S22B-01 INVITED 10:20h
Building the Southern California Earthquake Center
Kei Aki was the founding director of the Southern California Earthquake Center (SCEC), a multi-institutional collaboration formed in 1991 as a Science and Technology Center (STC) under the National Science Foundation (NSF) and the U. S. Geological Survey (USGS). Aki and his colleagues articulated a system-level vision for the Center: investigations by disciplinary working groups would be woven together into a "Master Model" for Southern California. In this presentation, we will outline how the Master-Model concept has evolved and how SCEC's structure has adapted to meet scientific challenges of system-level earthquake science. In its first decade, SCEC conducted two regional imaging experiments (LARSE I & II); published the "Phase-N" reports on (1) the Landers earthquake, (2) a new earthquake rupture forecast for Southern California, and (3) new models for seismic attenuation and site effects; it developed two prototype "Community Models" (the Crustal Motion Map and Community Velocity Model) and, perhaps most important, sustained a long-term, multi-institutional, interdisciplinary collaboration. The latter fostered pioneering numerical simulations of earthquake ruptures, fault interactions, and wave propagation. These accomplishments provided the impetus for a successful proposal in 2000 to reestablish SCEC as a "stand alone" center under NSF/USGS auspices. SCEC remains consistent with the founders' vision: it continues to advance seismic hazard analysis through a system-level synthesis that is based on community models and an ever expanding array of information technology. SCEC now represents a fully articulated "collaboratory" for earthquake science, and many of its features are extensible to other active-fault systems and other system-level collaborations. We will discuss the implications of the SCEC experience for EarthScope, the USGS's program in seismic hazard analysis, NSF's nascent Cyberinfrastructure Initiative, and other large collaboratory programs.
http://www.scec.org/
S22B-02 10:50h
Earthquake Prediction and Disaster Preparedness: Interactive Algorithms
Recent studies in prediction of destructive earthquakes months in advance lead to a reformulation of the interwined problems linking disaster disaster preparedness with earthquake prediction. Given a specific earthquake prediction including a time window, magnitude and geographic area, the disaster manager has to choose the optimal set of temporal preparedness measures, taking into account its level of uncertainty. The "predictor" has to ensure this optimization by setting up an appropriate tradeoff between different kinds of prediction errors. Both problems belong to the broad field of decision-making based on incomplete information. We hypothetically explore these problems using a simplified example, based on a small portion of a major water utility, and present concepts that can be practically and effectively utilized for decision support by disaster managers.
S22B-03 11:05h
Physics-Based Characterized-Source Modeling for Strong Motion Prediction
From recent development of the kinematic waveform inversion to investigate rupture process, we have understood that strong ground motion is relevant to slip heterogeneity inside the source rather than average slip in the entire rupture area. The most important parameters featuring strong ground motions are sizes of asperities and effective stress on each asperity, which characterize the amplitudes and periods of directivity pulses causing earthquake damage. Systematic analyses of slip inversions indicate that the asperity areas as well as the rupture areas scale with the seismic moment (Somerville et al., 1999). Using the source scaling of outer and inner fault parameters, we have developed characterized source model for predicting strong ground motions from future large earthquakes as a recipe (Irikura et al., 2004), Towards realistic ground motion prediction with dynamically-constraint parameters, we introduce (1) three-stage scaling of seismic moment and rupture area considering fault geometry, (2) stress drop distribution based on the multiple-asperity model, and (3) slip-velocity functions explaining surface and subsurface ruptures with difference fracture energy versus to the seismic moment. For (1), 3D dynamic simulations for both crack and asperity models naturally explain the circular-crack model with constant stress drop before fault width saturating the seismogenic zone, the L-model with increasing stress drop after the saturation of fault width, and the W-model with constant stress drop where the fault length becomes long enough. For (2), 3D dynamic simulations clarify the relationship between stress drop, slip, and number of asperities. Negative stress drop on the background area is proposed for the source model of large earthquakes with multiple asperities. For (3), we try to explain a paradox that the ground motions caused by subsurface rupture seem to be larger than the ones by surface rupture (Somerville, 2003) using slip-velocity functions considering larger fracture energy for surface rupture and smaller fracture energy for subsurface rupture, based on the dynamic simulations of recent large earthquakes.
S22B-04 11:20h
IMPROVED PREDICTION METHOD FOR TIME HISTORIES OF NEAR-FIELD GROUND MOTIONS WITH APPLICATION TO SOUTHERN CALIFORNIA
We have developed a new method and a computer code to simulate stochastically the kinematic faulting process. To simulate the kinematic rupture we divide the fault of the mainshock into subevents. For each subevent we prescribe the slip history. In our kinematic model each subevent represents a point source with parameters consisting of the local slip amplitude, rupture velocity, and rise time, all of which are poorly constrained for future earthquakes. In order to allow for our inadequate a priori knowledge we describe these parameters as random variables with probability distribution functions that are bound by estimates of the parameters based on past earthquakes. Dynamic modeling of complex rupture process (e.g. Oglesby and Day, 2002; Guatteri et al. 2003) shows that the areas of large slip correlate with high stress drop, fast rupture velocity and short rise time. But the correlation between rise-time and slip is not so strong as the correlation between rupture velocity and slip (Oglesby and Day, 2002). Based these studies, we assume that the correlation between rupture velocity and slip is about 80 percent and the correlation between rise-time and slip is -50 percent. We have applied this approach to generate kinematics source process for a scenario earthquake on Puente Hills thrust fault. We chose the same the fault geometry as that used by SCEC (https://srb.npaci.edu/cgi-bin/new/mysrb2.cgi, Graves 2003). Near-field synthetic ground motion velocities (up to 1 Hz) are calculated by using a 3D viscoelastic finite difference (FD) algorithm of Liu and Archuleta (2002) and the SCEC 3-D velocity model. This FD code allows for two separate regions: an upper one with finely spaced grid and a lower one with three-times the spacing of the fine grid. This feature allows for a finely spaced grid near the surface where shear-wave velocities are low without carrying the fine spacing to depth where it is not needed. As expected, ground velocities up-dip of the hypocenter (South and West of the fault) are larger due to the average directivity of the rupture. Note the long period late arriving phases due to the deep basin structure and low-velocity near surface material.
S22B-05 11:35h
Effect of Fault Zone on Strong Ground Motion
During the M=7.2 Duzce earthquake and its early aftershocks, we recorded the ground accelerations at a small array located within the fault zone of the Izmit earthquake. The array was about 25km away from the Duzce epicenter and consisted of 4 stations set a few hundred meters apart. The stations were all located within about 150m from the fault trace and one of them was installed on the fault trace itself. Peak horizontal acceleration during the main shock was 0.3g at the furthest station from the fault but drastically increased to about 1.0g on the fault itself. Likewise, during the many recorded aftershocks, peak acceleration shows a regular and drastic increase as the recording site is located closer and closer to the fault. During the strongest recorded aftershock, for instance, peak values continuously increase from 0.21g, to 0.34g, to 0.62g, and to about 1.0g, as the distance of the site to the fault trace decreases from about 150m to 0m. These observations show that the fault zone is a narrow structure which extends over a distance of about 100m from the fault trace and which considerably increases the ground shaking close to the fault. This result is in agreement with the study of fault-zone trapped waves made by Ben-Zion et al. (2003) along the same fault segment. The present results also show that the fault zone is not a homogeneous low-velocity wedge, but is a zone of continuously degrading elastic stiffness as one approaches the fault trace.
S22B-06 11:50h
Source-Averaged Basin Effects from 3D Ground Motion Simulations
We simulate long-period (0-0.5 Hz) ground motion time histories for a suite of sixty scenario earthquakes (Mw 6.3 to Mw 7.1) within the Los Angeles basin region. Fault geometries are based upon the Southern California (SCEC) Community Fault Model, and 3D seismic velocity structure is based upon the SCEC Community Velocity Model. The ground motion simulations are done using 5 different 3D finite difference and finite element codes, and we perform numerous cross-check calculations to insure consistency among these codes. The nearly 300,000 synthetic time histories from the scenario simulations provide a resource for ground motion estimation and engineering studies of large, long-period structures, or smaller structures undergoing large, nonlinear deformations. By normalizing spectral accelerations to those from simulations performed for reference hard-rock models, we characterize the source-averaged effect of basin depth on spectral acceleration. For this purpose, we use depth (H) to the 1.5 km/s S velocity isosurface as the predictor variable. The resulting mean basin-depth effect is period dependent, and both smoother (as a function of period and depth) and higher in amplitude than predictions from local 1D models. For example, relative to a reference hard-rock site, sites with H equal to 2.5 km (corresponding to some of the deeper L.A. basin locations) have a predicted mean amplification factor of approximately 5.5 at 2 s period, and approximately 7.5 at 10 s period. We compare long-period (5 s) spectral amplitudes from the reference simulations with standard regression relationships for sites nominally classified as "rock" in empirical studies. From that comparison, we infer that the average nominal rock site represented in the empirical regression models has response approximately a factor of 2 higher than our hard-rock reference model. Hence, the basin-depth factors can be scaled down by a factor of approximately 2 to convert them to correction factors to empirical rock-site relationships. For the H=2.5 km example, the resulting correction to rock-site relationships is about 2.75 at 2 s period, and 3.75 at 10 s period. The correction factors are lower for shallower basin sites, and the period dependence has the reverse sense. For example, for H=0.75 km, the corresponding factors are approximately 1.7 (at 2 s) and 1.4 (10 s). The factor of 2 long-period bias between hard-rock simulations and empirical rock-site regression models suggests that the standard rock-site classification actually incorporates sites in which relatively low S velocities (less than 2 km/s) extend, on average, to depths of the order of 0.5 km.
S22B-07 12:05h
Evidence for Variability in the Shape, not Just the Level, in the High Frequency Spectrum of Strong Ground Motions
The high-frequency spectrum of ground motions in Guerrero, Mexico and in Nevada and California has been examined from the perspective of the parameter "kappa". The main feature of the high frequency spectrum (i.e. above the corner frequency) can be approximated as proportional to exp(-pi*kappa*f). Based on a distance dependence, Anderson and Hough (1984) considered that kappa is a path effect related to Q. An alternative explanation is that it is predominantly affected by the seismic source. More recent studies indicate that both explanations are partially correct. In Mexico, Purvance and Anderson (2003) found that source and site terms have similar effectiveness in explaining and reducing the variability for multiple measurements. Similarly, in southern Nevada, the variability in kappa is reduced nearly as far as possible by a model with only site and event terms. The source contribution is an interesting phenomenon. In the simple omega-squared source model, the acceleration spectrum is flat above the corner frequency, although the level depends on stress drop. These results indicate that the shape also is variable. At least some of the shape variations above the corner depend systematically on event location and/or mechanism.