Figure 1: Frost Blanket
When I was in high school, I had a tomato plant that got to be 3½ years old. It was enormous. The stem where it came out of the ground was as big around as my arm. I had to use loppers to cut pathways through it to get to the tomatoes.
Most people treat tomatoes as annuals, but really they’re perennials; they just don’t tolerate frost. Some varieties are determinate, meaning that they set fruit once and then they’re done. Others are indeterminate, meaning they keep going and going and going. If you know what variety you have, you can try googling to see whether it is determinate or indeterminate.
A few nights each year, when it was forecast to get cold, I covered it with a tarp. The bits that stuck out beyond the edge of the tarp got killed, but the rest of the plant was fine. In this context “cold” means a forecast of 38∘ F or below with no wind, or 34∘ or below (with or without wind).
As a corollary of Murphy’s Law:
— If you cover your plants, the weather will be warmer than forecast.
— If you don’t cover them, the weather will be colder than forecast.
You can obtain a point forecast by going to https://forecast.weather.gov/MapClick.php?lon=-110.942261&lat=32.236327 and then clicking on the map (captioned Point Forecast, about halfway down the page) to give it your actual location. (If you bookmark the resulting page you can get back to it easily, without having to click the map again.)
Keep in mind that when NOAA talks about a temperature, they mean the air temperature. You don’t need to build a full-blown Stevenson instrument shelter (as discussed in reference 1), but you should put the thermometer in the shade, under some sort of awning, with unobstructed airflow, not too close to the ground, and not too close to the house, but close enough that you can see it. This is not always super-easy to arrange.
Once you have a reliable forecast of air temperature, keep in mind that the plants could be quite a bit colder than the air under some conditions, as discussed in section 2.
Gardeners distinguish between a frost and a freeze, as shown in figure 2.
Neither of those concepts is exactly what you care about for gardening purposes, because no plant on earth will drop dead when the temperature drops to exactly 32∘ F; they all contain “some” amount of antifreeze. (Some unusual tropical plants will get very unhappy if the temperature stays below 40∘ F for an extended period of time, but let’s not worry about that.) Some plants such as tomatoes are injured by temperatures only slightly below 32∘ F, while others tolerate much colder temperatures.
The main factors to consider are:
On a non-windy night with clear skies, you can have a temperature inversion. That means the air near the ground is colder than the air higher up. The coldest air will settle into hollows and stay there. (This stands in contrast to daytime conditions, where almost always the air is warmest near the ground and gets steadily cooler as you go up.)
The main operational implication is this:
| |
There are two notions of frost. One refers to hoarfrost, i.e. visible deposits of ice that form on cold surfaces. This requires the surface temperature to be below freezing and below the dew point.
The other possibility is black frost, where no ice forms, even though the surface is plenty cold, because the air is too dry. The surface temperature is below freezing, but the dew point is even lower than that, as shown in figure 3. This happens quite commonly in the desert.
For gardening purposes, the concept of frost includes black frost as well as hoarfrost. Frost does not have to be visible to injure the plants. High humidity makes the frost more visible but less dangerous.
Suppose the air is at 35∘ F. The night sky is at minus 455∘ F. The plant will split the difference. It will “mostly” be in equilibrium with the air, but it will have some tendency to come into equilibrium with deep space.
High humidity in the air protects plants in two ways:
Techniques to minimize frost damage include:
The cloth will get cold, so you get somewhat better protection if you space the cloth above the plant, using poles or carboard boxes or whatever. (If it’s a small seedling, spacers have the further advantage of preventing the the cloth from flattening the plant. On the other hand, if the plant is big, spacers aren’t worth the trouble. A few of the tips might get frostbitten, but the other 99.99% of the plant will be fine.)
Remove the frost blanket in the morning. If the sun is shining and the air temperature is above freezing, frost is not happening.
A loose and/or porous cloth will be effective at blocking radiative cooling when the air is not too cold. However, when the air itself is nasty cold, you need an airtight covering.
A single layer of aluminized mylar “space blanket” is adequate. Similarly, an opaque plastic tarp is adequate.
In contrast: Clear plastic of any kind (including bubble wrap) will not do, and may even make things worse. That’s because it’s transparent to IR radiation.
I’m not sure this is practical for the ordinary backyard garden, but it would be interesting to do the experiment.
It may seem counterintutive to warm something with a fan, but it actually makes sense. Note that the concept of wind chill that you see in weather reports applies when cold wind blows on big round humans who are trying to stay at 98∘. It does not apply to thin leaves that have already cooled to 33∘ and you are trying to warm them back up to 34∘ or so. A 35∘ wind cannot possibly cool a piece of dry wood below 35∘, no matter how hard it blows. That would violate the second law of thermodynamics.
We can apply the same idea in reverse: If you are situated in a hollow, IR radiation will make things get a lot colder where you are, compared to a nearby location that is more exposed to the wind. This is an example of a microclimate. The worst-case scenario is where the site is enclosed enough to block the wind, but open enough to give you a wide, clear view of the sky.
Consider the contrast:
| When the air is cold, you need to protect against the wind and the radiation both, as discussed in this section. If the air is below freezing, you have to use a covering that minimizes air leaks, especially if it’s windy. | When the air is above freezing, you need to protect against radiative cooling, but you do not need to make the covering airtight, as discussed back in section 2.1. |
Possible coverings include:
At the bottom, fold the flaps outward and put rocks on them to provide stability.
From the appliance store you can get boxes that are larger than the ones you typically get from Amazon. This helps for larger plants ... and for potted plants, where you have to account for the height of the pot. With a little ingenuity you can use tape and/or glue to combine two boxes to make something more than twice the height of a single box.
In contrast: Clear plastic of any kind is no good, because it is transparent to IR radiation.
Minor supplementary considerations include:
Do this only as a supplement to a good covering. The covering does most of the work. It would take thousands of watts of power to make up for the heat lost via IR radiation and air conduction.
I don’t recommend Christmas light strings, because they don’t put out very much heat. Also: Spreading them out evenly is unduly laborious, and if you don’t spread them out they have no advantage over one big bulb.
Again, this is only a minor supplement.
Conventional wisdom is that clouds protect against radiative cooling. That’s true for low clouds, but only partly true for high clouds. The situation is shown in figure 6.
The amount of power we are talking about is enormous: 650 watts per square meter. Twenty percent of enormous is still pretty big, so beware.
The fundamental issue is that high clouds are cold. They’re not as cold as outer space, but still cold enough to be a problem. The heat they give off goes like the fourth power of the absolute temperature, so the problem is worse than you might have guessed.
On the other hand: Invisible humdity in the air is almost as effective as a visible cloud. Humid air at low altitudes, where the air is relatively warm, gives considerable protection against radiative cooling. The presence of scattered low clouds is sometimes a proxy, indicating high humidity at lower altitudes. Rain pretty much guarantees high humidity at low altitudes.
Note: The graph stops at 36,000 feet because there are essentially never clouds higher than that.
We now consider the situation where we get a bit of rain immediately followed by a cold snap. This is unusual but not impossible, because there can be rainfall associated with passage of a cold front.
This poses some tricky questions. The answers depend on details.
The swamp-cooler temperature will be roughly halfway between the air temperature and the dewpoint. For example, you could have:
| Air temperature: | | 25∘ |
| Swamp-cooler temperature: | | 20∘ |
| Dewpoint: | | 15∘ |
In particular, a cardboard box makes a very effective cover, but doesn’t do well in the rain. The combination of a cardboard box covered by a plastic trash bag works better than either thing separately.
A wet cover is much better than nothing. A dry cover would be even better, but a wet cover is not bad. It may take some extra care to keep it from sagging and touching the plant, but that’s manageable.
