HR: 1340h
AN: S43A-1063    [Abstracts]
TI: A new global search inversion method to image earthquake kinematic rupture history: an application to the 2000 western Tottori earthquake
AU: * Piatanesi, A
EM: piatanesi@ingv.it
AF: Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, Rome, 00143 Italy
AU: Spudich, P
EM: spudich@usgs.gov
AF: US Geological Survey, 345 Middlefield Road, Menlo Park, CA 94025 United States
AU: Tinti, E
EM: tinti@ingv.it
AF: Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, Rome, 00143 Italy
AU: Cocco, M
EM: cocco@ingv.it
AF: Istituto Nazionale di Geofisica e Vulcanologia, Via di Vigna Murata 605, Rome, 00143 Italy
AB: We present a technique to invert strong motions records and geodetic data to retrieve the rupture history of an earthquake on a finite fault. To account for the actual rupture complexity, the fault parameters are spatially variable peak slip velocity, slip direction, rupture velocity and rise time. Our method belongs to the class of the so-called "single-window" approach for which the analytical form of the source time function is chosen a priori: we use a dynamically consistent source time function recently proposed to constrain the dynamic traction evolution on the fault plane [Piatanesi et al., 2004; Tinti et al., 2005]. The unknown parameters are given at the nodes of the subfaults, whereas the parameters within a subfault are allowed to vary through a bilinear interpolation of the nodal values. The forward modeling is performed with a discrete wavenumber technique, whose Green's functions include the complete response of the vertically varying Earth structure. We invert for the nodal values of the source parameters using the heat-bath simulated annealing technique that allows for a global and efficient search in a high-dimension parameter space. The recorded and modeled waveforms are compared in the frequency domain, using a cost function that is a hybrid representation between L1 and L2 norms: this cost function is less sensitive to the amplitude and more robust than classical least squares norm. Moreover, we split the inverted frequency band; we fit both amplitudes and phases at low frequencies (f < 1 Hz), which allows a good wave form fit, but we fit only amplitudes at high frequencies (f > 1 Hz). We present several applications performed to calibrate the inversion procedures. In particular, we have applied this method to the 2000 western Tottori, Japan, earthquake (Mw 6.8): this is one of the best recorded events in recent years, for which there are more than 30 surface and more than 10 borehole strong motions records as well as several GPS measurements
DE: 3260 Inverse theory
DE: 7209 Earthquake dynamics (1242)
DE: 7212 Earthquake ground motions and engineering seismology
DE: 7215 Earthquake source observations (1240)
DE: 8118 Dynamics and mechanics of faulting (8004)
SC: Seismology [S]
MN: Fall Meeting 2005