2017/03/20

Edmonton's Warmest & Coldest Winters

Today is officially the first day of spring, although we'll probably still have snow for a few more weeks yet.

But with the winter of 2016-2017 nearing its end, was it a warm one, a cold one, or pretty typical? I remember lots of coldsnaps and warmspells, but I don't have a great sense of how this winter compared to others. So today we're going to try to find out.

Temperatures for Winter 2016-2017

"Spring" technically starts today, March 20th, and "Winter" technically started on December 21st. But because the Canadian prairies take a slightly more relaxed approach to the seasons, here is a reminder of our temperatures going back to November:
  • November started very warm, with a few all-time records. And things stayed fairly warm through the end of the month.
  • December started with a long coldsnap, and after that the temperature jumped around.
  • January also started with a long coldsnap, followed by some very warm days and recent-records.
  • February also started with a coldsnap and a few recent-coldest days. That was followed by some very warm days, including an all-time record.
  • And March also started with a coldsnap, before warming up recently.

So we've had a lot of ups-and-downs, which isn't particularly usual. But how does this winter compare to others?

Warmest and Coldest Winters

This chart is a variation of one that I used last year in Edmonton's warmest years (or NOAA anomalies, NOAA problems), except this one is only looking at the months from November through March. If you're interested the details behind the chart it's probably easiest to go back to the original post.

But generally, this compares the temperatures for each winter to the 20th century average. Warm winters are in orange, and cold ones are in blue. Using this approach, Edmonton's warmest-ever winter was last year, and it was on-average 5°C warmer than the average for the 20th century.

2nd and 3rd place are a bit of a surprise, coming in as 1930-1931 and 1888-1889 respectively. But we have seen both of those years show up unexpectedly before - the winter of 1930-1931 had only a single day below -20°C; and 1889 was Edmonton's 8th-warmest year on record, even though it was more than a century ago.

2016-2017 is dashed-in on the right, because this chart is counting all the way to the end of March. Right now 2017 is sitting at about 2.5°C above the 20th century average, which is pretty much inline with the winter of 2014-2015 two years ago.

Lets take a closer look at the breakdown for each winter:

The Horserace

This chart is a variation of one that I used last year in How Warm is 2016? And for all of the details behind it, again it's probably easiest to go back to the original post.

The short version is that this tracks how warm or cold things were, on a day-to-day basis. Days that are warmer than the 20th century average for that date gain points, and colder days lose them. With this chart, if I show all 130+ winters it's an unreadable mess. So here we just have the recent winters going back to 2006-2007, and the colourful sections in the background show the historic ranges.

Looking at the redline for 2016-2017 we can see:
  • very warm and above-average November
  • colder-than-average early-December
  • fairly-average flat section from December through January
  • chinooks in late-January and mid-February
  • colder-than-average early-March

And here when I talk about the "average" that isn't just recent years, but is looking at the entire 20th century from 1901-2000.

As of late-March then, the winter of 2016-2017 is roughly tied with 2009-2010 and 2014-2015, ranking as Edmonton's 25th~ish warmest recorded winter. In terms of recent years, that makes this a fairly warm winter, with the only significantly warmer ones being 2015-2016 in 1st place overall, and 2011-2012 in 6th place.

That also means that 6-of-the-last-10 winters were colder than this year, with 2010-2011 and 2013-2014 both coming in below the 20th century average.

5 Warmest Winters

Here's a look at the 5 warmest winters of all-time, with 2016-2017 included as a reference. Our very warm November had us in 1st place for a little bit, but that tapered out by mid-November.

The Top-5 are all really close to one another, although they took different paths to get there:
  • 2015-2016 (in orange) followed a slow-and-steady path to the top spot.
  • 1930-1931 (in green) was out to a huge lead at the end of February, but then a cold March dropped it down into 2nd.
  • 1888-1889 (in dark red) started quite slow, but had huge gains in January and February to reach 3rd. 
  • 1976-1977 (in turquoise) also had a slow start, before taking off in late-January to finish in 4th. 
  • 1987-1988 (light blue) had a very warm start in November and December, before some drops in early January and early February.

One thing that we can see from this chart, is that the overall ranking depends on exactly which dates we choose. 1930-1931 was in first place right up until the final week of March. And so if we want to say that winter officially ends around March 20, then maybe 1930-1931 was truly Edmonton's warmest winter?

I'm not going to get too hung up on that though, because I think the definition of winter needs to be a bit more fluid in this part of the world. But it does bring up another question, which is when exactly does winter start?

October, November, or December?

This is the same chart  that we were looking at earlier which shows the last 10 winters, except here the starting point changes from October 1st, to November 1st (which we've already seen), to December 1st. Because this whole thing is cumulative, the final results depend on what day we choose to start from.

We've already looked at what happens when we start counting in November, and in that case 2016-2017 is about the 25th warmest winter. But we know that November was really warm, and that gives this winter a pretty big boost.

If we use December 1st as the beginning of our winter instead, then we're almost immediately into the negatives, because December was pretty cold. From there 2016-2017 doesn't hit the average again until the chinooks of late-January. And as of late-May it's roughly tied with 2007-2008, 2009-2010, and 2012-2013 for a ranking of about 40th-warmest winter.

If we use October 1st as the start of our winter, we are also immediately into the negatives, because last October was really cold. That gets offset by the warm November, and by late-May we're in roughly 40th spot again, and tied with 2007-2008 and 2009-2010 again.

So all of that is just a long way of saying that none of this is written in stone. Looking at the last 10 winters, depending on exactly how you count 2016-2017 would fall somewhere between typical and warmish.

I think November 1st makes the most sense as the unofficial start of our winter. With the switch from Daylight-Savings time it now makes a really distinct break from October. Even without the time change though, November is the point where snow starts to stick around. Winters with snow in early-November feel very different from the ones when we are snowfree until late-November or early-December. So I definitely think the heatwave of November 2016 is important to count here, because it was certainly a notable part of this winter.

And personally, I don't actually consider the winter to be over yet. The charts today are really just preparation for a full Winter 2016-2017 report card that I'll do in maybe mid-April. At that point we'll revisit this, but will also look at warm days, cold days, snowfall, and etc.

No comments:

Post a Comment