| Summary: | Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Earth, Atmospheric, and Planetary Sciences, February 2000. "August 1999." Includes bibliographical references (p. 72-73). Spectral methods are used to determine if there are any instances of narrow band oscillations in the climate signal over the past 1.4 million years. The study focuses on the suborbital and millennial-scale regime (~ 1 .5 - 20.0 kyr periodicities) at Site 980 and 983, two deep sea drill sites in the North Atlantic gathered by the Ocean Drilling Program in 1995. The study suggests one such narrow band oscillation in the sediment profile at Site 983, which is discussed from a variety of perspectives. Other questions addressed include a discussion of potential age-model inadequacies deriving from a potentially 'under-resolved' marine isotope map, as regards high-frequency climate variability. Site 983 has a 4-component sediment profile while Site 980 has a 2- component profile. This paper attempts to resolve the various independent inputs at Site 983 by the method described above. Four proxies have been used in this study; they are magnetic susceptibility, natural gamma, spectral reflectance, and GRAPE (Gamma Ray Attenuation Porosity Evaluator). These proxies are used to inform the millennial-scale issue by coherence methods and are also assessed to see which proxies may prove most useful in future high frequency research. Finally, a variety of mathematical and statistical methods have been employed including multi-taper analysis with different numbers of Slepian tapers, Singular Spectrum Analysis (SSA), noise background estimation methods (Mann and Lees, 1996), Akaike' s Information Criterion for AR(n) model fitting to a data set, and chi-square distribution confidence intervals. These various methods are discussed and evaluated according to their usefulness in sub-orbital research. by Darrell Ardon Coles. S.M.
|
|---|