The strong hurricanes bearing down on the Caribbean and Gulf of Mexico this year have caused renewed interest in predicting these violent tropical cyclones. Forecasting them has improved significantly in the last 25 years, providing more lead time for evacuation and other protective measures. But, it is still an imprecise science that needs a lot of fine-tuning.
The giant storms that spin out of the warm waters near the Equator take an enormous toll in death and destruction worldwide each year. The U.S. government's oceans and atmosphere agency NOAA says the number and intensity of hurricanes in the tropical north Atlantic Ocean has risen in the past decade after a 25-year lull. The director of the agency's National Hurricane Center, Max Mayfield, told a U.S. Senate committee that it is as if someone threw a switch in 1995.
"We've had a lot more hurricanes, not a record number of major hurricanes, but close," he said. "We've had a lot of activity again and the research meteorologists tell us that we are in for another 10 or 20 years or more of this active period here."
A recent study in the journal Science by a group of U.S. university researchers shows that the annual number of the strongest hurricanes - those with winds higher than 210 kilometers per hour - has almost doubled globally since 1990, from 10 to 18.
Some scholars blame global warming. They say the heating of the atmosphere by the trapped gasses from burning fossil fuel has caused warmer oceans and more moisture in the air, ideal conditions for tropical cyclones. But Max Mayfield and others say it is premature to make that connection.
"There are cycles with active periods and inactive periods," he said. "For example, the 1940s, '50s, and '60s were very, very active with lots of hurricanes, lots of major hurricanes, and in the '70s, '80s, and early '90s, the numbers really dropped down. Without invoking global warming, I think that just natural variability alone is what this can be attributed to."
Whatever the reason for the latest increase in hurricane activity, forecasting them is the key to protecting the population.
Mr. Mayfield's agency accurately predicted hurricane Katrina's path four to five days before it reached land in late August and devastated a wide swath of the U.S. Gulf of Mexico coast.
His boss, U.S. oceans and atmosphere agency chief Conrad Lautenbacher, says the prediction saved tens of thousands of lives and contrasts the situation to the loss of hundreds of thousands of lives in December's Indian Ocean tsunami.
"What makes the difference? Observations," he said. "There was nothing in place in the Indian Ocean to allow observations to be transmitted to an educated and ready-to-act public. In the United States, where we have global observations starting with satellite systems that detect things immediately as they form, ground based systems, air breathing systems, we had four or five days warning time for people to prepare and be ready to move out."
Sixty or 70 years ago, no advance warnings were possible. Then, in the 1940s, reconnaissance aircraft began hunting such storms regularly. Much later, orbiting satellites started monitoring hurricanes and scientists began putting the data, such as temperature, wind speed, direction and, precipitation into computers to make mathematical models for prediction. Still, just 25 years ago, forecasters could provide only 12 to 18 hours advance notice of an impending hurricane.
But hurricane Katrina was relatively easy to predict because it stayed on a steady track. Most storms are not so cooperative.
"Hurricanes by nature are notoriously difficult to predict. They are incredibly fickle beasts," said University of South Alabama meteorologist Keith Blackwell. He told the U.S. Senate committee that, unlike Katrina, many storms are unpredictable four to five days in advance. He argues that three day forecasts are currently the most reliable.
"We have come a long way with track forecasting of hurricanes, but there still are often severe limits to our skills several days into the future. Much more work remains," he said.
Mr. Blackwell also says forecasters still have little ability to predict important hurricane conditions such as intensity, size, rainfall amounts, and the tsunami-like surge of ocean water, which is often more devastating than high winds.
He points out that one of the biggest drawbacks is the lack of data collection inside and immediately around a developing storm.
University of South Carolina marine biologist Madelyn Fletcher says the density of U.S. coastal weather observations is relatively low compared to those over land. Only 140 sites collect data along the coasts compared to 14,000 on land, and she says that is not enough to make forecasts as reliable as possible.
"A lot of our weather comes from the ocean," she said. "That emphasizes the great need to have more measurements, more observations, and the serious need for a higher density of observations sites in coastal oceans. I see that in the United States, but I'm sure it is true globally."
Ms. Fletcher is optimistic that this will improve. She cites the growing network of partnerships among U.S. government forecasters and university researchers to improve coastal weather observations. In addition, the national oceans and atmosphere agency is working with the military and several other government agencies to achieve an integrated ocean observing system, a plan that Washington is seeking to expand with other nations into a global project.