Using empirical mode decomposition to correlate paleoclimatic time-series

J. Solé, A. Turiel, J. E. Llebot

Research output: Contribution to journalArticleResearchpeer-review

17 Citations (Scopus)

Abstract

Determination of the timing and duration of paleoclimatic events is a challenging task. Classical techniques for time-series analysis rely too strongly on having a constant sampling rate, which poorly adapts to the uneven time recording of paleoclimatic variables; new, more flexible methods issued from Non-Linear Physics are hence required. In this paper, we have used Huang's Empirical Mode Decomposition (EMD) for the analysis of paleoclimatic series. We have studied three different time series of temperature proxies, characterizing oscillation patterns by using EMD. To measure the degree of temporal correlation of two variables, we have developed a method that relates couples of modes from different series by calculating the instantaneous phase differences among the associated modes. We observed that when two modes exhibited a constant phase difference, their frequencies were nearly equal to that of Milankovich cycles. Our results show that EMD is a good methodology not only for synchronization of different records but also for determination of the different local frequencies in each time series. Some of the obtained modes may be interpreted as the result of global forcing mechanisms.
Original languageEnglish
Pages (from-to)299-307
JournalNatural Hazards and Earth System Science
Volume7
Issue number2
DOIs
Publication statusPublished - 1 Jan 2007

Fingerprint Dive into the research topics of 'Using empirical mode decomposition to correlate paleoclimatic time-series'. Together they form a unique fingerprint.

Cite this