November Statistical Fun and Games

Yesterday on my adventures across the northeastern part of PA, I was made keenly aware as to just how cold it has really been the last couple of weeks.  In what amounted to a bright and sunny and somewhat "gorgeous" mid autumn day, I observed a vast smattering of residual snows from earlier in the week.  Some of the snow was in deeper darker valleys while some sat vibrantly atop of peoples homes and roofs.  Still other evidence was small piles of snow that were clearly plowed and pushed about making the snow that more difficult to fade away.  All in all, it was impressive to see the evidence of the cold without a real snowstorm to deliver the true goods of a widespread snow event.  In addition, many of the smaller ponds that I passed by were frozen over with a rather distinct layer of ice.  Yes, it has been cold.

Which got me thinking early this Saturday morn.  Just how cold?  First, here is a graphic of the 1st half of November in terms of temperature anomaly for 2019 across the continental US (CONUS).

Contrast that with 2016, just 3 years ago!  What a difference...

So what I decided to do was the following.  Go back to 1981 and find the years where there was a cold starting November.  Not just cold, VERY COLD.  You see, over the last 39 years, this current year is FIVE standard deviations from the mean of that period. (Thanks to spreadsheets, that made this exercise a bit less laborious!)  That's quite anomalous.  So I found only 3 other years where the CONUS temps to this point were that far from the average.  The last time November has been this cold to date was 1991...take a look...now that is darn cold!

Next, I used the Earth Science Research Lab plotting tool to map what the subsequent December, January, February, March period resembled in terms of temperatures.  Since 1991 was such a large outlier, I triple weighted it as its SD was 3 times greater than 2019.  Then I also just double weighted it and then simply single weighted that year.  Here are the other years that were similar to this November:  1993, 1995, 2000.  Drumroll...here are the results.

Triple weighted 1991:

Double weighted 1991:

Equal weighted 1991, 93, 95, and 2000:

The signal that emerges is a cold northeast and a warm northwest...So only time will tell, but a cold November does seem to portend a colder than normal winter here in the eastern and especially the northeastern US.

OK...we're off to walk the dogs, check out the latest and greatest fitness equipment at our newly constructed PF, then visit our precious granddaughter.  Once all of the excitement eases, it's off to sling pork and sauerkraut for the annual Sauerkraut Supper.  As a result, our hair will hold the scent of fermented cabbage for at least a day or two to come!  Ha!

Go PSU and let's rebound from that hiccup called Minnesota!

I hope you enjoy this tune about November...

Smitty

AA:  A cold November looks as if it may signal an upcoming colder than normal winter in these parts.  I did the heavy lifting for you but I kinda like number crunching as you know.