Meteoritics amp Planetary Science 39 Nr 7 1233ndash1247 (2004)Abstract available online at httpmeteoriticsorg

1233 copy Meteoritical Society 2004 Printed in USA

Evidence for ocean water invasion into the Chicxulub craterat the CretaceousTertiary boundary

Kazuhisa GOTO1 Ryuji TADA1 Eiichi TAJIKA1 Timothy J BRALOWER2 Takashi HASEGAWA3 and Takafumi MATSUI4

1Department of Earth and Planetary Science The University of Tokyo Building 5 7ndash3ndash1 Hongo Tokyo 113ndash0033 Japan2Department of Geosciences Pennsylvania State University 503 Deike Building University Park Pennsylvania 16802 USA

3Department of Earth Sciences Kanazawa University Kakuma-machi Kanazawa 920ndash1192 Japan4Department of Complexity Science and Engineering The University of Tokyo 7ndash3ndash1 Hongo Tokyo 113ndash0033 Japan

Corresponding author E-mail gotoepssu-tokyoacjp

(Received 29 August 2003 revision accepted 2 March 2004)

AbstractndashThe possibility of ocean water invasion into the Chicxulub crater following the impact atthe CretaceousTertiary boundary was investigated based on examination of an impactite betweenapproximately 79463 and 89494 m in the Yaxcopoil-1 (Yax-1) core The presence of crosslamination in the uppermost part of the impactite suggests the influence of an ocean current at leastduring the sedimentation of this interval Abundant occurrence of nannofossils of late Campanian toearly Maastrichtian age in the matrices of samples from the upper part of the impactite suggests thatthe carbonate sediments deposited on the inner rim margin and outside the crater were eroded andtransported into the crater most likely by ocean water that invaded the crater after its formation Themaximum grain size of limestone lithics and vesicular melt fragments and grain and bulk chemicalcompositions show a cyclic variation in the upper part of the impactite The upward fining grain sizeand the absence of erosional contact at the base of each cycle suggest that the sediments were derivedfrom resuspension of units elsewhere in the crater most likely by high energy currents associationwith ocean water invasion

INTRODUCTION

High-energy CretaceousTertiary (KT) boundarydeposits have been reported at numerous locations near theimpact site at Chicxulub Some of these deposits are interpretedas resulting from tsunami (eg Bourgeois et al 1988 Smit etal 1996 Smit 1999) Recently tsunami deposits werediscovered at the KT boundary in western Cuba suggestingthat the tsunami were of significant scale to have affectedsedimentation in the deep proto-Caribbean Sea (Takayama etal 2000 Tada et al 2002 Forthcoming) However the originmagnitude and extent of tsunami associated with the KTboundary impact have not been determined in detail

Matsui et al (2002) proposed that ocean water invasioninto the Chicxulub crater and subsequent overflow of oceanwater from the crater was one of the possible mechanisms togenerate a giant tsunami at the KT boundary impactGeneration of tsunami by ocean water invasion has beenpostulated for other craters on the ocean floor (eg Lindstroumlmet al 1994 Ormouml and Lindstroumlm 2000 Dalwigk and Ormouml2001 Poag et al 2002) However ocean water invasion intothe crater immediately after the impact and consequent

generation of the tsunami at the KT boundary has beenquestioned (eg Ormouml and Lindstroumlm 2000) because waterdepth around the impact site was shallow (Sharpton et al1996 Pierazzo and Crawford 1997 Pierazzo et al 1998) andbecause the crater rim may have prevented ocean waterinvasion into the crater However collapse of the crater rimwas reported from the analysis of drilling cores recoveredfrom the inside and outside the crater (Sharpton et al 1996)and seismic reflection data (Morgan et al 1997)Furthermore according to numerical simulations thetransient crater rim collapsed within several minutes of theimpact (Morgan et al 2000 Collins et al 2002) suggestingthat ocean water invasion may have occurred rapidly

If ocean water invaded the crater immediately after theimpact sedimentation of the impactite defined as the varietyof rocks produced during an impact event (eg French 1998)within the Chicxulub crater should have been influenced byan ocean current However a detailed analysis of theimpactite within the Chicxulub crater has not been conductedbecause drilling core samples had not previously beenavailable for high-resolution analyses

In 2001 and 2002 the Chicxulub Scientific Drilling

1234 K Goto et al

Program (CSDP) sponsored by the International ContinentalScientific Drilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) cored the impactite within theChicxulub crater The drill hole Yaxcopoil-1 (Yax-1) wassituated approximately 65 km south of the crater center on thesouthern inside slope of the crater rim (Fig 1) and a drillingcore approximately 1500 m in length was recovered (Fig 2)

In this study we focused on the deposition of theuppermost part of the impactite in Yax-1 (79463ndash80802 m ofsub-bottom depth) to explore whether ocean water invadedthe crater immediately after the impact This interval wasdeposited during the latest stage of the formation of theChicxulub crater and thus is the most likely interval tocontain evidence of ocean water invasion The investigationintegrated quantitative analyses of grain compositionchemical composition maximum grain size and nannofossiland planktonic foraminiferal assemblages

STRATIGRAPHY

The recovered Yax-1 core consists of (from bottom totop) ~600 m-thick shallow water carbonate and sulfate rocks~100 m-thick impactite and ~800 m-thick Tertiary carbonates(Fig 2) The shallow water carbonate and sulfate rocks arethought to be mega-blocks of probable Cretaceous agedeposited in association with the KT boundary impact eventbased on the presence of shocked quartz melt fragments andimpact melt dikes in intervals at 910 m 916 m and 1347 m(Kenkmann et al 2003 Wittmann et al 2003) The impactiteis composed of an impact melt breccia and a suevite (Dressleret al 2003a) The impact melt breccia is defined as a depositcomposed of melt and rock fragments with a matrix of meltwhile suevite is a deposit composed of melt and rockfragments with a clastic matrix (eg French 1998)

Based on the lithology Dressler et al (2003a) divided thesedimentary sequence of the impactite at Yax-1 into 6 units indescending order (Fig 2) Unit 6 overlies the underlyinglimestone with irregular erosional surface is 1002 m thick(88492ndash89494 m) and is composed of a very coarse suevitewith abundant carbonate and silicate melt fragments (Dressleret al 2003a b) Unit 5 overlies unit 6 with gradational contact(Dressler et al 2003b) is 2386 m thick (86106ndash88492 m)and is composed of coarse suevitic melt agglomerate withmonomictly brecciated melt bodies (Dressler et al 2003a b)Unit 4 overlies unit 5 with sharp contact is 1526 m thick(84580ndash86106 m) and is composed of very coarse andheterogeneous melt agglomerate with fine-grained andhomogeneous brownish matrix (Dressler et al 2003a b) Unit3 overlies unit 4 with gradational contact is 2294 m thick(82286ndash84580 m) and is composed of coarse suevitic meltagglomerate with fine-grained and homogeneous brownishmatrix (Dressler et al 2003a b) Unit 2 overlies unit 3 withirregular erosional surface is 1484 m thick (80802ndash82286 m) and is composed of pebble-sized suevite (Dressler

et al 2003a b) The basal contact of the suevite in unit 1 wasnot observed due to the cracking of the core However thegrain composition and size of unit 1 seems to changegradually from unit 2 Unit 1 is 1339 m thick (79463ndash80802m) and is composed of coarse sand to pebble-sized suevite(Dressler et al 2003a) The suevite in units 1 and 2 isrelatively well- sorted and the grain size of the meltfragments and the carbonate lithics is small compared withthose in units 3ndash6 The litho-stratigraphical division ofDressler et al (2003a) is used in this study

Our own observations indicate that the lithology of anapproximately 50 cm-thick interval of dolomitic calcarenitelayer above unit 1 (79414ndash79463 m) is distinctly differentfrom the underlying impactite and the overlying Tertiarysediments However no stratigraphic subdivision of this layeris proposed and we refer to this layer as unit 0 (Fig 2)

SAMPLES AND ANALYTICAL METHODS

Thirty-five samples were taken from unit 1 in Yax-1 atapproximately 30 cm stratigraphic intervals between 79558and 80727 m (Fig 3) and subjected to detailed grain size andcomposition chemical composition and nannofossilbiostratigraphic analyses A similar number of samples was

Fig 1 A map showing locations of Yax-1 and other drilling sites inand around the Chicxulub crater (after Claeys et al 2003)

Evidence for ocean water invasion into the Chicxulub crater 1235

Fig 2 A columnar section its subdivision into units and nannofossil assemblage of the Yax-1 core Sampling horizons are also shown on theleft side of the column (modified from lithological profile on CSDP Web site httpicdpgfz-potsdamdehtmlsiteschicxulubnewsnewshtml) The litho-stratigraphical division is based on Dressler et al (2003a)

Evidence for ocean water invasion into the Chicxulub crater 1235

Fig 2 A columnar section its subdivision into units and nannofossil assemblage of the Yax-1 core Sampling horizons are also shown on theleft side of the column (modified from lithological profile on CSDP Web site httpicdpgfz-potsdamdehtmlsiteschicxulubnewsnewshtml) The litho-stratigraphical division is based on Dressler et al (2003a)

1236 K Goto et al

taken from the underlying carbonate and sulfate mega-blocksto determine their age using nannofossil biostratigraphy(Fig 2) Forty samples were taken from unit 2 and 12 sampleswere taken from units 3ndash6 to compare grain composition withunit 1 and to investigate nannofossil biostratigraphy (Figs 2and 3) Three samples were taken from unit 0 to determine theexact position of the boundary between the impactite and theoverlying Paleocene carbonate rock using nannofossil andplanktonic foraminiferal biostratigraphy All investigatedsamples were 10 cm3 in volume

Grain composition of 35 samples from unit 1 wasinvestigated by point counting We also investigated graincomposition of 40 samples from unit 2 to compare it with thatof unit 1 We investigated the components of the impactiteand their textures and composition of the matrix using thinsections Furthermore to investigate macroscopic variation ofthe grain composition we took close-up digital pictures of

rock surfaces in each sample and approximately 400 pointswere counted in these images Suevite in units 2 and 1 aremainly composed of large (gt2 mm) melt fragments and mostof them are recognizable on rock surface based on their colorsand textures

Mineral composition of suevite samples was investigatedby X-ray powder diffraction (XRD) Approximately 5 g ofbulk sample was powdered in an agate mortar The XRDanalysis was conducted using a MAC Science MXP-3 X-raypowder diffractometer at the Department of Earth andPlanetary Science the University of Tokyo The maximumsizes of carbonate grains and greenish silicate melt fragmentsin unit 1 were measured on each sample surface usingmicrometer calipers Nannofossil assemblage was examinedusing an optical microscope Planktonic foraminiferalassemblage was examined in thin sections

Non-destructive quantitative analyses of major element

Fig 3 A detailed columnar section vertical variation of bulk grain composition and nannofossil assemblage of units 2 and 1

Evidence for ocean water invasion into the Chicxulub crater 1237

composition and chemical mapping were conducted onpolished sample surface using a Horiba XGT-2700 X-rayscanning microscope (XGT) at the Department of Earth andPlanetary Science the University of Tokyo Analyticalprocedure is after Koshikawa et al (2003) Bulk chemicalcomposition of 5 major elements (Al Si Ca Ti and Fe) wasmeasured on an approximately gt15 cm2 area for the 35samples from unit 1 with XGT We excluded the chemicalcomposition data of Mg K and Mn due to the highlyheterogeneous distribution of these elements

RESULTS

Lithology and Petrography

We focused on the lithology and petrography of unit 1 toinvestigate the possibility of ocean water invasion into thecrater In this section we describe lithology graincomposition grain size and chemical composition of unit 1 indetail We further studied the lithology and petrography ofunits 2 and 0 to compare with those of unit 1

Lithology and Petrography of Unit 2

Unit 2 overlies unit 3 with an irregular erosional contact(Fig 4a) Unit 2 is approximately 15 m thick and is composedof highly heterogeneous suevite with abundant melt and lithicfragments The suevite in unit 2 is poorly-sorted and showsclast-supported fabric (Fig 4b) Unit 2 is basically composedof two normally-graded beds approximately 7 and 8 m thick(Fig 3)

The suevite in each bed is composed of pebble- tocobble-sized subangular to rounded yellowish green greenblack dark red and brown silicate melt fragments with smallamounts of limestone lithics and basement rock fragmentsLarge grayish melt fragments of up to 10 cm in diameteroccur in the basal part of this unit Content of yellowish greenmelt fragments is highly variable between 2 and 75 vol ofthe suevite (Fig 3) Most of the yellowish green meltfragments are larger than 1 cm

Two types of greenish silicate melt fragments arerecognized in this unit the first type is greenish alteredvesicular melt fragments which are also observed inunderlying unit 3 and are altered and replaced by smectite(Fig 5a) and the other type are silicate melt-coated calcitegrains (Fig 5b) The latter type is similar to the ldquocoredrdquoinclusion type melt fragment that is interpreted as fallbackejecta (French 1998) The abundance of greenish alteredvesicular melt fragments and ldquocoredrdquo inclusion type meltfragments is highly variable between 6 to 52 vol and 4ndash35vol of the bulk sample respectively Shocked quartz grainswith planer deformation features (PDFs) are observed in unit2 The PDFs are produced under pressure of 8ndash25 GPa (egFrench 1998) suggesting that deposition of unit 2 was

associated with the impact event The matrix of the suevite inthis unit is mainly composed of minute carbonate grains andcoccoliths with small amount of minute melt fragmentsquartz and plagioclase grains

Eight to ten small cyclic variations in grain compositionare observed in upward fining beds of unit 2 (Fig 3) Theseoscillations are characterized by the presence of extra largeyellowish green melt fragments of more than 1 cm indiameter Because unit 2 is highly heterogeneous and sizes ofanalytical samples are small (1 times 1 to 2 times 15 cm) it is difficultto justify the reliability of oscillations

Lithology and Petrography of Unit 1

The basal contact of unit 1 was not observed due to thecracking of the core However the grain composition and sizeof unit 1 sediments seems to change gradually from unit 2Unit 1 is approximately 13 m thick and is composed ofrelatively well-sorted suevite (Fig 4c)

Unit 1 is composed of 1ndash2 m scale slight normally andorinversely graded beds of suevite that repeated at least eighttimes (Fig 3) These graded beds are named cycles 1ndash8 inascending order The basal contacts of cycles 1ndash3 have notbeen observed due to cracking of the cores at the boundariesThe basal contacts of cycles 4ndash8 are conformable The basalcontact between the cycles 6 and 7 is relatively sharp(Fig 4d) while others are gradual The suevite in each cycleshows repetition of clast-supported and matrix-supportedfabrics The maximum size of greenish vesicular meltfragments shows inverse plus normal grading in cycles 1ndash3and normal grading in cycles 4ndash8 (Fig 6a) On the other handthe maximum size of limestone lithics shows normal gradingin cycles 1ndash8 (Fig 6a)

The major components of unit 1 are greenish vesicularmelt fragments ldquocoredrdquo inclusion type melt fragmentslimestone lithics and matrix These 4 components comprise80ndash90 of the suevite by volume (Fig 3) Large yellowishgreen melt fragments which are one of the major componentsof unit 2 are almost absent in unit 1 Limestone lithicsoccasionally contain small amount of planktonic foraminifera(Fig 5c) The matrix is mainly composed of minute carbonategrains coccoliths calcite grains small amount of minute meltfragments quartz and plagioclase grains The abundance ofgreenish vesicular melt fragments and limestone lithicsincreases upward while the abundance of ldquocoredrdquo inclusiontype melt fragments and calcareous matrix decreases (Fig 3)Oscillations in grain composition correspond to the individualcycles and are characterized by variation in the relativeabundance of fragments of limestone lithics silicate meltfragments and calcareous matrix (Fig 3) Shocked quartzgrains with PDFs (Fig 5d) are observed throughout unit 1suggesting that deposition of unit 1 was associated with animpact event

Table 1 shows the bulk chemical compositions of Al Si

1238 K Goto et al

Ca Ti and Fe in the suevite of unit 1 Oscillations in the bulkchemical compositions are also recognized in associationwith cycles 1ndash8 (Fig 6b) and are regarded as representing avariation in the ratio of the two components one ischaracterized by higher content of Ca and the other ischaracterized by higher contents of Si Fe Ti and Al(Fig 6b) Calcium is concentrated in the calcareous matrixalthough a small amount is held in the limestone lithics while

Si Fe Ti and Al are concentrated in the silicate meltfragments As a result higher Ca content implies largermatrix content This is consistent with the observation thathigher Ca samples in the upper part of each cycle show matrixsupported fabric (Figs 7a and b) while lower Ca samples inthe lower part of each cycle show clast-supported fabric(Figs 7c and d)

Approximately 20 cm interval of the uppermost part ofunit 1 (79463ndash79479 m) is composed of greenish mediumto coarse-grained sandstone with parallel to mono-directionalcross lamination (Fig 8) suggesting that an ocean currentinfluenced the deposition of this interval

Lithology and Petrography of Unit 0

Unit 0 is approximately 50 cm thick conformablyoverlies unit 1 (Fig 8) and is mainly composed of light graydolomitic very fine calcarenite The dolomitic calcarenite isinterbeded with seven ~2 cm-thick greenish suevitic beds(Fig 8) Dolomitic calcarenite is mainly composed of lightgray dolomite grains and micrite Rhombohedral dolomitegrains in unit 0 are of probable authigenic origin Crosslaminations which are unidirectional current cross-laminawith climbing structure (Fig 8) are observed in the lower20 cm of this unit while faint parallel lamination is observedin dolomitic calcarenite in the middle 20 cm of this unit(Fig 8) A 2 cm-thick black clay layer overlies this unit withprobable parallel unconformity (Fig 8 79412ndash79414 m)

Age Constraints of Carbonate Mega-Blocks Impactiteand Unit 0 in the Yax-1 Core

Sediments underlying the impactite (gt89494 m) arecomposed of probable mega-blocks of calcareniteslimestones dolostones and anhydrites which were probablyformed as a result of collapse of the rim caused by the impact(Kenkmann et al 2003 Wittmann et al 2003) Nannofossilassemblages of two calcarenite samples from 147449 m and145751 m in the lower part of carbonate mega-blocks containLithraphidites acutum Axopodorhabdus albianusMicrostaurus chiastius Eprolithus floralis and Rhagodiscusasper suggesting a late Albian to late Cenomanian ageFurthermore nannofossil assemblages of a calcarenite samplefrom 140744 m also contains Microstaurus chiastius whichis older than the latest Cenomanian Consequently the age oflower part of carbonate and sulfate mega-blocks are olderthan the latest Cenomanian (Fig 2) The middle to upper partof carbonate and sulfate mega-blocks is almost barren ofnannofossils so their ages are uncertain (Fig 2)

Units 6 to 4 are almost barren of nannofossils Samplesfrom 82324 and 82356 m in unit 3 contain a mixture ofspecies which combined suggest a late Campanian to earlyMaastrichtian age (Fig 2 Quadrum trifidum Quadrumgothicum and Eiffellithus eximius) The sample from

Fig 4 Core photographs of a) an irregular erosional contact betweenunits 3 and 2 (arrows) b) the poorly-sorted suevite in unit 2 (81778ndash81793 m) c) the well-sorted suevite in unit 1 (79921ndash79936 m)and d) a conformable and sharp contact between cycles 6 and 7 ofunit 1 Photo (d) is a photo of a piece taken by UNAM using anunrolled core scan (360deg)

Evidence for ocean water invasion into the Chicxulub crater 1239

82356 m also contains one specimen of M chiastius (Fig 2)M chiastius in unit 3 may have been derived from underlyingcarbonate mega-blocks because M chiastius is also found inthe lower part of carbonate mega-blocks On the other handthere is no late Campanian to early Maastrichtiannannofossils in carbonate mega-blocks suggesting thatnannofossils of late Campanian to early Maastrichtian agewas derived not from underlying carbonate mega-blocks butfrom outside the crater

Nannofossil assemblages of units 2 and 1 are the mixtureof diagnostic species the co-occurrence of which suggests arange from late Campanian to early Maastrichtian (Quadrumtrifidum Eiffellithus eximius Tranolithus orionatusAspidolithus parcus and Quadrum gartneri) Samples alsocontain numerous long-ranging Cretaceous taxa (Fig 3)Even though there could be numerous source horizons forthese reworked nannofossils (as discussed in Bralower et al(1998) for the ldquococktailrdquo assemblages in the Gulf of Mexico)

we interpret the minimum source of late Campanian to earlyMaastrichtian for the diagnostic species One specimen ofMicula murus which is the latest Maastrichtian in age isfound in a sample from the basal part of unit 2 (Fig 381996 m) Mixed nature of reworked microfossils includinglatest Maastrichtian age lithic fragments and impact derivedmaterials in the suevite of units 2 and 1 is similar to the KTboundary ldquococktailrdquo deposits in the Gulf of Mexico (Braloweret al 1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Unit 0 is barren of nannofossils and planktonicforaminifera (Fig 8) The lowermost Danian (P0) is absentbased on planktonic foraminifera (Smit et al 2003) and Iranomaly data has not been reported Therefore the boundarybetween the impactite and Paleocene may be anunconformity although we cannot still exclude the possibilitythat the boundary is within unit 0

Fig 5 Thin section photomicrographs of a) greenish altered vesicular melt fragments in unit 3 (8407 m open nicol) and b) a ldquocoredrdquoinclusion type melt fragment which is composed of silicate melt-coated calcite grains and is interpreted as fall back ejecta in unit 2 (81023m open nicol) c) a limestone lithic in unit 2 (81023 m open nicol) and d) a shocked quartz grain with three sets of PDFs in unit 1 (79490m open nicol)

1240 K Goto et al

The first appearance of Danian planktonic foraminifera(P eugubina) is just above the black clay layer (Fig 8according to Keller et al 2003) Smit et al (2003) reportedthat Paleocene fauna corresponding to G eugubina zoneappears 5 cm above the black clay layer (Fig 8 79407 m)The sample from 79394 m contains abundant biserialheterohelicids and globigerine specimens larger than 03 mmin size and a single specimen of Globigerinelloides-likeplanispiral genus These planktonic foraminifers suggestCretaceous age and are possibly reworked Very thin-walledindividuals are preserved among thick-walled specimensindicating that dissolution has not substantially modified thedepositional assemblage The sample from 79394 m alsocontains nannofossils Thoracosphaera and Braarudosphaerathat occur right after the impact In addition very smallspecimens of Cruciplacolithus primus are also observed inthis sample (Fig 8) suggesting an age younger than 648 Ma(C29r P1a) according to Berggren et al (1995)

DISCUSSION

The Possibility of Ocean Water Invasion and SedimentaryProcess of Units 2 to 1

Mixed nature of reworked microfossils including latestMaastrichtian age lithic fragments and impact derivedmaterials such as shocked quartz grains with PDFs in thesuevite of units 2 and 1 is similar to the KT boundaryldquococktailrdquo deposits in the Gulf of Mexico (Bralower et al1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Although the KT boundary sequence in Yax-1 core maynot be stratigraphically complete sedimentological dataprovide compelling evidence for events immediately after theimpact Reworked nannofossils of late Campanian to earlyMaastrichtian age are abundant in the matrix of units 2 and 1

Table 1 Bulk chemical composition of the suevite in unit 1 (wt) determined by XGTSamplenumber Cycle

Sub-bottomdepth (m)

Al2O3(wt)

SiO2(wt)

CaO(wt)

TiO2(wt)

Fe2O3(wt)

1277 8 79558 1018 3733 1854 044 459 1278 8 79587 1226 4454 949 054 564 1279 8 79618 1283 4681 1071 053 558 1280 8 79647 1106 4067 932 051 516 1281 7 79676 1061 3903 2050 046 439 1126 7 79752 1145 4019 1115 050 518 1127 7 79783 1182 4936 683 061 617 1128 7 79815 1296 4946 761 059 578 1129 6 79845 1113 4089 1550 049 496 1130 6 79874 1169 4540 859 056 562 1131 6 79896 1079 4358 1022 053 575 1133 5 79932 1172 4047 1657 046 492 1134 5 79962 1138 4356 1079 052 533 1135 5 79992 1172 4900 736 056 600 1137 5 80055 1209 4420 677 053 569 1138 5 80085 1162 4163 1081 051 523 1139 4 80115 941 3368 2029 041 423 1140 4 80134 1212 4639 817 053 554 1141 4 80161 1249 4715 666 058 594 1142 4 80196 1271 4798 678 057 588 1143 3 80231 1082 3519 2217 045 403 1145 3 80308 1072 3871 1779 052 440 1146 3 80338 1316 4446 1198 055 509 1147 3 80368 1121 4308 1354 051 521 1148 3 80398 1098 4521 870 056 567 1149 3 80428 1292 4919 782 059 587 1150 2 80458 1055 3704 1921 046 487 1151 2 80528 1176 4391 1170 052 548 1152 2 80547 1221 4641 787 056 610 1153 2 80577 1138 4456 897 052 550 1154 1 80607 1070 3803 2126 047 459 1155 1 80637 1049 3889 1692 045 449 1156 1 80667 1037 3768 1516 047 450 1157 1 80697 1222 4609 854 054 568 1158 1 80727 1235 4711 842 057 591

Evidence for ocean water invasion into the Chicxulub crater 1241

It is important to note that this age range is considerablynarrower than the range of the sediments excavated by thecrater formation This suggests that these nannofossils wereselectively incorporated in units 2 and 1 There is no lateCampanian to early Maastrichtian nannofossils in carbonatemega-blocks underneath the impactite in Yax-1 suggestingthat these nannofossils were not derived from underlyingcarbonate mega-blocks These nannofossils (also thelimestone lithics) cannot have been derived from the ejectaplume because they show no signs of recrystallization ormelting Therefore nannofossils (also the limestone lithics) inunits 1 and 2 were probably derived from unconsolidatedcarbonate sediments deposited outside the crater within theejecta curtain deposits or from the rim wall

Study of the Tvaumlren crater in Sweden which is 2 km indiameter and was formed on the floor of an ocean 100ndash150 mdeep (Ormouml and Lindstroumlm 2000) has documented resurgedeposits in the upper part of the impactite formed by the oceanwater invasion immediately after the impact (eg Ormouml andLindstroumlm 2000) These resurge deposits tend to show thereversed chronology (Ormouml and Lindstroumlm 2000) Namelythe lower part of the deposit consists of the particles derived

from younger sediments and the upper part consists of theparticles derived from older sediments (Ormouml 1994 Ormoumland Lindstroumlm 2000) This is because ocean floor sedimentsoutside the crater were eroded by currents and transportedinto the crater beginning with younger units and progressingto older units (Ormouml and Lindstroumlm 2000) In the case of thesuevite in Yax-1 core nannofossil assemblage of the majorpart of units 1 and 2 are of late Campanian to earlyMaastrichtian age while the only latest Maastrichtiannannofossil is observed in the basal part of unit 2 Thus themost probable mechanism of erosion and transportation of thenannofossils in the unconsolidated carbonate sediments of theinner rim margin andor outside the crater is via ocean waterinvasion

If carbonate sediments of Campanian to Maastrichtianage outside the crater were eroded by the surge of ocean waterinto the crater the sedimentary sequence outside the cratershould have a hiatus corresponding to this age rangeAlternatively collapse of the rim wall and ejecta curtaindeposits on the rim by the ocean water invasion or the impactseismic wave also could have supplied sediment particleswith nannofossils of Campanian to Maastrichtian age into the

Fig 6 a) A diagram showing vertical variations of maximum size of limestone lithics (mm) and greenish vesicular melt fragments The arrowsof dashed and solid lines indicate normal or inverse grading of limestone lithics and greenish melt fragments in each cycle respectively Theshades indicate intervals of high calcium content in each cycle b) diagrams showing vertical variations of chemical composition of the suevitein unit 1 (wt) determined by XGT

1242 K Goto et al

crater In this case the age range of nannofossils in the craterrim and ejecta curtain deposits outside the rim should bedominated by Campanian to Maastrichtian age However noinformation is available on the age of the impactite outside thecrater Further detailed research of the impactite outside thecrater is needed to determine the source of carbonatesediments of Campanian to Maastrichtian age in units 2 and 1

Unit 2 is composed of two normally graded beds (Fig 3)Irregular erosional basal contact poor sorting clast-supportedfabric and the occurrence of intraclast-like melt fragments atthe base of the lower bed suggest that this unit was depositedby gravity flows (eg Middleton and Hampton 1976) Thepresence of nannofossils probably derived from outside thecrater indicates that gravity flows were formed in associationwith ocean water invasion Alternatively if the collapse of therim wall and ejecta curtain deposits accumulated on the rimwere the sources of nannofossils impact seismic wave mayhave triggered gravity flows In this case gravity flows mayhave run down on the rim surface in association with theocean water invasion or under the hotdry conditions withoutinfluence of water invasion

The suevite in unit 1 is well-sorted and shows upwardfining of limestone lithics in each cycle This indicates thatthe suevite in each cycle in unit 1 was deposited as a result ofsettlement of re-suspended sediment particles Furthermorethere is no erosional surface at the base of each cyclesuggesting that coarse sediment particles which probablywere derived from outside the crater andor crater rimprobably have been transported in suspension by water massmovement Therefore repeated ocean water invasion is themost probable mechanism for the cyclic sedimentation of thesuevite in unit 1 Repeated forceful ocean water invasion isprobably required to transport mm-size grains in each cycle assuspension

Similar cycles are also reported from the impactite in theChesapeake Bay crater (Poag et al 2002) The ChesapeakeBay crater is 85 km in diameter and is buried 300ndash500 mbelow the floor of the southern part of the Chesapeake Bay(Poag et al 1994 2002) Poag et al (2002) mentioned that theupper part of the impactite in the Chesapeake Bay craterconsists of repeating upward fining beds They interpretedthese cycles to have been formed by repeated ocean water

Fig 7 a) A calcium distribution map and b) a rock surface photograph of higher calcium content sample from 80231 m in unit 1 Highercalcium sample shows matrix-supported fabric c) a calcium distribution map and d) a rock surface photograph of lower calcium contentsample from 80196 m of unit 1 The lower calcium sample shows clast-supported fabric

Evidence for ocean water invasion into the Chicxulub crater 1243

invasion based on facies analysis of cores and downholegeophysical logs Similarity in cyclic sedimentation pattern inthe upper part of the impactite both in the Chicxulub craterand the Chesapeake Bay crater supports our interpretationthat unit 1 in Yax-1 was probably formed by lateral supply ofcoarse sediments through repeated ocean water invasion intothe crater cavity

The thicknesses of cycles are 7ndash8 m in unit 2 and 1ndash2 min unit 1 at Yax-1 These thicknesses seem thin for repeatedforceful ocean water invasion (= resurge) when comparedwith the thicknesses of the resurge deposits in other marinetarget craters However this could be owing to the position ofYax-1 which is located near the edge of the crater rim and

well above the crater floor In fact the suevite reaches thethickness of ~300 m at C1 S1 and Y6 cores (Sharpton et al1996 Claeys et al 2003) that are located in the central part ofthe crater (Fig 1)

The presence of parallel and mono-directional crosslaminations in the uppermost part of unit 1 and the lower partof unit 0 suggests that the crater was already filled with waterand that sedimentation of this interval was under the influenceof currents Cross lamination is one of the typical charactersof the resurge deposits and is also found in other marine targetimpact craters such as the Lockne crater in Sweden (Dalwigkand Ormouml 2001) which is ~135 km in diameter and wasformed on the ocean floor in water over 200 m deep (Ormouml

Fig 8 A core photograph of the uppermost part of unit 1 unit 0 and the overlying Paleocene carbonate (79397ndash79463 m) Nannofossil andplanktonic foraminiferal assemblage are also shown The left-side photo is a photo of a piece taken by UNAM using an unrolled core scan(360deg) The right-side photos show half cut surfaces of the uppermost part of unit 1 and the basal part of unit 0 Parallel and mono-directionalcross laminations are observed in these intervals GS = greenish suevitic bed

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

Evidence for ocean water invasion into the Chicxulub crater 1237

composition and chemical mapping were conducted onpolished sample surface using a Horiba XGT-2700 X-rayscanning microscope (XGT) at the Department of Earth andPlanetary Science the University of Tokyo Analyticalprocedure is after Koshikawa et al (2003) Bulk chemicalcomposition of 5 major elements (Al Si Ca Ti and Fe) wasmeasured on an approximately gt15 cm2 area for the 35samples from unit 1 with XGT We excluded the chemicalcomposition data of Mg K and Mn due to the highlyheterogeneous distribution of these elements

RESULTS

Lithology and Petrography

We focused on the lithology and petrography of unit 1 toinvestigate the possibility of ocean water invasion into thecrater In this section we describe lithology graincomposition grain size and chemical composition of unit 1 indetail We further studied the lithology and petrography ofunits 2 and 0 to compare with those of unit 1

Lithology and Petrography of Unit 2

Unit 2 overlies unit 3 with an irregular erosional contact(Fig 4a) Unit 2 is approximately 15 m thick and is composedof highly heterogeneous suevite with abundant melt and lithicfragments The suevite in unit 2 is poorly-sorted and showsclast-supported fabric (Fig 4b) Unit 2 is basically composedof two normally-graded beds approximately 7 and 8 m thick(Fig 3)

The suevite in each bed is composed of pebble- tocobble-sized subangular to rounded yellowish green greenblack dark red and brown silicate melt fragments with smallamounts of limestone lithics and basement rock fragmentsLarge grayish melt fragments of up to 10 cm in diameteroccur in the basal part of this unit Content of yellowish greenmelt fragments is highly variable between 2 and 75 vol ofthe suevite (Fig 3) Most of the yellowish green meltfragments are larger than 1 cm

Two types of greenish silicate melt fragments arerecognized in this unit the first type is greenish alteredvesicular melt fragments which are also observed inunderlying unit 3 and are altered and replaced by smectite(Fig 5a) and the other type are silicate melt-coated calcitegrains (Fig 5b) The latter type is similar to the ldquocoredrdquoinclusion type melt fragment that is interpreted as fallbackejecta (French 1998) The abundance of greenish alteredvesicular melt fragments and ldquocoredrdquo inclusion type meltfragments is highly variable between 6 to 52 vol and 4ndash35vol of the bulk sample respectively Shocked quartz grainswith planer deformation features (PDFs) are observed in unit2 The PDFs are produced under pressure of 8ndash25 GPa (egFrench 1998) suggesting that deposition of unit 2 was

associated with the impact event The matrix of the suevite inthis unit is mainly composed of minute carbonate grains andcoccoliths with small amount of minute melt fragmentsquartz and plagioclase grains

Eight to ten small cyclic variations in grain compositionare observed in upward fining beds of unit 2 (Fig 3) Theseoscillations are characterized by the presence of extra largeyellowish green melt fragments of more than 1 cm indiameter Because unit 2 is highly heterogeneous and sizes ofanalytical samples are small (1 times 1 to 2 times 15 cm) it is difficultto justify the reliability of oscillations

Lithology and Petrography of Unit 1

The basal contact of unit 1 was not observed due to thecracking of the core However the grain composition and sizeof unit 1 sediments seems to change gradually from unit 2Unit 1 is approximately 13 m thick and is composed ofrelatively well-sorted suevite (Fig 4c)

Unit 1 is composed of 1ndash2 m scale slight normally andorinversely graded beds of suevite that repeated at least eighttimes (Fig 3) These graded beds are named cycles 1ndash8 inascending order The basal contacts of cycles 1ndash3 have notbeen observed due to cracking of the cores at the boundariesThe basal contacts of cycles 4ndash8 are conformable The basalcontact between the cycles 6 and 7 is relatively sharp(Fig 4d) while others are gradual The suevite in each cycleshows repetition of clast-supported and matrix-supportedfabrics The maximum size of greenish vesicular meltfragments shows inverse plus normal grading in cycles 1ndash3and normal grading in cycles 4ndash8 (Fig 6a) On the other handthe maximum size of limestone lithics shows normal gradingin cycles 1ndash8 (Fig 6a)

The major components of unit 1 are greenish vesicularmelt fragments ldquocoredrdquo inclusion type melt fragmentslimestone lithics and matrix These 4 components comprise80ndash90 of the suevite by volume (Fig 3) Large yellowishgreen melt fragments which are one of the major componentsof unit 2 are almost absent in unit 1 Limestone lithicsoccasionally contain small amount of planktonic foraminifera(Fig 5c) The matrix is mainly composed of minute carbonategrains coccoliths calcite grains small amount of minute meltfragments quartz and plagioclase grains The abundance ofgreenish vesicular melt fragments and limestone lithicsincreases upward while the abundance of ldquocoredrdquo inclusiontype melt fragments and calcareous matrix decreases (Fig 3)Oscillations in grain composition correspond to the individualcycles and are characterized by variation in the relativeabundance of fragments of limestone lithics silicate meltfragments and calcareous matrix (Fig 3) Shocked quartzgrains with PDFs (Fig 5d) are observed throughout unit 1suggesting that deposition of unit 1 was associated with animpact event

Table 1 shows the bulk chemical compositions of Al Si

1238 K Goto et al

Ca Ti and Fe in the suevite of unit 1 Oscillations in the bulkchemical compositions are also recognized in associationwith cycles 1ndash8 (Fig 6b) and are regarded as representing avariation in the ratio of the two components one ischaracterized by higher content of Ca and the other ischaracterized by higher contents of Si Fe Ti and Al(Fig 6b) Calcium is concentrated in the calcareous matrixalthough a small amount is held in the limestone lithics while

Si Fe Ti and Al are concentrated in the silicate meltfragments As a result higher Ca content implies largermatrix content This is consistent with the observation thathigher Ca samples in the upper part of each cycle show matrixsupported fabric (Figs 7a and b) while lower Ca samples inthe lower part of each cycle show clast-supported fabric(Figs 7c and d)

Approximately 20 cm interval of the uppermost part ofunit 1 (79463ndash79479 m) is composed of greenish mediumto coarse-grained sandstone with parallel to mono-directionalcross lamination (Fig 8) suggesting that an ocean currentinfluenced the deposition of this interval

Lithology and Petrography of Unit 0

Unit 0 is approximately 50 cm thick conformablyoverlies unit 1 (Fig 8) and is mainly composed of light graydolomitic very fine calcarenite The dolomitic calcarenite isinterbeded with seven ~2 cm-thick greenish suevitic beds(Fig 8) Dolomitic calcarenite is mainly composed of lightgray dolomite grains and micrite Rhombohedral dolomitegrains in unit 0 are of probable authigenic origin Crosslaminations which are unidirectional current cross-laminawith climbing structure (Fig 8) are observed in the lower20 cm of this unit while faint parallel lamination is observedin dolomitic calcarenite in the middle 20 cm of this unit(Fig 8) A 2 cm-thick black clay layer overlies this unit withprobable parallel unconformity (Fig 8 79412ndash79414 m)

Age Constraints of Carbonate Mega-Blocks Impactiteand Unit 0 in the Yax-1 Core

Sediments underlying the impactite (gt89494 m) arecomposed of probable mega-blocks of calcareniteslimestones dolostones and anhydrites which were probablyformed as a result of collapse of the rim caused by the impact(Kenkmann et al 2003 Wittmann et al 2003) Nannofossilassemblages of two calcarenite samples from 147449 m and145751 m in the lower part of carbonate mega-blocks containLithraphidites acutum Axopodorhabdus albianusMicrostaurus chiastius Eprolithus floralis and Rhagodiscusasper suggesting a late Albian to late Cenomanian ageFurthermore nannofossil assemblages of a calcarenite samplefrom 140744 m also contains Microstaurus chiastius whichis older than the latest Cenomanian Consequently the age oflower part of carbonate and sulfate mega-blocks are olderthan the latest Cenomanian (Fig 2) The middle to upper partof carbonate and sulfate mega-blocks is almost barren ofnannofossils so their ages are uncertain (Fig 2)

Units 6 to 4 are almost barren of nannofossils Samplesfrom 82324 and 82356 m in unit 3 contain a mixture ofspecies which combined suggest a late Campanian to earlyMaastrichtian age (Fig 2 Quadrum trifidum Quadrumgothicum and Eiffellithus eximius) The sample from

Fig 4 Core photographs of a) an irregular erosional contact betweenunits 3 and 2 (arrows) b) the poorly-sorted suevite in unit 2 (81778ndash81793 m) c) the well-sorted suevite in unit 1 (79921ndash79936 m)and d) a conformable and sharp contact between cycles 6 and 7 ofunit 1 Photo (d) is a photo of a piece taken by UNAM using anunrolled core scan (360deg)

Evidence for ocean water invasion into the Chicxulub crater 1239

82356 m also contains one specimen of M chiastius (Fig 2)M chiastius in unit 3 may have been derived from underlyingcarbonate mega-blocks because M chiastius is also found inthe lower part of carbonate mega-blocks On the other handthere is no late Campanian to early Maastrichtiannannofossils in carbonate mega-blocks suggesting thatnannofossils of late Campanian to early Maastrichtian agewas derived not from underlying carbonate mega-blocks butfrom outside the crater

Nannofossil assemblages of units 2 and 1 are the mixtureof diagnostic species the co-occurrence of which suggests arange from late Campanian to early Maastrichtian (Quadrumtrifidum Eiffellithus eximius Tranolithus orionatusAspidolithus parcus and Quadrum gartneri) Samples alsocontain numerous long-ranging Cretaceous taxa (Fig 3)Even though there could be numerous source horizons forthese reworked nannofossils (as discussed in Bralower et al(1998) for the ldquococktailrdquo assemblages in the Gulf of Mexico)

we interpret the minimum source of late Campanian to earlyMaastrichtian for the diagnostic species One specimen ofMicula murus which is the latest Maastrichtian in age isfound in a sample from the basal part of unit 2 (Fig 381996 m) Mixed nature of reworked microfossils includinglatest Maastrichtian age lithic fragments and impact derivedmaterials in the suevite of units 2 and 1 is similar to the KTboundary ldquococktailrdquo deposits in the Gulf of Mexico (Braloweret al 1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Unit 0 is barren of nannofossils and planktonicforaminifera (Fig 8) The lowermost Danian (P0) is absentbased on planktonic foraminifera (Smit et al 2003) and Iranomaly data has not been reported Therefore the boundarybetween the impactite and Paleocene may be anunconformity although we cannot still exclude the possibilitythat the boundary is within unit 0

Fig 5 Thin section photomicrographs of a) greenish altered vesicular melt fragments in unit 3 (8407 m open nicol) and b) a ldquocoredrdquoinclusion type melt fragment which is composed of silicate melt-coated calcite grains and is interpreted as fall back ejecta in unit 2 (81023m open nicol) c) a limestone lithic in unit 2 (81023 m open nicol) and d) a shocked quartz grain with three sets of PDFs in unit 1 (79490m open nicol)

1240 K Goto et al

The first appearance of Danian planktonic foraminifera(P eugubina) is just above the black clay layer (Fig 8according to Keller et al 2003) Smit et al (2003) reportedthat Paleocene fauna corresponding to G eugubina zoneappears 5 cm above the black clay layer (Fig 8 79407 m)The sample from 79394 m contains abundant biserialheterohelicids and globigerine specimens larger than 03 mmin size and a single specimen of Globigerinelloides-likeplanispiral genus These planktonic foraminifers suggestCretaceous age and are possibly reworked Very thin-walledindividuals are preserved among thick-walled specimensindicating that dissolution has not substantially modified thedepositional assemblage The sample from 79394 m alsocontains nannofossils Thoracosphaera and Braarudosphaerathat occur right after the impact In addition very smallspecimens of Cruciplacolithus primus are also observed inthis sample (Fig 8) suggesting an age younger than 648 Ma(C29r P1a) according to Berggren et al (1995)

DISCUSSION

The Possibility of Ocean Water Invasion and SedimentaryProcess of Units 2 to 1

Mixed nature of reworked microfossils including latestMaastrichtian age lithic fragments and impact derivedmaterials such as shocked quartz grains with PDFs in thesuevite of units 2 and 1 is similar to the KT boundaryldquococktailrdquo deposits in the Gulf of Mexico (Bralower et al1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Although the KT boundary sequence in Yax-1 core maynot be stratigraphically complete sedimentological dataprovide compelling evidence for events immediately after theimpact Reworked nannofossils of late Campanian to earlyMaastrichtian age are abundant in the matrix of units 2 and 1

Table 1 Bulk chemical composition of the suevite in unit 1 (wt) determined by XGTSamplenumber Cycle

Sub-bottomdepth (m)

Al2O3(wt)

SiO2(wt)

CaO(wt)

TiO2(wt)

Fe2O3(wt)

1277 8 79558 1018 3733 1854 044 459 1278 8 79587 1226 4454 949 054 564 1279 8 79618 1283 4681 1071 053 558 1280 8 79647 1106 4067 932 051 516 1281 7 79676 1061 3903 2050 046 439 1126 7 79752 1145 4019 1115 050 518 1127 7 79783 1182 4936 683 061 617 1128 7 79815 1296 4946 761 059 578 1129 6 79845 1113 4089 1550 049 496 1130 6 79874 1169 4540 859 056 562 1131 6 79896 1079 4358 1022 053 575 1133 5 79932 1172 4047 1657 046 492 1134 5 79962 1138 4356 1079 052 533 1135 5 79992 1172 4900 736 056 600 1137 5 80055 1209 4420 677 053 569 1138 5 80085 1162 4163 1081 051 523 1139 4 80115 941 3368 2029 041 423 1140 4 80134 1212 4639 817 053 554 1141 4 80161 1249 4715 666 058 594 1142 4 80196 1271 4798 678 057 588 1143 3 80231 1082 3519 2217 045 403 1145 3 80308 1072 3871 1779 052 440 1146 3 80338 1316 4446 1198 055 509 1147 3 80368 1121 4308 1354 051 521 1148 3 80398 1098 4521 870 056 567 1149 3 80428 1292 4919 782 059 587 1150 2 80458 1055 3704 1921 046 487 1151 2 80528 1176 4391 1170 052 548 1152 2 80547 1221 4641 787 056 610 1153 2 80577 1138 4456 897 052 550 1154 1 80607 1070 3803 2126 047 459 1155 1 80637 1049 3889 1692 045 449 1156 1 80667 1037 3768 1516 047 450 1157 1 80697 1222 4609 854 054 568 1158 1 80727 1235 4711 842 057 591

Evidence for ocean water invasion into the Chicxulub crater 1241

It is important to note that this age range is considerablynarrower than the range of the sediments excavated by thecrater formation This suggests that these nannofossils wereselectively incorporated in units 2 and 1 There is no lateCampanian to early Maastrichtian nannofossils in carbonatemega-blocks underneath the impactite in Yax-1 suggestingthat these nannofossils were not derived from underlyingcarbonate mega-blocks These nannofossils (also thelimestone lithics) cannot have been derived from the ejectaplume because they show no signs of recrystallization ormelting Therefore nannofossils (also the limestone lithics) inunits 1 and 2 were probably derived from unconsolidatedcarbonate sediments deposited outside the crater within theejecta curtain deposits or from the rim wall

Study of the Tvaumlren crater in Sweden which is 2 km indiameter and was formed on the floor of an ocean 100ndash150 mdeep (Ormouml and Lindstroumlm 2000) has documented resurgedeposits in the upper part of the impactite formed by the oceanwater invasion immediately after the impact (eg Ormouml andLindstroumlm 2000) These resurge deposits tend to show thereversed chronology (Ormouml and Lindstroumlm 2000) Namelythe lower part of the deposit consists of the particles derived

from younger sediments and the upper part consists of theparticles derived from older sediments (Ormouml 1994 Ormoumland Lindstroumlm 2000) This is because ocean floor sedimentsoutside the crater were eroded by currents and transportedinto the crater beginning with younger units and progressingto older units (Ormouml and Lindstroumlm 2000) In the case of thesuevite in Yax-1 core nannofossil assemblage of the majorpart of units 1 and 2 are of late Campanian to earlyMaastrichtian age while the only latest Maastrichtiannannofossil is observed in the basal part of unit 2 Thus themost probable mechanism of erosion and transportation of thenannofossils in the unconsolidated carbonate sediments of theinner rim margin andor outside the crater is via ocean waterinvasion

If carbonate sediments of Campanian to Maastrichtianage outside the crater were eroded by the surge of ocean waterinto the crater the sedimentary sequence outside the cratershould have a hiatus corresponding to this age rangeAlternatively collapse of the rim wall and ejecta curtaindeposits on the rim by the ocean water invasion or the impactseismic wave also could have supplied sediment particleswith nannofossils of Campanian to Maastrichtian age into the

Fig 6 a) A diagram showing vertical variations of maximum size of limestone lithics (mm) and greenish vesicular melt fragments The arrowsof dashed and solid lines indicate normal or inverse grading of limestone lithics and greenish melt fragments in each cycle respectively Theshades indicate intervals of high calcium content in each cycle b) diagrams showing vertical variations of chemical composition of the suevitein unit 1 (wt) determined by XGT

1242 K Goto et al

crater In this case the age range of nannofossils in the craterrim and ejecta curtain deposits outside the rim should bedominated by Campanian to Maastrichtian age However noinformation is available on the age of the impactite outside thecrater Further detailed research of the impactite outside thecrater is needed to determine the source of carbonatesediments of Campanian to Maastrichtian age in units 2 and 1

Unit 2 is composed of two normally graded beds (Fig 3)Irregular erosional basal contact poor sorting clast-supportedfabric and the occurrence of intraclast-like melt fragments atthe base of the lower bed suggest that this unit was depositedby gravity flows (eg Middleton and Hampton 1976) Thepresence of nannofossils probably derived from outside thecrater indicates that gravity flows were formed in associationwith ocean water invasion Alternatively if the collapse of therim wall and ejecta curtain deposits accumulated on the rimwere the sources of nannofossils impact seismic wave mayhave triggered gravity flows In this case gravity flows mayhave run down on the rim surface in association with theocean water invasion or under the hotdry conditions withoutinfluence of water invasion

The suevite in unit 1 is well-sorted and shows upwardfining of limestone lithics in each cycle This indicates thatthe suevite in each cycle in unit 1 was deposited as a result ofsettlement of re-suspended sediment particles Furthermorethere is no erosional surface at the base of each cyclesuggesting that coarse sediment particles which probablywere derived from outside the crater andor crater rimprobably have been transported in suspension by water massmovement Therefore repeated ocean water invasion is themost probable mechanism for the cyclic sedimentation of thesuevite in unit 1 Repeated forceful ocean water invasion isprobably required to transport mm-size grains in each cycle assuspension

Similar cycles are also reported from the impactite in theChesapeake Bay crater (Poag et al 2002) The ChesapeakeBay crater is 85 km in diameter and is buried 300ndash500 mbelow the floor of the southern part of the Chesapeake Bay(Poag et al 1994 2002) Poag et al (2002) mentioned that theupper part of the impactite in the Chesapeake Bay craterconsists of repeating upward fining beds They interpretedthese cycles to have been formed by repeated ocean water

Fig 7 a) A calcium distribution map and b) a rock surface photograph of higher calcium content sample from 80231 m in unit 1 Highercalcium sample shows matrix-supported fabric c) a calcium distribution map and d) a rock surface photograph of lower calcium contentsample from 80196 m of unit 1 The lower calcium sample shows clast-supported fabric

Evidence for ocean water invasion into the Chicxulub crater 1243

invasion based on facies analysis of cores and downholegeophysical logs Similarity in cyclic sedimentation pattern inthe upper part of the impactite both in the Chicxulub craterand the Chesapeake Bay crater supports our interpretationthat unit 1 in Yax-1 was probably formed by lateral supply ofcoarse sediments through repeated ocean water invasion intothe crater cavity

The thicknesses of cycles are 7ndash8 m in unit 2 and 1ndash2 min unit 1 at Yax-1 These thicknesses seem thin for repeatedforceful ocean water invasion (= resurge) when comparedwith the thicknesses of the resurge deposits in other marinetarget craters However this could be owing to the position ofYax-1 which is located near the edge of the crater rim and

well above the crater floor In fact the suevite reaches thethickness of ~300 m at C1 S1 and Y6 cores (Sharpton et al1996 Claeys et al 2003) that are located in the central part ofthe crater (Fig 1)

The presence of parallel and mono-directional crosslaminations in the uppermost part of unit 1 and the lower partof unit 0 suggests that the crater was already filled with waterand that sedimentation of this interval was under the influenceof currents Cross lamination is one of the typical charactersof the resurge deposits and is also found in other marine targetimpact craters such as the Lockne crater in Sweden (Dalwigkand Ormouml 2001) which is ~135 km in diameter and wasformed on the ocean floor in water over 200 m deep (Ormouml

Fig 8 A core photograph of the uppermost part of unit 1 unit 0 and the overlying Paleocene carbonate (79397ndash79463 m) Nannofossil andplanktonic foraminiferal assemblage are also shown The left-side photo is a photo of a piece taken by UNAM using an unrolled core scan(360deg) The right-side photos show half cut surfaces of the uppermost part of unit 1 and the basal part of unit 0 Parallel and mono-directionalcross laminations are observed in these intervals GS = greenish suevitic bed

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

1238 K Goto et al

Ca Ti and Fe in the suevite of unit 1 Oscillations in the bulkchemical compositions are also recognized in associationwith cycles 1ndash8 (Fig 6b) and are regarded as representing avariation in the ratio of the two components one ischaracterized by higher content of Ca and the other ischaracterized by higher contents of Si Fe Ti and Al(Fig 6b) Calcium is concentrated in the calcareous matrixalthough a small amount is held in the limestone lithics while

Si Fe Ti and Al are concentrated in the silicate meltfragments As a result higher Ca content implies largermatrix content This is consistent with the observation thathigher Ca samples in the upper part of each cycle show matrixsupported fabric (Figs 7a and b) while lower Ca samples inthe lower part of each cycle show clast-supported fabric(Figs 7c and d)

Approximately 20 cm interval of the uppermost part ofunit 1 (79463ndash79479 m) is composed of greenish mediumto coarse-grained sandstone with parallel to mono-directionalcross lamination (Fig 8) suggesting that an ocean currentinfluenced the deposition of this interval

Lithology and Petrography of Unit 0

Unit 0 is approximately 50 cm thick conformablyoverlies unit 1 (Fig 8) and is mainly composed of light graydolomitic very fine calcarenite The dolomitic calcarenite isinterbeded with seven ~2 cm-thick greenish suevitic beds(Fig 8) Dolomitic calcarenite is mainly composed of lightgray dolomite grains and micrite Rhombohedral dolomitegrains in unit 0 are of probable authigenic origin Crosslaminations which are unidirectional current cross-laminawith climbing structure (Fig 8) are observed in the lower20 cm of this unit while faint parallel lamination is observedin dolomitic calcarenite in the middle 20 cm of this unit(Fig 8) A 2 cm-thick black clay layer overlies this unit withprobable parallel unconformity (Fig 8 79412ndash79414 m)

Age Constraints of Carbonate Mega-Blocks Impactiteand Unit 0 in the Yax-1 Core

Sediments underlying the impactite (gt89494 m) arecomposed of probable mega-blocks of calcareniteslimestones dolostones and anhydrites which were probablyformed as a result of collapse of the rim caused by the impact(Kenkmann et al 2003 Wittmann et al 2003) Nannofossilassemblages of two calcarenite samples from 147449 m and145751 m in the lower part of carbonate mega-blocks containLithraphidites acutum Axopodorhabdus albianusMicrostaurus chiastius Eprolithus floralis and Rhagodiscusasper suggesting a late Albian to late Cenomanian ageFurthermore nannofossil assemblages of a calcarenite samplefrom 140744 m also contains Microstaurus chiastius whichis older than the latest Cenomanian Consequently the age oflower part of carbonate and sulfate mega-blocks are olderthan the latest Cenomanian (Fig 2) The middle to upper partof carbonate and sulfate mega-blocks is almost barren ofnannofossils so their ages are uncertain (Fig 2)

Units 6 to 4 are almost barren of nannofossils Samplesfrom 82324 and 82356 m in unit 3 contain a mixture ofspecies which combined suggest a late Campanian to earlyMaastrichtian age (Fig 2 Quadrum trifidum Quadrumgothicum and Eiffellithus eximius) The sample from

Fig 4 Core photographs of a) an irregular erosional contact betweenunits 3 and 2 (arrows) b) the poorly-sorted suevite in unit 2 (81778ndash81793 m) c) the well-sorted suevite in unit 1 (79921ndash79936 m)and d) a conformable and sharp contact between cycles 6 and 7 ofunit 1 Photo (d) is a photo of a piece taken by UNAM using anunrolled core scan (360deg)

Evidence for ocean water invasion into the Chicxulub crater 1239

82356 m also contains one specimen of M chiastius (Fig 2)M chiastius in unit 3 may have been derived from underlyingcarbonate mega-blocks because M chiastius is also found inthe lower part of carbonate mega-blocks On the other handthere is no late Campanian to early Maastrichtiannannofossils in carbonate mega-blocks suggesting thatnannofossils of late Campanian to early Maastrichtian agewas derived not from underlying carbonate mega-blocks butfrom outside the crater

Nannofossil assemblages of units 2 and 1 are the mixtureof diagnostic species the co-occurrence of which suggests arange from late Campanian to early Maastrichtian (Quadrumtrifidum Eiffellithus eximius Tranolithus orionatusAspidolithus parcus and Quadrum gartneri) Samples alsocontain numerous long-ranging Cretaceous taxa (Fig 3)Even though there could be numerous source horizons forthese reworked nannofossils (as discussed in Bralower et al(1998) for the ldquococktailrdquo assemblages in the Gulf of Mexico)

we interpret the minimum source of late Campanian to earlyMaastrichtian for the diagnostic species One specimen ofMicula murus which is the latest Maastrichtian in age isfound in a sample from the basal part of unit 2 (Fig 381996 m) Mixed nature of reworked microfossils includinglatest Maastrichtian age lithic fragments and impact derivedmaterials in the suevite of units 2 and 1 is similar to the KTboundary ldquococktailrdquo deposits in the Gulf of Mexico (Braloweret al 1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Unit 0 is barren of nannofossils and planktonicforaminifera (Fig 8) The lowermost Danian (P0) is absentbased on planktonic foraminifera (Smit et al 2003) and Iranomaly data has not been reported Therefore the boundarybetween the impactite and Paleocene may be anunconformity although we cannot still exclude the possibilitythat the boundary is within unit 0

Fig 5 Thin section photomicrographs of a) greenish altered vesicular melt fragments in unit 3 (8407 m open nicol) and b) a ldquocoredrdquoinclusion type melt fragment which is composed of silicate melt-coated calcite grains and is interpreted as fall back ejecta in unit 2 (81023m open nicol) c) a limestone lithic in unit 2 (81023 m open nicol) and d) a shocked quartz grain with three sets of PDFs in unit 1 (79490m open nicol)

1240 K Goto et al

The first appearance of Danian planktonic foraminifera(P eugubina) is just above the black clay layer (Fig 8according to Keller et al 2003) Smit et al (2003) reportedthat Paleocene fauna corresponding to G eugubina zoneappears 5 cm above the black clay layer (Fig 8 79407 m)The sample from 79394 m contains abundant biserialheterohelicids and globigerine specimens larger than 03 mmin size and a single specimen of Globigerinelloides-likeplanispiral genus These planktonic foraminifers suggestCretaceous age and are possibly reworked Very thin-walledindividuals are preserved among thick-walled specimensindicating that dissolution has not substantially modified thedepositional assemblage The sample from 79394 m alsocontains nannofossils Thoracosphaera and Braarudosphaerathat occur right after the impact In addition very smallspecimens of Cruciplacolithus primus are also observed inthis sample (Fig 8) suggesting an age younger than 648 Ma(C29r P1a) according to Berggren et al (1995)

DISCUSSION

The Possibility of Ocean Water Invasion and SedimentaryProcess of Units 2 to 1

Mixed nature of reworked microfossils including latestMaastrichtian age lithic fragments and impact derivedmaterials such as shocked quartz grains with PDFs in thesuevite of units 2 and 1 is similar to the KT boundaryldquococktailrdquo deposits in the Gulf of Mexico (Bralower et al1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Although the KT boundary sequence in Yax-1 core maynot be stratigraphically complete sedimentological dataprovide compelling evidence for events immediately after theimpact Reworked nannofossils of late Campanian to earlyMaastrichtian age are abundant in the matrix of units 2 and 1

Table 1 Bulk chemical composition of the suevite in unit 1 (wt) determined by XGTSamplenumber Cycle

Sub-bottomdepth (m)

Al2O3(wt)

SiO2(wt)

CaO(wt)

TiO2(wt)

Fe2O3(wt)

1277 8 79558 1018 3733 1854 044 459 1278 8 79587 1226 4454 949 054 564 1279 8 79618 1283 4681 1071 053 558 1280 8 79647 1106 4067 932 051 516 1281 7 79676 1061 3903 2050 046 439 1126 7 79752 1145 4019 1115 050 518 1127 7 79783 1182 4936 683 061 617 1128 7 79815 1296 4946 761 059 578 1129 6 79845 1113 4089 1550 049 496 1130 6 79874 1169 4540 859 056 562 1131 6 79896 1079 4358 1022 053 575 1133 5 79932 1172 4047 1657 046 492 1134 5 79962 1138 4356 1079 052 533 1135 5 79992 1172 4900 736 056 600 1137 5 80055 1209 4420 677 053 569 1138 5 80085 1162 4163 1081 051 523 1139 4 80115 941 3368 2029 041 423 1140 4 80134 1212 4639 817 053 554 1141 4 80161 1249 4715 666 058 594 1142 4 80196 1271 4798 678 057 588 1143 3 80231 1082 3519 2217 045 403 1145 3 80308 1072 3871 1779 052 440 1146 3 80338 1316 4446 1198 055 509 1147 3 80368 1121 4308 1354 051 521 1148 3 80398 1098 4521 870 056 567 1149 3 80428 1292 4919 782 059 587 1150 2 80458 1055 3704 1921 046 487 1151 2 80528 1176 4391 1170 052 548 1152 2 80547 1221 4641 787 056 610 1153 2 80577 1138 4456 897 052 550 1154 1 80607 1070 3803 2126 047 459 1155 1 80637 1049 3889 1692 045 449 1156 1 80667 1037 3768 1516 047 450 1157 1 80697 1222 4609 854 054 568 1158 1 80727 1235 4711 842 057 591

Evidence for ocean water invasion into the Chicxulub crater 1241

It is important to note that this age range is considerablynarrower than the range of the sediments excavated by thecrater formation This suggests that these nannofossils wereselectively incorporated in units 2 and 1 There is no lateCampanian to early Maastrichtian nannofossils in carbonatemega-blocks underneath the impactite in Yax-1 suggestingthat these nannofossils were not derived from underlyingcarbonate mega-blocks These nannofossils (also thelimestone lithics) cannot have been derived from the ejectaplume because they show no signs of recrystallization ormelting Therefore nannofossils (also the limestone lithics) inunits 1 and 2 were probably derived from unconsolidatedcarbonate sediments deposited outside the crater within theejecta curtain deposits or from the rim wall

Study of the Tvaumlren crater in Sweden which is 2 km indiameter and was formed on the floor of an ocean 100ndash150 mdeep (Ormouml and Lindstroumlm 2000) has documented resurgedeposits in the upper part of the impactite formed by the oceanwater invasion immediately after the impact (eg Ormouml andLindstroumlm 2000) These resurge deposits tend to show thereversed chronology (Ormouml and Lindstroumlm 2000) Namelythe lower part of the deposit consists of the particles derived

from younger sediments and the upper part consists of theparticles derived from older sediments (Ormouml 1994 Ormoumland Lindstroumlm 2000) This is because ocean floor sedimentsoutside the crater were eroded by currents and transportedinto the crater beginning with younger units and progressingto older units (Ormouml and Lindstroumlm 2000) In the case of thesuevite in Yax-1 core nannofossil assemblage of the majorpart of units 1 and 2 are of late Campanian to earlyMaastrichtian age while the only latest Maastrichtiannannofossil is observed in the basal part of unit 2 Thus themost probable mechanism of erosion and transportation of thenannofossils in the unconsolidated carbonate sediments of theinner rim margin andor outside the crater is via ocean waterinvasion

If carbonate sediments of Campanian to Maastrichtianage outside the crater were eroded by the surge of ocean waterinto the crater the sedimentary sequence outside the cratershould have a hiatus corresponding to this age rangeAlternatively collapse of the rim wall and ejecta curtaindeposits on the rim by the ocean water invasion or the impactseismic wave also could have supplied sediment particleswith nannofossils of Campanian to Maastrichtian age into the

Fig 6 a) A diagram showing vertical variations of maximum size of limestone lithics (mm) and greenish vesicular melt fragments The arrowsof dashed and solid lines indicate normal or inverse grading of limestone lithics and greenish melt fragments in each cycle respectively Theshades indicate intervals of high calcium content in each cycle b) diagrams showing vertical variations of chemical composition of the suevitein unit 1 (wt) determined by XGT

1242 K Goto et al

crater In this case the age range of nannofossils in the craterrim and ejecta curtain deposits outside the rim should bedominated by Campanian to Maastrichtian age However noinformation is available on the age of the impactite outside thecrater Further detailed research of the impactite outside thecrater is needed to determine the source of carbonatesediments of Campanian to Maastrichtian age in units 2 and 1

Unit 2 is composed of two normally graded beds (Fig 3)Irregular erosional basal contact poor sorting clast-supportedfabric and the occurrence of intraclast-like melt fragments atthe base of the lower bed suggest that this unit was depositedby gravity flows (eg Middleton and Hampton 1976) Thepresence of nannofossils probably derived from outside thecrater indicates that gravity flows were formed in associationwith ocean water invasion Alternatively if the collapse of therim wall and ejecta curtain deposits accumulated on the rimwere the sources of nannofossils impact seismic wave mayhave triggered gravity flows In this case gravity flows mayhave run down on the rim surface in association with theocean water invasion or under the hotdry conditions withoutinfluence of water invasion

The suevite in unit 1 is well-sorted and shows upwardfining of limestone lithics in each cycle This indicates thatthe suevite in each cycle in unit 1 was deposited as a result ofsettlement of re-suspended sediment particles Furthermorethere is no erosional surface at the base of each cyclesuggesting that coarse sediment particles which probablywere derived from outside the crater andor crater rimprobably have been transported in suspension by water massmovement Therefore repeated ocean water invasion is themost probable mechanism for the cyclic sedimentation of thesuevite in unit 1 Repeated forceful ocean water invasion isprobably required to transport mm-size grains in each cycle assuspension

Similar cycles are also reported from the impactite in theChesapeake Bay crater (Poag et al 2002) The ChesapeakeBay crater is 85 km in diameter and is buried 300ndash500 mbelow the floor of the southern part of the Chesapeake Bay(Poag et al 1994 2002) Poag et al (2002) mentioned that theupper part of the impactite in the Chesapeake Bay craterconsists of repeating upward fining beds They interpretedthese cycles to have been formed by repeated ocean water

Fig 7 a) A calcium distribution map and b) a rock surface photograph of higher calcium content sample from 80231 m in unit 1 Highercalcium sample shows matrix-supported fabric c) a calcium distribution map and d) a rock surface photograph of lower calcium contentsample from 80196 m of unit 1 The lower calcium sample shows clast-supported fabric

Evidence for ocean water invasion into the Chicxulub crater 1243

invasion based on facies analysis of cores and downholegeophysical logs Similarity in cyclic sedimentation pattern inthe upper part of the impactite both in the Chicxulub craterand the Chesapeake Bay crater supports our interpretationthat unit 1 in Yax-1 was probably formed by lateral supply ofcoarse sediments through repeated ocean water invasion intothe crater cavity

The thicknesses of cycles are 7ndash8 m in unit 2 and 1ndash2 min unit 1 at Yax-1 These thicknesses seem thin for repeatedforceful ocean water invasion (= resurge) when comparedwith the thicknesses of the resurge deposits in other marinetarget craters However this could be owing to the position ofYax-1 which is located near the edge of the crater rim and

well above the crater floor In fact the suevite reaches thethickness of ~300 m at C1 S1 and Y6 cores (Sharpton et al1996 Claeys et al 2003) that are located in the central part ofthe crater (Fig 1)

The presence of parallel and mono-directional crosslaminations in the uppermost part of unit 1 and the lower partof unit 0 suggests that the crater was already filled with waterand that sedimentation of this interval was under the influenceof currents Cross lamination is one of the typical charactersof the resurge deposits and is also found in other marine targetimpact craters such as the Lockne crater in Sweden (Dalwigkand Ormouml 2001) which is ~135 km in diameter and wasformed on the ocean floor in water over 200 m deep (Ormouml

Fig 8 A core photograph of the uppermost part of unit 1 unit 0 and the overlying Paleocene carbonate (79397ndash79463 m) Nannofossil andplanktonic foraminiferal assemblage are also shown The left-side photo is a photo of a piece taken by UNAM using an unrolled core scan(360deg) The right-side photos show half cut surfaces of the uppermost part of unit 1 and the basal part of unit 0 Parallel and mono-directionalcross laminations are observed in these intervals GS = greenish suevitic bed

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

Evidence for ocean water invasion into the Chicxulub crater 1239

82356 m also contains one specimen of M chiastius (Fig 2)M chiastius in unit 3 may have been derived from underlyingcarbonate mega-blocks because M chiastius is also found inthe lower part of carbonate mega-blocks On the other handthere is no late Campanian to early Maastrichtiannannofossils in carbonate mega-blocks suggesting thatnannofossils of late Campanian to early Maastrichtian agewas derived not from underlying carbonate mega-blocks butfrom outside the crater

Nannofossil assemblages of units 2 and 1 are the mixtureof diagnostic species the co-occurrence of which suggests arange from late Campanian to early Maastrichtian (Quadrumtrifidum Eiffellithus eximius Tranolithus orionatusAspidolithus parcus and Quadrum gartneri) Samples alsocontain numerous long-ranging Cretaceous taxa (Fig 3)Even though there could be numerous source horizons forthese reworked nannofossils (as discussed in Bralower et al(1998) for the ldquococktailrdquo assemblages in the Gulf of Mexico)

we interpret the minimum source of late Campanian to earlyMaastrichtian for the diagnostic species One specimen ofMicula murus which is the latest Maastrichtian in age isfound in a sample from the basal part of unit 2 (Fig 381996 m) Mixed nature of reworked microfossils includinglatest Maastrichtian age lithic fragments and impact derivedmaterials in the suevite of units 2 and 1 is similar to the KTboundary ldquococktailrdquo deposits in the Gulf of Mexico (Braloweret al 1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Unit 0 is barren of nannofossils and planktonicforaminifera (Fig 8) The lowermost Danian (P0) is absentbased on planktonic foraminifera (Smit et al 2003) and Iranomaly data has not been reported Therefore the boundarybetween the impactite and Paleocene may be anunconformity although we cannot still exclude the possibilitythat the boundary is within unit 0

Fig 5 Thin section photomicrographs of a) greenish altered vesicular melt fragments in unit 3 (8407 m open nicol) and b) a ldquocoredrdquoinclusion type melt fragment which is composed of silicate melt-coated calcite grains and is interpreted as fall back ejecta in unit 2 (81023m open nicol) c) a limestone lithic in unit 2 (81023 m open nicol) and d) a shocked quartz grain with three sets of PDFs in unit 1 (79490m open nicol)

1240 K Goto et al

The first appearance of Danian planktonic foraminifera(P eugubina) is just above the black clay layer (Fig 8according to Keller et al 2003) Smit et al (2003) reportedthat Paleocene fauna corresponding to G eugubina zoneappears 5 cm above the black clay layer (Fig 8 79407 m)The sample from 79394 m contains abundant biserialheterohelicids and globigerine specimens larger than 03 mmin size and a single specimen of Globigerinelloides-likeplanispiral genus These planktonic foraminifers suggestCretaceous age and are possibly reworked Very thin-walledindividuals are preserved among thick-walled specimensindicating that dissolution has not substantially modified thedepositional assemblage The sample from 79394 m alsocontains nannofossils Thoracosphaera and Braarudosphaerathat occur right after the impact In addition very smallspecimens of Cruciplacolithus primus are also observed inthis sample (Fig 8) suggesting an age younger than 648 Ma(C29r P1a) according to Berggren et al (1995)

DISCUSSION

The Possibility of Ocean Water Invasion and SedimentaryProcess of Units 2 to 1

Mixed nature of reworked microfossils including latestMaastrichtian age lithic fragments and impact derivedmaterials such as shocked quartz grains with PDFs in thesuevite of units 2 and 1 is similar to the KT boundaryldquococktailrdquo deposits in the Gulf of Mexico (Bralower et al1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Although the KT boundary sequence in Yax-1 core maynot be stratigraphically complete sedimentological dataprovide compelling evidence for events immediately after theimpact Reworked nannofossils of late Campanian to earlyMaastrichtian age are abundant in the matrix of units 2 and 1

Table 1 Bulk chemical composition of the suevite in unit 1 (wt) determined by XGTSamplenumber Cycle

Sub-bottomdepth (m)

Al2O3(wt)

SiO2(wt)

CaO(wt)

TiO2(wt)

Fe2O3(wt)

1277 8 79558 1018 3733 1854 044 459 1278 8 79587 1226 4454 949 054 564 1279 8 79618 1283 4681 1071 053 558 1280 8 79647 1106 4067 932 051 516 1281 7 79676 1061 3903 2050 046 439 1126 7 79752 1145 4019 1115 050 518 1127 7 79783 1182 4936 683 061 617 1128 7 79815 1296 4946 761 059 578 1129 6 79845 1113 4089 1550 049 496 1130 6 79874 1169 4540 859 056 562 1131 6 79896 1079 4358 1022 053 575 1133 5 79932 1172 4047 1657 046 492 1134 5 79962 1138 4356 1079 052 533 1135 5 79992 1172 4900 736 056 600 1137 5 80055 1209 4420 677 053 569 1138 5 80085 1162 4163 1081 051 523 1139 4 80115 941 3368 2029 041 423 1140 4 80134 1212 4639 817 053 554 1141 4 80161 1249 4715 666 058 594 1142 4 80196 1271 4798 678 057 588 1143 3 80231 1082 3519 2217 045 403 1145 3 80308 1072 3871 1779 052 440 1146 3 80338 1316 4446 1198 055 509 1147 3 80368 1121 4308 1354 051 521 1148 3 80398 1098 4521 870 056 567 1149 3 80428 1292 4919 782 059 587 1150 2 80458 1055 3704 1921 046 487 1151 2 80528 1176 4391 1170 052 548 1152 2 80547 1221 4641 787 056 610 1153 2 80577 1138 4456 897 052 550 1154 1 80607 1070 3803 2126 047 459 1155 1 80637 1049 3889 1692 045 449 1156 1 80667 1037 3768 1516 047 450 1157 1 80697 1222 4609 854 054 568 1158 1 80727 1235 4711 842 057 591

Evidence for ocean water invasion into the Chicxulub crater 1241

It is important to note that this age range is considerablynarrower than the range of the sediments excavated by thecrater formation This suggests that these nannofossils wereselectively incorporated in units 2 and 1 There is no lateCampanian to early Maastrichtian nannofossils in carbonatemega-blocks underneath the impactite in Yax-1 suggestingthat these nannofossils were not derived from underlyingcarbonate mega-blocks These nannofossils (also thelimestone lithics) cannot have been derived from the ejectaplume because they show no signs of recrystallization ormelting Therefore nannofossils (also the limestone lithics) inunits 1 and 2 were probably derived from unconsolidatedcarbonate sediments deposited outside the crater within theejecta curtain deposits or from the rim wall

Study of the Tvaumlren crater in Sweden which is 2 km indiameter and was formed on the floor of an ocean 100ndash150 mdeep (Ormouml and Lindstroumlm 2000) has documented resurgedeposits in the upper part of the impactite formed by the oceanwater invasion immediately after the impact (eg Ormouml andLindstroumlm 2000) These resurge deposits tend to show thereversed chronology (Ormouml and Lindstroumlm 2000) Namelythe lower part of the deposit consists of the particles derived

from younger sediments and the upper part consists of theparticles derived from older sediments (Ormouml 1994 Ormoumland Lindstroumlm 2000) This is because ocean floor sedimentsoutside the crater were eroded by currents and transportedinto the crater beginning with younger units and progressingto older units (Ormouml and Lindstroumlm 2000) In the case of thesuevite in Yax-1 core nannofossil assemblage of the majorpart of units 1 and 2 are of late Campanian to earlyMaastrichtian age while the only latest Maastrichtiannannofossil is observed in the basal part of unit 2 Thus themost probable mechanism of erosion and transportation of thenannofossils in the unconsolidated carbonate sediments of theinner rim margin andor outside the crater is via ocean waterinvasion

If carbonate sediments of Campanian to Maastrichtianage outside the crater were eroded by the surge of ocean waterinto the crater the sedimentary sequence outside the cratershould have a hiatus corresponding to this age rangeAlternatively collapse of the rim wall and ejecta curtaindeposits on the rim by the ocean water invasion or the impactseismic wave also could have supplied sediment particleswith nannofossils of Campanian to Maastrichtian age into the

Fig 6 a) A diagram showing vertical variations of maximum size of limestone lithics (mm) and greenish vesicular melt fragments The arrowsof dashed and solid lines indicate normal or inverse grading of limestone lithics and greenish melt fragments in each cycle respectively Theshades indicate intervals of high calcium content in each cycle b) diagrams showing vertical variations of chemical composition of the suevitein unit 1 (wt) determined by XGT

1242 K Goto et al

crater In this case the age range of nannofossils in the craterrim and ejecta curtain deposits outside the rim should bedominated by Campanian to Maastrichtian age However noinformation is available on the age of the impactite outside thecrater Further detailed research of the impactite outside thecrater is needed to determine the source of carbonatesediments of Campanian to Maastrichtian age in units 2 and 1

Unit 2 is composed of two normally graded beds (Fig 3)Irregular erosional basal contact poor sorting clast-supportedfabric and the occurrence of intraclast-like melt fragments atthe base of the lower bed suggest that this unit was depositedby gravity flows (eg Middleton and Hampton 1976) Thepresence of nannofossils probably derived from outside thecrater indicates that gravity flows were formed in associationwith ocean water invasion Alternatively if the collapse of therim wall and ejecta curtain deposits accumulated on the rimwere the sources of nannofossils impact seismic wave mayhave triggered gravity flows In this case gravity flows mayhave run down on the rim surface in association with theocean water invasion or under the hotdry conditions withoutinfluence of water invasion

The suevite in unit 1 is well-sorted and shows upwardfining of limestone lithics in each cycle This indicates thatthe suevite in each cycle in unit 1 was deposited as a result ofsettlement of re-suspended sediment particles Furthermorethere is no erosional surface at the base of each cyclesuggesting that coarse sediment particles which probablywere derived from outside the crater andor crater rimprobably have been transported in suspension by water massmovement Therefore repeated ocean water invasion is themost probable mechanism for the cyclic sedimentation of thesuevite in unit 1 Repeated forceful ocean water invasion isprobably required to transport mm-size grains in each cycle assuspension

Similar cycles are also reported from the impactite in theChesapeake Bay crater (Poag et al 2002) The ChesapeakeBay crater is 85 km in diameter and is buried 300ndash500 mbelow the floor of the southern part of the Chesapeake Bay(Poag et al 1994 2002) Poag et al (2002) mentioned that theupper part of the impactite in the Chesapeake Bay craterconsists of repeating upward fining beds They interpretedthese cycles to have been formed by repeated ocean water

Fig 7 a) A calcium distribution map and b) a rock surface photograph of higher calcium content sample from 80231 m in unit 1 Highercalcium sample shows matrix-supported fabric c) a calcium distribution map and d) a rock surface photograph of lower calcium contentsample from 80196 m of unit 1 The lower calcium sample shows clast-supported fabric

Evidence for ocean water invasion into the Chicxulub crater 1243

invasion based on facies analysis of cores and downholegeophysical logs Similarity in cyclic sedimentation pattern inthe upper part of the impactite both in the Chicxulub craterand the Chesapeake Bay crater supports our interpretationthat unit 1 in Yax-1 was probably formed by lateral supply ofcoarse sediments through repeated ocean water invasion intothe crater cavity

The thicknesses of cycles are 7ndash8 m in unit 2 and 1ndash2 min unit 1 at Yax-1 These thicknesses seem thin for repeatedforceful ocean water invasion (= resurge) when comparedwith the thicknesses of the resurge deposits in other marinetarget craters However this could be owing to the position ofYax-1 which is located near the edge of the crater rim and

well above the crater floor In fact the suevite reaches thethickness of ~300 m at C1 S1 and Y6 cores (Sharpton et al1996 Claeys et al 2003) that are located in the central part ofthe crater (Fig 1)

The presence of parallel and mono-directional crosslaminations in the uppermost part of unit 1 and the lower partof unit 0 suggests that the crater was already filled with waterand that sedimentation of this interval was under the influenceof currents Cross lamination is one of the typical charactersof the resurge deposits and is also found in other marine targetimpact craters such as the Lockne crater in Sweden (Dalwigkand Ormouml 2001) which is ~135 km in diameter and wasformed on the ocean floor in water over 200 m deep (Ormouml

Fig 8 A core photograph of the uppermost part of unit 1 unit 0 and the overlying Paleocene carbonate (79397ndash79463 m) Nannofossil andplanktonic foraminiferal assemblage are also shown The left-side photo is a photo of a piece taken by UNAM using an unrolled core scan(360deg) The right-side photos show half cut surfaces of the uppermost part of unit 1 and the basal part of unit 0 Parallel and mono-directionalcross laminations are observed in these intervals GS = greenish suevitic bed

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

1240 K Goto et al

The first appearance of Danian planktonic foraminifera(P eugubina) is just above the black clay layer (Fig 8according to Keller et al 2003) Smit et al (2003) reportedthat Paleocene fauna corresponding to G eugubina zoneappears 5 cm above the black clay layer (Fig 8 79407 m)The sample from 79394 m contains abundant biserialheterohelicids and globigerine specimens larger than 03 mmin size and a single specimen of Globigerinelloides-likeplanispiral genus These planktonic foraminifers suggestCretaceous age and are possibly reworked Very thin-walledindividuals are preserved among thick-walled specimensindicating that dissolution has not substantially modified thedepositional assemblage The sample from 79394 m alsocontains nannofossils Thoracosphaera and Braarudosphaerathat occur right after the impact In addition very smallspecimens of Cruciplacolithus primus are also observed inthis sample (Fig 8) suggesting an age younger than 648 Ma(C29r P1a) according to Berggren et al (1995)

DISCUSSION

The Possibility of Ocean Water Invasion and SedimentaryProcess of Units 2 to 1

Mixed nature of reworked microfossils including latestMaastrichtian age lithic fragments and impact derivedmaterials such as shocked quartz grains with PDFs in thesuevite of units 2 and 1 is similar to the KT boundaryldquococktailrdquo deposits in the Gulf of Mexico (Bralower et al1998) suggesting that units 2 and 1 in Yax-1 core wererelated to the KT boundary impact and were formed as aresult of reworking

Although the KT boundary sequence in Yax-1 core maynot be stratigraphically complete sedimentological dataprovide compelling evidence for events immediately after theimpact Reworked nannofossils of late Campanian to earlyMaastrichtian age are abundant in the matrix of units 2 and 1

Table 1 Bulk chemical composition of the suevite in unit 1 (wt) determined by XGTSamplenumber Cycle

Sub-bottomdepth (m)

Al2O3(wt)

SiO2(wt)

CaO(wt)

TiO2(wt)

Fe2O3(wt)

1277 8 79558 1018 3733 1854 044 459 1278 8 79587 1226 4454 949 054 564 1279 8 79618 1283 4681 1071 053 558 1280 8 79647 1106 4067 932 051 516 1281 7 79676 1061 3903 2050 046 439 1126 7 79752 1145 4019 1115 050 518 1127 7 79783 1182 4936 683 061 617 1128 7 79815 1296 4946 761 059 578 1129 6 79845 1113 4089 1550 049 496 1130 6 79874 1169 4540 859 056 562 1131 6 79896 1079 4358 1022 053 575 1133 5 79932 1172 4047 1657 046 492 1134 5 79962 1138 4356 1079 052 533 1135 5 79992 1172 4900 736 056 600 1137 5 80055 1209 4420 677 053 569 1138 5 80085 1162 4163 1081 051 523 1139 4 80115 941 3368 2029 041 423 1140 4 80134 1212 4639 817 053 554 1141 4 80161 1249 4715 666 058 594 1142 4 80196 1271 4798 678 057 588 1143 3 80231 1082 3519 2217 045 403 1145 3 80308 1072 3871 1779 052 440 1146 3 80338 1316 4446 1198 055 509 1147 3 80368 1121 4308 1354 051 521 1148 3 80398 1098 4521 870 056 567 1149 3 80428 1292 4919 782 059 587 1150 2 80458 1055 3704 1921 046 487 1151 2 80528 1176 4391 1170 052 548 1152 2 80547 1221 4641 787 056 610 1153 2 80577 1138 4456 897 052 550 1154 1 80607 1070 3803 2126 047 459 1155 1 80637 1049 3889 1692 045 449 1156 1 80667 1037 3768 1516 047 450 1157 1 80697 1222 4609 854 054 568 1158 1 80727 1235 4711 842 057 591

Evidence for ocean water invasion into the Chicxulub crater 1241

It is important to note that this age range is considerablynarrower than the range of the sediments excavated by thecrater formation This suggests that these nannofossils wereselectively incorporated in units 2 and 1 There is no lateCampanian to early Maastrichtian nannofossils in carbonatemega-blocks underneath the impactite in Yax-1 suggestingthat these nannofossils were not derived from underlyingcarbonate mega-blocks These nannofossils (also thelimestone lithics) cannot have been derived from the ejectaplume because they show no signs of recrystallization ormelting Therefore nannofossils (also the limestone lithics) inunits 1 and 2 were probably derived from unconsolidatedcarbonate sediments deposited outside the crater within theejecta curtain deposits or from the rim wall

Study of the Tvaumlren crater in Sweden which is 2 km indiameter and was formed on the floor of an ocean 100ndash150 mdeep (Ormouml and Lindstroumlm 2000) has documented resurgedeposits in the upper part of the impactite formed by the oceanwater invasion immediately after the impact (eg Ormouml andLindstroumlm 2000) These resurge deposits tend to show thereversed chronology (Ormouml and Lindstroumlm 2000) Namelythe lower part of the deposit consists of the particles derived

from younger sediments and the upper part consists of theparticles derived from older sediments (Ormouml 1994 Ormoumland Lindstroumlm 2000) This is because ocean floor sedimentsoutside the crater were eroded by currents and transportedinto the crater beginning with younger units and progressingto older units (Ormouml and Lindstroumlm 2000) In the case of thesuevite in Yax-1 core nannofossil assemblage of the majorpart of units 1 and 2 are of late Campanian to earlyMaastrichtian age while the only latest Maastrichtiannannofossil is observed in the basal part of unit 2 Thus themost probable mechanism of erosion and transportation of thenannofossils in the unconsolidated carbonate sediments of theinner rim margin andor outside the crater is via ocean waterinvasion

If carbonate sediments of Campanian to Maastrichtianage outside the crater were eroded by the surge of ocean waterinto the crater the sedimentary sequence outside the cratershould have a hiatus corresponding to this age rangeAlternatively collapse of the rim wall and ejecta curtaindeposits on the rim by the ocean water invasion or the impactseismic wave also could have supplied sediment particleswith nannofossils of Campanian to Maastrichtian age into the

Fig 6 a) A diagram showing vertical variations of maximum size of limestone lithics (mm) and greenish vesicular melt fragments The arrowsof dashed and solid lines indicate normal or inverse grading of limestone lithics and greenish melt fragments in each cycle respectively Theshades indicate intervals of high calcium content in each cycle b) diagrams showing vertical variations of chemical composition of the suevitein unit 1 (wt) determined by XGT

1242 K Goto et al

crater In this case the age range of nannofossils in the craterrim and ejecta curtain deposits outside the rim should bedominated by Campanian to Maastrichtian age However noinformation is available on the age of the impactite outside thecrater Further detailed research of the impactite outside thecrater is needed to determine the source of carbonatesediments of Campanian to Maastrichtian age in units 2 and 1

Unit 2 is composed of two normally graded beds (Fig 3)Irregular erosional basal contact poor sorting clast-supportedfabric and the occurrence of intraclast-like melt fragments atthe base of the lower bed suggest that this unit was depositedby gravity flows (eg Middleton and Hampton 1976) Thepresence of nannofossils probably derived from outside thecrater indicates that gravity flows were formed in associationwith ocean water invasion Alternatively if the collapse of therim wall and ejecta curtain deposits accumulated on the rimwere the sources of nannofossils impact seismic wave mayhave triggered gravity flows In this case gravity flows mayhave run down on the rim surface in association with theocean water invasion or under the hotdry conditions withoutinfluence of water invasion

The suevite in unit 1 is well-sorted and shows upwardfining of limestone lithics in each cycle This indicates thatthe suevite in each cycle in unit 1 was deposited as a result ofsettlement of re-suspended sediment particles Furthermorethere is no erosional surface at the base of each cyclesuggesting that coarse sediment particles which probablywere derived from outside the crater andor crater rimprobably have been transported in suspension by water massmovement Therefore repeated ocean water invasion is themost probable mechanism for the cyclic sedimentation of thesuevite in unit 1 Repeated forceful ocean water invasion isprobably required to transport mm-size grains in each cycle assuspension

Similar cycles are also reported from the impactite in theChesapeake Bay crater (Poag et al 2002) The ChesapeakeBay crater is 85 km in diameter and is buried 300ndash500 mbelow the floor of the southern part of the Chesapeake Bay(Poag et al 1994 2002) Poag et al (2002) mentioned that theupper part of the impactite in the Chesapeake Bay craterconsists of repeating upward fining beds They interpretedthese cycles to have been formed by repeated ocean water

Fig 7 a) A calcium distribution map and b) a rock surface photograph of higher calcium content sample from 80231 m in unit 1 Highercalcium sample shows matrix-supported fabric c) a calcium distribution map and d) a rock surface photograph of lower calcium contentsample from 80196 m of unit 1 The lower calcium sample shows clast-supported fabric

Evidence for ocean water invasion into the Chicxulub crater 1243

invasion based on facies analysis of cores and downholegeophysical logs Similarity in cyclic sedimentation pattern inthe upper part of the impactite both in the Chicxulub craterand the Chesapeake Bay crater supports our interpretationthat unit 1 in Yax-1 was probably formed by lateral supply ofcoarse sediments through repeated ocean water invasion intothe crater cavity

The thicknesses of cycles are 7ndash8 m in unit 2 and 1ndash2 min unit 1 at Yax-1 These thicknesses seem thin for repeatedforceful ocean water invasion (= resurge) when comparedwith the thicknesses of the resurge deposits in other marinetarget craters However this could be owing to the position ofYax-1 which is located near the edge of the crater rim and

well above the crater floor In fact the suevite reaches thethickness of ~300 m at C1 S1 and Y6 cores (Sharpton et al1996 Claeys et al 2003) that are located in the central part ofthe crater (Fig 1)

The presence of parallel and mono-directional crosslaminations in the uppermost part of unit 1 and the lower partof unit 0 suggests that the crater was already filled with waterand that sedimentation of this interval was under the influenceof currents Cross lamination is one of the typical charactersof the resurge deposits and is also found in other marine targetimpact craters such as the Lockne crater in Sweden (Dalwigkand Ormouml 2001) which is ~135 km in diameter and wasformed on the ocean floor in water over 200 m deep (Ormouml

Fig 8 A core photograph of the uppermost part of unit 1 unit 0 and the overlying Paleocene carbonate (79397ndash79463 m) Nannofossil andplanktonic foraminiferal assemblage are also shown The left-side photo is a photo of a piece taken by UNAM using an unrolled core scan(360deg) The right-side photos show half cut surfaces of the uppermost part of unit 1 and the basal part of unit 0 Parallel and mono-directionalcross laminations are observed in these intervals GS = greenish suevitic bed

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

Evidence for ocean water invasion into the Chicxulub crater 1241

It is important to note that this age range is considerablynarrower than the range of the sediments excavated by thecrater formation This suggests that these nannofossils wereselectively incorporated in units 2 and 1 There is no lateCampanian to early Maastrichtian nannofossils in carbonatemega-blocks underneath the impactite in Yax-1 suggestingthat these nannofossils were not derived from underlyingcarbonate mega-blocks These nannofossils (also thelimestone lithics) cannot have been derived from the ejectaplume because they show no signs of recrystallization ormelting Therefore nannofossils (also the limestone lithics) inunits 1 and 2 were probably derived from unconsolidatedcarbonate sediments deposited outside the crater within theejecta curtain deposits or from the rim wall

Study of the Tvaumlren crater in Sweden which is 2 km indiameter and was formed on the floor of an ocean 100ndash150 mdeep (Ormouml and Lindstroumlm 2000) has documented resurgedeposits in the upper part of the impactite formed by the oceanwater invasion immediately after the impact (eg Ormouml andLindstroumlm 2000) These resurge deposits tend to show thereversed chronology (Ormouml and Lindstroumlm 2000) Namelythe lower part of the deposit consists of the particles derived

from younger sediments and the upper part consists of theparticles derived from older sediments (Ormouml 1994 Ormoumland Lindstroumlm 2000) This is because ocean floor sedimentsoutside the crater were eroded by currents and transportedinto the crater beginning with younger units and progressingto older units (Ormouml and Lindstroumlm 2000) In the case of thesuevite in Yax-1 core nannofossil assemblage of the majorpart of units 1 and 2 are of late Campanian to earlyMaastrichtian age while the only latest Maastrichtiannannofossil is observed in the basal part of unit 2 Thus themost probable mechanism of erosion and transportation of thenannofossils in the unconsolidated carbonate sediments of theinner rim margin andor outside the crater is via ocean waterinvasion

If carbonate sediments of Campanian to Maastrichtianage outside the crater were eroded by the surge of ocean waterinto the crater the sedimentary sequence outside the cratershould have a hiatus corresponding to this age rangeAlternatively collapse of the rim wall and ejecta curtaindeposits on the rim by the ocean water invasion or the impactseismic wave also could have supplied sediment particleswith nannofossils of Campanian to Maastrichtian age into the

Fig 6 a) A diagram showing vertical variations of maximum size of limestone lithics (mm) and greenish vesicular melt fragments The arrowsof dashed and solid lines indicate normal or inverse grading of limestone lithics and greenish melt fragments in each cycle respectively Theshades indicate intervals of high calcium content in each cycle b) diagrams showing vertical variations of chemical composition of the suevitein unit 1 (wt) determined by XGT

1242 K Goto et al

crater In this case the age range of nannofossils in the craterrim and ejecta curtain deposits outside the rim should bedominated by Campanian to Maastrichtian age However noinformation is available on the age of the impactite outside thecrater Further detailed research of the impactite outside thecrater is needed to determine the source of carbonatesediments of Campanian to Maastrichtian age in units 2 and 1

Unit 2 is composed of two normally graded beds (Fig 3)Irregular erosional basal contact poor sorting clast-supportedfabric and the occurrence of intraclast-like melt fragments atthe base of the lower bed suggest that this unit was depositedby gravity flows (eg Middleton and Hampton 1976) Thepresence of nannofossils probably derived from outside thecrater indicates that gravity flows were formed in associationwith ocean water invasion Alternatively if the collapse of therim wall and ejecta curtain deposits accumulated on the rimwere the sources of nannofossils impact seismic wave mayhave triggered gravity flows In this case gravity flows mayhave run down on the rim surface in association with theocean water invasion or under the hotdry conditions withoutinfluence of water invasion

The suevite in unit 1 is well-sorted and shows upwardfining of limestone lithics in each cycle This indicates thatthe suevite in each cycle in unit 1 was deposited as a result ofsettlement of re-suspended sediment particles Furthermorethere is no erosional surface at the base of each cyclesuggesting that coarse sediment particles which probablywere derived from outside the crater andor crater rimprobably have been transported in suspension by water massmovement Therefore repeated ocean water invasion is themost probable mechanism for the cyclic sedimentation of thesuevite in unit 1 Repeated forceful ocean water invasion isprobably required to transport mm-size grains in each cycle assuspension

Similar cycles are also reported from the impactite in theChesapeake Bay crater (Poag et al 2002) The ChesapeakeBay crater is 85 km in diameter and is buried 300ndash500 mbelow the floor of the southern part of the Chesapeake Bay(Poag et al 1994 2002) Poag et al (2002) mentioned that theupper part of the impactite in the Chesapeake Bay craterconsists of repeating upward fining beds They interpretedthese cycles to have been formed by repeated ocean water

Fig 7 a) A calcium distribution map and b) a rock surface photograph of higher calcium content sample from 80231 m in unit 1 Highercalcium sample shows matrix-supported fabric c) a calcium distribution map and d) a rock surface photograph of lower calcium contentsample from 80196 m of unit 1 The lower calcium sample shows clast-supported fabric

Evidence for ocean water invasion into the Chicxulub crater 1243

invasion based on facies analysis of cores and downholegeophysical logs Similarity in cyclic sedimentation pattern inthe upper part of the impactite both in the Chicxulub craterand the Chesapeake Bay crater supports our interpretationthat unit 1 in Yax-1 was probably formed by lateral supply ofcoarse sediments through repeated ocean water invasion intothe crater cavity

The thicknesses of cycles are 7ndash8 m in unit 2 and 1ndash2 min unit 1 at Yax-1 These thicknesses seem thin for repeatedforceful ocean water invasion (= resurge) when comparedwith the thicknesses of the resurge deposits in other marinetarget craters However this could be owing to the position ofYax-1 which is located near the edge of the crater rim and

well above the crater floor In fact the suevite reaches thethickness of ~300 m at C1 S1 and Y6 cores (Sharpton et al1996 Claeys et al 2003) that are located in the central part ofthe crater (Fig 1)

The presence of parallel and mono-directional crosslaminations in the uppermost part of unit 1 and the lower partof unit 0 suggests that the crater was already filled with waterand that sedimentation of this interval was under the influenceof currents Cross lamination is one of the typical charactersof the resurge deposits and is also found in other marine targetimpact craters such as the Lockne crater in Sweden (Dalwigkand Ormouml 2001) which is ~135 km in diameter and wasformed on the ocean floor in water over 200 m deep (Ormouml

Fig 8 A core photograph of the uppermost part of unit 1 unit 0 and the overlying Paleocene carbonate (79397ndash79463 m) Nannofossil andplanktonic foraminiferal assemblage are also shown The left-side photo is a photo of a piece taken by UNAM using an unrolled core scan(360deg) The right-side photos show half cut surfaces of the uppermost part of unit 1 and the basal part of unit 0 Parallel and mono-directionalcross laminations are observed in these intervals GS = greenish suevitic bed

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

1242 K Goto et al

crater In this case the age range of nannofossils in the craterrim and ejecta curtain deposits outside the rim should bedominated by Campanian to Maastrichtian age However noinformation is available on the age of the impactite outside thecrater Further detailed research of the impactite outside thecrater is needed to determine the source of carbonatesediments of Campanian to Maastrichtian age in units 2 and 1

Unit 2 is composed of two normally graded beds (Fig 3)Irregular erosional basal contact poor sorting clast-supportedfabric and the occurrence of intraclast-like melt fragments atthe base of the lower bed suggest that this unit was depositedby gravity flows (eg Middleton and Hampton 1976) Thepresence of nannofossils probably derived from outside thecrater indicates that gravity flows were formed in associationwith ocean water invasion Alternatively if the collapse of therim wall and ejecta curtain deposits accumulated on the rimwere the sources of nannofossils impact seismic wave mayhave triggered gravity flows In this case gravity flows mayhave run down on the rim surface in association with theocean water invasion or under the hotdry conditions withoutinfluence of water invasion

The suevite in unit 1 is well-sorted and shows upwardfining of limestone lithics in each cycle This indicates thatthe suevite in each cycle in unit 1 was deposited as a result ofsettlement of re-suspended sediment particles Furthermorethere is no erosional surface at the base of each cyclesuggesting that coarse sediment particles which probablywere derived from outside the crater andor crater rimprobably have been transported in suspension by water massmovement Therefore repeated ocean water invasion is themost probable mechanism for the cyclic sedimentation of thesuevite in unit 1 Repeated forceful ocean water invasion isprobably required to transport mm-size grains in each cycle assuspension

Similar cycles are also reported from the impactite in theChesapeake Bay crater (Poag et al 2002) The ChesapeakeBay crater is 85 km in diameter and is buried 300ndash500 mbelow the floor of the southern part of the Chesapeake Bay(Poag et al 1994 2002) Poag et al (2002) mentioned that theupper part of the impactite in the Chesapeake Bay craterconsists of repeating upward fining beds They interpretedthese cycles to have been formed by repeated ocean water

Fig 7 a) A calcium distribution map and b) a rock surface photograph of higher calcium content sample from 80231 m in unit 1 Highercalcium sample shows matrix-supported fabric c) a calcium distribution map and d) a rock surface photograph of lower calcium contentsample from 80196 m of unit 1 The lower calcium sample shows clast-supported fabric

Evidence for ocean water invasion into the Chicxulub crater 1243

invasion based on facies analysis of cores and downholegeophysical logs Similarity in cyclic sedimentation pattern inthe upper part of the impactite both in the Chicxulub craterand the Chesapeake Bay crater supports our interpretationthat unit 1 in Yax-1 was probably formed by lateral supply ofcoarse sediments through repeated ocean water invasion intothe crater cavity

The thicknesses of cycles are 7ndash8 m in unit 2 and 1ndash2 min unit 1 at Yax-1 These thicknesses seem thin for repeatedforceful ocean water invasion (= resurge) when comparedwith the thicknesses of the resurge deposits in other marinetarget craters However this could be owing to the position ofYax-1 which is located near the edge of the crater rim and

well above the crater floor In fact the suevite reaches thethickness of ~300 m at C1 S1 and Y6 cores (Sharpton et al1996 Claeys et al 2003) that are located in the central part ofthe crater (Fig 1)

The presence of parallel and mono-directional crosslaminations in the uppermost part of unit 1 and the lower partof unit 0 suggests that the crater was already filled with waterand that sedimentation of this interval was under the influenceof currents Cross lamination is one of the typical charactersof the resurge deposits and is also found in other marine targetimpact craters such as the Lockne crater in Sweden (Dalwigkand Ormouml 2001) which is ~135 km in diameter and wasformed on the ocean floor in water over 200 m deep (Ormouml

Fig 8 A core photograph of the uppermost part of unit 1 unit 0 and the overlying Paleocene carbonate (79397ndash79463 m) Nannofossil andplanktonic foraminiferal assemblage are also shown The left-side photo is a photo of a piece taken by UNAM using an unrolled core scan(360deg) The right-side photos show half cut surfaces of the uppermost part of unit 1 and the basal part of unit 0 Parallel and mono-directionalcross laminations are observed in these intervals GS = greenish suevitic bed

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

Evidence for ocean water invasion into the Chicxulub crater 1243

invasion based on facies analysis of cores and downholegeophysical logs Similarity in cyclic sedimentation pattern inthe upper part of the impactite both in the Chicxulub craterand the Chesapeake Bay crater supports our interpretationthat unit 1 in Yax-1 was probably formed by lateral supply ofcoarse sediments through repeated ocean water invasion intothe crater cavity

The thicknesses of cycles are 7ndash8 m in unit 2 and 1ndash2 min unit 1 at Yax-1 These thicknesses seem thin for repeatedforceful ocean water invasion (= resurge) when comparedwith the thicknesses of the resurge deposits in other marinetarget craters However this could be owing to the position ofYax-1 which is located near the edge of the crater rim and

well above the crater floor In fact the suevite reaches thethickness of ~300 m at C1 S1 and Y6 cores (Sharpton et al1996 Claeys et al 2003) that are located in the central part ofthe crater (Fig 1)

The presence of parallel and mono-directional crosslaminations in the uppermost part of unit 1 and the lower partof unit 0 suggests that the crater was already filled with waterand that sedimentation of this interval was under the influenceof currents Cross lamination is one of the typical charactersof the resurge deposits and is also found in other marine targetimpact craters such as the Lockne crater in Sweden (Dalwigkand Ormouml 2001) which is ~135 km in diameter and wasformed on the ocean floor in water over 200 m deep (Ormouml

Fig 8 A core photograph of the uppermost part of unit 1 unit 0 and the overlying Paleocene carbonate (79397ndash79463 m) Nannofossil andplanktonic foraminiferal assemblage are also shown The left-side photo is a photo of a piece taken by UNAM using an unrolled core scan(360deg) The right-side photos show half cut surfaces of the uppermost part of unit 1 and the basal part of unit 0 Parallel and mono-directionalcross laminations are observed in these intervals GS = greenish suevitic bed

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

1244 K Goto et al

and Lindstroumlm 2000) According to Ormouml and Lindstroumlm(2000) the lower part of the resurge deposits in the resurgegullies of the Lockne crater is matrix-supported and is similarto debris flow deposit The upper part is clast-supported andhas a relatively high proportion of crystalline ejecta clasts(Ormouml and Lindstroumlm 2000) The uppermost fine-grained partof the resurge deposit displays current lineation cross-bedding and dewatering structures (Dalwigk and Ormouml2001) Upward fining character of the impactite and thepresence of cross-bedding in the uppermost part of the resurgedeposits of the Lockne crater is similar to those of units 2 to 0in Yax-1 implying that units 2 to 0 in Yax-1 were probablyformed by ocean water invasion similar to the resurgedeposits in the Lockne crater

Cross-lamination represents a lower energy flow regime(eg Middleton and Southard 1977) but its occurrence in theuppermost part of the impactite could suggest depositionduring the waning stages of current flow Similar cross-lamination is reported from the top of the impactite in UNAM5 (U5) core (Smit 1999) which is located approximately110 km to the south of the center of the Chicxulub craterprobably on the outside slope of the crater (Fig 1) Thissuggests that the influence of currents was not local butprevailed both inside and outside the crater

The Timing of Ocean Water Invasion

The maximum size of limestone lithics in unit 1 shows anumber of repeated depositional units with normal gradingwhile the maximum size of greenish vesicular melt fragmentsshows inverse plus normal grading in cycles 1ndash3 and normalgrading in cycles 4ndash8 (Fig 6a)

Large intact porous fragments such as pumice commonlyhave low bulk density and are likely deposited slowly throughthe atmosphere and water column while small pumiceprobably sink rapidly because most bubble walls are broken orcracked (eg Fisher and Schmincke 1984) Inverse grading ofpumice fragments results from the time lag between thedeposition of the large and lower density and small and higherdensity pumice fragments (eg Fisher and Schmincke 1984)Therefore formation of inverse grading of greenish vesicularmelt fragments in cycles 1ndash3 can be explained by a lag ofsedimentation between the larger and smaller melt fragmentsdue to the differences in their bulk densities On the otherhand normal grading of greenish vesicular melt fragments incycles 4ndash8 likely occurred after water infiltrated into the poresand increased their bulk densities

Infiltration of water into vesicular grains typically occurswithin minutes to hours of their contact with water althoughinfiltration time of water is dependent on the chemicalcomposition and the size of the grain (eg Fisher andSchmincke 1984) Greenish vesicular melt fragmentsobserved in units 2 and 1 is lt8 mm If we assume that thegreenish vesicular melt fragments were ejected to the upper

layer of the atmosphere by the impact and then settled throughthe atmosphere and the ocean water the greenish vesicularmelt fragments probably reached the ocean surface within anhour of the impact Therefore sedimentation of cycles 1ndash3may have started within several hours after the impact

Partial or complete collapse of the crater rim is requiredfor ocean water to enter the crater cavity immediately after theimpact The water depth of the impact site is estimated as lessthan 100 m at the time of the impact (Sharpton et al 1996Pierazzo and Crawford 1997 Pierazzo et al 1998) On theother hand the transient crater rim was estimated to beseveral km in height at the KT boundary impact (egMorgan et al 2000) Consequently if the transient crater rimhad not collapsed immediately after the impact it should haveprevented ocean water from invasion into the crater

However the transient crater rim is normally veryunstable especially in the case of large crater such as theChicxulub crater and thus likely collapsed immediately afterthe crater formation (eg Morgan et al 1997 2000 Collins etal 2002) Cretaceous mega-blocks (evidence for rimcollapse) are reported in Yax-1 (Kenkmann et al 2003Wittmann et al 2003) Furthermore a vertical offset of atleast 2 km is indicative of the collapse of the crater rim and ahigh elevation of the crater rim was not observed in theseismic reflection data (eg Morgan et al 1997) Accordingto numerical simulations the transient crater rim could havebeen temporarily elevated several km from the pre-impactsurface but it could have collapsed into the crater withinminutes after the impact (Morgan et al 2000 Collins et al2002) Furthermore the final crater shape of their numericalsimulation is similar to the crater shape of the Chicxulubcrater and a high elevation of the final crater rim was notobserved (Morgan et al 2000 Collins et al 2002)

In this way drilling core data seismic reflection dataand numerical simulation data suggest that the rim of theChicxulub crater may have partially or entirely collapsedimmediately after the impact Therefore ocean water couldhave surged into the Chicxulub crater immediately after theimpact in spite of the shallow water depth of the site beforethe impact

Implication for the Tsunami Generation Mechanism

Two different generation mechanisms have beenproposed for the giant tsunami of the KT boundary (egMatsui et al 2002) one is a tsunami generated by landslideson the slope of the Yucataacuten Platform triggered by the shockwave from the impact point (hereafter we call it landslide-generated tsunami) and the other is a tsunami generated byocean water invasion into the crater and subsequent overflowof water (hereafter we call it crater-generated tsunami)

Large-scale erosion of the Cretaceous sediments causedby the KT boundary impact (Bralower et al 1998 Tada et al2002) and over 250 m-thick KT boundary deposits of

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

Evidence for ocean water invasion into the Chicxulub crater 1245

probable gravity flow origin have been documented on thecontinental slope around the Yucataacuten Peninsula (Grajales-Nishimura et al 2000 Kiyokawa et al 2002) These large-scale gravity flows could have caused landslide-generatedtsunami which in turn could have affected the sedimentationof the KT boundary deposits around the proximal impact site(Bralower et al 1998 Kiyokawa et al 2002 Matsui et al 2002)

On the other hand the results of this study also suggestthat ocean water invasion into the crater probably occurredimmediately after the impact Cyclic sedimentation of unit 1implies repeated water mass movements within the crater thatwere forceful enough to carry mm-size grains as suspensionSuch forceful water mass movements could have been causedby repeated ocean water invasion and may have caused crater-generated tsunami Therefore the crater-generated tsunamipossibly could have also been another generation mechanismof tsunami at the KT boundary Numerical simulation of theKT boundary tsunami by Matsui et al (2002) suggests thatthe wave height of crater-generated tsunami could have beenseveral tens of meters around the coastal area of the Gulf ofMexico even if the water depth of the Yucataacuten platform werevery shallow The tsunami of this magnitude has enoughpower to cause suspension of sandy particles on the oceanbottom (eg Kastens and Cita 1981 Bryant 2001) Thereforecrater-generated tsunami may also have affected thesedimentation of the KT boundary deposits

However we need more information to evaluate theinfluence of ocean water invasion into the Chicxulub crater onthe ocean environment outside the crater The resultspresented here are based on the observations at a single butthe only available core Furthermore number direction andmagnitude of crater-generated tsunami is dependent onvarious parameters such as the target water depth and thescale and range of the crater rim collapse (Matsui et al 2002)Therefore further core drilling inside and outside the craterfor comparison with the results of Yax-1 and numericalcalculations based on stronger geological constraints arerequired to discuss the number direction and magnitude ofcrater-generated tsunami and their influence on thesedimentation of the KT boundary deposits The drilling intothe Chicxulub crater planned by the Integrated Ocean DrillingProgram (IODP) will enable us to discuss the role of oceanwater invasion into the crater just after the KT boundaryimpact

SUMMARY

Lithology grain composition chemical compositiongrain size and nannofossil assemblage were examined in theYax-1 core to investigate the possibility of the ocean waterinvasion into the Chicxulub crater immediately after theimpact An impactite occurs within the interval betweenapproximately 79463 and 89494 m depth and is divided into6 units in descending order In this study we mainly focused

on the sedimentary process of units 1 and 2 because this partis deposited in the uppermost part of the impactite andprobably was deposited during the latest stage of theformation of the Chicxulub crater

Unit 2 is composed of highly heterogeneous suevite withabundant melt and lithic fragments and contains two normallygraded beds On the other hand unit 1 is composed ofrelatively well-sorted suevite and contains at least eight 1ndash2m-scale normally andor inversely graded beds of the suevite

Abundant nannofossils of late Campanian to earlyMaastrichtian age in the matrix of units 2 and 1 suggests thatcarbonate sediments deposited in the inner rim margin oraround the crater were selectively eroded and transported intothe crater most likely by ocean water invasion into the craterThe occurrence of parallel and mono-directional crosslamination in the uppermost part of unit 1 and the lower partof unit 0 suggests the influence of current at least during thedeposition of this interval Given the infiltration time of waterinto the greenish vesicular melt fragments in unit 1 oceanwater invasion probably started immediately after the impactThis is supported by the presence of vertical offset and theabsence of a high crater rim suggesting the collapse of thecrater rim immediately after the impact

Irregular erosional basal contact normal grading poorsorting clast-supported fabric and the occurrence ofintraclast-like melt fragments at the base of the suevite in unit2 suggest that this sub-unit was formed by gravity flows thatwere probably triggered by ocean water invasion or an impactseismic wave On the other hand the upward fining grain sizeand the absence of erosional contact of each cycle in unit 1probably indicate that each cycle has been deposited from re-suspended sediment particles transported as suspension bywater mass-movement in association with ocean waterinvasion

AcknowledgmentsndashThis research is based on samples and dataprovided by the Chicxulub Scientific Drilling Program(CSDP) sponsored by International Continental ScientificDrilling Program (ICDP) and Universidad NacionalAutoacutenoma de Meacutexico (UNAM) We wish to thank J UFucugauchi J Smit and A M Soler Arechalde for theirsupport during sampling We also wish to thank staffs of theJapanese ICDP office for their support during submission ofour proposals We also thank P Claeys H Dypvik J OrmoumlD King F Imamura and F Masuda for their critical readingof the manuscript and for many valuable suggestions Thisresearch was partly funded by the Japanese Society for thePromotion of Science (JSPS) Fellowship provided to K Goto

Editorial HandlingmdashDr Philippe Claeys

REFERENCES

Berggren W A Kent D V Swisher C C and Aubry M P 1995 Arevised Cenozoic geochronology and chronostratigraphy In

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

1246 K Goto et al

Geochronology time scales and global stratigraphiccorrelation edited by Berggren W A Kent D V Aubry M Pand Hardenbol J Special Publication 54 Society of EconomicPaleontologists and Mineralogists pp 129ndash212

Bourgeois J Hansen T A Wiberg P L and Kauffman E G 1988A tsunami deposit at the Cretaceous-Tertiary boundary in TexasScience 241567ndash570

Bralower T J Paul C K and Leckie R M 1998 The CretaceousTertiary boundary cocktail Chicxulub impact triggers margincollapse and extensive sediment gravity flows Geology 26 331ndash334

Bryant E 2001 Tsunami The underrated hazard CambridgeCambridge University Press 320 p

Claeys P Heuschkel S Lounejeva-Baturina E Sanchez-Rubio Gand Stoumlffler D 2003 The suevite of drill hole Yucataacuten 6 in theChicxulub impact crater Meteoritics amp Planetary Science 381299ndash1317

Collins G S Melosh H J Morgan J V and Warner M R 2002Hydrocode simulations of Chicxulub crater collapse and peak-ring formation Icarus 15724ndash33

Dalwigk I and Ormouml J 2001 Formation of resurge gullies at impactsat sea The Lockne crater Sweden Meteoritics amp PlanetaryScience 36359ndash369

Dressler B O Sharpton V L Morgan J Buffler R Moran D SmitJ Stoumlffler D and Urrutia J 2003a Investigating a 65 Ma-oldsmoking gun Deep drilling of the Chicxulub impact structureEOS Transactions 84130

Dressler B O Sharpton V L and Marin L E 2003b ChicxulubYax-1 impact breccias Whence they come (abstract 1259)34th Lunar and Planetary Science Conference CD-ROM

Fisher R V and Schmincke H U 1984 Pyroclastic rocks Berlin-Heidelberg Springer-Verlag 472 p

French B M 1998 Traces of catastrophe A handbook of shock-metamorphic effects in terrestrial meteorite impact structuresLPI Contribution No 954 Houston Lunar and PlanetaryInstitute 120 p

Grajales-Nishimura J M Cedillo-Pardo E Rosales-DominguezC Moran-Zenteno D J Alvarez W Claeys P Ruiz-MoralesJ Garcia-Hernandez J Padilla-Avila P and Sanchez-Rios A2000 Chicxulub impact The origin of reservoir and sealfacies in the southeastern Mexico oil fields Geology 28307ndash310

Kastens K A and Cita M B 1981 Tsunami-induced sedimenttransport in the abyssal Mediterranean Sea Geological Society ofAmerica Bulletin 92845ndash857

Keller G Stinnesbeck W Adatte T Stuumlben D and Kramar U 2003Chicxulub impact predates K-T boundary Supports multipleimpact hypothesis (abstract) Third International Conference onLarge Meteorite Impacts p 4020

Kenkmann T Wittmann A Scherler D and Stoumlffler D 2003 TheCretaceous sequence of the Chicxulub Yax-1 drillcore What isimpact-derived (abstract) Third International Conference onLarge Meteorite Impacts p 4075

Kiyokawa S Tada R Iturralde-Vinent M A Tajika E YamamotoS Oji T Nakano Y Goto K Takayama H Delgado D GOtero C D Rojas-Consuegra R and Matsui T 2002Cretaceous-Tertiary boundary sequence in the Cacarajicaraformation western Cuba An impact-related high energygravity-flow deposit In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 125ndash144

Koshikawa T Kido Y and Tada R 2003 High resolution rapidelemental analysis using an XRF micro-scanner Journal ofSedimentary Research 73824ndash829

Lindstroumlm M Floden T Grahn Y and Kathol B 1994 Post-impactdeposits in Tvaumlren A marine Ordovician crater south ofStockholm Sweden Geological Magazine 13191ndash103

Matsui T Imamura F Tajika E Nakano Y and Fujisawa Y 2002Generation and propagation of a tsunami from the CretaceousTertiary impact event In Catastrophic events and massextinctions Impact and beyond edited by Koeberl C andMacleod G Special Paper 356 Boulder Geological Society ofAmerica pp 69ndash77

Middleton G V and Hampton M A 1976 Subaqueous sedimenttransport and deposition by sediment gravity flows In Marinesediment transport and environmental management edited byStanley D J and Swift D J P New York Wiley pp 197ndash218

Middleton G V and Southard J B 1977 Mechanics of sedimentmovement SEPM lecture notes for short course No 3 401 p

Morgan J Warner M and Chicxulub working group 1997 Size andmorphology of the Chicxulub impact crater Nature 390472ndash476

Morgan J V Warner M R Collins G Melosh H J Christeson GL 2000 Peak-ring formation in large impact cratersGeophysical constraints from Chicxulub Earth and PlanetaryScience Letters 183347ndash354

Ormouml J 1994 The pre-impact Ordovician stratigraphy of the TvaumlrenBay impact structure SE Sweden GFF 116139ndash144

Ormouml J and Lindstroumlm M 2000 When a cosmic impact strikes thesea bed Geological Magazine 13767ndash80

Pierazzo E and Crawford D A 1997 Hydrocode simulations ofChicxulub as an oblique impact event (abstract) Large MeteoriteImpacts and Planetary Evolution Conference p 40

Pierazzo E Kring D A and Melosh H J 1998 Hydrocodesimulation of the Chicxulub impact event and the production ofclimatically active gases Journal of Geophysical Research 10328607ndash28625

Poag C W Powars D S Poppe L J and Mixon R B 1994Meteoroid mayhem in Ole Virginny Source of the NorthAmerican tektite strewn field Geology 22691ndash694

Poag C W Plescia J B and Molzer P C 2002 Ancient impactstructures on modern continental shelves The Chesapeake BayMontagnais and Toms Canyon craters Atlantic margin of NorthAmerica Deep-Sea Research II 491081ndash1102

Sharpton V Marin L E Carney J L Lee S Ryder G SchuraytzB C Sikora P and Spudis P D 1996 A model of the Chicxulubimpact basin based on evolution of geophysical data well logsand drill core samples In Cretaceous-Tertiary event and othercatastrophes in Earth history edited by Ryder G Fastovsky Dand Gartner S Special Paper 307 Boulder Geological Society ofAmerica pp 55ndash74

Smit J Roep T B Alvarez W Montanari A Claeys P Grajales-Nishimura J M and Bermudez J 1996 Coarse-grained clasticsandstone complex at the KT boundary around the Gulf ofMexico Deposition by tsunami waves induced by the Chicxulubimpact In Cretaceous-Tertiary event and other catastrophes inEarth history edited by Ryder G Fastovsky D and Gartner SSpecial Paper 307 Boulder Geological Society of America pp151ndash182

Smit J 1999 The global stratigraphy of the Cretaceous-Tertiaryboundary impact ejecta Annual Review of Earth and PlanetarySciences 2775ndash113

Smit J Dressler B Buffler R Moran D Sharpton B Stoeffler DUrrutia J and Morgan J 2003 Stratigraphy of the Yaxcopoil-1drill hole in the Chicxulub impact crater (abstract) EGS-AGU-EUG Joint Assembly p 6498

Tada R Nakano Y Iturralde-Vinent M A Yamamoto S KamataT Tajika E Toyoda K Kiyokawa S Delgado D G Oji TGoto K Takayama H Rojas-Consuegra R and Matsui T 2002Complex tsunami waves suggested by the Cretaceous-Tertiary

Evidence for ocean water invasion into the Chicxulub crater 1247

boundary deposit at the Moncada section western Cuba InCatastrophic events and mass extinctions Impact and beyondedited by Koeberl C and Macleod G Special Paper 356 BoulderGeological Society of America pp 109ndash123

Tada R Iturralde-Vinent M A Matsui T Tajika E Oji T Goto KNakano Y Takayama H Yamamoto S Rojas-Consuegra RKiyokawa S Garciacutea-Delgado D Diacuteaz-Otero C and Toyoda KForthcoming KT boundary deposits in the proto-Caribbeanbasin American Association of Petroleum Geologists Memoir

Takayama H Tada R Matsui T Iturralde-Vinent M A Oji T

Tajika E Kiyokawa S Garcia D Okada H Hasegawa T andToyoda K 2000 Origin of the Pentildealver Formation innorthwestern Cuba and its relation to KT boundary impact eventSedimentary Geology 135295ndash320

Wittmann A Kenkmann T Schmitt R T Hecht L and StoumlfflerD 2003 Impact melt rocks in the ldquoCretaceous megablocksequencerdquo of drill core Yaxcopoil-1 Chicxulub craterYucataacuten Mexico (abstract) Third International Conferenceon Large Meteorite Impacts p 4125