10:45 〜 12:15
[SSS05-P01] Megathrust and tsunami hazard mitigation with Physical Wavelets
キーワード：Megathrust earthquake prediction, Tsunami prediction, Deterministic earthquake prediction , Physical Wavelets
Physical Wavelets (PWs) can describe the processes generating megathrust and significant earthquakes (EQs) in the Tohoku subduction zone using stochastic time series data collected from Japan's seismograph [1] and GPS [2] networks. The processes predict the locations, sizes, fault movements, and timings of the EQs up to three months in advance [3-6]. The PWs can also describe a tsunami genesis, allowing for real-time tsunami disaster warnings up to three months in advance [3,4]. Using a deterministic model instead of a solely probabilistic approach is crucial in effective hazard mitigations.
Deterministic models
Any physics-based approach to mitigations should have a deterministic model, as Keiiti Aki stated [5]: I attended an international meeting in Spain in October 2004, celebrating the centennial anniversary of an old observatory. Don Turcotte was there and gave a talk concluding that seismicity is a chaotic noise. I started my talk by saying, "I accept that completely as phenomena originating from the brittle part. Our data are dominated by the events from the brittle part. We need to find faint signals from the ductile part which can be modeled deterministically."
Equations with PWs
The well-known EQ size-frequency distributions of foreshocks and aftershocks of the 2011 Tohoku M9 EQ for M ≧ 3.5 show a subtle depth dependence of the ductile and brittle transition region [5]. The PWs describe the scale dependence as an equation of how changes in principal stress levels generate significant and megathrust EQs in a region. The equation models consecutive EQ events as a virtual EQ particle motion under crustal stress changes. The region's size and shape are selective, and it may be small or large and covered by a seismic network. Each EQ event detected by the seismic network has the property of the focus (in latitude LAT, longitude LON, and depth DEP), its origin time (event time), and magnitude MAG. The interval between consecutive event times is the inter-event interval (INT). In the c-coordinate space (c = LAT, LON, DEP, INT, and MAG), an event is a virtual EQ particle of unit mass that emerges and moves to a new location at the next event. A time history of the movement is {c}={d (c, 1 ), ., d (c, m ),. }. The d(c, m) is the particle position at time m that is the chronological event index m having no INT at m = 0. The movement is stochastic, like a Brownian particle [7]. However, d (c, m) is a function of the principal stress components of the region for which a selection of MAG ≧ Mc (Mc ≈ 3.5) reduces the stochastic noise level to about 15 - 25 % [5]. The cross-correlation of PWs with the {c} defines the noise-free displacement D (c, τ) and acceleration A (c, τ) at time τ, expressing the EQ scale-dependence by a periodic equation of A(c, τ) = −K(c)×D (c, τ) with a weakly-time-dependent positive constant K(c) at the index time τ. The A (c, τ) shows significant and megathrust EQ genesis processes, which can predict the EQs along with the strain-energy cycles [5].
Similarly, PWs describe the megathrust EQ and Tsunami genesis process with non-differentiable displacement {c} observed at the Tohoku GPS stations [4]. The c is the geographic axis denoted by E (eastward), N (northward), and h (upward) in (E, N, h), having {c}={d (c, 0 ), ., d (c, j ), .}. The j is the time in days.
Issues
Scientific and societal issues must be addressed to effectively mitigate EQ hazards and inform decision-making processes [8]. We must also solve the problems of the questionable media role and disaster countermeasures in Appx C of [4].
[1] https://www.hinet.bosai.go.jp/?LANG=en
[2] https://mekira.gsi.go.jp/index.en.html
[3] Takeda, F. (2015) https://patents.google.com/patent/JP5798545B2/ja
[4] Takeda, F. (2021) https://doi.org/10.48550/arXiv.2107.02799
[5] Takeda, F. (2022) https://doi.org/10.48550/arXiv.2201.02815
[6] Takeda, F. (2022) https://doi.org/10.48550/arXiv.2208.09486
[7] Takeo, M. (1999) Disperse systems, Wiley-VCH
[8] https://www.epi21.org
Deterministic models
Any physics-based approach to mitigations should have a deterministic model, as Keiiti Aki stated [5]: I attended an international meeting in Spain in October 2004, celebrating the centennial anniversary of an old observatory. Don Turcotte was there and gave a talk concluding that seismicity is a chaotic noise. I started my talk by saying, "I accept that completely as phenomena originating from the brittle part. Our data are dominated by the events from the brittle part. We need to find faint signals from the ductile part which can be modeled deterministically."
Equations with PWs
The well-known EQ size-frequency distributions of foreshocks and aftershocks of the 2011 Tohoku M9 EQ for M ≧ 3.5 show a subtle depth dependence of the ductile and brittle transition region [5]. The PWs describe the scale dependence as an equation of how changes in principal stress levels generate significant and megathrust EQs in a region. The equation models consecutive EQ events as a virtual EQ particle motion under crustal stress changes. The region's size and shape are selective, and it may be small or large and covered by a seismic network. Each EQ event detected by the seismic network has the property of the focus (in latitude LAT, longitude LON, and depth DEP), its origin time (event time), and magnitude MAG. The interval between consecutive event times is the inter-event interval (INT). In the c-coordinate space (c = LAT, LON, DEP, INT, and MAG), an event is a virtual EQ particle of unit mass that emerges and moves to a new location at the next event. A time history of the movement is {c}={d (c, 1 ), ., d (c, m ),. }. The d(c, m) is the particle position at time m that is the chronological event index m having no INT at m = 0. The movement is stochastic, like a Brownian particle [7]. However, d (c, m) is a function of the principal stress components of the region for which a selection of MAG ≧ Mc (Mc ≈ 3.5) reduces the stochastic noise level to about 15 - 25 % [5]. The cross-correlation of PWs with the {c} defines the noise-free displacement D (c, τ) and acceleration A (c, τ) at time τ, expressing the EQ scale-dependence by a periodic equation of A(c, τ) = −K(c)×D (c, τ) with a weakly-time-dependent positive constant K(c) at the index time τ. The A (c, τ) shows significant and megathrust EQ genesis processes, which can predict the EQs along with the strain-energy cycles [5].
Similarly, PWs describe the megathrust EQ and Tsunami genesis process with non-differentiable displacement {c} observed at the Tohoku GPS stations [4]. The c is the geographic axis denoted by E (eastward), N (northward), and h (upward) in (E, N, h), having {c}={d (c, 0 ), ., d (c, j ), .}. The j is the time in days.
Issues
Scientific and societal issues must be addressed to effectively mitigate EQ hazards and inform decision-making processes [8]. We must also solve the problems of the questionable media role and disaster countermeasures in Appx C of [4].
[1] https://www.hinet.bosai.go.jp/?LANG=en
[2] https://mekira.gsi.go.jp/index.en.html
[3] Takeda, F. (2015) https://patents.google.com/patent/JP5798545B2/ja
[4] Takeda, F. (2021) https://doi.org/10.48550/arXiv.2107.02799
[5] Takeda, F. (2022) https://doi.org/10.48550/arXiv.2201.02815
[6] Takeda, F. (2022) https://doi.org/10.48550/arXiv.2208.09486
[7] Takeo, M. (1999) Disperse systems, Wiley-VCH
[8] https://www.epi21.org