J. Nicholls. 1999. "Assessment of Coastal Vulnerability to Climate Change". Ambio, vol. 28, no. 2, March 1999, pp. 182-187.

King, R. 1993. "The Geographical Fascination of Islands", pp. 13-37 in D.G. Lockhart, D. Drakakis-Smith, and J. Schembri (eds.)., 1993, The Development Process in Small Island States, Routledge, London/New York.

Lewis, J. 1999. Development in Disaster-prone Places: Studies of Vulnerability. Intermediate Technology Publications, London, U.K.

Lewis, J. 2003. Island Anthology. Published on Island Vulnerability.

Mileti, D. and 136 contributing authors. 1999. Disasters by Design: A Reassessment of Natural Hazards in the United States, Joseph Henry Press, Washington, D.C.

Mossler, M. 1996. "Environmental Hazard Analysis and Small Island States: Rethinking Academic Approaches". Geographische Zeitschrift, vol. 84, issue 2, pp. 86-93.

Oliver-Smith, T. 1986. The Martyred City: Death and Rebirth in the Andes. University of New Mexico Press, Albuquerque, New Mexico, U.S.A.

Pelling, M. and J.I. Uitto. 2001. "Small Island Developing States: Natural Disaster Vulnerability and Global Change". Environmental Hazards, vol. 3, no. 2 (September), pp. 49-62.

Pyle, D. 1998. Volcanoes. Oceana / Quantum Books, London, U.K.

Ratter, B.M.W. and G. Sandner. 1996. "Small Islands, Large Questions: Introduction to Special Issue". Geographische Zeitschrift, vol. 84, issue 2, pp. 63-66.

Royle, S.A. 2001. A Geography of Islands: Small Island Insularity. Routledge, London, U.K.

SOPAC. 2002. Information from the website which is now SOPAC's Environmental Vulnerability Index. SOPAC (Pacific Islands Applied Geoscience Commission), Fiji.

Streeten, P. 1993. "The Special Problems of Small Countries". World Development, vol. 21, no. 2, pp. 197-202.

UN DHA. 1992. Internationally Agreed Glossary of Basic Terms Related to Disaster Management. UN DHA (United Nations Department of Humanitarian Affairs), Geneva, December 1992.

Wisner, B., P. Blaikie., T. Cannon, and I. Davis. 2004. At Risk: Natural Hazards, People’s Vulnerability and Disasters, 2nd ed. Routledge, London, U.K.

Traffic Islands in Wellington.

Do These Islands Count?
(Copyright Ilan Kelman 2004.)

Contact Island Vulnerability.

The material on the Island Vulnerability website is provided as only an information source. Neither definitive advice nor recommendations are implied. Each person or organisation accessing the website is responsible for making their own assessment of the topics discussed and are strongly advised to verify all information. No liability will be accepted for loss or damage incurred as a result of using the material on this website. The appearance of external links on this website does not constitute endorsement of the organisations, information, products, or services contained on that external website.