Constraining large earthquakes along the Andaman trench using deepwater turbidites: prospects and challenges

The 2004 Sumatra-Andaman earthquake was unprecedented in terms of its magnitude (M-w 9.2), rupture length along the plate boundary (1300 km) and size of the resultant tsunami. Since 2004, efforts are being made to improve the understanding of the seismic hazard in the Sumatra-Andaman subduction zone in terms of recurrence patterns of major earthquakes and tsunamis. It is reasonable to assume that previous earthquake events in the Myanmar Andaman segment must be preserved in the geological record in the form of seismo-turbidite sequences. Here we present the prospects of conducting deep ocean palaeoseismicity investigations in order to refine the quantification of the recurrence pattern of large subduction-zone earthquakes along the Andaman-Myanmar arc. Our participation in the Sagar Kanya cruise SK-273 (in June 2010) was to test the efficacy of such a survey. The primary mission of the cruise, along a short length (300 km) of the Sumatra Andaman subduction front was to collect bathymetric data of the ocean floor trenchward of the Andaman Islands. The agenda of our piggyback survey was to fix potential coring sites that might preserve seismo-turbidite deposits. In this article we present the possibilities and challenges of such an exercise and our first-hand experience of such a preliminary survey. This account will help future researchers with similar scientific objectives who would want to survey the deep ocean archives of this region for evidence of extreme events like major earthquakes.

Organic matter preservation in turbidites on the Madeira Abyssal Plain

Oxidized intervals of five organic-rich Madeira Abyssal Plain (MAP) turbidites deposited during the Miocene, Pliocene, and Pleistocene all displayed comparable major loss of total organic carbon (TOC) (84 ± 3.1%) accompanied by a negative isotopic (d13C) shift ranging from -0.3 to -2.9 per mil. Major but significantly lower loss of total nitrogen (Ntot, 61 ± 7.1%) also occurred, leading to a decrease in TOC relative to Ntot (C/Ntot) and a +1.3 to 2.7 per mil Ntot isotopic (d15N) shift. Compound specific isotopic measurements on plant wax n-alkanes indicate the terrestrial organic component in the unoxidized deposits is 13C-enriched owing to significant C4 contribution. Selective preservation of terrestrial relative to marine organic carbon could account for the d13C behavior of TOC upon oxidation but only if a 13C-depleted component of the bulk terrestrial signal is selectively preserved in the process. Although the C/Ntot decrease and positive d15N shift seems inconsistent with selective terrestrial organic preservation, results from analysis of a Modern eolian dust sample collected in the vicinity indicate these observations are compatible. Regardless of the specific explanation for these isotopic observations, however, our findings provide evidence that paleoreconstruction of properties such as pCO2 using the d13C of TOC is a goal fraught with uncertainty whether or not the marine sedimentary record considered is 'contaminated' with significant terrestrial input. Nonetheless, despite major and selective loss of both marine and terrestrial components as a consequence of postdepositional oxidation, intensive organic geochemical proxies such as the alkenone unsaturation index, UK'37, appear resistant to change and thereby retain their paleoceanographic promise.

Mineralogy, and composition of turbidites and background samples at ODP Sites 166-1003 and 166-1007

The lower slope and toe-of-slope sediments of the western flank of the Great Bahama Bank (Sites 1003 and 1007) are characterized by an intercalation of turbidites and periplatform ooze. In general, turbidites form up to 12% of the total mass of the sedimentary column. Based primarily on data from the Bahamas, it has been postulated that steep-sided carbonate platforms shed most of their sediments into the basin during sea-level highstands when the platforms are flooded. This highstand shedding is assumed to be less pronounced along platforms with a ramp-like depositional profile where sediment production is not restricted to sea-level highstand. Miocene to Pliocene sediments recovered in five drill holes during Leg 166 at the western margin of the Great Bahama Bank reveal that turbidite distribution follows a complex pattern that is dependent on several factors such as sedimentation rates, sea-level changes, and slope morphology. To identify the depositional sequences in the cores, the depths of seismic-sequence boundaries were used. The distribution of turbidites within sedimentary sequences varies strongly. Generally, turbidites are clustered at the upper and/or lower portions of the sequences indicating deposition of carbonate turbidites during both highstand and lowstand of sea level. Analyses of the Miocene turbidites show that (1) during high sea level, 60% of all turbidites were deposited at Site 1003 (309 out of 518 turbidites), while during low sea level, two thirds of all turbidites were deposited at Site 1007 (332 out of 486 turbidites); (2) the average thickness of highstand turbidites is 1.5 times higher than the average thickness of lowstand turbidites; and (3) the turbidites display slight differences in composition and sorting. In general, highstand turbidites are less sorted and contain an abundant amount of shallow-water constituents such as green algae, red algae, shallow-water benthic foraminifers (miliolids), and intraclasts. The lowstand turbidites are better sorted and contain abundant planktonic foraminifers and micrite. To complicate matters, highstand and lowstand turbidites seem to be deposited at different locations on the slope. At the lower slope (Site 1003), more turbidites were deposited during highstands, while at the toe of the slope, turbidites were dominantly deposited during sea-level lowstands. The result is a slope section with laterally discontinuous turbidite lenses within periplatform ooze, which is controlled by the interplay of sea-level changes, sediment production, and platform morphology.

Heavy and light minerals in turbidites from DSDP Hole 19-183, the Aleutian abyssal plain

The Aleutian abyssal plain is a fossil abyssal plain of Paleogene age in the western Gulf of Alaska. The plain is a large, southward-thinning turbidite apron now cut off from sediment sources by the Aleutian Trench. Turbidite sedimentation ceased about 30 m.y. ago, and the apron is now buried under a thick blanket of pelagic deposits. Turbidites of the plain were recovered at site 183 of the Deep Sea Drilling Project on the northern edge of the apron. The heavy-mineral fraction of sand-sized samples is mostly amphibole and epidote with minor pyroxene, garnet, and sphene. The light-mineral fraction is mostly quartzose debris and feldspars. Subordinate lithic fragments consist of roughly equal amounts of metamorphic, plutonic, sedimentary, and volcanic grains. The sand compositions are arkoses in many sandstone classifications, although if fine silt is included with clay as matrix, the sand deposits are feldspathic or lithofeldspathic graywacke. The sands are apparently first-cycle products of deep dissection into a plutonic terrane, and they contrast sharply with arc-derived volcanic sandstones of similar age common on the adjacent North American continental margin. The turbidite sands are stratigraphically remarkably constant in composition, which indicates derivation from virtually the same terrane through a time span approaching 20 m.y. Comparison of Aleutian plain data with the compositions of coeval sedimentary rocks from the northeast Pacific margin shows that the Kodiak shelf area includes possible proximal equivalents of the more distal turbidites. Derivation from the volcaniclastic Mesozoic flysch of the Shumagin-Kodiak shelf is unlikely; more probably the sediments were derived from primary plutonic sources. The turbidites also resemble deposits in the Chugach Mountains and the younger turbidites of the Alaskan abyssal plain and could conceivably have been derived from the coast ranges of southeastern Alaska or western British Columbia. The Aleutian plain sediment most likely was not derived from as far south as the Oregon-Washington continental margin, where coeval sedimentary deposits are dominantly volcaniclastic.

Characteristics and distribution of Neogene turbidites at ODP Site 108-657

Numerous and variable silty-sandy siliciclastic turbidites were observed in Neogene pelagic sediments (late Miocene to Holocene) at Site 657: (1) thick-bedded, coarse-grained and thin-bedded, fine-grained turbidites; and (2) turbidites composed of eolian dune sand and shallow-water bioclasts or of fluvial-sand or mixed sandy component assemblages. The stratigraphic distribution of these turbidites indicates five periods during which climatic conditions and material sources change. Turbidite occurrence prior to 6.2 Ma (late Miocene) is sparse; the deposits contain coarse and fine-grained turbidites with quartz grains of eolian or mixed origin suggesting the existence of arid conditions at about 8.5 and 6.5 Ma. A coarse-grained turbidite of fluvial origin, recording a humid climate, occurs at about 6.2 Ma. During the early Pliocene, turbidites are frequent (15/Ma); they contain only fine-grained sequences comprising material of mixed origin, which indicates a more humid climate perhaps. The late Pliocene starts with rare coarse-grained turbidites of wind-transported sand while the uppermost Pliocene deposits show a higher frequency of fine-grained sequences (10/0.7 Ma) composed mainly of fluvial material. During the early Pleistocene, similar high turbidite frequency was observed (20/1.3 Ma) but with a total lack of eolian supply. During the last 0.7 Ma, the frequency decreases and the sequences are characterized by highly variable sediment components that could be related to strong variations of climatic conditions. The sedimentary characteristics of turbidites are mainly controlled by sediment source and climate. The frequency must be influenced by sea-level variations, by cyclic processes of climatic origin, and possibly by variations in the continental slope morphology. Clay mineral assemblages suggest a south Saharan source of terrigenous material during the late Miocene and the Pliocene and a northwest Saharan source during the Pleistocene.

Geochemistry and position of turbidites in the Cap Timiris Canyon off Mauritania

This study of sediments from the Cap Timiris Canyon demonstrates that geochemical data can provide reliable age-depth correlation even of highly turbiditic cores and attempts to improve our understanding of how turbidite emplacement is linked to climatic-related sea-level changes. The canyon incises the continental margin off NW Africa and is an active conduit for turbidity currents. In sediment cores from levee and intrachannel sites turbidites make up 6-42% of sediment columns. Age models were fitted to all studied cores by correlating downcore element data to dated reference cores, once turbidite beds had been removed from the dataset. These age models enabled us to determine turbidite emplacement times. The Cap Timiris Canyon has been active at least over the last 245 kyr, with turbidite deposition seemingly linked to stage boundaries and glacial stages. The highly turbiditic core from the intrachannel site postdates to ~15 kyr and comprises Holocene and late Pleistocene sediments. Turbidite deposition at this site was associated especially with the rapid sea-level rise at the Pleistocene/Holocene transition. During the Holocene, turbidity current activity decreased but did not cease.