Evolutive methods in cyclostratigraphy: probing for Milankovic cycles in continental (complex) settings

Cyclostratigraphy is an integral part of many scientific studies on the age and duration of outcrop- and core material from sedimentary geoarchives. Yet, borehole data are not systematically assessed using cyclostratigraphic methods. This has various reasons, including (a) a specific resolution and commonly no possibility to increase data resolution after logging, (b) logging proxy data cannot be connected to the sedimentary environment as easily as core investigations, (c) commonly cyclostratigraphic studies focus on one lithostratigraphical unit, but borehole logs may comprise several (d) some data generated from core material (e.g. stable isotope ratios) cannot be acquired in boreholes directly.
To obtain a reliable understanding of (long) borehole logging datasets, a good understanding of the potential and specifics of relevant (time/depth) evolutive methods in cyclostratigraphy are an essential prerequisite. Therefore, initially we compare the suitability of several evolutive cyclostratigraphic methods using several artificial datasets consisting of modelled Milankovic signals and noise. The principles of spectral moments, and other types of signal characterizations, can be used for initial assessment of signal properties over the entire record. Wavelet analysis and evolutive harmonic analysis (EHA) represent windowed approaches of assessing cyclicity, where wavelet analysis can assess amplitude variations. Evolutive Astronomical Spectral Misfit (eAsm) and evolutive correlation analysis (eCOCO) assess the similarity of power spectra (eCOCO) and significant cyclic variations (ASM) in geological datasets against Milankovic targets. The timeOpt method investigates precession- and eccentricity amplitude modulations and aims at finding a best fit through assessing various sedimentation rates. The astronomical component estimation (ACE) approach can be used to extract Milankovic signals from datasets.
In a second step, we apply methods to several IODP and ICDP datasets, and compare the performance of selected methods in artificial and real cases.
Aim of this work is the comparison of different evolutive cyclostratigraphic methods for an understanding of which methods can resolve specific data challenges. This work represents a first step towards an assessment of method suitability for real cases.