A quantitative framework for CT-based characterization of sedimentary variability in core samples
Journal
Terrestrial, Atmospheric and Oceanic Sciences
Journal Volume
37
Journal Issue
1
ISSN
1017-0839
2311-7680
Date Issued
2025-12-07
Author(s)
Wu, Yen-Hsi
Degenhart, Gerald
Yu, Neng-Ti
Yen, Jiun-Yee
Chyi, Shyh-Jeng
Chen, Jia-Hong
Abstract
Reconstructing past environments requires accurate interpretation of sedimentary variability in cores. Identifying features such as grain-size and compositional changes is essential for understanding depositional processes, but surface-based observations and point measurements may miss internal heterogeneity, affecting sampling representativeness. X-ray computed tomography (CT) provides a non-destructive approach to visualizing internal structures based on X-ray attenuation. While CT has been widely adopted in sedimentological studies, many applications remain qualitative and dependent on interpreter experience. This study introduces a framework integrating multiple CT-derived parameters, applied to sediment cores from Dapeng Bay, southwestern Taiwan. CT intensity values were classified into six attenuation-based regions of interest (ROIs), and parameters including volume fraction, statistical metrics (mean, coefficient of variation), and morphological descriptors (mean particle thickness of selected ROIs) were extracted on a slice-by-slice basis. These parameters were integrated to produce the high-resolution (~ 100 μm) vertical profiles that capture internal textural variability. The results demonstrate how CT-based metrics can support visual core description by quantitatively delineating sedimentary structures, improving sampling design, and enabling systematic comparison between cores. Together, these advances yield a novel, parameter-based quantitative framework that offers a reproducible, three-dimensionally informed workflow for the numerical reading of full-core CT volumes, remains transferable across settings through CT-parameter initialization and refinement, and provides a foundation for future data-driven classification.
Subjects
CT-derived parameters
Quantitative core analysis
Sedimentary features
X-ray computed tomography (CT)
Publisher
Springer Science and Business Media LLC
Type
journal article