Post #G21-011: Last frost date, degree-days, and why I’m planting summer squash now.

Posted on April 16, 2021

I planted my summer squash yesterday.  That’s more than a bit too early, by almost any reckoning.  In this post I explain why I did that.

Briefly, planting later is not unambiguously better, in this case.  On the one hand, I want to avoid the last frost at the start of the growing season.  That part, everybody gets.  On the other hand, I want to avoid certain pests at the end of my growing season.  Based on my experience last year, I’m going to pull up my summer squash when the squash vine borer appears, rather than expend a lot of effort trying to fight it.

The key fact is that the warmer the spring, the earlier various pests will appear.  So the clock is ticking.  A warm spring gives me the opportunity to get these plants into the ground sooner, but it also brings on the mid-summer pests sooner.

It’s not so much that I have a potential start date for the season, and I can start growing any time after that.  It’s more that I have a defined window for the season.  And if it is going to be a warm spring, the longer I wait past the true last frost date this year, the closer I get to the other end of the window.

Tomatoes and peppers and such, they can still inside for a while yet.  But for varieties where mid-summer emergence of pests limits my season, I’m going to take a gamble.

Last frost dates and weather forecasts

In Post #G21-005, I took a look at “last frost” dates.  These show the likelihood that you’ll have a spring frost following that date.  So, for example, as it is typically presented, your 70th percentile last frost date means that you have a 30 percent change of having frost past that date.

These dates are based on a simple tabulation of the past three decades of actual, observed last frosts.  If you were to take actual, after-the-fact date of the last frost for the past three decades, sort those from earliest to latest, then (e.g.) the one that was squarely in the middle of those dates is your “50 percent last frost date”.  Because, simply enough, over the past 30 years, half the time, the last frost date occurred after that point.

The point of my prior post is that the labels on those “last frost” dates are correct if and only if you blindly put your plants into the ground on that day.  They aren’t correct if you look at the weather forecast.  For the simple reason that nobody is dumb enough to plant their frost-sensitive plants on a specific date of the forecast calls for frost in the near future.

More to the point, all of that “last frost” methodology was developed in an era with limited ability to do weather forecasts.  Back in the day, when the best you could hope for was a reliable three-day forecast, the fact that you had a forecast hardly made a difference.

As computer technology has improved, the reliability of longer-range weather forecasts has increased.  E.g., the National Oceanographic and Atmopheric Administration says that, on average, the seven-day forecast is now correct 80 percent of the time (though what, exactly, that means, I have no clue, and that’s clearly going to vary from place to place and time to time.)

Depending on what the forecast actually says, even a somewhat-unreliable ten-day forecast can provide useful information.   The key is whether or not you’re on the borderline of a freeze.   If predicted temperatures in the 10-day forecast are nowhere near freezing, then that forecast would have to be grossly incorrect for a freeze to occur.  You might be willing to gamble on planting, even if the typical accuracy of a 10-day forecast isn’t all that good, because it would take an extreme forecast error to result in a freeze.  By contrast, if the forecast shows temperatures just above freezing, you’d be taking a much larger risk by planting now on the basis of that highly-uncertain 10-day forecast.

Yesterday, per, Vienna VA passed its 70th percentile last frost date.  In theory, if I just blindly planted on that day, there would be a 30 percent chance that frost would kill my plants.

Source:  Adapted from,

But there’s nothing even approaching frost in the 10-day forecast.

Source:  Weather Underground.

And if you look at the last frost chart, another 10 days takes me literally off-the-charts.  The likelihood of frost is small enough that they don’t bother to tabulate it out of a mere 30 years’ worth of data.

So I went ahead and planted my summer squash.  And some of my cucumbers.  It would be less risk if I had a poly tunnel or a greenhouse or some such, to reduce the frost risk.  But I don’t.  And I’m planting anyway.

Degree days and insect emergence

In the past, I naively thought that insects emerged according to the calendar.  June bugs, well, they show up in June.  Japanese beetles show up mid-summer.  And so on.

But in fact, most bugs show up based not on the calendar, but on how warm it has been.  In practice, as measured by humans, they show up after a certain number of growing-degree-days in a season.

This is something that I had not quite grasped until I tried to deal with a bad infestation of squash vine borer, and decided to summarize everything I had found out about that pest.  (Post #G27, a treatise on the squash vine borer.)  That’s when I found out that the insects arrive regular-as-clockwork, based not on the calendar, but on the temperature.

I did a full write up of growing degree-days in that post.  I’ll repeat the key part here.

Growing degree days in a year are calculated as the cumulative time during which the air temperature exceeds 50F.  In Virginia,  Virginia Tech would be the place to get information.  A nice general reference is available from the American Public Gardens Association.  The most commonly-used basis is 50 degrees because that’s the temperature at which most plants and insects begin to grow.

For every major insect pest, you can look up their expected time of emergence in terms of growing degree-days.  For example, here’s a table of more-or-less every pest that attacks Christmas trees.

Elsewhere, you can look up individual pests.  For example, Japanese Beetles emerge at around 1000 GDD.  As of today, based on Cornell University’s tracking, we’re at just over 200 GDD in Vienna, VA (based on weather at Dulles Airport).

Last year, according to this site, we (Vienna, VA) passed 1000 GDD back on June 22. That’s the point where the squash vine borer begins to emerge.  There’s no telling when that will occur this year, but early July is a good guess for the end of my summer squash season this year.

Hence the hurry to get the plants into the ground.  This is early by almost any standard.  But there’s a reason for it.