Rainfall outlook for December 2002 to February 2003
The El Niño related convection and tendency towards above average rainfall is expected to persist in equatorial countries, including Tuvalu and Tokelau
Rainfall should trend below average from the Solomon Islands, southeast to the Southern Cook Islands
The region of enhanced convection in the NINO3 region should continue to persist from December 2002 through February 2003, resulting in a high likelihood of above average rainfall in both Western and Eastern Kiribati.
Average or above average rainfall is expected in Tuvalu and Tokelau. A tendency towards below average rainfall is forecast for a broad region from the Solomon Islands southeast to the Southern Cook Islands, including Vanuatu, New Caledonia, Fiji, Tonga and Niue.
Average or below average rainfall is likely in the Marquesas Islands. Near average rainfall is more likely elsewhere.
The forecast model skills for this outlook are generally moderate or high for most countries.
Rainfall outlook map for December 2002 to February 2003
Tropical cyclone update
‘Yolanda,’ the first tropical cyclone of the season, developed between Fiji and Samoa on 5 December, and was tracking southeast over Tonga on 6 December, with estimated maximum sustained winds of 75 km/h.
The chance of tropical activity for the December- April period still remains higher than normal for some countries east of the date line. They are still very likely about and west of the date line (but may occur with a lower than nomral frequency there).
Probabilities of rainfall departures from average
Broad-scale rainfall patterns and anomalies in the southern tropical Pacific area are estimated from the state of large-scale regional climate factors, such as La Niña or El Niño, their effect on the South Pacific and Tropical Convergence Zones, surface and sub-surface sea temperatures, and computer models of the global climate.
Rainfall estimates for the next three months for Pacific Islands are given in the adjacent table. The tercile probabilities (e.g. 20:30:50) are derived from the interpretation of several global climate models. They correspond to the odds of the observed rainfall being in the lowest (driest) one third of the rainfall distribution, the middle one third, or the highest (wettest) one third of the distribution. On the long-term average, rainfall is equally likely (33% chance) in any tercile.
The probabilities shown express the expected shift in the distribution from the long-term average, based on predictions of oceanic and atmospheric conditions. The amount of inter-model forecast consistency is indicated by the levels of confidence expressed in the table.
Island Group | Rainfall Outlook | Confidence in the Outlook | |
---|---|---|---|
Western Kiribati | 20:20:60 | (Above) | High |
Eastern Kiribati | 20:20:60 | (Above) | High |
Tuvalu | 20:45:35 | (Average or above average) | Moderate |
Tokelau | 20:45:35 | (Average or above average) | Moderate |
Wallis & Futuna | 20:50:30 | (Near average) | Moderate |
Samoa | 30:50:20 | (Near average) | Moderate |
Northern Cook Islands | 20:60:20 | (Near average) | High |
Society Islands | 25:50:25 | (Near average) | Moderate |
Austral Islands | 25:50:25 | (Near average) | Moderate |
Pitcairn Island | 30:50:20 | (Near average) | Moderate |
Papua New Guinea | 40:50:10 | (Average or below average) | Low |
Fiji | 45:35:20 | (Average or below average) | Moderate-high |
Tonga | 45:35:20 | (Average or below average) | High |
Southern Cook Islands | 40:40:20 | (Average or below average) | Moderate |
Marquesas | 45:35:20 | (Average or below average) | Moderate |
Solomon Islands | 50:30:20 | (Below average) | Moderate |
Vanuatu | 60:30:10 | (Below average) | High |
New Caledonia | 55:30:15 | (Below averge) | High |
Niue | 50:30:20 | (Below average) | Moderate – high |
Rainfall outcomes as estimated from models and historical records. The third column indicates the probability of bottom (below), middle (average) or top (above) tercile rainfall, where a percentage is given. The rainfall outlook (second column) is subjectively estimated from the probabilities of bottom, middle and top terciles.