OKLAHOMA CITY (AP) -- Mike Vescio reported for work as the lead forecaster at the Storm Prediction Center in Norman on the afternoon of May 3, 1999.
And even though the most damaging tornado outbreak in United States history was only hours away from sweeping through central Oklahoma, Vescio said signs of its severity were still sketchy.
"I don't think anyone on the planet could have predicted the magnitude of this event, and if they could I wouldn't have paid any attention to them," Vescio told an audience of forecasters and emergency officials at a national symposium Tuesday.
While hints of a possible severe tornado system began showing upon satellite images by early afternoon, Vescio said the patterns often weakened and dissipated. After one system finally looked conclusive, Vescio issued his first tornado watch at 4:30 p.m. He would issue seven more throughout his shift, as deadly tornadoes swept across the state that night, killing 44 people and churning up entire neighborhoods.
Vescio said even with a vast bank of the latest technology and one of the most weather-savvy staffs in the world, he wasn't certain of the storm's severity until he "saw the tornado on the ground at Chickasha."
Such is the world of weather forecasting -- even in a case as seemingly clear-cut as the May 3 tornadoes, said meteorologists who spoke at the symposium about their roles that day.
In the audience, forecasters from across the country shook their heads in disbelief at the descriptions of the day's circumstances.
Johannes Dahl, a meteorology student in Germany and self-described "severe weather freak," attended the symposium as part of a self-funded tornado-hunting visit to Oklahoma. The 20-year-old said it was encouraging to hear that even Oklahoma forecasters -- the cream of the world's crop -- are human.
"That motivates me to know that you may not know exactly what's going on, but that you can still get sensible results," he said.
A few attendees asked questions. Most sat quiet, no doubt putting themselves in the position of people like Dennis McCarthy of the National Weather Service office in Norman on the day of the tornadoes.
At the Norman NWS office, six meteorologists and three technicians tracked 59 tornadoes using radar, spotter reports and local news broadcasts.
"There were so many people involved, it was just a continuous loop of information," McCarthy said.
The staff at the office issued new tornado warnings about every 15 minutes, McCarthy said, and later won a national Department of Commerce award for the performance that night.
KWTV meteorologist Gary England called the effort by the research offices in Norman "nothing short of brilliant."
"I'm not here as a cheerleader...but these guys and gals did an incredible job," England said.
"The Storm Prediction Center may tell you their vision wasn't that good, but I'll tell you it was excellent."
England said the watches and warnings from the NWS and SPC offices allowed him to have eight Storm Trackers and 10reporter/photographer teams in place from Altus to just south of Ponca City by 2 p.m.
England said his news team fed as much off NWS research throughout the evening as that office gleaned from his station's live images.
"When tornadoes went across I-35, it was on our printer just like that," England said. But bottling chaos is never a flawless exercise. The Tulsa NWS office went into May 3 with its radar down, the victim of a gearbox failure that left it unable to rotate, said Tulsa meteorologist Chuck Hodges.
"When we saw the first warning from the Norman office, me and a coworker looked wide-eyed at each other," Hodges said.
The Tulsa office had to issue its own tornado warning at 9:45 p.m., as the storms pushed into eastern Oklahoma. Hodges said the staff relied on radars out of Oklahoma City, Wichita, Kan., and Fort Smith, Ark. But at 10 p.m., the Wichita radar went down as well.
Hodges said the office had no backup plan to further cover northern parts of the state, though he said the Fort Smith and Oklahoma City radars offered "good overlapping" coverage there. Hodges said the office was also getting acquainted with new software, causing data problems between Tulsa and Oklahoma City.
Despite the inevitable difficulties, Bill Proenza, director of the NWS Southern Region, had noting but praise for the prediction effort that he said saved literally hundreds of lives in Oklahoma on May 3.
To demonstrate, he compared the May 3 tornadoes with the 1947 tornado in Woodward, which killed 116 people in a largely rural area.
"Residents there got little warning besides seeing the tornado as it approached," Proenza said.
"Without doubt, the 1999 Oklahoma City tornado had the potential to be the most deadly tornado of all time. ... We onl yhave to look around at the technology in this area to see how we kept down the death toll in this major, major disaster," Proenza said.