It was at least a 12-inch mistake -- the extra foot of snow that dropped on the Washington area Sunday night and Monday morning that virtually no one had mentioned or forecast.
To professional weather forecasters, it may be remembered as one of the area's worst forecasting disasters in recent history.
What went wrong?
The National Weather Service (NWS) blames it on mistaken interpretation of computer data. One private forecaster says the weather service was paying more attention to computers and less attention to fast accumulating snowfall to the south of Washington.
A second private forecaster, a 24-year-old meteorology student in Lawrence, Kan., said it was not a matter of interpretation, but rather a "mechanical error" on the part of the computer.
Whatever it was, forecasters who made it to work Monday morning were slinging a little snow at each other and puzzling over the storm that left the greatest accumulation since 1922, according to the National Weather Service office at Suitland, Md.
"It was a judgment error, profesional judgment," said 36-year-old Chet Henricksen, the head NWS forcaster here.
"It was a small but intense low which I think made it much more difficult for the mathematical and physical equations which go into the computer... to handle it," said Henricksen.
The computer said the storm would move out to sea over Cape Hatteras, N.C., but instead it turned north and with a powerful surge of warmer, moist air, collided with frigid cold air from the Northeast.
"The system for analyzing the data and putting it into the computer may miss some initial data and not handle it well," said Henricksen. "Quite frankly, it was about as wrong as you can get."
However, in interpreting the computer's analysis of the storm, Henricken said his forecasters failed to recognize its massive snowfall potential.
"We look at it and say, 'What could be a problem,' and in many cases it's nothing and in a few cases something drastic happens," the forecaster said.
However, a spokesman for the Pennsylvania-based Accu-Weather service, Joel Myers, said the NWS "were just buying an obviously faulty computer prediction.
"The computer forecast did a really miserable job," said Myers. "We never bought it because much more snow was falling to the south earlier and it (the computer) was predicting much less down there in Washington."
Myers said his private service discovered that the computer prediction for areas south of Washington was three to five times less than what was actually falling. "So we multiplied our forecast three to five times," Myers said.
"They (the NWS forecasters) really didn't catch that trend," Myers said.
In reply, sand Henricksen, "He's full of it. I think if you look back and see what we did and what they did, it shows that we were more on the ball."
Henricksen said that at 4 a.m. Sunday when the weather service was predicting two-foot accumulations, Accu Weather was still predicting one-foot totals. "I don't think they were any more on it than we were."
More than 1,000 miles away, a 24-year-old meteorology student at the University of Kansas, Larry Cosgrove, said he was out in front of the forecasting pack by calling for one-foot or more accumulations in a 10 p.m. broadcast over WRBS-FM radio in Baltimore.
"I saw a very grave error very early on the long range prox (projection)," said Cosgrove. He said he recognized the combination of upper atmospheric changes and related weather patterns from South Carolina to Vermont that told him the storm would not only turn north, but would create much stronger conditions for snowfall.
"I would not blame them," Cosgrove said referring to the NWS forecasters. "It was instinct on my part versus computer mechanical error."
Cosgrove's recent claim to forecasting prowess was his six-day advance warning of the most recent blizzard that paralyzed Chicago.
In response, Henricksen said, "Let me set the record straight on this thing. It's very easy to be irresponsible and over-forecast, but it doesn't do the public any good."
Henricksen said he thinks most private forecasters believe that the NWS is too conservative in its forecasts.
"I don't think that's true," he said, "at least our verification records don't show we under-forecast."
Said Cosgrove: "I would imagine they were playing it conservatively. The last run of the computer yesterday (Sunday) morning was showing four to six inches in your area and there was not new information until before midnight."
Said Henricksen: "I'm not saying the guy's a jerk, but it's always easier after the fact..."