NOAA’s 2004 Atlantic hurricane season outlook indicates a 50% probability of an above-normal hurricane season, a 40% probability of a near-normal season, and a 10% chance of a below-normal season, according to a consensus of scientists at the National Oceanic and Atmospheric Administration's (NOAA) Climate Prediction Center (CPC), the Hurricane Research Division (HRD), and the National Hurricane Center (NHC). See NOAA’s definitions of above-, near-, and below-normal seasons.

The outlook calls for 12-15 tropical storms, with 6-8 becoming hurricanes, and 2-4 of these becoming major hurricanes. These numbers reflect a predicted ACE index in the range of 100%-160% of the median, and indicate a likely continuation of above-normal activity that began in 1995.

The predicted 2004 activity is based primarily on the ongoing active phase of the Atlantic multi-decadal signal, but also reflects expected warmer sea surface temperatures (SSTs) in the tropical Atlantic and Caribbean Sea than can be accounted for by the multi-decadal signal alone. The outlook also reflects the likelihood that ENSO-neutral conditions (no El Niño or La Niña) will continue through August-October, the peak months of the hurricane season. An updated hurricane outlook will be issued in early August.

DISCUSSION

1. Expected Activity- 50% chance above normal, 40% chance near normal, 10% chance below normal

An important measure of the total seasonal activity is NOAA’s Accumulated Cyclone Energy (ACE) index, which accounts for the collective intensity and duration of Atlantic tropical storms and hurricanes during a given hurricane season. The ACE index is also used to define above-, near-, and below-normal hurricane seasons (see Background Information). A value of 117% corresponds to the lower boundary for an above-normal season.

For the 2004 Atlantic hurricane season the current and expected climate conditions indicate an ACE index in the range of 100%-160% of the median. They also indicate a likely range of 12-15 tropical storms, with 6-8 becoming hurricanes, and 2-4 of these becoming major hurricanes [categories 3-4-5 on the Saffir-Simpson scale].  While it is reasonable to expect this range of tropical storms and hurricanes, the total seasonal activity measured by the ACE index can certainly be in the expected range without all three of these criteria being met. 

Many of the tropical storms and hurricanes in 2004 are likely to form during August-October over the tropical Atlantic and Caribbean Sea between 9°N-21.5°N (green box). These systems generally track westward toward the United States and Caribbean Sea as they strengthen. Historically, similar seasons have averaged 2-3 landfalling hurricanes in the continental United States and 1-2 hurricanes in the region around the Caribbean Sea. However, it is currently not possible to confidently predict at these extended ranges the number or intensity of landfalling hurricanes, or whether a given locality will be impacted by a hurricane this season.

2. Expected Climate Conditions – Active multi-decadal signal, above-average Atlantic Ocean temperatures, ENSO-neutral conditions

Beginning with 1995 all of the Atlantic hurricane seasons have been above normal, with the exception of the two El Niño years (1997 and 2002). This increased activity contrasts sharply with the generally below-normal activity observed during the previous 25-year period 1970-1994. The atmospheric and oceanic conditions controlling these very long-period fluctuations in hurricane activity are referred to as the Atlantic multi-decadal signal.

The active phase of the Atlantic multi-decadal signal began in 1995 (Goldenberg et al. 2001, Science), and has been a primary contributing factor to the increased hurricane activity observed since that time. This signal is a main factor guiding this year’s outlook. The signal is conducive to tropical storm and hurricane formation over the tropical Atlantic and Caribbean Sea, and includes 1) lower surface air pressure, warmer SSTs, and increased moisture across the central and eastern tropical Atlantic, 2) an amplified subtropical ridge at upper levels across the central and eastern North Atlantic, 3) reduced vertical wind shear in the deep tropics over the central North Atlantic, which results from an expanded area of easterly winds in the upper atmosphere (green arrows) and weaker easterly trade winds in the lower atmosphere (dark blue arrows), and 4) a configuration of the African easterly jet (wavy light blue arrow) that favors hurricane development from tropical disturbances moving westward from the African coast.

Also expected this season is a continuation of tropical Atlantic SSTs that are warmer than would be accounted for simply by the multi-decadal signal. This additional warmth is more conducive to hurricane formation than would be expected from the multi-decadal signal alone.

Another factor known to significantly impact seasonal Atlantic hurricane activity is ENSO (Gray 1984, Monthly Weather Review), with El Niño favoring reduced activity and La Niña favoring increased activity. Based on the most recent ENSO outlook issued by NOAA’s Climate Prediction Center, ENSO-neutral conditions are expected to continue in the tropical Pacific through July, and most likely through August-October as well. However, there is considerable uncertainty in the statistical and coupled model ENSO forecasts after July, and the comparatively wide range in the predicted ACE values reflects this uncertainty. A more confident ENSO forecast will be available for the updated Atlantic hurricane outlook to be issued in early August.

3. Multi-decadal fluctuations in Atlantic hurricane activity

Atlantic hurricane seasons exhibit prolonged periods lasting decades of generally above-normal or below-normal activity. These multi-decadal fluctuations in hurricane activity result nearly entirely from differences in the number of hurricanes and major hurricanes forming from tropical storms first named in the tropical Atlantic and Caribbean Sea.

Tropical cyclone activity during 1995-2003 has been considerably above normal in response to the active phase of the Atlantic multi-decadal signal. During 1995-2003 hurricane seasons have averaged 13 tropical storms, 7.6 hurricanes, 3.6 major hurricanes, and have an average ACE index of 159% of the median. NOAA classifies all but two of these nine seasons (the El Niño years of 1997 and 2002) as above normal. In contrast, during the preceding 1970-1994 period, hurricane seasons averaged 9 tropical storms, 5 hurricanes, and 1.5 major hurricanes, and had an average ACE index of only 75% of the median. NOAA classifies twelve (almost 1/2) of these 25 seasons as being below normal, and only three as being above normal (1980, 1988, 1989).

Time series of key atmospheric wind parameters and Atlantic SSTs associated with the Atlantic multi-decadal signal highlight the dramatic differences between these above-normal and below-normal periods. The active phase of the Atlantic multi-decadal signal was also present during the above-normal hurricane decades of the 1950s and 1960s, as indicated by comparing Atlantic SSTs and seasonal ACE values.

4. Uncertainties in the Outlook

There are two competing uncertainties in this outlook. The first is the possibility of an extremely active season similar to 2003, which resulted partly from near-record warmth across the deep tropical Atlantic, and partly from an amplified upper-level ridge across the western subtropical Atlantic and Gulf of Mexico in association with a persistent East Atlantic teleconnection pattern (Bell et al. 2004, Bull. Amer. Meteor. Soc., In press). In the event that the warmer tropical Atlantic SSTs and the subtropical ridge over the western Atlantic again become significantly enhanced, the ACE values could be at or even above the high end of the predicted range.

The second uncertainty involves the ENSO forecasts, which have exhibited low skill historically when issued at this time of the year and represent an ongoing source of uncertainty in the pre-season hurricane outlooks. There is currently considerably diversity in the statistical and coupled model forecasts of ENSO after July, which leads to a wider predicted ACE range in this outlook.

NOAA scientists will closely monitor the evolving climate conditions. A more confident ENSO forecast will be available for NOAA’s updated Atlantic hurricane outlook to be issued in early August, which is prior to the normal active portion of the Atlantic hurricane season.

CAUTIONARY NOTES

1) It is important to recognize that it is currently not possible to confidently predict at these extended ranges the number or intensity of landfalling hurricanes, or whether a particular locality will be impacted by a hurricane this season. Therefore, residents and government agencies of coastal and near-coastal regions should always maintain hurricane preparedness efforts regardless of the overall seasonal outlook.

2) Far more damage can be done by one major hurricane hitting a heavily populated area than by several hurricanes hitting sparsely populated areas or, of course, not making landfall at all. Therefore, hurricane-spawned disasters can occur even in years with near-normal or below-normal levels of activity. Examples of years with near-normal activity that featured extensive hurricane damage and numerous fatalities include 1960 (Hurricane Donna), 1979 (Hurricanes David and Frederic), and 1985 (Hurricanes Elena, Gloria and Juan). Moreover, the nation's most damaging hurricane, Andrew in 1992, occurred during a season with otherwise below normal activity.

FORECASTERS

NOAA's Climate Prediction Center
Dr. Gerald Bell, Meteorologist; Gerry.Bell@noaa.gov
Dr. Muthuvel Chelliah, Physical Scientist; Muthuvel.Chelliah@noaa.gov
Dr. Kingste Mo, Meteorologist; Kingste.Mo@noaa.gov

NOAA's Hurricane Research Division
Stanley Goldenberg, Meteorologist; Stanley.Goldenberg@noaa.gov
Dr. Christopher Landsea, Meteorologist; Chris.Landsea@noaa.gov

NOAA's National Hurricane Center
Eric Blake, Meteorologist; Eric.S.Blake@noaa.gov
Dr. Richard Pasch, Meteorologist; Richard.J.Pasch@noaa.gov