Skip to Content

First frost? When it typically happens and when we might see one

So far, we have managed to stay frost free across eastern Iowa which I’m sure all of the farmers can appreciate. Our coldest temperatures of the season were recorded yesterday morning (October 4th) with lows dipping into the upper 30’s to the middle 40’s.

Widespread frost has already been observed in parts of the Midwest, mainly in northern Minnesota and North Dakota.

The temperature needs to fall to 32 degrees in order for water to freeze and form frost. This is more likely to occur at the surface than 2 meters above the surface which just happens to be where temperatures are measured. Therefore, it is common for an air temperature to read out as a few degrees above freezing while there is frost on the ground. It isn’t out of the question to see a 2m temperature of 36° produce frost. Our typical first 36 degree night usually comes in late September.

Our typical first 32 degree morning comes in the last 10 days of September through the first 10 days of October, depending on your exact location.

Therefore, not only are we due for our first 36° low temperature, but also our first 32° low which would produce widespread frosty conditions.

A temperature of 28° is a hard freeze and would effectively end the growing season. That typically occurs in the first third of October north to the last third of October south. There has not been a widespread freeze in the Midwest to date this year.


Currently, a potent cold front is forecast to move through on Thursday which will cause temperatures to tumble through the day on Thursday. You can see the progression of the cold air here:

Currently, long range models peg lows to fall to the low 30’s (and upper 20’s) Thursday night and Friday night, although the best chance would come on Friday night if skies clear out and cold air continues to surge in.

Only the GFS is shown here

A meteorologist’s expertise is usually derived from experience. That experience tells me that:

A) models are rarely (if ever) spot on, even in the short term.

B) models in general struggle more than a few days out to paint an accurate picture

C) we are in a transition season where the models are trying to figure out what the heck is going on.

That’s why, in the last few days of the 7 day forecast, you don’t ever see large swings over a few degrees because if we relied solely on model data, the forecast would change wildly every 6 hours. Therefore, that far out we forecast more on trends than actual numbers.

The forecast as of Saturday morning doesn’t even show temperatures in the 30’s Thursday night and has a high of 55 for Friday. Those numbers are lower than the previous forecast based on the model trends. If the trend continues, the next forecast put out this afternoon will likely have lower numbers than currently. We will have a better idea of what we are looking at over the next few days.

Just keep in mind that a frost is possible by the end of the week but there are a lot of different factors that need to play out before we can say that with any confidence.

Also a quick note: if you come or came across any social media post referencing the potential for snow in Iowa by the end of this week, you are not following reliable weather sources. These pages should be blocked and deleted for spreading false weather information. Get your weather from reliable weather sources which are almost exclusively broadcast media or the National Weather Service.

Now, if you are wondering, the climatological date of a late first freeze, defined as the 90th percentile, ranges from early to late October.

For the date of a late first freeze of 28°, the dates range from mid October to early November.

Author Profile Photo

Brandon Libby


Skip to content