You are on page 1of 15

Comparison of Code-Based Design Spectra and Site-Specific Response

Spectra in San Francisco

Pawan Kumar, M. ASCE1, Jongwon Lee, Ph.D. P.E. M. ASCE2, Martin Walker, P.E. G.E.
M.ASCE 3 , Reza Baradaran, P.E., G.E.4, and Robert Chew, P.E., G.E. M. ASCE 5

1
Arup, 560 Mission St, Suite 700, San Francisco, CA 94105; e-mail: pawan.kumar@arup.com
2
Arup, 560 Mission St, Suite 700, San Francisco, CA 94105; e-mail: jongwon.lee@arup.com
3
Arup, 560 Mission St, Suite 700, San Francisco, CA 94105; e-mail: martin.walker@arup.com
4
San Francisco Public Works, 30 Van Ness Ave., 5th Floor, San Francisco, CA 94102; e-mail:
Reza.Baradaran@sfdpw.org
5
Robert Y Chew Geotechnical, 55 New Montgomery Street, San Francisco, CA 94105; e-mail:
Robert.chew@robertchewgeotech.com

ABSTRACT

This paper presents seismic site response studies performed in 2016 and 2017 for two projects in
San Francisco, California, carried out with San Francisco Public Works. The design rock input
motions developed for the projects are also discussed, which are used by a companion paper on
the structure-soil-structure interaction (SSSI) analyses. With a focus on how the surface response
spectra compare with the ASCE 7-10 code-based design response spectra, as well as the resulting
reduction of design spectral amplitudes allowed per the design code by performing a site-specific
study, the findings of the study indicate that the code minimum spectra tend to be conservative in
the short period range for soft clay sites. One of the project sites has the underlying Franciscan
Complex Bedrock sloping considerably across the project area, which results in different site
periods at either end of the proposed development. It was found that it is important to consider
multiple soil models to represent the varying site periods.

INTRODUCTION

The seismicity and the ground conditions in the San Francisco Bay Region (SFBR) offer a
unique set of challenges for performing seismic hazard assessment (SHA) and site response
analyses (SRA) for the design of geo-seismic components of the built environment. It is crucial
that the seismic community in the region understands the regulatory building code guidelines, in
particular ASCE 7-10 (ASCE, 2010) and future revisions, and the ramifications of prescriptively
using the simplified code-based approach for soft soil sites versus performing a site-specific
study. In this context, this study looks at two project sites in San Francisco, California: San
Francisco Police Department Traffic Company Forensic Services Division Site (TCFSD site) and

  – 1 –    
Mariposa Pump Station Improvements (Mariposa Site). See Figure 1 for the location of these
sites. The SHA and SRA analyses for these projects were completed under a San Francisco
Public Works (SFPW) contract for as-needed geotechnical services.

Figure 1. Location of the Mariposa and TCFSD sites in San Francisco, California

OVERVIEW OF PROJECT SITES

TCFSD Site. The TCFSD site is located in the Bayview District of San Francisco. The site is
roughly rectangular and measures approximately 75 m by 120 m. The project includes a new,
10,000 square-meter, three-story laboratory and office building, which is categorized as an
essential facility for design compliant with the 2016 California Building Code (2016 CBC).

Mariposa Site. The Mariposa Pump Station Improvements site is located in the San Francisco
Bay (Bay) margin area of San Francisco called Mission Bay. Mission Bay was successively infilled
from a marsh and tidal flat area to the dense residential, commercial, and medical facility zone it
is today. The area of reclaimed land extends west to the Union Pacific Railroad and Highway 280.
The Mariposa Pump Station includes an existing dry weather pump station and long wet
weather transport/storage structure (T/S box). The existing dry weather pump station constructed
in 1954 is to be completely demolished and replaced with a new station with a different location
and orientation. Thus, an oversized excavation will be required to facilitate both demolition of the
existing structure and construction of the new station. The design of the new pump station shows
a single-story concrete structure with a footprint of approximately 10.5 m by 10.5 m. The above-
grade structure is 5 m high and the basement is roughly 6.5 m deep. The existing T/S box located
a few meters to the north of the new pump station was constructed circa 1989. The companion
paper on the SSSI analyses (Ellison et al., 2018) provides a detailed description of the structural
components of the pump station system.

  – 2 –    
OVERVIEW OF GEOLOGY AND GROUND MODELS

The TCFSD and Mariposa sites are located within the north and the south quadrangle of the San
Francisco Geological Map, respectively (Schlocker, 1974; Bonilla, 1998). This region is
characterized by hills composed of Jurassic- to Cretaceous-aged bedrock surrounded by flat,
low-lying areas underlain by marine and non-marine Quaternary deposits.

TCFSD Site. The TCFSD site is located west of the Islais Creek Channel. Figure 2 shows the 
interpreted subsurface stratigraphy at the TCFSD site.

Figure 2. Northwest-southeast subsurface stratigraphic cross-section at the TCFSD site

The subsurface stratigraphy was interpreted based on recent and historical geotechnical
investigation, including exploratory boreholes, field vane shear tests, cone penetration testing
(CPTs), downhole PS suspension logging (PS logging) and associated laboratory testing. The site
stratigraphy includes a layer of artificial fill 2.5 to 5.5 m thick underlain by Bay Mud,
characterized as fat clay containing shell fragments and dark organic rich pockets, with thickness
varying from about 9 to 26 m. Beneath the Bay Mud, Undifferentiated Marine Deposits were
encountered at most locations across the site, overlying a stiff to very stiff fat clay of marine and
estuarine origin known as Old Bay Clay across the eastern portion of the site, which tapers out
towards the northern and northeastern portion of the site. A 3.5-meter-thick layer of colluvium
was encountered in the southeast portion of the site, overlying the weathered bedrock.
The Franciscan Complex Bedrock at the site is characterized as moderately weathered,
moderately soft to moderately hard, weak to medium strong serpentinite. The top of bedrock
slopes considerably at the site, with depths ranging from approximately 15 m in the northwest to

  – 3 –    
45 m in the southeast. For the SRA, two separate ground models were developed: northwest
(NW) and southeast (SE) to represent the varying depth to bedrock and thickness of the soft Bay
Mud. These two models are intended to capture the different frequency contents of surface
ground motions resulting from the site variability.
Figure 3 shows the two ground models (NW and SE), along with the key soil properties:
unit weight, overconsolidation ratio (OCR), shear strength, and shear wave velocity (Vs). Note
that the shear wave velocity (Vs) profile was derived from PS logging and CPT-based
correlations (Andrus et al., 2007), and the undrained shear strength of the clays was based on
field vane shear test data as well as CPT correlations (Lunne et al. 1997), whereas for sands it
was based on the vertical effective stress and friction angle. Unit weight and plasticity index (PI)
were derived from laboratory testing, and CPT correlations were utilized for the at-rest earth
pressure coefficient (K0; Robertson, 2009) and OCR (Kulhawy and Mayne, 1990). The results of
the correlations were compared to the knowledge of typical values for these parameters in the
region. The ground model for the NW profile was a truncated version of the SE profile, with
appropriate adjustments for layer thicknesses and overburden stress.

Figure 3. Ground model and soil properties for the TCFSD site.

Mariposa Site. The interpreted subsurface cross-section at the Mariposa site is shown in Figure
4. The soil units encountered are similar to those at the TCFSD site, with approximately 5 to 6 m
of loose to medium dense artificial fill and 3.5 to 5 m of soft to medium stiff Bay Mud. These
strata overlie alluvial deposits and bedrock. Bedrock at this site typically comprises a weathered
sheared shale matrix with variable sized blocks of greywacke sandstone and siltstone. Unlike the
TCFSD site, the inclination of the top of bedrock was not significant over the small footprint of
the site to warrant multiple ground models for free-field site response. Figure 5 shows the ground
model and soil properties, developed using an approach similar to the TCFSD site. 

  – 4 –    
Figure 4. North-south subsurface stratigraphic cross-section at the Mariposa site

Figure 5. Ground model and soil properties for the Mariposa site.

SEISMIC HAZARD ASSESSMENT

The seismic setting of the TCFSD and the Mariposa sites are similar to each other, including the
source-to-site distance from the governing San Andreas fault. Also, the measured bedrock shear
wave velocity (Vs) is similar at both sites. As a result, one single hazard assessment was
applicable to both the sites. The TCFSD site is classified as Site Class E, whereas the Mariposa

  – 5 –    
site was borderline between Site Class D and E based on the data from the geotechnical
investigation. The decision to classify the site as Class D (slightly lower simplified method
spectrum than Class E) was reasonable because the surface spectra from the SRA would be
lower than the code minimum spectra, as demonstrated in the results and conclusions sections.

Seismic Hazard. Earthquakes in the SFBR are induced by the northwestward movement of the
Pacific Plate relative to the North American Plate. The major active faults in SFBR include the
San Andreas, Hayward, Calaveras, and the San Gregorio faults. The fault governing the
earthquake hazard for the project site is the San Andreas Fault, located approximately 13 km
from both the project sites.
The San Andreas fault is a right-lateral strike-slip fault forming the tectonic boundary
between the North American Plate and the Pacific Plate. The San Andreas fault alone accounts
for more than half of the geologic slip rate across the SFBR and more than 90% of the SFBR
seismic moment sum in the historical record.
For the seismic hazard assessment, this study used the 2014 update of the National
Seismic Hazard Study (Petersen et al., 2014), which employed the Uniform California
Earthquake Rupture Forecast – Version 3 (UCERF3; Field, 2015) for the California region. The
UCERF3 was developed and reviewed by dozens of leading scientific experts from the fields of
seismology, geology, geodesy, paleoseismology, earthquake physics, and earthquake engineering
(UCERF3; Field, 2015). As such, it represents the latest earthquake-rupture forecast and the best
available science in the seismic hazard model. The code-compliant response spectra based on the
2014 National Seismic Hazard Study are discussed in the following subsection.
  Hazard deaggregation was also obtained from the USGS web-based tool (USGS, 2008),
but based on the 2008 National Seismic Hazard Study (Petersen et al., 2008) in lieu of the 2014
update. The deaggregation per the 2014 update was not available to the public at the time of this
study, but no significant change is expected considering the updates in the model. Figure 6
shows the hazard deaggregation for the TCFSD site for spectral period of 1.0 sec (close to the
structural period of 1.0 to 1.2 sec for the TCFSD) and a return period of 2,475 years for Site
Class B. As shown in Figure 6, most of the hazard is contributed by an MW and R (source-to-site
distance) pair of MW 8.0 and R=11.3 km (7.1 miles), corresponding to the San Andreas fault.
Note that a similar trend is observed for the Mariposa site as well, and hence not presented
herein. A similar trend was also observed for the deaggregation corresponding to the ground
surface level (Site Class E for TCFSD and Site Class D for Mariposa).  

Simplified Method Response Spectrum. The 2013 California Building Code (CFR, 2013)
refers to the design code by American Society of Civil Engineers ASCE 7-10 (2010) for the
development of seismic design response spectra. Using the USGS web-based tool (USGS, 2014)
with the 2014 Seismic Hazard Update, we determined the “simplified method” response spectra
for two different site conditions: bedrock and ground surface corresponding to Site Class B and
Site Class D/E, respectively. In accordance with Ch. 11.4.5 of ASCE 7-10, the simplified method

  – 6 –    
response spectrum is defined as the Risk-Targeted Maximum Considered Earthquake (MCER)
response spectrum. The MCER response spectrum represents the target 1% risk of structural
collapse in 50 years with risk targeting factors applied to the acceleration MCE response
spectrum. The MCE response spectrum is defined by the lower of the maximum direction 2%
probability of exceedance in 50 years (2,475-year return period) response spectrum limited by an
84th percentile deterministic cap.
Figure 7 shows the simplified method bedrock and surface (Site Class E for
TCFSD and Site Class D for Mariposa) MCER spectra. The bedrock MCER is used herein as a
target spectrum to generate rock input motion time histories for SRA, and 80% of the surface
MCER (80%MCER) is used as the code minimum surface spectrum per Ch. 21.3 of ASCE 7-10.

INPUT ROCK MOTION TIME HISTORIES

To account for the variability of ground motions in waveform, 11 pairs of horizontal motion time
histories were spectrally matched to the bedrock MCER for the TCFSD site (see Figure 7). The
matching was applied over the spectral period range of 0.2 to 2 sec, covering 0.2T1 to 1.5T1
(ASCE7), where T1 is the fundamental structural period.
Note that the same set of rock input motions is valid for the Mariposa site as well, which
required seven pairs of time histories for the SSSI analyses. Therefore, a subset of seven pairs
from the set of 11 pairs developed for TCFSD was used for Mariposa.

Seed Selection. The selected seed records are summarized in Table 1. The seed records for all
the eleven pairs were selected from the Pacific Earthquake Engineering Research Center (PEER)
Ground Motion Database – NGA-West2 (PEER, 2016). We selected only one record per
earthquake to incorporate variability of ground motions from different events. The selection
criteria considered herein are summarized below (in decreasing order of priority):
 Spectrum Shape – The response spectrum of seed motions comparable to the target
response spectrum is important to preserve the original waveform characteristics.
 Tectonic Setting and Fault Type – The site is in an active shallow crustal region. Six seed
records are strike-slip (like the San Andreas fault), and the rest are reverse or reverse-oblique.
 Earthquake Magnitude (MW) – MW 8.0 is the hazard-controlling magnitude. Most of the
selected seed records are from MW 7.0+ events, with some records from the MW < 7.0 range,
which was unavoidable due to the large data gap in large magnitude, near-field events.
 Source-to-Site Distance (R) – The hazard-controlling R is 11.3 km. Our selection includes
records with R less than 50 km except for one record from the 1989 Loma Prieta event,
which was included due to its close proximity to the project sites.
 Site Class (VS30) – Site Class B for bedrock was considered. We extended the boundary to
Class C (360 to 760 m/sec); extending up to one Site Class is generally acceptable, especially

  – 7 –    
given the lack of data that satisfy all the above search criteria. The two motions with VS30
closer to the lower bound of Class C were also deemed acceptable, as some of the site effects
will be modified through the spectral matching process (described in the next sub-section).

Figure 6. Hazard deaggregation for the TCFSD site for T = 1.0 sec, 2,475-year (USGS,
2008)

Figure 7. Bedrock and surface MCER spectra for the TCFSD and Mariposa sites

Spectral Matching Analysis. Spectral matching was completed using the computer program,
RspMatch09 (Al-Atik and Abrahamson, 2010). We adjusted the seed time histories so that their
response spectrum is compatible to the target design response spectrum. RspMatch09 uses a
time-domain spectral matching method, where adjustment of initial time series is made by
adding wavelet functions to the initial acceleration time history. The acceleration, velocity and
displacement time histories are compared before and after spectral matching to verify the

  – 8 –    
preservation of the original waveforms. Also, the energy distribution and significant duration of
the spectrally-matched time histories are examined to check that the original engineering
characteristics of the seed time histories are preserved. To facilitate the spectral matching, the
seed records are first scaled to match the peak ground acceleration of the target spectrum.

Table 1. Seed Records for the TCFSD and Mariposa sites


Record 5-95% Duration*
Earthquake Rrup VS30
Sequence Year Station Name MW Mechanism (sec)
Name (km) (m/sec)
Number Seed Matched
Yerba Buena Reverse
813** Loma Prieta 1989 6.9 75 660 17 ; 9 20 ; 9
Island Oblique
Twentynine
897** Landers 1992 7.3 Strike slip 41 635 31 ; 30 31 ; 31
Palms
Chi-Chi, Reverse
1230** 1999 CHY079 7.6 48 573 38 ; 31 38 ; 33
Taiwan Oblique
Sitka
1626** Sitka, Alaska 1972 7.7 Strike slip 35 650 28 ; 26 29 ; 30
Observatory
1633** Manjil, Iran 1990 Abbar 7.4 Strike slip 13 724 29 ; 29 28 ; 31
1111 Kobe, Japan 1995 Nishi-Akashi 6.9 Strike slip 7.1 609 10 ; 11 15 ;12
5818 Iwate, Japan 2008 Kurihara City 6.9 Reverse 13 512 13 ; 15 18 ; 21
El Mayor-
Rancho
5830** Cucapah, 2010 7.2 Strike slip 45 524 34 ; 33 34 ; 33
San Luis
Mexico
Darfield,
6891** 2010 CSHS 7.0 Strike slip 44 638 24 ; 29 31 ; 31
New Zealand
LA – Sepulveda Reverse
1004 Northridge-01 1994 6.7 8 380 8;9 11 ; 9
VA Hospital Oblique
4228 Niigata, Japan 2004 NIGH11 6.6 Reverse 9 375 10 ; 10 8;9
* The two values of D5-95 correspond to the two components of the earthquake record
** Subset seed records used for Mariposa Site

To ensure the matched time histories contain the energy content comparable to the target
design spectra, we confirmed that the average spectrum of the matched motions does not fall
below the target response spectrum at any spectral period over the period range of interest (0.2 to
2 sec). Figure 8 shows the average response spectrum of all 22 matched time histories along with
the target response spectrum and the period range of interest. Also, the significant duration, D5-95
of the matched motions is compared with the median D5-95 predicted by the empirical
relationship presented by Kempton and Stewart (2006). As inputs to the empirical relationship,
we used the hazard-controlling MW of 8.0, R=11.3 km, and bedrock average shear wave velocity
(VS30) of 945m/s. The median D5-95 of the matched motions is 31 sec, comparable to the
predicted median of 33 sec.

SITE RESPONSE ANALYSIS

This site response analysis for the TCFSD site employed the 1-dimensional site response
analysis program, DEEPSOIL (version 6.1; Hashash et al. 2016). For the Mariposa site, since
SSSI analyses were required, the site response was performed in the finite element program, LS-
DYNA. The freefield surface response from LS-DYNA was validated with selected DEEPSOIL

  – 9 –    
runs for the Mariposa soil profile. This section will mainly focus on the methodology adopted for
TCFSD. See the companion paper (Ellison et al., 2018) for details on the LS-DYNA analyses.

Figure 8. Average response spectrum of matched motions and the target spectrum and
period of interest for the TCFSD site (unshaded).

DEEPSOIL allows users to perform linear, equivalent-linear, and nonlinear SRA. To


account for soil nonlinearity, the equivalent-linear analysis and/or nonlinear time domain
analysis can be used. However, the equivalent-linear approach may not be appropriate for a soft
site subject to high intensity motions because the short-to-intermediate period content may be
underestimated (Hashash et al., 2010). We compared the empirical index threshold (Iγ)
established by Kim et al. (2016) to those calculated for the two TCFSD site profiles (see Figure
9). As the Iγ values exceeded the threshold, we adopted the nonlinear time domain SRA for our
analysis. We selected the non-Masing option for reloading/unloading stress-strain curves since
hysteretic damping can be overestimated when using the Masing rules (Kwok et al., 2007;
Hashash et al., 2010).

Figure 9. Shear strain index threshold and index values for the TCFSD site

  – 10 –    
The SRA input nonlinear backbone curves representing the shear modulus degradation
(G/Gmax) and damping for the soil layers were based on Darendeli (2011). The G/Gmax curves
were adjusted to have implied shear strengths comparable to the design shear strength profiles.
Further details on the site response models have been excluded due to the page length limitation
for the conference proceedings.

RESULTS

The DEEPSOIL SRA for the TCFSD site consisted of a total of 44 analyses: 22 time histories
(11 pairs) propagated through two soil models (NW and SE). The LS-DYNA SRA for the
Mariposa site used 7 pairs of time histories, with the two orthogonal components applied
simultaneously to simulate bi-directional shaking (Ellison et al., 2018).

TCFSD Site. The spectral amplification at the TCFSD site’s SE and NW profiles for all 11 pairs
of motions are presented in Figure 10. The elastic site periods for NW and SE are 0.7 and 1.4 s,
respectively. As shown in Figure 10, the elongated site periods due to the soil nonlinearity are 2
and 3.2 sec for NW and SE, respectively. This difference in site periods is due to the bedrock
depth being much deeper and the much thicker Bay Mud for the SE profile.

Figure 10. Spectral amplification for the NW and SE profiles of the TCFSD site

Figure 11 presents the maximum shear strain profiles. The high maximum shear strains
occur within the Bay Mud as expected, considering its low shear wave velocities. The large shear
strains damp out most of the short-period content of the bedrock input motions.
Figure 12 presents the major axis response spectra for the two TCFSD site profiles, the
80%MCER (Class E), and the code-compliant envelope spectrum of the SRA result and
80%MCER. The major axis response spectra were determined as the maximum response
spectrum of a pair of horizontal motions over all orientations. As shown in the figures, the
80%MCER spectrum governs for most of the periods except for the elongated site periods. Figure
12 also shows the 100% MCER spectrum at surface per Ch. 11.4.5 of ASCE 7-10 (Class E),
along with the three resulting site-specific surface MCER spectra calculated per Ch. 21 of

  – 11 –    
ASCE 7-10: SE, NW, and envelope spectra. These spectra were obtained by smoothing out the
spectra from the SRA. As evident from the figure, the SRA developed spectra are generally
lower than the simplified method MCER spectrum except for the site fundamental periods.

Figure 11. Maximum shear strain profiles for the NW and SE profiles of the TCFSD site
 

Figure 12. (a) Major Axis Spectra for the NW and SE profiles of the TCFSD site,
80%MCER and Code-Compliant Envelope; (b) Proposed site-specific MCER spectra at
surface level and simplified method surface MCER for Site Class E

  – 12 –    
Mariposa Site. The SRA results for the Mariposa site for all 7 pairs of motions are presented in
Figure 13. In general, the response is amplified at periods greater than about 0.5 sec, particularly
in the 1.0 to 2.0 sec range. The reported surface motions are generally attenuated at periods less
than about 0.5 sec, similar to those for the TCFSD site. The maximum shear strain profiles, like
those for the TCFSD site show high shear strains in the Bay Mud.
Figure 14 shows the resulting smoothened site-specific surface MCER spectrum
calculated per Ch. 21 of ASCE 7-10, with the 80%MCER spectrum governing for most of the
periods except for the elongated site periods at this site.

Figure 13. SRA results for the Mariposa site: (a) Maximum shear strain profile; (b)
Spectral amplification; (c) Surface spectra

Figure 14. Proposed site-specific MCER spectra at surface level and code-compliant surface
MCER for the Mariposa site (Site Class D)

  – 13 –    
CONCLUSIONS

This study presents a comparison of the codified minimum response spectra with those obtained
from site-specific response analyses for two San Francisco Public Works project sites in San
Francisco, California.
We conclude that the codified minimum spectra tend to be conservative in the short
period range for Bay Mud sites as the soft clay significantly damps out the short period content.
Also, for a site with significant inclination of bedrock (or any other significant unit), it is
important to consider multiple ground models to capture the range of frequency contents.
The SRA for the TCFSD and Mariposa sites indicated that the 0.2 second spectral
ordinate was far below the code minimum. This has direct ramifications for structural design,
particularly calculation of base shear, potentially leading to increased construction costs. For
projects that employ peer-review of structural performance, the soil-structure interaction and
time history analysis of below- and above-grade structures may ultimately save significant costs
by departing from code minimums that do not adequately capture common soft-soil site effects.

ACKNOWLEDGEMENTS

The authors would like to thank Frank Greguras for inputs on the geotechnical interpretation, and
Michael Rossetto for his guidance and overall management of the TCFSD project, and Sue Chen,
William Leung, and Rich Graham for their guidance on the Mariposa project. We would also
like to thank Abraham Gebrezgiabhier and Dee Dee Maggi for assistance in generating graphics
for the paper.

REFERENCES

Andrus, R.D., Mohanan, N.P., Piratheepan, P., Ellis, B.S., and Holzer, T.L. (2007). “Predicting
shear-wave velocity from cone penetration resistance”. Proc. 4th Int Conf Eqk Geotech
Engg, Thessaloniki, Greece.
ASCE (American Society of Civil Engineers). (2010) Minimim Design Loads for Buildings and
Other Structures, Standard ASCE/SEI 7-10. ASCE, Reston, VA.
Atik, L.A. and Abrahamson, N. (2010). “An improved method for nonstationary spectral
matching”. Earthquake Spectra, Vol. 26 (3), pp. 601–617.
Bonilla, M. G. (1998). “Preliminary geologic map of the San Francisco South 7.5'quadrangle and
part of the Hunters Point 7.5'quadrangle, San Francisco Bay area, California: A digital
database”. USGS Report, Open-File Report 98-354.
California Code of Regulations (CFR) (2013). Title 24, Part 2, California Building Code (CBC).
California Code of Regulations (CFR) (2016). Title 24, Part 2, California Building Code (CBC).

  – 14 –    
Ellison, K., Masroor, A., Walker, M., Liang, W., Chen, S., Kwan, T. and Tam, B. (2018). “SSI
versus SSSI for adjacent pump stations in San Francisco”. Submitted to Geotechnical
Earthquake Engineering and Soil Dynamics V, Austin, TX, May 2018.
Field, E. H. (2015). “UCERF3: A new earthquake forecast for California's complex fault
system”. USGS, No. 2015-3009.
Hashash, Y.M.A., Phillips, C., and Groholski, D.R. (2010). Recent advances in non-linear site
response analysis. 5th International Conference on Recent Advances in Geotechnical
Earthquake Engineering and Soil Dynamics, San Diego, California, May 24–29.
Hashash, Y.M.A., Musgrove, M.I., Harmon, J.A., Groholski, D.R., Phillips, C.A., and Park, D.
(2016). DEEPSOIL 6.1, User Manual.
Kempton, J.J. and Stewart, J.P. (2006). “Prediction Equations for Significant Duration of
Earthquake Ground Motions Considering Site and Near-Source Effects”. Earthquake
Spectra, 22 (4), 985–1013.
Kim, B., Hashash, Y.M.A., Stewart, J.P., Rathje, E.M., Harmon, J.A., Musgrove, M.I.,
Campbell, K.W., and Silva, W.J. (2016). “Relative differences between Nonlinear and
equivalent-linear 1D site response analyses”. Earthquake Spectra, 32 (3).
Kulhawy, F.H. and Mayne, P.H. (1990). “Manual on estimating soil properties for foundation
design”. Report EL-6800 Electric Power Research Institute, August.
Kwok, A.O.L., Stewart, J.P., Hashash, Y.M.A., Matasovic, N., Pyke, R., Wang, Z., and Yang, Z.
(2007) Use of exact solutions of wave propagation problems to guide implementation of
nonlinear, time-domain ground response analysis routines. J Geotech and Geoenv Engg,
133 (11).
Lunne, T., Robertson, P.K., and Powell, J.J.M. (1997). Cone penetration testing in geotechnical
practice. Blackie Academic, EF Spon/Routledge Publ., New York.
PEER (2016). PEER Ground Motion Database, last accessed on August 19, 2016,
http://ngawest2.berkeley.edu/spectras/new?sourceDb_flag=1
Petersen, M.D., Frankel, A.D., Harmsen, S.C., Mueller, C.S., Haller, K.M., Wheeler, R.L.,
Wesson, R.L., Zeng, Yuehua, Boyd, O.S., Perkins, D.M., Luco, Nicolas, Field, E.H.,
Wills, C.J., and Rukstales, K.S. (2008). “Documentation for the 2008 update of the
United States national seismic hazard maps”. U.S. Geological Survey Open-File Report
2008–1128, 128 p.
Petersen, M.D., Moschetti, M.P., Powers, P.M., Mueller, C.S., Haller, K.M., Frankel, A.D.,
Zeng, Yuehua, Rezaeian, Sanaz, Harmsen, S.C., Boyd, O.S., Field, Ned, Chen, Rui,
Rukstales, K.S., Luco, Nico, Wheeler, R.L., Williams, R.A., and Olsen, A.H., (2014).
“Documentation for the 2014 update of the United States national seismic hazard maps”.
U.S. Geological Survey Open-File Report 2014–1091, 243 p.
Schlocker, J. (1974). “Geologic Map of the San Francisco North Quadrangle, San Francisco and
Marin Counties, CA”. USGS Report, Professional Paper 782.
USGS (2008). 2008 Interactive Deaggregations, http://geohazards.usgs.gov/deaggint/2008/
USGS (2014). U.S. seismic design maps, http://earthquake.usgs.gov/designmaps/beta/us/

  – 15 –    

You might also like