RESOLVING MILANKOVITCH IN THE PALEOCLIMATE RECORD: CONSIDERATION OF SIGNAL AND NOISE

Orbital-climate theory has had a profound impact on the field of paleoclimatology. In fact, the orbital paradigm has become a standard component of the Earth System model that is taught in Earth Science classrooms. Yet the orbital hypothesis is not without its detractors. Although the identification and quantification of orbital rhythms in paleoclimate archives has become commonplace, a number of criticisms of the theory have been advanced that cast doubt upon the importance of an astronomical driver. If such criticisms are valid, they will require a major revision of the established Earth System model. Part of this debate stems from problems associated with accurate quantification of orbital cyclicity in paleoclimate data. That orbital-insolation cycles occur is not an issue. That they vary both temporally and spatially in the extent that they influence the climate system is a given. The variable fidelity of the sedimentary record of climatic change is also well established. A key procedural challenge in any attempt to quantify orbital influence in the geologic past is the identification and filtration of noise from the climatic signal. In this talk, we outline a theoretical framework to systematically analyze the sources of noise in paleoclimate records, and build a methodology (employing spectral and statistical techniques) for the accurate quantification of orbitally-forced climate variability. We apply this methodology to a recently revised temperature proxy record from the Vostok ice core. Once the sources of noise and distortion of the orbital-insolation signal have been constrained, our analysis indicates that the Milankovitch periods account for >50% of the variance observed in the temperature proxy record, as opposed to a recent estimate of ~7%. The results highlight the importance of this methodological approach in the analysis of noisy geological archives, and provide further support for the orbital paradigm.