Fill out this form to email this article to a friend
Weather
Hindsight on the weather guys
A few days before Sunday night's cold blast, five TV forecasts were too warm by nearly 20 degrees.
By TOM ZUCCO
Published January 25, 2005
Predicting high and low temperatures in summer is like child's play: Highs in the low 90s, lows in the high 70s.
Predicting winter cold fronts, particularly a few days out, can be trickier.
We gathered weekend forecasts of five major Tampa Bay area TV stations, filed at 6 p.m. Wednesday, to see how Denis, Paul, Steve, Dick and Alan would fare against actual temperatures.
The answer: Not bad on Friday and Saturday. Not so good on Sunday.
Every station predicted highs a few degrees below what they actually were on Friday, and lows a few degrees above what they actually were on Saturday.
But Sunday night, when the strongest cold front of the season barreled in a few hours ahead of schedule, the predicted lows were off by nearly 20 degrees. All five stations predicted the low Sunday at 54 or 55 degrees.
The actual low was 37.
WTSP-Ch. 10 chief meteorologist Dick Fletcher came the closest to the actual high and low temperatures for Friday through Sunday. He was off by 26 degrees, largely because of the front.
Close behind was WTVT-Ch. 13 chief meteorologist Paul Dellegatto (27 degrees), followed by chief meteorologists Alan Winfield at Bay News 9 (30 degrees), Denis Phillips at WFTS-Ch. 28 (31), and Steve Jerve at WFLA-Ch. 8 (32).
What does all this mean?
Forecasters say that while temperatures can vary more in winter than in summer, long-range predictions are merely a guide.
"With the low temperatures, most people think in terms of the morning hours," Winfield said. "Sunday morning, we hit our high for the day and the temperature dropped all day. We wanted people to know the temperatures would be in the 50s in the morning and afternoon, and they were. We were kind of giving the flavor of the day.
"It's no different than the tracking cones with hurricanes. The areas get wider the further out in time you go. There is greater error further out."
Still, people want to know how cold it will get next week. So forecasters go to their computer models and make the best predictions they can.
"We offer the seven-day forecast knowing the last two days are tough to nail," said Ch. 8's Jerve. "What we look for is a trend, and as we get closer, we're going to fine-tune that forecast.
"The hope is that by doing this, we get better with the forecast every year. There is error long-term, but the more we do it, the better we get."
It is arguably a success that TV forecasters missed only by several hours the arrival of the latest cold front.
"The thing that busts more forecasts is timing," said Ch. 10's Fletcher. "We'll generally have a good idea when a cold front will arrive. But sometimes, that far out, it doesn't play out right."
More often than not, forecasters come remarkably close. "Every year, the forecast accuracy should be going up," Winfield said. "Maybe half a percent a year. And we do get a few busts every year. Years ago, we might have a bust every two weeks. That's when you call for a sunny day and it's cloudy, raining and much colder."
Even being right most of the time is still not good enough.
"The downside of good forecasts is that when you do get a bust, people are extremely unforgiving," Winfield said. "They forget we still mess up a few times."
[Last modified January 25, 2005, 06:10:30]
Share your thoughts on this story
|