Arthur D. Frankel

Rupture History of the 2011 M 9 Tohoku Japan Earthquake Determined from Strong‐Motion and High‐Rate GPS Recordings: Subevents Radiating Energy in Different Frequency Bands

Abstract Strong‐motion records from KiK‐net and K‐NET, along with 1xa0sample/s Global Positioning System (GPS) records from GEONET, were analyzed to determine the location, timing, and slip of subevents of the M xa09 2011 Tohoku earthquake. Timing of arrivals on stations along the coast shows that the first subevent was located closer to the coast than subevent (2), which produced the largest slip. A waveform inversion of data from 0 to 0.2xa0Hz indicates that the first subevent primarily ruptured down‐dip and north of the hypocenter and had an M of 8.5. The areas of this subevent that generated the low ( 0.2u2009u2009Hz) frequency energy are located in the same vicinity. The inversion result for the second subevent ( M xa09.0) has large slip on the shallow part of the fault with peak slip of about 65xa0m above about 25xa0km depth. This slip generated the tsunami. The preferred inversion has initiation of subevent 2 on the shallow portion of the fault so that rupture proceeded down‐dip and mainly to the south. Subevent 2 started about 35xa0s after subevent 1, which allows for the possibility of dynamic triggering from subevent 1. The slip model predicts displacements comparable to those found from ocean‐bottom transducers near the epicenter. At frequencies that most affect tall buildings (0.1–0.5xa0Hz), there is a strong pulse (subevent 3) in the strong‐motion records that arrives after the near‐field ramp from subevent 2. High‐frequency subevent 3 was located down‐dip and south of the high‐slip portion of subevent 2 and was initiated as rupture from subevent 2 proceeded down‐dip. The compact pulse for subevent 3 is modeled with an M xa08.0 source in a 75 by 30xa0km area that ruptured down‐dip and to the south with a high slip velocity, indicating high stress drop.

Modeling the Effects of Source and Path Heterogeneity on Ground Motions of Great Earthquakes on the Cascadia Subduction Zone Using 3D Simulations

Abstract We ran finite‐difference earthquake simulations for great subduction zone earthquakes in Cascadia to model the effects of source and path heterogeneity for the purpose of improving strong‐motion predictions. We developed a rupture model for large subduction zone earthquakes based on a k −2 slip spectrum and scale‐dependent rise times by representing the slip distribution as the sum of normal modes of a vibrating membrane. Finite source and path effects were important in determining the distribution of strong motions through the locations of the hypocenter, subevents, and crustal structures like sedimentary basins. Some regions in Cascadia appear to be at greater risk than others during an event due to the geometry of the Cascadia fault zone relative to the coast and populated regions. The southern Oregon coast appears to have increased risk because it is closer to the locked zone of the Cascadia fault than other coastal areas and is also in the path of directivity amplification from any rupture propagating north to south in that part of the subduction zone, and the basins in the Puget Sound area are efficiently amplified by both north and south propagating ruptures off the coast of western Washington. We find that the median spectral accelerations at 5xa0s period from the simulations are similar to that of the Zhao etxa0al. (2006) ground‐motion prediction equation, although our simulations predict higher amplitudes near the region of greatest slip and in the sedimentary basins, such as the Seattle basin.

A Scenario Study of Seismically Induced Landsliding in Seattle Using Broadband Synthetic Seismograms

Abstract We demonstrate the value of utilizing broadband synthetic seismograms to assess regional seismically induced landslide hazard. Focusing on a case study of an M w xa07.0 Seattle fault earthquake in Seattle, Washington, we computed broadband synthetic seismograms that account for rupture directivity and 3D basin amplification. We then adjusted the computed motions on a fine grid for 1D amplifications based on the site response of typical geologic profiles in Seattle and used these time‐series ground motions to trigger shallow landsliding using the Newmark method. The inclusion of these effects was critical in determining the extent of landsliding triggered. We found that for inertially triggered slope failures modeled by the Newmark method, the ground motions used to simulate landsliding must have broadband frequency content in order to capture the full slope displacement. We applied commonly used simpler methods based on ground‐motion prediction equations for the same scenario and found that they predicted far fewer landslides if only the mean values were used, but far more at the maximum range of the uncertainties, highlighting the danger of using just the mean values for such methods. Our results indicate that landsliding triggered by a large Seattle fault earthquake will be extensive and potentially devastating, causing direct losses and impeding recovery. The high impact of landsliding predicted by this simulation shows that this secondary effect of earthquakes should be studied with as much vigor as other earthquake effects. Online Material: High‐resolution maps of relative seismically induced landslide hazard for an M w xa07.0 Seattle fault earthquake for dry and saturated soil conditions.

Modeling Strong‐Motion Recordings of the 2010 Mw 8.8 Maule, Chile, Earthquake with High Stress‐Drop Subevents and Background Slip

Strong‐motion recordings of the M wxa08.8 Maule earthquake were modeled using a compound rupture model consisting of (1)xa0a background slip distribution with large correlation lengths, relatively low slip velocity, and long peak rise time of slip of about 10xa0s and (2)xa0high stress‐drop subevents (asperities) on the deeper portion of the rupture with moment magnitudes 7.9–8.2, high slip velocity, and rise times of slip of about 2xa0s. In this model, the high‐frequency energy is not produced in the same location as the peak coseismic slip, but is generated in the deeper part of the rupture zone. Using synthetic seismograms generated for a plane‐layered velocity model, I find that the high stress‐drop subevents explain the observed Fourier spectral amplitude from about 0.1 to 1.0xa0Hz. Broadband synthetics (0–10xa0Hz) were calculated by combining deterministic synthetics derived from the background slip and asperities (≤1u2009u2009Hz) with stochastic synthetics generated only at the asperities (≥1u2009u2009Hz). The broadband synthetics produced response spectral accelerations with low bias compared to the data, for periods of 0.1–10xa0s. A subevent stress drop of 200–350 bars for the high‐frequency stochastic synthetics was found to bracket the observed spectral accelerations at frequencies greater than 1xa0Hz. For most of the stations, the synthetics had durations of the Arias intensity similar to the observed records.