Journal of Civil Engineering and Architecture 9 (2015) 410-418

doi: 10.17265/1934-7359/2015.04.005

DAVID

PUBLISHING

Proposal of Indonesia Seismic Hazard Deaggregation

Maps for Sumatra, Indonesia
Achmad Fauzi1, Usama Juniansyah Fauzi2, Hendriyawan2, Wayan Sengara2 and Masyhur Irsyam2
1. Faculty of Civil Engineering and Earth Resources, University of Malaysia Pahang, Pahang 26300, Malaysia
2. Faculty of Civil and Environmental Engineering, Bandung Institute of Technology, Bandung 40132, Indonesia
Abstract: The new method for determining ground-motion parameters in the Indonesian Earthquake Resistant Building Code SNI
(Indonesia National Standard) 03-1726-2012 has significant changes than the previous code. The maps of mean and modal of
magnitude and distance presented here are intended to convey information about the distribution of probabilistic seismic sources and
to provide prescriptions or suggestions for seismic sources to use in developing artificial ground motion in building design or retrofit
projects. This paper presents deaggregation of Indonesia Seismic Hazard Map 2010 for Sumatra. Deaggregation for 0.2-s and 1.0-s
pseudo SA (spectral acceleration) is performed for 10% PE (probability of exceedance) in 50 years (475-year mean return period)
and 2% PE in 50 years (2,475-year mean return period). The information of deaggregation analysis can and perhaps should be
considered in a complex seismic-resistant design decision-making environment.
Key words: Seismic hazard deaggregation, 3D seismic sources, seismic sources contribution, Indonesia Building Code.

1. Introduction

The new Indonesia Seismic Hazard Maps for

revising the previous map in SNI 03-1726-2002 was
published in July 2010 [1-3] as based for next edition
of SNI-03-1726-2012. Seismic source model used in
hazard maps was developed based on various
publications, previous researches, and the latest
information that have been summarized and integrated
by Team for Revision of Seismic Hazard Maps of
Indonesia. Seismic sources were represented by
subduction, fault and background zones that were
represented by three-dimensional models. The new
maps were developed based on PSHA (probabilistic
seismic hazard analysis) method proposed by Cornell
[4].
A principal advantage of the probabilistic method is
combining all the possible earthquakes affecting the
site [4]. Unfortunately, the integrative process of
PSHA also has some disadvantages. One major
Corresponding author: Usama Juniansyah Fauzi, M.Sc.,
research field: geotechnical earthquake engineering. E-mail:
[email protected].

weakness is the lost concept of real earthquake. To

resolve this problem, deaggregation of seismic hazard
can be used to display the contributions to a hazard
value of fundamental parameters such as magnitude
and distance.
The Team for Revision of Seismic Hazard Maps of
Indonesia has produced several new seismic hazard
maps for Indonesia. The final model and maps were
issued in 2010 as summary of Study Team for
Revision of Seismic Hazard Maps of Indonesia [3].
The method and results given in this summary are the
basis for BSN (National Standardization Agency)
recommended seismic design provisions for the
Indonesian Earthquake Resistant Building Code SNI
(Indonesia National Standard) 03-1726-2012. This
summary presented seismic hazard maps computed
for sites on bed rock (shear wave velocity, Vs = 760
m/s2) at the 10% PE (probability of exceedance) in 50
years and 2% PE in 50 years.
Seismic hazard parameters for subduction
considered recurrence relationship that includes
truncated exponential model and pure characteristic

Proposal of Indonesia Seismic Hazard Deaggregation Maps for Sumatra, Indonesia

model. For fault sources, truncated exponential model

and characteristic model with aleatory uncertainty in
the magnitude using a normal distribution sigma of
0.12 were used. For background source, only
truncated exponential model was used in the
development of hazard maps. Several attenuation
functions including NGA (next generation attenuation)
and logic-tree were used. The detail information on
seismic source models and seismic parameters for
development seismic hazard maps appear in
Refs. [5, 6].
This paper, in conjunction with Team for Revision
of Seismic Hazard Maps of Indonesia, intends to
convey information about the distribution of
probabilistic seismic sources with explaining typical
values of earthquake magnitude and distance that are
making the largest contributions to the seismic hazard
maps. Identifying the predominant sources of hazard
will lead to better choices for the design earthquakes
characteristics. Performing deaggregations at more
than one period will help to determine if one source
dominates at all periods and clarify the need for one,
or more than one, design earthquake [7].

2. PSHA and Deaggregation of PSHA

PSHA was developed by Frankel et al. [8] which is
based on the probability concept developed by Cornell
[4]. It is assumed that the earthquake magnitude M,
and distance R, are continuous independent random
variables. In general, form of total probability theorem
can be expressed in the following:
H(a) = viP[A > am,r]fMi(m)fRiMi(r,m)drdm (1)
where, vi is annual rate of earthquakes (with
magnitude higher than some threshold value of Moi) in
source I, and fMi(m) and fRi|Mi(r,m) are probability
density functions on magnitude and distance,
respectively. P[A > a|m,r] is the probability that an
earthquake of magnitude m at distance r produces a
peak acceleration A at the site that is greater than a.
Software for PSHA used in this study was obtained
from the USGS (United States of Geological Survey)

411

[9]. A site spacing of 0.1 in latitude and longitude

and area between 94 E to 142 E longitudes and
12 S to 8 N latitude were used in the analysis. The
ground motion parameters obtained from this study
were computed for sites on bed rock (Vs = 760 m/s2).
The verification seismic models and parameters in this
research with Team for Revision of Seismic Hazard
Maps of Indonesia are shown in Ref. [6].
The method of deaggregation of hazard is to
separate the contributions into a limited number of
bins of (annular) distance, magnitude, and
ground-motion uncertainty [8]. For this research, the
distance annular width, R, is 5 km and the magnitude
bins is 0.5. For subduction sources, the maximum
considered source to site distance is 1,000 km. For
fault and background sources, the maximum
considered source to site distance is 200 km. Using
PSHA result, the relative contributions of sources to
the overall hazard results at the given site are
deaggregated in different types of bins to determine
and understand. The integration of the PSHA is
carried out and the final results are presented often in
terms of 3D M-R- (magnitude-distance-epsilon) bins
or even geographical deaggregation (4D) [10]. Typical
deaggregation result can be seen in Fig. 1.
The mean distance and mean magnitude presented
here are the weighted mean values of R and M,
respectively, for all sources that contribute to hazard
at each grid location, for instance, the mean value of
magnitude and distance in Fig. 1. The modal distance
and modal magnitude are the (R, M) pair having the
largest contribution in the hazard deaggregation at
each grid location, for instance, the peak value of
magnitude and distance in Fig. 1. The maps develop
using the grid increment of 0.1 in both latitude and
longitude and in area between 94 E to 142 E
longitudes and 12 S to 8 N latitude so that
deaggregations seismic hazards are performed for
more than 96,600 sites. Software for deaggregation
PSHA used in this study was obtained from the
USGS.

Proposal of Indonesia Seismic Hazard Deaggregation Maps for Sumatra, Indonesia

Probability density

412

Distance (km)
Fig. 1

Deaggregation hazard result of 500 years return period for Palembang (T = 0.2 s).

3. Earthquake Catalog, Seismotectonic

Model, Recurrence Relations and Ground
Motion Prediction Equation
Seismic parameters used in this study were derived
from published journals, proceedings, previous
researches conducted by team members, and latest
information obtained during this study. This study has
then compiled and integrated previous and current
studies. Earthquake source parameters were
determined based on earthquake catalog, geological,
and seismological information of active faults. The
earthquake catalog covered earthquake period
between 1900 to 2009, relocated catalog by the year
2005, and area between 90 E to 145 E longitudes
and 15 S to 15 N latitudes.
Seismic sources use recurrence relationship that
includes truncated exponential model, pure
characteristic model, and both models. Geometry of
fault and subduction were represented by 3D models
based on the result of tomography and slip-rates of
faults were determined by considering the results of
GPS (global positioning system) measurement as
shown in Fig. 2. Background source zones were modeled

Fig. 2

Seismotectonic model for analysis.

Proposal of Indonesia Seismic Hazard Deaggregation Maps for Sumatra, Indonesia

using gridded seismicity based on spatially smoothed

earthquake rates. The earthquake catalog was used for
developing gridded seismicity starting from 1900 to
2009 and the updated Engdahl catalog up to 2009 was
used for control geometry of subduction [11]. Several
well-known attenuation functions were selected in
accordance with the mechanism of seismic source
including the NGA. Logic tree was also applied to
account for epistemic uncertainty including recurrence
model, maximum magnitude and several attenuation
functions.

4. Results and Discussions

One purpose of deaggregation analysis is to find
plausible (R, M) pairs from which we can choose
accelerograms, A(t), for input to seismic design
programs for structural response. The maps of mean
and modal magnitude and distance for 10% PE in 50
years are shown in Figs. 3-6. Maps of mean and
modal magnitude and distance for 2% PE in 50 years

413

are shown in Figs. 7-10.

The analysis result showed that the maps of modal
results are associated with the highest contribution of
earthquake source. In case of areas near the fault (<
200 km), magnitude and distance are controlled by
fault source. However, in the area located far from the
fault source (> 1,000 km), most magnitude and
distance are controlled by gridded seismicity source.
Then, in the area between 200~1,000 km, most
magnitude and distance are controlled by subduction
source. The mean magnitude and distance are
associated with the average of contribution from
multimodal/multi scenario earthquake. The mean and
modal maps showed different values of pair
magnitude and distance so that most sites receive
contributions from a broad distribution of source
magnitudes and distances. There is a discontinuity in
the contour in Figs. 6 and 10 in their left upper parts.
It is due to higher contribution of Sumatra subduction
zone than faults and background zones.

Mean R (km)
Mean Mw

(a)
(b)
Fig. 3 Maps of 0.2 s spectral acceleration of Sumatra for 10% probability of exceedance in 50 years: (a) mean magnitude; (b)
distance.

414

Proposal of Indonesia Seismic Hazard Deaggregation Maps for Sumatra, Indonesia

Mean Mw

Mean R (km)

(a)
(b)
Fig. 4 Maps of 0.2 s spectral acceleration of Sumatra for 10% probability of exceedance in 50 years: (a) modal magnitude;
(b) distance.

Mean Mw

Mean R (km)

(a)
(b)
Fig. 5 Maps of 1.0 s spectral acceleration of Sumatra for 10% probability of exceedance in 50 years: (a) mean magnitude; (b)
distance.

Proposal of Indonesia Seismic Hazard Deaggregation Maps for Sumatra, Indonesia

Mean Mw

415

Mean R (km)

(a)
(b)
Fig. 6 Maps of 1.0 s spectral acceleration of Sumatra for 10% probability of exceedance in 50 years: (a) modal magnitude;
(b) distance.

Mean Mw

Mean R (km)

(a)
(b)
Fig. 7 Maps of 0.2 s spectral acceleration of Sumatra for 2% probability of exceedance in 50 years: (a) mean magnitude; (b)
distance.

416

Proposal of Indonesia Seismic Hazard Deaggregation Maps for Sumatra, Indonesia

Mean Mw

Mean R (km)

(a)
(b)
Fig. 8 Maps of 0.2 s spectral acceleration of Sumatra for 2% probability of exceedance in 50 years: (a) modal magnitude; (b)
distance.

Mean R (km)
Mean Mw

(a)
(b)
Fig. 9 Maps of 1.0 s spectral acceleration of Sumatra for 2% probability of exceedance in 50 years: (a) mean magnitude; (b)
distance.

Proposal of Indonesia Seismic Hazard Deaggregation Maps for Sumatra, Indonesia

417

Mean R (km)

Mean Mw

(a)
(b)
Fig. 10 Maps of 1.0 s spectral acceleration of Sumatra for 2% probability of exceedance in 50 years: (a) modal magnitude;
(b) distance.

5. Conclusions
For most locations, the deaggregation reveals that
more than one design earthquake will be required for
engineering purposes. The maps of mean and modal
of magnitude and distance presented here are intended
to convey information about the distribution of
probabilistic seismic sources and to provide
prescriptions or suggestions for seismic sources to use
in building design or retrofit projects. The information
of deaggregation analysis can and perhaps should be
considered in a complex seismic-resistant design
decision-making environment.

D.) for providing the authors with technical

discussions and the software for PSHA and PSHA
deaggregation. Financial support for the research was
provided by Bandung Institute of Technology,
Indonesia.

References
[1]

[2]

Acknowledgments
The authors wish to show their sincerest gratitude
to Team for Revision of Seismic Hazard Maps of
Indonesia, for allowing authors to incorporate the
valuable data when the team developed new Indonesia
Seismic Hazard Maps. The authors are also deeply
indebted to USGS (Harmsen, S. D., and Petersen, M.

[3]

[4]

Irsyam, M., Asrurifak M., Hendriyawan, H., Budiono, B.,

Triyoso, W., and Firmanti, A. 2010. Development of
Spectral Hazard Maps for Proposed Revision of
Indonesia Seismic Building Code. Geomechanic and
Geoengineering 5 (1): 35-47.
Irsyam, M., Asrurifak M., Hendriyawan, H., Budiono, B.,
Triyoso, W., and dan Hutapea, B. 2008. Proposal of
Indonesia Seismic Hazard Maps with Three Dimensional
Earthquake Source Using Probabilistic Method. In
Proceeding of Seminar Indonesian Society of
Geotechnical Engineering, CD-ROM. (in Indonesian)
Team for Revision of Sesismic Hazard Maps of Indonesia.
2010. Summary of Study. Research report of Ministry of
Public Works.
Cornell, C. A. 1968. Engineering Seismic Risk
Analysis. Bulletin of the Seismological Society of
America 58 (5): 1583-606.

418
[5]

[6]

[7]

[8]

Proposal of Indonesia Seismic Hazard Deaggregation Maps for Sumatra, Indonesia

Fauzi, U. J. 2011. Deaggregation Maps of Indonesia
with Three Dimensional Earthquake Source Using
Probabilistic Method. Master thesis, Bandung Institute
of Technology. (in Indonesian)
Halchuk, A., and Adams, J. 2004. Deaggregation of
Seismic Hazard for Selected Canadian Cities. In
Proceeding of 13th World Conference on Earthquake
Engineering, No. 2470.
McGuire, R. K. 1995. Probabilistic Seismic Hazard
Analysis and Design Earthquakes: Closing the Loop.
Bulletin of the Seismological Society of America 85:
1275-84.
Frankel, A. D., Petersen, M. D., Mueller, C. S., Haller, K.
M., Wheeler, R. L., Leyendecker, E. V., Wesson, R. L.,
Harmsen, S. C., Cramer, C. H., Perkins, D. M., and
Rukstales, K. S. 2002. Documentation for the 2002

Update of the National Seismic Hazard Maps. US

Geological Survey, Open file report.
[9] Harmsen, S. C., and Frankel, A. 2001. Geographic
Deaggregation of Seismic Hazard in the United States.
Bulletin of the Seismological Society of America 91 (1):
13-26.
[10] Engdahl, E. R., Villasenor, A., de Shon, H. R., and dan
Thurber, C. H. 2007. Teleseismic Relocation and
Assessment of Seismicity (1918-2005) in the Region of
the 2004 Mw 9.0 Sumatra-Andaman and 2005 Mw 8.6
Nias Island Great Earthquakes. Bulletin of the
Seismological Society of America 97: S43-61.
[11] Harmsen, S. C., Frankel, A. D., and Petersen, M. D.
2003. Deaggregation of US Seismic Hazard Sources:
The 2002 Update. US Geological Survey, Open file
report.