Journal
Journalof
ofCoastal
CoastalResearch
Research
SI 64
pg -- pg
865
869
ICS2011
ICS2011 (Proceedings)
Poland
ISSN 0749-0208
Hazards Associated with Tsunami Waves in the Gulf of Oman
M.R. Akbarpour Jannat, M. Noranian Esfahani, V. Chegini and K. Rezanejad
Iranian National Institute for
Oceanography, Tehran
1411813389, Iran
[email protected]
ABSTRACT
Akbarpour Jannat, M.R., Noranian Esfahani, M., Chegini, V. and Rzanejad, K., 2011. Hazards Associated with
Tsunami Waves in the Gulf of Oman. Journal of Coastal Research, SI 64 (Proceedings of the 11th International
Coastal Symposium), – . Szczecin, Poland, ISSN 0749-0208
The Iranian Makran coast with a length of about 500 km lies near the Makran subduction zone and is a
susceptible area to the tsunamigenic waves. Although northwestern Indian Ocean has experienced some deadly
tsunamis in the past, this region remains one of the least studied regions in the world and little research work has
been devoted to its tsunami hazard assessment. The tsunami life is usually divided into three phases: The
generation, the propagation and finally the inundation of coastal shores. In this study, different source scenarios
along Makran subduction zone are considered and for each scenario, tsunami propagation is performed by
numerical modeling. In each case, the maximum positive offshore tsunami heights at various water depth along
the Makran coasts are calculated. The maximum positive tsunami height at the coast produced by this study will
provide a preliminary estimation of tsunami hazard and show which locations face the greatest threat from a
large tsunami. The results show the existence of more than 10 m wave height at some locations near the coasts
around Chabahar bay. Tsunamis are categorized as long waves; therefore, tsunami travel times can be computed
using water depth as the only variable. Here, tsunami travel time software is applied to calculate the arrival time
of the tsunami to the Iran coasts in the Oman Sea. According to the 1945 Makran earthquake source, it takes 15
minutes which the first tsunami wave reaches the Iran coasts in the Oman Sea.
ADITIONAL INDEX WORDS: Numerical modeling , Makran coast, Tsunami travel time
INTRODUCTION
Makran subduction zone (MSZ) and Andaman-NicobarSumatra Arc are two major important tsunamigenic zones in the
Indian Ocean which have noticeable underwater seismic activities
for creating tsunamis. The MSZ as triggered the second deadliest
tsunami in the Indian Ocean is an important source for large
tsunamis, as we know from the magnitude 8.1 earthquake and
tsunami that occurred there in 1945 with a death toll of about 4000
people (Pararas and Carayannis, 2006).
Following the massive Indian Ocean tsunami in December 2004
and threatened to more than 225,000 human live and leaving at
least a million homeless (Geist and Titov, 2006) was cause more
serious attention to this phenomenon in assessing the risks of
vulnerability on different coasts around the world, especially for
marginal Indian Ocean countries. Definitely, Indonesian’s
powerful tsunami in 2004 is not the first one in the region.
Although several tsunamis is occurring in the Indian Ocean
region, the most recent full-size predecessor to the 2004 tsunami
occurred about 550–700 years ago (Jankaew et al., 2008).
Tsunami hazard assessment in any particular region and its
irreparable losses requires compilation and analysis of the past
tsunami records. Better understanding of the tsunami on any
region will be effect on the development of urban activities and
also to estimate losses caused by criminal and financial of such
events (Heidarzadeh et al., 2008). Therefore, studies the history of
tsunami occurrence in coastal vulnerability around the world is
collected and cataloged.
There are many documents based on historical reports and
numerical modeling which show many tsunamis have been
occurring with different sources until 1945 into the MSZ
(Ambraseys and Melville, 2005). Heck (1947) prepared a list of
global tsunamis generated from 479 BC to 1946, that some of
them were related to tsunami occurrences in the Arabian Sea and
adjacent regions. Berninghausen (1966) gathered another list of
tsunami including 27 tsunamis reported from regions neighboring
the Indian Ocean which only three events were from the west
coast of India and the Arabian Sea. As an important tsunami
capacity, Page (1979) presented evidence for the recurrence of
large-magnitude earthquakes along the MSZ and predict the 1945type earthquake can occur every about 150–250 years in eastern
Makran. Ambraseys and Melville (1982) studied historical
earthquakes in Iran. Other list compiled by Murty and Rafiq
(1991) used a variety of sources in the Indian Ocean region from
326 BC to 1974. Byrne et al. (1992) studied great thrust
earthquakes along the plate boundary of the MSZ. Threshold
tsunami generation potential was studied by Pararas-Carayannis
(2006) along the MSZ.
There are many different ways for studying tsunami, such as using
paleotsunami data, historical tsunami recorded reports, online
DART systems, local tide gages and numerical modeling. In point
of view tsunami behavior after occurrence, tsunami simulation
using historical data and numerical modeling due to low cost, is
still common. Numerical modeling of tsunami is very important
for understanding past events and simulating future ones.
Determination of the potential run-ups and inundation by using
numerical modeling from a local or distant tsunami is recognized
as useful and important tool, since data from past events are
usually insufficient (Rafi and Mahmood, 2010).
Journal of Coastal Research, Special Issue 64, 2011
865
Numerical Modeling of Tsunami Waves
Figure 1. Tectonic feature and major earthquakes location of
northern continent part of Indian Ocean.
The Makran Subduction Zone (MSZ) of the coast of Iran and
Pakistan is a part of an accretionary wedge of late Cretaceous to
Holocene sediments (Arthurton et al., 1982). The MSZ is a
tsunamigenic narrow region due to reconstruction among three
tectonic Eurasian, Indian and Arabian plates (Figure 1) which
extends east from the Strait of Hormuz at the end of ZindanMinab fault in south of Iran to the Ornach-Nal fault near Karachi
in Pakistan with the length of about 900 km (Heidarzadeh et al.,
2008). According to other subduction rate of subduction zones,
MSZ is a relatively slow moving subduction zone. In point of
view of seismicity, MSZ demonstrates that large and shallow
earthquakes are common in this region which can be regarded as
evidence for tsunami generation potential with approximately
average return period 100–250 years of magnitude order of 8+
earthquakes (Heidarzadeh et al., 2008).
The purpose of this work is to investigate numerical modeling of
tsunami propagation and inundation in Chabahar coast which is
closed to MSZ based on four different tsunami scenarios, 326 BC,
1008, 1897 and 1945.
associated velocity field. The next phase of the problem solves the
equations of motion to propagate the initial condition from the
source region across the computational domain. Finally the runup
phase considers the interaction of the wave front with the
shoreline and the propagation over dry land.
For this study we used the numerical model MOST (Method Of
Splitting Tsunami) and ComMIT (Community Model Interface
for Tsunami) interface to simulate each of the three processes in
the tsunami modeling problem. The aim of ComMIT is to provide
an interface which allows for the selection of model input data
(initial condition, bathymetry grids etc) as well as a platform to
display model output through a graphical user interface (GUI). For
wave generation in the fault scenarios, the initial disturbance is
assumed to be of tectonic origin. MOST uses a version of
Okada’s (1985) model for surface deformations due to a fault
rupture below the surface of the earth. The resulting deformation
on the sea floor is translated directly to the water surface and
imposed as an initial condition to the wave propagation part of the
process. For wave propagation, MOST solves the 2+1 (two
horizontal spatial dimensions plus time) dimension non-linear
shallow water wave equations. At the shoreline, MOST employs a
moving shoreline algorithm to move the wave front across dry
land (Titov and Synolakis, 1998). The MOST code has been
extensively tested and validated against laboratory and field data
and has been shown to be accurate and reliable across a wide
range of spatial dimensions (Titov and Gonzales, 1997). The
MOST model is currently in use at NOAA’s Pacific Marine
Environmental Laboratory as the primary tool for creating tsunami
forecasting and hazard assessment tools and in the State of
California for producing inundation and evacuation maps.
Numerical Grids
The numerical grids used in this study are designed to capture all
potential inputs to the study locations (Figure 2, Table 2). The model
was set up with a system of three nested grids as shown in Figure 2.
The outermost grid was sampled to 1-arc minute, the intermediate
grid to 10-arcsec while the full resolution (1-arcsec) data was used
for the innermost grid. Model output is designated at the specific
locations (Table 1).
Tsunami Subduction Zone Sources
METHODS
This section describes the component of the model used in the
present study. We selected a suite of sources with the potential to
impact Makran coast based on historic events, Pacific basin
tectonics, and regional seismicity to initialize our numerical
simulations. Only far-field sources were considered in this study.
We used the numerical model MOST (Titov and Gonzales, 1997;
Titov and Synolakis, 1998) to simulate tsunami generation,
propagation and runup. MOST uses an elastic deformation model
to initialize the computation then solves the 2 dimensional shallow
water wave equations to propagate the disturbance across an ocean
basin. Four scenarios were modeled for a magnitude 8.1
earthquake in Makran subduction zone. The model results allow
an estimation of the maximum credible tsunami water heights at
the Chabahar sites and a sensitivity study of sources likely to
produce the greatest impacts.
Numerical Method and Model
Numerical modeling of tsunamis is composed of three parts;
generation, propagation and runup. The generation phase involves
modeling a geophysical source such as a landslide, earthquake or
volcanic eruption into an initial water surface displacement and
The first step of each tsunami modeling is generation of the waves
with concerning the initial disturbance of ocean floor and
resultantly ocean surface due to earthquake caused deformation at
ocean floor using seismic parameters. To investigate the tsunami
effects at study locations, several far-field sources were used. A
far-field source is one where the source region is located a great
distance from the region considered for the tsunami effects. If the
source is an earthquake, this means that the affected area lies
outside of the region where ground shaking is felt or where there
is coseismic surface deformation. A near-field tsunami on the
other hand is one where the tsunami source is close to the area
affected, where ground shaking or surface deformation for
earthquake sources would be experienced. Potential far-field
sources include large earthquakes on the various subduction zones
around the Pacific Rim.
Table 1: Model grids
MOST Stage
Grid dimensions
Grid A
1 arc minute (~1800m)
Grid B
10 arc second (~300m)
Grid C
1 arc second (~30m)
Journal of Coastal Research, Special Issue 64, 2011
866
Akbarpour Jannat et al.
Table 2: The list of tsunami in the MSZ (Heidarzadeh et al.,
2008)
No.
1
2
3
4
Year
Longitude (°E)
Latitude (°N)
326 BC
1008
1897
1945
67.30
60.00
62.30
63.0
24.00
25.00
25.00
24.50
In order to model the far-field events, we used NOAA’s FACTS
(Facility for the Analysis and Computation of Tsunami
Simulations) database. This database contains the full transoceanic simulations for tsunamis generated on segments of the
major subduction zones along the Pacific Rim. The database was
created by subdividing each subduction zone into 2 parallel rows
of 100 km long by 50 km wide fault segments (Figure 2). A pure
thrust earthquake mechanism with unit slip (1 m) is then imposed
on each segment and the resultant trans-oceanic wave propagation
is computed and stored. Larger earthquakes and tsunamis are then
created from this database by combining segments and
multiplying the output by an appropriate scalar to reach the
desired earthquake magnitude.
Sources can be constructed by adding 100 km long and 50 km
wide segments. A magnitude 8.9 earthquake with a fault length of
800 km and a fault width of 100 km (from the two rows of 50 km
wide segments) requires an average slip of 9.3 m (30.5 ft) on each
fault segment. The computed tsunami from eight adjacent 100 km
segments is multiplied by 9.3 and the results linearly combined
into one resultant wave field. Times series of wave height and
velocity are interpolated at the boundary of the outermost local
grid (see Figure 2) and used as the initial condition to the local
tsunami inundation model (Borrero et al., 2004). A summary of
all the fault scenarios used in this study is given in Table 3 and
Figure 3.
Figure 2. Nested grids of A, B and C and available tsunami source
cells along MSZ.
selected time steps and provides the initial and boundary
conditions for Inundation Phase.
RESULTS AND DISCUSSION
The oldest tsunami which has been occurred in 326 BC is modeled by
choosing mkszb2 as source cell of earthquake (Figure 2). According
to model results, the first peak of the tsunami with a 9 cm height is
sensed after 1.45 hour near Chabahar port. The maximum wave
height (19 cm) due to this type of tsunami receives to Chabahar
port after 2 hours (Figure 4-a).
Based on 1897 tsunami position, source of occurrence tsunami is
located in mkszb2 as shown in Figure 2. The first peak was sense
after 45 minutes by height of 0.66 meter near Chabahar port
(Figure 4-b). The maximum available magnitude of tsunami
height is nearly 1 meter that occurred after 1.0 hour.
Set up Tsunami Initial Conditions
The initial conditions for the run are specified from the previously
computed tsunami propagation runs. ComMIT provides access to
pre-computed propagation model databases from NOAA. The
interactive geographic information window is used to choose the
source fault planes that correspond to associated propagation run.
If several faults are selected, a linear combination of those faults is
assumed. The magnitude of the earthquake can be adjusted
manually in the source specification frame. If adjusted, the whole
propagation solution is scaled accordingly. Either magnitude or
slip can be specified. For a combination source, the slip or
percentage of magnitude can be specified for individual subfaults.
Regard s to Figure 2, the earthquake source of 1008 AD tsunami
is located in mksza8. The first peak was sense after 33 minutes by
0.3 meter height near Chabahar port, but the maximum tsunami
wave available is 1.36 meter that occurred after 1.2 hour. The
received wave’s shows drying coast around 1.5 hour after tsunami
occurred (Figure 4-c).
In order to 1945 earthquake modeling that its source is located
in mkszb4, the first peak of the tsunami wave occurred after 65
minutes with a height of less than 60 cm. Although the source
position of earthquake is far from Chabahar Bay but the height of
wave recorded near Chabahar port is noticeable than the
maximum wave height obtained from other source (Figure 5).
The height at the coast produced by this study will provide a
preliminary estimation of tsunami hazard and show which
locations face the greatest threat from a large tsunami. The results
show the existence of more than 10 m wave heigh at some
locations near the coasts around Chabahar Bay. Although
northwestern Indian Ocean has experienced some deadly tsunamis
Propagation Phase Modeling
The Propagation Phase models the open-ocean evolution of a
tsunami using a depth-integrated version of Nonlinear Shallow
Water (NSW) wave equations in two spatial and one temporal
dimension. The output of a Propagation Phase calculation, the
wave's height, and zonal and meridional velocities is saved for
Table 3: Details of the initial conditions used in the scenario database
Scenario
1~3
2
Magnitude
(Mw)
8.1
8.1
Segment
Length
(km)
1
3
100
300
Width
(km)
50
50
Depth
(m)
Dip
(deg.)
Strike
(deg.)
Slip
(deg.)
25
25
7
7
265
280
90
89
Journal of Coastal Research, Special Issue 64, 2011
867
Numerical Modeling of Tsunami Waves
40
30
Makran
1945
Tsunami Wave Height (cm)
20
10
0
−10
−20
−30
−40
−50
−60
0
Figure 3. Schematic of plate boundary and source locations.
Thick line represents rupture segment, W is the width, D is the
depth and δ is the dip angle.
in the past, this region remains one of the least studied regions in
the world and little research work has been devoted to its tsunami
hazard assessment. In the studies done so far, different issues
regarding tsunami in this region have been investigated. Another
scenario for tsunami propagation along the Makran coast with a
combination of three adjacent sources was studied. The effects of
a wide-integrated cell of seismic earthquake including mkszb6,
mkszb7 and mkszb8, show a new situation of incoming waves
toward Chabahar Bay (Figure 6). While this type of earthquake is
generated along MSZ, the incoming tsunami wave will be more
intensively than the all of other single scenario as mentioned
above.
Tsunami Travel Time Calculation
Tsunamis are categorized as long waves, therefore, tsunami travel
times can be computed using water depth as the only variable.
Long waves are those in which the distance between crests of the
wave is much greater than the water depth. Wave speed is
computed from the square root of the quantity water depth times
the acceleration of gravity. Thus, tsunami travel times can be
computed without any knowledge of the tsunami's height,
wavelength, etc. Here, TTT software is applied to calculate the
arrival time of the tsunami to the Iran coasts in the Oman Sea.
20
(a)
Tsunami Wave Height (cm)
0
−20
0
100
0.5
1
1.5
2
2.5
(b)
200
0.5
1
1.5
2
Time (hour)
1.5
2
2.5
Figure 5. Time series of incoming tsunami waves in Chabahar
port associated with 1945 earthquake
According to the 1945 Makran earthquake and using a non-point
earthquake source (24.5N, 63.0E, length 200km, width 100km,
strike 270), it takes 15 minutes which the first tsunami wave
reaches the Iran coasts in the Oman Sea (Figure 7). It implies the
importance of the tsunami warning systems establishment in the
Indian Ocean to prevent the people life along the Iran coasts in the
Oman Sea.
CONCLUSIONS
In this study, different source scenarios along Makran subduction
zone are considered and for each scenario, numerical modeling of
tsunami propagation is performed by ComMIT numerical software
which developed by IOC/UNESCO. In each case, the maximum
positive offshore tsunami heights at various water depth along the
Makran coasts are calculated. The Makran coast of Iran is exposed
to the long waves generated in the Indian Ocean. The coast
experienced long waves induced by tsunami during the past
decades. Using the impact of the past events we present potential
inundation areas for the great waves on the Makran coast. Based
on the value of long wave height, three flood zones are specified
on the inundation map. The first zone associated with no
inundation on uplifted rocky shore, the second zone would
experience up to 2 km inundation on the coast of Chabahar Bay
and Jask areas. Areas of sandy beaches that backed with high
cliffs fall between two extreme regions.
Results show that the Makran subduction zone can be the site of
large earthquakes, as evidenced by the 1945 event. The tsunami
reached Oman, and may have penetrated into the Persian Gulf.
ACKNOLEDGEMENT
2.5
(c)
LITERATURE CITED
100
0
−100
0
1
Founding for this research was provided by the Iranian National
Institute for Oceanography (INIO) and IOC-UNESCO. Thanks are
due to Charitha Pattiaratchi for his thoughtful comments.
0
−100
0
0.5
0.5
1
1.5
2
2.5
Ambraseys, N.N. and Melville, C.P., 1982. A History of Persian
Earthquakes, Cambridge University Press.
Ambraseys, N.N. and Melville, C.P., 2005. A History of Persian
Earthquakes, Cambridge Univrsity Press.
Figure 4. Time series of incoming tsunami waves in Chabahar
port from three historical earthquakes a)326 BC, b)1008 and
c)1987
Journal of Coastal Research, Special Issue 64, 2011
868
Akbarpour Jannat et al.
Figure 6. Incoming waves toward Chabahar bay
Arthurton, R.S., Farah, A., et al., 1982. The Late CretaceousCenozoic history of western Baluchistan Pakistan-the northern
margin of the Makran subduction complex, Geological Society
,10 , 373-385.
Berninghausen, W.H., 1966. Tsunamis and seismic seiches
reported from regions adjacent to the Indian Ocean, Bulletin of
the Seismological Society of America, 56(1), 69-74.
Byrne, D., Sykes, L.R., et al., 1992. Great Thrust Earthquakes of
the and Aseismic Slip Along the Plate Boundary of the
Makran Subduction Zone, Journal of Geophysical research
97(B1): 449–478.
Geist, E.L., Titov, V.V., et al., 2006. Tsunami: Wave of Change,
Scientific American(January), 56-63.
Heck, N.H., 1947. List of seismic sea waves, Bulletin of
Seismological Society of America, 37(4), 269-286.
Heidarzadeh, M., Pirooz, M.D., et al., 2008. Historical tsunami in
the Makran Subduction Zone off the southern coasts of Iran
and Pakistan and results of numerical modeling, Ocean
Engineering, 35, 774–786.
Jankaew, K., Atwater, B.F., et al., 2008. Medieval forewarning of
the 2004 Indian Ocean tsunami in Thailand, Nature 455,
1228-1231.
Murty, T. and Rafiq, M., 1991. A tentative list of tsunamis in the
marginal seas of the north Indian Ocean, Natural Hazards 4,
81-83.
Okada, Y., 1985. Surface deformation due to shear and tensile
faults in a half-space, Bulletin Seismological Society of
America, Vol. 75, 1135-1154.
Page, W., 1979. Evidence for the recurrence of large-magnitude
earthquakes along the Makran Coast of Iran and Pakistan,
Tectonophysics, 52(1-4), 533-547.
Pararas-Carayannis, G., 2006. The Potential of Tsunami
Generation along the Makran Subduction Zone in the
Northern Arabian Sea. Case Study: The Earthquake and
Tsunami of November 28, 1945, Science of Tsunami Hazards
24(5), 358-384.
Rafi, Z. and Mahmood, N., 2010. Numerical Modeling of
Tsunami Inundation for Potential Earthquake at Makran
Subduction Zone - A case study for Gwadar Coastal Area.
Islamabad, Pakistan Meteorological Department. PMD-45.
Titov, V.V. and Gonzales, F., 1997. Implementation and testing of
the method of splitting tsunami (MOST) model. Technical
Memorandum ERL PMEL 112, NOAA, 11 pp.
Figure 7. Contour intervals of tsunami travel time, 15min= o.25
Titov, V.V. and Synolakis, C.E., 1998. Numerical modeling of
tidal wave runup. Journal of Waterway, Port, Coastal, And
Ocean Engineering, Vol. 124(4), 157–171.
Journal of Coastal Research, Special Issue 64, 2011
869