Hurricane Allen -- one of the most intense tropical cyclones on record -- is showing up the strengths and limitations of a modern hurricane forecast and warning system.
Meteorologists at the US National Hurricane Center (NHC) at Coral Gables, Fla., are able keep close track of Allen using satellites, aircraft, surface observations, and radar. They can fine-tune their warnings as the storm approaches populated areas. But they cannot adequately anticipate when and how much it will intensify or weaken or in which direction it may suddenly jump.
This is not a failure of weather observing technology, but a lack of fundamental scientific understanding. The greatest storms on Earth represent a frontier of meteorological knowledge. Hurricanes are self-sustaining circulation systems whose inner dynamics and interaction with the rest of the atmosphere are poorly known. Because of this, says NHC forecaster Joseph M. Pellissier, hurricane forecasting involves a good deal of informed guessing.
He explains that the forecast has three components -- storm development, changes of intensity, and changes of motion. Forecasters show least skill in anticipating hurricane birth.
Most North Atlantic hurricanes develop from distrubances called "waves in the easterlies." These are troughs of low pressure, often accompanied by showers, that move eastward from the African coast in the general Trade Wind flow. In a typical season, NHC may track around 100 of these waves. About 11 them become tropical storms with well developed circulations around low pressure centers and with up to gale force winds. Only around eight make it to the hurricane stage. More rarely, a hurricane may develop from a tropical depression that forms from a low pressure feature in the upper air.
In any case, the crucial combination of factors that determines which tropical disturbance or storm will become a hurricane is unknown. NHC meteorologists have to monitor all the candidates and see what happens.
Forecasters do show a little more skill in predicting changes in intensity once a hurricane has formed. Nevertheless, there are major uncertainties. Hurricanes feed on energy represented by heat and moisture drawn from the warm sea surface. Meteorologists have only a sketchy understanding of how this energy flows through the storm and powers it.
For example, the moisture cutoff and the roughness of terrain experienced when a hurricane moves over land can disrupt and weaken its circulation. But the storm may reintensify when it moves out over the ocean again. Meteorologists cannot accurately forecast these and other intensity changes. This is a continuing concern when a hurricane is moving among islands in the Caribbean.
Forecasters show most skill in predicting hurricane motion. this is relatively easy, Dr. Pellissier motion. This is relatively easy, Dr. Pellissier explains, because the storms move more or less with the large scale atmospheric circulation in which they are embedded like a cork in a stream. For a hurricane such as Allen moving across the southern North Atlantic toward the Carribbean, this means drifting with the clockwise flow of air around the "Bermuda High" -- the high pressure region that dominates North Atlantic weather during the hurricane season.
Again there are uncertainties. Dr. Pellissier points out that the force steering the storm is the resultant of forces due to winds blowing at different heights, all trying to steer in different directions. Also, there are no river banks to channel the main flow as in teh case of the cork in the stream. Thus it is important to predict the large scale wind field at several levels.
That is hard to do. Over the ocean, weather data are sparse. Cloud motions in satellite photos help analysts estimate winds 7 to 14 kilometers high and near the surface. but there is a gap at middle levels where clouds are few in the tropics. Yet hurricane movement correlates best with middle level winds.
Furthermore, Dr. Pelissier notes, and unlike the metaphorical cork, the hurricane is a vigorous circulation in its own right that interacts with the larger wind flow trying to steer it. The forecaster has to try to anticipate that interaction, which accounts for something like 15 percent of hurricane movement, and judge how the storm will move in relation to the steering current. "What a storm does seems to depend upon its fine detail and that's not well understood," Dr. Pelissier says.