Chemical-Based Event-Stratigraphic Correlation along the Japan Trench by XRF-CS Chemical Fingerprint and Multivariate Statistics

Megathrust earthquakes in subduction zones, such as the AD 2011 Tohoku-oki earthquake, are known to generate turbidity currents that transport sediment into deep marine trenches, creating distinctive event deposits. These deposits are pivotal in submarine paleoseismology, which seeks to extend earthquake records and assess disaster potentials by meticulously analyzing the spatiotemporal distribution of these deposits through precise distinction and correlation. In our research, over 800 meters of sediment cores from 15 sites along the Japan Trench were collected during the International Ocean Discovery Program (IODP) Expedition 386, focusing on these event deposits to trace earthquake history. Utilizing X-ray Fluorescence Core Scanning (XRF-CS), we performed efficient and non-destructive high-resolution analysis of the chemical characteristics of these deposits. Our methodology integrates XRF-CS data with multivariate statistical techniques, such as Principal Component Analysis (PCA) and Cluster Analysis (CA). This integration enables us to objectively differentiate and correlate event deposits based on their unique chemical properties. Significant discoveries were made at Site M0083, where we detected variations in background sediments and event deposits associated with major historical earthquakes, including the AD 1454 Kyotoku and AD 869 Jogan events. These unique chemical fingerprints were also traced to adjacent trench-fill basins at sites M0089 and M0090, revealing consistent event-stratigraphic sequences across the basins and affirming the efficacy of our chemical-based correlation technique. In our ongoing analysis, we aim to explore the diverse chemical fingerprints of background sediments, turbidite bases, and turbidite tails. This investigation seeks to uncover potential spatial and stratigraphic provenance changes, and includes examining mineral and biogenic compositions, such as clay minerals and smear slides, to elucidate the reasons behind variations in chemical signals. Our findings underscore the effectiveness of combining chemical analysis with statistical methods for event-stratigraphic correlation. This novel approach not only sheds light on provenance changes but also helps establish a detailed spatiotemporal distribution framework for event deposits along the Japan Trench. Moreover, this integrated methodology could inform subsequent sampling strategies by objectively selecting sampling locations based on the chemostratigraphy framework. It also has the potential to be adapted to other research areas that require event-stratigraphic correlation and comprehensive spatiotemporal analysis.