Climate Outlook - Rainfall

Statement from the Caribbean Regional Climate Outlook Forum
21-22 May 1998, Kingston, Jamaica

THE CLIMATE OUTLOOK FORUM

A Climate Outlook Forum was convened on May 21-22, 1998 to formulate and communicate a consensus precipitation forecast for the Caribbean for the period of June-July-August 1998. In addition, the Forum intended to identify gaps in information and technical capability; facilitate research cooperation and data exchange within and between regions, and improve coordination within the climate forecasting community.

The Forum was comprised of climate researchers and representatives of meteorological and hydrological services and disaster preparedness officials from the Bahamas, the British Virgin Islands, Barbados, Puerto Rico, Antigua, Dominican Republic, Haiti, Cuba, Saint Lucia, Trinidad and Tobago, Netherlands Antilles, and Jamaica. Additional participating institutions included the Florida State University, the University of Maryland, the Hurricane Research Division of the National Oceanic and Atmospheric Administration (NOAA), the Center for Ocean-Land-Atmosphere Studies (COLA), the NOAA Climate Diagnostics Center (NOAA/CDC), the International Research Institute (IRI) for climate prediction, the NOAA Climate Prediction Center (NOAA/CPC), Service National des Ressources Naturaelles et Ressources en Eau, the Caribbean Meteorological Institute, the University of the West Indies, the Office of Disaster Preparedness and Emergency Management (ODPEM), the Office of Foreign Disaster Assistance of the US Agency for International Development (USAID/OFDA), the NOAA Office of Global Programs (NOAA/OGP), the Inter-American Institute for Global Change Research (IAI), the World Meteorological Organization (WMO), and the Caribbean Disaster Emergency Response Agency.

The Climate Outlook Forum was co-sponsored by the University of the West Indies, ODPEM, IAI, NOAA/OGP, WMO, USAID/OFDA.

METHODOLOGY

Sea surface temperature (SST) anomalies in the Pacific and Atlantic Oceans and Caribbean Sea are among the most important predictors of rainfall anomalies in the Caribbean region. The present SST anomalies in the central and eastern Pacific are among the largest ever recorded, with positive anomalies exceeding 4° C. Current predictions call for diminishing SST anomalies over the next few months in the equatorial Pacific; however, it is felt that this El Niño episode will continue to affect the Caribbean region during the period of June-August, 1998. Therefore, this Climate Outlook takes into consideration the lingering effects of the current El Niño event and the local topography of this region. The Climate Outlook was generated through analysis of historical climate records throughout the region, data available from the global climate monitoring system, and computer models of the evolution of the global SST field.

The Caribbean region has diverse topography and high regional variability in precipitation regimes. The wet seasons in the region are primarily of a bi-modal distribution with an early and late season peak, the exception being the Northern Leeward islands. This is reflected in the range of probabilities that were arrived at in the consensus. The forecast area was divided into 8 regions, with Cuba further subdivided into 2 regions. Due to this large scale nature of the forecast, there are likely to be local rainfall patterns that vary significantly from the outlook, particularly in mountainous regions.

An analogy for climate forecast probability is the rolling of dice. If we pretend that a six-sided die represents the range of possibilities in the climate system, and designate two sides of the die as above-normal precipitation, two sides as average precipitation, and two sides as below-normal precipitation, each climate possibility has a one in three (or 33.3%) chance of being rolled. This probability distribution represents the historical climate record in any given area. During an El Niño event, the climate die is loaded. Depending on the area considered, it may be either wetter or drier than the historical average. To simulate a wet scenario, we can replace a below-normal precipitation side of the die with one of above-normal precipitation. In this case, the probability that precipitation will be above-normal is 1 in 2 (or 50%), the average is 1 in 3 (or 33%), and below-normal is 1 in 6 (or 17%). In other words, the likelihood of a wetter climate increases, and that of a drier climate decreases, although dry conditions are still possible.

Seasonal climate forecasting is still a relatively new science and future forecasts can be improved through increased local and regional exchange of data and knowledge. This is of particular importance in the Caribbean region, where there is high spatial and temporal variability. Because of the relatively small spatial scale of most of the islands in the region, there is a need to explore the use of mesoscale models in climate prediction.

OUTLOOK

The Climate Outlook addresses the June to August 1998 period for the Caribbean region. The experts provided probability distributions to indicate the likelihood of below-, near-, or above normal rainfall for each sub-region (see attached map). For this purpose, "normal" is defined as the climatological mean. Users of this Outlook are strongly advised to contact participating institutions and other climate information sources for interpretation of this Outlook and for additional guidance. It is emphasized that the locations of the boundaries between the subregions are only qualitatively defined, and should be considered as transition zones rather than sharp boundaries.

Wetter than normal conditions are expected in the Western two-thirds of Cuba, while near normal conditions are expected in the eastern part of the island. Near normal conditions are also expected for Haiti and the Dominican Republic. For the Bahamas and Jamaica, a moderate tendency toward drier conditions is anticipated, while conditions in Puerto Rico and the Virgin Islands are expected to be normal to slightly above normal. The northernmost Leeward Islands are expected to be affected by the lingering effects of El Niño with lower than normal rainfall projected. This is thought to be in part due to the persistent strength of the North Atlantic high-pressure ridge. The southern Leeward Islands and the Windward Islands may experience rather normal conditions due to the possibility of more normal activity in the Inter-Tropical Convergence Zone (ITCZ). Rainfall from ITCZ disturbances have already started to affect Trinidad and Tobago by mid-May. Rainfall in the dry zone off the coast of South America has been well below normal since March 1997. Tercile analysis for the island of Curacao shows that the driest third of historical years include all strong El Niño events in the last 60 years. Consequently, continuing drier conditions are anticipated for the Outlook period.

The best educated consensus is that the effects of the current El Niño event on the Atlantic basin tropical cyclone activity will persist at least through the month of August. We expect the strong vertical shear over the Atlantic tropics associated with El Niño will inhibit tropical cyclone development. This effect is most pronounced over the Caribbean region and therefore, we would expect minimal hurricane activity in the region at least through the month of August. Although an occasional system might develop and even reach tropical storm strength, there is a low probability of any system reaching hurricane strength. However, it must be recognized that even inactive years can produce disasters. Firstly, enhanced precipitation from a weaker system can still cause flooding. Also, there can still be an occasional exception (i.e. a hurricane affecting the Caribbean region even during an overall inactive period). Although the exceptions are rare, normal hurricane preparedness efforts should still be maintained.