| Subevent 2 | Subevent 1 |
---|

Reference point latitude [°N] | 37.77^{c} | 37.6–38.1^{d} |

Reference point longitude [°E] | 143.81^{c} | 143.8–144.5^{d} |

Reference point depth [km]^{a} | 19.5 | 57.8 |

Strike [°]^{a} | 189 | 158 |

Dip [°]^{a} | 50 | 59 |

Rake [°]^{a} | −90 | 48 |

*L*_{1} [km]^{b} | 0, 5, 10, ..., 50^{e} | 15 |

*L*_{2} [km]^{b} | 0, 5, 10, ..., 50^{e} | 15 |

*W*_{1} [km]^{b} | 10, 15, 20, ..., 40^{e} | 10 |

*W*_{2} [km]^{b} | 10, 15^{e} | 10 |

Slip amount [m]^{b} | Adjusted so that the VR values become maximal |

^{a}Strike, dip, rake angles, and reference point depth were fixed to the GCMT value^{b}Rectangular fault model with uniform slip is assumed. The fault length (*L* = *L*_{1} + *L*_{2}) was assumed to be greater than the fault width (*W* = *W*_{1} + *W*_{2})^{c}A point 2 km west of the GCMT centroid^{d}Horizontal location of reference point was searched within the range of ± 30 km in the EW and NS direction from the centroid location of Lay et al. (2013) (37.82°N, 144.13°E), with increments of 5 km^{e}Increments for the fault length and width are 5 km