I am not a gardener, and the biggest clue that I am not a gardener is the fact that it has taken more than 2 years for this blog to get around to the subject of frost-free days. They're clearly an important part of Edmonton weather, but they're also totally in my blindspot. We're going to try to fix that oversight today, but with the warning that because I don't garden I don't necessarily know how all this stuff works.
Last Frosts Each Spring
We're going to start by looking at the last frosts that occur every spring.
"Firsts" and "Lasts" of anything are always a bit variable, and so this chart shows the last 5 frosts for each year, going back to 1881. The number 5 has no special significance, but there are some years with a really late, outlier frost and this gives a sense of how big the gap is between those and the rest of winter.
The typical measurement used for frost-free days is 30-Year Averages, and the orange lines in this chart show how that has changed over time.
- At Blatchford right now the average date for the final frost is May 7th, but a century ago it was almost 3 weeks later on May 27th.
- The 2nd frost is on-average 5 days before the last frost, and is on May 2nd. A a century ago that gap averaged 10 days.
- April 24th is the average date of our 5th-last frost, which is about 2 weeks before the final frost. A century ago the average date of the 5th-last frost was May 7th, and that lines up with where our final frosts are today.
One thing to keep in mind with these averages is that they only mark the midpoint - half of the time the last frost will occur before the average, but the other half of the time it will occur after. Looking at some recent years, 2012 had one of the earlier final-frosts on April 18th. On the other hand, in this chart for both 2009 and 2004 we can see that the dark-blue lines cross the average, and that's because 2009 had 6 frosts after May 7th and 2004 had 5.
For a bit more trivia, the year with the biggest gap between that 2nd-last and the last frosts was 1915, with 33 days between a frost on May 12 and the final frost on June 14. Looking at more recent years: 1997 had a gap of 21 days from April 30th to the final frost on May 21th; and 2000 went 19 days from April 18th to May 7th.
We'll talk a bit about the probability of avoiding an extra-late last frost later on.
First Frosts Each Fall
Moving to the other end of the summer, this chart shows the first 5 frosts for each fall.
Looking at the 30-Year averages again:
- Right now the average date for the first frost each fall is September 24th. That's 18 days later than the September 6th date from a century ago. That shift is also similar to the 20 days that we saw when we looked at the last frost for each spring.
- On average the 2nd frost occurs 5 days after the first, and then by October 10th we've averaged 5 frosts. Both of those dates have also shifted by about 20 days over the last century.
In this chart 1992 stands out with a really early frost on August 22nd, and it actually had a total of 7 frosts before the average date. To find another year with a frost that was before September we need to go all the way back to 1934. On the other end of things 2011 was a really late year, with the first frost waiting until October 15th. The latest first frost was in 1975, on October 21st.
Frost-Free Days
When we combine the last spring frosts with the first fall frosts this is where we end up for the year.
(for clarity there's the fancy, animated chart, as well as a non-fancy version)
Over the last century the frosts have moved by about 18~20 days at both the spring & fall ends of the summer, and so we've gained more than a month of frost-free days. That big shift happened between the 1940's and 1980's though, and since then things have been fairly constant.
In this chart we've added probabilities with the dashed, orange lines which are labeled "25% Chance" and "10% Chance." The 30-Year average line is the point where half of the first/last frosts will have occurred, but that means that 50% haven't occurred yet, and that is still pretty risky. The 25% and 10% lines add a safety factor to help avoid frosts that are early/late. They're calculated using the standard deviations for that last 30 years.
In the spring the 25% Chance line occurs 8 days after the average, and in the fall it's 7 days before. So if you use that as your guide then that shortens the season by 2 weeks, although for some years it will be more and for others it will be less. In the years since 2000, if you had followed the 25% line you would have been caught by late spring frosts 3 times: in 2015, 2009 (there were 5 late frosts that year) and 2002. And in the fall there were 2 years with early frosts: 2014 and 2007.
If a 25% chance of frost is too risky, then there's also the 10% line. That shortens the season by another 2 weeks compared to the 25% line, or by about a month compared to the average. If you had followed the 10% line since 2000 you would have been caught by one late spring frost, on May 23 2002. And the last time that a fall frost occurred before the 10% line was back in 1992.
The choice of the average, the 25% line or the 10% line is all about the risk of late frosts, and since I'm not a gardener I can't give any advice on that. But right now if we use the 10% risk as our criteria the frost-free season lasts from May 21st to September 10th, while a century ago it only lasted from June 18th through August 22nd.
Total Number of Frost-Free Days Each Year
In the previous charts today we've looked at the frosts in the spring and in the fall, and here we're looking at the gap between those two to measure the total length of Edmonton's frost-free season.
Right now the average length of the frost-free season is 140 days, up from 102 days a century ago. Some recent years like 2011 & 2012 were up at 174 and 169 days, while others like 1992 and 1989 were down around 92 and 104. The year with the shortest frost-free period was 1884 with 44 days, and the longest was in 1980 with 185.
This chart also shows the length of the season based on the 25% and 10% Chances of frost. As we saw earlier, using 25% is about 2 weeks shorter than the average, and 10% is about a month shorter. Looking at recent years, the only two which fell below the 111 days for a 10% chance were 1992 and 1989.
If you are more adventurous this chart also shows the average gaps between the 2nd frosts, and the 5th frosts. Risking the 2nd frost would add 10 days to the season, and stretching things all way to the 5th would add more than a month.
One thing to notice in this chart is the cluster of really high frost-free days that occurred in the late-1970s, with 7 years in a row all up above 160 days. Do all those long frost-free seasons mean that that was a particularly warm time period for Edmonton? Let's take a look:
Frost-Free Days and Edmonton's Warmest & Coldest Years
The background of this chart - the blue & orange bars - is one that we've used before, most recently in 2017 in Review - Temperatures. We won't go into all the details again today, but the short version is that it compares the overall temperature for each year to the 20th century average, with the warm years shown in orange while the cold years are in blue. Overlaid on top of that we've added the number of frost-free days for each year as a red dot.
Eye-balling off the chart it looks like there is a bit of a relationship between the number of frost-free days and how warm or cold each year was. For the consistently colder years of 1880-1940 the frost-free days were usually down between 80-120. And for the warmer years of 1970-today the range is more typically 120-160.
They're not a perfect match though, and the best demonstration of that is the years 1980 and 1981. As we saw earlier, 1980 had the longest period of frost-free days at 181, but it wasn't a particularly warm year coming in at 0.8°C above the 20th Century average. On the other hand, 1981 was Edmonton's warmest year at 3°C above average, but it's frost-free season was only a little bit high at 159 days. Earlier we'd also talked about 1992 which had a very low 92 frost-free days, but it was actually a fairly warm year at 1.7°C above average.
To really dig into this here's another chart:
This chart is just plain nerdery. It plots the number of Frost-Free Days each year against how warm or cold the year was. It doesn't have any real, physical meaning, but it lets us compare the relationship between the two pieces of data.
If frost-free days and warmness were a perfect match then all of these values would align along a single line, but there's some variability and so instead they form a cloud. There's definitely a relationship between the two, but the correlation is 0.21 so it's not particularly strong.
If you watch the dotted best-fit line that is in the middle of the cloud you can see that as time passes it has rotated counter-clockwise a bit, which just reflects the relationship between yearly temperatures and the frost-free days. As we've had more warm years recently the increase in the length of the frost-free season hasn't quite kept pace, and that has caused the line to tilt up a bit more.
Growing Season?
Finally, in doing research for today I found some definitions of "Growing Season" which were distinct from frost-free days. There were actually a few different definitions, and they didn't all agree, and so in this chart we are using the one from Alberta Agriculture and Forestry. They've defined it as:
The growing season period is based on an average daily temperature of at least 5°C.So the green lines in this chart show the 30-year average date on which our mean daily temperatures (High & Low) reaches 5°C. And through a fluke of averages, and the fact that our shoulder-seasons haven't changed much over the years, it turns out that the growing season today is about the same as it was a century ago. Using this method the season starts on April 17th, and it ends on October 14th. That works out to a total of 180 days (and that matches Alberta Agriculture and Forestry's number, which is a relief).
I'm not sure what this growing season measurement is useful for - I think that it applies to trees and perennials which don't care about frosts and which are here for the long haul? But since it is quite different from frost-free days I thought it was worth a quick mention.
So that's it for today. Because I don't garden I may have made some mistakes, or maybe I missed some useful information? If that's the case please leave a note in the comments, or you can reach me on twitter @yegwxnerdery.
Next week we'll be back with our standard April month-in-review.
Then in 2 weeks time we'll have Frost-Free Days: Part 2 - versus Round. We'll compare Blatchford to some of its neighbours, and we also might take a look at climate hardiness zones, although that might wait until a Part 3.
No comments:
Post a Comment