S12A-02:
Strong Effects of Vs30 Heterogeneity on Physics-Based Scenario Ground-Shaking Computations

Monday, 15 December 2014: 10:40 AM
John N Louie, Univ of Nevada, Reno, NV, United States and Satish K Pullammanappallil, Optim Software and Data Solutions, Reno, NV, United States
Abstract:
Hazard mapping and building codes worldwide use the vertically time-averaged shear-wave velocity between the surface and 30 meters depth, Vs30, as one predictor of earthquake ground shaking. Intensive field campaigns a decade ago in Reno, Los Angeles, and Las Vegas measured urban Vs30 transects with 0.3-km spacing. The Clark County, Nevada, Parcel Map includes urban Las Vegas and comprises over 10,000 site measurements over 1500 km2, completed in 2010. All of these data demonstrate fractal spatial statistics, with a fractal dimension of 1.5-1.8 at scale lengths from 0.5 km to 50 km. Vs measurements in boreholes up to 400 m deep show very similar statistics at 1 m to 200 m lengths. When included in physics-based earthquake-scenario ground-shaking computations, the highly heterogeneous Vs30 maps exhibit unexpectedly strong influence. In sensitivity tests (image below), low-frequency computations at 0.1 Hz display amplifications (as well as de-amplifications) of 20% due solely to Vs30. In 0.5-1.0 Hz computations, the amplifications are a factor of two or more. At 0.5 Hz and higher frequencies the amplifications can be larger than what the 1-d Building Code equations would predict from the Vs30 variations. Vs30 heterogeneities at one location have strong influence on amplifications at other locations, stretching out in the predominant direction of wave propagation for that scenario. The sensitivity tests show that shaking and amplifications are highly scenario-dependent. Animations of computed ground motions and how they evolve with time suggest that the fractal Vs30 variance acts to trap wave energy and increases the duration of shaking. Validations of the computations against recorded ground motions, possible in Las Vegas Valley due to the measurements of the Clark County Parcel Map, show that ground motion levels and amplifications match, while recorded shaking has longer duration than computed shaking. Several mechanisms may explain the amplification and increased duration of shaking in the presence of heterogeneous spatial distributions of Vs: conservation of wave energy across velocity changes; geometric focusing of waves by low-velocity lenses; vertical resonance and trapping; horizontal resonance and trapping; and multiple conversion of P- to S-wave energy.