HR: 1340h
AN: T53B-0484 [Abstracts]
TI: Investigation of the Influence of Different Velocity Models on the Focal Mechanism Solutions Derived
from Inversion of Regional Broadband Waveforms for Earthquakes in Greece
AU: * Konstantinou, K
EM: kostas@earth.sinica.edu.tw
AF: Institute of Earth Sciences
Academia Sinica, POB 1-55
Nankang, Taipei, 115
Taiwan
AU: * Konstantinou, K
EM: kostas@earth.sinica.edu.tw
AF: Institute of Geodynamics
National Observatory of Athens, POB 20048, Athens, 118 10
Greece
AU: Sokos, E
EM: esokos@gein.noa.gr
AF: Institute of Geodynamics
National Observatory of Athens, POB 20048, Athens, 118 10
Greece
AU: Melis, N
EM: nmelis@gein.noa.gr
AF: Institute of Geodynamics
National Observatory of Athens, POB 20048, Athens, 118 10
Greece
AU: Stavrakakis, G
EM: g.stavr@gein.noa.gr
AF: Institute of Geodynamics
National Observatory of Athens, POB 20048, Athens, 118 10
Greece
AU: Boukouras, K
EM: kbouk@admin.gein.noa.gr
AF: Institute of Geodynamics
National Observatory of Athens, POB 20048, Athens, 118 10
Greece
AB:
The routine use of regional broadband data for the determination of moment tensors for even
very small (Mw $\sim$ 3.5) events, has considerably enhanced our understanding of tectonic
processes in many active regions around the world. However, this advance comes at the expense
of having to know an adequate regional one-dimensional model that can successfully predict
the source-to-receiver path effects, thus minimizing the differences between observed and
synthetic data during the source inversion process. This is especially true for the data
recorded by the Greek national network, due to the limited bandwidth of its stations that
record periods of up to 20 s, increasing the dependency of the source inversions on the
velocity model used. The network is operated by the Institute of Geodynamics of the National
Observatory of Athens and consists of 22 digital, three-component seismometers that cover
most parts of the country. In order to investigate this dependency, we selected five moderate
to large (Mw $\sim$ 5.0-6.4) earthquakes based on the following two criteria: (1) the available
moment tensor solutions for these events reported by the Harvard CMT group, the USGS and/or
the Swiss Seismological Service showed good agreement with each other, (2) the events should
occur in geographically different parts of the Greek region. As a first step before the
inversion we gathered all the published velocity models for the Greek region resulting
from teleseismic/local earthquake/surface-wave tomography, created average one-dimensional
models and did forward modelling of the waveforms of each event. If the calculated misfit
of the forward-synthetic and observed waveform was small, then this velocity model was
further used in the source inversion in the frequency band 0.05-0.08 Hz. The results from
the inversions showed that in all five cases the focal mechanism was exhibiting only minor
differences when compared to the other available solutions. However, the amount of CLVD
was high ($\sim$ 70%) for events in the central and northern Aegean Sea, while the misfit
versus depth curve showed no clear minimum for a range of depth values. The situation was
better for events occuring in western and southern Greece with lower CLVD values ($\sim$
2-34%) and clear minima in the corresponding misfit versus depth curves. These observations
suggest that most published models for the Greek region are probably insufficient for
matching all path effects for the frequency bandwidth used in our inversions, but at the same time
seem adequate for the recovery of a reliable focal mechanism. The results of this study
have great importance in view of future efforts to routinely invert regional waveforms
for the determination of focal mechanism solutions for earthquakes in Greece.
DE: 7230 Seismicity and seismotectonics
DE: 8123 Dynamics, seismotectonics
DE: 7209 Earthquake dynamics and mechanics
SC: Tectonophysics [T]
MN: 2004 AGU Fall Meeting