Electronic Supplement to
Predominant-period Site Classification for Response Spectra Prediction Equations in Italy
by
Carola Di Alessandro, Luis Fabian Bonilla, David M. Boore, Antonio Rovelli, and Oona Scotti
The files included in this electronic supplement contain the coefficients of Eq. (1) [Table S1],
material showing variations of H/V spectral ratios as function of magnitude [Table S2 and Figure S1]
and of hypocentral distance [Table S3 and Figure S2], dependence of intra-event residuals for a variety
of periods with respect to magnitude [Figure S3 and Figure S5] and distance [Figure S4 and Figure S6],
and comparison of our predicted spectra with respect to other global GMPEs [Figure S7].
The tables are comma-separated-variable (csv) files and the figures are jpg files.
-
Table S1. The coefficients of the ground-motion
prediction equations giving log10 Sa in terms of predictor variables, including our predominant-period-based site
classes.
The meaning of the columns of the table are obvious,
except perhaps for the site-class columns (e.g., CL-VI).
These correspond to the Sj in equation (1). Note that CL-I does not appear, because it is the
reference site class.
-
Table S2. The number of records in each moment magnitude group.
-
Table S3. The number of records in each hypocentral distance group.
-
Figure S1. Variations of the H/V spectral ratios for our classes as a
function of magnitude ranges. The shape of the average H/V spectral ratios
in each class is remarkably stable and the period of the predominant peaks
does not shift with magnitude, thus implying that nonlinearity does not
introduce a systematic bias at the largest magnitudes of our dataset. Only
class CL-IV exhibits moderately larger peak ratios for records with Mw
between 5.0 and 5.4, in comparison to those of other magnitude ranges,
probably because of the particular combination of events and soft-soil
stations that recorded these events.
-
Figure S2. Variations of the H/V spectral ratios for our classes
as a function of hypocentral distance ranges. They are not relevant
to the purpose of site classification, therefore there is no evident
hypocentral distance dependence for the site classes.
-
Figure S3. Dependence of intra-event residuals on moment
magnitude Mw at 0.2 sec period for the different predominant-period
classes proposed in this study.
-
Figure S4. Dependence of intra-event residuals on hypocentral
distance Rhypo at 0.2 sec period for the different predominant-period
classes proposed in this study.
-
Figure S5. Dependence of intra-event residuals on moment
magnitude Mw at 1.0 sec period for the different predominant-period
classes proposed in this study.
-
Figure S6. Dependence of intra-event residuals on hypocentral
distance Rhypo at 1.0 sec period for the different predominant-period
classes proposed in this study.
-
Figure S7. Spectral amplitudes at 0.1 sec, 0.3 sec and 1.0 sec
period with respect to Joyner and Boore distance (Rjb) for the same
GMPEs considered in Fig. A5. Different distance metric compatibility
is provided by the distance conversion scheme proposed by Scherbaum et al. (2004).
[ Back ]
