Publication Date:
2020-10-26
Description:
We model interevent times and Coulomb static stress transfer on the rupture segments
along the Corinth Gulf extension zone, a region with a wealth of observations on
strong-earthquake recurrence behavior. From the available information on past seismic activity,
we have identified eight segments without significant overlapping that are aligned along the
southern boundary of the Corinth rift. We aim to test if strong earthquakes on these segments are
characterized by some kind of time-predictable behavior, rather than by complete randomness.
The rationale for time-predictable behavior is based on the characteristic earthquake hypothesis,
the necessary ingredients of which are a known faulting geometry and slip rate. The tectonic
loading rate is characterized by slip of 6 mm/yr on the westernmost fault segment, diminishing
to 4 mm/yr on the easternmost segment, based on the most reliable geodetic data. In this study,
we employ statistical and physical modeling to account for stress transfer among these fault
segments. The statistical modeling is based on the definition of a probability density distribution
of the interevent times for each segment. Both the Brownian Passage-Time (BPT) and Weibull
distributions are tested. The time-dependent hazard rate thus obtained is then modified by the
inclusion of a permanent physical effect due to the Coulomb static stress change caused by
failure of neighboring faults since the latest characteristic earthquake on the fault of interest. The
validity of the renewal model is assessed retrospectively, using the data of the last 300 years, by
comparison with a plain time-independent Poisson model, by means of statistical tools
including the Relative Operating Characteristic diagram, the R-score, the probability gain and
the log-likelihood ratio. We treat the uncertainties in the parameters of each examined fault
source, such as linear dimensions, depth of the fault center, focal mechanism, recurrence time,
coseismic slip, and aperiodicity of the statistical distribution, by a Monte Carlo technique. The
Monte Carlo samples for all these parameters are drawn from a uniform distribution within their
uncertainty limits. We find that the BPT and the Weibull renewal models yield comparable
results, and both of them perform significantly better than the Poisson hypothesis. No clear
performance enhancement is achieved by the introduction of the Coulomb static stress change
into the renewal model.
Description:
Published
Description:
3655–3673
Description:
5T. Modelli di pericolosità sismica e da maremoto
Description:
6T. Variazioni delle caratteristiche crostali e precursori
Description:
JCR Journal
Keywords:
04.06. Seismology
;
05.08. Risk
Repository Name:
Istituto Nazionale di Geofisica e Vulcanologia (INGV)
Type:
article
Permalink