Andrew caused an unprecedented $26 bn in damage in 1992, even though the storm struck south of the densely populated areas of Miami, Miami Beach, and Fort Lauderdale. To mark the anniversary and assess progress in hurricane prediction, the National Hurricane Center (NHC) in Miami assembled a panel of scientists who were active then and now.
It was clear that great progress has been made in tracking storms and predicting their onward paths. Max Mayfield, who worked at NHC when Andrew hit, and directed the centre at the time of Katrina, recalled that Andrew, the first named storm of 1992, formed on 14 August off the African coast. It became a hurricane on 22 August, just two days before hitting the Bahamas and south Florida.
Mayfield distinguished two types of hurricane forecast – track and intensity. Neither was particularly well developed in 1992. Over the past 20 years, he said, great progress has been made in predicting a hurricane's track. Television viewers are used to seeing a storm's projected path and its error bars. Still, he said, even now, the projected tracks are not as precise as he would like, and the models need to be improved.
As for intensity, "no one that I know of thought [Andrew] was going to be a Category 5 hurricane before it hit," Mayfield said. "We did not have very good models for guidance on intensity forecasting."
We still do not, acknowledged Richard Knabb, who was recently appointed director of NHC. Forecasters are typically off by one category of intensity, even a day before landfall, he said, adding that "rapid intensification continues to be a challenge", as it was with Andrew.
The Hurricane Forecast Improvement Project (HFIP), which is currently underway, is intended to improve intensity forecasting among other parameters, said Knabb. According to NOAA, the specific goals of the HFIP are to reduce the average errors of both hurricane track and intensity forecasts by 20% within five years and 50% in 10 years, with a forecast period out to seven days.
Frank Marks, head of NOAA's Hurricane Research Division (HRD), who flew into Andrew to assess its strength as it approached Florida, agreed that as track forecasts improve, intensity and structure have to become the focus of research. According to the panellists, even in the age of sophisticated satellites, instrument-laden aircraft flying into the heart of hurricanes remain an indispensable tool for forecasters.
The scientists reckoned that improved predictions of storm surge would also be helpful, but they said that there are simply too many variables to develop good models with our currently available data and understanding of the dynamics of surge.
environmentalresearchweb asked how climate change might affect the improvements in forecasting that researchers are seeking. Hugh Willoughby of Florida International University, who was HRD director in 1992, said that statistical models of the type used back then would no longer be representative after significant climate change, as they were based on history during a different climate.
Based on recent research, Willoughby said, it is likely that "the strongest hurricanes will get stronger, because the oceanic heat source is stronger, but because of increased shear of the surrounding winds, the numbers will go down, and the locus of activity in the Atlantic is more likely to move to the open Atlantic from the Gulf [of Mexico]." By the time the threat has altered enough to affect insurance rates, he added, "the numerical predictions may have improved enough that we will know how it is changing as it changes."