复制成功
  • 图案背景
  • 纯色背景

笔记

  • 2019-11-16
    为大人带来形象的羊生肖故事来历 为孩子带去快乐的生肖图画故事阅读
    谈谈怎样学好数学_苏步青-中学生文库
个人中心

上传于:2015-12-28

粉丝量:154

网络技术员,毕业于北大青鸟!!现有自己的网络工作室!



A mid-Holocene candidate tsunami deposit from the NW Cape (Western Australia)

下载积分:3000

内容提示: A mid-Holocene candidate tsunami deposit from the NW Cape(Western Australia)Simon Matthias Maya, ⁎ , Simon Falvard b , Maike Norpoth a , Anna Pint a , Dominik Brill a , Max Engel a ,Anja Scheffersc , Manuel Dierick d , Raphaël Paris b , Peter Squire c , Helmut Brückner aaInstitute of Geography, University of Cologne, Albertus-Magnus-Platz, 50923 Cologne, GermanybLaboratoire Magmas et Volcans, Université Blaise Pascal, CNRS–IRD, OPGC, 5 rue Kessler, 63038 Clermont-Ferrand, FrancecSouthern Cross GeoScie...

文档格式:PDF| 浏览次数:424| 上传日期:2015-12-28 07:17:25| 文档星级:
A mid-Holocene candidate tsunami deposit from the NW Cape(Western Australia)Simon Matthias Maya, ⁎ , Simon Falvard b , Maike Norpoth a , Anna Pint a , Dominik Brill a , Max Engel a ,Anja Scheffersc , Manuel Dierick d , Raphaël Paris b , Peter Squire c , Helmut Brückner aaInstitute of Geography, University of Cologne, Albertus-Magnus-Platz, 50923 Cologne, GermanybLaboratoire Magmas et Volcans, Université Blaise Pascal, CNRS–IRD, OPGC, 5 rue Kessler, 63038 Clermont-Ferrand, FrancecSouthern Cross GeoScience, Southern Cross University, PO Box 157, Lismore, NSW 2480, AustraliadDepartment of Physics and Astronomy, Centre for X-ray Tomography, Ghent University, Proeftuinstraat 86, 9000 Ghent, Belgiuma b s t r a c t a r t i c l e i n f oArticle history:Received 15 October 2015Received in revised form 19 November 2015Accepted 20 November 2015Available online 23 November 2015Editor: Dr. J. KnightAlthough extreme-wave events are frequent along the northwestern coast of Western Australia and tsunamis in1994 and 2006 induced considerable coastal flooding locally, robust stratigraphical evidence of prehistoric trop-ical cyclones and tsunamis from this area is lacking. Based on the analyses of X-ray computed microtomography(μCT) of oriented sediment cores, multi-proxy sediment and microfaunal analyses, optically stimulated lumines-cence (OSL) and14 C-AMS dating, this study presents detailed investigations on an allochthonous sand layer ofmarine origin found in a back-barrier depression on the NW Cape Range peninsula. The event layer consists ofmaterial from the adjacent beach and dune, fines and thins inland, and was traced up to ~400 m onshore.Although a cyclone-induced origin cannot entirely be ruled out, the particular architecture and fabric of thesediment, rip-up clasts and three subunits point to deposition by a tsunami. As such, it represents the firststratigraphical evidence of a prehistoric, mid-Holocene tsunami in NW Western Australia. It was OSL-dated to5400–4300 years ago, thus postdating the regional mid-Holocene sea-level highstand.© 2015 Elsevier B.V. All rights reserved.Keywords:Palaeotsunami depositSediment fabricX-ray microtomographyOptically stimulated luminescence1. IntroductionThe NW coast of Western Australia is impacted by 1–2 tropicalcyclones per year, and ten tsunamis have been recorded since 1858,including those following the 1883 Krakatoa eruption and the earth-quakes off the coast of Indonesia in 1977, 1994 and 2006 (Goff andChagué-Goff, 2014). Although observations during these tsunamis arerather sporadic, wave heights of ~6 m (Cape Leveque) and a run-up of~8 m Australian Height Datum (AHD) (Shelter Bay, Shark Bay) locallyoccurred during the 1977 Sumba and 2006 Java tsunamis, respectively(Prendergast and Brown, 2012; Goff and Chagué-Goff, 2014). At theNorth West Cape, the 1994 Java tsunami overwashed the dune barrierat Baudin (Fig. 1), 3.6 km S of the study area. Inundation extended upto 300 m inland, and a run-up of ~7 m AHD was inferred (Gregsonand van Reeken, 1998). In addition, tropical cyclones cause extensivecoastal flooding. As one of the most powerful cyclones in Australia'shistorical record, category 5 tropical cylone Vance (March 18th–22nd,1999) resulted in water levels of ~7 m above event tide in the ExmouthGulf (Nott and Hubbert, 2005). However, little is known about thegeological imprint of historical (Prendergast and Brown, 2012; Mayet al., 2015a) and prehistorical extreme-wave events in northwesternWestern Australia. So far, prehistorical tsunamis or storms wereinferred from corals or large molluscs in washover deposits and dunesup to 1 km inland, or from marine organisms attached to wave-emplaced boulders (Scheffers et al., 2008, and references therein).These findings lack stratigraphic contexts, and uncertainties related toradiocarbon dating of reworked marine organisms (May et al., 2015a)or the marine palaeo-reservoir effect may bias the inferredchronologies.This paper provides, for the first time, detailed stratigraphicevidence of a prehistoric extreme-wave event from northwesternWestern Australia. Based on multi-proxy sediment and microfaunalanalyses,opticallystimulatedluminescence(OSL)and14 C-dating,anal-lochthonous sand layer of marine origin was identified in a mud-filledback-barrier depression at the NW coast of the Cape Range peninsula.Asa novel approach in thecontextof tsunamideposits, X-raycomputedmicrotomography (μCT) scans of oriented cores reveal characteristicsediment fabrics, which have been described from modern (Wassmeret al., 2010) and historical (Cuven et al., 2013) tsunami deposits butare rarely preserved in palaeotsunami deposits. We discuss the originof the event layer against the background of its sedimentary character-istics and the local sedimentary environment.Sedimentary Geology 332 (2016) 40–50⁎ Corresponding author.E-mail address: mays@uni-koeln.de (S.M. May).http://dx.doi.org/10.1016/j.sedgeo.2015.11.0100037-0738/© 2015 Elsevier B.V. All rights reserved.Contents lists available at ScienceDirectSedimentary Geologyjournal homepage: www.elsevier.com/locate/sedgeo 2. Physical settingInvestigationswere carried outin thelow-lyingback-barrier zoneatNinety Mile Beach access (21°50′28.03″S, 114°2′54.06″E),located in thenorthwestern part of the NW Cape (Western Australia). The site com-prises a circular mudflatsome200 m from thesea,and slightly elevatedlow and open shrublands (Fig. 1). A dune belt of ~5 m AHD separatesthe site from the Ningaloo Reef, the local producer of bioclastic skeletalfragments. Heavy winds, sustained rainfall and severe flooding aremainly related to tropical cyclones during summer. The post-glacialmarine transgression approached its maximum at ~7000–6000 yearsBP, when the formation of the Ningaloo Reef off the NW Cape had juststarted (Collins et al., 2003). A relative sea-level (RSL) highstand of1–2 m above present level followed by a marine regression is docu-mented for several coastal sections of Western Australia (Lewis et al.,2013; Engel et al., 2015) and is also assumed for the NW Cape.The continuously falling RSL resulted in low sedimentation rates incoastal settings since c. 5000 years, impeding the accumulation ofsediment archives suitable for storing palaeo-wave events (Nott,2004). However, although its deposits have not yet been presented,the 1994 Java tsunami explicitly showed that considerable flooding ofnear-coast sediment archives may occur (Gregson and van Reeken,1998), and palaeotsunami and palaeocyclone deposits may be storedin topographical lows.3. Methods3.1. Sedimentary analysesSamples were taken from cores and trenches at 16 locations along ashore-perpendicular transect crossing a circular mudflat (Fig. 1) atNinety Mile Beach access, starting at c. 200 m from the coast. The strat-igraphicaltransectcomprisestrenchesT7,T20andT35,T36and38,per-cussion cores VC31–34 and push cores C3 (master core; Fig. 2), 22–25,27 and 37. Percussion cores were taken using an Atlas Copco CobraPro percussion corer and probes of 6 cm in diameter. Push cores weretakenbypushingplastictubesalsowithadiameterof6cmintothesed-iment by hand. Cores and trenches reached depths between 50 and150cmbelowsurface(b.s.),andsedimentrecoveryamountedtoamin-imum of 90% during coring. The distance between sampling sites wasbetween 5 and 50 m. Percussion cores and trenches were photographi-cally documented and significant facies were sampled. Push cores wereopened and sampled in the lab in steps of 0.5–1 cm. Elevation of thecores and profiles were measured using a Topcon HiPer Pro differentialglobal positioning system (DGPS) with an altimetric accuracy of ~2 cm.Elevations given in AHD (Australian Height Datum) are based on DGPSmeasurements using the static mode and AUSPOS post-processing.Modern samples were taken from the present beach (EXM B), thedune (EXM D) and the lagoon some 450 m off the beach (i.e., theHolocene reef platform, TTB).The coarse- and fine-grained fractions (N2 mm and b2 mm) wereseparated using a 2 mm sieve and weighed; percentage of the gravelfraction (N2 mm) is separately illustrated for master core C3. The air-dried fine-grained fraction was carefully hand-pestled for particledisaggregation. Grain size distribution of the fine-grained fraction wasmeasured using a Beckman Coulter LS 13320 laser particle analyser(b2mm).Sampleswerepre-treatedwithH 2 O 2 (30%)toremoveorganiccarbon and 0.5 N Na 4 P 2 O 7 (55.7 g/l) for aggregate dispersion. Statisticalmeasures were calculated after Folk and Ward (1957) using theGRADISTAT software (Blott and Pye, 2001). In order to determinesedimentary environments and sediment source areas, microfaunalanalyses (foraminifers, ostracods) were carried out for samples ofmaster core C3 and for the modern samples. Samples (10 cm 3 ) werepre-treatedwithH 2 O 2 (30%)fordispersionandsievedtoisolatefractionsofN100andb100μm.Themicrofaunalcontentwasinvestigatedunderabinocular microscope and quantitatively recorded. A minimum of ~300(foraminifers) and ~100 (ostracods) specimens per sample were count-ed where possible. We distinguished reworked and non-reworkedFig. 1.(A) Overviewof NW Western Australia [based on NASA Worldview data (NASAEOSDIS)]. Study site ismarkedbya red frame. (B)Panorama photo of studysite (Ninety Mile Beachaccess)asseenfromthetopoftheforedune(viewtowardsSW).Locationofsamplingsitesalongthe~400mlongshore-perpendicularsedimenttransect(Fig.1C)aswellasofthemodernsamples are depicted. (C) Sediment transect with landward thinning and fining of unit B, and grain size distribution of cores/trenches.41 S.M. May et al. / Sedimentary Geology 332 (2016) 40–50 foraminifers based on the degree of fragmentation and abrasion/rounding. For determination taxa and ecological preferences, Yassiniand Jones (1995), Hayward et al. (1999), De Deckker and Yokoyama(2009), and Karanovic (2012) were used.Principal component analysis (PCA) was carried out using thePAST software (Hammer et al., 2001) for samples of master core C3(Fig. 3). For PCA, the grain size parameters mean, sorting, skewnessand kurtosis as well as the percentage of microfaunal contents wereincluded. Spearman's rank correlation coefficient was used to avoidautocorrelations.We interpreted μCT scans of an oriented push core (C37) in order toinfer information about sediment fabric as well as flow directions andcharacteristics. Scans were performed at HECTOR, a custom-built CTscanner setup at UGCT(Centre ForX-ray Tomography, GhentUniversity,Belgium) (cf., Masschaele et al., 2013). The CT scanner was specificallydesigned for multidisciplinary CT applications and is built around a240 kV X-RAY WorX microfocus source (XWT 240-SE) and a Perkin-Elmer flat panel detector (1620 CN3 CS). Based on the sample size andcomposition, optimal scanner settings were determined to be 220 kVtube voltage, 50 W beam power and 1 mm of aluminium filtration toreduce beam hardening while preserving a sufficiently high signal-to-noise ratio. A total of 2000 projections was recorded, with a 1-secondexposure per projection. The isotropic voxel pitch of the reconstructedvolume was 50 μm. The tomographic datasets were reconstructedusing the Octopus software package, and the reconstructed 3D volumeswere rendered using the visualisation software package VGStudio MAX.Visual interpretation was performed using myVGL 2.1 software.Inaddition,quantificationofsedimentfabricssuchastheorientationanddipofthelongestaxesofparticlesandparticleshapemeasurements(Figs. 4, 5) was performed using ImageJ, Blob3D (Ketcham, 2005) andStereo32 (Röler and Trepmann, 2013) software. Three sections of thereconstructed core (lower parts of subunits I and II) were chosen forsubsampling and sediment fabric analyses. The original scans (images)were cropped and oriented towards the north using ImageJ software.Subsequentprocessingincludedadjustmentsofbrightnessandcontrastas well as corrections of the signal-to-noise ratio in order to extractlarger particles and remove smaller ones from the images. Subsampleswere extracted from the whole dataset and transferred into Blob3D.Fig. 2. Sediment characteristics, granulometry and microfaunal data of master core C3 and the modern samples. * — reworked (fragmented/rounded); b.s. — below surface.42 S.M. May et al. / Sedimentary Geology 332 (2016) 40–50 Subsample I (basal layer of unit B; base of subunit I, trc; cf., Fig. 4)included 360 images, subsample II (lower to middle part of subunitI) 500 images, and subsample III (lower part of subunit 2) 400 images.The grayscale images were converted into a binary image and thethreshold value between black and white was determined to 172.Subsequently, particle correction and separation was done by usingthe “remove islands” and “remove holes” function of Blob3D and,finally, manually with the separation module of Blob3D. Analyseswere stopped when a total of at least 300 particles per sample weremeasured. Finally, the shape of the ellipsoidal particles was calculatedusing Stereo32 for each subsample using the major axes of particlesextracted in Blob3D. Particle isotropy (I) and elongation (E) are basedon eigenvectors S1 (max), S2 (intermediate) and S3 (min), calculatedusing Stereo32 software.3.2. Dating techniquesThree OSL (optically stimulated luminescence) samples were takenfrom trenches T36 (T36 OSL 1, 15 cm b.s.) and T38 (T38 OSL 1; 36 and70 cm b.s.). OSL dating was performed at the Cologne LuminescenceLab (CLL; Institute of Geography, University of Cologne). To estimateenvironmental dose rates (Table 1), assessment of U, Th and K concen-trations by means of high-resolution gamma spectrometry and mea-surement of in situ water contents were performed on grab samplesof the material within a 30-cm radius around the dated sample. Incase of T36 OSL 1, where the sampled layer has a thickness of only30 cm, an additional dose rate sample was taken from the mudflatbelow and considered in the calculation of the total dose rate. In addi-tion, the contribution of the cosmic dose rate was calculated on thebasis of geographical position, altitude above sea level, and thicknessand density of the sediment overburden (Prescott and Hutton, 1994).Comparison of the decay activities of different nuclides within the238 U and 232 Th chains reveals disequilibria in the range of 12–22% forthe232 Thchain and 5–15% for the 238 U chain.For thefinal luminescenceages these disequilibria are insignificant compared to 1-sigma uncer-tainties of about 10%, since the related age differences are significantlysmaller, as already documented (e.g., Olley et al., 1996).Samples for burial dose determination were collected in opaquealuminium tubes and processed under subdued red light. Sampleswere sieved to fractions of 150–200 mm and chemically pre-treated with HCl, H 2 O 2 , sodium oxalate and HF toremove carbonates,organics, clay and the alpha-affected rim of the grains. Pure quartzwas extracted by heavy liquid separation with densities of 2.62 and2.68 g/cm 3 . Small 1-mm aliquots of quartz were measured on RisøTL/OSL devices with90 Sr/ 90 Y beta sources delivering ~0.08 to0.15 Gy/s at the sample position. Signals were stimulated by meansof blue LEDs for 40 s and a Hoya U340 filter (7.5 mm) was used forsignal detection.EquivalentdosemeasurementsfollowedtheSARprotocolof Murrayand Wintle (2003). Since a combined dose-recovery-preheat-plateauand thermal transfer test (measurement of four aliquots per tempera-ture interval after bleaching by blue LEDs for 100 s at room tempera-ture) exemplarily performed on T36 OSL 1 showed no dose–dependence for temperatures between 180 and 280 °C (transfer ofcharge b0.5% of natural D e and dose–recovery values of 0.94–0.98), apreheat temperature of 200 °C was selected for all D e measurements(Fig. 6). The general suitability of the applied SAR protocol for the sam-pleswasverifiedby(i)doserecoverytestswith measuredtogiven doseratios of 0.94–0.98 that revealed reproducibility; (ii) linear-modulatedluminescence (with linearly increasing stimulation energy over ameasurement interval of 3600 s) which demonstrates the dominanceofaneasilybleachablefastcomponent(photon-ionisationcross-sectionof 1.6–1.8 × 10 −17 ) comparable to the one described by Singarayer andBailey (2003); and (iii) constant D e (t) plots that point to thermallystable signals (Bailey, 2003).Outof48measuredaliquotspersample,between 37and45aliquotspassedtheSARacceptancecriteriainrespectofrecyclingratio(0.9–1.1),recuperation (b5% of D e ), depletion ratio (0.9–1.1), and signal to back-ground ratio (N3) and could be considered for the calculation of burialdoses. Due to the approximately normally distributed (skewness b2)and moderately scattered (overdispersion of 16–25%) D e distributions(Fig. 7), all samples were assumed to be well bleached and, in conse-quence, the central age model (CAM) of Galbraith et al. (1999) wasused for burial dose calculation. The respective dose values, statisticalcharacteristics of the D e distributions, and ages for each sample aresummarised in Table 2.A mollusc and a plant remain were dated by14 C (Table 3). Themollusc test has been transported prior to deposition.14 C-ages werecorrectedforamarinereservoireffectofΔR=65±18wherenecessary(O'Connor et al., 2010) and calibrated using CALIB 7.0 software (Hogget al., 2013; Reimer et al., 2013) (Table 3).Fig. 3. (A) PCA results show clustering of sedimentary units. Unit B separates from otherunits and has the same characteristics as the modern beach and dune. (B) NegativecorrelationofmeangrainsizeandsortingforsamplesfromunitBinC3.FortopofsubunitsB-I and B-III (fining- and sorting-up) this trend is parallel to sample depth (sample no.).(C) Relation of mean grain size and sorting for all samples from sampling sites T7,C22–25, C27, T35 and VC34. Event unit B clearly separates from basal unit A andsubsequent units C, D and E. A negative correlation of mean grain size and sorting is alsovisible for unit B.43 S.M. May et al. / Sedimentary Geology 332 (2016) 40–50 4. Sedimentology4.1. Trench-scale grain size and microfaunal characteristicsAt the base of C3, laminated sandy mud (31–24 cm b.s., Figs. 2, 3)shows layers of one to few mm thickness and mean grain size valuesbetween4and~100μm.Theforaminiferalassemblage(Fig.2)ismainlycomposed of well-preserved individuals of Adelosina longirostra,Quinqueloculina seminula and Quinqueloculina sp. Juvenile and adultCyprideis australiensis, Paracypris sp., Paracytheroma sp., Xestoleberis sp.compose the ostracod assemblage.Unconformably overlying is coarse to medium sand with abundantforaminifers, shell debris and well-rounded gravel (24–6 cm b.s., unitB). A medium sand layer (24–22 cm b.s.; base of subunit B-I) comprisesthe bottom section. Coarse sand and gravel contents increasefrom 22–20 cm b.s. and decrease between 20–17 cm b.s (top of subunitB-I). Between 17–11 cm b.s., massive sand and gravel was found,coarsening-up at 17–12 cm b.s. and fining-up at 12–11 cm b.s. (subunitB-II). The sediment lacks any gravel component and is fining-up at10–6 cm b.s. (subunit B-III).Samples from unit B cluster based on the PCA and the sorting/meangrain size ratio (Fig. 3). Granulometric characteristics are similar to thebeach and dune (EXM B and D), although the latter are slightly bettersorted (Figs. 2, 3). Reefal sediments (TTB) are poorly sorted and showincreased mud contents (Fig. 2).Reworked (i.e., fragmented/rounded) Amphistegina spp. (60–90%)and other inner shelf taxa dominate the foraminiferal assemblages ofEXM B and D as well as unit B (Figs. 2, 3A). Sample TTB has a higherdiversity and is dominated by less or non-reworked individuals ofinner shelf taxa (Fig. 2), which are also present in lower abundance insections of unit B and samples EXM B and D. Fresh individuals ofPeneroplis pertusus are only found in unit B and TTB. While samplesEXM B and D are void of ostracods (Fig. 2), several adult specimens ofC. australiensis are present in unit B.Fig. 4. μCT scans of core C37. (A) Similar to C3, subunits are present in unit B of C37. Sediment fabric (particle orientation) and normal grading are visible. (B) Rip-up clasts in subunit B-I.(C) Fining-up and fabric of subunit B-II. (D) Traction carpet (?) at the base of subunit B-I and flame structures at the boundary between units A and B. (E) Dominant (shore-parallel)orientation and dip of longest particle axes at the base of subunits B-I and B-II (red colours indicate highest, blue lowest densities); flow direction is perpendicular (landward) (OS1-3,Fig. 4A). az. — azimuth.44 S.M. May et al. / Sedimentary Geology 332 (2016) 40–50 Mean grain size again increases until 2 cm b.s., before fining-upfrom 2–0 cm b.s. A higher mud content is evident in the uppermost6 cm (unit C), and a thin mud drape overlies the sand sequence.4.2. Sediment fabric in μCT scansSimilar to unit B in C3, at least two subunits with a distinctorientation of coarser shell and gravel components and graded bed-ding are visible in the μCT scans of oriented core C37 (Figs. 1, 4).However, in C37, subunit B-II is fining-up (Fig. 4C). For the base ofsubunits B-I and B-II (OS1-3 in Fig. 4A,E), the longest axes of themainly prolate coarse particles (Fig. 5) exhibit a distinct shore-parallel orientation (SW–NE to WSW–ENE) with a dominant dip of~6–9° both towards SW and NE. Well-rounded, 15 cm-large Pleisto-cene reef rock clasts floating in unit B were found in T20 (Fig. 8).Micro-cavities (~0.5 cm) disrupt the sharp erosive boundary tounit A, resembling flame structures (Fig. 4D).Mudclasts floatingin the sandyto gravelly matrix wereidentified insubunit B-I (Figs. 4B, 8). The (palaeo-)surface of the underlying mud isoverlain by a ~1 cm-thick layer of medium to coarse sand with loweramounts of shell fragments and gravel (Fig. 4D), analogous to the baseof unit B in C3 (base of subunit B-I; Fig. 2).4.3. Transect-scale sediment characteristics and chronologyThe sand sheet of unit B was detected in trenches (T) and furthercores (VC, C) along the entire transect, except for the most landwardcore VC31 (Fig. 1). It thins and fines landwards and is partly cementedin VC32–34 and T35. The basal ~1 cm-thick layer of medium to coarsesand with reduced contents of larger mollusc fragments and gravel(base of subunit B-I; Figs. 2, 4C–E) varies in thickness; it was notfound in the landward cores. While unit B is overlain by unit C inthe cores from the mudflat, it is followed by laminated sandy mud(unit D, C27) and well-sorted fine sand (unit E, C25) landwards. In thelatter core, unit E gradually changes to muddy fine sand in its upperpart (50-0 cm b.s.). Landwards of VC32, the weathered Pleistocenereef rock emerges at the surface; VC31 penetrated deeply weatheredcemented sand and red loam.Based on the weighted mean of two OSL ages from T36 and T38[4715 ± 453 years, 4997 ± 631 years], unit B was deposited 4833 ±549 years ago (Tables 1, 2; Figs. 6, 7). This is in agreement with theFig. 5. Shape of the ellipsoids calculated from the orientations (longest axis) of theparticles separated for each subsample using the shape triangle diagram of Benn (1994).Particle isotropy (I) and elongation (E) are calculated using the eigenvectors of theellipsoids, S1 (max eigenvalue), S2 (intermediate) and S3 (min); see Benn (1994) fordetails. The eigenvalues have been calculated using Stereo32 software. All subsamplesexhibit a rather prolate shape, which supports the idea of a traction-based transportationmode of the particles.Table 1Dosimetrydata.Notes:depth b.s.— samplingdepth below surface,water — in-situ watercontent,U — uranium, Th — thorium, K — potassium, RDE— radioactive disequilibria. OSL datingwas carried out in the Cologne Luminescence Lab (CLL).Sample Depthb.s. [m]Water[%]U[ppm]Activities U-chain[Bq/kg]RDE[%]Th[ppm]Activities Th-chain[Bq/kg]RDE[%]K[%]Dose rate[Gy/ka]T36 OSL1 0.15 12 ± 5 1.06 ± 0.09 Pb 214: 13.6 ± 0.36Bi 214: 13.3 ± 0.2610 0.43 ± 0.04 Th 228: 1.7 ± 0.10Ac 228: 2.2 ± 0.36Tl 208:1.9 ± 0.2412 0.12 ± 0.00 0.55 ± 0.06T36 basal mud 0.35 63 ± 5 0.48 ± 0.03 0.49 ± 0.07 0.17 ± 0.00T38 OSL 1 0.36 10 ± 5 1.32 ± 0.06 Pb 214: 16.3 ± 0.25Bi 214: 17.5 ± 1.955 1.03 ± 0.17 Th 228: 3.9 ± 0.11Ac 228: 4.3 ± 0.34Tl 208: 5.4 ± 0.2818 0.22 ± 0.01 0.57 ± 0.09T38 OSL 2 0.70 7 ± 5 1.06 ± 0.14 Pb 214: 12.8 ± 0.21Bi 214: 13.5 ± 0.2515 0.36 ± 0.08 Th 228: 1.3 ± 0.09Ac 228: 1.7 ± 0.34Tl 208: 1.7 ± 0.2122 0.12 ± 0.00 0.74 ± 0.10Fig. 6. Results of preheat-plateau and dose–recovery test performed on T36 OSL1 (n = 4per temperature interval). Since no significant temperature dependency of dose valuesis visible, a preheat temperature of 200 °C was selected.45 S.M. May et al. / Sedimentary Geology 332 (2016) 40–50 14 C-age of a bivalve shell from unit B in C3 (5582–5757 cal years BP) at9 cm b.s. (Table 3), giving a maximum age for the sand sheet. Plantremains from unit A yielded an age of 5922–6177 cal years BP at46 cm b.s. The well-sorted fine sand of unit E in T38 (36 cm b.s.) givesan age of 3773 ± 445 years.5. Data interpretation and discussion5.1. Sediment source and origin of unit BWhile the foraminiferal assemblage of unit A, dominated by innershelf species A.longirostra, Q. seminula and Quinqueloculina sp., suggestsinter- to subtidal lagoonal conditions, juvenile C. australiensis (prefer-ring sandy substrate), Paracytheroma sp. (preferring organic-richmud), Paracypris sp. and Xestoleberis sp. indicate a protected marginal-marine palaeoenvironment with salinity fluctuations and temporarilyhypersaline conditions (e.g, Yassini and Jones, 1995), probably a semi-enclosed lagoon (Table 4). These conditions persisted until ~6050 calBP or later.Theinverse gradingin the basal (B-I, B-II) and normal gradingin theupper (B-I, B-III) part of its subunits (Figs. 1, 2), the distinct fining andthinning landward pattern, and the erosive boundary to underlyingmudflat sediments point to an event-induced origin of unit B(e.g., Morton et al., 2007). Although a suite of diagnostic characteristicsfor the identification of palaeotsunami (and palaeostorm) deposits hasbeen discussed in the aftermath of the tsunamis in 2004 (Dec 26th,Indian Ocean Tsunami), 2010 (Feb 27th, Chile Tsunami) and 2011(Mar 11th, Tohoku-oki Tsunami, Japan) (e.g., Morton et al., 2007; Pariset al., 2010; Szczuciński et al., 2012), the detection and differentiationof storm and tsunami deposits are still associated with a number ofchallenges, including site-specific sediment characteristics, anarchive-specific preservation potential, or post-depositional changes(e.g., Szczuciński, 2012). However, tsunamis tendto generate extensivesand sheets, generally fining inland and upwards, and comprising dis-tinct subunits as well as an erosive lower contact. They often containrip-up clasts of reworked material, and sediment fabrics may reflectpalaeo-current directions (Paris et al., 2010). While storm depositsmay produce similar trench-scale sediment characteristics (Mortonet al., 2007), they generally tend to be more narrow and thick, andthin landward rather abruptly on a transect scale (Paris et al., 2010).All mentionedtsunami characteristicsarefoundin unitB,thereasonwhy we tentatively conclude that it was deposited by a tsunami 5400–4300 years ago. However, Supertyphoon Haiyan on the Philippinesrecently demonstrated that powerful tropical cyclones may result incoastal flooding with tsunami-similar hydrodynamic characteristics,where infragravity waves with periods of N1 min are caused by non-linear wave interactions with shallow reef platforms (May et al.,2015b; Roeber and Bricker, 2015; Shimozono et al., 2015). In thesecases, pulses of sustained (i.e. N1 min) landward-directed coastalflooding may reach flow velocities of N5 m/s, which is in the range ofthose inferred for recent major tsunamis at the coast (Fritz et al., 2006,2012). Since wave-reef interactions and the generation of infragravitywaves at our study site cannot be excluded, a storm generation of unitB cannot entirely be ruled out.The sediment source of the candidate tsunami deposit is predomi-nantly the beach and dune directly NW of the transect based on unitB's similarities in granulometry, microfaunal composition and taphono-my (Figs. 2, 3), which reflects conditions at the sediment source and ishardly modified by tsunami transport (Szczuciński et al., 2012). Well-preserved individuals of Elphidium sp., Amphistegina sp. and Eponidescribrorepandus, and some species found only in the modern sample TTBFig. 7.Abanicoplots ofD e distributions(Dietzeetal.,2015) forsamples from trenches T36(T36OSL 1)andT38(T38OSL1 and2).All samplesshow approximatelynormallydistrib-uted dose values. Hence, the burial doses were calculated using the central age model(dark grey bars).46 S.M. May et al. / Sedimentary Geology 332 (2016) 40–50 (P. pertusus) indicate a minor component from the reef lagoon (Table 4).Some individuals of Pararotalia stellata as well as Quinqueloculina spp.suggest the entrainment of material eroded from unit A.5.2. Implications on hydrodynamic conditions and depositional processesThe orientation of particles deposited from suspension or tractionis preserved when current velocities are sufficiently high. A flow-perpendicular orientation of the longest axes of (ellipsoidal) particles,as observed in unit B (SW–NE to WSW–ENE, Fig. 4E), is typical forbedloadtransportbyrollingorslidinginstrongandhighlyconcentratedcurrents (Wassmer et al., 2010; Schneider et al., 2014) and indicatesflow directions towards the SE. Inverse grading, observed at the baseof subunits B-I and B-II, is associated with traction carpets (trc) anddominant bedload transport (tr; Figs. 2–4) also at the base of tsunamideposits from the 2006 Java Tsunami and the 2011 Tohoku-oki Tsuna-mi, where finer but dense sedimentgrainssettle in a highly concentrat-ed, collision-dominated flow (Moore et al., 2011; Szczuciński et al.,2012). In addition, the flame structures between units A and B indicaterun-up-associated,synsedimentarydeformationandtruncationofbasalstrata (i.e., unit A) (Matsumoto et al., 2008) due to high-velocityunidirectional (landward) flow generating strong shear stress.Thepoorsortingofbasallayerscanbeattributedtothepresenceofacoarser fraction (N700 μm), which causes the bimodal or close tobimodal grain size distribution of, for example, the lower samples ofsubunits B-I and B-II (Figs. 2, 3). The correlation between mean grainsize and sorting and the fining-upward trend may suggest a change indeposition from bedload to suspension within a decelerating flowparticularly in subunits B-I and B-III (Moore et al., 2011) (Fig. 2).While sediment particles in subunit B-II of C3 may have almost entirelybeen deposited bytraction,flow deceleration maybeinferred atthetopof this subunit at C37 (Fig. 4C).We tentatively interpret the deposit to represent three major wavesof onetsunami, with thebase of subunits (Figs. 2, 4) representinginitialhigh-velocity inundation pulses. In addition, sediments at its base showa trend towards better sorting between C27 and T35, before mudcontents within unit B cause very poor sorting in VC33. Deceleratingflow velocities during deposition of unit B and dominant depositionfrom suspension are inferred for the landward section of the transect(Moore et al., 2011).5.3. Comparison with regional event data and RSL changePrevious studies found major far-field tsunamis of similar age fromgeologic evidence in the northern Indian Ocean (e.g., Jackson et al.,2014). However, the regional tsunami history and numerical modelssuggest that the strongest events affecting Western Australia arisefrom earthquakes along the Java and Sumbawa segments of theSunda megathrust (Burbidge et al., 2008). While a correlation of thesepalaeoeventswiththeonepresentedherethusseemsequivocal,itisre-markable that age clusters of corals attached to wave-emplaced boul-ders at the NW Cape and of shells found in dunes in SW Exmouth Gulf(Scheffers et al., 2008, and references therein) coincide with the age ofunit B.Table 2Burial doses andluminescence ages. Notes: N acc/meas — number of accepted versusmeasured aliquots, OD— over-dispersion,CAM — central age model(Galbraith et al.,1999). OSL datingwas carried out in the Cologne Luminescence Lab (CLL). The weighted mean age of unit B is 4833 ± 549 years.Sample Lab code N (acc/meas) OD[%]Skewness Age model Burial dose[Gy]Age[a]T36 OSL1 C-L3514 42/45 16.1 0.6 CAM 2.57 ± 0.14 4715 ± 453T38 OSL 1 C-L3515 37/48 25.9 −0.22 CAM 2.83 ± 0.26 3773 ± 445T38 OSL 2 C-L3516 45/48 16.3 1.8 CAM 2.79 ± 0.16 4997 ± 631Table 314 C-AMS dating results. Notes: unid. plant remains — unidentified plant remains; unid. m. fragments ...

关注我们

关注微信公众号

您选择了以下内容