10:45 〜 12:15
[SCG45-P26] What is behind the heterogeneities in the migration patterns of deep LFE activity in Nankai?
キーワード：PINN model, migration pattern, LFE swarms
LFE’s usually occur as dense groups or bursts which propagate along the subduction
zone, mostly horizontally (along-strike), sometimes vertically (along dip) at distances of
up to 200 km, with average speeds of 10-15 km/day. In the last 20 years, deep LFE’s
activity was consistently detected within a belt like structure at 20-40 km deep, while at
shallower depths, closer to the Nankai Trench area, shallow LFEs and SSEs were
discovered more recently (Araki et al., 2017; Nakano et al., 2018; Takemura et al.,
2019, Yamamoto et al., 2021).
Deep-learning algorithms have the potential to model and predict complex systems
which might be difficult to achieve using large temporal and spatial data. Using this
approach, we try to distinguish between two hypotheses that can explain what
mechanism controls the migration of bursts. For the first model, we assume that small
patches of ductile material (Ando et al., 2010, Nakata et al., 2012) are responsible for
the cascade propagation of LFE bursts via stress diffusion (Ariyoshi et al., 2014). Thus,
the variations in the propagation of LFE fronts are controlled by structural
heterogeneities at the plate interface. Considering the Philippines plate was sliding
about 1 meter in the last 20 years, if these heterogeneities control LFE propagation, we
can obtain a map of their geographical location.
The second model implies fluid diffusion propagation, which is well known from fluid-
induced and volcanic seismicity as a parabolic propagation. New research suggests
that the migration patterns of fluid induced micro-earthquake swarms is directly related
to how close from failure is the stress state of a region (De Baros et al., 2021). The
front will be controlled by the accumulated and released stress in the slow earthquake
cycle, lasting 3 to 6 months, not by the location of the heterogeneities.
For the first model, if the structural heterogeneities are responsible for the variations in
the migration of swarms, the heterogeneities within the tremor belt are corresponding
to a consistent increase/decrease in the rupture velocity or diffusivity. In the case of
stress/fluid diffusion-mediated transport, such heterogeneities do not control the
propagation pattern, but rather the stress does. If the area is under high stress, the
migration will show accelerating front (linear pattern). The swarm front will be mostly
diffusive if the stress state is low. Using the data-driven application of a PINN model
(Camporeale et al., 2022), we aim to clarify which model best describes the migration
of LFE swarms to contribute to the understanding of the heterogeneities across Nankai
subduction zone for producing better forecasting models.
zone, mostly horizontally (along-strike), sometimes vertically (along dip) at distances of
up to 200 km, with average speeds of 10-15 km/day. In the last 20 years, deep LFE’s
activity was consistently detected within a belt like structure at 20-40 km deep, while at
shallower depths, closer to the Nankai Trench area, shallow LFEs and SSEs were
discovered more recently (Araki et al., 2017; Nakano et al., 2018; Takemura et al.,
2019, Yamamoto et al., 2021).
Deep-learning algorithms have the potential to model and predict complex systems
which might be difficult to achieve using large temporal and spatial data. Using this
approach, we try to distinguish between two hypotheses that can explain what
mechanism controls the migration of bursts. For the first model, we assume that small
patches of ductile material (Ando et al., 2010, Nakata et al., 2012) are responsible for
the cascade propagation of LFE bursts via stress diffusion (Ariyoshi et al., 2014). Thus,
the variations in the propagation of LFE fronts are controlled by structural
heterogeneities at the plate interface. Considering the Philippines plate was sliding
about 1 meter in the last 20 years, if these heterogeneities control LFE propagation, we
can obtain a map of their geographical location.
The second model implies fluid diffusion propagation, which is well known from fluid-
induced and volcanic seismicity as a parabolic propagation. New research suggests
that the migration patterns of fluid induced micro-earthquake swarms is directly related
to how close from failure is the stress state of a region (De Baros et al., 2021). The
front will be controlled by the accumulated and released stress in the slow earthquake
cycle, lasting 3 to 6 months, not by the location of the heterogeneities.
For the first model, if the structural heterogeneities are responsible for the variations in
the migration of swarms, the heterogeneities within the tremor belt are corresponding
to a consistent increase/decrease in the rupture velocity or diffusivity. In the case of
stress/fluid diffusion-mediated transport, such heterogeneities do not control the
propagation pattern, but rather the stress does. If the area is under high stress, the
migration will show accelerating front (linear pattern). The swarm front will be mostly
diffusive if the stress state is low. Using the data-driven application of a PINN model
(Camporeale et al., 2022), we aim to clarify which model best describes the migration
of LFE swarms to contribute to the understanding of the heterogeneities across Nankai
subduction zone for producing better forecasting models.