2:00 PM - 2:15 PM
[SSS05-02] Constraint on the background stress field in the source region of the 2016 Kumamoto earthquake based on changes in elastic strain energies by aftershocks
Keywords:stress, elastic strain energy, the 2016 Kumamoto earthquake
In this study, we examined the deviatoric stress magnitude of the background stress field prior to the 2016 Kumamoto earthquake sequence, examining changes in elastic strain energies caused by lots of aftershocks. For this purpose, we modeled three background stress fields in the source region of the earthquake sequence with the parameter of effective friction coefficients (m’= 0.38, 0.18, and 0.04), using the 3-D Mohr diagram together with the stress orientations in the region. We also computed coseismic stress changes fields caused by the largest foreshock and mainshock, and combined them with the background stress fields to obtain the absolute stress fields immediately after the primary events.
In order to estimate changes in elastic strain energies caused by earthquakes, we need to determine the fault plane and the coseismic slip distribution of the events. However, it is not straightforward to determine the information for small to middle events. On the other hand, the changes in elastic strain energies by earthquakes can be represented by using the inner product of stress and moment tensors of earthquakes where the stress tensor represents the average stress state during the rupture (Matsu’ura and Terakawa, 2021). This representation is advantageous because we can directly observe moment tensors of earthquakes without determining the source. Using this representation and assuming the stress drop to be 5 MPa, we evaluated changes in elastic strain energies caused by 2628 aftershocks DEa (16 April 2016 to 30 April 2019) that occurred within 10 km from the source region (Mitsuoka et al., 2020; NIED Moment Tensor Catalog) under the three models of absolute stress fields immediately after the mainshock.
A part of elastic strain energies released by events is consumed as the radiated energies. The scaled energy, which is defined by the radiated energies normalized by the seismic moment, takes nearly constant value of 10-5 over the wide range of magnitudes (e.g., Ide and Beroza, 2001). Considering the energy balance that changes in elastic strain energies must be larger than the radiated energies, changes in elastic strain energies normalized by the seismic moment should be larger than the scaled energy (10-5). The ratios of events with DEa/Mo > 10-5 are 95, 91 and 33 %, and the average values of DEa/Mo are 1.0E-3, 4.4E-4 and -1.9E-5, for m’ = 0.38, 0.18, 0.04, respectively. In the model with m’ = 0.04 more than 60 % events are not only inconsistent with the energy balance but also increased elastic strain energies after the events. The results are contradictory to the fact that earthquakes are a physical process that releases elastic strain energies. We concluded that the realistic value of effective friction coefficient of background stress field is much larger than those estimated in previous studies (m’ < 0.1).
In order to estimate changes in elastic strain energies caused by earthquakes, we need to determine the fault plane and the coseismic slip distribution of the events. However, it is not straightforward to determine the information for small to middle events. On the other hand, the changes in elastic strain energies by earthquakes can be represented by using the inner product of stress and moment tensors of earthquakes where the stress tensor represents the average stress state during the rupture (Matsu’ura and Terakawa, 2021). This representation is advantageous because we can directly observe moment tensors of earthquakes without determining the source. Using this representation and assuming the stress drop to be 5 MPa, we evaluated changes in elastic strain energies caused by 2628 aftershocks DEa (16 April 2016 to 30 April 2019) that occurred within 10 km from the source region (Mitsuoka et al., 2020; NIED Moment Tensor Catalog) under the three models of absolute stress fields immediately after the mainshock.
A part of elastic strain energies released by events is consumed as the radiated energies. The scaled energy, which is defined by the radiated energies normalized by the seismic moment, takes nearly constant value of 10-5 over the wide range of magnitudes (e.g., Ide and Beroza, 2001). Considering the energy balance that changes in elastic strain energies must be larger than the radiated energies, changes in elastic strain energies normalized by the seismic moment should be larger than the scaled energy (10-5). The ratios of events with DEa/Mo > 10-5 are 95, 91 and 33 %, and the average values of DEa/Mo are 1.0E-3, 4.4E-4 and -1.9E-5, for m’ = 0.38, 0.18, 0.04, respectively. In the model with m’ = 0.04 more than 60 % events are not only inconsistent with the energy balance but also increased elastic strain energies after the events. The results are contradictory to the fact that earthquakes are a physical process that releases elastic strain energies. We concluded that the realistic value of effective friction coefficient of background stress field is much larger than those estimated in previous studies (m’ < 0.1).