|
E. Linacre and B. Geerts |
3/’02 |
Much research has been dedicated to the discovery of cycles in time series of climate data. Any record can be interpreted as the summation of simultaneous cycles of change, each cycle having its own individual cycle duration (ie ‘period) and amplitude. This is called spectral analysis. Spectral curves display frequency of a cycle against its period. The Madden Julian Oscillation, for instance, was discovered through the spectral analysis of upper tropospheric winds in the equatorial Pacific.
The only outstanding cycles in the climate system are the daily and yearly rhythms. The remaining cycles explain less than 1% of the variation. Because the daily and annual cycles are exact, they can be removed from the spectral curve, and the remaining variation can be examined. A signal with no cycles appears as a slowly descending spectral curve, known as red noise. Cycles are identified as the positive departures from this red noise curve.
Amongst the many alleged cycles of climate, only a few are credible. These are a) based on a long record, b) statistically significant, and c) explicable in terms of known physical processes. Others less believable may well be the result of poor statistical treatment of the data (1,2).
sunspots
Sunspots have a periodicity of about 11 years and recently evidence has emerged that the number of sunspots affects climate. For instance, sunspots were virtually absent during the Little Ice Age between 1450 to 1820.
solar day
Sunspot movements indicate that high-latitude parts of the Sun rotate once in each 25 of our days, but in 28 days near the Sun’s equator. Those figures define the ‘solar day’. The rotation makes rays of charged particles from the Sun sweep over the Earth in a regular fashion. Each passage of the electromagnetic fields associated with these particles reduces the strength of the circumpolar flow and hence the vorticity within the Earth's upper atmosphere at middle and high latitudes, lasting a few days. Such a meteorological response to solar activity is most evident in winter, and tends to be opposite over land from over the oceans.
Incidentally, the solar day is about as long as the lunar cycle, which affects the height of the tides. In coastal areas where large areas lay between the high and low tidal marks, the lunar cycle may affect the intensity of the sea breeze and other aspects of weather.
effects due to the Earth’s orbit
Variations of climate over the past 500,000 years have been analysed from fluctuations in the composition of sediments at different depths below the bed of the Indian Ocean (2). The spectral peak with a period of 100,000 years was found. The next most important swings have maxima approximately each 43,000, 24,000 & 19,000 years, respectively, again largely explained by the Milankovic variations of the Earth’s orbit.
The so-called ‘Milankovic hypothesis' explains the Ice Ages in terms of fluctuations of terrestrial solar radiation due to features of the Earth’s orbit around the Sun. The hypothesis was previously put forward by Adhemar in 1842 (3). It is supported by the history of sea-surface temperature changes indicated by the analysis of layers of deposits on the ocean floor. But the observed changes are greater than are likely from the variation of solar radiation alone.
A complementary cause might be the fluctuation of Sun/Earth distance (due to the Earth’s orbit) affecting the Sun’s gravitational pull on the swirling of the Earth’s liquid, magnetic core. The resulting changes in the Earth’s magnetism would alter our atmosphere’s susceptibility to the solar wind of charged particles from the Sun.
References