"Even in Buffalo, where the ability to endure severe weather is a source of civic pride, few expected a snowstorm like this in mid-October including those who predict them.
Don Paul, a meteorologist at WIVB-TV4 in Buffalo, has worked in the area for more than 20 years. 'Of all the events I’ve seen here, this storm involves the most widespread devastation in the most populated area,' he said. 'It’s absolutely an historic storm.' "
On Oct. 14, 2006, the New York Times published a lengthy story on our just-passed disastrous storm, including comments from then. Gov. George Pataki, National Grid, Mayor Byron Brown, the National Weather Service, Buffalo citizens and me (see above). The storm, still called by many the “October Surprise,” remains the single most destructive weather event in my 37 years here in Buffalo. It didn’t seem likely it would turn out to be of such incredible magnitude heading into the storm.
Some snow had been in the forecast for six to seven days ahead of the rare early lake-effect outbreak. That far in advance, the thinking was the best chance for accumulating snow would be on higher elevations inland from Lake Erie, as is usually the case with early season lake snow. In those first few days of the advance forecast, low-level wind direction could not be accurately determined on a local basis. But the models of the time were already projecting a deep, unseasonably cold low pressure system to drop down near Lake Superior from the north. Such a low would be pumping exceptionally cold air aloft over warm Lake Erie somewhere into Western New York.
Scanning past climate records I could only find limited archived evidence of heavy lake snow in northern Erie County so early in the season, nearly a century ago. Since I knew such snow to be exceedingly rare, I leaned heavily on Buffalo climatology in making my final forecast the night before.
The models of the time were painting a picture of extraordinarily cold air about a mile up arriving by Oct. 12. The temperature difference between the lake surface and the 5,000-foot level was going to be in the range of 22-24 degrees. With that big a lapse in temperatures from the warm lake to the cold layer, the warm, moist air at the surface would rise rapidly and form deep convection.
When Oct. 11 arrived, Lake Erie was still 62 degrees, two degrees above normal for the date. By then, it had become apparent the low-level winds would be focusing mainly on northern Erie County, including the metro area. Even so, there was no way to make a higher confidence forecast on accumulating snow so early in the season. I assumed the layer of air near the surface would be warmed by the lake and the warm ground. Heavy lake-effect precipitation would be inevitable, with thunder, but the extreme cold aloft would be modified by the surface warmth, or so we thought.
On the evening of Oct. 11, I had leaned toward a mix, with a “possible” slushy accumulation of up to 2 to 4 inches in parts of northern Erie County. I made it plain this possibility was not high probability. (I was aware the then-worst case scenario of 2 to 4 inches would produce considerable damage if it occurred.)
On the morning of Oct. 12, the Detroit National Weather Service weather balloon sounding showed an unprecedented minus-8 Celsius temperature at 5,000 feet, upwind from Buffalo creating a huge temperature lapse rate and extreme instability.
By late in the afternoon and early evening on Oct. 12, courtesy of that lapse rate, the first 2 to 4 inches of slushy, water-laden snow already began taking its destructive toll on fully foliated trees. Branches were crashing down from the weight of the slush on the foliage. All of us meteorologists already knew we were in serious trouble. Even so, none of us imagined we’d end up with more than 22 inches of accumulation by the next morning.
It turned out 4 inches of slushy snow was enough to cause a great deal of the destruction that occurred. By the time we reached that range in the early evening, power outages and tree damage were well underway.
The worst damage came early in the event, as recently noted by the Buffalo NWS.
An additional factor tied to the extreme snow was the drier air southwest of the metro area drawn into the convective precipitation, allowing for more evaporation and cooling as rain fell into drier air with lower dew points. This evaporative cooling enabled more conversion from rain into graupel pellets first, then snow. The mechanism went into high gear. Thundersnow, generally associated with heaviest snowfall rates, became nearly constant for up to 12 hours. Within the convective band, cloud tops reached 25,000 to 30,000 feet, almost never seen in lake effect. The dashes represent lightning when this radar image was recorded.
Although many of you have heard “1 inch of rain equals 12 inches of snow,” that is a gross oversimplification. The temperature profile of the lower atmosphere makes all the difference. In a deep layer of arctic air, the ratio can be 1 inch liquid = 25 to 30 inches snow. In more marginal temperatures, such as those the evening of Oct. 12, 2006, the ratio was 1 inch liquid = 6 inches snow. That made for extremely dense, slushy snow added to the weight of the foliage, exacerbated by the wind causing limbs to sway.
In addition, the soil was already saturated from a previous rainy period, further weakening trees in the face of this onslaught. The ratio expanded to 1 inch liquid = 12 inches snow by morning, but most of the damage was already done before that.
For those of you who crave more technical details on the meteorology preceding and underlying this extreme weather disaster, Buffalo NWS Meteorologist and Winter Weather Team Lead Bob Hamilton has compiled an excellent YouTube narrative on the evolution and impacts of the storm.
Incidentally, we will be turning cooler this coming weekend. That’s “seasonably” cooler, and there will be no surprise in that. There probably will be other extreme October events in our future, but we are safe this year. (I can almost hear the skeptics muttering "or so he thought.")