Speleothem stable isotope records for east-central Europe: resampling sedimentary proxy records to obtain evenly spaced time series with spectral guidance

István Gábor Hatvani,Attila Demény,Zoltán Kern,Szabolcs Leél-Őssy

doi:10.5194/essd-10-139-2018

Abstract

Abstract. Uneven spacing is a common feature of sedimentary paleoclimate records, in many cases causing difficulties in the application of classical statistical and time series methods. Although special statistical tools do exist to assess unevenly spaced data directly, the transformation of such data into a temporally equidistant time series which may then be examined using commonly employed statistical tools remains, however, an unachieved goal. The present paper, therefore, introduces an approach to obtain evenly spaced time series (using cubic spline fitting) from unevenly spaced speleothem records with the application of a spectral guidance to avoid the spectral bias caused by interpolation and retain the original spectral characteristics of the data. The methodology was applied to stable carbon and oxygen isotope records derived from two stalagmites from the Baradla Cave (NE Hungary) dating back to the late 18th century. To show the benefit of the equally spaced records to climate studies, their coherence with climate parameters is explored using wavelet transform coherence and discussed. The obtained equally spaced time series are available at https://doi.org/10.1594/PANGAEA.875917.

Highlights

With more than a hundred speleothem-related studies published every year, it is trivial to state that speleothems are one of the most important objects of paleoclimate research
Data preprocessing was necessary to ensure a time series is evenly spaced in time (Sect. 2.2.1) in order to find the areas with common powers between the speleothem stable isotope time series and the climatic data in the time–frequency space
From a practical point of view, the power spectrum density (PSD) graphs of the Wavelet transform coherence (WTC) analysis between the stable isotope time series and the monthly climate data were calculated to find the months with the highest response

Summary

Introduction

With more than a hundred speleothem-related studies published every year, it is trivial to state that speleothems are one of the most important objects of paleoclimate research. Hatvani et al.: Evenly spaced speleothem δ13C–δ18O records obtained with spectral guidance environment and carbonate precipitation are determined in a multi-year study (e.g., Breitenbach et al, 2015; Mattey et al, 2008, 2016; Suricet al., 2018; Riechelmann et al, 2011) These can be used in forward modeling to quantify and understand the role of the dominant factors regulating the geochemical signal in speleothems (Baker and Bradley, 2010). Despite all efforts, uneven spacing is a common feature of sedimentary paleoclimate records This characteristic usually biases the results of classical statistical methods and prohibits the application of time series analysis tools in many cases (Cosford et al, 2008; Mudelsee, 2010).

Data sources

Interpolation and resampling with spectral guidance

Wavelet transform coherence

Data preprocessing before WTC analysis