News » Features

Gentlemen, Start Your Barometers

BW's inaugural olympics of forecasting


With the advances in satellite imaging and computer graphics (especially computer graphics) that have characterized the last decade in television meteorology, it seems indisputable that ours is a golden age of weather forecasting. Unfortunately, nobody is keeping tabs. "Kids [in other towns] used to do science fair projects about tracking accuracy all the time--and of course the winner would do his forecast from the science fair," recalls Vin Crosby, meteorologist at both Boise's KBCI Channel 2 and KTRV Fox 12. "For some reason nobody really seems to do that here in Boise." KBCI's Scott Dorval agrees with Crosby's diagnosis, but here's the kicker: they're sad about it. Weathermen want to be tested, paired up and scrutinized. They're just sick that way.

The reason for this lack of juvenile accountability is doubtless that a project requiring such independent research and motivation could not easily be matched up to soulless standardized tests--but I digress. All on-screen "NWS" (National Weather Service) and "AMS" (American Meteorological Society) seals aside, Idaho's weathermen, women and androgynous "meteorologists" (whose numbers seem to have suspiciously boomed over the last few years--am I right, fellow news junkies?) don't officially have to answer to any higher authority for their weather prophecies. Dorval doesn't send his predictions away to a secret laboratory in the clouds at 10:30 every night. KTVB weekender Bob Anthony doesn't receive a satellite-antenna-lash from an NWS thug for every degree that he is amiss in an overnight forecast--lucky for him. In fact, even with all the official-looking, hey-Ma-you-can-actually-see-the-snow-falling 3D computer gimmicks that populate local weathercasts, the only entity to which forecasters remain even remotely accountable is their viewers.

With this in mind, Boise Weekly's crack Meteorological Cynicism Department recently formulated a science project of its very own: tracking the forecasting accuracy of three local stations (as Crosby has the unprecedented task of covering two stations, Fox 12 is lumped in with CBS) over 10 fairly typical days of Idaho spring weather from April 17 to April 26. Rather than worry about decorations like "precipitation" or "dewpoints," we used only temperature as our measuring stick, divining rod and if necessary, charlatan-exposing sword (it wasn't necessary).

This test comes in the nick of time, because a change could be in the wind for weather forecasting. In late April the remote Russian province of Irkustk, long known to Americans only as a crucial junction on the Risk board, received global attention when Emergency Situations Minister Sergei Shoigu promised to "squeeze compensation" and even jail time out of inaccurate forecasters whose mistakes affect his flood-ravaged constituents. Likewise in the U.S. both the government agency NWS and The Weather Channel have been sued in recent years for accidents that allegedly resulted from inaccurate weather prediction. Although both Crosby and Dorval maintain that Idaho, and the Treasure Valley in particular, contains enough finicky microclimates that forecasting accurately is a perpetually daunting task, don't expect climate-centered incarcerations to make local headlines any time soon. Concerns over accuracy manifest very differently for these perfectionist forecasters by way of a very unique sentiment--weather guilt.

"I used to feel weather guilt all the time," Crosby admits. "I would beat myself up for getting anything wrong, but I just had to realize: You don't have to be perfect to be liked, and you can't control the weather. If you're off a degree, it's OK."

Dorval's version of the emotion is a little more internal. "More than anything I get angry with myself," he explains, "I've been trying to predict the weather since I was 3-years-old, so if I'm off, I'm the one who has to live with it--but of course everybody always thinks we're more off than we are." We shall see.

Weathering the Weather Test

Two times a day, a weather balloon released from the Boise Airport travels 20 miles into the atmosphere and explodes. Once an hour, a single thermometer at the same location gives the NWS a temperature reading. These archaic methods may not be as well televised as the infrared satellites that pick up proto-El Niños and Niñas months before they moisten the shores of Oregon and Washington, but they remain benchmarks of Idaho weather forecasting. As such, the daily highs and lows from the airport thermometer are the standards to which local one-, five- and even seven-day forecasts, as well as The Weather Channel's overnight Boise forecasts, were held. The accuracy of each forecaster was gauged by two different standards: First, by a straightforward figuring of their average daily error in degrees, and second by the percentage of forecasts in which each station was just plain wrong in an evening forecast. "Wrong" is considered in terms of decades--10 cumulative degrees, or the difference between "The '70s" and "The '80s." The numbers were snatched from evening broadcasts, so if daytimers Larry Gebert and Jim Duthie would have fared better, sorry guys. It's your lot as people who get up at 6 a.m. to be occasionally overlooked.

Channel 2/12

Vin Crosby's CBS tagline is "Idaho's Most Accurate Meteorologist," but this Weather Channel alum prefers to think of himself as the most adventurous. "In meteorology we have computer forecasts that most guys will just stick to," Crosby explains, "but I'm very, very aggressive, and I'll go away from those numbers if the long range [weather] pattern suggests it." As such, Crosby, who is joined by newcomer Dan Barth at KBCI, had both big hits and big misses during the test period. On Thursday, April 22, for instance, Crosby precisely divined a temperature dip to 39 degrees six days in the future, while neither of his rivals were within 10 degrees of the truth. The day before, he also penned the only totally accurate forecast of any of the test subjects, a stellar 82/48 high/low bull's-eye forecasted six days out on Wednesday April 21. In the same newscast, however, Crosby also had the biggest miss of any subject: a 32-degree seven-day overshot. In all, the team fared well in both the six- and seven-day forecasts, barely losing to Channel 6 in the total five- to seven-day average. Indeed, Crosby's average was better for five- to seven-day forecasts than for three or four, with only 18 percent of his forecasts straying over 10 degrees in error for both categories.

Channel 6

Thirteen-year Treasure Valley weather veteran Dorval and newcomer Steve Frazier combined forces for KIVI Channel 6 during the 10-day test, with Frazier taking weekends and Dorval weekdays. The team started strong, with Dorval nailing two one-day lows and a high like cans on a fencepost. Channel 6's one-day average of 2.4 degrees of error per high and low was the best among locals, only falling short of champion The Weather Channel by one-tenth a degree.

The team also ran away with the five- to seven-day forecast title, barely surviving a mammoth 32-degree seven-day miss on Wednesday, April 21 (prediction 82/48, actual high/lows 60/39--yikes!). Dorval had more right-on predictions than any other station's forecaster over the test period, showing particular prowess at pegging the long-term weather patterns. He even goes so far as to offer a brash 10-day forecast on his weather reports, which is especially notable considering that when the New England native started in Boise in 1991, "I only did a four-day planner. I figured, it just wasn't worth going out past four days, because the accuracy just wasn't there."

Channel 7

KTVB Channel 7 has the largest local staff of on-air meteorologists with four, as well as claiming to have Idaho's "Chief" meteorologist Rick Lantz--a term which, as I understand it, means that Lantz is responsible for leading the weathermen in any future wars with sportscasters. While the team was in the middle of the pack in short forecasts, Lantz and Co. thrived in the crucial but often overlooked middle range of three to five days. Lantz, Deborah Smith and Bob Anthony all had at least one low temperature exactly right in that spread with Sunday stud Anthony nailing both a three-day and a five-day. That's some sharp shootin', and it earned Channel 7 the middleweight crown.

In five- to seven-day predictions, however, Channel 7 got a little roughed up. Although Lantz is Idaho's longest-running weathercaster, the spike of 13.8 average degrees of error in seven-day broadcasts looked to be his Achilles heel--especially when compared with Crosby and Barth's 10 and Dorval and Frazier's 11.8. In one fateful two-day set, Channel 7 was a record 46 degrees over the actual temperatures in seven-day forecasts, and the crew didn't get within five degrees in a single seven-dayer. Maybe the station's number isn't a good luck charm after all.


If the accuracy achieved by the weather teams seems unexpectedly consistent across the board--only Channel 7 had a worse five- to seven-day total than their three- to four-day--the cause is undoubtedly a rather dry, uneventful 10-day stretch of springtime weather. No weathercasters struggled in the weeklong categories nearly as much as I hoped, er ... feared, and they matched one another far more than they differed. With few exceptions, Dorval's best days were also Lantz's, and Crosby's worst were also found in the company of his peers. Indeed, the occasions where any weathercaster differed from his or her rivals by more than five degrees on a forecast of any length could be counted on one's rain-pelted hands.

All competition aside, accuracy doesn't buy viewers and nobody knows it better than the weathercasters themselves. Their wording, poise and especially graphics are what win awards and on-screen endorsements from meteorological organizations. Crosby sums up the situation simply, "Graphics and numbers really just help you to tell a great story. You can have the looks, but the proof is in the pudding. If you can't tie it all together, the graphics, accuracy and a forecast into an actual 'weather story,' the public will definitely understand--'click!'"