Climate change in the South Pacific
In 1998 and 1999, the South Pacific Regional Environment Programme (SPREP) commissioned the Commonwealth Scientific and Industrial Research Organisation (CSIRO) Atmospheric Research to carry out two projects on climate change and sea-level rise for the South Pacific.
The first of these projects, Identification of Latent Sea-Level Rise within the Climate System at 1995 and 2020, simulated the amount of sea-level rise that would occur if human emissions of greenhouse gases were to cease abruptly at those times. Although applying totally unrealistic scenarios, this exercise was useful for assessing the commitment, in terms of sea-level rise, by emissions now and in the near future.
The second project, Regional Climate Change Scenarios and Risk Assessment Methods, constructed regional scenarios of climate based on a number of climate models and a broad overview of climate impacts for the region. The project also introduced a risk assessment methodology for the assessment of impacts and adaptation options.
Both of these reports were recently released in time for the Pacific Islands Climate Change Conference held at Raratonga in the Cook Islands from April 3rd-7th 2000.
Latent sea-level rise
Latent sea-level rise is the increase in sea level likely to occur due to emissions of greenhouse gases already in the atmosphere. That is, if all human-induced emissions of greenhouse gases stopped tomorrow, sea level would continue to rise due to warming, ice-melt and other factors, until atmospheric and sea-level equilibrium was reached.
Although this is an artificial concept because it is impossible to turn off human-induced greenhouse gas emissions like a tap, it is useful for policy makers who wish to determine the impact of historical or projected greenhouse gas emissions on sea-level rise.
Two latent sea-level rises were investigated.
Two scenarios were constructed to simulate these rises. Scenario 1 applied estimated emissions of the major greenhouse gases in 1990, estimated carbon dioxide from fossil fuel emissions and IS92a estimates for the other greenhouse gases in 1995, then natural emissions from 2000 until 2100.
Scenario 2 applied greenhouse gas emissions based on the IS92a scenario, with reductions to anthropogenic emissions based on the Kyoto Protocol, to 2020, then natural emissions from 2025 until 2100.
These scenarios were applied to the simple model used to provide the Intergovernmental Panel for Climate Change (IPCC) 1996 projections of sea-level rise. Assuming a carbon dioxide doubling temperature of 2.5 degrees Celsius or more (i.e. the climate sensitivity will be in the upper half of the range of uncertainty), the projections of latent sea-level rise for both scenarios are as given in Table 1.
These figures indicate that within several decades, global emissions will be sufficient to produce the risk of a 30 cm sea-level rise. This is a level that Nakibae Teuatabo, climate change coordinator from Kiribati in the South Pacific, was quoted in New Scientist recently as saying would have very serious implications for his entire country.
Regional climate change scenarios
Six climate change simulations from four research groups were analyzed and used to construct projected ranges of temperature and rainfall change for the SPREP region. Each was scaled to produce local patterns of change per degree of global warming over the region.
The individual model patterns from all six models were then aggregated into a range of local change per degree of global warming. The upper and lower limits of this range were multiplied by projections of global warming from the IPCC Second Assessment Report to produce regional projections of temperature and rainfall. See Tables 2 to 4.
In general, being ocean-dominated areas, temperature increases across the South Pacific were slightly less than the projected global average. The key feature produced by all the models analyzed was in the equatorial central eastern Pacific where increases in both temperature and rainfall were the largest for the Pacific region.
The Intertropical Convergence Zone (ITCZ) and the South Pacific Convergence Zone (SPCZ) were reasonably well represented under current climate, but there were no consistent shifts of these features between models under climate change, save in the central eastern Pacific. This is reflected in the accompanying tables.
Analyses of model output and observed climate histories suggest that the El Niño Southern Oscillation (ENSO) phenomenon will continue as a major climatic feature in the future. Patterns and analyses of temperature and rainfall produced for this study show that the global climate models produce a more El-Niño-like mean state over the Pacific under climate change. Rainfall increases are also distributed in an El-Niño-like pattern; they generally occur over most of the Pacific, except where warming is least.
Changes to rainfall variability from selected models were analyzed. Both interannual and decadal-scale variability are represented within the climate models but there are too few results available for projections to be made. Increases in daily rainfall intensity are expected in most regions where rainfall increases, remains the same or decreases slightly. Even where there is a large decrease in average rainfall, reductions in daily rainfall intensity may be negligible.
Tropical cyclone simulations showed no evidence of change in their numbers or regions of formation, although there was some evidence tropical cyclones may track further polewards. A general increase in tropical cyclone intensity, expressed as possible increases in wind speed and central pressure of 10-20 per cent at the time of carbon dioxide doubling, now appears likely. How this affects the risk posed by severe storms needs to be determined on a regional basis.
A risk assessment framework was introduced along with a simple example of risk analysis. As can be seen from the scenario tables, if the upper and lower ranges of climate change are followed through to impacts with their added uncertainties, that range can become very large. This information is very difficult to apply in a policy or management sense.
Instead, risk analysis identifies a level of outcome that is important for the activity being analyzed, called an impact threshold. An impact threshold marks a level that can be desired (for example, sustainability) or avoided (for example, large human, financial or environmental costs). By linking an impact threshold with climate variables likely to change, the range of uncertainty for those climatic variables is identified, and the risk of exceeding that threshold is assessed.
The process of setting thresholds requires input from stakeholders, that is, those who manage and use the resources impacted upon. Stakeholders are also central to the introduction and analysis of adaptation options. Therefore, the framework utilizes a risk analysis methodology that has the potential to be integrated with existing adaptation programmes in the Pacific, such as integrated coastal zone management.
As an example, a risk analysis, comparing risk before and after the implementation of the Kyoto Protocol, was carried out on two thresholds for a hypothetical coastline: a threshold of 50 cm sea-level rise, and an atmospheric concentration of carbon dioxide of 560 ppm (associated with a possible reduction of calcification rates in reef communities). The risk of both these thresholds being exceeded in 2075 was 16 per cent and in 2100 was 44 per cent. When analyzed for combined risk under IS92a-f scenarios and Kyoto Protocol-modified scenarios in 2075 and 2100, the risk under the Kyoto Protocol was reduced by 9 per cent and 6 per cent respectively. This shows that, for these two impacts, implementing the Kyoto Protocol reduces the risk by less than 10 per cent, or alternatively, delays it by just under a decade. The results are shown in Table 5.
The two project reports prepared by the CSIRO present the most comprehensive analysis of climate change projected from climate models yet undertaken for the South Pacific.
More modelling will be undertaken in the future, reducing the ranges of established uncertainties and illuminating others. However, the reports argue that it is time to establish links between the modellers, impact researchers, planners and users of impact information in order to carry out adaptation assessments.
Although there are traditionally developed skills for coping with existing climate impacts in the South Pacific, the process of ongoing change is placing a great deal of pressure on traditional knowledge systems. Adaptation options that manage uncertainty and that are culturally, economically and scientifically suitable are required to cope with such changes. The risk assessment methodology presented in the second report is suggested as one way of doing this.
Roger Jones, Climate Risk and Integrated Assessment Project, Climate Impact Group, CSIRO Atmospheric Research, Private Bag No. 1, Aspendale, VIC 3195, Australia. Fax: +61-3-92394688. Email: firstname.lastname@example.org. Web: www.dar.csiro.au/res/cm/impact.htm.