THE NEWS FROM DENMARK

It has been suggested many times that the cycles of solar activity are correlated with cycles of climate on the earth. For example, the proponents of this hypothesis pointed to the rings in the cross sec-tion of old trees, which often also showed an 11-year periodicity, the thicker rings corresponding to greater growth in a wetter year. But the skeptics pointed to trees where the pattern was missing or where its 11-year periodicity was inverted with respect to trees elsewhere.

The proponents had other data, correlating temperature with sunspot number. But the skeptics shot them down as statistically insignificant and as based on unreliable temperature records, and indeed, the records of the sunspot number over the last 240 years are a lot more solid than the unreliable data on temperature, which have been plentiful enough for a (dubious) calculation of a global average only for the last 130 years or so. So the question remained open, but was regarded as unproven; for in good old unpoliticized science, the burden of proof is on the one who upsets the status quo, and until he does so, his hypothesis is considered, but not ac-cepted. This is in stark contrast to the media monkeys who first heard of ozone thinning in the arctic in 1984, had not the slightest inkling of the data for 1974 or 1984 B.C., and instantly broke unto screams of "Avenge the widows!"

Conversely, for the correlation with solar activity they used not the sunspot number itself, but the cycle length, defined as the time elapsed from one maximum (or minimum) to the next. This parameter also turns out to have periodic cycles, but their duration is about 90, rather than 11, years. (Another indictment of the media monkeys who observe something for 6 weeks and run to the Sunday supplements.)

The correlation between these two quantities, solar cycle length and temperature anomaly, shows excellent correlation (see top fig-ure). The comparison of duration of sea ice and the sunspot num-ber is shown in the bottom two figures. It will be seen that each maximum of the cycle length is accompanied by a minimum in the weeks-of-ice count in the period since 1750. From 1865 on, the mean temperature in the Northern Hemisphere has also been added to the figure, and this again closely follows the cycle length.

Of course, we have not heard the skeptics yet. But this time the evidence is incomparably more convincing than the statistics often presented up to now [More: E. Friis-Christensen, K Lassen, "Length of the solar cycle: an indicator of solar activity closely associated with climate," Science, 11/1/91, pp.698700.]