[S-SS06] New trends in data acquisition, analysis and interpretation of seismicity

[SSS06-P02] Deep fault geometries in the Hidaka collision zone in southern Hokkaido revealed by hierarchical clustering of hypocenter distributions.

*Admore Mpuang¹, Yasunori Sawaki¹, Takahiro Shiina¹, Takahiko Uchide¹ (1.Geological Survey of Japan, AIST)

Keywords:Hokkaido , Fault plane geometries, Hidaka collision zone, Seismicity, Hierarchical clustering

The Hidaka collision zone in southern Hokkaido hosts anomalously deep focus crustal earthquakes, with large magnitude (~M7) events reportedly occurring every 40 years (Kita et al., 2014). Most recently, a magnitude 6.7 earthquake occurred at a depth of 37 km (Japan Meteorological Agency (JMA)) in the eastern Iburi region near an active Ishikari-Teichi-Toen fault zone (ITTFZ). The location and geometry of the fault plane associated with this earthquake, determined from InSAR and GNSS data, is reported to have no connection to any known surface traces of the ITTFZ (Kobayashi et al., 2019). However, with the complex crustal structure in this region (e.g., Gou et al., 2019; Iwasaki et al., 2004), it remains controversial whether this event was part of the ITTFZ. To answer this question, and thus better understand the rupture processes in the source region and potential implications on future seismic hazards and regional seismotectonics, detailed knowledge of the causative fault geometries is needed. Sawaki et al. (under review) developed a new method to reconstruct intricate fault models based on hierarchical clustering of hypocenter distributions, integrating a geometry-based feature in hypocenter distributions. We apply this method to the relocated hypocenters in the source region of the 2018 M6.7 Eastern Iburi earthquake using the HypoDD program (Waldhauser and Ellsworth, 2000) and the JMA2001 1-dimensional velocity model. Our preliminary determined fault plane results suggest a fault system with multiple fault segments, connecting to the southern sections of the ITTFZ surface traces. The strike and dip of the main fault planes are consistent with the focal mechanism solutions of 4 events with magnitude ~5 or larger, including the mainshock. At least 3 fault segments seem to be associated with the events of magnitude ~5 in the aftershock sequence between September 2018 and February 2019. While the location of the major fault plane associated with the mainshock deviates from the fault surface traces, its southern end is located beneath part of an estimated fault surface trace. Two M~5 events occur on the southern-most segment that seems to connect two estimated surface traces with approximately NW-SE strikes, suggesting that stress loading on this section of the ITTFZ, with an estimated long-term earthquake probability within 30 years of 0.2% (Kobayashi et al., 2019), is likely to have been altered. Since the relocated hypocenters are at depths greater than 10 km, the fault plane geometries in the shallower part of the crust could not be determined in the current analysis. To improve our results, we apply machine learning-based phase picking to detect and locate any additional small events that may delineate the upper extensions of the fault planes in the source region and improve fault plane determination accuracy. Our results provide key insights in understanding the intricacies of the ITTFZ fault zone and its future potential seismic risk.
Acknowledgement: This study is supported by MEXT Project for Seismology toward Research Innovation with Data of Earthquake (STAR-E) Grant Number JPJ010217. We used data from the JMA Unified Earthquake Catalog.