2017/12/31

New Year's

Today we're going to look at the weather history of New Year's Eve and New Year's Day in Edmonton:

New Year's Eve Temperatures

In the last 30 years the average temperatures for both New Year's Eve and New Year's Day have been a High of -8°C and a Low of -16°C. Just last week we looked at Christmas Day, which has an average that's a little bit higher with a High of -4°C and Low of -13°C.

One of the important things about New Year's Eve is always hoping not to freeze at -20°C or -25°C while waiting for the fireworks in Churchill (or at any other outdoor activity). And that's a bit hit-or-miss. In the last ten years 2008, 2009, 2010 and 2013 all had Lows below -22°C.

This morning the temperature dropped to -27.9°C, and if that ends up being the Low for the day then that would make 2017 the coldest New Year's Eve since 1984 which was at -29.3°C. But the forecast for tonight is calling for the temperature to fall further to -29°C, and so the city has announced that it's cancelled most of the outdoor events. New Year's Eve 1984 was -29.3°C, and 1981 hit -29.4°C, and to find anything colder than that we would have to go back to 1964 & 1965 which both dropped to -31.1°C.

On the bright side though, the last three New Year's Eves were fairly warm with Highs of 3°C, 2.5°C and -1.5°C respectively for 2014, 2015 and 2016, and then Lows of -3.6°C, -14.4°C and -9.7°C.

New Year's Eve Temperatures

Unsurprisingly, New Year's Day has been very similar to New Year's Eve.

In recent years 2015 and 2016 were both warm with Highs of 3.2°C and 1.4°C, but 2016 was cooler with a High of -8.8°C and Low of -17.1°C.

New Year's Eve & Day Precipitation

When we looked at Christmas last week I combined the precipitation numbers for Christmas Eve and Christmas Day, and I think that's a fair approach again since New Year's Eve and Day are so closely tied.

We saw that Christmas hasn't gotten new snow very often in the years since 1992, but for New Years it's about 50/50. (please note: Blatchford is missing data for 2007-2014, so in the chart above the International's data is correct)

Jewelry stores used to do a promotion where if it snowed X amount (I can't remember how much) on New Year's Eve or Day (also can't remember which) then anyone who'd bought something during the holiday season would get it for free. Across the two days the most snow we've gotten recently (ie. since the 1930s) was 11.8cm at the International for the 1997-1998 New Year's.

And since the 1880's rain has been recorded twice - 0.5mm in 1958-1959 and 0.6mm in 2000-2001.

New Year's Day Snowdepth

And finally, here we have the snowdepth at Blatchford and at the International on New Year's Eve.

The average hovers around 15cm. In 2014, 2015 & 2016 we were on the low side of that, but in 2013 the International was way up at 41cm (Blatchford data is missing for that year).

We recently looked at Low-Snow (or brown) Christmases, and here we see several of those years were still lacking snow on New Year's: 1959, 1989, 1993, 1997, 1999 & 2005.

2017/12/30

It's Cold

Earlier this week we looked at the Blatchford weather station's 1,080+ day streak without a temperature of -30°C.

That came to an end on December 30, 2017, having lasted from January 4, 2015 for a total of 1,089 days. That made it the 4th longest streak, following a 1,109 day streak that ended in 2009, a 1,138 day streak that ended in 1987, and the 1,835 day streak that lasted almost 5 years from January 13, 1998 through January 3, 2003.

Today we're just going to look at some charts.


This is an updated version of the chart from a few days ago, which shows all of Blatchford's Lows below -30°C. Our two new additions appear on the right side.

And as I said last time, the big thing to take from this is just how much sparser the right side of the chart is compared to the left.

...and now in 3d!

Here we have the same data in a slightly format. These 3d charts aren't particularly clear, and so I don't use them very often, but sometimes they're fun to play with.

This shows every day since 1880 that had a Low of -20°C or below, and the focus again is really is on the left side versus the right side. Almost all of Blatchford's -40s and -45s occurred before 1950. As we move into the -35s and -30s we get some points appearing on the right side, but they're pretty sparse.

Recent History

Things can get a little busy going all the way back to 1880 so here we're just looking at recent history back to 1980. The previous chart showed Lows below -20°C, but this one includes everything below freezing.

Each of the decades here recorded 3 or 4 Lows below -35°C...except for the 2010s which hasn't had any yet. 

And for Lows of -30°C or below the 1980s had 33, the 1990s had 42, the 2000s had 18, and so far the 2010s are at 10.

Edited to add:

And here's the same chart for cold temperatures at the International Airport, going back to when its records began in 1960.

2017/12/28

1,080ish Days without -30°C

In the last week we've looked at the First -20°C of the Winter and the First -25°C of the Winter. Today we're going to be following that up with a look at -30°C.

The last time that the Blatchford weather station recorded a temperature of -30°C or lower was back on January 4, 2015 with a Low of -30.2°C. As I write this on December 28th we've gone 1,088 days without a -30°C, but that streak might be ending because we have a few cold really night in the forecast. 1,088 without a -30°C sounds impressive, but is it?

Really Cold Days

This chart shows every day since 1880 that had a recorded Low temperature of -30°C or below. The dots are colour-coded so that the colder days have darker dots, although it's a little hard to tell them apart here, but we'll take a closer look at that later on.

There are a few things to take from this chart, but the best place to start is to compare the left and the right sides. The left side of the chart - especially the 1880s - is much more dense because there were many more cold days recorded. And in recent years - from about 1980 onwards - temperatures below -30°C have been fairly rare. Of the last 20 winters there were 9 that didn't have any -30°C temperatures.

Another thing to notice is exactly when those -30°C temperatures show up. Since 1997-1998 we've had 7 -30°C days in December, 16 in January, only 1 in February, and 4 in March. The years that had -30°C in March were 2003, 2009 and 2011, and those type of temperatures that late in the year just aren't fair.


This is an old chart that we first looked at way back in one of this blog's earliest posts, and it provides some context for the difference between the left and right side of the first chart. 130ish years ago Edmonton would average 24 days below -30°C each winter, but now that number has fallen to around 1.

Trivia

This chart digs a bit more into the various levels of extreme cold.
  • -45°C is ridiculously cold, and even a century ago it was pretty rare. We haven't had a -45°C since February 8, 1939, and there were only ever 17 recorded from 1880-1939. The coldest temperature ever recorded was -49.4°C on January 19 & 21, 1886; and then again on February 3, 1893.

    The International Airport still occasionally sees temperatures in that range, including -46.1°C on December 13th, 2009.

  • The city hasn't recorded a -40°C since January 25, 1972, and before that the previous -40°C was 22 years earlier, on January 26, 1950.

  • Currently we haven't hit -35°C since December 13, 2009. That's more than 2930 days, which is the longest we've ever gone without a -35°C. The previous record was 2,556 days from January 27, 1997 through January 26, 2004. And before that the longest record had been 1,856 days from December 25, 1983 through January 31, 1989.

  • Our current streak of 1,080ish days without a -30°C is the 4th longest. It's about a month shorter than a 1,109 day streak that ended in 2009, and two months shorter than a 1,138 day streak that ended in 1987. But it's still well below the 1,835 day streak that lasted almost 5 years from January 13, 1998 through January 3, 2003.

Cold Highs

For a change of pace, here are the coldest High temperatures recorded since 1880:
  • Edmonton has never recorded a High below -45°C, and there was only ever one at -40°C or lower, right at -40°C on January 30, 1893.

  • Highs at or below -35°C were also pretty uncommon, with only 13 of them ever recorded. The most recent was a -35°C on January 20, 1943.

  • Edmonton has had 110 days with Highs of -30°C or below, although since 1980 there have only been 6, and only 1 since 2000. The most recent was a High of -30.6°C on January 27, 2004.

  • Since 2000 we've had 10 days with Highs of -25°C or below. Most of them have come in groups: January 26-28, 2004; January 28 & 29, 2008; and December 13 & 14, 2009. The most recent was December 6, 2013 with a High of -25°C

  • Lows at or below -20°C are fairly common. Recently both 2001 & 2016 didn't have any, but some years had lots: 2008 had 10, 2009 had 10, 2011 had 12, etc.

Lows Below -20°C

And finally, here we're back to Low temperatures, but this time the Lows below -20°C and -25°C have been added. This is a bit of a companion piece to First -20°C of the Winter and the First -25°C of the Winter, since it shows the "season" of cold temperatures in Edmonton.

The -20°C days start around mid-November, and that hasn't changed much over the years. The cold days end by about mid-March, and that's maybe a week or two earlier than they used to. But the big difference here is again the density, with the left side having a lot more dots, and a lot more of the extra-cold dots compared to what we've seen recently.

Today has had a bit of a high-level, historical perspective. If you're interested in the specifics of some recent coldsnaps we looked at those in more detail here: Coldsnaps & Deepfreezes

2017/12/27

First -25°C of the Winter

A few days ago we looked at when Edmonton gets it's First -20°C of the Winter.  Today we're doing the sequel, as we look at when -25°C first appears.

Recent Years

This year we hit -20°C for the first time on December 24th, and that was late compared to some recent years, although it's not unusual. Then one day later we hit -25°C, on December 25th.

Of the 23 years shown here, there were 11 where the first -25°C showed up within a few days of first -20°C: 2017, 2011, 2010, 2009, 2008, 2006, 2003, 2000, 1998, 1997 & 1996.

And then there are some other winters where the first -25°C didn't show up until late-January or early-February, like in 2005-2006 and 2007-2008.

History

And here is the entire recorded history.

There was one winter in Edmonton that did not have any temperatures below -25°C, and that 1986-1987. You can't really see it in the chart though, because it's pretty buried.

The largest gap here was 1930-1931, with -20°C showing up early on October 19, and then 144 days later -25°C hitting very late on March 11. 1991-1992 was similar, getting it's first -20°C on November 1, and then -25°C arrived 110 days later on February 20. After that 2005-2006 had a 74-day gap, and then there were a few winters at around 60-days: 1888-1889, 1982-1983, 2007-2008.

Looking at the 10 Year averages in red, we can see that the first -25°C has crept up a little bit over the years, and now sits at mid-to-late December.


This histogram shows how things typically break down. November is most common for -20°C, while December is for -25°C. But the two months of November & December combined are where the vast majority of our cold days first hit: 90% of the time for the first -20°C and 79% for the first -25°C.

10-Year Average

Taking a closer look at the averages, on-average there's about a two week gap between the first -20°C and -25°C. That isn't really representative of what's actually happening, though.

As we saw at the start, almost half of the time -20°C and -25°C show up during the same coldsnap, and there's very little separation between them. But then in other years the gap might be two weeks or even two months. The average of all of that is 13 days, but in reality only about 10-out-of-138 have been close to that average.

Now we've looked at the first -20°C of each winter, and the first -25°C. There won't be a version of this for -30°C, because there have been 20 winters without a -30°C, and this type of chart doesn't handle gaps well. There is the threat of -30°C temperatures this week though, so in a few days we'll try to visualize that a different way.

2017/12/24

First -20°C of the Winter

Christmas Eve has brought us the first temperatures below -20°C for the winter of 2017-2018. Today is going to be a trivia day, as we look at when the Lows first drop below -20°C. (we also looked at this a bit last year in 'Tis the season of -20°C)

Date of the First -20°C

Here we have the recent history of the first -20°C Low of each winter.

This year we had 3 days at -18°C during November, including one way back on November 4th. But we didn't break the -20°C barrier until December 24, which makes 2017 the latest since 2002.

Recently mid-to-late November or early-December has been pretty common for the first -20°C, but in 1997-2002 we had a string of late years with the dates in December and even into January.

History

And here we have the entire history, going back to 1880.

November has been pretty popular all along, although there have always been some years that stretched into December, and even 7 that waited until January. In the 1920s-1940s there were also half-a-dozen years years where -20°C showed up in late-October.

The red line that appears in this animation is the 10-year average date.

10-Year Average

Looking at the 10-year average, right now we actually average about the same date that we did 120 years ago.

The average dipped into mid-November in the late-1920s and early-1930s, with a string of really early years. And in the early-2000s it shifted into mid-December, after the 1997-2002 stretch of late years. But 2010-2014 were all between November 13-19, and that's pretty early even compared to history, and so our average is sitting just a few days behind where it was in 1890.

We shouldn't necessarily read too much into that, though. At the top of this post I said that today was about trivia, and that's just because -20°C is an arbitrary cutoff. I like to use -20°C as a reference, but we got really close to it almost two months ago now. The fact that -20°C didn't show up in November is as much about luck as anything, and that could apply to any of the years here to one degree or another.

Now does seem like a good time to mention that while the date that -20°C first arrives hasn't changed too much, the number of times that we hit -20°C has dropped:

Total Number of Cold Days

This is a chart I've used several times before - most recently in 2016-2017 Winter in Review - The Really Cold Days.

It shows how the number of really cold days each year has been falling since the 1880s. Right now we average about 23 Lows below -20°C each winter, about 7 below -25°C, and maybe 1 below -30°C. It's always important to keep in mind that Blatchford is not the International, though. The temperatures from the International Airport are often much scarier than what we see in the city, and we looked at that in detail in Versus - The Edmonton International Airport: Part 2.

Christmas Day

The forecast for Christmas Day this year is pretty chilly, but lets take a look back at history:

Temperature

Over that last 30 years the average temperatures for Christmas have been fairly warm, with a High of -4°C and a Low of -13°C.

Recently things have ranged from warms years like 2005 & 2011 with Highs above 6°C, to cold years like 2008 & 2012 with Lows down around -25°C. And during 2001 through 2007 all 7 of those Christmas Days had a High above freezing.

In the last few years 2015 and 2016 were both at the cold end of things, with Highs of -12.9°C and -14.4°C, and Lows of -20.1°C and -22.9°C respectively.

Precipitation

Here is the precipitation for December 24th & 25th. I've combined both days here, since Christmas Eve snow and Christmas Day snow are pretty interchangeable. We just looked at this chart in Low-snow (or brown) Christmases.

Last year we had quite a bit of snow, with 6.6cm at Blatchford and 7.5 at the International. 2014 & 2015 both also had skiffs of snow, but before that we have to go back to 2004 to find a snowy Christmas. And before that all the way back to 1992.

Updated for 2017

And here is where we ended up for 2017.

Our High of -19.3°C was just a bit warmer than the -20.6°C from 2012, but 2012's Low was "only" -26.3°C compared to -26.7°C. To find a colder Low we need to go back to 1996, which had a High of -16.2°C but a Low way down at of -29.5°C.

2017/12/20

Low-snow (or brown) Christmases

Right now it's less than a week before Christmas, and there isn't much snow on the ground, and so people have been asking about "brown" Christmases.

Global just released this forecast for Christmas day, which is entertainingly specific:
Source: White Christmas Expected In Most of Canada

The typical definition of a "White" Christmas seems to be 1" (2.5cm) or more snow on the ground. And so therefore it follows that a "Brown" Christmas isn't necessarily completely snow-free, but just low-snow.

Is 2017 low-snow in Edmonton?


Edmonton has two main weather stations - Blatchford, and the International Airport.

Following the big December heatwave there's a bit of a discrepancy between the two stations, with 1cm of snow on the ground measured at Blatchford, and 8cm at the International. So as of December 19th in the city we're definitely near zero, but at the International things are just a bit below average.

History of Low-Snow Christmases

This chart shows the history of snowdepth on Christmas day at both Blatchford and at the International. Years that had 3cm or less are highlighted.

Both stations recorded low-snow Christmases in 1987, 1988, 1997, 1999, 2002 and 2005. Blatchford also recorded a few extra in 1959, 1962, 1963, 1988 and 1993, while the International recorded an extra one in 1985.

There is a caveat with the data today, because sometimes Environment Canada will show blanks, and it's not always clear if that is meant to be a 0 or if  data is missing. I've tried to use a consistent approach to clean those up, and looking at the records for these two stations we can see that they usually track each other pretty closely.

The current gap of 7cm between the two stations isn't unusual - there are some gaps in here as large as 10cm or 16cm - but it is unusual for the low-snow years. 1985 is the only similar one, which had the International at 1cm while Blatchford was at 8cm. But for all the other low-snow years here the two stations are separated by at most 2cm of snow.

Fresh Snow (or Rain)

Here we have the amount of fresh snow recorded on December 24th or 25th. There have also been 3 years with rain - 1929 at 0.9mm, 1975 at 0.3mm, and more notably 1989 at 2.4mm.

In 2016 we received 6.6cm of snow at Blatchford and 7.5cm at the International, and that made it a pretty snowy Christmas. That's been uncommon recently though, with 2004 being the last notable Christmas Eve or Christmas Day snow, and before that we have to go back to 1992.

What happens after Christmas?

This chart shows how snowdepth changes over the course of a winter.

It starts by showing all of the winters since 1995 for some context, and then it focuses on the recent low-snow winters: 1997-1998, 1999-2000, 2002-2003 and 2005-2006. 2015-2016 and 2016-2017 are also included here because they're fresh in our memory, and they were relatively low too.

In 1997-1998 the snow took off right after Christmas, hitting 20cm by January 5th. 1999-2000 and 2002-2003 were both a bit slower, but both were around 20cm by January 20th. And 2005-2006 stayed low through January and February, before jumping up to 27cm on March 3rd.

Summary

At this point it's been more than a decade since we've had a low-snow or brown Christmas.

But in 61 years of recordings at Blatchford there have been 11 low-snow Christmases (18% of the time), and in 55 years at the International there were 7 (13% of the time). And the peak era for low-snow Christmases was 1997-2005, with 4-out-of-9 years recording 1cm-or-less of snow at both stations.

2017/12/18

Warm Decembers

November of 2017 was pretty cold, and so a few weeks ago I asked the question: Does a cold November mean a cold Winter? And the answer appeared to be "not really." We looked at the winters since 1980 which had cold Novembers, and about half of them were warm and half were cold.

But since late-November the temperatures have flipped, and December has been very warm. I'm a pessimist, and so now I'm worried that a warm December means that winter is really going to wallop us in January and February.

Today we ask the question: does a warm December mean a cold winter?

December 2017 at the mid-point

This chart shows how warm or cold a month is, relative to the 20th century average. The calculation is very simple - it takes the average temperature for all of the days in a month, and compares that to the average for that month for 1901-2000. We first looked at it in How Warm is 2017 - the Months, and then again in Cold Novembers.

We're only halfway through the month and so this chart isn't really fair, but right now December 2017's average temperature is more than 10°C above the 20th century average. There is still plenty of time for that number to drop before the end of the month, though. Assuming that the rest of the month is just average then that number would fall to about 5°C above the 20th century average.

So for right now we shouldn't make a big deal of the number, but we do know that December 2017 has been very warm, and that even if we do get a coldsnap it will still have been a warm month.

Other Warm Decembers Since 1980

This chart shows 10 other Decembers since 1980 that were 4°C or more warmer than the 20th century average.

The years here are 1985, 1986, 1987, 1993, 1997, 1999, 2002, 2005, 2006 & 2011. They range from 2005 at 4.4°C above the 20th century average, up to 1999 at 7.8°C and 1997 at 7.9°C.

We'll break these up by decade to take a closer look.

2000s & 2010s

Since 2000 we've had 4 other particularly warm Decembers, and they are shown here.

The main reference point of these charts - the "0" line - is to the 20th century average. But the dotted line shows the 5-year average, and that is maybe a more relevant comparison because it shows how these months compared to their neighbours.

  • 2002-2003: January-May were all a little bit cooler than the recent average, but nothing extreme.

  • 2005-2006: A very warm January, a cool~ish March, and the rest were pretty typical.

  • 2006-2007: Another warm January, cool February, warm March, cool April.

  • 2011-2012: Really warm across the board.

1990s

This chart shows the 3 winters with warm Decembers from the 1990s:

  • 1993-1994: Of the 10 winters here, this is the only one that took revenge for its warm December, with a cold January and a really cold February.

  • 1997-1998: Cold January but warm after that.

  • 1999-2000: January was a little cool, and everything else was pretty typical.

1980s

And finally, the 1980's had a string of warm Decembers in 1985, 1986 and 1987.

  • 1985-1986: Really warm January, cool February, warm March.

  • 1986-1987: Really warm January and February.

  • 1987-1988: Pretty warm across the board.

Summary

When we get unseasonably warm winter weather in Edmonton it's natural to worry that we will pay for it again later, and that winter will have its revenge.

But of the 10 recent, warm Decembers that we've looked at here, the only one that was followed by a particularly cold winter was 1993-1994. For the rest of the winters there were 5 that were genuinely warm (1985-1986, 1986-1987, 1987-1988, 2005-2006, 2011-2012), 2002-2003 was just a bit on the cool side, and the others were pretty average for circa 1980-2016.

And in these 10 winters there was only one "revenge" month, with February 1994 at -7°C below the 20th century average. A few of the other months here were a bit on the cool side, but there was nothing else extreme.

When we'd looked at cold Novembers we saw that about half of the time the rest of the winter was cold, and for the other half it was warm. I'd sort of expected to see the same even split today, but at least based on these recent examples it's pretty rare for a warm December to be followed by a cold winter.

I've said before that this blog is not about predictions, and I will not claim to know what 2018 has in store for us. This is winter, and this is Edmonton, and so undoubtedly the temperatures will cool off, and undoubtedly there will be some coldsnaps in our future. But if you're worried that winter will demand payback for this warm December, you probably don't need to worry.

2017/12/16

Mid-December Update

We don't normally do mid-month updates around here, but this December feels like it's worthy of some discussion.

50 Warmest Days for December 7-13

December has all been warm so far, but the peak of the heatwave was really from the 7th through the 13th.

The forecasts had made it sound like we might break some all-time records, but in the end that didn't happen. Notably, Edmonton's longest-standing record of 10.6°C from December 12, 1880 remained unbroken with 2017 coming in just a touch cooler at 10.0°C.

We did have 4 days in the Top-50 though, which is the most for any year. And that brings the 2010's to a total of  8 days in the Top-50. That's still below the 10 days during the 1890's, but the 2010's still have two more years to go.

High Temperatures

The heatwave actually started on November 23rd, so right now it's been going for just over 3 weeks. We haven't set any all-time records, but so far in December we've set 3 recent records, going back to 1996.


I often like to stress that warm temperatures during Edmonton's winters are not unheard of, and in any "typical" December we'll have at least a week of Highs above freezing. There are some years like 2010 and 2009, with 0 and 1 day respectively, but there are other years like 2011 and 1999 which had 19 and 22.

At the mid-way point of December 2017 we've had 15 days above freezing and 8 days above 5°C, and that's definitely on the warm side of things. We're well above the average, and in roughly the same range as the monthly totals for 2006, 2005 and 2004. We're still below 2011 and 1999 though, but with two weeks left in the month who knows what the weather has in store for us?

Days in a Row Above Freezing

On December our High was -3.7°C, but since then every day has had a High above freezing. As of December 16th that makes a string of 13 days, and from the forecast it looks like that might continue a bit longer.

This chart shows other long stretches of December days above freezing:

  • 1898: December 15-28, 14 days
  • 1939: December 1-20, 20 days (although that streak actually started in November and lasted a total of 41 days)
  • 1954: December 10-22, 13 days
  • 1985: December 13-26, 14 days
  • 2002: December 5-18, 14 days

We're still a week from the all-time record of 20 days in 1939, but otherwise we're right near the top of the list.

Low Temperatures

Low temperatures aren't as much fun as Highs, but we have set 3 recent-records for warmest-Lows


Even with the warm temperatures this year, so far we haven't had any overnight Lows above freezing. Some of the other Decembers here - 1999, 2004, 2005 - had a few of those.

Snowdepth

The warm temperatures have brought some significant melting, and on the 14th Blatchford dropped to "trace" amounts of snow on the ground, and then back up to 1cm on the 15th. But that's down from 10cm on December 5th, and from 12cm when the heatwave started back on November 23rd.

The International has seen some melting too, falling from 20cm on November 23rd down to 9cm on December 13th and staying level for the 14th and 15th.

2017/12/06

Record Watch: Mid-December

The forecast is calling for some very warm days in the next week, and there's a possibility that we might even break some records. Today we're going to look back at other warm temperatures from the week of December 7th through 13th.

50 Warmest Days

Here are the 50 warmest days for December 7-13, going back to 1880. In this top 50 we have temperatures ranging from the extremely warm December 10, 1913 at 15.6°C, down to seven days which hit 6.7°C.

It's maybe a bit surprising to see 2014 and 2015 in here, but they both had mini-heatwaves (or "don't-call-them-chinooks") at this point in December. 2014 even set an all-time record of 12°C on December 9th. The grey shading in this chart groups things by decade, and so far in the 2010s we have had 4 days in the top 50, but the real standout is the 1890s with 10 days.

We took an in-depth look at all-time records earlier this year in How Often Does Edmonton Break Temperature Records and Broken Records - Part 2. There are a few notable records for this week of December:

Recent History

Here we have December High temperatures for the last 20 years. From the colours we can see that High temperatures throughout December jump around a lot. It's pretty typical for December to have a cold week, a warm week, and then a bunch of days in the middle.

At this point last winter we had just plunged into a deepfreeze, which certainly happens sometimes. 2016 had one High down at -20°C, 2013 had one at -18°C, and 2009 had one High of -28°C. We talked about December 13, 2009 a bit here, but it set all-time records for coldest-High at -28.3°C, and coldest-Low at -36.7°C, and the International dropped to -46°C.

Thinking warmer thoughts, in both 2015 and 2014 this week of December had some days up at 7°C and 12°C. And that isn't particularly rare, with 2008, 2006, 2005, 2002 and 1998 all having strings of warm days. The only years here that had no days above freezing during this week were 2016, 2013, 2010, 2009 and 2003.

Finally for today, we'll end off with a reminder of how often Edmonton breaks temperature records:


December 12 Update

The initial forecasts for this week hadn't made it sound like the December 12, 1880 record of 10.6°C was at risk.

But now here we are on December 12th, and the forecast for today is 10°C. So we'll have to see how things go.


After this whole warmspell is over we'll do a full recap for some context.

2017/12/04

Does a cold November mean a cold Winter?

We just had a fairly cold November, which made me wonder if now we need to prepare for a similarly cold winter?

On this blog we do not try to forecast or predict weather. But today we are going to look back at some recent cold Novembers, just to see how the rest of those winters went.

Winter 2017-2018

This chart shows how warm or cold a month is, relative to the 20th century average. The calculation is very simple - it takes the average temperature for all of the days in a month, and compares that to the average for that month for 1901-2000. From this chart we can't tell why a month was cold - if it was just one extreme coldsnap, or if it was cold throughout - we just know if it was above or below average.

We first looked at this chart in How Warm is 2017 - the Months, although this time it's been flipped around to start in July and it has New Year's as the midpoint. Since we're going to be looking at winters this approach will make more sense.

In 2017 July, August and September all averaged about 2°C warmer than the 20th century. October was flat, and then November was -3.9°C colder than the 20th century average. The entire 20th century is a pretty broad reference though, and so the dotted line here is the average of just the last 5 years to provide some recent context.

In the last 5 years our Novembers have actually averaged about 1°C colder than the 20th century, but then the January's and February's have been 4-6°C warmer.

Other Cold Novembers Since 1980

This chart cycles through the 12 other notably cold Novembers that have happened since 1980: 1982, 1984, 1985, 1986, 1990, 1995, 1996, 2003, 2006, 2012, 2013, 2014, and then 2017.

These years range from a small difference of -2.3°C below the 20th century average for 2013, down to a big difference of -8.9°C for 1985. Of these Novembers, the ones that were colder than 2017 were 1984 at -4.5°C, 1985 at -8.9°C, 1996 at -6.6°C, and 2006 at -5.6°C.

As this chart cycles through the years one thing to notice is that they aren't all a sea of below-average blue, and there are quite a few warm months in there. Lets take a closer look, by decade:


The 2010s: 2012, 2013, 2014

Recently we had a string of cold Novembers, in 2012, 2013 & 2014.
  • 2012-2013: warm January and February, but the rest were cold
  • 2013-2014: warm January, but the rest were cold
  • 2014-2015: warm across the board, except for that cold November.
One problem with a cold November is that an early start to winter can make the winter seem extra long. And the winters of 2012-2013 & 2013-2014 not only had cold Novembers, but also cold Marches and Aprils. I don't remember either of them specifically, but those years would have been pushing up against the dreaded "6 months of winter."


The 1990s & 2000s: 1990, 1995, 1996, 2003, 2006

The 1990s and 2000s didn't have very many cold Novembers, but there were a few notable ones here:
  • 1990-1991: cold November & December, but a warm January and a very warm February
  • 1995-1996: pretty cold across the board. 1996 was the coldest year in Edmonton since the 1950's, possibly due to some volcano eruptions?
  • 1996-1997: the second half of 1996, and again pretty cold. February 1997 was warm, though.
  • 2003-2004: generally warm, with November as the only cold outlier.
  • 2006-2007: cold November followed by a warm December & January.

The 1980s: 1982, 1984, 1985, 1986

And the 1980s had a group of cold Novembers all in quick succession, in 1982, 1984, 1985 & 1986.
  • 1982-1983: generally warm, with November as a cold outlier.
  • 1984-1985: cold November & December, but warm January & March
  • 1985-1986: really cold November, but really warm December, January & March
  • 1986-1987: really warm January, February & March.

Summary

So does the cold November of 2017 mean that we need to brace for an extra-cold winter this year? The good news is that there doesn't seem to be a connection.

Of the 12 winters that we've looked at here, 7 of them were warmer during December-April than the average for 1980-through-2016. The 5 that were colder were 1984-1985, 1995-1996, 1996-1997, 2012-2013 and 2013-2014 (although unfortunately those last 2 were very recent). 7-out-of-12 warm versus 5-out-of-12 cold is basically a flip of the coin.

And all of the winters here (except 1995-1996) had at least one genuinely warm month after November, and most winters had a few of them. "Warm" is relative of course, and it really means "warm for Edmonton", but 4°C or 6°C above average at least takes the edge off of things.

We shouldn't treat this as a prediction, but just because we had a cold November doesn't mean that we need to worry that the rest of the winter will follow suit.