T43B-01
Sedimentary Signature of the 26 December 2004 Mega Tsunami on the Eastern Coast of Banda Aceh, Indonesia P. Wassmer, F. Lavigne, J. Sartohadi, Ph. Baumert, R. Paris
The eastern part of Banda Aceh was hit by a series of waves with flow depth measured up to 15 meters. In the Kajuh district, at least seven waves were observed by eyewitnesses. They emplaced sand deposits up to 80cm in thickness. We carried out a sedimentological study of these deposits along a 2km long transect oriented according to the main run up flow direction. The sedimentary signature of the tsunami moving landward shows a succession of sequences of normally graded deposit, except for the ungraded basal layer. Close to the shoreline only one or two sequences can be observed. The number of sequences increases progressively landward to reach a maximum of seven up to 2km from the former shoreline. Further inland, the thickness of the deposits is very limited. The sequences are attributed to sediment pulses related to each new wave reaching our investigation sites. Each pulsation corresponds to the arrival of a wave front. The high turbulence led to the deposit of the coarse basal material of each sequence. At the end of each pulse the decreasing of the energy allowed deposition of the finer material. The basal layer, which is undoubtedly related to the first wave, seems to have been truncated by the second and highest wave. This explains the absence of the fine material at the top of the basal layer. Close to the shoreline, the high energy of each new wave, and probably of the backwash, may explain the small amount of sedimentary sequences.The energy decreasing 2km from the shoreline allows each wave to sign his passage, emplacing a new sequence. This sedimentary record is of high interest for its exceptional number of sequences rarely observed so far. Such recording is to be related to several factors: the succession of several high waves progressing far inland, the presence on the shore of a stock of heterometric sediments easily removable, a flat topography which reduced the turbulence of the incoming flows. However, we put in light the key role played by the micro topography on the number of sequences through a spatial analysis of the deposits based on 150 outcrops
T43B-02
2004 Indian Ocean tsunami sediment characteristics along the Malaysia-Thailand Peninsula
In spring 2005, we analyzed stratigraphy, grain size distribution and foraminiferal assemblages of pre-tsunami and tsunami sediment from Pulau Penang and Langkawi, Malaysia, and Khao Lak, Koh Phi Phi and Koh Lanta, Thailand to gain information on tsunami sediment source and deposition style. Between three and five stratigrahic units identified at each site, contain between one and three fining upward sequences. Fining upward sequences likely reflect waning flow and/or pre-backwash deposition. Coarser sand layers reflecting high-energy forewash flow separate the fining upward sequences, with the exception of Pulau Penang, which contained a coarsening upward sequence. All contacts between the pre-tsunami and tsunami sediments were sharp. Medium and coarse sand dominated all of the tsunami sediment deposits and often contained shell fragments. Between two and four foraminiferal zones were identified at each of the five sites and provided estimates of sediment source and wave characteristics. Backwash flow and deposition occurred at Pulau Penang. Foraminiferal assemblages were also capable of distinguishing pre-tsunami sediment from tsunami sediment where no stratigraphic or lithologic boundary was evident.
T43B-03
Luminescence Dating of Tsunami sediments : Residual Signal Levels in Sediments from the 26th December 2004 Indian Ocean Event in Thailand.
Luminescence dating has been applied in the past to several tsunami deposits, notably from Northern Australia, the Scilly Isles, Scotland and Chile. Providing initial signals levels are close to zero, optically stimulated luminescence (OSL) dating is a promising method for dating palaeo-tsunami sediments, and providing an indication of past frequencies of occurrence of at least large scale events. The time resolution available, however, is critically dependent on initial signal levels in zero age deposits. With that in mind we have undertaken a study of sediments deposited in Phang Nga province and Phuket in Thailand by the 26th December 2004 tsunami. Approximately 250 sediment samples were collected from a diverse range of settings early in 2005 and prior to monsoonal disturbance. A newly developed portable OSL reader was used to assist fieldwork by making real- time measurements during the fieldwork period of more than 100 samples. This proved invaluable in confirming that all tsunami deposited material retained measureable luminescence signals, and guiding sampling towards an examination of the underlying factors that influence residual levels. Samples were collected from pre-tsunami ground surfaces, from basal and superficial tsunami deposits, from stratified sediments showing multiple sequences of coarse and fine sediments associated with different parts of the wavetrain, from a series of open and shaded back-beach locations, and from a transect reaching inland for approximately 1 km. Laboratory studies have now been conducted to extend initial screening measurements to all samples, and to perform calibrated analyses of mineral separates from selected samples. The work has confirmed that non-zero initial signal levels are present in all samples. The magnitude of such signals ranges over 4 orders of magnitude apparently controlled by the inherited age of the source sediment being entrained in the tsunami wave prior to redeposition. The lowest residual signal levels are associated with the surfaces of old ground surfaces; the initial coarse fractions, and the finer fractions settling out late in the depositional sequence. In these cases residuals ages can be decadal, which would not inhibit accurate dating of palaeo-tsunami sediments where these can be identified. Recommendations can thus be made for sampling tsunami deposits for chronometric studies.
T43B-04
Searching for Pre-2004 Tsunami Deposits in Thailand
We found only one candidate for a pre-2004 tsunami deposit during a ten-day search in July 2006 of four coastal sites in Phangnga Province. Although our initial field effort was limited, the paucity of pre-2004 tsunami deposits suggests that either there have been few Late Holocene tsunamis like the 2004 event or that the identification of paleotsunami deposits will be challenging in this region. Our investigations at Ban Bang Neang, Ban Lang Ong, Ban Nauk Nai, and Khlong Phru Sai involved examining soils and sediments 30 to 250 cm below the surface in cutbank exposures, gouge-core samples, and excavations. We targeted swales between beach ridges in areas undisturbed by tin mining and where tsunami deposits might have accumulated and been preserved. As shown in previous studies, the deposits of the 2004 tsunami extend as much as 1.5 km inland, thin over high ground, and thicken in swales. The deposits are composed of 1 to 4 beds, ranging from coarse to very fine sand, that commonly fine upward and locally contain parallel laminations. Where estuaries are relatively unprotected by mangroves, the 2004 tsunami deposits extend farther inland. Where mangroves fringe estuaries for 100s of meters, the deposits are concentrated in areas of mangrove damage. Crabs have already destroyed much of the tsunami deposits by mixing it with underlying peaty soil. At Ban Nauk Nai, we found a candidate for a pre-2004 tsunami deposit about 1.1 km inland at the back edge of the coastal plain adjacent to a steep hillslope. The deposit, identified over a distance of 60 meters, consists of a 10-cm-thick, silty, fining-upward coarse to fine sand about 25 cm below the bottom of a shallow pond and the adjoining area. In the same area, the overlying 6 to 13- cm-thick 2004 tsunami deposit consists of a normally graded fine to very fine sand. At the other sites, we found coarse, medium, and fine sand beds typical of coastal plain beach deposits. Although tsunami deposits may occur within the beach deposits, we were not able to identify any with confidence. In future searches, it may be beneficial to target natural ponds near upland areas where accumulation of organics and fines may help to separate tsunami deposits from other sandy deposits. Our findings suggest that the apparent scarcity of pre- 2004 tsunami deposits in Phangnga Province may be due in part to extensive placer mining of the coastal plain, bioturbation of mangrove soils, and difficulty in distinguishing tsunami from beach deposits.
T43B-05
Identifying tsunami deposits using shell taphonomy: Sur lagoon, Oman
On November 28th, 1945 an 8.1 magnitude earthquake focused in the eastern portion of the Makran subduction zone (Arabian Sea) generated a powerful tsunami that destroyed many coastal villages in Pakistan and India. Reports indicate that the tsunami also caused significant damage in Muscat, Oman, although its effects elsewhere in Oman are unknown. A thick bivalve dominated shell horizon was discovered inside the Sur lagoon, which is located on the eastern promontory of Oman (200 km south of Muscat). This shell deposit is significant because it is laterally extensive (> 1 km2), extends deep within the lagoon (>2 km), ranges in thickness from 5 - 25 cm at the sample localities, contains numerous subtidal and offshore bivalve species, and articulated subtidal and offshore bivalve species are abundant. Although there is an absence of typical tsunami indicators such as allochthonous sediment in and around the lagoon, verbal accounts, cultural evidence recovered during coring, and the absence of strong storms during the past 100 years indicates that this shell unit was caused by the 1945 tsunami. In this setting, it would be advantageous to have another proxy for tsunami detection and risk prediction. The use of shell taphonomy is one of the potential indicators and here we present new evidence of its utility. We sampled this unit in eight locations, and compared the shell taphonomy to surface shell samples collected from beach and reworked horizons in the lagoon, and to shell samples from a known tsunami and corresponding storm/ballast deposit in Israel (Reinhardt et al., 2006). Taphonomic analysis yielded promising results, as the two tsunami horizons shared excellent agreement between the amount of fragmented shells, and the percentage of shells displaying angular breaks. Both of these categories were significantly different from the percentage of fragments and angular fragments recovered from the reworked, beach, and storm/ballast deposits, indicating different environmental factors acting upon the shell assemblages. These results suggest that tsunamigenic shell deposits can be identified based on their taphonomic characteristics when compared to beach and storm deposits in the same setting. Our data indicates that the following diagnostic taphonomic characteristics may indicate a tsunamigenic deposit: 1) presence/absence of articulated bivalves, 2) increased percentage of fragmented valves, 3) increased percentage of angular fragments, and 4) the large number of offshore bivalves in the lagoon. This study highlights the potential benefits and opportunity for using bivalve taphonomy as a useful indicator of paleotsunami deposits, particularly in protected coastal embayments in arid regions where stratified deposits might not readily preserve. This technique holds potential, as shell deposits are easy to identify in the field by non-experts, is very low cost, and analysis uses simple, easily applicable and recognizable taphonomic characteristics. Furthermore, this proxy can be used for risk assessment purposes in coastal areas with a seismic history but lacking a paleotsunami record. Further testing of this hypothesis should be conducted along the Omani coastline, particularly in the lagoons along the Eastern promontory. Key Words: tsunami, taphonomy, Oman, bivalve
T43B-06
Identification of Potential Tsunamogenic Areas From Multibeam and Seismic Data in the Greater Antilles South Slope
As it is known, tsunamis have various general generating mechanisms associated with earthquakes, landslides/slumps and volcanic processes. Active tectonics along the north-eastern Caribbean plate is the principal cause de tsunami in the eastern Greater Antilles islands. In this area, earthquake and tsunami record suggest that the main tsunami sources are large shallow earthquakes (6.5) in submarine areas, which can yield a significant vertical motion over seafloor, and sudden landslides/slumps maybe triggered by this earthquakes generally in steeped bathymetry areas. In the spring of 2005 was carried out in the north-eastern Caribbean plate the marine geophysical survey GEOPRIO-DO aboard of Spanish R/V Hesp¦rides. This marine survey explored the north-eastern area of Puerto Rico and Virgin Islands and the southern slope of Dominican Republic and Puerto Rico. The survey included diverse geophysical techniques to research from deep to shallow structure, but with special emphasis on multibeam systematic mapping carried out in the Muertos Accretionary Prism and Muertos Trough, in the south of Dominican Republic and Puerto Rico. This accretionary complex is occupying a broad band of active compression regime with an east-west trend. It is limited to the south by the Muertos Trough (>5600 m depth), where the Venezuelan Basin oceanic crust is being underthrusted beneath Puerto Rico and Hispaniola islands. We have used multibeam systematic data to get detailed bathymetry grids in which we have made geothematic mapping (contours, slopes, slope direction, channel network). The processing of multibeam data and their integration with shallower seismic data has allowed us to make a detailed morphotectonic analysis. We have identified many seafloor structures like landslides/slumps and active fault zones with high tsunamogenic potential. Our results suggest that the combination of very active tectonics with steep seafloor slopes in the vicinity of highly populated islands yields high tsunami hazard for the eastern Greater Antilles. We consider the integration of high resolution multibeam data with shallow seismic data essential tools to identify sources with high tsunamogenic potential.
T43B-07
Using Tsunami Deposits and Observed Tsunami Heights to Test Source Models of 1960 Chile Earthquake
Near-field tsunami simulations of the giant 1960 Chile earthquake (Mw 9.5) were performed with several source models. The computed inundation areas are compared with tsunami deposits localization and observed tsunami heights. Tsunami simulations were focused on the Pacific coasts of the south-central Chile, examining the Rio Maullin estuary in detail, localized at 41.5° S midway along the 1960 rupture; in this place the 1960 earthquake lowered the area by 1.5 m, and the ensuing tsunami spread sand across lowland soils. The tsunami heights at the coast were greater than 9 m above sea level, and in much of this flooded area, the tsunami reached a height of 4.5 m. Fault parameters of source models examined in this study consider six sources from Kanamori and Cipar (1974) with the same length, width, slip, strike and dip but different rake and depth; six sources proposed by Cifuentes (1989) with the same length, strike and rake but different width, slip, dip and depth; and one source from Barrientos and Ward (1990). The results allow propose the earthquake source models for tsunami generation that better fit with real evidences. Research supported by Fondecyt 1060227 and 1060745.
T43B-08
A Millenium of Great, and Perhaps Lesser, Earthquakes Recorded by Unusual Landslide and Tsunami Deposits in Chiloe, South-Central Chile
A stratigraphic sequence at Cocotue Bay, Chiloe Is., south-central Chile, containing both earthquake-triggered debris flows and tsunami deposits, yields an earthquake history that likely includes the giant earthquakes and tsunamis inferred previously from a combination of geology, ecology, and written records in south-central Chile (AD ~1000, ~1300, 1575 and 1960). However, the Cocotue sequence also contains evidence for at least four additional earthquakes and tsunamis that are intermediate in age and perhaps lesser in size. The most recent of the additional earthquakes recorded at Cocotue probably corresponds to the poorly understood Chilean earthquake and associated trans-Pacific tsunami of 1837. The tsunami, though small on the Chilean coast, ran as high as 6 m in Hawaii and flooded fields in Japan. At Cocotue, the 1837 earthquake, like other three earlier events, is probably represented by both a debris-flow deposit and a tsunami deposit. To study the sedimentary record of intermediate and lesser events is important because this is the more probable kind of event to occur in the near future of south-central Chile. Research funded by Proyecto Fondecyt 1060227