Information Technology Reference
In-Depth Information
different renewal models (such as lognormal, Brownian Passage Time, etc.) used in
PSHA studies. The workflows differ in the estimation techniques employed. SSF2
workflow offers a choice between two standard estimation methods, namely least
squares and maximum likelihood. SSF3 workflow, on the other hand, uses a robust
estimation method called the modi
ed maximum likelihood (Tiku 2004), and
additionally provides an option to choose the error distribution and related
parameters that are to be used in the estimation process. For both workflows the
estimation process requires an earthquake catalog and a set of seismic sources as
inputs. The parameter estimates are computed, using the provided earthquake cat-
alog, for each seismic source provided as an input.
SSF4 work fl ow performs PSHA computations for a speci
ed site region of
interest. A given set of seismic sources, for which analysis parameters are estimated
beforehand (most probably using SSF2 or SSF3 workflows), is used in PSHA
calculations. It is possible to choose among different PSHA models; namely
Poisson (Cornell 1968), renewal (Matthews 2002), and renewal hybrid (Wu 1995)
models. It is also possible to specify the attenuation model to be used for the
analysis study and choose among available magnitude models; namely exponential
(Gutenberg 1944), truncated-exponential (Cornell 1969), and characteristic earth-
quake (Youngs 1985) magnitude distributions. In addition, various parameters for
the analysis computations can be speci
ed before execution.
SSF5 work fl ow evaluates the behavior of a chosen ground motion prediction
equation (GMPE), i.e., attenuation model, and enables the users to examine the
effects of using that speci
c GMPE in the PSHA studies. Both generic GMPEs, such
as the one proposed by Boore and Atkinson (2008), and region-speci
c GMPEs,
such as the one developed for Turkey by Kalkan and G
lkan (2004), are provided as
alternatives. The workflow further offers a chance to evaluate the chosen model for
different output types it supports and also for the different input parameters.
SSF6 work fl ow performs logic-tree based PSHA computations (Krinitzsky 1995)
for a speci
ΓΌ
ed site region of interest. Similar to the case in SSF4 workflow, a given
set of seismic sources, for which analysis parameters are estimated beforehand, is
used in PSHA calculations. It is possible to select multiple different PSHA, mag-
nitude, and attenuation models for the same analysis study. For each selected model
a subjective probability, which will be used in logic-tree calculations, is assigned.
Furthermore, various parameters for the analysis computations can be speci
ed
before execution. A sensitivity analysis (Giner 2002) is also conducted by this
workflow.
SSF7 work fl ow performs seismic risk calculations depending on the PSHA
results obtained by using SSF4 or SSF6 workflows. Expected annual damage ratios
(EADRs) are calculated for a given arbitrary number of different structure types
provided together with corresponding damage probability matrices (DPMs), which
are constructed from observational and estimated data using past earthquake data or
subjective judgments of experts (Deniz 2009).
Advanced-use work fl ow provides a wrapper job for executing user programs,
written in different programming languages, which involve many calls to SSFs
arbitrarily. The workflow performs Remote API calls for executing SSFs requested
Search WWH ::




Custom Search