What is a disaster?

Hazards and vulnerabilities combine to form disasters (UNISDR 2009). Many environmental phenomena are simultaneously hazards and resources for society, supporting livelihoods and making living in a place viable. For instance, volcanic slopes and river floodplains are frequently productive agricultural lands, encouraging settlement. Whilst no location could be free from all environmental hazards, the smaller the land area, the more difficult it is to find settlement locations far from the most devastating environmental hazards, identifying a challenge faced by many islanders. All of Sicily (Italy) is in range of Mount Etna's volcanic ash. All of Tongatapu (Tonga) would be severely shaken by a nearby subduction zone earthquake, with all of the main settled areas in the potential tsunami inundation zone.

Rather than hazards causing disasters, instead social and political processes, circumstances and characteristics lead societal groups to be potentially harmed by hazards (vulnerability) or to be able to deal with those hazards (resilience) without being harmed (Lewis Reference Lewis1999; UNISDR 2009; IPCC 2013–2014). Examples are (in)equity, (in)justice, lack/presence of livelihoods and lack/presence of access to resources. Vulnerability and resilience are not strictly opposites, because both can exist simultaneously due to the same social process (Box 1).

Despite this paper referring to ‘island communities’, DRR including CCA accepts that communities are not homogenous, but have various groupings, sectors, power structures and differences within themselves (Lewis Reference Lewis1999). One consequence is that one community group might reduce vulnerability for some while (or through) increasing vulnerability for others. Different community groups could also work at cross-purposes in enacting DRR including CCA.

Because social and political processes occur over multiple time and space scales, the groups within society creating and perpetuating vulnerability and resilience to environmental hazards and hazard drivers are not always those experiencing any disasters. For instance, international trade regimes incentivizing livelihoods dependent on external forces, such as tourism, alongside decisions by insurers in world financial capitals to increase disaster premiums, affect Barbados's vulnerability and resilience to environmental hazards and hazard drivers, despite Barbados having minimal influence on these decisions (Pelling & Uitto Reference Pelling and Uitto2001).

Given that vulnerability is a necessary input into a disaster, alongside hazard, and given that vulnerability is entirely an anthropogenic process (constructed socially and politically), few ‘natural disasters’ exist, because most disasters are human-caused through vulnerability. In fact, most disasters could be averted through long-term vulnerability reduction and resilience building. Many hazards are also influenced by human activities, such as wildfire regimes affected by forest management and flood regimes affected by river and coastal engineering, making those hazards not quite natural (e.g. Tobin (Reference Tobin1995) for floods and Johnson et al. (Reference Johnson, Miyanishi and Weir1998) for wildfires). Some environmental hazards, however, can have planetary-wide consequences irrespective of DRR, such as ice ages and large-magnitude volcanic eruptions.

Within these varied human and environmental influences on hazards, contemporary climate change is one global hazard influencer with a significant proportion that is caused by human activities and yielding considerable concerns for island communities (IPCC 2013–2014). Climate change primarily affects weather-related hazards, changing the frequencies, intensities and extents of potential hazards such as storms, precipitation-related floods and droughts, and landslides (IPCC 2013–2014). Warming and rising oceans further affect ecosystems, including through coral bleaching and salinification of coastal lowlands (IPCC 2013–2014). The changing environment can lead to increased numbers of invasive species and can contaminate freshwater supplies with saltwater, and cannot always be addressed through traditional, local knowledge (e.g. Nunn Reference Nunn2009).

Climate change effects are particularly acute for island communities, from the small island developing states (SIDS) to the Arctic (IPCC 2013–2014). The emerging patterns of island vulnerability and resilience are complex. As will be further explored below, much depends on how resources are allocated and managed, cementing the human cause of disasters.

What do and could island studies offer in dealing with disasters?

Islands have long contributed significant knowledge to dealing with environmental hazards and hazard drivers, including climate change. These contributions provided many foundational theoretical and empirical works for dealing with environmental hazards and hazard drivers, including climate change. Many disciplines have contributed in this area, including anthropology (Belshaw Reference Belshaw1951), human and physical geography (McLean et al. Reference McLean, Bayliss-Smith, Brookfield and Campbell1977), seismology (Angenheister Reference Angenheister1921) and development studies (Lewis Reference Lewis1981). The 1970s yielded seminal literature alongside the beginning of concerns about climate change and the founding of contemporary theories of vulnerability and resilience to environmental hazards and hazard drivers. Two island-related projects stand out from this decade (Boxes 2 and 3).

The themes from Boxes 2 and 3 continued through the ensuing decades of research, policy and practice related to island communities dealing with environmental hazards and hazard drivers. By using Antigua, Lewis (Reference Lewis1984) designed a methodology of constructing a multi-hazard history, breaking down the silos separating hazards by discipline. Using Tuvalu and sea-level rise, Lewis's (Reference Lewis1989) work became one of the first peer-reviewed journal papers to connect vulnerability theory and climate change.

Through the Malé Declaration on Global Warming and Sea Level Rise (1989), the Alliance of Small Island States (AOSIS) was founded to lobby for SIDS’ interests in international climate change negotiations. AOSIS has become a powerful group within UN contexts (UN 1994, 2005, 2014), highlighting small countries’ vulnerabilities to external forces and keeping ocean topics prominent in UN agendas (Betzold Reference Betzold2010; Betzold et al. Reference Betzold, Castro and Weiler2012). The 2015 climate change negotiations in Paris leading to the UNFCCC (2015) agreement demonstrated not only AOSIS's influence in the intense discussions surrounding a global mean temperature target of 1.5 °C above pre-industrial levels, but also their ultimate lack of power through their failure to set the agreement's baseline below 2 °C (Fry Reference Fry2016). SIDS have been less prominent in international DRR negotiations, not because of a lack of interest, but because of a lack of resources and of bargaining power. Their interests nonetheless end up being expressed in voluntary international agreements on DRR (e.g. UNISDR 2015), mainly via mentions of islands as being particularly vulnerable.

With the founding of island studies followed by expanding research into the particular vulnerabilities and resiliences of islands (Journal of Coastal Research 1997; Sustainable Development 2006), island communities have continued to be leaders in developing and testing innovative methods for DRR including CCA. Examples are:

• • Combining different knowledge forms to ensure that neither local, traditional knowledge nor external, scientific knowledge is side-lined in DRR including CCA (Nunn Reference Nunn2009).

• • Community members using local materials to build three-dimensional desktop maps for identifying their hazards, vulnerabilities and resiliences (Maceda et al. Reference Maceda, Gaillard, Stasiak, Le Masson and Le Berre2009; Leon et al. Reference Leon, Hardcastle, James, Albert, Kereseka and Woodroffe2015).

• • Historical reconstructions to intuit islander decision-making regarding, and influences on, local environmental changes (Nunn Reference Nunn2003).

All of these examples indicate the balance of internal and external factors contributing to hazards, hazard drivers, vulnerabilities and resiliences, exemplified by island communities and their characteristics.

Population

Islands are generally assumed to have small communities with strong kinship networks. This characterization has plenty of truth, especially for more remote locations, along with plenty of exceptions, notably for cities comprising islands and cities on islands such as Copenhagen, Jakarta, Manila, Mumbai and New York (Grydehøj Reference Grydehøj2015). For dealing with environmental hazards and hazard drivers, no specific population characteristic is inevitably a panacea or a detriment; population characteristics have both advantages and disadvantages.

A smaller population has fewer total requirements for dealing with environmental hazards and hazard drivers, but also might lack the skills or resources to deal with them internally. Consequently, islands frequently pool resources through organizations (such as AOSIS) in order to harness the best expertise from across island communities while providing a focal point for island interests and advocacy/action power, which might otherwise be diluted. Countries such as the Faroe Islands and St Kitts/Nevis, with populations of approximately 50,000 each, would have trouble finding in-country individuals with the deep technical expertise across all scales and activities needed for dealing with environmental hazards and hazard drivers.

Multilateral cooperation overcomes limited population size by bringing together experts from around a region to support all countries within that region on a specific topic, such as the UNFCCC (2015) and the UNISDR (2015) agreements. Pooling resources is further advantageous in drawing on multiple, diverse perspectives rather than producing a cloistered, inward-looking framing, which could overlook lessons and advice from others’ experiences and perspectives.

Small populations, especially with stronger kinship-based connections, display nimbleness and swiftness in preparing for and responding to environmental hazards due to trust and ease of communication. Impediments are seen through petty disputes and loss of trust precisely due to tightness, smallness and familiarity. While the island literature examines such issues of kinship and trust (e.g. Randall et al. Reference Randall, Kitchen, Muhajarine, Newbold, Williams and Wilson2014), there is little empirical research examining a population's kinship, internal trust and coherence for dealing with environmental hazards and hazard drivers.

Furthermore, a small population does not necessarily mean an ignorant population. Island communities often have extensive knowledge about their local environmental hazards and hazard drivers and are sometimes able to respond well through local warning systems (examples are given in the next section). Nevertheless, no knowledge system could ever be complete. External knowledge should contribute, provided that it supplements and complements, rather than displaces or supersedes, local knowledge – and vice versa. Local knowledge should neither dominate nor disparage external knowledge.

Island communities have led research and application regarding combining knowledge systems for dealing with environmental hazards and hazard drivers. Cronin et al. (Reference Cronin, Gaylord, Charley, Alloway, Wallez and Esau2004a, Reference Cronin, Petterson, Taylor and Biliki2004b) brought together community members, government representatives and external scientists in order to deal with volcanic hazards in Vanuatu and Solomon Islands. Facilitated by external parties, an open, participatory process for applying everyone's knowledge, for identifying gaps and for filling in the gaps overcame distrust and political conflict.

Gaillard and Maceda (Reference Gaillard and Maceda2009) adopted a multi-hazard, multi-vulnerability, multi-resilience approach for Filipino communities, piloting participatory 3D mapping (P3DM). Minority and majority community members joined with local government, local and external scientists and non-governmental organization workers to build scale models of the community using locally sourced materials in order to identify hazards, vulnerabilities and resiliences. P3DM combines knowledge forms from different societal groups, yielding original data for risk analysis while leaving behind a legacy of the map and increased awareness. This knowledge and the map's data can be shared externally in order to maintain dialogue and to continue seeking external collaboration for action. Leon et al. (Reference Leon, Hardcastle, James, Albert, Kereseka and Woodroffe2015) used this method for BoeBoe village (Solomon Islands) with similar success, focusing on the hazard driver of climate change, especially with regards to sea-level rise.

A population's dispersiveness can be advantageous in dealing with environmental hazards and hazard drivers when island diasporas mobilize to assist their affected home. Island communities receive remittances for post-disaster assistance, with kinship networks meaning that remittances usually exceed official aid in terms of effectiveness, speed and reaching the most affected people (Le De et al. Reference Le De, Gaillard, Friesen and Smith2015). Islanders have long used economic migration and remittances as risk management and livelihood strategies (Bertram & Watters Reference Bertram and Watters1985). Migrants reduce an island community's population, thus taxing local resources less, while developing their own, external resources to assist their home communities.

Island communities have also been prominent regarding discussions of forced migration as a response to environmental hazards and hazard drivers. Islanders have long undergone forced migration with no guarantee of return due to environmental hazards including volcanic eruptions (e.g. Niuafo'ou (Tonga) in 1946) (Lewis Reference Lewis1979) and hazard drivers including climate and sea-level changes (e.g. around the Pacific in the fourteenth century) (Nunn Reference Nunn2007).

In contemporary work, the most notable manifestation is the rhetoric on islanders becoming ‘climate refugees’ due to the hazard driver of climate change. Some island communities are planning and undertaking relocation due to only climate change, such as from the Carteret Islands (Papua New Guinea) (Yamamoto & Esteban Reference Yamamoto and Esteban2014; Connell Reference Connell2016) and from Kivalina and Shishmaref (Alaska) (Bronen & Chapman III Reference Bronen and Chapin III2013). Some entire island countries are considering migration due to climate change, such as Kiribati and Maldives. Even though they consider this migration to be forced, the islanders tend to reject the label of ‘refugees’ because they wish to control the manner, mode and timing of their movement as much as is feasible, even while recognizing the need for external assistance in effecting their own migration-related decisions (McNamara & Gibson Reference McNamara and Gibson2009).

The discussions regarding islander migration for dealing with environmental hazards and hazard drivers occurs within the context of many islanders having long been migrants (Hau'ofa Reference Hau'ofa, Naidu, Waddell and Hau'ofa1993). Pacific exploration over the last few millennia and the extensive Caribbean communities in North America and Europe evidence reasons for migration being livelihood, education, family, health and adventure/exploration. These reasons do not justify forced migration due to contemporary human-induced environmental changes. They do indicate that stereotypical assumptions regarding island populations have truths and exceptions, with the populations’ characteristics providing advantages and disadvantages in dealing with environmental hazards and hazard drivers.

Area

As noted in the introductory section, islands are frequently characterized as having small land areas, but their water areas play important roles. Even islands that are comparatively large in land area frequently look towards their water. Greenland is one of the largest islands in the world by land area. It is mainly covered by an ice sheet, so communities remain coastal, small and dispersed – half of Greenland's population lives in settlements of fewer than 3000 people – while using comparatively little land area for living and livelihoods. Instead, traditional livelihoods are based on ocean hunting and fishing, while more recent livelihoods, such as administration, tourism and small businesses, are based within the settlements (Statistics Greenland 2016).

Despite the contribution of water area to island life and livelihoods, few islanders live on or in the water. Some nomadic peoples live in boats, such as throughout the Mergui Archipelago (Burma) (Dancause et al. Reference Dancause, Chan, Arunotai and Lum2009). Consequently, most islander homes sit on the limited island land area, and dealing with environmental hazards and hazard drivers generally occurs in this same area, including when evacuating elsewhere. For locations without higher ground or without much land to evacuate to, such as atolls, with enough warning, tsunamis can be ridden out by travelling to the deep sea.

The generally small land areas of islands thus impose significant constraints on dealing with environmental hazards and hazard drivers. Tsunamis can inundate 100% of such land area, salinating freshwater supplies and agricultural land for years, as occurred for some Maldivian atolls on 26 December 2004 (Orłowska Reference Orłowska2015). The eruption of Laki, Iceland, from 1783 to 1784 led to a famine killing approximately 25% of Iceland's population (Grattan & Charman Reference Grattan and Charman1994). Ireland's population required more than a century to recover to the levels seen prior to the main nineteenth century famine and emigration period (Boyle & Grada Reference Boyle and Grada1986). Many volcanic eruptions have led to entire island evacuations, such as Niuafo'ou (Tonga) (Lewis Reference Lewis1979) and Vestmannaeyjar (Iceland) (Chester Reference Chester1993). An environmental change or trend can make living on an island unviable, as occurred for Pacific communities in the fourteenth century (Nunn Reference Nunn2007). Such difficulties were also seen during droughts in the twenty-first century affecting Tuvaluan islands, for which continued habitation might not have been feasible without importing water and desalination equipment (Kuleshov et al. Reference Kuleshov, McGree, Jones, Charles, Cottrill, Prakash, Atalifo, Nihmei and Seuseu2014).

Nevertheless, islanders have had many successes in dealing with environmental hazards within the small land areas of their communities (Box 4). In the western Pacific, many islands have been continuously occupied for over three millennia (Hung et al. Reference Hung, Carson, Bellwood, Campos, Piper, Dixon, Bolunia, Oxenham and Chi2011). Inhabitants of Simeulue (Indonesia) have oral traditions of tsunami warning and response, so they evacuated to high ground after feeling an earthquake on 26 December 2004, saving their lives during the subsequent tsunami (Gaillard et al. Reference Gaillard, Clavé, Vibert, Azhari, Denain, Efendi, Grancher, Liamzon, Sari and Setiawan2008). In all of these cases, despite the small land areas of the islands, sufficient locations existed that enabled the population to move out of harm's way for the particular environmental hazards experienced.

Geomorphology

Observations of island geomorphological responses to changing sea levels are mixed at present (Woodroffe Reference Woodroffe2008; Rankey Reference Rankey2011; Kench et al. Reference Kench, Thompson, Ford, Ogawa and McLean2015; Albert et al. Reference Albert, Leon, Grinham, Church, Gibbes and Woodroffe2016; Nunn et al. Reference Nunn, Runman, Falanruw and Kumar2016). Depending on localized parameters such as waves, currents and human activities including sea walls and sand mining, islands have grown, shrunk, changed longitudinally or not been significantly affected in locations with measurable sea-level rise. Future responses, as sea levels increase and perhaps the rate of rise accelerates, are hard to project, except to note that island disappearance is a possible but not inevitable outcome.

Another uncertainty is ocean acidification. Ocean water absorbs some of the increased atmospheric carbon dioxide, yielding carbonic acid, which increases oceans’ acidity. The impacts of acidification on coastlines including shingle beaches are not well understood. Coral reefs experience the two-fold stress of increased acidity and increased temperatures, which can kill them through coral bleaching. While coral reefs across previous millennia rebounded from massive bleaching events as well as large changes in sea level and ocean temperature, it is unclear how well they will survive under contemporary climate change projections (Hoegh-Guldberg Reference Hoegh-Guldberg2014). Coral reefs protect land from currents and waves, meaning that massive coral die-offs could expose island shorelines to the ocean's full power, leading to accelerated geomorphological changes (Hoegh-Guldberg et al. Reference Hoegh-Guldberg, Mumby, Hooten, Steneck, Greenfield, Gomez, Harvell, Sale, Edwards, Caldeira, Knowlton, Eakin, Iglesias-Prieto, Muthiga, Bradbury, Dubi and Hatziolos2007).

Even if geomorphological changes do not destroy islands, dealing with some environmental hazards seems likely to become more difficult under climate change. Freshwater management will become challenging as rising seas salinate groundwater and freshwater lenses. As ecosystems change, subsistence food will be affected due to invasive species and species extinctions (Wetzel et al. Reference Wetzel, Kissling, Beissmann and Penn2012; Betzold Reference Betzold2015). The potential impacts on fisheries are particularly concerning due to many island communities’ reliance on this sector (Nurse Reference Nurse2011). Changing biota, in turn, affects island geomorphology (e.g. if coastal vegetation no longer traps sediment or if inland vegetation no longer anchors soil during rainfall).

Human responses that alter an island's geomorphology, such as raising islands above the sea or building floating settlements, have been proposed (Yamamoto & Esteban Reference Yamamoto and Esteban2014). While the engineering appears to be technically feasible, the funds required are so far not available. Consequently, much discussion has emerged regarding migrating from island communities as a method of dealing with environmental hazards, or forced migration as a failure to deal with environmental hazards, including both climate-related and volcano-related hazards (see above).

Many islands are volcanoes, and this fact shapes livelihoods, due to, for example, the fertile soil resulting from volcanic ash or the lack of arable land due to hardened lava. There can be long periods between a volcano's eruptions, so the current human population that is settled on an island might have limited knowledge of a volcano's potential activity, as occurred for Montserrat in the Caribbean in 1995 (Pattullo Reference Pattullo2000). Lack of knowledge and experience can inhibit responses to environmental hazards, underscoring the importance of combining local and external knowledge forms (see above). The town of Sete Cidades, Azores, sits in a caldera with volcanic walls rising over 150 m above the settlement. One potentially apocryphal story from the town, as told to this author during field work there, is that, prior to modern transportation, people could be born in the town and never leave the caldera; they would never have seen the sea despite being just a few kilometres from the coastline. The island's geomorphology precluded experience with the ocean and associated environmental hazards.

Connectedness

The stereotype of islands as isolated, insular and marginalized, which then creates difficulties for dealing with environmental hazards and hazard drivers, appears often (e.g. Boxes 3 and 4). A lack of connectedness can, however, also lead to striving for self-sufficiency, thereby enabling islanders to deal with environmental hazards and hazard drivers (Box 5).

For sea-level rise (see above), engineering-based approaches could make low-lying atolls inhabitable under many scenarios. These approaches are expensive to construct and maintain in isolated locations, partly due to the need for transporting all of the necessary materials and expertise long distances. Many donors have invested in the engineering of islands, such as Japan's construction of sea walls in Tonga and Maldives. So far, no one has been willing to commit the full resources necessary to guarantee century-scale inhabitation of the countries that are most expected to require population migration due to climate change.

Other mechanisms overcoming island isolation include physical connections (fixed links of bridges, tunnels and causeways), transportation connections incorporating ferries and aeroplanes and communication connections, mainly involving the internet and phones. These mechanisms assist in dealing with environmental hazards and hazard drivers, but can also lead to dependency. If an assumption is made that a mobile phone and a bridge could be relied on for requesting and bringing in disaster aid, then the maintenance of local caches and local skills could be neglected. An environmental hazard might then render the phone networks or bridge unusable, causing response problems.

This situation is a concern in the Pacific, where many islanders prefer imported, cheap, unhealthy foods, leading to reduced interest in and acquisition of local foods, thus increasing disaster vulnerability (Campbell Reference Campbell2009). The expectation of post-disaster assistance has also diminished some local Pacific populations’ capabilities for dealing with environmental hazards (Lewis Reference Lewis2009). In Fiji, more isolated communities that have previously received less aid than less isolated communities have developed more local capabilities for dealing with cyclones (Johnston Reference Johnston2015).

Nonetheless, not all islands are isolated, insular or marginalized. Island diasporas can be important connectors (see above), or they might reinforce island isolation and marginalization. For Cuba, parts of the diaspora, especially in Florida, became vociferously opposed to Fidel Castro, deliberately seeking to increase Cuba's isolation in order to bring down his government (García Reference García1996).

Islands can also be more connected than they seem, particularly when dealing with environmental hazards. When Montserrat's volcano first started erupting in 1995, it was seen by the UK government as a minor crisis in a faraway, small, irrelevant place. The inept governance of the crisis, especially by the UK's government, contributed to a major political scandal followed by an overhaul of how the UK Overseas Territories are governed (Pattullo Reference Pattullo2000). Political connectedness meant that the events in Montserrat had major ramifications for London. When Eyjafjallajökull volcano in Iceland erupted in 2010, sending volcanic ash across Europe and stopping tens of thousands of commercial flights for several days, the implications of the eruption and the ensuing crisis management were felt worldwide (Alexander Reference Alexander2013).

Island connectedness and separateness have many levels. At times, stereotypes exist in reality, influencing the dealing with environmental hazards and hazard drivers. Many exceptions exist as well, affecting plans based on assumptions regarding island connectedness.