000019798 001__ 19798
000019798 005__ 20170118182337.0
000019798 04107 $$aeng
000019798 046__ $$k2017-01-09
000019798 100__ $$aCrempien, Jorge
000019798 24500 $$aThe Ucsb Method for Simulating Broadband Ground Motion Using Correlated Kinematic Rupture Parameters on a Finite Fault

000019798 24630 $$n16.$$pProceedings of the 16th World Conference on Earthquake Engineering
000019798 260__ $$b
000019798 506__ $$arestricted
000019798 520__ $$2eng$$aThe UCSB method produces broadband ground motion (0-25 Hz) for earthquakes with magnitudes between 5 and 8. For each earthquake scenario we generate kinematic rupture scenarios (50 stochastic realizations or more) that are characterized by Kostrov-like [1] slip-rate functions densely spaced on a finite fault area. The slip-rate functions are parameterized by the total slip, time to reach the maximum slip-rate (peak-time), the total time of slipping (rise-time), and time when slip starts (this determines the local rupture velocity). For each stochastic simulation each kinematic rupture parameter we filter white noise with a Von Karman power spectrum in the wavenumber domain to produce the spatial correlation. We use the correlations between each kinematic source parameter pair based on the dynamic rupture scenarios [2]. For each synthetic earthquake rupture scenario, we compute ground motion using the representation theorem. For this, we separate the wave propagation problem into a low- and high-frequency components separated at 1.0 Hz. The high-frequency Green's functions capture the geometrical decay with a simplified velocity structure. The high-frequency Green’s functions will be later convolved with scattering functions. The high-frequency amplitude is modified using the detailed 1D velocity model for the velocity and density and the quarter-wavelength impedance [3]. The low-frequency ground motion is propagated from the source to the site using either a 1D or 3D velocity structure. We merge the low- and high-frequency ground motion by stitching these two in the wavelet domain [4]. Important to note is that the source is the same for both the low- and high-frequency components of ground motion. So far we have successfully validated our method against well-recorded data produced by earthquakes in different tectonic regions such as California, Eastern United States, and Japan [5]. The metric for comparing the quality of the computed ground motion is acceleration response spectrum at distance up to 150 km from the fault. The comparisons look at the total bias between observed and computed response spectral ordinates at different periods as well as the dependence of the spectral ordinate on distance from the fault at various periods. The UCSB method is an attractive solution to generate physics-based synthetic ground motion with minimal computational cost.

000019798 540__ $$aText je chráněný podle autorského zákona č. 121/2000 Sb.
000019798 653__ $$akeywords1, keywords2, (max. one line) Keywords should use Times New Roman 10 pt. font; Italic; separated by semicolon; Maximum 5

000019798 7112_ $$a16th World Conference on Earthquake Engineering$$cSantiago (CL)$$d2017-01-09 / 2017-01-13$$gWCEE16
000019798 720__ $$aCrempien, Jorge$$iArchuleta, Ralph
000019798 8560_ $$ffischerc@itam.cas.cz
000019798 8564_ $$s6464674$$uhttps://invenio.itam.cas.cz/record/19798/files/4433.pdf$$yOriginal version of the author's contribution as presented on USB, paper 4433.
000019798 962__ $$r16048
000019798 980__ $$aPAPER