89-9 Quantitative assessment of stratigraphic correlation with pyCoreRelator: Application to turbidite paleoseismology in Cascadia

Session: Advancing Earthquake Geology and Surficial Deformation from Geologic Provinces to Political Entities through Multidisciplinary High-Resolution Data

Presenting Author:

Authors:

Abstract:

Physical stratigraphic and chronological correlations are fundamental to sediment-based paleoseismology. However, objectively assessing event bed correlations between sediment cores collected from different depositional environments or across significant distances remains challenging. To address this, we introduce a new Python tool, pyCoreRelator, designed to systematically and objectively identify all feasible bed-to-bed correlation solutions between core pairs, while honoring available age constraints and accounting for lateral bed thickness variability. The tool employs the dynamic time warping (DTW) algorithm to automatically align log and other profile data and to compute signal pattern similarity metrics between core pairs. This enables evaluation of correlation quality and identification of optimal correlation solutions. To determine whether correlation solutions are geologically meaningful or simply coincidental, we assess whether these similarity metrics differ statistically from those expected for randomly stacked event beds. Building on Nieminski et al. (2024), we apply pyCoreRelator to Cascadia offshore sediment cores to test the correlatability of marine turbidites previously interpreted as synchronous, earthquake-triggered deposits. Physical property logs and image color profiles are used in multidimensional DTW correlations and for lithofacies classification via cluster analysis. For each core pair, the algorithm seeks optimal solutions by minimizing lateral bed thickness change and maximizing compatibility with reservoir-corrected radiocarbon ages. Our findings show that correlation quality improves markedly for core pairs collected in proximity and within the same depositional environment. Notably, only a small number of core pairs collected nearby within distal abyssal plain channels correlate more confidently than synthetic, randomly stacked core pairs, regardless of whether age constraints are considered. Interestingly, correlation strength of turbidites over long distances and across different depositional settings (e.g., distal channel, lower fan, slumping slope base) is not improved, and sometimes even degraded, by the inclusion of additional age constraints. These results highlight the uncertainty of event bed correlation across different depositional environments and call into doubt the assumptions of the long-distance correlatability of some Cascadia turbidites. While further work is necessary to extend our analysis across the full Cascadia margin, our approach enables robust and reproducible evaluation of stratigraphic correlations between event beds and offers new insights into the effectiveness and limitations of combining age data with stratigraphy and physical property logs.

Geological Society of America Abstracts with Program. Vol. 57, No. 6, 2025

doi: 10.1130/abs/2025AM-8466

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.