10:45 AM - 11:00 AM
[SSS10-01] Source rupture process of the 2016 Rieti Italy earthquake evaluated from waveform inversion with empirical Green’s functions
Keywords:the 2016 Rieti Italy earthquake, rupture process, empirical Green’s functions, waveform inversion, strong ground motions
The 2016 Rieti, Italy, earthquake (ML 6.0) on August 24th 01:36:32 (UTC), 2016, occurred in the central Italy, a portion of the Central Apennines (Italy) between Norcia and Amatrice towns. The moment tensor solution, immediately released by INGV, shows normal faulting with planes striking along the Apenninic direction, the hypocenter is located at a depth of 8 km. Tragically, it caused devastating loss in the villages of Amatrice and Accumoli, and almost 300 people lost their lives due to the collapse of several buildings in the towns and villages closest to the epicenter. Therefore, it motivated us to study its rupture features, especially focusing on the generation mechanism of the strong ground motions.
This event was well recorded by permanent strong-motion networks of the RAN and INGV closest to the epicenter, forming a rich waveform database, which enables us to investigate the rupture process of the earthquake in detail. Therefore, in this analysis, the source rupture process of this event was evaluated through the waveform inversion of strong-motion data with the empirical Green’s functions (EGFs). The site conditions near the seismic region are rather complicated, applying the EGFs could avoid the possible uncertainties caused by the assumption of subsurface velocity structures, along with considering relatively higher frequency ranges of strong ground motions typically up to 2.0 Hz.
In this work, the linear least-squares waveform inversion (Nozu, 2007; Nozu and Irikura, 2008; Nozu and Nagasaka, 2017) was adopted on the strong-motion data. The inversion scheme follows the multi-time-window approach (Hartzell and Heaton, 1983). Three small events, event 1 (2016/8/24, 17:46:09, Mw 4.2), event 2 (2016/10/30, 13:34:54, Mw 4.1) and event 3 (2016/11/12, 14:43:33, Mw 4.1) were selected as EGFs, and 8 near-fault strong-motion stations within hypocenter distance of 40km are selected, which are depicted in Figure 1 (a). The source parameters and moment tensors are based on the information from the INGV.
Our preferred slip model is illustrated in Figure 1 (b~c), and the comparisons of observed and synthesized velocity waveforms (0.2 - 2.0 Hz) are illustrated in Figure 2. The final slip distribution shown in Figure 1(a) indicates that, one main large slip region with a maximum slip of approximately 1.6 m was centered ~ 4km SW of the hypocenter, and it is with a very shallow depth of ~ 1.5 km; two regions with large peak slip velocity were identified as shown in Figure 1(b), the first one with peak value of 1.6 m/s was center very close to the NE of the hypocenter, and the secondary one was located coincide with the main large slip region. The slip distributions in our preferred source model might explain that why several damages occurred in the villages of Amatrice. A rupture velocity of 3.0 km/s was identified, and the preferred source model corresponds to Mw 6.4 for this event.
This event was well recorded by permanent strong-motion networks of the RAN and INGV closest to the epicenter, forming a rich waveform database, which enables us to investigate the rupture process of the earthquake in detail. Therefore, in this analysis, the source rupture process of this event was evaluated through the waveform inversion of strong-motion data with the empirical Green’s functions (EGFs). The site conditions near the seismic region are rather complicated, applying the EGFs could avoid the possible uncertainties caused by the assumption of subsurface velocity structures, along with considering relatively higher frequency ranges of strong ground motions typically up to 2.0 Hz.
In this work, the linear least-squares waveform inversion (Nozu, 2007; Nozu and Irikura, 2008; Nozu and Nagasaka, 2017) was adopted on the strong-motion data. The inversion scheme follows the multi-time-window approach (Hartzell and Heaton, 1983). Three small events, event 1 (2016/8/24, 17:46:09, Mw 4.2), event 2 (2016/10/30, 13:34:54, Mw 4.1) and event 3 (2016/11/12, 14:43:33, Mw 4.1) were selected as EGFs, and 8 near-fault strong-motion stations within hypocenter distance of 40km are selected, which are depicted in Figure 1 (a). The source parameters and moment tensors are based on the information from the INGV.
Our preferred slip model is illustrated in Figure 1 (b~c), and the comparisons of observed and synthesized velocity waveforms (0.2 - 2.0 Hz) are illustrated in Figure 2. The final slip distribution shown in Figure 1(a) indicates that, one main large slip region with a maximum slip of approximately 1.6 m was centered ~ 4km SW of the hypocenter, and it is with a very shallow depth of ~ 1.5 km; two regions with large peak slip velocity were identified as shown in Figure 1(b), the first one with peak value of 1.6 m/s was center very close to the NE of the hypocenter, and the secondary one was located coincide with the main large slip region. The slip distributions in our preferred source model might explain that why several damages occurred in the villages of Amatrice. A rupture velocity of 3.0 km/s was identified, and the preferred source model corresponds to Mw 6.4 for this event.