Forecasting the first cold snap of 2015-16


I’m not one to pat myself on the back. Not too much at least. The progression of this year’s first cold snap, the strong forecast for it, is worth noting though. And as I wrote this I thought maybe I’ll try to do these “forecast recaps” when I think to do so.

During the first week of the month, as immediate concerns of Joaquin and the Southeast/South Carolina flood began to wane, the idea of the season’s first cold snap started to show itself.

By October 5, the Euro was teasing with long-range 0c 850 lines.

Given pattern progression, and a tendency for strong El Ninos to have ‘early’ freezes (see pasted data at bottom), I was already sort of on the lookout for it. Of course, western parts of the D.C. area see their average first freeze around mid-October anyway.

The signs persisted, and grew with time. But there was still over a week before the cold air reached the region.

As the cold snap reached the near-term, about 3 days out from happening, details became increasingly apparent.

It was perhaps something of a win for the GFS in its constant battle versus the Euro. The GFS was both arguably the first to consistently show the cold shot (though the Euro was hinting at it in long range for a while) and it also generally won the short-term battle of actual temperatures.

While the Euro ran “warm” much of the time, a down-scaled version concocted by Ryan Maue over at did pretty well in capturing more of the actual temperature variations over the region. It was also fairly accurate with its temperature “improvements.”

As the cold snap set in, CWG took a look at what was coming and how it stacked up. Noted above, it actually came about when it should have for Dulles. Though in the end, places like BWI got there a good 10 days earlier than typical.

The cold push roughly produced highs of upper 40s to mid-50s at the low end and lows in the upper 20s to lower 30s for “peak morning cold.”

Amidst the cold, the first wintry weather of the season came to the local area as well (note: DC saw neither wintry precip or a freeze, and might not until March. #floridaing).

Up north, it was even more awesome. If you like that sort of thing.

Top image via Weatherbell Models.


I pasted some research I did below while conjuring up my guesscast for the first freeze contest we do over on AmericanWx every year. The strong Nino sample is so small it’s kind of embarrassing to even say I considered it as a reliable guide, but eh… It helped keep me in third place there so far.

==DCA== Nov-18 (1985-2014)

Strong+ (before normal, all older)

Nov-11 (1957)
Oct-29 (1965)
Oct-20 (1972)
Nov-06 (1982)
Nov-16 (1997)

Mod (near normal, 09 very late)

Nov-13 (1963)
Nov-13 (1986)
Nov-11 (1987)
Nov-04 (1991)
Nov-27 (2002)
Dec-06 (2009)

==IAD== Oct-17 (1985-2014)

Strong+ (before normal, all older)

Oct-06 (1965)
Oct-10 (1972)
Oct-17 (1982)
Oct-02 (1997)

Mod (near normal)

Oct-14 (1963)
Oct-07 (1986)
Oct-09 (1987)
Oct-08 (1991)
Oct-18 (2002)
Nov-04 (2009)

==BWI== Oct-29 (1985-2014)

Strong+ (before normal, all older)

Oct-13 (1957)
Oct-26 (1965)
Oct-20 (1972)
Oct-18 (1982)
Oct-23 (1997)

Mod (near normal)

Nov-04 (1963)
Nov-13 (1986)
Oct-26 (1987)
Oct-21 (1991)
Nov-01 (2002)
Nov-04 (2009)

Leave a Reply

Your email address will not be published. Required fields are marked *