Reproducible event‐bed correlation with uncertainty: pyCoreRelator

ABSTRACT Stratigraphic correlation is fundamental to Earth science, yet conventional approaches rely heavily on subjective visual data matching. Automated algorithms such as dynamic time warping (DTW) provide reproducibility, but when sequences are subdivided into clearly defined stratigraphic units, DTW alone struggles to explore a vast range of correlation possibilities among these units and evaluate the optimal alignment. To address this, we introduce pyCoreRelator , an open‐source Python package designed for quantitative and reproducible pairwise stratigraphic correlation. The package can handle core imagery and geophysical log data, and systematically explores all feasible inter‐unit correlations, honouring geochronological constraints and accounting for possible lateral discontinuities of the beds. Correlation scenarios are ranked using quantitative similarity metrics, including normalised DTW cost and Pearson's r , to identify the optimal solution that minimises the degree of variation in sedimentation rates and erosion while maximising similarity. Furthermore, pyCoreRelator statistically evaluates the geological significance of these correlations by comparing them to the expected background lithological similarity in the natural environment, represented by stochastic baselines generated through Markov Chain Monte Carlo stratigraphic simulations. We demonstrate this workflow for an event‐bed correlation case study with radiocarbon‐dated marine turbidites used previously to derive a chronology of large earthquakes in the Cascadia Subduction Zone. Pairwise analysis of closely spaced core sites yielded robust, statistically significant correlations that align with prior human interpretations. However, within the same channel system, the quality and certainty of correlations degrade as the distance between core sites increases and, in some cases, worsen with the inclusion of age constraints. These results indicate that current data cannot unambiguously confirm margin‐wide synchronous turbidite deposition, urging caution when relying on visual, long‐distance correlations to define regional earthquake histories. These results underscore the efficiency and effectiveness of pyCoreRelator in identifying meaningful stratigraphic correlations through a transparent, probabilistic approach that can be applied across diverse geoscience fields.

Authors

Institutions

Publication Details

Journal
Sedimentology
Published
2026-11-17
DOI
https://doi.org/10.1111/sed.70163
Primary Topic
Geological formations and processes
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Reproducible event‐bed correlation with uncertainty: pyCoreRelator

Zoltán Sylvester, Larry Syu‐Heng Lai, Joan S. Gomberg, Jacob A. Covault et al.
Sedimentology
Geological formations and processes
article

Reproducible event‐bed correlation with uncertainty: pyCoreRelator

Zoltán Sylvester, Larry Syu‐Heng Lai, Joan S. Gomberg, Jacob A. Covault, Nora M. Nieminski
article en

Abstract

ABSTRACT Stratigraphic correlation is fundamental to Earth science, yet conventional approaches rely heavily on subjective visual data matching. Automated algorithms such as dynamic time warping (DTW) provide reproducibility, but when sequences are subdivided into clearly defined stratigraphic units, DTW alone struggles to explore a vast range of correlation possibilities among these units and evaluate the optimal alignment. To address this, we introduce pyCoreRelator , an open‐source Python package designed for quantitative and reproducible pairwise stratigraphic correlation. The package can handle core imagery and geophysical log data, and systematically explores all feasible inter‐unit correlations, honouring geochronological constraints and accounting for possible lateral discontinuities of the beds. Correlation scenarios are ranked using quantitative similarity metrics, including normalised DTW cost and Pearson's r , to identify the optimal solution that minimises the degree of variation in sedimentation rates and erosion while maximising similarity. Furthermore, pyCoreRelator statistically evaluates the geological significance of these correlations by comparing them to the expected background lithological similarity in the natural environment, represented by stochastic baselines generated through Markov Chain Monte Carlo stratigraphic simulations. We demonstrate this workflow for an event‐bed correlation case study with radiocarbon‐dated marine turbidites used previously to derive a chronology of large earthquakes in the Cascadia Subduction Zone. Pairwise analysis of closely spaced core sites yielded robust, statistically significant correlations that align with prior human interpretations. However, within the same channel system, the quality and certainty of correlations degrade as the distance between core sites increases and, in some cases, worsen with the inclusion of age constraints. These results indicate that current data cannot unambiguously confirm margin‐wide synchronous turbidite deposition, urging caution when relying on visual, long‐distance correlations to define regional earthquake histories. These results underscore the efficiency and effectiveness of pyCoreRelator in identifying meaningful stratigraphic correlations through a transparent, probabilistic approach that can be applied across diverse geoscience fields.

Sedimentology
United States Geological Survey (US), Bureau of Economic Analysis (US), Alaska Department of Natural Resources (US)
Geological formations and processes
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon