Science: Climate follows the 'double sunspot cycle'
A 22-YEAR cycle linked with changes in the Sun’s activity has influenced trends in temperature this century more than any other factor. This is the conclusion of Nicholas Newell, a scientist in Arlington, Massachusetts. He and his colleagues at the Massachusetts Institute of Technology have analysed temperature variations from 1856 to 1986. The idea that changes in solar activity affect the weather is not new. Unfortunately, researchers have always found it difficult to pin down the link because they lack any good data on climate in the past. Recently, however, several groups of researchers have collated and analysed temperature measurements from around the world. The result of their work is the best set of global annual averages that researchers have ever had. The data go back to the middle of the 19th century. The most important of these data are scientists’ measurements of the air temperature, made at night above the oceans. These are called Marine Air Temperatures, or MATs. Newell and his colleagues analysed the variations in these temperatures and found strong evidence of a periodic fluctuation with a cycle that is 21.8 years long (Geophysical Research Letters, vol 16, p 311). Statistical tests that the researchers have applied to the data show that there is less than one chance in a thousand that this periodicity is simply an accumulation of random changes. The cycle of 21.8 years matches a fundamental cyclic change in the Sun, the so-called ‘double sunspot cycle’, over which the magnetic polarity of the Sun first reverses, then reverts to its former state. Most people are more familiar with the cycle of sunspot activity that lasts for 11 years. The average length of the double sunspot cycle over the past 237 years has been very close to 22 years. When Newell and his colleagues match up the solar magnetic cycle and the temperature cycle, they find that these march almost perfectly in step, with alternate peaks of the 11-year sunspot cycle corresponding to alternate upward and downward swings of temperature, over a range of just a couple of tenths of a degree celsius. ‘A possible interpretation,’ say the researchers, ‘is that cooling in MAT takes place during one phase of solar hemispheric magnetic polarity, and warming during the other.’ There is one exception, however. In the 1870s, there was an interval of sustained ‘high’ temperature. But this may be the exception that proves the rule, because there was a hiccup in the solar cycles at that time, producing two consecutive peaks in sunspot activity without a change in solar magnetic polarity. It was the only such occasion between 1856 and 1986. Tantalisingly, the analysis stops in the mid-1980s. Since then, the world has experienced record-breaking temperatures: 1987 and 1988 were the warmest on record. If there is more than a coincidental connection between solar magnetic variations and trends in temperature, the world should now be cooling slightly as we head towards a peak in sunspot activity, which is due within the next few months. Relatively cool conditions in the late 1960s and 1940s show up clearly in the analysis and contrast with relative warmth in the late 1950s and 1970s (see figure). But whatever the implications for future weather (which Newell does not discuss) the team has no doubts about the need to reconsider analyses of past trends in the light of their discovery. ‘After subtraction of the periods less than 26 years, which are dominated by this variation (the 22-year cycle), there is no appreciable difference between temperatures at the beginning and end of the record.’ They also point out that although the record of the 20th century alone does show a warming trend, this is largely because temperatures fell to the lowest point in the 131-year record in the first decade of the present century,