A Bayesian approach to age estimation of marine terraces and implications for the history of the great Kanto earthquakes, central Japan
2021
Abstract We quantitatively evaluated the emergence ages of the tectonically uplifted marine terraces, called the Numa terraces, in the southernmost part of the Boso Peninsula, central Japan. The combination of the complete dataset of the geological dating survey and the model inversion method is newly proposed. Along the Sagami Trough, the subduction boundary of the Philippine Sea Plate, M8 class interplate earthquakes are known to have repeatedly occurred and caused huge crustal deformation in adjacent regions, leaving tectonic landforms behind, such as the Numa terraces. Although many geological studies have investigated the Numa terraces, the estimation of their emergence ages contains ambiguity due to employed qualitative judgments and separate evaluations at different survey points. In this study, we first construct a unified dataset for dating throughout the southern Boso area by compiling the existing data and our new data. We also examin the lateral continuity of the terraces using a new technique to amplify subtle topographic changes in digital elevation data. Second, we propose a new method to estimate the emergence ages of terraces using the temporal distribution of the sample ages via a model of the sedimentation process and Bayesian statistic. The emergence ages of the three levels of the Numa terraces, uplifted in the prehistorical era, are determined as 5855–5455 yBP, 3345–3025 yBP, and 2125–1415 yBP, in descending order. This result robustly shows that the recurrence intervals of terrace-forming great Kanto earthquakes can vary by more than a factor of two. The robustness of the result is supported by the quantitative evaluations of the formation ages and the errors in the sedimentation process, made possible by our method for the first time. This result will provide inevitable information for further subduction zone researches and future hazard assessments.
Keywords:
- Correction
- Source
- Cite
- Save
- Machine Reading By IdeaReader
33
References
0
Citations
NaN
KQI