Study: weather patterns affect frost timing more than global warming


Weather patterns’ influence on frost timing

Air circulation affects frost more than global warming — for now

May 23, 2017– Gardeners know the frustration of a false spring. Coaxed outside by warm weather, some people plant their gardens in the spring only to see a sudden late frost strike at the plants with a killer freezer burn. Grumbling green thumbs, along with farmers and water supply managers, would benefit from more accurate predictions of the first and last frosts of the season.

This image shows the average day of the year with the last spring frost (left) and first fall frost (right). Cooler colors indicate an earlier day of the year. CREDIT Courtesy of Court Strong/University of Utah.

Such timing is in flux, however. The frost-free season in North America is approximately 10 days longer now than it was a century ago. In a new study, published today in Nature Communications, researchers from the University of Utah and the U.S. Geological Survey parse the factors contributing to the timing of frost in the United States. Atmospheric circulation patterns, they found, were the dominant influence on frost timing, although the trend of globally warming temperatures played a part as well.

“The frost-free season has been lengthening over the past century, and now we understand the changes in atmospheric circulation that are extremely strong in frost timing, even stronger than global warming,” says University of Utah atmospheric sciences professor Court Strong.

Weather and climate are complex systems, with many factors affecting what the particular weather conditions might be in a certain place at a certain time. Previous research, says Gregory McCabe, of the USGS in Denver has focused on the role of large-scale phenomena like El Niño. “I don’t think anyone has broken it down to look at the circulations patterns specific to the timing of frost,” McCabe says.

Strong and McCabe set out to investigate the relative contributions of the global warming trend and local atmospheric circulation patterns to the century-long lengthening of the frost-free season.

“If you ask a U.S. forecaster what determines the first fall frost, they’ll say a cold air mass coming down out of Canada, clearly due to circulation,” Strong says. “There’s a role for warming, but on the other hand forecasters will tell you there’s clearly a role for circulation as well.”

To more accurately capture regional, relatively small-scale circulation patterns, Strong and McCabe divided the United States into four regions, and examined separately how frost timing patterns varied in each region over 93 years of weather data.

The researchers found that atmospheric circulation patterns accounted for between 25 and 48 percent of the variation in frost timing. To put that in context, Strong says, remember that the frost-free season has lengthened by an average of 10 days over the past century. Three to five of those days can be accounted for by atmospheric circulation, while three days can be chalked up to global warming. Other factors, such as local cloud cover, may account for the remaining two to four days.

Although the results show that atmospheric circulation is the primary driver of frost timing, the warming trend exerts an influence over circulation beyond the general trend of warming temperatures. “We also found evidence that these circulation patterns themselves have been altered by global warming, especially in the Western U.S. and the Northwest,” Strong says. “Warming is an important part of this narrative despite this finding that circulation is a stronger driver historically.”

Next, Strong and McCabe will evaluate how well climate models capture the drivers of frost timing and look for ways the models can be improved. Better modeling of atmospheric patterns leads to more accurate forecast of future frost timing. “The year-to year variability in climate is controlled by these changes in atmospheric circulation,” McCabe says. “On top of that you have the warming trend. If you don’t get these patterns right then the simulations are going to have a lot of uncertainty in them.”


Full disclosure, the lead scientist used to be a roommate of mine, he’s a sensible and affable guy and former television meteorologist who worked at the same TV station as I did.

The paper:

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
May 23, 2017 7:51 am

Each year is a variation on previous year’s climate. And it shifts suddenly and often with little warning, some winters are colder than usual and others are hotter than usual, we go from rain cycles to drought cycles. Each time this happens, it is a shock to people who want the weather to be like indoor heating/cooling systems.
Accepting this is part of being a real adult. And geology has given us all many reasons to worry about the next Ice Age. We still don’t know why Ice Ages happen like clockwork though many thinkers believe it has something to do with our sun/planet/galaxy relations with each other.

Reply to  emsnews
May 23, 2017 8:11 am

“The frost-free season in North America is approximately 10 days longer now than it was a century ago.”
The frost-free season in North America is approximately 10 days longer now than it was at the end of the LIA.
..UofU is f’in with us

May 23, 2017 7:53 am

although the trend of globally warming temperatures played a part as well.
oh please….call me when you can grow bananas

Rhoda R
Reply to  Latitude
May 24, 2017 1:49 pm

I’d settle for oranges and avacados.

May 23, 2017 8:01 am

This image shows the average day of the year with the last spring frost which is nonsense. At 6600′ in west central Colorado, it is stupid to plant before 150 days, last year it was 165 days. The scale only goes to 140 days.

Reply to  stevekeohane
May 23, 2017 8:53 am

Shh, there is no accounting for location/altitude. For the Denver/Front range area, normally, one shouldn’t plant before Mother’s Day. This year, depending on elevation along the Front Range, several locations had a hard freeze less than a week after Mother’s Day. Thankfully, only thing I had planted was potatoes.

Pop Piasa
Reply to  stevekeohane
May 23, 2017 8:31 pm

Their chart must be referring to urban heat islands and sunny springs. We never plant frost-sensitive plants until 2nd week of May here at the Illinois/Mississippi confluence. I guess if you average in years like 2012 with early springs you might come up with that optimistic of a chart.

May 23, 2017 8:07 am

The colder and further north the climate, the more dangerous it is to set a specific day as ‘safe to plant’ anything. We have to be very wary about when freezes happen. For example, it snowed on my mountain just over a week ago, in May, and we had exactly two warm days this month. It is cold now, went to 40 degrees F last night.

May 23, 2017 8:09 am

“Warming is an important part of this narrative despite this finding that circulation is a stronger driver historically.”
Finding? Are they claiming that they just now discovered that ‘weather’ was an important factor in determining the date of the first and last frost? I am pretty sure that most everyone who has graduated 4th grade was already aware of this fact. This seems more like an exercise in ‘writing for funds’, more than an exercise in science. Step 1: take a page from any ‘Meteorology 101’ text book. Step 2. Put a ‘warming’ spin on it. Step 3: collect money.
My favorite line was: “If you don’t get these patterns right then the simulations are going to have a lot of uncertainty in them.” Of course, if you DO get the patterns right, the simulations will still have a lot of uncertainty in them. They are simulations of a non-linear, chaotic, very large system. Uncertainty is inevitable.
I actually like the fact that he refers to the value of patterns in forecasting. ‘Pattern recognition’ was nearly lost when the idea that you could determine everything from initial conditions and a big enough computer came into fashion. This idea was one of those ‘availability cascades’ mentioned in a previous post; a belief that was never true or accurate, but became widely accepted as true because it was repeated over and over again as if it were true. Now we have a whole globe full of people who believe that climate models have significant value!

Reply to  jclarke341
May 24, 2017 4:59 am

+10 jclarke341
These weathermen have rediscovered the weather. Perhaps now other climate scientists will rediscover that climate is weather, which will remain unpredictable in the long term.
It’s totally true that using analytical math models for prediction became the dominant theory of today. Searchers in every realm of discovery are rediscovering patterns and the dominant way of discovering them seems to be training neural network programs to find them. Neural network programs are not any sort of analytical math, since there is no way to predict how a NN will weight various inputs for the final result.

Dave Irons
May 23, 2017 8:19 am

Here in Maine all gardeners know that planting anything before Memorial Day is risky at best. And it has snowed twice this month.

Reply to  Dave Irons
May 23, 2017 9:29 am

The humming birds showed up at their feeder this week, finally. They know best when it is warm enough.

Richard M
May 23, 2017 8:55 am

So, the changes due to the AMO, PDO, etc. which affect circulation patterns are a bigger factor than warming. In reality, they are also a big factor in the warming.

Reply to  Richard M
May 23, 2017 9:48 am

And in the COOLING cycles, too. Duh.

May 23, 2017 10:41 am

Tend to watch the lunar cycles more than anything else here north of Minnesota. If you get a full moon in the first ten days of June it’s risky to plant prior to the long weekend in May. If it comes later than that you’re fine to plant on that weekend. It’s a truism to be sure, but about 90% accurate.

John F. Hultquist
May 23, 2017 11:24 am

With elevation comes frost danger.
At 2,240 feet a clear night gets cold in a hurry.
Being rural, we do not experience urban heat effects.
Squash and tomato plants have been replaced once — so far.
Strawberry plants look great — so far.
[Can you not use plastic sheathing laid at night over the plants? Or are there too many acres to make that viable? Smog burners? (More CO2!) .mod]

Joel Snider
May 23, 2017 12:19 pm

I guess the question then is ‘how is 10 extra frost-free days going to cause global Armageddon?’
I’ll bet the answer comes in CGI.

tony mcleod
Reply to  Joel Snider
May 23, 2017 3:47 pm
Pop Piasa
Reply to  tony mcleod
May 24, 2017 12:00 pm

Quite the coincidence that population is greatest in places where it is always warm, isn’t it? Is that the case when you look at where the majority of farming takes place? Cold is the only viable threat to the survival of humanity. The food supply has only profited from the natural warming of this late interglacial. You are our “Debbie Downer troll” (banned from the Optimist Club).

Pop Piasa
Reply to  tony mcleod
May 24, 2017 3:16 pm

You know, if it got cold like the LIA again, there might quickly be only 500 million Indians to ponder.

Frederik Michiels
May 23, 2017 2:50 pm

is that why on 10 of july 2015 we still had ground frost after a heatwave?
in very exceptional dry conditions this can happen. note this only happened 7 times since records began here in 1833….

May 23, 2017 4:25 pm

From the article: “Three to five of those days can be accounted for by atmospheric circulation, while three days can be chalked up to global warming.”
Global warming? They said they studied the last 93 years of data, which means they started at 1924. Ok, the globe warmed from 1910 to 1940, then the globe cooled from 1940 to 1978, then the globe warmed from 1978 to the present. The warming from 1910 to 1940 was equivalent in magnitude to the warming from 1978 to the present.
The period from 1910 to 1940 is officially considered not to be affected by human-caused CO2 (IPCC). Why would you assume warming from 1978 to present *is* affected by human-caused CO2 (besides the obvious)? Natural processes could explain both periods of warming without the need to introduce CO2 into the equation.
And then there is the rising CO2 levels versus the temperatures which are *not* rising, to consider. The speculation says the more CO2 put into the atmosphere, the hotter it should get, but here we are putting more CO2 in the atmosphere, but it’s not getting hotter, and that’s been going on for almost 20 years now. So why assume CO2 is a driver of the Earth’s climate? I think you are assuming too much if you do.

May 23, 2017 7:07 pm

Weather goes in cycles, just like climate does. I’m still running my furnace. Did that last year until the thermometer reached and stayed at 65F overnight. I have pictures of snow on my lawn on March 30 this year, and same-same last year. Wanna see ’em?
It’s just weather. Is there any way to stop this obsession with climate change and not hurt people’s feelings?

May 23, 2017 8:17 pm

The researchers found that atmospheric circulation patterns accounted for between 25 and 48 percent of the variation in frost timing.
Acute, precise. Palm reading.

May 23, 2017 8:28 pm

So these researchers will forecast next year’s frost with error bars between 25 and 48 percent of timing variation.

Kevin Quitberg
May 23, 2017 8:37 pm

6″ of snow in my lawn last Thursday, 11″ at my daughter’s ranch house just 45 miles north of us. Bullet-proof ice on my windshield Monday morning, just simple global warming to scrape off this A.M. 40 degrees inside my greenhouse this morning but I think it is now safe to plant our tomatoes and peppers inside the greenhouse. Nothing outside until after the first week of June.
In 1985 we had a plant-killing frost in the garden on the 29th of June. It was followed by one that got everything the June frost didn’t on the 8th of August. We do have a short growing season here at 6200′ in western Wyoming. Heat radiates away pretty quickly. Did I mention the wind? A tourist once asked: “Does the wind always blow this way?” The short answer: “No, sometimes it blows the other way.”
During the winter of ’85 the water lines under the street to our house here in town froze. They were 9′ deep. Our next door neighbor’s lines ruptured and had to be excavated with a track hoe. I built a fire with railroad ties and lump coal to thaw the ground enough for the excavator to get through. I think a lot about global warming at those times. Although our sustained cold has not reappeared we do on occasion still get cold. At my work place about 20 miles west of town we had a -49 degree morning followed by -51 the next morning. It didn’t hang around much, however. It was a pretty brutal winter on the wildlife here in western Wyoming.

May 24, 2017 4:35 am

Here in the Texas Panhandle, each winter, on average, our risk of frost is from October 16 through April 13.
Almost certainly, however, we will receive frost from October 31 through March 30.
We are almost guaranteed that we will not get frost from April 28 through October 2.
The uncertainty in the ‘average’ dates is +/- 15 days or so.
With an error margin of about 30 days on each end, talking about freeze dates creeping by a day or two either way is pure foolishness.

%d bloggers like this:
Verified by MonsterInsights