| In the first chapter, I test the hypothesis that the phases between Milankovitch climate forcing and climate response has remained constant over long time periods (105 to 106 years). With the exception of sites influenced by atmospheric dust transport associated with the African monsoon, the phases of obliquity and precession have remained remarkably stable for the last six million years. In the eastern equatorial Pacific, the eccentricity phase and dominant period changed approximately three million years ago.; In the second chapter, I focus on the generation of spurious harmonics in power spectra and bispectra by non-sinusoidal data, which are common in paleoclimatic time series. As the focus of paleoclimatic studies moves to sub-Milankovitch variability it is important to understand the conditions that generate spurious harmonics in power spectra and bispectra. The presence of normally-distributed data or sub-Milankovitch variability in raw time series or in carefully filtered time series minimizes the possibility of misinterpreting harmonics and combination tones.; In the third chapter, I investigate the relationship between sub-Milankovitch climate variability and the transition from the 41,000 year (41 ky) glacial cycles of the Pliocene to the 100 ky year cycles of the Pleistocene. Results indicate that the 100 ky glacial cycles began to build near 1.5 million years ago (Ma), coincident with the propagation from the tropics to the high latitudes of a phase-locked ∼10 ky semiprecession cycle. We suggest that expansion of tropical semiprecession cycles to the northern latitudes provides the 10 ky oscillation required by numerical models to produce high amplitude 100 ky glacial cycles.; In the final chapter I applied Empirical Orthogonal Function (EOF) analysis to 16 benthic and 14 planktic oxygen isotope records. Results indicate that the combined benthic records (stack) contain a global ice volume signal whereas the planktic stack contains a combination of ice-volume and temperature effects. Planktic isotope residuals suggest that increased cross-equatorial heat transport was associated with a relatively warmer northern tropical Atlantic during periods of decreasing ice volume. |
