What drives a storm's force?Scientists say that when it comes to predicting hurricane intensity, their models are `running blind.' But they think they have the tools to change that.By Maya BellSentinel Staff Writer August 1, 2005 MIAMI -- This summer, scientists are peering into hurricanes in ways they never have before, all in hopes of solving one of nature's greatest mysteries: Why do some tropical cyclones undergo rapid growth spurts and others suddenly fizzle? Last week, researchers flew into a disturbance off the Yucatán Peninsula, for the first time measuring the winds, rain and temperature as the system sprouted into a tropical depression. Come mid-August, they will shift their focus to mature hurricanes, observing the interaction between a hurricane's spiraling rain bands and dangerous inner core. If the right opportunity knocks, one day soon an unmanned aircraft weighing just 33 pounds will try to corkscrew into the eye wall of a hurricane and dive close to the sea. If it survives, the Aerosonde will go where pilots dare not fly. All three experiments are small players in an arduous effort aimed at conquering the next frontier of hurricane forecasting: accurately predicting a hurricane's intensity. Compared with their tracking forecasts, meteorologists have very little skill in predicting a hurricane's intensity, which provides vital information for emergency managers who order evacuations based on the size and strength of a looming storm. But they're counting on faster computers, better data from a hurricane's inner core and the "unsung heroes" of the nation's hurricane-warning system, who are busy cranking out the next-generation hurricane-forecast model, to crack the intensity code. "It's our No. 1 priority," said Max Mayfield, director of the National Hurricane Center west of Miami. "If people go to bed preparing for a Category 1 hurricane and wake up for a Category 4 hurricane, or even a Category 5, the consequences could be catastrophic." That scenario is not just the stuff of nightmares. In just 36 hours, the Labor Day storm of 1935 exploded from a pesky tropical storm in the Bahamas to the most intense hurricane to strike the United States. The top-of-the-chart Category 5 killer took more than 400 lives in the Florida Keys. Even with the improved warnings and forecasts of today, a similar disaster is not far-fetched. During a relative three-decade drought of hurricanes, the Sunshine State filled its cities with skyscrapers and its coastlines with condos, both with millions of people. Now the Atlantic Basin has entered a renewed cycle of hurricane activity which, characterized by more major storms of Category 3 or higher, could last 20 years. For proof, doubters need look no further than last month, which not only spawned five named storms -- the most for any July on record -- but two of the most intense hurricanes recorded this time of year. Storms' strength in doubt In theory, scientists understand what energizes hurricanes -- heat from the ocean, moisture in the air -- but have little idea how atmospheric and oceanic processes conspire to make them stronger or weaker, bigger or smaller. For example, they can't explain why Hurricane Andrew, a tiny storm that leveled much of south Miami-Dade in 1992, and Hurricane Floyd, a behemoth that threatened Florida's east coast in 1999, were so different in size yet so similar in intensity. But thanks to improved forecast models, they accurately predicted that Floyd would veer north at the last minute, sparing Florida. Those improvements helped forecasters cut their tracking errors in half during the past 15 years and break records each of the past three. Last year, for example, at 24 hours before landfall, projected storm paths were off by an average of just 67 miles. But, at the same time, improvements in the center's intensity predictions are negligible. Last year, the average error for maximum winds at 24 hours before landfall was almost 12 mph -- about the same for the previous decade, according to forecaster James Franklin. "In terms of intensity, we're very primitive," he said. The reason is as simple as the solution is complicated: Hurricanes are small compared with the large-scale weather features -- the systems of high and low pressures -- that push them around. So, as satellite observations, computers and numerical models used to simulate a storm's motion improved, the steering currents became easier to predict. But intensity changes depend not only on the huge steering currents surrounding a hurricane, but also on the much smaller structure of a storm's inner core. For now, forecasters don't have a good idea how the two interact. The reason? They have never had a detailed description of the core to plug into their numerical models. While the National Oceanic and Atmospheric Administration's reconnaissance planes provide valuable details about humidity, temperature and winds inside the core, the turboprops fly at low levels. As a result, their observations don't capture the whole landscape. "To a great extent, our intensity models are running blind," Franklin said. "We have to make assumptions about what the structure inside the core is." But that is about to change. State-of-the-art radar Come 2007, NOAA's hurricane-surveillance jet will be outfitted with Doppler tail radar, enabling a high-speed, high-altitude Gulfstream IV to give scientists unprecedented data about a storm's circulation -- from the top of the clouds to the sea surface. "As with any system in nature, if you want to be able to predict what a hurricane is going to do tomorrow or the day after, you have to have an accurate description of what it's doing now," said Naomi Surgi, a NOAA scientist who has been pushing the upgrade for more than a decade. "That's what the G-IV Doppler is going to give us." Today, though, such information would be all but useless. That's because neither the current models nor the supercomputer that NOAA's Environmental Modeling Center in Camp Springs, Md., uses to simulate and project storm motion can make sense of such minute details. The models are too coarse, and the computer is not sophisticated enough even though it performs 1.3 trillion calculations a second. But that, too, is about to change, thanks to a planned five-fold increase in the computer's capacity, and to modelers like Surgi, whom Mayfield calls "the unsung heroes of the nation's warning system." Working in relative obscurity, they have spent four years developing the next-generation hurricane forecast model. If all goes as planned, the Hurricane Weather Research and Forecast model will debut in 2007, eventually assigning the correct weight to all the variables that affect intensity and accurately predicting changes in a storm's strength and structure. "Which part of the storm is the most important for intensity? We don't know," said Surgi, the HWRF project leader. "But we're going to find out." Groundbreaking experiments To aid the effort, field researchers are doing their part, embarking on unprecedented experiments this summer. This month, meteorologists at NOAA's hurricane-research division flew into Hurricane Dennis and Tropical Storm Gert from their infancy to decay -- something that had never been done. Usually, research flights are reserved for intense storms that threaten land. Robert Rogers, the principal investigator, is fairly sure the pilots were bored by the exercise. They're used to flying into hurricanes. But he found it exhilarating to fly through the system that would become Gert when it was declared a tropical depression. He knows Surgi will use that information to give HWRF the initial descriptions it must have to solve the intensity puzzle. "I was whooping in the cockpit," he said. For six weeks starting in mid-August, Rogers will join researchers at the Universities of Miami and Washington to investigate the influence of rain bands on intensity, for the first time flying three Doppler-equipped airplanes into a storm to take simultaneous measurements. "We hope to contribute a piece to the intensity puzzle," said Shuyi Chen, the UM meteorologist who dreamed up RAINEX, a water repellent used on glass. So does Joe Cione, another researcher at NOAA's hurricane division. He thinks the 5-foot-long Aerosonde, which launches from the roof of a sport utility vehicle, can survive its journey and collect data where pilots dare not fly: the layer of the hurricane core where air meets sea. He admits, however, "there are no shortage of skeptics." Many skeptics also doubt scientists can unlock the complicated intensity code, but after 33 years at the hurricane center, Mayfield is confident it will happen in his lifetime. "We improved track forecasting as a result of improvements in numerical modeling, and we'll improve intensity forecasting the same way," Mayfield said. "You just can't expect miracles from the models without good data." Maya Bell can be reached at mbell@orlandosentinel.com or 305-810-5003. Copyright © 2005, Orlando Sentinel |
