By Mitch Lansky
Some plants, such as tomatoes, peppers, melons or squash, are frost sensitive. Well, I'm sensitive to frost, too. I live in a frost pocket in Wytopitlock, Maine. We have a growing season that averages fewer than 100 days a year, and we can have frosts at any month of the year. Indeed, we have had frosts on such unlikely dates as June 26, July 4, July 16 and even July 31. Our last "spring" frost has come as late as June 16, and our first "fall" frost has come as early as the 23rd of August.
We can, to some extent, protect our frost sensitive plants by covering them with blankets and tarps. Our gardens, however, are quite large, and we prefer not to cover and uncover all the sensitive crops unless a strong chance of frost exists.
I have been frustrated by unreliable forecasts. Too many times we have heard forecasts for overnight temperatures in the low 40s, with no mention of possibility of frost, only to awaken and find our crops crisped in a layer of ice.
Since I can't rely on forecasts, I use weather clues, such as clear skies, no winds, no expectation of warm fronts, temperatures of 54 degrees or less at 9 p.m., with fast dropping temperatures, low dew point (the temperature at which water condenses), and a gut feel that frost is in the air. I also sometimes call a friend who has a Ph.D. in meteorology for a second opinion, but he lives over 100 miles away, so observations from his area aren't totally reliable in ours. If I think there is a chance of frost, I cover my plants.
Other friends warn me that I need to watch for frosts around the time of the full moon. Occasionally, a frost does occur on the full moon, and on such nights, the moon is very clear and bright, because the air is clear, with little moisture to block the view. But does the bright moon cause the frosts or does the frosty air make the moon bright?
A wealth of weather lore relates the moon to frost. Some of this lore, regarding rings around the moon, the color of the moon (sallow or red), or the clarity of the moon, has some scientific basis, indicating the coming of high cirrus clouds, the approach of a front, or the amount of moisture in the atmosphere. Others, regarding the phases of the moon as a cause for weather changes, are regarded with skepticism by scientists.
These friends, however, insist that the connection between full moons and frosts is based on science. They argue that the light of the full (or nearly full) moon can heat the upper atmosphere and evaporate cloud moisture that would ordinarily trap heat escaping from earth, thus facilitating radiational cooling that could lead to frosts. Though this argument sounds slightly plausible, I have never seen studies that back it up. For example, how much heat in the light reflected from the moon reaches earth? Scientists I have talked to don't think it's enough heat to evaporate clouds in the upper atmosphere that are composed of ice crystals. If the sun can't do that during the day, how can the moon do that at night?
I also wonder at the logic of the full moon "causing" frosts, given that frosts often happen in a spotty fashion; hitting one part of town, in the valleys, but skipping the hilly sections. Frosts also hit some towns and not others. And the moon shines over much of the hemisphere at the same time. Surely no frost during the growing season happens over the entire northern hemisphere at the same time. Weather is far too variable.
Lately, I read an article by a respected author who stated the moon/frost connection as if it were an established fact that "everyone knows." This raised my curiosity. If an established, scientifically-based connection exists, then I ought to use it in my arsenal of tools for anticipating frosts.
I called the National Weather Service in Caribou, Maine, but they were very skeptical of such a connection and had little interest in researching it. The Service did, however, direct me to weather-related Web sites, so I did the research myself by checking frost data from Houlton at this web site.
I used data from the Houlton, Maine, airport, a reporting station in a low-lying area that is likely to get frosts and that is the closest reporting station to Wytopitlock. I looked at all 38 years of data available from that station for the period of May 18 to September 20, the official growing season for Houlton, and I looked at only the first 30 and last 30 days of the growing season, since frosts are much less likely from late June to late August. I also found a web site that showed moon phases for thousands of years, including the 38-year period for my analysis.
I decided to look at temperatures 34 degrees or lower, rather than 32 or lower, because my own experience has been that temperatures of 34 degrees in Houlton often mean frosts on my land in Wytopitlock. Part of this is due to lower temperatures on my land than in Houlton, but sometimes frosts can occur on the ground even when the air temperature is 34 degrees, as the ground can cool faster than the air. I also decided to look at not just full moons, but also the days before and after full moons, since these nearly full moons should have a similar effect on the atmosphere. These extra days also made my sample base larger, thus improving statistical reliability.
The results surprised me. I was open to the possibility that the moon might have some slight effect (but probably not enough to usefully predict frosts), but I found that only 8 out of the 80, or 10%, of full moons during this period coincided with frosts. Only 26 out of the 232 days (11.2%) of full moons plus one day on either side coincided with frosts. The same period had 267 frosts out of 2280 days, so the chance of a frost on any day was 11.7 percent. Thus the coincidence of frosts with full moons seems to be random. A supporter of the moon/frost hypothesis said my analysis was too simplistic. The relationship only works, he told me, where a cold air mass coincides with the full moon. The full moon cannot evaporate clouds in a warm-air mass. If the incidence of cold-air masses during a full moon is random, however, then the influence of the moon on that fraction of air masses, leading to increased frosts, should still increase the probability of total frosts over random levels. This was not the case in my data sample.
He responded that the relationship did not show because my sample base was too small. If 38 years of data are inadequate to show a relationship, however, then any relationship that does exist is too subtle to be useful for gardeners with a normal human life span.
Others have suggested that I look at different weather stations. It seems highly unlikely, however, that the moon would influence frosts in Bangor, Halifax or Albany, but not Houlton.
When I am wondering if a frost is coming, I continue to take many factors into account, including air masses, wind speed and direction, dew point and cloud cover. But I don't pay any attention to the phase of the moon, except, perhaps, to the extent that moonlight will be bright enough that I won't need a flashlight to cover my crops at night.
About the author: Mitch Lansky is the author of Beyond the Beauty Strip: Saving What's Left of our Forests and editor of Low-Impact Forestry: Forestry as if the Future Mattered. The article first appeared in Rural Delivery.
An article relating several factors (including the phase of the moon) simultaneously to frost appeared in the Sept.-Nov. 2004 issue of The MOF&G and elicited many comments and questions. For example, MOFGA member Beedy Parker wonders: If the full moon makes water tides in the soil water table rise as they do in the sea, could water rise more in plants during a full moon, too, causing more tissue damage as plants freeze … or, alternatively, can transpiring that extra water protect plants from frost the way spraying water on them can? Weather and gardening can generate no end of questions! – Ed