Rupture Propagation and Segmentation of the 2017 Pohang Earthquake

The 2017 Mw 5.4 Pohang earthquake has been suggested to be induced by fluid injection from the nearby Pohang enhanced geothermal system (EGS). We installed eight temporary seismic stations around the EGS site five days before the mainshock for a precise seismic monitoring. We report here rupture propagation and segmentation of the Pohang earthquake inferred from temporal and spatial distribution of hypocenters (depth range: 2.5-7 km) including 6 foreshocks and 2,829 aftershocks as well as the mainshock recorded by our local seismic array for 168 days. Using 3-hour seismic data after the mainshock, Kim et al. (https://doi.org/10.1126/science.aat6081) showed that the initial rupture consists of a main segment (N35E (strike)/55NW (dip), 3.5 km long; MS) and a subsidiary one (N20E/50NW, 1.5 km long; SS1) extending from the northeastern tip of MS. The 167-day aftershock sequence shows that the frequency decreases with the reciprocal of time, following Omori's law, until 83 days after the mainshock and then there was a sudden surge of aftershock frequency followed by another Omori's-law decay. The second subsidiary segment (N13W/75SW, 2 km long; SS2) has propagated from the northern tip of SS1. Toward southwest, the main segment (MS) has propagated further ( 400 m more) from the first 3-hour configuration. Then, the third subsidiary segment (N83W/77NE, 3 km long; SS3) has developed at the southwestern tip of MS, almost perpendicular to MS which does not extend across SS3. That is, the propagation of MS was truncated by SS3. The timing of SS3 development corresponds to the aftershock surge at 83 days after the mainshock described above. One possible explanation for this pattern of rupture propagation and segmentation is that fluid migrating along MS encountered a fluid barrier (gouge layer along NW-striking fault?) at the southwestern tip of MS, leading to buildup of pore pressure there followed by SS3 rupturing.