Numerical experiment and a case study of sediment transport simulation of the 2004 Indian Ocean tsunami in Lhok Nga, Banda Aceh, Indonesia
© The Society of Geomagnetism and Earth, Planetary and Space Sciences (SGEPSS); The Seismological Society of Japan; The Volcanological Society of Japan; The Geodetic Society of Japan; The Japanese Society for Planetary Sciences; TERRAPUB. 2012
Received: 31 October 2010
Accepted: 21 October 2011
Published: 24 October 2012
We use a two-dimensional tsunami sediment transport model to study the source of the 2004 earthquake. To test the model behavior, numerical experiment on sediment deposition and erosion is performed using various hypothetical parameters of tsunami wavelength, topographic slope, and sediment supply. The numerical experiment results show that erosion and deposition are strongly influenced by the tsunami wavelength and the topographic slope. The model is used to compute the spatial distribution of tsunami deposit thickness produced by the 2004 Indian Ocean over an actual elevation datasets in the coastal area of Lhok Nga, Banda Aceh, Indonesia. The model produced simulated tsunami deposits that have similar thicknesses with the measured data along a surveyed transect. Then we estimate a simple fault model for the southern portion of the 2004 earthquake using tsunami sediment transport simulations. The simulated tsunami run-up from the fault model is very close to the measured run-up. This result indicates that a source process of a large earthquake that generates a large tsunami has a potential to be estimated using sediment deposit distribution data.
Large tsunamis can deposit sand layers up to several tens centimeters and distributed sand layers several kilometers inland. For example, in case of the 2004 Indian Ocean tsunami, the maximum tsunami height at Lhok Nga, west of Banda Aceh, Indonesia, was 35 m, and the inundation distance reached 6 km inland (Tsuji et al., 2005; Paris et al., 2007). The tsunami deposited sand up to 80 cm thick and left mud up to 5 km inland (Moore et al., 2006).
Recent tsunami sediment deposits have been studied by Nishimura and Miyaji (1995) in Hokkaido, Dawson (1994) in Java, Gelfenbaum and Jaffe (2003) in Papua New Guinea, Moore et al. (2006) and Paris et al. (2007) in Aceh, Sumatra, and MacInnes et al. (2009) in Kuril Islands. The interactions between the tsunami, the topography, and the sediment source affect the spatial distribution and characteristics of tsunami deposits (Dawson, 1994; Gelfenbaum and Jaffe, 2003). Jaffe and Gelfenbuam (2007) developed a simple model for tsunami sedimentation that can be applied to calculate tsunami flow speed from the thickness and grain size of a tsunami deposit. Two-dimensional simulation models that simulate topographic change induced by tsunami previously developed by Takahashi et al. (1999, 2000), Fujii et al. (2008), and Nishihata et al. (2005). A three-dimensional hydrostatic shallow water model, the C-HYDRO3D, was developed to study tsunami sediment transport at coastal areas near a harbor (Kihara and Matsuyama, 2010). Another three-dimensional numerical model, the Deflt3D, was used to simulate the inundation and sediment transport of tsunami over measured and idealized coastal morphologies at Fagafue Bay, American Samoa (Apotsos et al., 2011a) and Kuala Meurisi, Sumatra, Indonesia (Apotsos et al., 2011b).
On the other hand, when those tsunami deposits were buried and preserved, they became the geological records of past tsunamis (Paris et al., 2007). Paleotsunami deposits have been used to estimate the recurrence interval of great earthquakes (Atwater, 1987; Minoura et al., 2001; Jankaew et al., 2008; Monecke et al., 2008). Paleo-tsunami deposits have been used to study the type and size of a pre-historical earthquake (Minoura et al., 2001; Namegaya et al., 2010). Namegaya et al. (2010) compare the simulated inundation area with the distribution area of the 869 Jogan tsunami deposits to determine that the earthquake was an interplate earthquake with magnitude of Mw 8.4. Those previous studies assumed that the tsunami inundation area is the same as the tsunami deposits distribution area. However, the tsunami inundation area can be much larger than the deposited area (Goto et al., 2011).
In this paper, we try to show that the tsunami deposits data are useful to find the source processes of pre-historical large earthquake. Source processes of pre-historical large earthquakes are one of key data need to be found for earthquake prediction researches. Numerical simulation of tsunami sediment transport from the co-seismic deformation using the earthquake fault model is needed for explaining the tsunami deposit data. First, to understand the interaction between tsunami hydraulics and sediment transport with a given input parameters, we perform one-dimensional sediment transport numerical experiment. Then, we test our technique to explain the actual tsunami deposit data at Lhok Nga, Banda Aceh, Indonesia, due to the 2004 Sumatra-Andaman earthquake using a two-dimensional sediment transport simulation. Finally, we try to find the possibility to estimate the slip amount on an assumed simple fault model from sediment transport simulations.
2. Tsunami Sediment Transport Model
Among the sediment transport models mentioned in the previous section, we choose to build and use a two-dimensional model based on the method by Takahashi et al. (2000) because it is rather simple compare to the three-dimensional models. Another reason is that a two-dimensional model requires less computer time per run compare to a three-dimensional model. The magnitude of the 2004 tsunami required us to simulate sediment transport over a large area, so a two-dimensional model is preferred. Takahashi et al. (2000) assumed two distinct layers of suspended load layer and bed load layer in their model and used the concept of exchange rate that connect the two layers. The exchange rate is a balance between sediment settling down from suspended load layer and the sediment rising up to the suspension. This method was applied to study the sediment transport of the 1960 Chile tsunami in Kesennuma Bay, Japan (Takahashi et al., 2000).
In the above equations, ZB is bed level, λ is porosity, QBx and QBy are bed load rate in −x and −y directions, Wex is exchange load, and C̄ is the mean concentration of sediment in the suspended load layer. The mean concentration of the suspended load is limited up to 1%.
The settling velocity of a particle is related to the particle shape, particle size, specific gravity, and the kinematic viscosity of the sediment. The settling velocity of a sphere in a fluid at rest can be estimated by solving the balance between the gravitational force and the drag resistance.
A simple formula to estimate the settling velocity of natural sediment particles has been obtained from the previous work of Dietrich (1982), which can be used for a given sediment diameter, shape factor, and roundness. In case of no information on shape and roundness factor, the shape factor of 0.7 and roundness value of 3.5 can be used for natural sediment particles (Jiménez and Madsen, 2003). The submerged specific gravity is assumed to be 1.65. The kinematic viscosity corresponds to fresh water at specified temperature.
3. Numerical Experiment
3.2.1 Effect of tsunami wavelength
3.2.2 Effect of topographic slope
The simulated sand deposits thickness increase near the steeper slope topography (hill) that stops the tsunami for inundating farther inland. This is because more sand particles fall on the bed due to accumulation and saturation of suspended concentration in the water columns near the hill. Comparison of bed changes using the three topography shows that the sand layer is distributed more smoothly when using milder topographic slope. The comparisons using tsunami waveforms with different wavelengths also show the same result.
3.2.3 Effect of sediment supply
4. Case Study of the 2004 Indian Ocean Tsunami
Tsunami sand deposits were collected by Moore et al. (2006) along a 400 m transect in Lhok Nga, Banda Aceh, Indonesia. The measured transect originates at the shoreline at 5°26′32.5″N and 95°14′22.8″E. Along the transect they measured topographic profile and thickness of sediment deposited by the 2004 tsunami. The sand layer is distributed along the profile from 100 to 400 m from shoreline with maximum sediment thickness of 20 cm. The mean grain diameters of sediment samples range from 0.3 to 0.9 mm. In this study, these data are used for comparison to verify the simulation result. The tsunami run-up height of about 14 m at 400 m from shoreline is estimated from the limit of sediment deposit.
Elevation datasets that include both bathymetry and topography are required for tsunami inundation and sediment transport simulations. For bathymetry data, the General Bathymetric Chart of the Oceans (GEBCO) dataset with 30 arc-sec resolution and nautical charts around west coast of Banda Aceh are used. To reproduce the topography dataset around the surveyed transect, the BAKOSURTANAL topographic contour map and the measured ground elevations by Moore et al. (2006) are combined. Then the elevation datasets for tsunami inundation and sediment transport simulation is obtained by combining the bathymetry and the topography datasets.
The sediment transport simulation of the 2004 tsunami is done using different sediment grain diameters of 0.3 mm, 0.5 mm, and 0.8 mm that are within the range of mean grain diameters in the study area measured by Moore et al. (2006). Each of sediment grain diameters is simulated separately. Computation time interval (Δt) of 0.5 sec is selected for the sediment transport simulation in the finest grid system. Detailed topography data before the tsunami is required in order to properly compute the bed change due to the tsunami. Unfortunately, such data in this study area does not exist. Because detail topography data and the supply of sediment along the surveyed transect before the tsunami are unknown, therefore we assume the basal limit as the initial topography that is not movable by the tsunami. Movable bed layer is set to be from 30 m inland to offshore until the limit of modeling domain.
To verify the sediment transport model, we run numerical simulation of sediment transported by the 2004 Indian Ocean tsunami in Lhok Nga, Banda Aceh, Indonesia. Slip distribution for the 2004 Sumatra-Andaman earthquake was estimated using tsunami waveforms data and sea surface heights captured by satellite by previous studies (Fine et al., 2005; Hirata et al., 2006; Tanioka et al., 2006; Tanioka and Iwasaki, 2006; Fujii and Satake, 2007). In this study we use the source model estimated by Tanioka and Iwasaki (2006) because it can well explain the tsunami waveform data and satellite altimetry data.
The bottom deformation of the 2004 earthquake is calculated by Okada (1985) formula using slip distribution from Tanioka and Iwasaki (2006). The rupture area of the earthquake extends all the way to the trench axis, so the generated bottom deformation has sharp uplift near the trench axis. In this case, it is not appropriate to assume that the sea-surface deformation is the same as the sea-bottom deformation. Instead of using the assumption, the sea surface deformation is calculated by Kajiura (1963) formula using the bottom deformation.
4.3.1 The 2004 tsunami inundation and sediment deposits simulations
In this study we demonstrate a way to estimate slip amount on an assumed fault model of the 2004 earthquake using tsunami deposit data. A simple fault model of the 2004 earthquake is assumed to be a shallow dipping thrust fault (strike = 340°, dip = 10°, rake = 90°) with the shallowest part of the fault in contact with the ocean bottom. The fault length and width are 300 km and 150 km, respectively (Fig. 7). The fault is located at the southern portion of the 2004 earthquake source area. The location and size of the fault is based on previous studies on source model of the 2004 earthquake (e.g. Lay et al., 2005; Hirata et al., 2006; Tanioka and Iwasaki, 2006; Chlieh et al., 2007; Fujii and Satake, 2007). The sediment transport simulations in the study area using different fault lengths produce similar results, using sediment deposits at only one location is not enough to estimate a fault length. There is trade-off between fault width and slip amount, but such analysis is beyond the scope of this study.
To estimate the slip amount on the simple fault model by sediment transport simulations, we run simulations of sediment transport using five slip amounts of 5 m, 10 m, 15 m, 20 m, and 30 m. Then each of the simulated sediment deposits is compared with the measured sediment deposits thicknesses along the surveyed transect by Moore et al. (2006). A range of slip amount on the single fault model of the 2004 earthquake can be estimated by comparing the measured sediment deposits thickness with the simulated ones. The sediment transport simulations are done using sediment grain diameter of 0.8 mm as suggested in the previous section.
The above result suggests that comparisons between the measured and the simulated sand layer thickness distribution can be used to estimate the slip amount on the single fault of the 2004 earthquake, although some disagreements are still exist between the simulated and the measured sand deposit distribution. The disagreements can be caused by at least three factors, which are unknown detailed topography data before the tsunami, unknown sediment supply geometry before the tsunami, and the limitation in the sediment transport model. The numerical experiment shows that topography data influence strongly on the location of erosion and deposition. It also shown that sand supply near-shore and onshore strongly influences the volume of deposition inland. These suggest that, to simulate sediment transport properly, it is important to have detailed topography data and geomorphology data of coastal areas before and after tsunami.
The effect of tsunami wavelength, topographic slope, and sediment supply are examined using the numerical simulation. Tsunami with longer wavelength distributes sediment layer more smoothly along the coastal plain, while tsunami with shorter wavelength causes more erosion at the beach and deposit more sediment near a hill. Comparison of bed changes at the three different hypothetical topographies with different slopes shows that the volume of erosion around the beach is higher when using topography with sharper slope break. The erosion usually occurs near the slope break where the spatial flow acceleration is large. The depositions on land seem to be influenced significantly by supply of sediment near-shore and onshore, but not significantly by supply of sediment offshore.
The observed tsunami inundation in Lhok Nga, Banda Aceh, due to the 2004 great Sumatra earthquake is explained well by the simulated tsunami inundation from the slip distribution estimated by Tanioka and Iwasaki (2006). The simulated sediment layer distributed along the profile from 80 to 400 m inland with maximum sediment deposits thickness of 35 cm, which is close to the measured sediment deposits thickness. The simulation result shows that the deposited sediment deposits were mostly transported by the tsunami in suspension.
Slip amount on the assumed fault model of the 2004 earthquake is estimated to be between 10 m and 15 m from the sand deposits along the transect in Lhok Nga. The fault models with slip amount of 10 m and 15 m can generate tsunami run-ups of 12 m and 15 m, respectively, which are close to the measured run-up of 14 m. This indicates that tsunami run-up heights can be estimated from sand deposit distribution data. Furthermore, if tsunami sand deposit distributions for many transects are available, the source process of a large earthquake that generates a large tsunami can be estimated.
We thank Kazuhisa Goto and two anonymous reviewers for their helpful comments and suggestions.
- Apotsos, A., G. Gelfenbaum, B. Jaffe, S. Watt, B. Peck, M. Buckley, and A. Stevens, Tsunami inundation and sediment transport in a sediment-limited embayment on American Samoa, Earth-Sci. Rev., 107(1–2), 1–11, 2011a.View ArticleGoogle Scholar
- Apotsos, A., G. Gelfenbaum, and B. Jaffe, Process-based modeling of tsunami inundation and sediment transport, J. Geophys. Res., 116(F1), 1–20, 2011b.Google Scholar
- Atwater, B. F., Evidence for Great Holocene earthquakes along the outer coast of Washington State, Science, 236, 942–944, 1987.View ArticleGoogle Scholar
- Bird, E. C. F., Coastal Geomorphology, 411 pp., John Wiley & Sons, Ltd, England, 2008.Google Scholar
- Chlieh, M., J.-P. Avouac, V. Hjorleifsdottir, T.-R. A. Song, C. Ji, K. Sieh, A. Sladen, H. Hebert, L. Prawirodirdjo, Y. Bock, and J. Galetzka, Co-seismic slip and afterslip of the great Mw 9.15 Sumatra-Andaman Earthquake of 2004, Bull. Seismol. Soc. Am., 97, S152–173, 2007.View ArticleGoogle Scholar
- Dawson, A. G., Geomorphological effects of tsunami run-up and backwash, Geomorphology, 10, 83–94, 1994.View ArticleGoogle Scholar
- Dietrich, W., Settling velocity of natural particles, Water Resour. Res., 18(6), 1615–1626, 1982.View ArticleGoogle Scholar
- Fine, I. V., A. B. Rabinovich, and R. E. Thomson, The dual source region for the 2004 Sumatra tsunami, Geophys. Res. Lett., 32, L16602, 2005.View ArticleGoogle Scholar
- Fujii, N., M. Ikeno, T. Sakakiyama, M. Matsuyama, M. Takao, and T. Mukohara, Hydraulic experiment on flow and topography change in harbor due to tsunami and its numerical simulation, Ann. J. Coast. Eng., JSCE, 56, 291–295, 2008 (in Japanese).Google Scholar
- Fujii, Y. and K. Satake, Tsunami source of the 2004 Sumatra-Andaman earthquake inferred from tide gauge and satellite data, Bull. Seismol. Soc. Am., 97, S192–S207, 2007.View ArticleGoogle Scholar
- Gelfenbaum, G. and B. Jaffe, Erosion and sedimentation from the 17 July 1998 Papua New Guinea tsunami, Pure Appl. Geophys., 160, 1969–1999, 2003.View ArticleGoogle Scholar
- Goto, C, Y. Ogawa, N. Shuto, and F Imamura, Numerical method of tsunami simulation with the leap-frog scheme, IUGG/IOC TIME Project, IOC Manuals and Guides, UNESCO, 35, 1997.Google Scholar
- Goto, K., C. C. Goff, S. Fujino, J. Goff, B. Jaffe, Y. Nishimura, B. Richmond, D. Suguwara, W. Szczucinski, D. R. Tappin, R. Witter, and E. Yulianto, New insights of tsunami hazard from the 2011 Tohoku-oki event, Mar. Geol., 2011 (submitted).Google Scholar
- Hirata, K., K. Satake, Y Tanioka, T. Kuragano, Y Hasegawa, Y Hayashi, and N. Hamada, The 2004 Indian Ocean tsunami: Tsunami source model from satellite altimetry, Earth Planets Space, 58, 195–201, 2006.View ArticleGoogle Scholar
- Imamura, F, Review of tsunami simulation with a finite difference method, in Long-Wave Runup Models, edited by H. Yeh, P. Liu, and C. Synolakis, 403 pp., World Scientific, Singapore, 1996.Google Scholar
- Jaffe, B. E. and G. Gelfenbaum, A simple model for calculating tsunami flow speed from tsunami deposits, Sediment. Geol., 200,347–361,2007.View ArticleGoogle Scholar
- Jankaew, K., B. F Atwater, Y Sawai, M. Choowong, T. Charoentitirat, M. E. Martin, and A. Prendergast, Medieval forewarning of the 2004 Indian Ocean tsunami in Thailand, Nature, 455, 1228–1231, 2008.View ArticleGoogle Scholar
- Jiménez, J. A. and O. S. Madsen, A simple formula to estimate settling velocity of natural sediments, J. Waterw. Port Coast. Ocean Eng., 129(2), 70–78, 2003.View ArticleGoogle Scholar
- Kajiura, K., The leading wave of a tsunami, Bull. Earthq. Res. Inst., 41, 535–571,1963.Google Scholar
- Kihara, N. and M. Matsuyama, Three-dimensional hydrostatic numerical simulations on tsunami-induced sediment transport in a model harbor, Proceedings of the 3rd International Tsunami Field Symposium, 2010.Google Scholar
- Lay, T., H. Kanamori, C. J. Ammon, M. Nettles, S. N. Ward, R. C. Aster, S. L. Beck, S. L. Bilek, M. R. Brudzinski, R. Butler, H. R. DeShon, G. Ekstrom, K. Satake, and S. Sipkin, The great Sumatra-Andaman earthquake of 26 December 2004, Science, 308(5725), 1127–1133, 2005.View ArticleGoogle Scholar
- MacInnes, B. T, J. Bourgeois, T. K. Pinegina, and E. A. Kravchunovskaya, Tsunami geomorphology: Erosion and deposition from the 15 November 2006 Kuril Island tsunami, Geology, 37, 995–998, 2009.View ArticleGoogle Scholar
- Minoura, K., F Imamura, D. Sugawara, Y Kono, and T. Iwashita, The 869 Jogan tsunami deposit and recurrence interval of large-scale tsunami on the Pacific coast of northeast Japan, J. Nat. Disaster Sci., 23(2), 83–88, 2001.Google Scholar
- Monecke, K., W. Finger, D. Klarer, W Kongko, B. G. McAdoo, A. L. Moore, and S. U. Sudrajat, A 1,000-year sediment record of tsunami recurrence in northern Sumatra, Nature, 455, 1232–1234, 2008.View ArticleGoogle Scholar
- Moore, A., Y. Nishimura, G. Gelfenbaum, T. Kamataki, and R. Triyono, Sedimentary deposits of the 26 December 2004 tsunami on the northwest coast of Aceh, Indonesia, Earth Planets Space, 58, 253–258, 2006.View ArticleGoogle Scholar
- Namegaya, Y., K. Satake, and S. Yamaki, Numerical simulation of the AD 896 Jogan tsunami in Ishinomaki and Sendai plains and Ukedo rivermouth lowland, Ann. Rep. Active Fault Paleoearthq. Res., 10, 1–21, 2010 (in Japanese).Google Scholar
- Nishihata, T., Y. Tajima, and Y. Moriya, Study on the dynamic topography change due to tsunami attacks at kirinda port—the December 26, 2004 earthquake tsunami of Indian Ocean, Ann. J. Coast. Eng., JSCE, 52, 1386–1390, 2005 (in Japanese).View ArticleGoogle Scholar
- Nishimura, Y. and N. Miyaji, Tsunami deposits from the 1993 Southwest Hokkaido earthquake and the 1640 Hokkaido Komagatake eruption, Northern Japan, Pure Appl. Geophys., 144, 719–733, 1995.View ArticleGoogle Scholar
- Okada, Y., Surface deformation due to shear and tensile faults in halfspace, Bull. Seismol. Soc. Am., 75(4), 1135–1154, 1985.Google Scholar
- Paris, R., F. Lavigne, P. Wassmer, and J. Sartohadi, Coastal sedimentation associated with the December 26, 2004 tsunami in Lhok Nga, west Banda Aceh (Sumatra, Indonesia), Mar. Geol., 238(1–4), 93–106, 2007.View ArticleGoogle Scholar
- Takahashi, T., N. Shuto, F. Imamura, and D. Asai, A movable bed model for tsunamis with exchange rate between bed load layer and suspended layer, Proc. Coatal Eng., JSCE, 46, 606–610, 1999 (in Japanese).View ArticleGoogle Scholar
- Takahashi, T., N. Shuto, F. Imamura, and D. Asai, Modeling sediment transport due to tsunamis with exchange rate between bed load layer and suspended load layer, Coast. Eng., 1508–1519, 2000.Google Scholar
- Tanaka, H., Y. Yoshitake, and N. Shuto, Numerical simulation of bottom deposit classification in wave and flow coexisting field, Proc. Coast. Eng., 36, 264–268, 1989 (in Japanese).View ArticleGoogle Scholar
- Tanioka, Y. and S. Iwasaki, Rupture processes of the 2004 Sumatra earthquake estimated using the tsunami waveform inversion, Earth Monthly, 56, 19–24, 2006.Google Scholar
- Tanioka, Y., Y. Yudhicara, T. Kususose, S. Kathiroli, Y. Nishimura, S. I. Iwasaki, and K. Satake, Rupture process of the 2004 great Sumatra-Andaman earthquake estimated from tsunami waveforms, Earth Planets Space, 58, 203–209, 2006.View ArticleGoogle Scholar
- Tsuji, Y., H. Matsutomi, Y. Tanioka, Y. Nishimura, T. Sakakiyama, T. Kamataki, Y. Murakami, A. Moore, and G. Gelfenbaum, Distribution of the tsunami heights of the 2004 Sumatra tsunami in Banda Aceh measured by the tsunami survey team, http://www.eri.utokyo.ac.jp/namegaya/sumatera/surveylog/eindex.htm, 2005.
- Watanabe, A., Y. Maruyama, T. Shimizu, and T. Sakakiyama, Numerical prediction model of three-dimensional beach transformation due to installed structures, Proc. Conf. Coast. Eng., 31, 406–410, 1984 (in Japanese).View ArticleGoogle Scholar