After the loss of life, and confusion, following Hurricane Charley, an interesting op-ed piece was written by Bryan Norcross, Chief Meteorologist from WFOR in Miami. You can read it here now, or click the ‘continue’ link at the end of this posting.
Norcross makes some interesting points, many of which I agree with.
Though we make our own forecasts at the TV station, we respect the Weather Service’s watches and warnings (though there are times I mention them, followed by what I think will actually happen).
The bigger problem occurs when watches and warnings are contradictory. Uncoordinated watches, warnings and statements for hurricanes, severe storms… even winter weather, is a continuing weakness of The Weather Service. All hurricane watches, warnings and statements should come from one place – period.
This certainly led to the disservice done to the people for Florida.
When local offices speak, they address problems from their own perspective, which is not necessarily the public’s. And, the public and media are probably concentrating their attention on the Storm Prediction Center (Whose idea was it to change this from the much more meaningful Hurricane Center?), which is where most people would expect to find hurricane info.
I work in Connecticut, a small state served by three NWS offices. Their statements often mislead the public because each only refers to the region for which they forecast.
Here’s an example. If Boston says a watch has been canceled for Connecticut, they mean their counties. No one in Connecticut could read a statement like that and understand that half the state is still under a watch.
During the winter, Litchfield County, our ‘snowbelt,’ might be under a lesser category of alert because the Albany office uses somewhat different criteria than the New York or Boston offices. When I post a map which shows a Winter Weather Advisory for Litchfield while there’s a Winter Storm Warning for our other counties (even though Litchfield has the more wintry forecast) it does nothing but confuse.
I have been to NWS ‘customer’ conferences in Washington, and have tried to sensitize them to this confusion. As you see – no change.
Forecast on target, but the media mangled message
BY BRYAN NORCROSS
Hurricane Charley was expected to hit Tampa, but instead came ashore in Punta Gorda and Port Charlotte. Many Floridians, if not most, have the perception that the forecast was wrong and that the storm bamboozled the experts.
But that’s not true, and the misperception harms the hurricane-warning program.
Admittedly, the National Weather Service does itself no favors by choosing to communicate the basic issues of ”who might get hit and how badly” in a fragmented format including graphics that present a mixed message. But, bad format or not, reporters and television meteorologists should know how to piece the information together to present an accurate picture of the forecast. That’s their job.
The National Hurricane Center puts out a full ”advisory package” four times a day, made up of four separate bulletins. Besides the watches, warnings and current storm attributes, such as center location, top wind speed and motion, the most important pieces of information are the forecast locations for the center of the eye for the next five days, the forecast wind speeds and the often overlooked ”strike probabilities.” There is also an intrinsic set of numbers that are part of every forecast advisory: the 10-year average forecast errors for each future time period.
All of this information must be taken together to get a full understanding of the hurricane forecast. The average-error numbers are combined with the forecast coordinates to create a ”forecast cone.” About 70 percent of the time, based on past forecasts, the center of the storm stays inside that cone. In the case of Charley, Fort Myers, Tampa and Orlando were all in the cone for days before the storm made landfall. No one in those cities should have been unaware that they were at risk.
The Hurricane Center further fine-tunes the chances of specific locations being affected by issuing ”strike probabilities.” On the afternoon before Charley hit, Fort Myers’ strike probability was 42 percent and Tampa’s was 39 percent. On the morning of landfall, Fort Myers was at 50 percent and Tampa was at 47 percent.
So that Friday morning when the hurricane was still south of Key West, Fort Myers was within the forecast cone, had a strike probability of 50 percent and was under a hurricane warning. Calling a hit in that area a mistaken forecast or a surprise is like flipping a coin and being surprised it comes up tails. And reporting the final landfall point as ”unexpected” is a misrepresentation of the facts.
Concerning the strength of the storm, the Hurricane Center predicted that the storm would be a 120-mph, Category 3 hurricane. It ended up going through what’s known as ”rapid intensification,” a phenomenon that was acknowledged as possible but is poorly understood. The bottom line, however, is that preparations are the same for the Category 3 storm that was forecast or the Category 4 storm that hit.
As noted, the National Weather Service is partially responsible for the communications boondoggle. It spread critical information out over multiple bulletins issued at different times, in different formats, from different offices. And within the ”cone” graphic on its website, there is an eye-catching line down the middle, partially negating the important information conveyed by the cone and misguiding those uninitiated in the nuances of the forecast process. Strike probabilities are on another similar-looking graphic, further confusing the issue.
But the National Weather Service is not, in general, the final link in the communications chain. Most people still get their emergency information through television, radio and newspapers. This is where the majority of the misinformation and misimpressions come from.
All Miami television stations now draw a forecast cone. But most stations still highlight the forecast points on the ”center line” and ignore the strike probabilities creating an erroneous impression. This, in spite of TV meteorologists having recently been presented research demonstrating how these practices dramatically change a viewer’s perception and leave most people with an erroneous impression. This problem is widespread on television stations in the hurricane zone.
In The Herald, statements such as ”expected landfall in Tampa” in a post-hurricane editorial and ”track” maps like the one in the paper the Sunday after the storm, with a misleading accompanying story, compound the confusion and misunderstanding of the hurricane-warning process.
An examination of the list of fatalities from Hurricane Charley shows that almost everyone, except for heart-attack victims, who died during the storm was doing something that they were clearly told not to do: staying in mobile homes, driving or walking outside.
Did they somehow think that it would be OK to do these things because the message was not clear? We’ll never know.
The misrepresentations of the Charley forecast in the media point to a serious national problem that reaches beyond hurricanes. The government puts out important emergency information, but does so in a less-than-ideal format; and the media, whose job it is to make that information clear and useful, modifies and mangles the message so that there are widespread misperceptions of the risks involved. We should waste no time in doing better.
Bryan Norcross is director of meteorology at WFOR/CBS, Channel 4.