PP33C-1258:
Late Pleistocene Sea Level Stack
Wednesday, 17 December 2014
Rachel M. Spratt and Lorraine E Lisiecki, University of California Santa Barbara, Santa Barbara, CA, United States
Abstract:
Sea level reconstructions have been created using wide variety of proxies and models. The accuracy of individual sea level reconstructions is limited by measurement, noise, local variations in salinity and temperature, and the assumptions particular to each reconstruction. To address these limitations, we have created a sea level stack (average) which increases the signal-to-noise ratio of sea level estimates by combining 5-7 sea level reconstructions over the last 800 kyr. Principal Component analysis (PCA) of seven sea level records from 0-430 kyr ago shows that 82% of the variance in these records is explained by their first principal component (i.e., the stack). Additionally, a stack of just the 5 longer records that extends to 800 kyr closely matches the timing and amplitude of our seven-record mean. We find that the mean sea level estimate for Marine Isotope Stage (MIS) 5e is 0-4 m above modern, and that the standard deviation of individual estimates is 11 m. Mean sea level estimates for MIS 11 are 12-16 m above modern with a standard deviation of 30 m. Due to the large variability between individual reconstructions, our sea level stack may provide more robust sea level estimates than any single technique.
