There's good news for hurricane forecasters. They soon should get a better take on where these powerful storms will track.
Beginning this year, aircraft will drop probes, called dropwindsondes, into a hurricane and the surrounding atmosphere. The probes measure winds, air pressure, temperature, and humidity between the aircraft altitude and the sea surface. Their data give an unprecedented view of the wind field in which the storm is embedded. Using these data, computer forecasting programs then can better project where those external winds will steer the storm.
As hurricane Bertha, the first Atlantic hurricane of the season, tears through the Caribbean, forecasters should have plenty of storms to work with this year. Hurricane forecaster William Gray at Colorado State University in Fort Collins says he expects "about an average hurricane season" from June 1 to Nov. 30. Dr. Gray's forecast, issued June 6, calls for 10 tropical storms strong enough to be named, including 6 hurricanes, 2 of which will be major storms. Last year there were 19 named storms in the second most active season on record. Gray's forecast will be updated Aug. 7 when the most active part of the Atlantic hurricane season usually begins.
Gray uses a statistical forecast method that takes account of several factors influencing Atlantic hurricanes. They include such influences as the direction of global stratospheric winds, West African rainfall, sea-level air pressures in the Caribbean and off West Africa, and the presence or absence of El Nio warm-water conditions in the equatorial Pacific Ocean.
Robert Burpee, director of the National Oceanic and Atmospheric Administration (NOAA) National Hurricane Center in Miami, calls the new hurricane forecasting technology "a major breakthrough in hurricane forecasting." He notes that the accuracy in forecasting for the Atlantic and eastern Pacific regions for which he is responsible has improved at an average rate of 1 percent a year for the past couple of decades. Tests of the new forecasting scheme using dropwindsonde data show it can bring, at one stroke, as much improvement as has been accumulated over the past 20 years, he says. That means that Dr. Burpee's team expects to cut errors by 16 percent to 30 percent for storm track forecasts made 12 to 60 hours in advance.
That would be a badly needed accuracy gain. Hurricane meteorologists take pride in the fact that, as storm surveillance and forecasting have improved, hurricane-related fatalities have gone down as a result of timely warnings. But NOAA administrator James Baker points out that, even though the fatality rate has dropped, hurricane damage costs are shooting up. He notes that, in the United States alone, more than half (129 million) of the population now lives within 50 miles of the coast. He adds that this means that an increasing share of the country's population is becoming exposed to hurricane danger along the coasts of the Atlantic and the Gulf of Mexico. The same can be said for other areas of the Gulf and Caribbean undergoing extensive development.
Robert Sheets, Burpee's predecessor at the National Hurricane Center, has estimated that population has grown by 3 to 4 percent a year in hurricane-prone areas. That growth rate could overwhelm the relatively slow improvements in forecasting accuracy.
In a paper published in the May Bulletin of the American Meteorological Society, Burpee and several colleagues warn: "Unless the rate of forecast improvements can be accelerated, the downward trend of hurricane casualties is not likely to continue."
This is the kind of accuracy gain that Burpee thinks his team now has in hand. He explained during an interview that realizing that gain is, for him, a personal crusade. After working 18 years researching the new observing scheme, he says, "I'm very emotionally involved [in seeing] it transitioned into practice."
He says the practical payoff could be substantial. Warnings currently are issued 18 to 24 hours before hurricane landfall for a coastline zone averaging 555 kilometers (344 miles). That's no longer enough lead time for orderly evacuation of some the now intensively developed coastal regions. It also overwarns a large area to allow for forecast errors. The actual path of damaging winds and storm surges for a given storm is usually less than 185 km (115 miles). If forecasters sharpen their storm track projections and reduce the overwarned zone, that would cut down congestion as people evacuate a smaller area more likely to be in real danger. Improved forecast accuracy could also lead to more confident longer-range forecasts so that warnings could be issued earlier. That would allow more evacuation time.
Burpee says that economic savings could also be substantial. Preparation costs can run about $346,000 per kilometer of coastline warned, based on a recent estimate, he says. He estimates that, if the new forecasting scheme using the dropwindsonde data lets forecasters narrow their warning zone by only 5 percent, the average savings in local storm preparation costs could run to 80 times the added cost of getting the dropwindsonde data.
NOAA will gather those data with its own aircraft fleet, including a new jet airplane. They will complement the "hurricane hunter" reconnaissance missions flown by the US Air Force. The new data may also give meteorologists insight into what Burpee calls the other "big uncertainty" in his forecasts - storm intensity. He explains that "we have very little skill in intensity forecasting ... basically because we don't understand the physics."
Sudden changes in hurricane intensity can take forecasters by surprise. For example, Hurricane Opal intensified explosively and accelerated toward the Florida panhandle last October. Burpee notes that this happened at night when people were sleeping. It was hard to get public attention for the new danger which forecasters hadn't anticipated.
The probes and the aircraft instruments will give new information on how a hurricane interacts with its atmospheric and sea-surface environment. This may help elucidate the physical processes involved in hurricane evolution.