LETTERS TO THE EDITORS
Groundhog Day Science
Groundhog Day Science To the Editors:
We enjoyed the fascinating article by Benjamin Orlove, John Chiang and Mark Cane about folk weather prediction in the September–October issue ("Ethnoclimatology in the Andes"). But the negative remarks about Groundhog Day reminded us of a speculation that we had about its origin that suggests the tradition may be more plausible than commonly supposed. At the time we were spending our winters in Cambridge, England (for our sins), and we noticed that winter weather in East Anglia was of two types, which tended to be quite persistent. Either the prevailing winds were from the west, which led to clouds, light rain and mist—often a freezing fog whose nastiness had to be experienced to be comprehended—or they were from the east, originating in the Russian steppes, warming up incrementally over Holland and the North Sea, but generally leading to a persistence of sunny, very cold, clear weather. The former condition was good for the snowdrops and crocuses and gradually led to a warming trend as the Gulf Stream warmed the Atlantic, whereas the east winds warmed up considerably later.
Thus in East Anglia, Punxsutawney Phil would have been a pretty good weatherman. Our ears pricked up when we read in the local paper of an ancient East Anglian superstition—that if a hedgehog (there being no English equivalent of a groundhog) awoke and saw the sun on Candlemas, it would be a long, cold winter. Candlemas happens to be the Sunday closest to February 4.
To continue the speculation, New England was in large part settled from East Anglia—the town names attest to that—and from that earliest settlement of independent farmers, many migrated down the Atlantic seaboard. So it's reasonable that the tradition was imported by East Anglian farmers, its origin forgotten—as well as that of Candlemas, with the Anglican church calendar discarded—and the closest substitute for the animal appropriated.
Philip and Joyce Anderson
Hopewell, New Jersey Dr. Orlove replies:
My coauthors and I agree that it would be interesting to know if there is an empirical basis to the beliefs about Candlemas in East Anglia. It might well turn out to be another case of an ethnoclimatological belief that is significantly better than random guessing.
You propose to identify the particular atmospheric processes that link the predictor and the outcome. One might begin by looking simply for a statistical link. One could use weather station data to categorize years by whether Candlemas was "sunny" or "shady" and whether the onset of spring (as defined by some measurable indicator) was early or late. It would be simple to see whether the prediction deviates from random. If it does, then there should be dynamic explanations of these associations, which might draw on the characteristics of persistent wind patterns, as you suggest.
We note as well that Candlemas falls at a critical point in the calendar, much as the festival of San Juan does in the Andes. It lies close to the date midway between an equinox and a solstice. Such dates carry an importance in many European and European-derived calendars. Halloween and May Day also occur at such points. Overlaps such as this, and the Indian case that we noted in our article in which observations are made at the festival of Holi, suggest a direction for future research. Do the dates when forecast predictors are noted correspond to the times of greatest forecast accuracy, or do these dates also reflect cultural salience—their position in religious or cosmological calendars? Such questions will be answered only as ethnoclimatological research advances.