Probabilistic assessment of earthquake insurance rates for buildings - Seismic Risk Analysis and Management of Civil Infrastructure Systems - page 784

ment of the earthquake insurance rates consists of the following basic steps
(Yucemen and Ozturk, 2008):

1. A seismic database is compiled using catalogs which list information on

past earthquakes that have occurred in the region of interest. The raw
data in the earthquake catalogs have to be processed considering the
following: magnitudes reported in different scales have to be converted
to a common scale, preferably to the moment magnitude ( M w ) scale.
Earthquake clusters are identifi ed and dependent events (fore- and
after-shocks) are eliminated by defi ning appropriate space and time
windows (Deniz, 2006). Finally, an analysis of catalog completeness is
performed to obtain complete rates (i.e. complete number of events
over a particular time period).

2. The region is divided into discrete seismic sources in the form of lines

and areas. This involves preparation of a fault map, identifi cation of
active faults and assessment of the parameters of active faults. The basic
information required for an active fault is as follows: actual location,
coordinates of starting and ending points, activity rate, type, orientation
(strike, dip amount, and direction), length, segmentation, age, total hori-
zontal/vertical slip amount, annual slip rate, the magnitude and return
period of the characteristic earthquake, and date of the last character-
istic earthquake.

3. A seismotectonic map showing the distribution of earthquake epicen-

ters and their relationships with active faults should be plotted. Such a
map also displays distribution of past earthquake data to the different
seismic sources consistent with their epicentral locations. Based on the
data, recurrence relationships that provide information on the relative
frequency of occurrence of earthquakes with different magnitudes are
obtained and the seismicity parameters are assessed for each seismic
source. Areal background seismic sources are identifi ed/developed to
include the contribution of earthquakes that could not be associated
with any one of the faults. For this purpose, a spatially smoothed seis-
micity model (Frankel, 1995) and background areal seismic source with
uniform seismicity can be used.

4. Selection of the appropriate probabilistic and stochastic models for the

earthquake magnitude distribution and the earthquake occurrence in
the time domain. In the PSHA, earthquake occurrences are assumed to
exhibit a Poisson process and magnitudes to be distributed exponen-
tially (i.e. Gutenberg-Richter law). These two assumptions are still
widely employed in current seismic hazard studies. However, the
renewal process (Ellsworth et al. , 1999) and the characteristic earth-
quake model (Schwartz and Coppersmith, 1984) can be considered as
alternatives to the Poisson process and the Gutenberg-Richter relation-