1998 Hurricane Season Probably a Taste of Seasons to Come; Near-Average Predictions Fell Short of Actual Storm Activity

Note to Editors: Forecast totals are in the attached chart. The complete hurricane forecast, plus related research and press releases, are available on the World Wide Web at: http://tropical.atmos.colostate.edu/forecasts/index.html

If the theory of Colorado State University hurricane forecaster William Gray and his associates is correct, the 14 Atlantic Basin storms that occurred this year are just a harbinger of hurricane seasons to come in the next few decades.

Gray believes certain climate phenomena indicate a return to conditions prevalent from the early 1930s to the late 1960s, when hurricane seasons had more intense or major storms along the U.S. East Coast. The 25 years between 1970-94 were relatively quiet, records show, but the upsurge in numbers and intensity since 1995 is likely to continue.

A professor of atmospheric sciences who has been issuing his forecasts for 15 years, Gray and his colleagues underestimated the number of storms for this season. However, beginning with the December 1997 forecast, the updated forecast continually increased the number of storms and overall hurricane activity.

"We forecast a slightly-above-average year for 1998 based on climate data that has proven accurate in the past," Gray said. "Easterly stratospheric winds, west African rainfall and the remnants of last year’s El Niño all led us to think this season would be only slightly above average."

Current climate signals indicate that 1999 also will be very active. The initial 1999 forecast will be issued Dec. 4 with updates in early April, early June and early August of 1999.

This season was the most destructive on record in terms of loss of life and property damage in the Caribbean basin because of hurricanes Georges and Mitch, according to Gray. In addition, an unusual number of storms – three hurricanes and four tropical storms – made landfall along the U.S. coastline. Only 1916, with nine, and 1985, with eight, had more named storms impacting the United States, he said.

"We were lucky because none of these seven named storms was of major hurricane intensity," Gray said, although hurricanes Bonnie and Georges approached major-storm status.

Despite greatly reduced hurricane activity in 1997 due to the strongest El Niño on record, the years 1994-98 have been the most active four-year period of hurricane activity this century. The four consecutive years have spawned 53 named storms, 32 hurricanes and 15 major hurricanes.

Gray believes this stems in part from a phenomenon called the Atlantic Ocean thermohaline circulation, which moves Atlantic waters northward from the vicinity of the Caribbean to an area east of Greenland. There, the current sinks to deep levels, moves southward and flows into the South Atlantic Ocean and beyond.

Warm water in the conveyor belt – perhaps one degree Fahrenheit above the average North Atlantic sea surface temperature of about 45 Fahrenheit in the regions west of the British Isles – is associated with more intense hurricanes and more major (Saffir-Simpson category 3-5 storms with winds of 111 mph or above) landfalling storms along the eastern seaboard, Gray said. Cool water and low salinity tend to dampen this ocean circulation and hurricane activity.

Despite the team’s belief that hurricane activity is increasing, team members called for numbers only slightly above average this year. The 1998 season predicted 10 named storms, six hurricanes and two intense hurricanes in their third and final update issued in August. The actual season total was 14 named storms (Nicole was a late-season Eastern Atlantic anomaly), nine hurricanes and three intense hurricanes. Long-term statistical averages yield 9.3 tropical storms, 5.8 hurricanes and 2.2 intense hurricanes annually.

"We didn’t anticipate such an active season," Gray said. "Given the variables we had and prior historical records, there is no way we could have predicted such an active hurricane season.

"Hurricane Mitch, which wreaked such devastation on Central America, was the strongest late-season storm on record," he said.

Gray and research team members Chris Landsea, Paul Mielke and Kenneth Berry rely primarily on the following factors to issue their forecasts: the strength or weakness of El Niño the direction of equatorial stratospheric winds at 68,000-75,000 feet; rainfall in the West African Sahel region; temperature and pressure readings in West Africa; Caribbean Sea-level pressure readings; Atlantic sea-surface temperature readings; tropospheric winds at 40,000 feet; and pressure readings in the northeast Atlantic.

Gray said three key signals led the team to expect a much less active season in 1998 than what actually took place, including:

  • Easterly winds from the Quasi-Biennial Oscillation, a stratospheric, equatorial wind pattern that shifts direction every 26 to 30 months. Easterly winds tend to inhibit tropical storm formation.
  • A very dry season in the Sahel region of West Africa in both 1997 and through July 1998. Dry conditions historically are linked with relatively few tropical storms.
  • Residual El Niño effects in the eastern equatorial Pacific through July. Westerly winds from El Niño tend to shear the tops off tropical depressions and prevent them from becoming full-fledged storms.

"Our research shows that while we improve on climatology most of the time, 10 to 20 percent or so of our forecasts will fail," Gray said. "The 1998 forecast fell short because global circulation variables that had good predictive value in prior years just didn’t work this time.

"Next year we’ll be calling for above-average hurricane activity, and we believe this is a reflection of the new hurricane era we entered in 1995. We expect this trend to continue for the next few decades."


11/98 8/98 6/98 4/98 12/97
Named Storms (9.3)* 14 10 10 10 9
Named Storm Days (46.9) 66 50 50 50 40
Hurricanes (5.8) 9 6 6 6 5
Hurricane Days (23.7) 47 25 25 20 20
Intense Hurricanes (2.2) 3 2 2 2 2
Intense Hurricane Days (4.7) 9 5 4 4 3
Hurricane Destruction Potential (70.6)** 139 75 70 65 50
Net Tropical Cyclone Activity (100%) 161% 110% 100% 95% 90%

* Number in ( ) represents average year totals based on 1950-1990.

** 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.