Bayesian analysis of rapid climate change during the last glacial using Greenland δ¹⁸O data

Abstract

We present statistical methods to determine climate regimes for the last glacial period using three temperature proxy records from Greenland: measurements of δ18O from the Greenland Ice Sheet Project 2 (GISP2), the Greenland Ice Core Project (GRIP) and the North Greenland Ice Core Project (NGRIP) using different timescales. A Markov Chain Monte Carlo method is presented to infer the number of states in a latent variable model along with their associated parameters. By using Bayesian model comparison methods we find that a model with 3 states is sufficient. These states correspond to a gradual cooling during the Greenland Interstadials, more rapid temperature decrease into Greenland Stadial and to the sudden rebound temperature increase at the onset of Greenland Interstadials. We investigate the recurrence properties of the onset of Greenland Interstadials and find no evidence to reject the null hypothesis of randomly timed events.