Increases in the frequency of extreme weather and climate events and the severity of their impacts on the natural environment and society have been observed across the globe in recent decades. In addition to natural climate variability and greenhouse-induced climate change, extreme weather and climate events produce the most pronounced impacts. In this paper, the climate of three island countries in the Western Pacific: Fiji, Samoa and Tuvalu, has been analysed. Warming trends in annual average maximum and minimum temperatures since the 1950s have been identified, in line with the global warming trend. We present recent examples of extreme weather and climate events and their impacts on the island countries in the Western Pacific: the 2011 drought in Tuvalu, the 2012 floods in Fiji and a tropical cyclone, Evan, which devastated Samoa and Fiji in December 2012. We also relate occurrences of the extreme weather and climate events to phases of the El Niño-Southern Oscillation (ENSO) phenomenon. The impacts of such natural disasters on the countries are severe and the costs of damage are astronomical. In some cases, climate extremes affect countries to such an extent that governments declare a national state of emergency, as occurred in Tuvalu in 2011 due to the severe drought’s impact on water resources. The projected increase in the frequency of weather and climate extremes is one of the expected consequences of the observed increase in anthropogenic greenhouse gas concentration and will likely have even stronger negative impacts on the natural environment and society in the future. This should be taken into consideration by authorities of Pacific Island Countries and aid donors when developing strategies to adapt to the increasing risk of climate extremes. Here we demonstrate that the modern science of seasonal climate prediction is well developed, with current dynamical climate models being able to provide skilful predictions of regional rainfall two-three months in advance. The dynamic climate model-based forecast products are now disseminated to the National Meteorological Services of 15 island countries in the Western Pacific through a range of web-based information tools. We conclude with confidence that seasonal climate prediction is an effective solution at the regional level to provide governments and local communities of island nations in the Western Pacific with valuable assistance for informed decision making for adaptation to climate variability and change.
Increases in the frequency of extreme weather and climate events and in the severity of their impacts on the nat- ural environment and society have been observed across the globe in recent decades [
Tropical cyclones, floods and droughts are the most destructive severe weather and climate extreme events which affect countries in the Western Pacific. In this region, floods are typically associated with tropical distur- bances and droughts are usually related to the El Niño-Southern Oscillation (ENSO). The ENSO is a large-scale climate phenomenon that occurs across the tropical Pacific Ocean and has two distinctly different phases: warm (El Niño) and cold (La Niña). There is also a neutral phase which is a transition mode from the ENSO warm phase to its cold phase and vice versa. Relationship between the ENSO and tropical cyclone activity in the Western Pacific (e.g. [
Analysing historical records from the archives of the National Meteorological Services (NMSs) of Fiji, Samoa and Tuvalu for the past few decades, we identified a number of severe weather and climate extreme events and traced their relationship to phases of the ENSO. Rainfall distribution is closely related to the ENSO variability, and the position and intensity of the key climate features namely the Intertropical Convergence Zone (ITCZ), West Pacific Monsoon (WPM) and South Pacific Convergence Zone (SPCZ; [
El Niño is often associated with significant rainfall deficits in countries southwest of the SPCZ including Fiji and Samoa which occasionally results in a meteorological drought. On the northern side of SPCZ, droughts are associated with prolonged southwest displacement of the SPCZ, which occurs during La Niña events. If a tropical cyclone or a tropical depression occurs in the vicinity of Fiji and Samoa, especially if a system is moving slowly and staying stationary for a few days, then even during El Niño episodes it can temporarily break a drought or even cause flooding.
Droughts have severe impacts on the natural environment and society of Fiji, Samoa, Tuvalu and other Pacific Island Countries. In 1997/1998 (El Niño episode), severe drought affected Samoa, and then another drought in 2002/2003 when electricity shortages occurred due to lack of water for hydro-electricity. During these dry con- ditions, fire caused significant damage to forests in Samoa. According to the Samoa Fire Service, 800 hectares of Asau and Aopa forests were destroyed by fire in 2011 (also an El Niño episode).
In extreme cases, droughts affect countries to such an extent that they cause devastating water crisis (e.g. the 2011 drought in Tuvalu). In this paper, a case study for the 2011 Tuvalu drought is discussed, along with de- tailed analysis of seasonal climate prediction for the region as one of the efficient ways for producing early warnings about climate extremes, including expected dry or wet conditions. It is demonstrated that seasonal cli- mate prediction could be a potentially effective solution to adapt to climate variability and change in small is- land countries with vulnerable economies.
The paper is organised as following. The current climate of Fiji, Samoa and Tuvalu, temperature and rainfall changes over the past decades and association of rainfall with the ENSO phases is presented in Section 2. In Section 3, we describe the impacts of recent severe weather and climate extremes on island countries: A tropical cyclone Evan which devastated Samoa and Fiji in December 2012; severe floods in Fiji in January and March 2012; and the 2011 drought in Tuvalu. The current status of the science of seasonal climate prediction and case studies which demonstrate the high skill of a dynamical climate model for forecasting dryer or wetter than aver- age (climatology) conditions are presented in Section 4. In Section 5, we discuss the obtained results and present concluding remarks.
Fiji, Samoa and Tuvalu are island countries located in the Western South Pacific Ocean with a tropical maritime climate characterized by warm temperatures all year around. Fiji is an island country in Melanesia. The country comprises an archipelago of 332 islands, of which 111 are permanently inhabited, amounting to a total land area of circa 18,333 km2. The two major islands, Viti Levuand Vanua Levu, account for 95% of the population of 837,271 [
The countries have two distinct seasons—a warm and wet season (November to April) and a cooler but rela- tively dry season (May to October). In this section, detailed analysis of climate in Fiji, Samoa and Tuvalu is presented. The analysis has been performed based on long-term historical climate records obtained from their respective NMSs. Meteorological data have been analysed using the Pacific Climate Change Data Portal avail- able through the Australian Bureau of Meteorology web site http://www.bom.gov.au/climate/pccsp/. Seasonal and interannual variability of temperature and rainfall have been examined and long-term trends have been de- rived. Significant climate variability in the countries has been identified and has been related to the ENSO. His- torical tropical cyclone data used in this study have been recently revised [
The annual average temperature across Fiji varies from around 27˚C in coastal regions to about 20˚C in elevated inland regions. Seasonal changes in the temperature are relatively small and strongly related to changes in the surrounding ocean temperature. Around the coast, the average night-time temperatures can be as low as 18˚C and the average maximum day-time temperatures can be as high as 32˚C. At elevated altitudes of the central parts on the main islands, average night-time temperatures can be as low as 15˚C.
Rainfall across Fiji is highly variable. On Fiji’s two main islands, Viti Levu and Vanua Levu, rainfall is strongly influenced by high mountain peaks up to 1300 m. On the south-eastern slopes of Viti Levu, near Suva, the average annual rainfall is about 3000 mm. In contrast, the lowlands on the western side of Viti Levu, near Nadi, are sheltered by the mountains and have an annual average rainfall of about 1800 mm with a well-defined dry season. Fiji’s climate is also influenced by the trade winds blowing from the east or south-east directions associated with the Hadley Circulation. The trade winds bring moisture onshore causing heavy showers in the interior and eastern parts of larger islands.
Analysis of historical climate records at two Fijian meteorological stations with long term high quality records, Nadi Airport and Laucala Bay (Suva), demonstrated that annual maximum and minimum temperatures have in- creased in both Suva and Nadi since the 1950s (
Rainfall data for Suva and Nadi Airport since the 1950s shows no clear trends in annual rainfall; but decadal variability is evident, especially at Suva (
Temperatures in Samoa are generally consistent throughout the year, with the annual average temperature around 26˚C - 27˚C. Seasonal differences are very small, with temperatures being coolest in July, when the cool, dry south-east trade winds are strongest. The warmest month is March. The country has distinct wet and dry seasons. On average 75% of Samoa’s total annual rainfall occurs in the wet season. Samoa’s mountains have a significant effect on rainfall distribution. Wetter areas are located in the south-east and relatively sheltered, drier areas-in the north-west.
In Apia, which is located in coastal area of the Upolu Island, annual maximum and minimum temperatures have increased at a rate of about 0.08˚C and 0.05˚C per decade respectively since the 1960s (
Apia’s rainfall shows no clear trends in annual rainfall since the 1960s (
In Funafuti, the capital of Tuvalu, there is little variation in temperature throughout the year. The maximum temperature is between 31˚C - 32˚C and the minimum temperature is between 25˚C - 26˚C all year round. Air
temperatures are strongly related to the ocean temperatures surrounding the islands and atolls of the country. While Tuvalu also has two distinct seasons, wet and dry, seasonal rainfall variability in Funafuti is not as strongly pronounced as in some other Pacific Island Countries, with rainfall averages being in a range between 200 and 400 mm each month of the year (
Annual average maximum and minimum temperatures have increased in Funafuti since the 1960s at rates of about 0.13˚C and 0.27˚C per decade, respectively (
year to year (
Recent research undertaken under the Pacific Climate Change Science Program found that in general climate of the Pacific region is undergoing gradual transition which is driven by both natural and anthropogenic forces. These changes are clearly manifested in steady air and ocean temperatures warming, sea level rise and shift in rainfall patterns [
Changes in historical climate extremes in the Western Pacific have been also analysed under the Pacific Aus- tralian Climate Change Science and Adaptation Planning Program (PACCSAP; [
Weather and climate extremes produce significant impacts on Pacific Island Countries. Torrential rain associ- ated with tropical cyclones and rainfall variability associated with the ENSO and the position of the SPCZ can result in heavy rainfall and subsequent flooding. On the other hand, prolonged rainfall deficit often results in droughts in the region. Such weather and climate extremes can cause economic and social hardship affecting ag- ricultural productivity, infrastructure and economic development which can persist for many years after the ini- tial impact. In this section, we describe extreme events (a tropical cyclone, floods and droughts) which have re- cently affected Fiji, Samoa and Tuvalu and have caused major economic and social disruption in these countries.
Recently, severe tropical cyclone Evan affected both Samoa and Fiji in December 2012. Severe tropical cyclone Evan was considered to be the worst tropical cyclone to affect Samoa since cyclones Ofa and Val devastated the country in 1990 and 1991, respectively. Cyclone Ofa produced gales or high winds, resulting in widespread damage due to a combination of storm surge and high seas (
The system was first noted on the 9 December, as a weak tropical depression about 700 km to the northeast of Suva, Fiji. Over the next couple of days the depression gradually developed further before it was named, by Re- gional Specialised Meteorological Centre (RSMC) Nadi, Evan on the 12 December, as it had developed into a tropical cyclone. During that day the system moved towards Samoa (
Significant progress in operational tropical cyclone forecasting has been achieved over the recent years which has resulted in improved early warning systems, but the death toll attributed to cyclones is still high. In Samoa, at least 14 deaths have been confirmed and 10 sailors were missing as a result of the impact of cyclone Evan. In Fiji, more than 400 people were evacuated by authorities from isolated resorts on small islands and from low-lying areas to emergency shelters in preparation for cyclone’s impact. Such timely measures yielded excel-
lent results-no death reports in Fiji related to cyclone Evan. However, with the cyclone passing in close vicinity of the Fijian islands, damage to vegetation and man-made structures was significant (see examples in
Significant progress in operational tropical cyclone forecasting has been achieved over the recent years which have resulted in improved early warning systems, but the death toll attributed to cyclones is still high. In Samoa, at least 14 deaths have been confirmed and 10 sailors were missing as a result of the impact of cyclone Evan. In Fiji, more than 400 people were evacuated by authorities from isolated resorts on small islands and from low- lying areas to emergency shelters in preparation for cyclone’s impact. Such timely measures yielded excellent results—no death reports in Fiji related to cyclone Evan. However, with the cyclone passing in close vicinity of the Fijian islands, damage to vegetation and man-made structures was significant (see examples in
Cyclone-related economic losses also remain very high, causing significant negative impacts on the econo- mies of both countries. Assessment of the cost of damage caused by cyclone Evan to Fiji indicates total losses of about FJ $194.9 million (US $108 million) or about 2.6% of country’s gross domestic product (GDP) [
Severe floods frequently affect major infrastructure, transportation, business and property in Pacific Island Countries. Often significant flooding is attributed to a tropical cyclone or a tropical depression. In this section, we present information on the significant floods in January and March 2012 which affected Fiji and were caused by tropical depressions.
On the 19 January the Fiji Meteorological Service’s RSMC Nadi reported that a tropical disturbance had de- veloped along an active trough of low pressure to the north of Fiji. Over the next couple of days, the tropical disturbance moved towards the south before it developed into a tropical depression and moved across Vanua Levu during the 21 January. Once over Vanua Levu, the depression remained stationary causing widespread heavy rainfall and strong winds over Fiji’s Northern and Eastern divisions until the 25 January, when it weakened and started to move away to the southeast.
Above average rainfall in the preceding three months across most the country and heavy rains from the mid- dle of January lead to unprecedented flooding and landslides, which led to a state of emergency being declared late in the month. The flooding was worst in the western part of the main island of Viti Levu. Eight people were reported to have died in the floods, most of them on the main island, and thousands of people were displaced, with about 3500 people placed in temporary shelters. Power and water supply disruptions were experienced in many areas, and roads were washed away. Further damage was caused to crops and to other infrastructure. Estimated losses due to the January 2012 flood were about FJ $40 million [
Just two months later, another severe flood devastated the country. The March 2012 flood was the worst flood to affect Fiji in recent times in terms of both the magnitude and the damage to properties (
This resulted in massive devastation across the western portion of Viti Levu. The towns of Ba and Nadi which have major rivers running through them were the worst affected during this natural disaster, with the flood height apparently setting a new record. The maximum flood water level recorded by Nadi River gauge before it malfunctioned was 7.6 m. The actual flood level is likely to be just over 8.0 m.
All towns in the Western Division suffered huge losses to business and residential properties. Torrential rain and subsequent flooding resulted in landslides and destruction of major roads and bridges (e.g. the Queens Road from Nadi to Lautoka was damaged by strong currents at several sections, and some bridges were damaged). The total damage caused by the March 2012 flood was estimated at around FJ $70 million [
A brief examination of the rainfall record reveals the Nadi Airport 2012 maximum 1-day rainfall of 291.9 mm (24 January) to be the fourth highest on record. This amounts to a 1-in-25 year rainfall event for data record that
begins in 1942. Of greater interest is the maximum 5-day rainfall of 677.0 mm recorded between 28 March to 1 April, the second highest 5-day event since 1942, amounting to a 1-in-70 year event. Both the highest 5-day record event in January 2009 and the March 2012 event are associated with prolonged SPCZ activity over Fiji and slow moving tropical depressions over the island nation. These events highlight the high chances of extreme flooding when high intensity rainfall events coincide with saturated ground conditions, especially in short, steep river catchments.
On the other side of hydrological extremes are droughts which also frequently affect islands of the Pacific (Fig- ure 15) impacting on the countries to a various degrees of severity depending on the drought’s length. For ex- ample, the 2011 Tuvalu drought severely impacted a small island nation living on coral atolls in the Pacific. Water resources in Tuvalu are heavily dependent on an abundant rainfall. If the country experiences a prolonged rainfall deficit, it results in critical shortage of water.
This prolonged drought episode was related to a strong La Niña event which affected the region in 2010-11. Typically, during La Niña years, oceanic waters are cooler than climatology (average conditions) in the region of the Western Central Pacific where Tuvalu is located. As a result of oceanic cooling, atmospheric convection is suppressed and hence the rainfall amount in Tuvalu during La Niña episodes is below average (see
In 2011, a number of countries and territories in the region, including Samoa, Tokelau and Tonga were af- fected by La Niña-induced rainfall deficit; however, Tuvalu was particularly seriously impacted. On the 28 Sep- tember 2011, the government of Tuvalu declared a state of emergency due to critically low water supplies. Households were rationed to about 40 litres (two buckets) of fresh-water a day as some parts of Tuvalu had just a two day supply of water left [
At Funafuti, rainfall was the lowest on record with only 515 mm (36% of the long-term average of 1430 mm) received between May to October. Severe drought in Tuvalu in the mid-1970s was also associated with a strong La Niña event [
Severe weather events and climate extremes often affect island nations in the Western Pacific and in this paper
we brought a number of recent examples of devastation caused by cyclones, floods and droughts. Damage to the natural environment and economy could be enormous especially taking in consideration that these countries are among the least developed countries. In extreme cases economic loses from such natural disasters could take away a large part of country’s annual GDP, as we demonstrated in Section 3.
To implement effective early warning systems with the aim of reducing the impacts of extreme events, cli- mate variability and climate change on the natural environment and society, authorities of island countries de- velop various action plans as well as adaptation measures. One of the most cost-effective and easy-to-implement ways to adapt to climate variability and change which we discuss here is for authorities to make informed deci- sions based on the latest achievements of climate science and to efficiently use seasonal climate prediction products which forecast expected rainfall and temperature over the coming season (next three months). Under the Pacific Adaptation Strategy Assistance Program and the Pacific Australia Climate Change Science and Ad- aptation Planning Program research has been conducted to evaluate the skill of the dynamical climate model POAMA (Predictive Ocean Atmosphere Model for Australia) for predicting seasonal rainfall [
In this section, we discuss how the timely utilization of seasonal climate prediction information could assist in adaptation to climate change and reduce the hardship which society experiences as the result of increased fre- quency in climate extremes. To illustrate, we use a case study for the 2011 Tuvalu drought.
The probability of accumulated rainfall falling in the lower climatological tercile (i.e. an elevated risk of be- low average rainfall) for the Western Central Pacific in 2011 was predicted by POAMA. The high skill of POAMA’s rainfall prediction in this region is based on the model’s ability to accurately forecast phases of the ENSO. The rainfall deficit in 2011 was related to one of the strongest La Niña episodes in recent history, which lasted for about two years (2010-2012). Observations from meteorological stations from December 2010 to January 2011 already demonstrated that accumulated rainfall in the Western Central Pacific region where Tu- valu is located was more than 500 mm below the climatological average. Here we verify that POAMA could produce reliable seasonal forecasting, to better understand the tendency for expected rainfall in the region.
The POAMA-based seasonal forecast issued in February 2011 for the three-month period February-March- April predicted that the rainfall deficit in the area of Tuvalu would continue, with the ensemble mean forecast quantity of seasonal rainfall about 225 mm below average (
POAMA demonstrated high skill in predicting seasonal rainfall over the Pacific region and the model’s out- puts could be used with confidence for forecasting both rainfall deficit leading to droughts and excessive rainfall potentially leading to floods. In
Collaborative efforts of the National Meteorological Services of Australia and Pacific Island Countries are un- dertaken through a number of Pacific programs (the Pacific Climate Change Science Program, the Pacific Adap- tation Strategy Assistance Program and the Pacific Australia Climate Change Science and Adaptation Planning Program), and have improved the understanding of current and future climate of the Pacific Island Countries. While some recent studies describe impacts of tropical cyclones, floods and droughts on South China [
The most pronounced impacts on the countries come from severe weather and climate extremes and in this paper we presented examples of such impacts caused by tropical cyclones, floods and droughts. Tropical cy- clones pose a major risk to the wellbeing and economies of Pacific Island Countries. The impacts of tropical cy- clones on the countries are severe and the costs of damage are astronomical. By the end of this century, projec- tions suggest decreasing numbers of tropical cyclones but a possible shift towards more intense categories [
This should be taken into consideration by the authorities of Pacific Island Countries and aid donors when developing adaptation strategies to reduce the increasing risk of climate extremes. Climate science could signifi- cantly assist with this task. As we demonstrated here, the modern science of seasonal climate prediction is well developed and could provide assistance to informed decision making in adaptation to climate variability and change. A range of web-based information tools to deliver seasonal climate predictions based on outputs from state-of-the-art dynamical climate model have been developed, to provide National Meteorological Services in the Western Pacific with information about expected climate extremes two-three months in advance.
We conclude with confidence that skilful seasonal climate prediction is an effective solution at the regional level to provide governments and local communities of island nations in the Western Pacific with valuable assis- tance in informed decision making for adaptation to climate variability and change.
The research discussed in this paper was conducted under the Pacific Climate Change Science Program (PCCSP), the Pacific Adaptation Strategy Assistance Program (PASAP) and the Pacific Australia Climate Change Science and Adaptation Planning Program (PACCSAP) supported by the Australian Agency for International Develop- ment (AusAID), in collaboration with the Department of Climate Change and Energy Efficiency (DCCEE).