Continental Shelf Research 79 (2014) 46–53

Contents lists available at ScienceDirect

Continental Shelf Research

journal homepage: www.elsevier.com/locate/csr

Research papers

Tsunami flood modelling for Aceh & west Sumatra and its application for an early
warning system
B.A.D. Van Veen a,n, D. Vatvani b,1, F. Zijl b
a
WitteveenþBos, Banda Aceh, Indonesia
b
Deltares (WL9Delft Hydraulics), Banda Aceh, Indonesia

article info Summary

Article history: For implementation of a regional Tsunami Early Warning System (EWS) in Sumatra island in Indonesia, a set of detailed and
Received 26 July 2011 accurate tsunami propagation and flooding models using Delft3D were developed. The purpose of the models was not only to
Received in revised form reproduce the 2004 Indian Ocean tsunami, but also to determine tsunami flood hazard maps with different return periods. The
9 March 2012
model outputs have then been used to build a tsunami flooding database covering 1250 hypothetical sources for different
Accepted 20 August 2012
Available online 4 September 2012 earthquake parameters along the Sunda Trench for an EWS called RiskMap. The model simulations produced detailed
information of near-shore tsunami wave height, tsunami inundation length and run-up. Smart storage of computational results,
Keywords: in a geo-referenced database, allows quick access to the requisite information. The result is a system capable of issuing a
Tsunami modelling
warning within few minutes after a detection of an earthquake. The system has been successfully installed and tested in the last
Tsunami flooding
two years at national and regional emergency coordination centres, National Agency for Meteorology, Climatology and
Early warning system
Geophysics (BMKG) and at Tsunami Disaster Mitigation Research Centre (TDMRC) in Banda Aceh.
Flood hazard map
Risk map

& 2012 Elsevier Ltd. All rights reserved.

1. Introduction – Producing tsunami flood hazard maps with different return periods.

Tsunami restoration and recovery activities in the Aceh province in
Indonesia, caused by the Indian Ocean tsunami of 26th December 2004, have
been concluded by the closure of the responsible agency, known as the
‘Agency for Rehabilitation and Reconstruction’ (BRR, Badan Rehabilitasi
dan Rekonstruksi). The Agency coordinated all the recovery
activities/projects of the Government of Indonesia, local governments,
NGO’s, foreign donors and UN organisations. One of the projects coordinated
by the BRR was a project that had a primary objective of putting in place an
appropriate strategy for coastal protection, flood protection, multifunction
refuge construc-tion and a regional early warning system that is to linked to
the national EWS. The project was funded by the Netherlands govern-ment
and carried out by a consortium called Sea Defence Consul-tants (SDC),
consisting of DHV Consultants, WitteveenþBos and Deltares. The project
focused, among others, on:
– Developing a set of models to reproduce the tsunami event of 2004.
n Corresponding author. Present address: Van Twickelostraat 2, 7411 SC Deventer,
– Studying different coastal/tsunami protection options for the city of Banda
Aceh.
– Implementing a regional tsunami EWS that is to be linked with the national
tsunami EWS.
Tsunami Early warning systems in other places in the world, such as
PTWC, are principally based on two steps approach:
– Issue an alert when an earthquake event is recorded that has a potential to
generate tsunami and
– issue a warning in all countries bordering the ocean when a tsunami has
been confirmed by one or more data buoys and or tidal gauges.
This method works when areas are located are far away from the tsunami
source and sufficient time is available to evaluate the data before taking the
decision to issue a warning. Tsunamis in Aceh and West-Sumatra arrive form
nearby source with arrival times in the order of 15–30 min, just like in Japan.
So evaluation of tsunami hazard can only be based on earthquake information
available shortly after the event occurred.

The Netherlands, Tel.: þ31 652 665 756. The national system in Indonesia evaluates the tsunami hazard potential
E-mail addresses: b.vveen@witteveenbos.nl (B.A.D. Van Veen),
deepak.vatvani@deltares.nl (D. Vatvani), firmijn.zijl@deltares.nl (F. Zijl). based on various information sources leading to two types of warning levels
1
Present address: Rotterdamseweg 185, 2629 HD Delft, The Netherlands, Tel.: þ31 88 335 at provincial levels: ‘tsunami watch’ or ‘tsunami warning’. The system in
Aceh and West-Sumatra
8273.
0278-4343/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.csr.2012.08.020
 B.A.D. Van Veen et al. / Continental Shelf Research 79 (2014) 46–53 47

described here provides a more detailed warning on village com-plements the

national system.

2. Tsunami modelling

Initially in 2005 and early 2006 the project was requested to develop a
strategy and determine priorities for reconstruction after taking the
vulnerability of the coast for future tsunami occurrences into account. As
inundation data were then scarcely available, except at larger cities, flooding
simulations generated by numerical model were deemed necessary to get a
good overall picture of areas that had been hit in the province.

2.1. 2004 Case

A comprehensive set of models using Delft3D, online coupled, were then

set-up to compute the tsunami propagation and inundation caused by the
earthquake of 2004. These set of models contained over a million grid points
covering the entire Indian Ocean, with a resolution of 1000 m and above, and
the coastal areas in the province of Aceh and Nias with a resolution of 50–200
m (SDC (2007)), (see also). The digital elevation model is based on SRTM
data; SRTM data has a vertical error of upto

þ / 16 m (Berthier et al. (2006)). In the city of Banda Aceh the data was
supplemented by surveyed spot heights data. Experi-ments with preliminary
models, with similar coverage but con-taining fewer amounts of grid cells,
were able to yield very good approximations of the initial tsunami source
(Vatvani et al. (2005a)), which is essential for the reproduction of the tsunami
wave, arrival times and flooding pattern (Fig. 1).
The tsunami wave heights simulated by the propagation model compared
well with sea level heights measured by the radar altimeter satellites that
recorded the running tsunami wave in the Indian Ocean on December the 26th
2004 as shown in Fig. 2 (Vatvani et al. (2005a)). Similarly this also holds for
the computed arrival times when compared to observed data at various sites in
Indian Ocean (see Table 1 and Vatvani et al., 2005b).
Furthermore the simulated tsunami flooding in the city of Banda Aceh has
been compared with the surveyed data (see Fig. 3). Overall the resuts
compare well with observed data, with under prediction of the flooding east
of the city (near the village of LhokNga). Overall, we can conclude that
results obtained are quite accurate. Most of the inaccuracies could be
afterwards primarily attributed to inaccurate DEM and bathymetry. This is
especially true for the stretch of coast near the village of LhokNga.
Accurate inundation was obtained due to application of a special
numerical scheme in solving the Non-Linear Shallow Water equa-tions
(NLSW) (Stelling and Duinmeijer (2003); Hesselink et al. (2003)). This
solver is a finite difference scheme that combines the efficiency of staggered
grids with momentum conservation proper-ties to ensure accurate results for
rapidly varying flows. The Delft3D model has been validated against
numerous laboratory experiments with very good results (Apotsos et al.
(2010)). The models have also been applied to simulate sediment transport
and morphological changes near Calang (West Aceh) due to the tsunami with
encoura-ging results (Gelfenbaum et al. (2007)).
With respect to the neglection of the dispersive effects, despite earlier
indication from Horrillo et al. (2006) on the importance of this effect, later
studies (Shigihara and Fujima (2006); Pujiraharjo and Hosoyamada (2008))
have concluded that the dispersion effect is negligible. They also show that
run-up computation with NLSW models (at least in Northern Sumatra) is
more reliable for practical computation purposes as they give more consistent
results (to observed data) compared to Boussinesq type of models.
Fig. 1. Flooding model area embedded in the overall Indian Ocean model (blue coloured, with a
resolution of approximately 1 km near the coast). Other colours separate the different
inundation models. All inundation models have a resolution of 200 m except in the city of
Banda Aceh (red coloured area) that has a resolution of 50 m). (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this
article.)
2.2. Flood hazard maps
The next challenge for the project was the determination of the hazard of
the Aceh and Nias coasts for future tsunami events. Various preliminary
studies then suggest a return period of a tsunami with a magnitude
comparable to the Indian Ocean tsunami as being anywhere between 500 and
1000 years (Gelfenbaum et al. (2007); Wilkinson (2005); Stein and Okal
(2005); Bellier et al. (1997)). Based on these articles it was concluded that
occurrence of earthquakes with magnitudes (Mw) of 7.5, 8.0, 8.5, 9 and 9.5
could be associated with return periods of approximately 100, 150, 200, 500
and 1000 years respectively.
Subsequently, tsunami sources for these hypothetical earth-quakes with
different epicentres and focal depths were compiled using an adapted version
of the programme RNGCHN by Feigl and Dupre´ (1999), which is based on
the Okada model (Okada (1985)). For these sources uniform slip distribution
is assumed, which is the commonly accepted practise when no actual data is
available. These sources were defined by specifying epicentres of potential
earth-quakes along the Sunda Trench in such a manner that two succeeding
earthquake scenarios overlap each other by about half a rupture length. By
overlapping the earthquake sources it is ensured that a maximum flood impact
along the entire coast can be computed. In total approximately 400 scenarios
were simulated for earthquakes with magnitudes (Mw) of 6.5, 7.0, 7.5, 8.0,
8.5, 9.0 and 9.5. The maximum flooding computed for scenarios with equal
magnitude were then combined to produce flood hazard maps for the province
of Aceh and Nias that can be associated with the assumed recurrence intervals
of an earthquake magnitude.
The computed tsunami wave heights along the coast using this approach
compared very well with the results computed using an entirely different
method (Thio et al. (2005)). The computed
48 B.A.D. Van Veen et al. / Continental Shelf Research 79 (2014) 46–53

Fig. 2. Comparison of the simulated tsunami wave height for the Indian Ocean tsunami event with observed satellite data along their flight trajectories. Top: Topex/
Poseidon satellite and bottom: Envisat.

Table 1
Computed vs. observed arrival times at various coastal stations.

Station Lat. Lon. Arrival times (in minutes) Source

(decimal degrees) Observed Computed

Pasir Panjang 5.45 105.28 200 205 Bakosurtanal—Indonesia

Tuticorin 8.75 78.22 208 197 Nat. Inst. of Oceanography India
Vizakhapatnam 17.70 83.30 156 155 Nat. Inst. of Oceanography India
Colombo 6.98 79.85 180 (135–150)a 180 (160) UHSLC Indian Ocean Data
Male 4.18 73.52 190 210 UHSLC Indian Ocean Data
Diego Garcia 7.23 72.43 220 243 UHSLC Indian Ocean Data
Hanimadhoo 6.77 73.17 220 215 UHSLC Indian Ocean Data

a
The numbers between brackets represents the arrival time of the first smaller wave. This was mentioned by the eyewitness interviewed by Japanese surveyors at Galle face green, near Colombo
(Kyoto Univ. 2005). This wave arrived approximately1 h earlier.

tsunami wave heights for an 1–1000 year event, along the 15 m depth contour
line around the Aceh province shows a wave height development from 5 m in
south Aceh to 14 m in the East and Northern part of the province. Similar
figures were found by Thio et al. (2005). Having computed these 400
scenarios, the results were subsequently stored in a database forming the basis
of an EWS that is described in the following sections.
3. The database
A tsunami early warning system has to be reliable to have the desired
effect (i.e. reduction in number of victims). Uncertainties in the event that
generates the tsunami and actual local condi-tions (inaccuracies in the
bathymetry due outdated use of bathymetry or due to constant changing of the
morphology) should be taken into account in advance.

2.3. Extension of the models covering west Sumatra
In order to provide a tsunami EWS that also covers the coast of west and
south west Sumatra, the approach was extended with a second set of models.
The entire west coast of Sumatra is now covered in detail with another set of
ten inundation models using Delft3D. The new models covering Sumatra
have a resolution of 1000 m and above in the overall model and
approximately 200 m in the nested near-coast models. The scenario approach
for the earthquakes and tsunami sources as described earlier was applied here
as well. These models have not been validated in West Sumatra against an
actual event as no such event have occurred in this area yet. In total, including
the earthquake scenarios for Aceh and Nias, about 1250 scenarios were run.
All the results were then stored in the database.
The warning need be consistent to avoid confusion. So although, in
principle, highly localised separate warning informa-tion could be
disseminated, issuing the warnings at this detailed manner, i.e. at household
level, would probably result in more confusion. On the other hand, warnings
issued by current cen-tralised systems are usually very general and valid for
all coastal areas far away from the source.
These factors together with the extremely short arrival time of tsunamis in
Sumatra (15-30 min) lead to the idea of designing a database comprising
tsunamigenic earthquake scenarios and model results.
The current database provides detailed information on most likely affected
areas along the Sumatran coast within seconds after a tsunamigenic
earthquake is reported. The original simula-tion results contain over 1 Tb of
data (1 Gb per scenario). To allow
 B.A.D. Van Veen et al. / Continental Shelf Research 79 (2014) 46–53
Fig. 3. Computed 26-12-2004 tsunami inundation in Northern Aceh (dots represents observed data).
quick processing and accessing of information to speed up decision making,
only the most essential data are stored. Data reduction was achieved by
dividing the coast into 10 km long sections. One siren has an approximate
range of 3 km, a distance of 10 km was partly based on the assumption that a
village has at least 2 sirens. Each 10 km segment contains the results of all
grid cells along the segment (50–200 m grid resolution). For each tsunami
scenario only highest inundation and water levels (in each grid cell) along the
sections are stored and used. This has the advantage that adjacent areas
(within a section) are warned with the same level to avoid any confusion.
In the database therefore only the following information is stored.
– Maximum wave heights along the Sumatran coast are stored for warning
purposes and
– extent of inundated areas is important for emergency response
coordination.
Due to extremely short available response time (15–30 min) in Aceh and
West-Sumatra, this warning system (also known as RiskMap) is triggered by
earthquake parameters only. It then combines the results from two simulations
produced by two adjacent earthquakes that has a similar characteristics as the
detected earthquake. The highest water level values from these two
simulations are considered for the maps and for warning issuance.
A warning, issued at local level, is aimed to be effective for evacuation
and therefore it is kept straightforward and concise: ‘evacuate now’.
Information provided is limited to expected tsunami wave heights. Additional
information like inundation
49
extent, though useful for some other reasons and purpose, is not included in
the warning.
For preparedness and planning, maximum tsunami heights and inundation
related to recurrent intervals of earthquake magnitudes are needed. This
information is provided on flood hazard maps and communicated during
mock-up drills and is therefore not part of the early warning system as
described here.
3.1. Data retrieval
After a tsunamigenic earthquake with a certain magnitude and location is
reported (e.g. by BMKG), this system retrieves the maximum expected wave
heights from the database. The system is designed to be conservative as the
preliminary earthquake parameters are usually subject to adjustments. Maps
do not show spatial distribution of tsunami heights or run-ups but only the
stretch of coast that need to be warned.
The system retrieves the two nearest available scenarios by rounding up
the magnitude to the nearest scenario and rounding down the focal depth to
the nearest scenario available. For example, the scenarios that would have
been identified by the database for the 2004 event is shown below:
– 2004 case (initially reported): magnitude (Mw) 8.7; location
95.851E, 3.321N; focal depth 3075.6 km and
– a second scenario: magnitude (Mw) 9.0; location 95.451E, 3.471N; focal
depth 10 km.
Warning and inundation maps are then prepared for emer-gency response
coordination purposes using the maximum inun-dation simulated from those
two events.
50 B.A.D. Van Veen et al. / Continental Shelf Research 79 (2014) 46–53
The (configurable) trigger levels for warnings differentiate three zones for
each 10 km stretch of coastline:
– a warning zone, indicated as red, when the retrieved water level anywhere
along the coastline is higher than 0.50 m.
– a wet-feet zone, shown in orange, when the retrieved water level anywhere
along the coastline lies between 0.25 and 0.50 m.
– a safe zone, shown in green, when the retrieved water level on the coastline
is less than 0.25 m.
There have been some recent developments on tsunami detec-tion. In the
Indian Ocean, off the coast of Sumatra, buoys have been
Fig. 4. Scenario approach.
Fig. 5. Example of an output of the Riskmap EWS for a hypothetical tsunami.
deployed and across Sumatra GPS devices have been installed. These systems
are currently not linked to RiskMap.
3.2. Tsunami warning
RiskMap was installed at the regional incident command centre for Aceh
(Pusdalops) in Banda Aceh. An example of a warning map is shown in Fig. 4
for a hypothetical 8.0 Mw earthquake scenario (with focal depth of 10 km
near the island of Simeulue.
The local staff was trained to operate the system and the system has been
tested during provincial mock-up drills. RiskMap has been directly linked to
sirens at different locations in the province of Aceh. It can be operated in an
off-line as well as in an on-line mode (directly linked to Indonesian
meteorology site, BMKG, for obtaining the earthquake parameters in near
real time) so that it is triggered automatically. The system can be linked to
other sites as well (e.g. USGS) to double check the earthquake parameters
that has been reported by BMKG (Fig. 5).
3.3. Multi-hazard EWS and response coordination
Demographic information as well as maps and infrastructure information,
all GIS based, is required in the coordination of emergency response. These
data is included in the databases of the EWS. The possible affected
administrative units and number of people that may be affected can be shown
immediately on the screen. The system can also store names of the officials
and agencies that need to be contacted in case of an approaching tsunami
event.
Finally, the EWS has been extended and made a multi hazard Early
Warning System in 2009 by coupling it to a tide database and including
results from wave models to forecast flood events caused by extreme winds.
Further expansion is possible to include other hazards such as river floods,
landslides and volcano erup-tions. Due to these features this EWS is also
known as RiskMap.
4. Results and verification: Real events
Since the completion of the tsunami database for Sumatra several strong
earthquakes (47.0 Mw) have occurred along the
 B.A.D. Van Veen et al. / Continental Shelf Research 79 (2014) 46–53 51

Sunda Trench that has been used to verify the applicability and reliability of
the tsunami models (TMs) and RiskMap. In Table 2 an overview is presented
of recent strong earthquakes along the Sunda Trench off the coast of Sumatra.
We have chosen to use the USGS parameters here only because on the USGS
website all historic earthquake events are retrievable and the website also
provides observed tsunami height data.
communication and manual input). RiskMap was installed on the request of
BRR and the regional government at the regional incident command centre
for Aceh (Pusdalops) in Banda Aceh. Staff was trained to operate the system
and the system has been tested during provincial mock-up drills. Local
communities were trained on a separate parallel preparedness programme co-
executed with the Indonesian Red Cross (Palang Merah Indonesia).

The results show that the models and the past tsunami scenarios chosen
performed very well in predicting the tsunami heights. In the appendix the
results and screen dumps for each of the event have been included.
5. Discussion
The main challenge of this project was to develop a simple yet reliable
system to forecast the tsunami hazards. As was shown the TMs and the early
warning system RiskMap are very well capable of providing a quick and
reliable forecast with a high resolution. This is especially helpful for institutes
and agencies close to the area under potential tsunami risk.
The system described here was developed to run on simple basic
infrastructure; no high-end computing systems are required to use RiskMap. It
runs both online and offline and can receive data and provide information in
several ways (internet, sms, satellite
The system is capable of issuing warnings for a specific coastal area. A
length of 10 km has been selected as it is equivalent to the length of 1–2
coastal villages and it therefore minimises confu-sion by not issuing different
warnings to a single or neighbouring villages. At the time of developed of the
system it was envisaged that along the entire coast of Sumatra sirens would be
installed. There are several installed already in the coastal areas of Meula-
boh, Calang, and around Banda Aceh. RiskMap has been directly linked to
these sirens. The system can be linked to other dissemination systems if so
required. This local EWS ams to warn people in a bottom-up manner (villages
and then regional level).
At the time RiskMap was developed and became operational in the region,
the national system was not ready for external evalua-tion purposes. As a
result two systems are now operational in Aceh and Sumatra. We are aware
this could lead to local inconsistencies between the national and regional
EWS results. A comparison between the national and regional system is no
longer possible under this project. We are recommending to carry-out a
comparison

Table 2
Recent strong earthquake events along the Sunda Trench.

Date Location Magnitude (Mw) Tsunami heights

Location Measured (m) Predicted (m)

30 09 2009 Padang 7.6 Padang (0.9S–100.4E) 0.27 0.29

07 04 2010 Sinabang 7.5 Teluk Dalam (97.8E, 0.6N) 0.14 0.15–0.20
09 05 2010 Meulaboh 7.5 No tsunami observed – –
25 10 2010 Mentawai (Pagai) 7.7 Padang (0.9S–100.4E) 0.33 0.38
Tanahbaleh (0.5S–98.5E) 0.23 0.23

Pagai 3 3.06

Fig. A1. RiskMap screenshot of the 30 September 2009 Padang earthquake (USGS data).
52 B.A.D. Van Veen et al. / Continental Shelf Research 79 (2014) 46–53

of RiskMap with the national EWS, especially to avoid confusions on the B.2. RiskMap
warnings issued.
The initial magnitude was estimated at 7.2 SR at 05:20 local time by
BMKG and 7.5 Mw by USGS. Hypocentre was estimated at 30– 40 km
depth.
Appendix A. September 30 2009, Padang earthquake Both BMKG and USGS issued a tsunami alert. BMKG mentioned the
alert in the earthquake sms (05:20 WIB) and USGS issued a tsunami watch
A.1. Introduction bulletin (05:22 WIB) (Fig. B1).
Based on RiskMap (based on a 7.5 Mw scenario with a hypocentre at 10
In this appendix we present the results of RiskMap for the recent km depth), using the BMKG earthquake para-meters, there was no threat for a
earthquake of Wednesday afternoon September the 30th 1009 at 17:16 local tsunami. Most coastal segments near to the the epicentre (Nias, Singkil and
time near Padang (Sumatra Barat, Indonesia) by using the USGS earthquake Simeulue) had a maximum tsunami heights of 0.00–0.20 m. In some specific
parameters. The Riskmap results are compared with the observed and locations RiskMap predicted extreme values upto 0.40 m.
reported data.
At 06:01 WIB a tsunami was registered at Teluk Dalam (South Nias,
97.81E, 0.61N) with a height of 0.14 m. In this area RiskMap estimated a
maximum tsunami height of about 0.15–0.20 m (Fig. B2).
A.2. RiskMap results

An earthquake occurred at 0.91S 99.91E with a magnitude initially

estimated to be 7.7 Mw at 17:16 local time by USGS. USGS issued a tsunami
watch bulletin (17:26 WIB).
Based on RiskMap output (based on a 8.0 Mw scenario with a hypocentre
at 75 km depth), it was found that there was no threat for a large tsunami.
Most coastal segments close the epicentre (Padang, Siberut) had a maximum
value of 0.00–0.40 m. In some specific locations RiskMap predicted extreme
values upto approximately 0.70 m (Fig. A1).

Around the area of Padang, RiskMap estimated the maximum tsunami

height of about 0.15–0.30 m near Padang and 0.29 m at the observed site. The
reported height at Padang (17:46 local time, 100.41E, 0.91S) was 0.27 m.

Appendix B. April 7 2010, Sinabang earthquake

B.1. Introduction

In this appendix we present the results of RiskMap for the earthquake

occurred on Wednesday morning April 7 2010 at 05:15 am local time (6 April
2010 22:15 UTC) near Sinabang (Simeulue, Aceh, Indonesia) by using the
BMKG earthquake parameters/data. The results are compared with measured
data. Fig. B2. Sinabang earthquake.

Fig. B1. RiskMap screenshot of the 07 April 2010 Sinabang earthquake (BMKG data).
 B.A.D. Van Veen et al. / Continental Shelf Research 79 (2014) 46–53
Fig. C1. RiskMap screenshot of the 25 October 2010 Pulau Pagai earthquake (USGS data).
Appendix C. October 25, 2010 Pulau Pagai earthquake
C.1. Introduction
In this appendix we present the results of RiskMap for the earthquake near
Pulau Pagai (West Sumatra, Indonesia) that occurred on 25 October 2010 at
21:42 pm local time (25 October 2010 14:42 UTC) by using the USGS
earthquake parameters. The RiskMap results are compared with observed and
reported data.
C.2. RiskMap
An earthquake occurred at 3.11S 100.11E with a magnitude initially
estimated at 7.5 Mw at 21:42 local time by USGS. The hypocentre was
estimated at 33 km depth. USGS issued a tsunami watch bulletin (21:49 WIB)
(Fig. C1).
Based on RiskMap output (8.0 Mw scenario with a hypocentre at 10 km
depth), it was concluded that there was a threat for a local tsunami along the
coast of Pulau Pagai Utara and Pulau Pagai Selatan. At Pagai Islands
RiskMap predicted a maximum tsunami height of approximately 3 m.
At several locations tsunami waves were registered.
– 22:58 WIB at Padang (0.91S–100.41E) 0.33 m
– 23:05 WIB at Tanahbaleh (0.51S–98.51E) 0.23 m
– On Pagai people reported tsunami waves of approximately 3 m. There was
no gauge installed.
In the area RiskMap estimated maximum tsunami heights of
– 0.15–0.40 m near Padang and 0.38 m at the observed site.
– 0–0.25 m near Tanahbaleh and 0.23 m at the observed site.
– 0.20–3.20 m around Pagai Utara.
References
Apotsos, A., Buckley, M., Gelfenbaum, G., Jaffe, B., Vatvani, D., 2010. Nearshore tsunami
inundation model validation: towards applications. Pure and Applied Geophysics, 1–23,
March 2011.
53
Bellier, O., Se´brier, M., Pramumijoyo, S., Beaudouin, Th., Harjono, H., Bahar, I., Forni, O.,
1997. Paleoseismicity and seismic hazard along the Great Sumatran fault zone (Indonesia).
Journal of Geodynamics 24 (1), 169–183.
Berthier, E., Arnaud, Y., Vincent, C., Re´my, F., 2006. Biases of SRTM in high-mountain areas.
Implications for the monitoring of glacier volume changes. Geophysical Research Letters
10.1029.
Feigl, K.L., Dupre´, E., 1999. RNGCHN: a program to calculate displacement components from
dislocations in an elastic half-space with applications for modeling geodetic measurements
of crustal deformation. Computers and Geosciences 25, 695–704.
Gelfenbaum, G., Vatvani, D., Jaffe, B., Dekker, F., 2007. Tsunami Inundation and Sediment
Transport in Vicinity of Coastal Mangrove Forest. In: Proceedings of Coastal Sediments. 7
May 2007, New Orleans, LA, pp. 1117–1128.
Hesselink, A.W., Stelling, G.S., Kwadijk, J.C.J., Middelkoop, H., 2003. Inundation of a Dutch
river polder, sensitivity analysis of a physically based inundation model using historic data.
Water Resources Research 39 (9), 17.
Horrillo, J., Kowalik, Z., Shigihara, Y., 2006. Wave dispersion study in the Indian Ocean-
Tsunami of December 26, 2004. Marine Geodesy 29, 149–166.
Okada, Y., 1985. Surface deformation to shear and tensile faults in a half-space.
Bulletin of the Seismological Society of America 75 (4), 1135–1154.
Pujiraharjo, A., Hosoyamada, T., 2008. Numerical study of dispersion and non-linearity effects
on tsunami propagation. Coastal Engineering, 1275–1287.
SDC, 2007. Coastal Baseline Studies Aceh & Nias. Volume III, Tsunami Modelling and Risk
Assessment. /http://seadefenceconsultants.com/wp-content/ uploads/2009/12/SDC-R-
70041A_Vol-III-Tsunami-Modelling-EN.pdfS.
Shigihara, Y., Fujima, K., 2006. Wave dispersion effect in the Indian Ocean tsunami.
Journal of Disaster Research 1 (1), 142–143.
Stein, S., Okal, E., 2005. Long Period Seismic Moment of the 2004 Sumatra Earthquake and
Implications for the Slip Process and Tsunami Generation.
www.earth.northwestern.edu/people/seth/research/sumatra2.html.
Stelling, G.S., Duinmeijer, S.P.A., 2003. A staggered conservative scheme for every Froude
number in rapidly varied shallow water flows. International Journal for Numerical Methods
in Fluids 43 (12), 1329–1354.
Thio, H.K., Ichinose, G., Somerville, P., 2005. Probabilistic Tsunami Hazard Analysis.
In: Risk Frontiers Quarterly Newsletter, Vol. 5, Issue 1.
Vatvani, D., Schrama, E.J.O., van Kester, J., 2005a. Hindcast of Tsunami Flooding in Aceh–
Sumatra. In: Proceedings of the Fifth International Symposium on Ocean Wave
Measurement and Analysis—Waves, 2005.
Vatvani, D., Boon, J., Ramanamurty, P.V., 2005b. Flood Risk due to Tsunami and Tropical
Cyclones and the Effect of Tsunami Excitations on Tsunami Propaga-tions. In: Proceedings
of the IAEA Workshop on External Flooding Hazards at Nuclear Power Plant Sites,
Kalpakkam, Tamil Nadu, India.
Wilkinson, F., 2005. Coastal Design and Tsunami Mitigation for the United Nations High
Commissioner for Refugees. Shelter/House Reconstruction West Coast Province of
Nanggroe Aceh Darussalam. UNHCR/RedR, Australia.
Websites
/http://earthquake.usgs.govS.