Anomaly Detection in Paleoclimate Records Using Permutation Entropy

Joshua Garland,Elizabeth Bradley,Michael Neuder,James W C White,Valerie Morris,Tyler R Jones

doi:10.3390/e20120931

Abstract

Permutation entropy techniques can be useful for identifying anomalies in paleoclimate data records, including noise, outliers, and post-processing issues. We demonstrate this using weighted and unweighted permutation entropy with water-isotope records containing data from a deep polar ice core. In one region of these isotope records, our previous calculations (See Garland et al. 2018) revealed an abrupt change in the complexity of the traces: specifically, in the amount of new information that appeared at every time step. We conjectured that this effect was due to noise introduced by an older laboratory instrument. In this paper, we validate that conjecture by reanalyzing a section of the ice core using a more advanced version of the laboratory instrument. The anomalous noise levels are absent from the permutation entropy traces of the new data. In other sections of the core, we show that permutation entropy techniques can be used to identify anomalies in the data that are not associated with climatic or glaciological processes, but rather effects occurring during field work, laboratory analysis, or data post-processing. These examples make it clear that permutation entropy is a useful forensic tool for identifying sections of data that require targeted reanalysis—and can even be useful for guiding that analysis.

Highlights

Paleoclimate records, such as ice and sediment cores, provide us with long and detailed accounts of Earth’s ancient climate system
We identified abrupt changes in the complexity of these isotope records using sliding-window calculations of the permutation entropy [11] (PE) and a weighted variant of that method known as weighted permutation entropy (WPE) that is intended to balance noise levels and the scale of trends in the data [12]
As a proof of concept for the claim that information theory can be useful in detecting anomalies in paleorecords, we focused on water-isotope data from the West Antarctic Ice Sheet (WAIS) Divide ice core [9]

Summary

Introduction

Paleoclimate records, such as ice and sediment cores, provide us with long and detailed accounts of Earth’s ancient climate system Collection of these data sets can be very expensive, and the extraction of proxy data from them is often time consuming, as well as susceptible to both human and machine error. Ensuring the accuracy of these data is as challenging as it is important Most of these cores are only fully measured one time, and most are unique in the time period and region that they “observe”, making comparisons and statistical tests impossible. These records may be subject to many different effects—known, unknown, and conjectured—between deposition and collection. These challenges make it very difficult to understand how much information is present in these records, how to extract it in a meaningful way, and how best to use it while not overusing it

Methods

Results

Conclusion