Time series analysis of Cenozoic era sea level and paleotemperature data

A statistical analysis of Cenozoic era sea level and paleotemperature data was performed to determine the cycles of each data set and the correspondence between them. Accordingly, each of the four time series were first analyzed independently in the univariate mode of a spectral analysis. The two basic data sets were then analyzed in a paired cross-spectral analysis.

The prominent periodic cycles remaining in the data sets after linear trend removal, were: sea level surface from seismic stratigraphy--9.6 million years, updated version of sea level surface from seismic stratigraphy--9.5 million years, continental paleotemperatures from paleobotanical interpretations--9.6 million years, and marine paleotemperatures from foraminiferal isotopic data--12.7 million years. The cross-correlation properties between the data sets of continental paleotemperatures from paleobotanical interpretations and sea level surface from seismic stratigraphy at the common prominent period of 9.6 million years were: (1) The squared coherency value which measures cross correlation between the two data sets has the value 0.30, and (2) the amount by which the continental paleotemperatures from paleobotanical interpretations data lags the sea level surface from seismic stratigraphy data is 2.70 million years.