Effect of Spectra Scaling on Site Specific Design

Earthquake Characteristics Based on 1D Site

Response Analysis

A. Ansal,
Ozyegin University, Engineering Faculty, Istanbul, Turkey

G. Tönük, A. Kurtuluş & B. Çetiner

Boğaziçi University, Kandilli Observatory and Earthquake Research Institute, Istanbul

SUMMARY
A parametric study was conducted to evaluate the effects of different strong motion scaling options to be used
for preparing input acceleration time histories for 1D site response analyses. The first alternative is to use
outcrop PGA from the hazard study to scale strong motion acceleration records that are compatible with the
selected outcrop NEHRP hazard spectrum in addition to the hazard compatibility with respect to fault type,
earthquake magnitude, source distance and site conditions of the recording station. The second alternative is to
scale these selected records by an optimization routine to obtain the best fit mean acceleration spectra with
respect to the outcrop NEHRP hazard spectra. The third option is modify the amplitude and frequency content
of the selected strong motion records to have better outcrop NEHRP spectrum compatibility. The results of site
response analysis using 22 hazard compatible strong motion records with different scaling options are compared
for a site specific study based on 25 soil profiles.

Keywords: site response, design spectra, spectra scaling, input motion

1. INTRODUCTION

The purpose of a probabilistic site specific earthquake hazard analysis is to estimate peak ground
acceleration and uniform hazard acceleration spectrum for performance levels of Collapse Prevention,
Life Safety, and Immediate Occupancy that may correspond to exceedance levels of 50%, 10%, and
2% in 50 years or 72, 475 and 2475 year return periods. In the case of new engineering structures, it is
preferable to adopt a probabilistic earthquake hazard assessment, however in evaluating possible
response for the existing structures; deterministic approach may also be suitable. Independent of the
methodology adopted for the earthquake hazard analyses, strong motion (SM) records are needed to
conduct site response analyses and to estimate possible earthquake characteristics on the ground
surface. It was demonstrated by Ansal and Tönük (2007) that if limited number of acceleration time
histories (e.g. 3 records as specified in some earthquake codes) is used even with scaling to the same
PGA amplitude for site response analysis, the results on the ground surface can be different for
different sets of input acceleration records. One possible option to overcome this variability is to
conduct site response analyses using large number of hazard compatible acceleration time histories
and by adopting a probabilistic approach to estimate earthquake characteristics on the ground surface
for design and vulnerability assessments.

Assessment of site-specific design earthquake characteristics may be conducted as composed of

statistically independent two consecutive stages. The first stage involves the seismic hazard study to
assess the design earthquake characteristics on rock outcrop for the selected performance levels. The
second stage involves site response analysis to estimate design earthquake characteristics on the
ground surface based on geotechnical and geological site conditions. A probabilistic approach may be
adopted to estimate the overall exceedance probability for the design earthquake characteristics on the
ground surface for the selected performance levels (Ansal et al. 2011).
The uncertainties arising from the source characteristics may be taken into account by using large
number of seismic hazard compatible (fault mechanism, earthquake magnitude, and source distance)
strong motion acceleration records for site response analyses. In using a selected set of strong motion
records for site specific study, even though they may be recorded under similar tectonic conditions and
were selected based on hazard compatibility with respect to the probable earthquake magnitude, fault
type, source distance and site conditions at the recording station, it would still be necessary to scale
them to have a better compatibility with the site specific earthquake hazard study. This scaling can be
with respect to different earthquake parameters such as PGA, PGV, Arias Intensity and others without
altering the frequency content (Ansal, et al., 2006, Tönük and Ansal, 2010).

The other alternative is to adopt uniform hazard spectra from the earthquake hazard study on rock
outcrop for scaling strong motion records for site response analysis. Spectral scaling approaches
include methods developed in the time domain (Abrahamson, 1992; Hancock et al., 2006) and in the
frequency domain (Gasparini and Vanmarcke, 1976; Silva and Lee, 1987). Both approaches can be
used to modify existing time-histories to match the design response spectrum. One important issue is
not to modify significantly the basic time domain character of the recorded strong motion records.
The scaling process would be more efficient if the overall shape of the acceleration spectrum of the
selected strong motion record is not very different from the specified design acceleration spectrum and
PGA scaling may be adopted first so that the spectrum is approximate at the same level of the target
spectrum before initiating spectrum scaling. It was observed that time domain scaling gives better fits
with respect to target spectrum.

One dimensional (1D) site response analysis were conducted using slightly modified version of
Shake91 (Idriss and Sun, 1992) site response analysis code (Ansal, et al., 2009) to evaluate design
earthquake characteristics with respect to exceedance levels of 10% and 2% in 50 years or return
periods of 475 and 2475 years for a specific site where 25 soil borings were conducted.

A parametric study was carried out to evaluate the effects of different scaling options for input strong
motion set used in 1D site response analyses. The first option is to use outcrop hazard PGA scaled set
of strong motion records that are most compatible with the selected outcrop NEHRP hazard spectrum
obtained from regional hazard study in addition to hazard compatibility with respect to fault type,
earthquake magnitude, distance and average shear wave velocity of the recording station. The second
option is to scale these selected hazard and spectra compatible records by an optimization routine to
obtain the best fit for the mean acceleration spectra with respect to the outcrop NEHRP hazard
spectrum. The third option is to modify the amplitude and frequency content of the input motion set to
have better outcrop NEHRP hazard spectra compatibility using the available spectrum scaling
methodologies (Abrahamson, 1992, 1993, Hancock et al., 2006).

2. SITE RESPONSE ANALYSIS

There are different options for the estimation of earthquake characteristics on the ground surface at the
selected site to be used for engineering analyses. The simplest option is to use contemporary ground
motion prediction relationships formulated in terms of site and source classifications (Abrahamson et.
al., 2008) or to use empirical formulations such as amplification factors suggested by Borcherdt (1994)
based on equivalent shear wave velocity. The more comprehensive approach is based on site response
analyses using detailed site characterization. Taking into consideration possible differences in soil
profiles even within relatively short distances and observations in previous earthquakes that indicated
site conditions are important, and as demonstrated based on parametric studies (Ansal, et al., 2010),
the use of empirical amplification factors may not yield results on the conservative side. Thus, it is
preferable to adopt the comprehensive option for the assessment of site-specific ground motion
characteristics.

Site characterizations at the selected site are based on 25 soil borings where shear wave velocity
profiles for each boring were modelled by averaging the measured or calculated values for each soil
layer. The variations of shear wave velocities with depth were determined from SPT blow counts
using the empirical relationship proposed by Iyisan (1996) and based cross-hole, down-hole and
MASW measurements conducted at the site.

2.1. Input acceleration time histories

As suggested by Bommer and Acevedo (2004) and Bommer et al., (2000), it would be preferable to
use strong ground motion records for site response analysis and as observed by Ansal and Tonuk
(2007) use of simulated acceleration records may yield overconservative resuls thus may be
considered not very suitable in some cases.

Based on regional geological and hazard studies in the region, the investigated site may be affected
from earthquakes that may occur in different fault zones basically with two different fault
mechanisms; strike-slip and normal faulting. 11 pairs of acceleration time histories were selected
using the PEER data base that are compatible with the outcrop NEHRP Hazard spectrum in addition to
hazard compatibility with respect to probable magnitude (M=6.0-7.0), epicentre distance (R=5-20km)
and fault mechanisms (strike slip, normal, and reverse) recorded on stiff site conditions with average
shear wave velocities Vs30  546m/s as input acceleration time histories. The purpose is to account for
the variability arising from the differences in the source characteristics observed in the acceleration
time histories. Special effort was spend using PEER database to select as much as possible strong
motion records with similar acceleration spectra in comparison to target NEHRP hazard spectrum.

It is assumed that the selected strong motion records (Table 2.1) represent the characteristics of future
possible earthquakes that may take place in the near vicinity of the investigated site since the records
were obtained under similar tectonic conditions. Thus the effects of differences in the characteristics
of probable earthquakes can be taken into account with respect to the required design level.

Table 2.1. Set of strong motion records (SM) with outcrop NEHRP spectra compatibility (PEER)
Fault
Earthquake Station Date YMD Mw Vs30 (m/s) Repc (km) Components PGA (g)
type
GAZ000 0.608
Gazli, USSR Karakyr 1976-09-15 6.8 660 5.5
GAZ090 0.718
Victoria, CPE045 0.621
Cerra Prieto SS 1980-06-09 6.33 660 14.4
Mexico CPE315 0.587
A-STU000 0.251
Irpina Sturno N 1980-11-23 6.9 1000 10.8
A-STU270 0.358
Nahanni, S1010 0.978
Site 1 RV 1985-12-23 6.8 660 9.6
Canada S1280 1.096
Nahanni, S2240 0.489
Site 2 RV 1985-12-23 6.8 660 4.9
Canada S2330 0.323
Beverly Hills MU2035 0.617
Northridge RV 1994-01-17 6.7 546 18.4
Mulhol MU2125 0.444
LA0000 0.261
Northridge LA 00l RV 1994-01-17 6.7 706 19.1
La0090 0.388
Simi Valley KAT000 0.877
Northridge RV 1994-01-17 6.7 557 13.4
Katherine Rd KAT090 0.640

San Jose Santa RV- SJTE225 0.275

Loma Prieta 1989-10-18 6.9 672 14.7
Teresa Hills OBL SJTE315 0.228
Gilroy Gavilan RV- GIL337 0.325
Loma Prieta 1989-10-18 6.9 730 10
Coll. OBL GIL067 0.357
G06000 0.222
Morgan Hill Gilroy Array #6 SS 1987-01-10 6.2 663 9.9
G06090 0.292
3. INPUT MOTION SCALING

A parametric study was conducted to evaluate the effects of different input ground motion scaling
options for 1D site response analyses. The first option which happens to be the most widely adopted
and simpler is PGA scaling of the selected hazard compatible set of SM acceleration records (Ansal et
al., 2006). The acceleration spectra of these PGA scaled strong motion records are shown in Fig.1 in
comparison to the uniform outcrop hazard spectra for return periods of 2475 and 475 years. In this
case the mean spectral accelerations of all PGA scaled strong motion records in general are less than
the outcrop NEHRP spectral accelerations.

4 2.5
Scaled acceleration spectra for the selected SM records
3.5 Outcrop NEHRP Hazard Spectrum
SPECTRAL ACCELERATIONS (g)

2 Median acceleration spectrum for selected SM records

3 Median + 1 standard deviation
Median - 1 Standard Deviation
2.5
1.5

2 RP = 2475 years RP = 475 years

1
1.5

1
0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
PERIOD (s) PERIOD (s)

Figure 1. Acceleration spectra of for PGA scaled input acceleration records

The second option is to modify the strong motion acceleration set to have better outcrop NEHRP
hazard spectra compatibility by using a simple optimization scheme based on PGA scaling to have the
best fit of the mean acceleration spectrum with respect to the target outcrop NEHRP hazard spectra.
The spectra of all scaled strong motion records and the mean spectrum with respect to outcrop hazard
spectrum are shown in Fig. 2.

3.5 2.5
Scaled acceleration spectra for the selected SM records
Outcrop NEHRP Hazard Spectrum
SPECTRAL ACCELERATIONS (g)

3
2 Median acceleration spectrum for selected SM records
Median + 1 standard deviation
2.5
Median - 1 Standard Deviation
1.5
2
RP = 2475 years RP = 475 years

1.5
1

0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
PERIOD (s) PERIOD (s)

Figure 2. Acceleration spectra of for input acceleration records obtained for best fit mean spectra

The third option is to modify the input motion set by spectra scaling using the available two
methodologies that also modifies the frequency content of the input acceleration records to have better
fit to the outcrop NEHRP hazard spectra. The spectra of all scaled strong motion records by Method
A using SeismoMatch (2011) and the mean spectrum in comparison to the outcrop hazard spectrum
are shown in Fig. 3.
 3.5 2.5
Scaled acceleration spectra for selected SM records
SPECTRAL ACCELERATION (g)

3 Outcrop Hazard Spectra

2 Mean acceleration spectrum for selected SM records
2.5 Mean + 1 standard deviation
Mean - 1 standard deviation
1.5
2

RP=2475 years
1.5
1
RP=475 years
1

0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
PERIOD (s) PERIOD (s)

Figure 3. Acceleration spectra of for input acceleration records by one spectra scaling using Method A

The spectra of all spectra scaled earthquake records by Method B based on Abrahamson (1993)
methodology and the mean spectrum are shown in Fig. 4 in comparison to outcrop hazard spectrum.

3.5 2.5
Scaled acceleration sppectra for selected SM records
3 Mean acceleration spectrum for selected SM records
SPECTRAL ACCELERATION (g)

2
Outcrop Hazard Spectra
2.5
Mean + 1 standard deviation

1.5 Mean - 1 standard deviation

2

RP=2475 years
1.5
1 RP=475 years
1

0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5
PERIOD (s) PERIOD (s)

Figure 4. Acceleration spectra of for input acceleration records by one spectra scaling using Method B

The approach based on Abrahamson (1993) gave the best fits with respect to spectra scaling for the
outcrop NEHRP hazard spectra with very limited scatter in the individual acceleration spectra of the
22 scaled strong motion records and with very low standard deviation. On the other side, the
differences between spectral fitting using Method A and mean spectrum matching by an optimization
scheme can be considered negligible with respect to the mean spectral curves. However, the scatter
with respect to spectral accelerations of scaled strong motion records are more in the case of mean
spectrum matching since this approach only involved amplitude scaling and without altering the
frequency contents of input strong motion records as in the case of spectral scaling using Method A.

4. EARTHQUAKE CHARACTERISTICS ON THE GROUND SURFACE

Site specific earthquake design characteristics were investigated for one site in. Site response analyses
were carried out for 25 borings conducted as a part of the geotechnical investigation that was
supplemented with shear wave velocity profiles determined based on SPT blow counts, seismic
surface and in-hole tests. The above summarized four sets of input strong motion acceleration records
scaled with respect to four options are used for site response analyses. The results of site response
analysis using outcrop hazard PGA scaled strong motion records are shown in Fig. 5 as mean and
mean + 1 standard deviation acceleration response spectra. The best fit envelop as NEHRP design
spectrum is plotted to reflect the design levels corresponding to return periods of 2475 and 475 years.
 4 3
Site response analysis
3.5
Best fit NEHRP design spectra
SPECTRAL ACCELERATION (g)

2.5

3 Mean

2 Mean+1 standard deviation

2.5

2 RP=2475 years 1.5 RP=475 years

1.5
1

0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4
PERIOD (s) PERIOD (s)

Figure 5. Acceleration spectra on the ground surface calculated using PGA scaled acceleration records

The results of site response analysis using set of strong motion records scaled for mean spectra
matching by an optimization based on PGA scaling to have the best fit of the mean acceleration
spectra with respect to outcrop hazard spectrum are shown in Fig. 6.

4 3
Site response analysis
3.5
Best envelope NEHRP design spectrum
SPECTRAL ACCELERATION (g)

2.5

3 Mean acceleration spectrum

2 Mean+1 standard deviation

2.5

2 RP = 2475 year 1.5 RP = 475 year

1.5
1
1

0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4

PERIOD (s) PERIOD (s)

Figure 6. Acceleration spectra on the ground surface calculated using mean spectra scaled SM records

The results of site response analysis using spectra scaled strong motion records using Method A to
match the target NEHRP hazard spectrum are shown in Fig.7 in terms acceleration spectra on the
ground surface for 2475 and 475 year return periods.
4 3
Site response analysis
3.5
Best envelope NEHRP design spectrum
SPECTRAL ACCELERATION (g)

2.5

3 Mean acceleration spectrum

2 Mean+1 standard deviation

2.5

2 RP = 2475 year 1.5 RP = 475 year

1.5
1

0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4
PERIOD(s) PERIOD (s)

Figure 7. Acceleration spectra on the ground surface calculated using Code A spectra scaled SM records
The results of site response analysis using set of records that are spectra scaled using Method B
approach to match the target NEHRP hazard spectrum are shown in Fig.8 with respect to acceleration
spectra on the ground surface for 2475 and 475 year return periods.
4 3
Site response analysis
3.5
SPECTRAL ACCELERATION (g)

2.5 Best envelope NEHRP design spectrum

3 Mean acceleration spectrum

2
2.5 Mean +1 standard deviaiton

2 RP = 2475 year 1.5

1.5
1 RP = 475 year
1

0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4
PERIOD (s) PERIOD (s)

Figure 8. Acceleration spectra on the ground surface calculated using Code B spectra scaled SM records

The results of site response analyses indicate slight to significant deamplification with respect to mean
spectral accelerations (Fig.9) on the ground surface with respect to outcrop NEHRP hazard spectrum
depending on the approach adopted for input motion scaling. One of the reasons for lower
amplification in the case of spectra scaling is most likely due to frequency changes applied to
acceleration time histories to match the target acceleration spectrum. On the other hand, the spectra
compatibility for the selected SM records that was improved by PGA scaling to have better fit with the
outcrop NEHRP hazard spectrum gave slightly higher spectral accelerations with respect to
conventional PGA scaling approach.

2.5 2.5

Outcrop NEHRP hazard spectrum

Spectral Scaling Method B
SPECTRAL ACCELERATION (g)

2 2
Spectral scaling Method A
Mean spectra fitting
1.5 1.5 PGA scaling

RP = 2475 year
1 1 RP = 475 year

0.5 0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4
PERIOD (s) PERIOD (s)

Figure 9. Comparison of mean acceleration spectra calculated on the ground surface using four different
approaches for scaling the hazard compatible input SM records

It may not suitable to adopt mean acceleration spectra for design purposes; one option preferred by the
authors is to adopt mean + 1 standard deviation as possible design spectrum corresponding to return
periods of 475 and 2475 year return periods (Ansal and Tonuk, 2009). In that case as shown in Fig.10
and Fig.11, acceleration spectra obtained for mean +1 standard deviation and corresponding envelope
NEHRP design spectra are more suitable for design and vulnerability assessment purposes.
 4 4
Outcrop NEHRP hazard spectrum
SPECTRAL ACCELERATION (g)

3.5 3.5
Spectral Scaling Method B

3 3 Spectral scaling Method A

Mean spectra fitting
2.5 2.5 PGA scaling

2 RP = 2475 year 2

1.5 1.5
RP = 475 year
1 1

0.5 0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4
PERIOD (s) PERIOD (s)

Figure 10. Comparison of mean + 1 standard deviation acceleration spectra on the ground surface using four
different approaches for scaling the hazard compatible input SM records

4 3
Outcrop NEHRP hazard spectrum
SPECTRAL ACCELERATION (g)

3.5
2.5 Spectral Scaling Method B
3 Spectral scaling Method A
2 Mean spectra fitting
2.5
PGA scaling
2 RP = 2475 year 1.5

1.5
1 RP = 475 year
1
0.5
0.5

0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4
PERIOD (s) PERIOD (s)

Figure 11. Comparison of NEHRP design spectra on the ground surface using four different approach for
scaling the hazard compatible input SM records

4. CONCLUSION AND RECOMMENDATIONS

For the case of site specific assessment of earthquake ground motion characteristics on the ground
surface, earthquake hazard determination is based on regional scale to take into account earthquake
source zones and characteristics in a more comprehensive manner. In order to account for the
variability of the ground motion characteristics due to source mechanism and path effects large
number of hazard compatible strong motion acceleration records need to be used as input for site
response analysis to determine the possible earthquake characteristics on the ground surface. The first
requirement for the selection of strong motion acceleration records is the hazard compatibility with
respect to fault type, earthquake magnitude, source distance and average shear wave velocity at the
recording station. The second criteria may be defined with respect to similarity to the regional outcrop
acceleration hazard spectra and estimated peak ground acceleration. The third and maybe the most
important stage in this type of analyses and the methodology adopted to scale the hazard compatible
strong motion acceleration records to be used for site response analyses as outcrop motion. A
parametric study was conducted using a very widely used 1D site response analysis, Shake91 code, to
evaluate the effects of four possible options of scaling.

The geotechnical data comprised of 25 soil profiles also indicating possible variability at one site was
adopted to study the site specific design earthquake characteristics for two performance levels
corresponding to 475 and 2475 years. The results obtained indicate that the highest level of ground
shaking in terms of acceleration spectrum was calculated by using mean spectra matching approach.
However, the conventional PGA scaling approach yielded very similar or slightly lower results. In the
case of spectra scaling approaches the calculated acceleration spectra on the ground surface were
lower in comparison to conventional PGA scaling option, interestingly for the case studied; both
spectra matching schemes did not lead to spectral amplifications. Considering the extra effort and
time for spectral scaling and in the light of these preliminary results it may not be feasible and
sufficiently conservative to adopt spectral scaling schemes for site response analyses.

REFERENCES

Abrahamson, N., Atkinson,G., Boore,D., Bozorgnia,Y., Campbell, K., Chiou,B., Idriss, I.M., Silva, W., Youngs,
R. (2008). Comparisons of the NGA Ground-Motion Relations. Earthquake Spectra 24:1, 45-66.
Abrahamson, N.A. (1992). Non-stationary spectral matching, Seismological Research Letters, 63:1, 30
Abrahamson, N. A. (1993) Non-Stationary Spectral Matching Program RSPMATCH, User Manual
Ansal, A. ve Tönük, G. (2009). Short Comment on Site specific earthquake ground motion characteristics for
performance based design. International Conference on Performance-Based Design in Earthquake
Geotechnical Engineering, Ed. T.Kokusho, Y.Tsukamoto, and M.Yoshimine, CRC Press, 157
Ansal, A. ve Tönük, G. (2007). Source and Site Effects for Microzonation. Theme Lecture. 4th International
Conference on Earthquake Geotechnical Engineering, Earthquake Geotechnical Engineering, Editor:
K.Pitilakis, Ch.4, 73-92, Springer.
Ansal,A., Tönük,G. & Kurtuluş,A. (2011). Site Specific Earthquake Characteristics for Performance Based
Design. Proc. of the 5th Int. Conf.on Geotechnical Earthquake Engineering, Santiago, Chile.
Ansal,A., Kurtuluş,A. & Tönük,G. (2009). Earthquake Damage Scenario Software for Urban Areas.
Computational Structural Dynamics and Earthquake Engineering, Ch.2, 377-391, Structures and
Infrastructures Series, Editör(ler): Papadrakakis, M; Charmpis, DC; Lagaros, ND; Tsompanakis.
Ansal, A., Kurtuluş, A., Tönük, G. (2010). Seismic microzonation and earthquake damage scenarios for urban
areas. Soil Dynamics and Earthquake Engineering 30, 1319-1328.
Ansal, A., Durukal, E, Tönük,G. (2006). Selection and Scaling of Real Acceleration Time Histories for Site
Response Analyses. Proc. of ETC12 Workshop, Athens, Greece, 93-98.
Borcherdt, R. D. (1994). Estimates of Site Dependent Response Spectra for Design (Methodology and
Justification). Earthquake Spectra 10:4, 617-654.
Bommer, J. J., Scott, S. G., Sarma, S. K. (2000). Hazard-Consistent Earthquake Scenarios, Soil Dynamics and
Earthquake Engineering, 19, 219-231
Bommer, J. J. and Acevedo, A. B. (2004) “The use real earthquake accelerograms as input to dynamic analysis,”
Journal of Earthquake Engineering, 8(1):43-91
Bradley B.A. (2010). A generalized conditional intensity measure approach and holistic ground-motion
selection, Earthquake Engineering and Structural Dynamics, Vol. 39.
Hancock J., Watson-Lamprey J., Abrahamson N.A., Bommer J.J., Markatis A., McCoy E., Mendis R. (2006). An
improved method of matching response spectra of recorded earthquake ground motion using wavelets.
Journal of Earthquake Engineering, 10: 67–89.
Hancock J., Bommer J.J. (2007). Using spectral matched records to explore the influence of strong-motion
duration on inelastic structural response, Soil Dynamics and Earthquake Engineering, 27: 291-299.
Idriss, I. M. ve J. I. Sun (1992). Shake91, A Computer Program for Conducting Equivalent Linear Seismic
Response Analysis of Horizontally Layered Soil Deposits, Modified based on the original SHAKE program
by Schnabel, Lysmer and Seed, 1972.
Iyisan, R. (1996) “Correlations between Shear Wave Velocity and In-situ Penetration Test Results”, Technical
Journal of Turkish Chamber of Civil Engineers, Vol. 7, No. 2, pp.1187-1199 (in Turkish)
PEER, Pacific Earthquake Engineering Research Center, Strong Motion Database, http://peer.berkeley.edu/
NEHRP (2003). Recommended Provisions for New Buildings and other Structures, FEMA-450, prepared by the
Building Seismic Safety Council for the Federal Emergency Management Agency, Washington, DC.
SeismoMatch (2011), Seismosoft Earthquake Engineering Software Solutions
Tönük,G and Ansal,A (2010). Selection and Scaling of Ground Motion Records for Site Response Analysis, 14th
European Conference of Earthquake Engineering, Ohrid, paper no.1386.