The Grinch is Real

[ariof]

So based on some Twitter threads I got to looking at how climate data at KBOS anyway (1936-2021) has not been changing over the years at the same rate at different times of year. After abusing Excel (as usual, I got to the point where I should have stuck this into R or py or something, but had too much invested) I came out with the attached chart.

What this is showing is the change in temperature, per year, for each calendar day of the year (excluding February 29). There's some weird stuff going on that I can not begin to explain.

• Warming is not taking place at all evenly over the course of the year. It varies by quite a bit, but if you look at the orange line (21 day moving average; longer periods don't really change things) it's clear that there are several local minima and maxima to the rate of warming.
• Specifically: 
  • Late fall—from basically Columbus Day to Veterans Day—is not warming at all. 
  • Same for early May to Mid June, although there is a slight upward trend here (<1˚/100 years), and a period in late Morch
  • Jul-Sep is warming at a rate of about 2˚/100 years, with late September slightly faster.
  • Met winter is warming at a rate closer to 3˚/100 years, with lots of variability
  • Late December is warming at a rate of 6˚ per 100 years, which is by far the fastest warming of the year.
• Looking at the variability, there seems to be a ~21-day period from late Nov to early Mar between rapidly warming cycles (7˚/100 years) and periods really without any perceptible warming. The cycles peak on 29-Nov, 22-Dec, 11-Jan, 1-Feb, 22-Feb, 9-Mar.
• The 7-day average's highest five days, warming at a rate of 7.5˚ per century, are December 20 to December 25. December 22 and 23 are two of the five highest days in the year. (The others: 12-Jan, 2-Feb and 1-Dec.) So basically the Grinch is very real and has been getting worse over time.
• The seasonal variation probably has something to do with the ocean+sun, given that the two minima are about 6 months apart when the sine curve of the sun flattens. But the local minima (3-ish week period) during the winter probably has to do with something else entirely?

Anyway, IANAMeteorologist and IANAClimatologist and I may have buggered up these data completely. But if not, a) it's interesting and b) the Grinch is real!

 

[sheahunter]

interesting stuff, thanks for posting

[tamarack]

Only 24 years of record here, but there's a distinct difference for the period 12/21-28 (centered on 24/25) compared to the 8-day periods before and after.

Numbers are temp departure, daily snow and number of 10"+ storms for each period.
12/13-20:    -0.9°    0.74"    2
12/21-28:   +1.5°    0.49"    1 (on 21-22/2008)
12/29-1/5:  -0.5°     0.94"    5

Looking just at 12/23-25, it's even worse, temp avg +3.7°, daily snow only 0.34" and apart from the 8" on Christmas 2017 there's no storms reaching 5".

[Whineminster]

Cool data, confirms that ginxys COC fetish in early summer is real

[powderfreak]

I know the Fairbanks Museum in St Johnsbury, VT has seen the same thing.  I have to see if I can find it but like the 10 days around Christmas have a temperature rise that’s not seen before or after that period.  IIRC they had like an almost 2F rise lately.

[powderfreak]

Couple off the cuff thoughts…

As cold air presses south in winter, it’s fluid like the ocean tide coming in.  The overall tide comes in but there are waves that push forward and then retreat back (just not as far back as the previous wave).  Is there some climatological bunching that we are seeing the “tide/wave” retreat a bit in that period after pressing in early/mid Dec?

Almost like late Spring has a well documented secondary snowpack maximum in the NNE mtns?  The tide/cold starts heading back north by March but by late March/early April there’s another surge south and some renewed cold, just not as cold as before?  Trying to conceptualize air mass movements as ocean tide in or out, you see some fluidity back and forth despite an overall movement of cooling or warming.

When I first moved up here to VT a couple decades ago, everyone talked about the “January thaw.”  That hasn’t been as big of a deal the past decade… it feels like it’s moved to the Holiday thaw or the “Grinch” systems.  Has there been any reason for a pivot there in timing?  Is it just random bunching on short time scales?  

[beavis1729]

Interesting data. While I don’t live in the region, I have a theory as to why the week before Christmas has seen the most warming. I think it’s a few things:

(1) Some randomness of the data, as powderfreak mentioned 

(2) SSTs are warming, and this time of year could be when coastal temps are highly influenced by SSTs
 

(3) This is a time of year when snow pack in the immediate source region is highly variable from one year to the next. And due to the general background warming trend and the sensitivity of upstream snow cover at this time of the year, there are many more years with little/no snow on the ground upstream (and in SNE) than there used to be. 
 

Once you get into January, there is probably a good amount of upstream snow cover regardless of how mild December has been…which is why the warming in January isn’t as great.