The time was 3:35 p.m.
That's when the EF5 tornado of May 20 ended.
For various people and different reasons, it will not end.
Among those are the meteorologists of the National Weather Service, Norman Forecast Office.
In the days after the tornado, questions were fielded from state, national and international media by David Andra, the meteorologist in charge, and Rick Smith, the warning coordination meteorologist, both of that office.
But now they want answers, both meteorological and social. They want to know how the storm worked and why it did, what it did, when it did. They also want to know the same about the response of the public. What actions did they take? What did they do differently on May 20, versus the F5 of May 3, 1999?
They already know of one of the differences: social media.
“We're trying some new things here to communicate with people,” Smith said. “The weather service since the '60s has been yelling in all capital letters to people with these standard looking tornado warnings. But with Twitter and Facebook and some of these other technologies we're able to talk to people more directly, more simply.”
At least 19 tornadoes occurred on May 19-20, including two EF2s, one EF3, one EF4 and one EF5 with more information still being collected by the National Weather Service.
In looking at the Norman Forecast Office's Facebook numbers, it gained just short of 10,000 likes from May 18 to May 28. And its Twitter account went from 14,373 followers to 19,626.
“One of the questions would be just how did people get their information?” Smith said. “We know typically a lot of people watch television to get their information, but social media is so big now and we put such a big emphasis on it this time around more than we ever have.
“Did that work? Did anybody see a tweet or a Facebook post that really made a difference, did that really save someone's life?”
Andra said they take this “Monday morning quarterback” approach whether they've just come through a winter storm, hailstorm, or wildfire. But on tornadoes they've really looked into numbers that maybe aren't typically associated with weather.
After a May 24, 2011, tornado outbreak, which included an EF5 in central Oklahoma, they studied traffic patterns, normal versus that day. There were forecasts days before stating that large damaging tornadoes were possible May 24.
“It had never really happened before where people were taking so many actions at 9 and 10 o'clock in the morning based on a forecast,” Smith said. “There were businesses that closed early that day. And many schools were rescheduling events.
“So we have to ask ourselves, ‘If we're going to tell people this far in advance, what are they going to do with that information?'”
In their study they found that traffic was well below average at several locations at rush hour in the Oklahoma City metro on May 24. At 3 p.m., there were 1,000 or more vehicles more than usual on the roads at six of those 11 sites. At 5 p.m., there were about 5,000 to 7,000 fewer vehicles at six of the 11 sites.
Mike Foster, who since has retired as the meteorologist in charge, said then that they were not sure if the low traffic count in each case was because vehicles weren't on the roads or because they were bumper to bumper and couldn't move past the sensors. One concern is that people waited too long to drive somewhere to seek shelter.
So they continued to study behaviors. Andra said there are many scientists at the National Weather Center who can provide help in studying the meteorological questions.
“The other part about what people did can be the more difficult question because it comes right down to the individual basis,” Andra said. “And that will be hard information to get, but if we can understand why people respond the way they do, maybe we can tailor our information to be better on days like May 20.”