"This season was kind of a wake-up call for us. We can do a lot better than we are doing," said Dr Hugh Willoughby, Miami-based director of the United States Hurricane Research Centre. Noting that Hurricane Mitch's devastation was caused mostly by floods and landslides, not by its winds, he added: "We focused on wind speed. Now it's time for a new direction."
It has to be an urgent new direction, according to leading hurricane expert William Gray, of Colorado State University, since he predicts an equally bad season next year.
Dr Willoughby, experts at the National Hurricane Centre - also in Miami - and top forecasters at Colorado State University all agree on one thing: they blew it. All predicted a fairly average June to November season this year of 10 tropical storms, with six of them becoming hurricanes.
In fact, 1998 saw 14 tropical storms, of which nine grew to hurricane strength. More important, they included the most destructive hurricane in 200 years, Mitch, which killed at least 11,000 in Honduras, Nicaragua and the rest of Central America, left thousands more missing and destroyed hundreds of thousands of homes.
"These deaths are not just numbers. They are real people, they are kids, moms and dads, and friends," said Jerry Jarrell, director of the National Hurricane Centre. "To say we did a great job sounds a bit hollow. We might have done more." He admitted that the traditional method of forecasting hurricanes - using the five-point Saffir-Simpson scale to measure wind speed - had proved woefully inadequate this year.
"The Saffir-Simpson scale was designed to give a visual picture of winds. It doesn't talk about rainfall," Mr Jarrell said. He suggested a feasibility study into a new system that would predict how much rain a hurricane would carry.
The experts admitted that they may have underestimated the effect of the so-called La Nina weather phenomenon - a cooling of tropical waters in the Pacific that affects air circulation around the world. Last year, when the opposite phenomenon - El Nino, or warmer Pacific waters - prevailed, the number of hurricanes was below average.
Mr Jarrell also implied that Third World communications problems may have played a role in the Honduran disaster. "In the initial warning we put out, we had a problem contacting Honduras," he said.
Other experts conceded that they had expected Hurricane Mitch to continue due west in late October, instead of swerving south to slam into Honduras and its Bay Islands.
Referring to Hurricane Georges, which battered the Florida Keys in late September and killed hundreds in Haiti and the Dominican Republic, Mr Jarrell said: "We will never know how close we came to a comparable [to Hurricane Mitch] disaster in Key West. We predicted Georges would be a category three hurricane. It hit land as a two but it could just as easily have been a category four."
He attacked what he said was the complacency of Florida Keys residents, 60 per cent of whom refused urgent warnings to evacuate, although he admitted that many had been afraid to flee across the only road out - a narrow two-lane, sea-level highway and causeways to mainland Florida.
The head of the National Hurricane Centre also criticised television reporters who competed for the most dramatic images by going out in hurricane force winds, clinging to lampposts or wading in floods. "This is the worst possible example we can set. Sooner or later, we're going to lose some of these reporters," Mr Jarrell said. Speaking to media representatives, he said: "If any of you or your organisations are responsible for that, shame on you."