2017/02/06

Snowstorms

I've had this one written for awhile, but I like to be topical and was really hoping to post it during a snowfall warning or a nice, 3-day snowstorm. But since October we haven't had much in the way of snowpocalypses.

We might get a bit of snow today, but southern Alberta is getting a true snowpocalypse right now. And that's as good an excuse as any to talk about snowstorms. This is Part 5 of my recent(ish) look at Edmonton snowfall:

The first question is: when it snows, how much does Edmonton typically get? 1cm? 10cm? More?

Typical Snowfall Amounts

This shows the breakdown of how much snow we get in a typical day's snowfall. I'm including data from Blatchford and the International here, because Blatchford data goes all the way back to 1880 but stops in 2007, and the International data is current but it only goes back to 1961.

For both Blatchford and the International, almost half of the time snowfall will be less than 1cm. About 85% of the time it will be less than 5cm, and 95% of the time less than 10cm. And days above 30cm are very rare.

Looking at daily snowfall is a bit problematic though, because snow will often start in the evening and continue on through the morning. Environment Canada's numbers would split that across the two days, but I'd rather look at that as a total. So for the rest of today I'm going to look at snowfall "Events" and try to combine the totals when there were several days in a row with snowfall.

I'm going to define a Snowfall Event as:

  • the total snow on consecutive days where each day received at least an inch of snow. 

I've put that 1" cutoff in there, because I don't want to combine a bunch of days which only received a light dusting. 1" is completely arbitrary, but using 1cm as the cutoff seemed too low, and setting it at 2.5cm moves things beyond the 0-2cm range which makes up about 65% of snowfalls.

So with all of that said, here is the chart revised for these multi-day snowfall Events:

Typical Multi-day Snowfall Events

It doesn't look much different, but things have shifted around a bit. In particular, for snowfall of 10cm or more we're now seeing slightly higher numbers. What's happening is that some of the 2-5cm and 5-10cm days are now combined to give larger multi-day totals.

For a little bit more explanation of what grouping snow into these Events actually does:

Here we have hypothetical examples of two different 3-day sequences - in both cases the total snow for the 3 days is 20cm.

Example 1 is counted as 3 small snowfalls, because the second day only received 1cm of snow, which is below the 2.5cm cutoff.

Example 2 is grouped together as one big event, because all three days cross the 2.5cm threshold.


This is just an approximation that I've made up, and it's certainly not perfect or scientific. But for trying to count snowstorms I think it makes sense to group snowfall into a multi-day total like this, rather than just looking at the 24-hour amounts.

So what qualifies as a lot of snow? Personally, I think 10cm is a nice starting point. With 10cm of fresh snow on the ground walking and biking gets challenging, and I'm sure that the Quesnell bridge will be a mess.

More than 10cm (4") of Snow

Here we have the count of snowfall events for each year that are greater than 10cm.

I haven't included 2016-2017 in any of the charts today because the winter isn't over yet. But for the record, back in October the International got 11.5cm from the 7th through the 9th, and then 9.5cm on the 14th. And those are the only really notable storms so far.

This also isn't perfect, because it shows 2015-2016 with no 10cm events, while I'm pretty sure that the downtown had one major snowstorm. For November 23 & 24, 2015 the International recorded 9.4cm of snowfall, but Environment Canada also has a station at the UofA which recorded 15cm. I think that's a better representation of what actually hit the city, but I'm not going make things even more complicated by mixing in the data from a 3rd station. For today we'll just have to accept a bit of imprecision.

Generally, we can see that in most winters there are from 2-to-5 10cm snowfall events. Some winters have 6-or-7, and some have 0-or-1, but 2-to-5 is pretty typical. And outside of a few really snowy years (1903-1904 and 1906-1907) that hasn't changed much in the last century.

When are these 10cm+ snowfalls most likely to happen?


Of the roughly 2-to-5 snowfalls greater than 10cm in a typical winter, here is the likelihood of them occurring in each month.

If I had been forced to guess, I would have said that November gets a disproportionate number of big snowfalls. Looking at the numbers though, things are spread fairly evenly. Of the real winter months January is the highest and February is the lowest, but none are really a huge standout.

For a little bit more context, for the rest of today we're going to include Montreal as a comparison.


Here we have the same count of yearly 10cm snowfall events, but with Montreal added on top in purple. Central Canada and the northeastern US gets much larger yearly snowfall totals than we do, and I wanted to see how things compare.

Where Edmonton typically gets 10cm snowfalls 2-to-5 times per winter, Montreal is more in the range of 4-to-7. And their historic extremes are between 15 and 18 times per year.

(if it looks like Montreal's numbers are trending downward, that is correct. Montreal's overall snowfall has fallen by about 50% over the last century. But that is a story for another time...)

For a little bit more detail, here we'll look at recent history:


This is the same chart as above, but now only going back to 1980. (Montreal's data ends in 2013, so the last few years might look low on this chart, but that is because they are actually missing).

Hopefully this makes it a bit easier to see what is going on. Blatchford and the International are pretty similar, although it looks like Blatchford typically got a few more snowy days (before the data ended in 2007). And Montreal generally gets about twice as many 10cm snowfalls as the Edmonton stations (before that data ended).

10cm is pretty snowy, but it's not really snowy. Next we'll look at snowfall events of more than 20cm.

More than 20cm (8") of Snow

Here is the same basic chart, but now we're looking at snowfall events of 20cm or greater.

The numbers have dropped off, with Edmonton now typically in the range of 0 or 1 events each year. And going back a century it hasn't changed much - the average has fluctuated around a 60% probability each year since the records started. Montreal is much higher, frequently seeing 3 or more 20cm snowfalls, although there are a few years that have only seen 1.

Now we'll focus on recent history again:


Here we have a closer look at the 20cm snowfalls.

The standout is 1992-1993 with 3 20cm snowfalls at Blatchford, but that whole 1990-1994 period had at least one 20cm event each year. More typically though they happen about every-other-year. And just in the last 10 years Edmonton's numbers have dropped, so that we've only seen a 20cm snowfall every 3 or 4 years.

Montreal's numbers bounce around quite a bit, but they've almost always had at least 1 20cm snowfall, and 2007 had 7. And then from 2013-2014 there are no records.

Finally we'll up the numbers one more time and take a look at 30cm snowfalls.

More than 30cm (1') of Snow

At least according to Environment Canada's measurements, 30cm snowfalls are very rare in Edmonton - there have only been about 20 of them recorded going back to the 1880s. There was a period in the 1940s where they happened almost every-other-year, and there were two in the winter of 1984-1985, but other than that things have been much more spread out.

The most recent was 35.2cm on May 5th & 6th 2003, and before that 40.6cm on April 6th & 7th 1991.

Montreal also doesn't typically get a huge number of 30cm snowfalls, but they are more frequent and occur roughly every-other-year. And 2007-2008 was a really snow year, with 3 of them.

I chose to include Montreal in these charts today for a few reasons:
  1. I was curious how central Canada compares to Alberta.

  2. Adding Montreal actually makes the 20cm and 30cm charts easier to interpret - they don't look very good with Edmonton data alone since it just bounces between 0 and 1. 

  3. Having a second city gives a bit more context. I know that the absolute numbers today - 10cm, 20cm, 30cm - are a little questionable. As I said, last November there was one day downtown that definitely had more than 10cm, but Environment Canada didn't record it that way at the airport. And in the last 5 years the media have reported 30cm at least a few times, but again those don't show up here. With Montreal here as a comparison, we can see that whichever exact "10cm" or "20cm" we're talking about, Montreal obviously has more of those than Edmonton.
For that last point then, lets go back to the beginning and look at the breakdown of typical snowfalls - but this time with Montreal added to the mix:

Typical Snowfall Events redux

And here we can see that Montreal's distribution is a little different than Edmonton's.

In particular, about 7% of their snowfalls are 20cm or more, compared to 1-2% of the time for Edmonton. And 18% of their snowfalls will be 10cm or more, compared to about 8% for Edmonton. Montreal has a shorter snow-season than Edmonton does (which we saw here), but when they get snow they are more likely to get more of it.

Finally for today, just for the sake of trivia lets look at the biggest snowfall events for Edmonton and Montreal:

Largest Snowfall Events

There's a lot of information here, but this is basically a scatterplot of the largest recorded snowfalls for Edmonton and Montreal. It shows the Top-5 largest snowfalls for each city, and I've broken things down into the modern era from 2000-onwards, as well as the Top-5 ever recorded.

Because of the way that I've grouped snowfall into events, some of these totals were just one snowy day, some were 2 or 3, and there's one which is most of a week. 25cm spread over January 17-22 is maybe stretching the idea of a "snowfall event" a little bit, but the rest are okay. And I'm still happier using these multi-day groupings rather than individual 24-hour periods.

Historically, Montreal's 5 largest multi-day snowfalls were in the range of 57cm to 68cm. But since 2000 the largest was 50cm in 2000, with the rest between 36cm and 46cm.

Edmonton largest historic snowfalls are between 42cm and 47cm. And the recent Top-5 ranges from 21cm up to 35cm in 2003.

Ultimately though, what all of this means for giant snowfalls in Edmonton is:
  • 10cm (4"): usually about 2-5 times per year
  • 20cm (8"): maybe every-other-year
  • 30cm (1'): none in the last decade, 3 in the last 30 years.
And as I said earlier, for the winter of 2016-2017 the two big snowfalls were back in October: 11.5cm on the 7th through the 9th, and then 9.5cm on the 14th.

No comments:

Post a Comment