Finite difference simulations of seismic wave propagation for understanding earthquake physics and predicting ground motions: Advances and challenges

Aochi, Hideo; Ulrich, Thomas; Ducellier, Ariane; Dupros, Fabrice; Michéa, David

doi:10.1088/1742-6596/454/1/012010

Cited by 20 publications

(14 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this study, we calculate the ground motions in the 3‐D structure using a finite difference method [ Aochi and Madariaga , ; Aochi et al ., ] according to different finite source scenarios dynamically simulated. For the purpose mentioned above, we adopt PGV as the primary intensity measure as we find that our tests on moderate earthquakes ( M of about 5) show a consistency among the simulation, the observations, and the GMPEs (section 2.3) and also because PGV is sensible at lower frequencies (around 1 Hz), which we calculate than PGA.…”

Section: Methodsmentioning

confidence: 99%

“…The ground‐motion simulations are carried out in a 3‐D structural model using a finite difference method, as in AU15. The finite difference is based on approximations that are fourth order in space and second order in time for a staggered grid scheme [ Aochi and Madariaga , ; Aochi et al ., , and references therein]. In the model domain of 200 km (EW) by 120 km (NS) by 40 km (UD), the 3‐D structural model combines a 3‐D tomographic model [ Bayrakci et al ., ], bathymetry from the General Bathymetric Chart of the Ocean, and a 1‐D layered model (H. Karabulut, personal communication, 2012; see Figure ).…”

Section: Methods and Modelmentioning

confidence: 99%

See 1 more Smart Citation

Stress accumulation in the Marmara Sea estimated through ground‐motion simulations from dynamic rupture scenarios

2017

Self Cite

View full text Add to dashboard Cite

We compare ground motions simulated from dynamic rupture scenarios, for the seismic gap along the North Anatolian Fault under the Marmara Sea (Turkey), to estimates from empirical ground motion prediction equations (GMPEs). Ground motions are simulated using a finite difference method and a 3‐D model of the local crustal structure. They are analyzed at more than a thousand locations in terms of horizontal peak ground velocity. Characteristics of probable earthquake scenarios are strongly dependent on the hypothesized level of accumulated stress, in terms of a normalized stress parameter T. With respect to the GMPEs, it is found that simulations for many scenarios systematically overestimate the ground motions at all distances. Simulations for only some scenarios, corresponding to moderate stress accumulation, match the estimates from the GMPEs. The difference between the simulations and the GMPEs is used to quantify the relative probabilities of each scenario and, therefore, to revise the probability of the stress field. A magnitude Mw7+ operating at moderate prestress field (0.6 < T ≤ 0.7) is statistically more probable, as previously assumed in the logic tree of probabilistic assessment of rupture scenarios. This approach of revising the mechanical hypothesis by means of comparison to an empirical statistical model (e.g., a GMPE) is useful not only for practical seismic hazard assessments but also to understand crustal dynamics.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Methods and Modelmentioning

confidence: 99%

Stress accumulation in the Marmara Sea estimated through ground‐motion simulations from dynamic rupture scenarios

2017

Self Cite

View full text Add to dashboard Cite

show abstract

“…In the next section we present the standard sequential algorithm for seismic wave propagation simulations. Aochi et al [20] provide a more detailed description of the equations governing these simulations in the case of an elastic material.…”

Section: Seismic Wave Propagationmentioning

confidence: 99%

Seismic wave propagation simulations on low-power and performance-centric manycores

et al. 2016

Self Cite

View full text Add to dashboard Cite

The large processing requirements of seismic wave propagation simulations make High Performance Computing (HPC) architectures a natural choice for their execution. However, to keep both the current pace of performance improvements and the power consumption under a strict power budget, HPC systems must be more energy e cient than ever. As a response to this need, energye cient and low-power processors began to make their way into the market. In this paper we employ a novel low-power processor, the MPPA-256 manycore, to perform seismic wave propagation simulations. It has 256 cores connected by a NoC, no cache-coherence and only a limited amount of on-chip memory. We describe how its particular architectural characteristics influenced our solution for an energy-e cient implementation. As a counterpoint to the low-power MPPA-256 architecture, we employ Xeon Phi, a performance-centric manycore. Although both processors share some architectural similarities, the challenges to implement an e cient seismic wave propagation kernel on these platforms are very di↵erent. In this work we compare the performance and energy e ciency of our implementations for these processors to proven and optimized solutions for other hardware platforms such as general-purpose processors and a GPU. Our experimental results show that MPPA-256 has the best energy e ciency, consuming at least 77 % less energy than the other evaluated platforms, whereas the performance of our solution for the Xeon Phi is on par with a state-of-the-art solution for GPUs.

show abstract

“…For ground motion simulations, we use a fourth-order finite-difference method (Aochi et al 2013a , b ) with a grid spacing of Δs = 80 m for a dimension of 110 km (EW) × 120 km (NS) × 30 km (UD). The ground surface is approximated as flat, and the Sea of Japan is not taken into account.…”

Section: Numerical Simulations Of Ground Motionmentioning

confidence: 99%

Seismological asperities from the point of view of dynamic rupture modeling: the 2007 Mw6.6 Chuetsu-Oki, Japan, earthquake

Aochi

Yoshimi

2016

J Seismol

Self Cite

View full text Add to dashboard Cite

We study the ground motion simulations based on three finite-source models for the 2007 Mw6.6 Niigata Chuetsu-oki, Japan, earthquake in order to discuss the performance of the input ground motion estimations for the near-field seismic hazard analysis. The three models include a kinematic source inverted from the regional accelerations, a dynamic source on a planar fault with three asperities inferred from the very-near-field ground motion particle motions, and another dynamic source model with conjugate fault segments. The ground motions are calculated for an available 3D geological model using a finite-difference method. For the comparison, we apply a goodness-of-fit score to the ground motion parameters at different stations, including the nearest one that is almost directly above the ruptured fault segments. The dynamic rupture models show good performance. We find that seismologically inferred earthquake asperities on a single fault plane can be expressed with two conjugate segments. The rupture transfer from one segment to another can generate a significant radiation; this could be interpreted as an asperity projected onto a single fault plane. This example illustrates the importance of the fault geometry that has to be taken into account when estimating the very-near-field ground motion.

show abstract

Finite difference simulations of seismic wave propagation for understanding earthquake physics and predicting ground motions: Advances and challenges

Cited by 20 publications

References 28 publications

Stress accumulation in the Marmara Sea estimated through ground‐motion simulations from dynamic rupture scenarios

Stress accumulation in the Marmara Sea estimated through ground‐motion simulations from dynamic rupture scenarios

Seismic wave propagation simulations on low-power and performance-centric manycores

Seismological asperities from the point of view of dynamic rupture modeling: the 2007 Mw6.6 Chuetsu-Oki, Japan, earthquake

Contact Info

Product

Resources

About