1:45 PM - 2:00 PM
[SCG70-01] Physics of Near-Source Strong Ground Motions
★Invited papers
Several theories have been proposed for the cause of near-source strong ground motion. Among them, the rupture directivity effect is one that has been long studied and widely accepted. Its physics is clear and the mechanism described in this theory is that seismic ground motions radiated from subfaults along the progress of fault rupture yield a strong ground motion in the direction of rupture propagation due to constructive interference. Meanwhile, the fling step effect has also been proposed, but its physics is unclear. It is written that “The other (cause) is due to the movement of the ground associated with the permanent offset of the ground” in Bolt and Abrahamson (2003). However, since both the permanent offset and strong ground motion are results of earthquake faulting, one of them cannot be a cause of the other. Also, as Hisada (2003) showed without intention, they are naturally contained in seismic ground motions calculated from earthquake faulting. If we dare to distinguish it as a special one, it can be related to the near-field term of seismic ground motion that contributes greatly near the source region. Nevertheless, it is necessary to present a reason if the near-field term becomes strong ground motion. In addition, it has been explained that the earthquake faulting motion itself appears in seismic ground motion. There is no objection on this explanation, but this can apply to all kinds of seismic ground motion, because all of them are the results of earthquake faulting. Therefore, this cannot explain why strong ground motion occurs, like the fling step effect.
It has been thought that rupture directivity occurs if a strike-slip or dip-slip rupture propagates to a site in the along-strike or updip direction, respectively. Its effects appear in the fault-normal component of seismic ground motion. However, though the 2015 Gorkha earthquake was a dip-slip event, the along-strike rupture propagation generated large ground motion pulses. For the 2016 Kumamoto earthquake, large ground motion pulses were found in the fault-parallel component. Accordingly, the fling step effect and faulting motion theory have been advocated, but our studies showed the rupture directivity effect to be a main cause of the large pulses. The first key point is that a ‘large’ ground motion pulse has to occur for the identification of rupture directivity. Although constructive interference can occur in any rupture direction, a large ground motion pulse is not generated if subfaults along the direction radiate only small ground motions or their focal mechanisms vary largely. For a typical dip slip of 45 degree, large ground motions are not generated along the strike direction because of a nodal plane in the S-wave radiation pattern, and therefore, the rupture directivity is not visible during the along-strike rupture propagation of a typical dip-slip earthquake. Since the Gorkha earthquake is a low-angle dip-slip event of 10 degree, the strike direction got away from the nodal plane and into a zone of large ground motion in the radiation pattern, causing the rupture directivity effect. The second key point is that an overall rupture direction is not always similar to the rupture direction around a zone of large slip. The Kumamoto earthquake is this case so that the upward rupture propagation around its large-slip zone generated directivity pulses in the fault-parallel component.
To compare the rupture directivity and fling step contributions to ground motions in the near field of the Gorkha earthquake, we calculated ground motions using the far- and intermediate-field terms, or only the near-field term of the analytical solution in an infinite medium. The results of this calculation in the figure show that, even in the near field, the rupture directivity effect is mostly larger than the fling step effect. It is noted the most that the shape of a fling step pulse is mainly controlled by constructive interference like a rupture directivity pulse.
It has been thought that rupture directivity occurs if a strike-slip or dip-slip rupture propagates to a site in the along-strike or updip direction, respectively. Its effects appear in the fault-normal component of seismic ground motion. However, though the 2015 Gorkha earthquake was a dip-slip event, the along-strike rupture propagation generated large ground motion pulses. For the 2016 Kumamoto earthquake, large ground motion pulses were found in the fault-parallel component. Accordingly, the fling step effect and faulting motion theory have been advocated, but our studies showed the rupture directivity effect to be a main cause of the large pulses. The first key point is that a ‘large’ ground motion pulse has to occur for the identification of rupture directivity. Although constructive interference can occur in any rupture direction, a large ground motion pulse is not generated if subfaults along the direction radiate only small ground motions or their focal mechanisms vary largely. For a typical dip slip of 45 degree, large ground motions are not generated along the strike direction because of a nodal plane in the S-wave radiation pattern, and therefore, the rupture directivity is not visible during the along-strike rupture propagation of a typical dip-slip earthquake. Since the Gorkha earthquake is a low-angle dip-slip event of 10 degree, the strike direction got away from the nodal plane and into a zone of large ground motion in the radiation pattern, causing the rupture directivity effect. The second key point is that an overall rupture direction is not always similar to the rupture direction around a zone of large slip. The Kumamoto earthquake is this case so that the upward rupture propagation around its large-slip zone generated directivity pulses in the fault-parallel component.
To compare the rupture directivity and fling step contributions to ground motions in the near field of the Gorkha earthquake, we calculated ground motions using the far- and intermediate-field terms, or only the near-field term of the analytical solution in an infinite medium. The results of this calculation in the figure show that, even in the near field, the rupture directivity effect is mostly larger than the fling step effect. It is noted the most that the shape of a fling step pulse is mainly controlled by constructive interference like a rupture directivity pulse.