Wednesday, November 28, 2012

Oct-Nov Temperature Correlation to Dec-Jan-Feb Found

An article connecting temperatures in October and November to the following 3 winter months (December, January and February) seems to have proved the long-held theory that fall trends can influence the upcoming winter.

The correlation is cited in the summary of this article:


In summary, temperature at the Blue Hill Observatory during October-November
correlates to some degree to the following winter temperature more than half the time, with an
overall correlation of 0.45
and with much higher and much lower correlation over some ten-year
periods during the 20th century. Both precipitation and snowfall, which are much more highly
variable than temperature, are in general essentially uncorrelated between ON and the following
winter, though some correlation is occasionally apparent over some years due to random
variations. As a result, ON precipitation and snowfall provide little predictive capability of the
subsequent winter precipitation and snowfall. Although ON temperature appears to provide some
prognostic information about the following winter temperature, the extent to which the
correlation is effective very likely depends on the specific atmospheric dynamical processes than
dominate the weather conditions during particular fall and winter seasons.

 0.45 may not seem like much, but that's saying that, out of 100 winters, 45% of those winters had temperature connections to the previous fall's temperatures. That's a pretty significant correlation. Let's take a look back at our October 2012 temperatures to see what we could see this winter.

This map showing October 2012 temperatures displays a very below-normal October was observed, especially across the Ohio Valley, Mississippi Valley, South Plains, Central Plains and Northern Plains. It does appear that the Northern Plains achieved the lowest temperature for the month of October- a good sign for the coming winter if you like cold weather.

Of course, this article mentions October AND November, so I would like to see November temperatures before I make a conclusion on what fall temperatures could mean for this winter.

Andrew

2 comments:

WolfForever said...

Likely November finish = colder than normal in region bounded on west by states IN-KY-TN-MS, and then extending south and east to the coast from that. Also colder than normal in New York and parts of Southern New England. North/west of that near normal for a small band of areas than warmer than normal except in parts of the extreme northern Plains states.

Anonymous said...

So,are we going to have an neutral enso winter?