The past two weeks turned out to be a fine time for me to go on a wide-ranging crop tour across seven states. For one thing, it's just a lovely time of year for driving. For another thing, I didn't really want to sit in front of a computer screen watching the market carnage. I couldn't bring myself to sell into a short-term bearish market when I believe the fundamentals will turn the thing around in the longer term. (Of course I say that with all the usual disclaimers about the dollar, the European economy, the risks of commodity investing, etc.)

So ... speaking of things I dread, and which aren't expected under ordinary circumstances, yet which go ahead and happen anyway ... how about that frost last week? Calling it "unexpected" is a bit of a stretch. Weather happens. And for an arbitrarily-selected NOAA weather station relatively close to the 45-degree line of latitude, a first freeze date of September 15 is fairly probable. Aberdeen, S.D., for instance, has a 50% probability of reaching 32 degrees Fahrenheit by September 21.

The people who were particularly unsurprised by last week's frost were the ones who've inherited the myth that the first frost of the fall is likely to hit around the same time as the Harvest Moon. This kind of legend -- with absolutely no basis in actual science -- drives me crazy. And what drives me even crazier is the fact that it was right ...this year. First frost hit the northern tiers of the Corn Belt just after 6 a.m. on September 15, which was just three days after the Harvest Moon waxed on September 12 at 5:27 a.m. (Central Time).

The closest scientific explanation for this moon-frost myth anyone could find for me was that when there's a full moon (a "Harvest Moon" is the full moon which occurs closest to the Autumnal Equinox, which is usually September 22 or 23), the light from the moon is strong enough to heat the upper atmosphere and burn off the clouds which would otherwise insulate our surface temperatures. Forget it. The moon doesn't put off that much heat.

And anyway, I am just pedantic enough to run the statistics and finally disprove this notion once and for all. In the 35 years between 1969 and 2003, the first freeze (32 degrees) in Aberdeen, S.D., occurred on the same date as the Harvest Moon only once: September 23, 1983. Giving a three-day grace period on either side of the Harvest Moon (a six-day period), the first freeze occurred within that Harvest Moon timeframe eight times in those 35 years (only 23% of the time). Since the moon has a 28-day cycle, you can pick any random six-day period during the moon's September cycle and get roughly the same result for the probability of a first frost (6 divided by 28 is 21%). The phase of the moon has nothing to do with it.

We get much better results if we look at the probability of a first freeze around the timeframe of the actual Fall Equinox. Eighteen times (51%) in those 35 years, the first freeze along the 45-degree line of latitude occurred within three days of the equinox. But that should be no surprise, because the equinox, by definition, arrives at a specific time of year which is almost exactly in line with the 50% probability level for the first fall freeze.

This whole subject became more of an issue in 2011 than in some other years, because of the prevalence of late-planted soybeans across the Upper Midwest and Northern Plains. Recall that fewer than 40% of soybean acres in Michigan, Minnesota, North Dakota, South Dakota and Wisconsin were planted by the end of May. It's a shame, because some of the nicest soybean fields I saw on my multi-state crop tour were in Minnesota, but they were just starting to turn colors and were a long way from being ready for a freeze.

The potential for an early freeze was the whole reason the market was concerned about that late planting (not that traders are acting terribly concerned right now). It seems like anything you could imagine going wrong with U.S. crops did go wrong ... somewhere. I can find fields for you this year to demonstrate the effects of hail, green snap, flooding, poor pollination, weed pressure, Goss' wilt, lack of early heat units, vomitoxin, frost and/or a crippling lack of rain. We didn't know these things were going to happen at the start of the season, but certainly we knew they were possible.

Back on June 1, a farmer from Macoupin County, Ill., pointed out to me that Brood XIX of the "13-year" cicada was hatching in 2011, and it would be the largest brood of cicadas ever, by geographical area. Their population stretched across Missouri, Arkansas, southern Illinois, Mississippi, Tennessee, Alabama, and eastward. This was relevant, in farmer-lore, because supposedly the arrival of the cicadas leads to above-average corn yields.

Well, buzz-kill that I am, I crunched those numbers, too. In the six cycle years since 1925 that these insects have flourished (1933, 1946, 1959, 1972, 1985 and 1998), corn yields in Macoupin County, Ill., have averaged nearly 5% below their local trendline yields. Three of the years were above trendline (by an average of 11% above the Macoupin County trendline), and three of the years were below trendline (by an average of 20% below trendline). The percentage above or below trendline during these cycle years was random enough to be considered statistically insignificant. And 2011's miserable yields aren't going to help that particular myth.

Anyway, these little exercises are a long way off from being what we would call "trade-able" ideas, particularly in light of the fact that soybean prices have done the exact opposite of what fundamental traders would expect after the Harvest Moon freeze of 2011. And corn prices, too, seem oddly unconcerned with the cicadas' failure to deliver a bumper crop to southern Illinois. But it's a big world out there, and there are a lot of things I don't understand. I'm just grateful for the opportunity to get out there and find out about them.

Elaine Kub can be reached at ekub@agrisk.net

(AG/ES)

© Copyright 2011 DTN/The Progressive Farmer, A Telvent Brand. All rights reserved.