The astronomical theory and data compared - Ice Ages and Interglacials: Measurements, Interpretation, and Models - page 308

Figure 10.23. Frequency distribution corresponding to the function in Figure 10.22 .
10.3.2.1 Spectral analysis of solar data
We know from the analysis provided in Section 9.2.1 and Figures 9.2 , 9.3 , and 9.4
that the primary cycle periods that characterize variability of the longitude of
perihelion, obliquity, and eccentricity are roughly 22,000, 41,000, and 100,000
years, respectively. Therefore, one would expect that the frequency spectrum of
solar intensity at high latitudes would show peaks at frequencies of about 0.045,
0.024, and 0.01 (kyr) 1 . Figure 9.7 shows that the dominant factor in determining
the oscillations of solar intensity is the longitude of precession. The heights of the
peaks in the upper part of this figure are not directly correlated with the heights of
the peaks of solar intensity in the lower half of the figure. This suggests that
obliquity is more important than eccentricity in determining the heights of solar
peaks. Nevertheless, the effect of eccentricity is discernible in Figures 9.10 and
9.11 . The amplitude of oscillations appears to maximize at roughly regular
intervals of about 100,000 years at 800,000, 700,000, 600,000, etc. years before the
present although 400,000 ybp is anomalous. Yet, this clearly cyclical behavior does
not seem to be reflected in the spectrum. This suggests that spectral analysis
may not properly identify low frequencies against the background of a rapidly
oscillating time series.
As Figure 9.13 shows, solar intensity at 65 S is similar to that at 65 N, except
that they are out of phase by 11,000 years. Thus, we expect the frequency
spectrum for 65 S to be essentially the same as that for 65 N, and this spectrum is
expected to have a dominant frequency corresponding to precession of the
longitude of perihelion, a secondary peak corresponding to obliquity, and a
weaker peak corresponding to eccentricity. Figure 10.24 shows the spectra
reported by M&M for 65 N and 65 S. On the left-hand side of this figure (65 N),
we clearly observe the primary peak corresponding to precession of the longitude