«APPROACHES TO VULNERABILITY ASSESSMENT ON PACIFIC ISLAND COASTS: EXAMPLES FROM SOUTHEAST VITI LEVU (FIJI) AND SOUTH TARAWA (KIRIBATI) D.L. Forbes1 & ...»
3rd SPREP Meeting on Climate Change & Sea Level Rise in the Pacific, ORSTOM, Nouméa, 1997
APPROACHES TO VULNERABILITY ASSESSMENT
ON PACIFIC ISLAND COASTS: EXAMPLES FROM
SOUTHEAST VITI LEVU (FIJI) AND SOUTH TARAWA (KIRIBATI)
D.L. Forbes1 & S.M. Solomon1
South Pacific Applied Geoscience Commission [SOPAC],
PMB/GPO, Suva, Fiji 1
Geological Survey of Canada, Bedford Institute of Oceanography, 1 Challenger Drive (POB 1006), Dartmouth, Nova Scotia, Canada B2Y 4A2 [firstname.lastname@example.org & email@example.com]
Interannual variations in water level related to El Niño/ Southern Oscillation and other processes may be comparable to predicted water-level trends under climate change and must be incorporated in the analysis. Existing models for coastal response to rising sea level are inappropriate; new models, incorporating reef response and sand supply, are urgently needed for realistic vulnerability assessment. Improved models of reef-lagoon circulation, storm-surge, setup, runup, and wave transmission over reefs are needed for hazard zone delineation and risk assessment in settings such as Suva. Large uncertainties in the specification of physical impacts, including the qualitative response of atoll beaches to rising sea level, have obvious implications for the assessment of socioeconomic impacts and response options. Monitoring is required to detect changes and facilitate ongoing VA as an iterative process. With serious deficiencies in the baseline data required for effective VA, and significant uncertainties in the prediction of physical impacts, these issues demand attention as critical aspects of capacity building for climatechange preparedness and mitigation in the Pacific. Furthermore, the quantitative GIS approach to coastal vulnerability adopted in this study is equally appropriate for coastal hazard assessment under present conditions, a necessary prerequisite for realistic 3rd SPREP Meeting on Climate Change & Sea Level Rise in the Pacific, ORSTOM, Nouméa, 1997 climate-change VA, and provides the basic management tool for integrated coastal-zone management and planning.
INTRODUCTIONA number of recent developments have highlighted the need for coastal vulnerability assessment [VA] in the context of anticipated global climate change and accelerated sealevel rise [ASLR]. These include the latest evaluations of the Intergovernmental Panel on Climate Change (IPCC, 1996), multinational and regional initiatives such as the Global Environment Facility [GEF], the Pacific Island Climate Change Assistance Program [PICCAP], and the International Coral Reef Initiative [ICRI] (SPREP, 1995a, 1995b; SOPAC, 1995), coastal evaluations under national programs such as the Canadian Climate Program and Green Plan (e.g. Shaw et al., 1994; Solomon et al., 1994; Forbes et al., 1997), and specific international assistance and coordination efforts such as the US Country Studies Program (1994). Under this program, the government of the USA has committed a total of US$25 million to assist developing countries and those with economies in transition to meet their obligations under the Framework Convention on Climate Change, to carry out greenhouse gas inventories and develop plans for responding to climate change.
Under the umbrella of the Country Studies Program, the governments of Fiji and Kiribati requested the South Pacific Applied Geoscience Commission [SOPAC] to undertake surveys and assessments of coastal vulnerability in and around Suva (Solomon & Krüger, 1996;
Solomon et al., 1997) and South Tarawa (Solomon, 1997). This paper summarizes the results of these and related SOPAC assessments of coastal vulnerability in the South Pacific (e.g. Carter, 1990a; Rearic, 1990; Forbes & Hosoi, 1995; Forbes, 1996).
Coastal vulnerability assessment in Pacific Island countries
This work built on a number of earlier resource, hazard, and vulnerability assessments in both Fiji and Kiribati (e.g. Carter, 1990b; Nunn et al., 1993, 1994; and Shorten, 1993 [in the Suva region]; McLean, 1989; Woodroffe & McLean, 1992 [in Kiribati]). These and other studies in the region have highlighted limitations of standard methodologies for VA in the Pacific Islands (Holthus et al., 1992; Nunn & Waddell, 1992; Kay et al., 1993; Yamada et al., 1995; Harvey, 1997). In this context, the present work required new strategies appropriate to the specific environmental settings and data limitations of Pacific Island nations. It also demonstrates wide regional diversity in coastal environments and in the availability of comprehensive and systematic environmental and socioeconomic data sets required for effective VA. The two examples highlighted in this paper employed different methodologies to deal with significantly different environmental settings, but both demonstrate the utility and value of integrating data sets within a geographic information system [GIS], to enable georeferenced queries and common mapping of physical conditions, environmental impacts, and socioeconomic data.
In the Suva area, located on a high volcanic island at 18°S latitude (Fig. 1), tropical cyclones and storm surges are the dominant environmental hazards. Emphasis in the VA was placed on the potential enhancement of these hazards under rising sea level. Furthermore, the coast is extensively modified by seawalls, so that natural shore adjustment to sea-level rise is a relatively minor issue here. Much of the remaining shoreline is mangrove, requiring yet another set of approaches to impact assessment (Woodroffe, 1987; UNEP-UNESCO, 1992).
The socioeconomic setting involves a highly developed urban centre, but suitable topographic, land value, and population data were not readily available.
[SOPAC Miscellaneous Report 277 – Page 2] 3rd SPREP Meeting on Climate Change & Sea Level Rise in the Pacific, ORSTOM, Nouméa, 1997 In the South Tarawa (Kiribati) study, focusing on a densely populated atoll islet at 1°N latitude (Fig. 1), tropical cyclones were of little significance but interannual variations in sea level, winds, and sand transport related to El Niño/ Southern Oscillation (ENSO) effects are a major factor in coastal response. Here, as in the Suva case, the health and response of the reef to rising sea level is of major importance but poorly known. Furthermore, in this atoll setting, it remains a matter of controversy and uncertainty whether the shore will respond to higher sea levels by erosion and retreat (as postulated by conventional methods developed for continental settings and promoted in the Common Methodology and other international VA guidelines) or by progradation and vertical accretion (if enhanced reef growth supplies adequate sediment to the shore).
Existing VA methodologies and limitations A set of guidelines has been provided for vulnerability and adaptation assessments under the US-funded program (US Country Studies Program, 1994). For coastal assessments, these follow closely the Common Methodology for assessing vulnerability to sea-level rise in coastal regions, set out by the IPCC Response Strategies Working Group (IPCC, 1992).
The Common Methodology consists of seven components as follows:
1- Delineate case study area and specify ASLR and climate-change scenarios;
2- Inventory study area conditions, including natural system and socioeconomic data;
3- Identify relevant development factors;
4- Assess physical changes and natural system responses;
5- Formulate response strategies including potential costs and benefits;
6- Assess vulnerability profile and interpret results;
7- Carry out needs assessment and develop action plan.
[SOPAC Miscellaneous Report 277 – Page 3] 3rd SPREP Meeting on Climate Change & Sea Level Rise in the Pacific, ORSTOM, Nouméa, 1997 The IPCC Common Methodology and the US Country Studies Guidelines have limitations in their application to tropical reef coasts. Their applicability is further limited where international assistance and traditional subsistence activities form significant components of the economy (Woodroffe & McLean, 1992) and where there is a strong cultural attachment to the land (Nunn, 1992; Yamada et al., 1995). Other issues include the choice of appropriate temporal and spatial scales for VA and the need for better evaluation of coastal hazards, both natural and human-induced, independent of sea-level rise (Harvey, 1997). The semi-quantitative methodology proposed by Yamada et al. (1995) can provide a broad-based overview of vulnerability and resilience, but is not suitable for coastal hazard delineation or for practical planning applications.
The methodology suggested in the Country Studies guidance volume for estimation of erosion losses is based on processes operating along open sandy coasts without reefs. In particular, the Bruun (1962) model of erosional response related to adjustment of an equilibrium shoreface profile (recommended for use in these guidelines) has been shown to have very limited validity, except under stringent assumptions (cf. Forbes et al., 1989; Pilkey et al. 1993). It is not generally applicable to supply-limited carbonate coasts, where landward losses may be expected (Thom & Roy, 1988), new sediment can be produced on the reef (Hopley & Kinsey, 1988) and where there is a non-erodible underlying reef pavement which limits profile adjustment.
In atoll countries, such as Kiribati and Tuvalu, there is as yet no scientific consensus on the medium-term (10-100 years) response of islet beaches to accelerated sea-level rise (e.g. Roy & Connell, 1989). The possibilities include beach erosion and inundation or beach growth and upward accretion. The extent to which these responses occur will depend on shoreline orientation, exposure, reef width, reef health (including sand productivity), and weather (wind, storm, wave) patterns. Different responses can be expected on ocean and lagoon shorelines and both will be important to the vulnerability assessment. Reef health and productivity may be affected by natural climate change, including sea-surface temperature and coral bleaching (Brown, 1990), by predation and disease (Devaney & Randall, 1973; Littler & Littler, 1994), by pollution and sedimentation (Zann, 1982; Cortes & Risk, 1985), or by sea-level rise, leading either to drowning or to renewed growth and accretion (e.g. Neumann & MacIntyre, 1985; Buddemeier & Smith, 1988; Woodroffe & McLean, 1990).
In the present study, the IPCC Common Methodology and the Country Studies guidelines were adapted in different ways, as appropriate to the different settings in Viti Levu and South Tarawa. Emphasis in the Suva study was placed on determining flooding, overtopping, and erosion risks along urban natural and artificial shorelines (Suva & Lami) and rural mangrove and deltaic shorelines (Laucala Bay and Rewa Delta, east of Suva). In the Betio (South Tarawa) study, emphasis was placed on shoreline stability, using historical trends of shore recession or accretion, beach-profile monitoring to determine the shoreline response to medium-term (ENSO) variations in mean sea level, and observed changes in sediment distribution on the adjacent reef flat. The risk assessment in both studies was based on existing wave and storm climatology superimposed on various scenarios for sea-level rise, combined with a geographic information system [GIS] for vulnerability assessment.
Vulnerability was determined in relation to shore type and condition, reef morphology, backshore elevation, historical trends, and socioeconomic factors such as population density, land values, land use, infrastructure and other urban development factors.
SUVA & VICINITY, VITI LEVU, FIJI Approach and objectives The objectives of the study were to assess the physical effects of accelerated sea-level rise [ASLR] on the Suva Peninsula (Fig. 2) and to investigate potential response options, which would mitigate these effects. During the early stages of project implementation, it was recognised that there were significant information gaps, particularly in the area of physical characterisation. These included a lack of topographic data required for inundation analysis and of qualitative information on shore types.
Fig. 2. Suva Peninsula and Rewa Delta study area, Viti Levu, Fiji.
The vulnerability assessment of the Suva port facilities by Nunn et al. (1994) utilised existing information, much of it from design water levels for the port of Suva developed by Carter (1990b). They concluded that the wharves themselves are only in danger of overtopping if sea level rises by 1 m. Other parts of Suva were identified as being flood-prone even without accounting for sea-level rise.