Random-Field Modeling of Subsurface Stratigraphy and Geo-properties: Applications in the Taipei Basin
DOI:
https://doi.org/10.64862/Keywords:
Uncertainty, Stratigraphic random field modeling, Markov random fields, Seismic site characterization.Abstract
Regional‐scale characterization of subsurface uncertainty is essential for basin-wide seismic assessment, yet integrating stratigraphy and geo-properties in a single probabilistic model remains uncommon in practice. To address this gap, we adopt a stratigraphic-based geo-property random field (SGRF) framework that couples a stochastic Markov random field (SMRF) for categorical stratigraphy with a hidden Markov random field (HMRF) for continuous geo-properties (e.g. void ratio , unit weight). Fractional Brownian motion (FBM) captures scale-dependent variability, and parameters are inferred via Bayesian calibration. Applied to the Taipei Basin, the framework generates ensembles of stratigraphy and geo-property fields and propagates them to shear-wave velocity (VS) and VS30 through soil-type-specific estimation functions. Stratigraphic uncertainty is quantified with information entropy, while geo-property uncertainty is summarized by ensemble moments. The resulting VS30 maps are geologically plausible and correspond well to an independent kriging baseline; where differences occur, they coincide with zones of elevated modeled uncertainty. Basin-wide analyses show that uncertainty in lithologic architecture is the dominant driver of dispersion in VS and VS30, with void-ratio variability contributing more than unit-weight variability. Overall, SGRF offers a computationally tractable route from heterogeneous geological data to probabilistic VS/VS30 mapping at a regional scale and clarifies how uncertainty propagates from stratigraphy to engineering quantities, supporting targeted data acquisition and more robust seismic design decisions.
References
Juang, C. H., Dong, J. J., and Shen, M. (2025, August 25–28). Characterisation and assessment of engineering geological model uncertainty: Geotechnical engineer’s perspective (Lacasse Lecture). In Proceedings of the 9th International Symposium on Geotechnical Safety and Risk (ISGSR 2025), Oslo, Norway.
Li, Z., Wang, X., Wang, H., and Liang, R. Y. (2016). Quantifying stratigraphic uncertainties by stochastic simulation techniques based on Markov random field. Engineering Geology, 201, 106–122. https://doi.org/10.1016/j.enggeo.2015.12.017
Lu, Y. C., Lin, Y. C., Dong, J. J., Chien, W. Y., Hung, W. Y., and Tien, Y. M. (2023). Mapping VS30 in Taipei Basin: Considering stratigraphic and geo-property spatial variabilities. Sino-Geotechnics, 178, 27–36.
Lu, Y. C., Chien, W. Y., Nicholas, S. C., Wang, H., Dong, J. J., and Juang, C. H. (2025, August 25–28). Probabilistic stratigraphic and geo-property models at a regional scale: A case study of the Taipei Basin. In Proceedings of the 9th International Symposium on Geotechnical Safety and Risk (ISGSR 2025), Oslo, Norway (Paper ID 221).
Nicholas, S. C., Nguyen, L. T. M., Kuo, C. L., Gao, J. C., Kuo, C. H., Tran, D. H., Wang, S. J., and Dong, J. J. (2025). Enhancing VS30 mapping in the Taipei Basin: Integrating new Vs estimation functions, extrapolated Vs profiles, and kriging with varying local means. Bulletin of the Seismological Society of America, 115(4), 2439–2463. https://doi.org/10.1785/0120250020
Wei, X., and Wang, H. (2022). Stochastic stratigraphic modeling using Bayesian machine learning. Engineering Geology, 307, 106789. https://doi.org/10.1016/j.enggeo.2022.106789
Downloads
Published
Issue
Section
License
Copyright (c) 2025 Nepal Society of Engineering Geology (NSEG)
This work is licensed under a Creative Commons Attribution 4.0 International License.
