The Pacific Islands region includes the northwest tropical Pacific, and the tropical southwest Pacific. North of the Equator, the wet season occurs from May to November. In the south, the wet seasons occurs from November to April.
The phenomena mainly responsible for climate variations in the Pacific Islands are ENSO (Section 14.4), the SPCZ (Section 18.104.22.168), the ITCZ (Section 22.214.171.124) and the WNPSM (Section 126.96.36.199). During El Niño events, the ITCZ and SPCZ move closer to the equator, rainfall decreases in western regions and increases in the central Pacific, and tropical cyclone numbers tend to increase and to occur farther east than normal (Diamond et al., 2012). During La Niña, the western tropical Pacific tends to experience above-average numbers of tropical cyclones (Nicholls et al., 1998; Lavender and Walsh, 2011).
The seasonal evolution of the SPCZ has a strong influence on the seasonality of the climate of the southern tropical Pacific, particularly during the wet season. The SPCZ moves northward during moderate El Niño events and southward during La Niña events (Folland et al., 2002; Vincent et al., 2011). During El Niño events, southwest Pacific Island nations experience an increased occurrence of forest fires and droughts (Salinger et al., 2001; Kumar et al., 2006b), and an increased probability of tropical cyclone damage, as tropical cyclogenesis tends to reside within 6° to 10° south of the SPCZ (Vincent et al., 2011).
Nauru experiences drought during La Niña as the SPCZ and ITCZ move to the west (Brown et al., 2012c). During strong El Niño events (e.g., 1982/1983, 1997/1998) the SPCZ undergoes an extreme swing of up to 10 degrees towards the equator and collapses to a more zonally oriented structure (Vincent et al., 2011; Section 14.3.2). The impacts from these zonal SPCZ events are much more severe than those from moderate El Niño events (Vincent et al., 2011; Cai et al., 2012b), and can induce massive droughts and food shortages (Barnett, 2011).
Temperatures have increased at a rate between 0.1°C and 0.2°C per decade throughout the Pacific Islands during the 20th century (Folland et al., 2003). Changes in temperature extremes have followed those of mean temperatures (Manton et al., 2001; Griffiths et al., 2005). During 1961–2000, locations to the northeast of the SPCZ became wetter, with the largest trends occurring in the eastern Pacific Ocean (east of 160°W), while locations to the southwest of the SPCZ became drier (Griffiths et al., 2003), indicative of a northeastward shift of the SPCZ. Trends in the frequency of rain days were generally similar to those of total annual rainfall (Manton et al., 2001; Griffiths et al., 2003). Since 1980, western Pacific monsoon- and ITCZ-related rain during June to August has decreased (Hennessy et al., 2011).
Future projections for tropical Pacific Island nations are based on direct outputs from a suite of CMIP3 models, updated using CMIP5 wherever available (Brown et al., 2011; Hennessy et al., 2011; Irving et al., 2011; Moise and Delage, 2011; Perkins, 2011; Perkins et al., 2012). These projections carry a large uncertainty, even in the sign of change, as discussed below and as evident in Table 14.1.
Annual average air and sea surface temperature are projected to continue to increase for all tropical Pacific countries. By 2055, under the high A2 emissions scenario, the increase is projected to be 1°C to 2°C. A rise in the number of hot days and warm nights is also projected, and a decline in cooler weather, as already observed (Manton et al., 2001). For a low-emission scenario, the lower range decreases about 0.5ºC while the upper range reduces by between 0.2°C and 0.5°C.
To a large extent, the response of the ITCZ, the SPCZ, and the WNPSM to greenhouse warming will determine how rainfall patterns will change in tropical Pacific. In northwestern and near-equatorial regions, rainfall during all seasons is projected to increase in the 21st century. Wet season increases are consistent with the expected intensification of the WNPSM and the ITCZ (Smith et al., 2012a). For the southwestern tropical Pacific, the CMIP3 and CMIP5 ensemble mean change in summer rainfall is far smaller than the inter-model range (Brown et al., 2012b; Widlansky et al., 2013). There is a projected intensification in the western part of the SPCZ and near the equator with little mean change in SPCZ position (Brown et al., 2012a; Brown et al., 2012b).
For the southern group of the Cook Islands, the Solomon Islands, and Tuvalu, average rainfall during the wet season is projected to increase; and for Vanuatu, Tonga, Samoa, Niue, Fiji, a decrease in dry season rainfall is accompanied by an increase in the wet season, indicating an intensified seasonal cycle.
Extreme rainfall days are likely to occur more often in all regions related to an intensification of the ITCZ and the SPCZ (Perkins, 2011). Although the intensification appears to be reproduced in CMIP5 models (Brown et al., 2012a), it has recently been questioned (Widlansky et al., 2013; see Section 14.3.1). There are two competing mechanisms, the ‘wet regions getting wetter’ and the ‘warmest getting wetter, or coldest getting drier’ paradigms. These two mechanisms compete within much of the SPCZ region. Based on a multi-model ensemble of 55 greenhouse warming experiments, in which model biases were corrected, tropical SST changes between 2°C to 3°C resulted in a 5% decrease of austral summer moisture convergence in the current SPCZ region (Widlansky et al., 2013). This projects a diminished rainy season for most Southwest Pacific island nations. In Samoa and neighbouring islands, summer rainfall may decrease on average by 10 to 20% during the 21st century as simulated by the hierarchy of bias-corrected atmospheric model experiments. Less rainfall, combined with increasing surface temperatures and enhanced potential evaporation, could increase the chance for longerterm droughts in the region. Such projections are completely opposite to those based on direct model outputs (Figure 14.27).
Recent downscaling experiments support the above conclusion regarding the impact of biases on the SPCZ change, and suggest that the projected intensification of the ITCZ may have uncertainties of a similar nature (Chapter 7 of Hennessy et al., 2011). In these experiments a bias correction is applied to average sea surface temperatures, and the atmosphere is forced with the ‘correct’ climatological seasonal cycle together with warming derived from large-scale model outputs. The results show opposite changes in much of the SPCZ and some of the ITCZ regions, resulting in much lower confidence in rainfall projections. Despite the uncertainty, there is general agreement in model projections regarding an increase in rainfall along the equator (Tables 14.1 and 14.2), and regarding a faster warming rate in the equatorial Pacific than the off-equatorial regions (Xie et al., 2010b). A potential consequence is an increase in the frequency of the zonal SPCZ events (Cai et al., 2012b).
In summary, based on CMIP3 and CMIP5 model projections and recently observed trends, it is very likely that temperatures, including the frequency and magnitude of extreme high temperatures, will continue to increase through the 21st century. In equatorial regions, the consistency across model projections suggests that rainfall is likely to increase. However, given new model results and physical insights since the AR4, the rainfall outlook is uncertain in regions directly affected by the SPCZ and western portion of the ITCZ.