We can now apply the age scale to the color series to obtain a first impression of the relationships between color and climatic change. A comparison with the oxygen-isotope record of the Milankovitch Chron is instructive (Fig. 8). It is the most familiar portion of the Quaternary and the sole one dominated by 100-k.y. climatic cycles. The detrended color data that have been stacked in the manner described for magnetic susceptibility are plotted in Figure 8. All values are standardized and offset for clarity.
It is obvious that the color cycles have much higher frequencies than those in the isotope curve, and that an influence of the ice-age cycles (100 and 41 k.y.) is not readily apparent (Fig. 8A). There is a hint of more reddish sediments being associated with the glacial stages. For the total reflectance, a similar relationship holds, except that high reflectance seems distinctly tied to interglacial conditions (Fig. 8B).
The overall negative correlation between color and reflectance within the 100-k.y. band holds true, especially for the Milankovitch Chron (Fig. 9A), but it tends to break down at times of minimum eccentricity of the Earth's orbit (near 400 and 800 k.y.). This points to the importance of eccentricity—or, rather, precession effects modulated by eccentricity—in the production of color cycles. The curve labeled "model" is the near-100-k.y. component of the ice-age template of Berger et al. (1995), which contains a saw-tooth oscillation feeding off precession. It is in phase with color and reflectance for the Milan-kovitch Chron, but not in the Croll Chron, showing that different rules of sedimentation govern the two periods.
The relationship of color and reflectance to changes in obliquity of the Earth's axis is strong only for the late Brunhes (last 350 k.y.), but not earlier (Fig. 9B). This suggests that climatic messages from the southern polar region became stronger at that time in the Congo Fan region. The change is reminiscent of the mid-Brunhes climatic shift observed by Jansen (1990), but it moves in the opposite sense. Interestingly, within the obliquity-related cycle, color and reflectance tend to be in a phase opposite to that in the 100-k.y. cycle. This strongly suggests that different sediment components determine color and reflectance within different spectral bandwidths. The curves labeled "model" and "obliquity" in Figure 9B are derived from the ice-age template of Berger et al. (1995) and from the orbital data of Berger and Loutre (1991), respectively.
In comparing the obliquity-related color cycles with orbital forcing, we are at the limit of what can be done with the present age model, which has an uncertainty in phase of about 10 k.y. It is moot, therefore, to discuss the phase of precessional cycles. However, we can study the overall pattern of the precessional forcing on the sedimentary cycles as expressed in color (Fig. 10). Precessional amplitudes in color and reflectance follow the amplitudes of orbital forcing in many respects. For example, we see large amplitudes around 100, 200, 600, 700, and 950 ka, as expected. However, the match is not good in some intervals. Perhaps the most striking mismatch is around 400 ka, where orbital forcing is subdued, and yet the response of the system is quite strong. This means that the sensitivity of the system changes perhaps as a function of conditions provided by the 100-k.y. oscillations.
It is apparent that the precessional amplitudes in the reflectance cycles are much weaker than those in the color cycles, especially within the Milankovitch Chron (Fig. 10). A plot of the pertinent spectra brings out the pattern more clearly (Fig. 11). The curves shown depict the amplitude of Fourier components of the autocorrelation series for each of the two color-related variables and for three different periods (0–400, 400–800, and 800–1200 ka). Outstanding features are the substantial power of the reflectance near 100 k.y. throughout the record and the impressive spike of power in the precessional band in the red/blue ratio for the last 400 k.y. Also of interest is the fact that in both variables, obliquity-related cycles are important only within the last 400 k.y., but not earlier. The offset in precessional power in the spectrum of both variables for the period from 0.8 to 1.2 Ma and from the position near 23 k.y. to one just below 20 k.y. suggests that the sedimentation rate was set too high (by ~10%), which would throw the age scale off by 40 k.y. at 1.2 Ma.