A Solar-Cycle Influence on New England’s Climate?
Richard Wolfson [wolfson@middlebury.edu], Middlebury College, VT

This year’s SORCE meeting site inspires a search for solar-cycle influence on New England’s climate. Our dataset combines individual station data for the 73 New England and New York (NE-NY) stations of the Historical Climatology Network that have continuous data for the 98-year period 1903 through 2000. We construct a function consisting of sequential sinusoids whose periods are the (varying) solar-cycle lengths observed for the twentieth century; when we Fourier transform this function, the result is a broadened peak centered at approximately the mean cycle length of 10.65 years. When we Fourier transform the NE-NY temperature data, we find a feature with similar shape located at the same point in frequency space. Correlation between the two transforms in the region of this feature is greater than 90 percent. A Monte Carlo simulation shows that correlation as high as that of the actual data occurs with a probability of only about 0.2%. We corroborate this seeming evidence for a solar-cycle influence with two other studies, one involving the ability of our method to pick the solar-cycle signal out of noise, and the other an autocorrelation of the NE-NY temperature series with varying phase lags. The statistical methodology used here is novel and perhaps questionable, so the author—not a data analyst but a theorist specializing in the solar corona—invites comments on the method and on any significance of the results.