Note to Editors: Forecast totals are in the attached chart. The complete hurricane forecast and related research and press releases are available on the at http://tropical.atmos.colostate.edu/forecasts/index.html
Data since the official June 1 start of the 2000 hurricane season have convinced William Gray and his hurricane prediction team that this year, while active, won’t be quite as bad as they had thought in early June.
In the fourth forecast and final update for the year, the Colorado State University team is calling for 11 named storms, seven hurricanes and three major hurricanes.
That’s how the year started when the initial prediction was issued back in December 1999, and that’s what they anticipated when their April update was issued. By June 1, however, the official beginning of the June 1-Nov. 30 Atlantic Basin hurricane season, Gray and colleagues saw reason to raise the number in each category of storm by one.
They’re back to 11, seven and three, due to recently observed cooler Atlantic Ocean surface water temperatures and a weakening La Nina that should not enhance Atlantic Basin hurricane formation quite as much as the team’s expectations in early June.
An average year during the period 1950-1990 has seen 9.3 tropical storms, 5.8 hurricanes and 2.2 major hurricanes (those with wind speeds above 110 mph).
The fact that two months have passed without so much as a named storm is irrelevant, Gray said. Last year, a very active one, produced only one tropical storm, Arlene, before Aug. 19, generally acknowledged as the start of the intense part of the season, and hurricane activity in 1998 also started very late.
"The fact that we haven’t had an early season storm doesn’t mean anything," Gray said. "There is no correlation between June and July storms and what may take place later in the season; in fact, there’s some evidence for a slight negative correlation (i.e., the absence of early season storms is made up for by late-season activity)," according to Gray.
He expects the severe part of the season to occur about when it usually does, for 60 days beginning around Aug. 20 and continuing until roughly that date in October.
What led Gray, professor of atmospheric science, and his colleagues to return to their original prediction- Geophysical phenomena, what Gray calls "climate signals," seemed ready to take an upturn based on April-May data but now have weakened to some extent.
"Sea surface temperatures in the Atlantic Basin have not warmed quite as much as expected," said Gray (the Atlantic Basin consists of the North Atlantic, Caribbean and Gulf of Mexico). Because tropical depressions that could grow into named storms and beyond require a pool of warm water to draw energy from, and because estimates say that pool must be at least 100-150 feet deep, cooler surface water temperatures in the Atlantic could inhibit storm formation.
So, too, has La Niña slowly weakened. This mass of cold water in the equatorial eastern Pacific, when strong, produces weak westerly or even easterly winds that promote hurricane formation. At the beginning of summer Gray and his team saw evidence that the cool water would persist, but recent readings indicate that La Niña has slightly weakened over past weeks, making it a little less favorable for hurricane formation.
What coastal dwellers should anticipate, Gray said, is for an active hurricane season but one likely to be slightly weaker than those that occurred in 1995, 1996, 1998 and 1999. The five years beginning in 1995 constitute the five most active consecutive years of hurricane activity in history, producing 65 named storms, 41 hurricanes and 20 major hurricanes.
Gray and colleagues believe that the current upturn in storms has nothing to do with global warming but instead is a result of a stronger Atlantic Ocean thermohaline system, or Atlantic conveyor belt. Moving warn, relatively salty water to the cold seas of the North Atlantic produces a condition where the salty, hence relatively heavy, water sinks, carrying with it its warmth.
The result of a strong (warm and salty) Atlantic conveyor belt does not cause many more weaker storms, Gray said. Its main influence is to produce more major storms (in Saffir-Simpson categories 3-5 with wind speeds of 111 mph or higher). Such enhanced conditions existed from the late 1920s to the late 1960s, causing numerous major storms to make landfall on the U.S. Atlantic coast. This tendency has again been picking up since 1995. The periods 1900-25 and 1970-74 saw weak (i.e., relatively cool and unsalty water) in the Atlantic conveyor belt and the number of major landfalling storms during these periods were reduced from the long-term average.
Gray and his team’s final forecast update for 2000 (he’ll issue an assessment of the year’s predictions in late November) means a lot of storms will be packed into a relatively short period of time, August-October (few storms take place in November).
Last year’s forecast-which did not change numbers throughout the season-was one of the most successful for the Colorado State team, who’ve been predicting the number and intensity of Atlantic Basin hurricanes since 1984. The team called for nine hurricanes and 40 hurricane days. There were eight hurricanes and 43 hurricane days. They forecast four major hurricanes and five were observed.
Gray and co-authors Chris Landsea, Paul Mielke and Kenneth Berry, with the assistance of John Shaeffer, Todd Kimberlain, Eric Blake and Bill Thorson, use a variety of climatic factors in their forecasts. In addition to stratospheric winds, North Atlantic sea surface temperatures and La Niña-El Niño conditions, they look at a ridge of barometric high pressure called the Azores High, warmer sea-surface temperatures throughout the Atlantic, two measures of west African rainfall and mid-latitude oceanic wind patterns in the Atlantic and Pacific oceans.
Although Gray argues strongly that the public and media talk only of the influences of La Niña and El Niño affecting hurricane activity, a number of other climate factors, such as Atlantic sea surface temperatures and surface pressure also are involved. These latter factors do note look quite as favorable as they did in early June but still more favorable than average.
"We thought we had verification that La Niña would continue relatively strong throughout the season, but it appears to be fading a little more rapidly than we expected," Gray said. "Under those circumstances, we’re reducing our early June forecast numbers a little, although people should remember that there still is the possibility of this being a very active season, especially compared with the 1970-1994 period."
Predictions continue to place the the probability of one or more major hurricanes coming ashore somewhere along the U.S. coastline from Brownsville to the Maine-Canada border at 60 percent (the long-term mean is 52 percent).
The probability of one or more major hurricanes making landfall on the U.S. East Coast, including the Florida peninsula, is 39 percent, compared with an average the past century of 31 percent. The Gulf Coast faces a 34 percent probability of facing one or more Saffir-Simpson 3-4-5 storms (the last century’s average is 30 percent).
The Caribbean basin faces a possibility of landfall by one or major storms about 10 percent above the past century’s average.
GRAY RESEARCH TEAM
HURRICANE FORECAST FOR 2000 SEASON
|Dec 1999||April 2000||June 2000||August 2000|
|Named Storms (9.3)*||11||11||12||11|
|Named Storm Days (46.9)||55||55||65||55|
|Hurricane Days (23.7)||25||25||35||30|
|Intense Hurricanes (2.2)||3||3||4||3|
|Intense Hurricane Days (4.7)||6||6||8||6|
|Hurricane Destruction Potential (70.6)**||85||85||100||90|
|Maximum Potential Destruction (61.7)||70||70||75||70|
|Net Tropical Cyclone Activity (100%)||125||125||150||130|
* Number in ( ) represents average year totals based on 1950-1990 data.
** Hurricane Destruction Potential measures a hurricane’s potential for wind- and ocean-surge damage. Tropical Storm, Hurricane and Intense Hurricane Days are four, six-hour periods where storms attain wind speeds appropriate to their category on the Saffir-Simpson scale.