Depth Reporting story on hurricane forecasting

Written in early September 2005, this is me writing about a topic I'd like to think I know a lot about. And, to be fair with myself, I know my share. If only I had the math skills necessary to botch seven forecasts a week, I'd have been a on-air met.

This was written for one of my favorite college professors of all time, Chris Waddle, who taught depth journalism at UA. He's a Pulitzer winner, two times over, and hey, he gave me an A on it! How 'bout that? Praise from Caesar...

A Fine Line?

As a record hurricane season rages on, forecasters are trying to convince coastal residents that a hurricane is more than meets the eye.

By Matt Hooper

It was to be a city under water.

Downtown streets turned into canals, high-rise buildings vented by exploding windows, major thoroughfares and access arteries cut off from rescue aid. A city of nearly a half million displaced.

Tampa, Fla., on Friday, Aug. 13, 2004.

***

It had been a full day since the National Hurricane Center’s forecast map placed the residents of Tampa and its sibling city of St. Petersburg in the crosshairs of a violent storm.

Every six hours in the life of a tropical storm or hurricane, the NHC releases an updated forecast map to news outlets across the country. The map details the storm’s current position, denoting it with a brown bulls-eye, and quantifies the system’s proximity to land. From the bulls-eye, there extends a skinny black line that directs the most violent part of the storm, the eyewall, to its forecasted destination. That skinny black line splits a cone of probability, colored in white, which widens as the forecast extends further into the future; serving as a visual representation of the unpredictability of these prodigious storms and marking the extent of the area within which the skinny black line could wobble left or right.

Just before sunrise on Thursday, Aug. 12, Tampa Bay was bisected on the newest forecast map by a skinny black line.

Charley, the third child of the 2004 hurricane season, was born on the afternoon of Monday, Aug. 9, just north of the Venezuelan coast. The forecast called for the storm to move northwest toward an area off the Cuban coast. It would muscle up past tropical storm status to become a hurricane, as happens to all tropical systems that achieve constant wind speeds at or greater than 74 mph.

Monday morning’s skinny black line missed the western coast of Cuba by roughly 90 miles.

Monday afternoon’s skinny black line was shifted nearly 150 miles east.

Now central Cuba was to expect the brunt of Charley and much sooner than expected as almost a full day had been shaved off expected arrival times. Six hours later the skinny black line inched further past Cuba and into the eastern Gulf of Mexico.

Just before sunrise on Wednesday, Aug. 11, Charlotte Harbor, an inlet more than 90 miles south of Tampa’s famous crescent, was bisected on the newest forecast map by a skinny black line.

One day later, the track shifted north to Tampa.

Landfall was expected on Friday afternoon. Charley could possibly be a Category 2 hurricane, as measured by intensity scale for these storms, known as the Saffir-Simpson.

A plan for the evacuation of the Tampa region commenced immediately. According to an Aug. 12, 2004 report by CBS News, it was estimated that 380,000 residents of the coastal city were told to pack up and head north. Cars and trucks inched slowly north on Interstate 75 throughout the day on Thursday, while Charley’s swirling winds and drenching downpours walloped the island nation of Cuba and skirted past the Florida Keys.

Lisa Kaminski, the manager of a Days Inn Hotel in Key West told CBS News that while guests were told to high-tail it to Hialeah or motor up to Miami to escape the pounding surf and howling winds, her staff was going nowhere.

“We’re staying,” she said. “This isn’t a big one.”

At 11 a.m. on Friday, Aug. 13, Category 2 Charley and his 110 mph winds were knocking on Tampa’s door, its speed increasing, its eye tightening, its winds strengthening. Tampa was doomed for a direct impact. Port Charlotte, once the forecasts’ focal point, was bumped down to the outermost fringe of the cone of probability.

Around noon, the storm’s energy exploded in a fit of strengthening. What was once only a formidable system became apocalyptic. Its winds held steady at 145 mph with higher gusts, strong enough to rip apart well-built homes, move cars without turning a key, shift a beach hundreds of yards inland or just erase it altogether. Now Category 4 royalty on a scale that halts at five, Tampa’s scenario got bleaker, despite the fact that what first appeared to be a wobble to the right of the hurricane’s set path, was looking more like an out-and-out detour.

The skinny black line stayed put.

John Emmert was the news director at WINK-TV in Fort Myers on that Friday the 13. He tells “The Rundown,” a Web site devoted to the coverage of local television news, that his meteorologists predicted a more southern landfall.

“The Hurricane Center’s forecast has a center line which is the consensus line, which was Tampa,” Emmert explained. “Our guys didn’t think [the storm’s track shift] was a wobble, we went on the air right after that and told people it looked like the hurricane was headed closer to Fort Myers.”

At 2 p.m., the NHC’s skinny black line finally shifted southward, bisecting Charlotte Harbor, staring down Punta Gorda and Fort Myers.

At 3:45 p.m., Charley’s devastating winds, flooding rains and pounding surf permanently altered the landscape of Port Charlotte and Punta Gorda and the lives of its residents.

Ninety miles away from Tampa Bay were 31 dead bodies and more than 11,000 uninhabitable homes and pieces of homes. Much of downtown Punta Gorda was reduced to rubble. Citizens complained that they were unprepared; the NHC blamed the media.

“We’re kind of surprised that people were surprised,” said Robbie Berg, an NHC meteorologist speaking to the Associated Press.

Berg said that the problem lies not with the Center’s forecasted path, but with the media’s fixation on “Tampa, Tampa, Tampa.”

“We always wish we could have more, better guidance,” Berg said. “But with what we had, we did the best we could. Errorwise, we really weren't that bad. It's just that the storm happened to be so intense, that it made a big difference in landfall.”

***

More than 30 years have passed since hurricanes were first tagged with a number on the original Saffir-Simpson scale and first led to shore by a skinny black line. In the late 1970s, hurricane forecasts extended only one day into the future. In the early 1980s and 1990s, the prognostications pushed outward to three days. Now, the Hurricane Center issues a five-day outlook that includes the skinny black line, the cone of probability, a likely cone of probability and the storm’s Saffir-Simpson classification.

Forecasters have become more knowledgeable of the storms they track, computer models are more exact and confidence in forecast accuracy is increasing. Chances are that no tropical meteorologist would argue that the forecasts have not improved markedly over the past 30 years.

But Punta Gorda’s people still claim they were unprepared for Charley. The National Hurricane Center still claims that not only should Punta Gorda have been prepared, despite being 90 miles away from the pinpoint forecast, but that the average margin of error of these storms at 48 hours from predicted landfall is 110 miles.

How does that skinny black line communicate danger to that 110 mile radius? Why not focus only on the cone of probability? Is a five-day outlook in the best interest of the public? Will the confusion that took place on Florida’s western coast happen again if a similar storm threatens another major metropolitan area?

***

It is a city under water.

Downtown streets are canals, high-rise buildings are vented by exploded windows, major thoroughfares and access arteries are cut off from rescue aid. A city of a half million is displaced.

New Orleans, August 29, 2005.

***

New Orleans, a city long-feared to one day wind up in the crosshairs of a deadly cyclone, awoke on Aug. 26, 2005 and the skinny black line was staring at the mouth of Mobile Bay, over 100 miles east of the city.

By lunch time, the line and The Big Easy got cozier and the weather service got nervous. At supper time, Crescent City residents could theoretically have stepped outside their homes, glanced over the skyline, and seen a skinny black line silhouetted in the eastern sky against a rising crescent moon.

The line punctured the southern Louisiana barrier island of Grand Isle before snaking dangerously close to the New Orleans’ ports, to its rebellious French Quarter, to its disadvantaged Ninth Ward. Parked a handful of miles east of the city, the black line was the harbinger for Katrina, the largest and most powerful hurricane to threaten the Gulf Coast since 1969.

Born out of a jumbled mass of thunderstorms and tropical showers, Katrina quickly morphed into a Category 2 spitfire, slicing through South Florida and emerging into the simmering Gulf waters. There it flourished, leaping from two to five on the Saffir-Simpson scale and chugging swiftly north toward the coastline shared by Louisiana and Mississippi.

Evacuations were ordered as hurricane warnings stretched the length of the Magnolia State shoreline and into the Louisiana bayou. A day before the forecasted landfall, the city of New Orleans was forced to flee. Most did, several did not.

News networks across the country sprang into action, with the onus of activity centering on The Big Easy. What was going to happen when a city surrounded by water and resting almost completely under sea level became engulfed in storm surge and tropical downpours?

At 6:10 a.m. on Monday Aug. 29, a slightly-weakened Category 4 Katrina pierced the early morning calm at Grand Isle before barreling east into the Mississippi towns of Waveland and Gulfport. The most despicable winds, the heaviest rains and a 30-foot wall of storm surge ravished the Delta State coast.

New Orleans survived the winds; few structures were flattened by Katrina’s breezes. It succumbed to the rain.

The forecast was not exactly perfect, as the skinny black line was about 15 or 20 miles west of the storm’s actual path, but it was pretty darn close according to a report filed by staff reporters at MSNBC, who declared that “two federal agencies got it right: the National Weather Service and the National Hurricane Center.

“They forecast the path of the storm and the potential for devastation with remarkable accuracy.”

But Jack Crochet of Biloxi, who rode out the storm in his now splintered home, told staff reporters at The (Jackson, Miss.) Clarion-Ledger that “we thought everything was going to New Orleans.”

President George W. Bush, during a question and answer session in Biloxi, himself acknowledged that the coast of Mississippi seemed to pale in comparison to The Crescent City, saying “Mississippi people have got to understand that I know a lot of the focus is on New Orleans.”

This time it was a near perfect forecast, but still there are complaints of citizens who suffered ungodly devastation and felt that they were unprepared, despite hurricane warnings and attempted evacuations. Why? Was everyone too fixated on that skinny black line?

***

Neil Johnson of The Tampa Tribune published an article, “Hurricane Forecast Maps May Lose Direct-Hit Line,” on Oct. 22, 2004, a little over a month following Charley’s unexpected detour into Charlotte Harbor. In it, Johnson addresses this argument by citing Max Mayfield, the director of the National Hurricane Center.

“Don’t focus on that skinny line,” said Mayfield, which also happens to be what he told Congress on Sept. 22, 2005 as he addressed the NHC’s role in forecasting Katrina. Mayfield told Johnson that he wants the line to be removed, but the consensus of the meteorological community was that the skinny black line did more good than harm.

After the 2004 hurricane season, the NHC conducted an email poll to determine whether or not forecasts had reached the end of the skinny black line. According to the AP’s Bill Kaczor, of the 971 email responses from media representatives, meteorologists and local emergency management agencies, 63 percent are in favor of keeping the line in the forecast packages. So the NHC succumbed to the will of the majority.

Chris Landsea is one of the Center’s most seasoned forecasters and was a part of the group of meteorologists that paced up and down the office floor in Miami as Charley hiccupped into Punta Gorda and Katrina careened into the easternmost coast of Louisiana. He described the inaccuracies accompanying that skinny black line.

“A two-day forecast has an average margin of error of 125 to 150 miles or so,” Landsea said. “So it’s something that, if you give an exact point, can be misleading.”

One hundred fifty-six miles separate Birmingham, Ala., from Atlanta, 142 miles separate Cleveland from Columbus, Ohio and 121 miles separate Los Angeles from San Diego.

Massachusetts is 152 miles wide.

Hurricanes themselves can be hundreds of miles in diameter, with its peak winds packed into the eyewall and gale force winds extending up to 100 miles from the center of circulation, but the intensity of the storm is not uniform.

Storms in the Northern Hemisphere have a natural tendency to spin counterclockwise, which takes the circulation from left to right and back again and secures that right side of the storm (if the hurricane is moving north) will always be stronger than the left. Katrina, for example, made landfall with winds of at least 145 mph, moving north at 15 mph. On the left side of the storm, you would subtract the motion of the storm from the wind total. In New Orleans, the winds spun at 130 mph, which is a Category 3 measurement.

On the right side of the storm, you would add the storm’s motion to the wind speed. On the Mississippi Gulf Coast, 160 mph winds sandblasted homes, ripped pavement off of roads, displaced bridges and moved boats miles into dry dock. Those are Category 5 winds, or what Saffir and Simpson referred to as “catastrophic.”

If a 20 mile jog in the forecasted track of a storm that is hundreds of miles wide can mean the difference between moderate damage and catastrophic damage, imagine what 150 miles could do.

One thing that a 150 mile forecast mistake does is give people a false sense of security. Brian Peters, the former chief of Birmingham’s National Weather Service office and now an on-camera forecaster at WBMA ABC 33/40 in The Magic City, says that one bad forecast can contaminate coastal residents for generations.

“Human nature says: ‘Well, they missed it before, and they’ll miss it again,’” Peters said. “Human nature also says: ‘I survived the last one, I’ll survive this one.’”

Peters’ obsession with all things tropical began, ironically, with Hurricane Donna, which slammed into Charlotte Harbor on Sept. 10, 1960, blazing the trail that Charley would follow 44 years later. Since then, he has not only studied some of history’s worst storms, he’s been hunkered down in the midst of them. If anyone knows why there can be 150 miles of forecast error for these systems, he does.

“Well, forecasting is not completely an exact science,” Peters said with a chuckle. “We’d like to think it is, but I don’t think it will be in my lifetime.

“There are many factors. For instance, the eye in a hurricane or tropical storm is constantly changing, constantly changing its character. They don’t move in a straight line, they wobble.”

A wobble, a burp in the not-so-straight path of a hurricane might be normal, leaving the forecasted track unchanged. Or it might signify a shift in the track, as it did for Charley. But even if Charley’s eye had stayed on its original track, the damage was still likely to peak south of the point of landfall. Again, the onus falls on the skinny black line.

“It’s important if you’re a Gulf Coast resident to understand that there is more to a hurricane than the eye,” Peters said. “Historically, that’s what we’ve looked at, statistically it’s an identifiable thing that meteorologists use in trying to evaluate a forecast.”

Dr. Jennings Bryant, a professor and media researcher at the University of Alabama in Tuscaloosa, has studied the role of the media in covering disasters. He believes that the media’s focus on the skinny black line can also give some people that false sense of security.

“People tend to think in terms of central tendencies, always using means and averages,” Bryant said. “People feel like: ‘That line is 20 miles to the east of me, well, I’m safe.’ In terms of public safety, that’s a problem.”

A hurricane is more than the eye, that’s a fact you don’t have to sweat out of forecasters. But the line tracks the eye and only the eye. So how do you sway people away from what Bryant calls “central tendencies?”

The NHC is trying to tackle that problem.

For 2005, a new, experimental map accompanies each tropical system’s forecast package, showing the probability of hurricane or tropical storm force winds striking a particular area with a certain time frame. In addition, the NHC continues to issue hurricane watches and warnings for areas that find themselves within the cone of probability. All of the areas along Florida’s western coast that weathered Charley’s fury were under warnings, the same for Mississippi’s coast during Katrina.

The Weather Channel, America’s 24-hour weather television resource, has eliminated the skinny line completely from their forecasts, relying solely on a “cone of concern.” The same applies at Accuweather.com.

Media, however, still have a tendency to focus on that one point, that skinny black line. Remember forecaster Robbie Berg and his complaint of “Tampa, Tampa, Tampa?” Remember Dr. Bryant’s “central tendencies?”

It appears that the skinny black line is also becoming a money-making obsession. According to “The Rundown’s” story which first told of meteorologists in Fort Myers disagreeing with the NHC’s forecast track, the local television stations in Tampa and St. Petersburg were in a “hotly contested marketing battle” to predict Charley’s correct path, their own skinny black line.

Back in Birmingham, Brian Peters, whose ABC station has been involved in a meteorological arms race with the other major networks in the city since its inception, acknowledges that the media’s role in covering these storms is an important cog in a system designed to save lives.

“If the NHC forecast is good…and I don’t know of anyone who can truly fault their forecasts over the past couple of years…the warnings are for the right areas, the media does their job of conveying that and not focusing on the line, and people respond by being informed, I think the system works fine.”

Peters adds that while he is not in favor of completely eliminating the skinny black line from the forecast package, he is in favor of directing the public’s eyes away from it.

“If there was something that could be used to get people away from the line, I certainly would be in favor of it,” Peters said. “That is, as long as we could continue to convey to people the dangers that they are likely to experience.

“We need to do more education to deemphasize the line.”

The NHC’s Chris Landsea agrees, and emphasizes that the most important forecast tools his office has are hurricane watches and warnings.

“You know, hurricane warnings are put out there when the storm is over a day away to get the point across that there is a region that is going to be impacted,” Landsea said. “With regard to Katrina, all of Mississippi’s coast was warned that: ‘Hey, there’s a hurricane coming and you’d better get ready now.’

“It’s unfortunate that people came away with a wrong notion.”