This page intentionally left blankA GEOLOGI C TI ME SCALE 2004An international team of over 40 stratigraphic experts, manyactively involved in the International Commission of Stratig-raphy (ICS), have helped to build the most up-to-date in-ternational stratigraphic framework for the Precambrian andPhanerozoic. This successor to A Geologic Time Scale 1989by W. Brian Harland et al. (Cambridge, 1989) begins with anintroduction to the theory and methodology behind the con-struction of the new time scale. The main part of the book isdevoted to the scale itself, systematically presenting the stan-dard subdivisions at all levels using a variety of correlationmarkers. Extensive use is made of stable and unstable isotopegeochronology, geomathematics, and orbital tuning to producea standard geologic scale of unprecedented detail and accuracywith a full error analysis. A wallchart summarizing the wholetime scale, with paleogeographic reconstructions throughoutthe Phanerozoic is included in the back of the book. The timescale will be an invaluable reference source for academic andprofessional researchers and students.FELI X GRADSTEI N is currently Professor of Stratig-raphy and Micropaleontology at the Geological Museum ofthe University of Oslo where he leads the offshore relationalstratigraphic database funded by a petroleum consortium. Heis the current chair of the International CommissiononStratig-raphy, whichis workingonthe formal classicationof the globalPrecambrian and Phanerozoic rock record and the interna-tional time scale.JAMES OGG is Professor of Stratigraphy at Purdue Uni-versity, West Lafayette, Indiana, and has research interestsin magnetochronology, cyclostratigraphy, sedimentology, andstratigraphic databases. He is presently Secretary General ofthe International Commission on Stratigraphy (ICS).ALAN SMI TH is Reader in Geology at the University ofCambridge and a Fellow of St. Johns College. His principalresearch interests are paleogeographic reconstruction and re-lated software development.A Geologic Time Scale 2004Edited byFelix M. Gradstein, James G. Ogg, and Alan G. Smithcaxniioci uxiviisir\ iiissCambridge, New York, Melbourne, Madrid, Cape Town, Singapore, So PauloCambridge University PressThe Edinburgh Building, Cambridge cn: :iu, UKFirst published in print format isnx-:, ,;-c-,::-;::-isnx-:, ,;-c-,::-;o;,-;isnx-:, ,;-c-,::-c:,o-, F. M. Gradstein, J. G. Ogg, and A. G. Smith 20042005Information on this title: www.cambridge.org/9780521781428This book is in copyright. Subject to statutory exception and to the provision ofrelevant collective licensing agreements, no reproduction of any part may take placewithout the written permission of Cambridge University Press.isnx-:c c-,::-c:,o-:isnx-:c c-,::-;::-oisnx-:c c-,::-;o;,-Cambridge University Press has no responsibility for the persistence or accuracy ofuiis for external or third-party internet websites referred to in this book, and does notguarantee that any content on such websites is, or will remain, accurate or appropriate.Published in the United States of America by Cambridge University Press, New Yorkwww.cambridge.orghardbackpaperbackpaperbackeBook (EBL)eBook (EBL)hardbackDedicationWe dedicate this third edition of the Geologic Time Scalebook to W. B. (Brian) Harland. He was an inspiring leaderin practical stratigraphy, its philosophical roots, and its primeproduct: The Geologic Time Scale!Deceased.With the acceptance of a reliable time scale, geology will havegained an invaluable key to further discovery. In every branchof science its mission will be to unify and correlate, and withits help a fresh light will be thrown on the more fascinatingproblems of the Earth and its Past.Arthur Holmes, 1913, The Age of the EarthContentsList of contributors page xiiPreface xvAcknowledgments xviiList of abbreviations and acronyms xixPART I I NTRODUCTI ON 11 Introduction 3r. . oa+ns+rr1.1 A Geologic Time Scale 2004 31.2 How this book is arranged 41.3 Conventions and standards 61.4 Historical overview of geologic time scales 72 Chronostratigraphy: linking time and rock 20r. . oa+ns+rr, J. o. ooo, +n +. o. sr+n2.1 Time and rock 202.2 Standardization of the chronostratigraphic scale 212.3 Case examples of GSSPs 422.4 Major subdivisions of the geologic time scale 432.5 Examples of stratigraphic charts and tables 45PART I I CONCEPTS AND METHODS 473 Biostratigraphy: time scales fromgraphic and quantitative methods 49r. . oa+ns+rr, a. +. coorra,+n r. . s+nrra3.1 Introduction 493.2 Graphic correlation 493.3 Constrained optimization 513.4 Ranking and scaling 524 Earths orbital parameters and cyclestratigraphy 55r. +. nrov4.1 Introduction 554.2 Earths orbital parameters 554.3 Orbitally forced insolation 574.4 Orbital signals in cycle stratigraphy 594.5 Estimating orbital chronologies 615 The geomagnetic polarity time scale 63J. o. ooo +n +. o. sr+n5.1 Principles 635.2 Late CretaceousCenozoic geomagneticpolarity time scale 655.3 Middle JurassicEarly Cretaceousgeomagnetic polarity time scale 745.4 Geomagnetic polarity time scale for MiddleJurassic and older rocks 855.5 Superchrons and polarity bias 865.6 Summary and conclusions 866 Radiogenic isotope geochronology 87. vrrrrrtvr6.1 Introduction 876.2 Types of uncertainties 876.3 Dating methods 886.4 Summary and conclusions 957 Strontium isotope stratigraphy 96J. . c+a+nta +n a. J. now+a+n7.1 Introduction 967.2 Materials for strontium isotope stratigraphy 987.3 A Geologic Time Scale 2004 (GTS2004) database 987.4 Comments on the LOWESS t 1028 Geomathematics 106r. r. +o+raurao8.1 History and overview 1068.2 Paleozoic applications 1108.3 Late Cretaceous and Paleogene applications 1178.4 Concluding remarks 124PART I I I GEOLOGI C PERI ODS 1279 The Precambrian: the Archean andProterozoic Eons 129r. J. aouu, +. n. korr, k. +. rrtu,o. +. snrrrns, n. s+a+tss, +n J. vrrzra9.1 Introduction 1299.2 History and recommended subdivision 130ixx Contents9.3 Nomenclature of the subdivisions 1329.4 The Neoproterozoic 1329.5 Isotope stratigraphy in the Precambrian 1339.6 Biostratigraphy in the Neoproterozoic 1369.7 Neoproterozoic ice ages and chronometricconstraints 1399.8 Summary 14010 Toward a natural Precambriantime scale 141w. urrrkra10.1 Introduction 14110.2 Current Precambrian subdivisionsand problems 14110.3 A natural Precambrian time scale 14210.4 Conclusions 14611 The Cambrian Period 147J. n. snraoorn +n a. +. coorra11.1 History and subdivisions 14711.2 Cambrian stratigraphy 15711.3 Cambrian time scale 15912 The Ordovician Period 165a. +. coorra +n r. . s+nrra12.1 History and subdivisions 16512.2 Previous standard divisions 16912.3 Ordovician stratigraphy 17112.4 Ordovician time scale 17613 The Silurian Period 188. J. rrcnr, a. +. coorra,+n r. . s+nrra13.1 History and subdivisions 18813.2 Silurian stratigraphy 19313.3 Silurian time scale 19814 The Devonian Period 202. a. notsr +n r. . oa+ns+rr14.1 History and subdivisions 20214.2 Devonian stratigraphy 20814.3 Devonian time scale 21314.4 Appendix 22015 The Carboniferous Period 222v. n+vvnov, u. a. w+anr+w, +nr. . oa+ns+rr15.1 History and subdivisions 22215.2 Carboniferous stratigraphy 23315.3 Carboniferous time scale 23716 The Permian Period 249u. a. w+anr+w, v. n+vvnov, +nr. . oa+ns+rr16.1 History and subdivisions 24916.2 Regional correlations 25616.3 Permian stratigraphy 26316.4 Permian time scale 26417 The Triassic Period 271J. o. ooo17.1 History and subdivisions 27117.2 Triassic stratigraphy 28017.3 Triassic time scale 28818 The Jurassic Period 307J. o. ooo18.1 History and subdivisions 30718.2 Jurassic stratigraphy 32618.3 Jurassic time scale 33919 The Cretaceous Period 344J. o. ooo, r. r. +o+raurao, +nr. . oa+ns+rr19.1 History and subdivisions 34419.2 Cretaceous stratigraphy 36519.3 Cretaceous time scale 37120 The Paleogene Period 384n. r. rt+rau+cnra, J. a. +rr,n. uarkntrs, r. . oa+ns+rr,J. J. nookra, s. orcnr, J. o. ooo,J. rowrrr, t. aonr, +. s+rrrrrro,+n u. scnr+z20.1 History and subdivisions 38420.2 Paleogene biostratigraphy 38920.3 Physical stratigraphy 40120.4 Paleogene time scale 40321 The Neogene Period 409r. rotars, r. nrror, . J. sn+ckrr+o,J. r+sk+a, +n n. wrrso21.1 History and subdivisions 40921.2 Neogene stratigraphy 41921.3 Toward an astronomically tuned Neogenetime scale (ATNTS) 43022 The Pleistocene and Holocene Epochs 441r. oruu+an +n +. v+ korrscno+r22.1 Pleistocene series 44122.2 Terrestrial sequences 44322.3 Ocean sediment sequences 448Contents xi22.4 Landsea correlation 44922.5 PleistoceneHolocene boundary 45122.6 Holocene Series 451PART I V SUMMARY 45323 Construction and summary of thegeologic time scale 455r. . oa+ns+rr, J. o. ooo, +n +. o. sr+n23.1 Construction of GTS2004 45523.2 Future trends in geologic time scales 462Appendix 1 Recommended color coding of stages 465r. . oa+ns+rr +n J. o. oooAppendix 2 Orbital tuning calibrations andconversions for the Neogene Period 469r. rotars, r. nrror, . J. sn+ckrr+o,J. r+sk+a, +n J. wrrsoAppendix 3 Geomathematics 485r. r. +o+rauraoBibliography 487Stratigraphic Index 587General Index 589Contributorsrrrrx . oa+ns+rrGeological MuseumUniversity of OsloPO Box 1172 BlindernN-0318 OsloNorwayfelix.gradstein@nhm.uio.noJ+rs o. oooDepartment of Earth and AtmosphericSciences550 Stadium Mall DrivePurdue UniversityWest Lafayette, IN 47907-2051USAjogg@purdue.edu+r+ o. sr+nDepartment of Earth SciencesUniversity of CambridgeDowning StreetCambridge CB2 3EQUKags1@esc.cam.ac.ukrar+s r. +o+rauraoGeological Survey of Canada601 Booth StreetOttawa, Ontario K1A OE8Canadawot+ra urrrkraGeological Survey of Canada601 Booth StreetOttawa, Ontario K1A OE8Canadaaoora +. coorraGeological Time SectionInstitute of Geological and Nuclear SciencesPO Box 30368Lower HuttNew Zealandvr+nrra n+vvnovPermian Research InstituteBoise State University1910 University DriveBoise, ID 83725-1535USArnrr oruu+anGodwin Institute of Quaternary ResearchDepartment of GeographyUniversity of CambridgeDowning StreetCambridge CB2 3ENUKrrn+ +. nrovDepartment of Earth and Planetary SciencesThe Johns Hopkins UniversityBaltimore, MD 21218USArcn+rr a. notsrDepartment of GeologySouthampton Oceanographic CentreSouthamptonUKrtc+s rotarsFaculty of Earth SciencesDepartment of GeologyUtrecht UniversityBudapestlaan 43508 TA UtrechtThe Netherlandsn+s-rr+ra rt+rau+cnraInstitut und Museum f ur Geologie und PalaontologieEberhard-Karls Universit atSigwartstrasse 10D-72076 TubingenGermanyDeceased.xiiList of contributors xiiiJon c+a+ntaInstitute of Geological SciencesUniversity College LondonGower StreetLondon WC1E 6BTUKrkr J. rrcnrDepartment of Earth SciencesSt. Francis Xavier UniversityPO Box 5000Antigonish, NS B2G2W5Canadar+tarcr J. aouuEconomic Geology Research InstituteHugh Allsopp LaboratoryUniversity of the WitwatersrandPrivate Bag 3, Wits 2050South AfricaJon snraoornLa FreunieBenayes19510 MasseretFrancerkr vrrrrrtvrGeological Survey of Canada601 Booth StreetOttawa, ON K1A OE8Canadauatcr a. w+anr+wUS Geological Survey926A National CenterReston, VA 20192-0001USAJ+so +rrDepartment of Earth SciencesUniversity of Hong KongPokfulam RoadHong KongPeoples Republic of Chinanrk uarkntrsLaboratory of Paleobotany and PalynologyFaculty of BiologyUtrecht UniversityBudapestlaan 43584 CD UtrechtThe Netherlandsrarnrark J. nrrorFaculty of Earth SciencesDepartment of GeologyUtrecht UniversityBudapestlaan 43584 CD UtrechtThe NetherlandsJraav nookraThe Natural History Museum, PaleontologyCromwell RoadLondon SW7 5BDUKarcn+an J. now+a+nInstitute of Geological SciencesUniversity College LondonGower StreetLondon WC1E 6BTUK+narw n. korrDepartment of Earth and Planetary SciencesHarvard University24 Oxford StreetCambridge, MA 02138USAJ+cqtrs r+sk+aAstronomie et Systemes DynamiquesBureau des Longitudes77 Av. Denfert-RochereauF-75014 ParisFrancesror+++ orcnrDipartimento di Scienze della TerraUniversit` a di Firenze4, Via La PiraI-50121 FirenzeItalyJ+rs rowrrrDinosystems105 Albert RoadRichmondSurrey TW10 6DJUKkrr+n +. rrtuPO Box 102Hawker, ACT 2614Australiaxiv List of contributorsrs+urrr+ a+rrrDipartimento di Scienze della TerraUniversitario G. DAnnunzio66013 Chieti ScaloItalytastr+ aonrGeowissenschaftenUniversit at BremenPO Box 33 04 40D-28334 BremenGermanyrr+ra s+nrraDepartment of Earth SciencesUniversity of California at RiversideRiverside, CA 92521USA+rk+ s+rrrrrroScripps Institution of OceanographyUniversity of California at San DiegoLa Jolla, CA 92033USAuraora scnr+zMarine Geology, Earth Science CentreG oteborg UniversityBox 7064S-41381 G oteborgSwedenrcnor+s J. sn+ckrr+oGodwin LaboratoryDepartment of Earth SciencesUniversity of CambridgeCambridge CB2 3SAUKoa+n+ +. snrrrnsSchool of Earth SciencesJames Cook UniversityTownsville, Old. 4811Australian+a+rn s+a+tssGeologisch-Pal aontologisches InstitutWestf alische Wilhelms-Universit at M unster48149 M unsterGermanyJ+ vrrzraOttawaCarleton Geoscience CentreUniversity of OttawaOttawa, Ontario K1N 6N5CanadaJ. v+ n+Faculty of Earth SciencesUtrecht UniversityBudapestlaan 43584 CD UtrechtThe Netherlands+nrJs v+ korrscno+rFaculty of ArchaeologyLeiden UniversityReuvenplaats 4, 2300 RA LeidenThe Netherlandsnoto wrrsoDepartment of Geological SciencesUniversity of CaliforniaSanta Barbara, CAUSAPrefaceThis study presents the science community witha newgeologictime scale for circa 3850 million years of Earth history. Thescale encompasses many recent advances in stratigraphy, thescience of the layering of strata on Earth. The newscale closelylinks radiometric and astronomical age dating, and providescomprehensive error analysis on the age of boundaries for amajority of the geologic divisions of time. Much advantagein time scale construction is gained by the concept of stageboundary denition, developed and actively pursued underthe auspices of the International Commission on Stratigraphy(ICS), that co-sponsors this study.It was in 1997 that Alan Smith approached two of us(F.M.G. andJ.G.O.) withthe request toundertake a neweditionof AGeologic Time Scale 1989 (GTS89) for Cambridge Univer-sity Press. This was just after the Phanerozoic Time Scalewith the A3 format time scale colour chart as insert, spon-sored by Saga Petroleum in Norway, had appeared in Episodes.Although we realized this newrequest was a tall order, we opti-mistically accepted. A proposal was formulated and improvedthrough peer review. As with GTS89, the new edition of thebook would not necessarily give the very latest developmentsin any eld, but would present a balanced overview designedto be educational and useful for advanced university students.In particular progress with the concept and dening of stageboundaries had delineated most international geologic stages.Initially, a rather limited slate of specialists was engaged,andwe optimistically projectedcompletionof a revisedGTS89at the turnof the Millennium. Slightly after, F.M.G. andJ.G.O.became executive ofcers of ICS for the 20002004 term, andthe GTS project was incorporated in ICSs formal objectives.Creatinga newGTSin2000, 2001, or even2002, turnedoutto be rather optimistic. The more we involved ourselves in themyriad of challenges in stratigraphy and the Phanerozoic geo-logic time scale the more we realized that a major overhaul wasinorder. Rather thanupdating andrevising chapters of GTS89we set out to re-write the book from scratch and expand geo-logic period chapters along a xed, and ambitious format oftext and gures. Advances in time scale methodology involvingcycle stratigraphy, mathematics and statistics, stable isotopestratigraphy, and the formidable progress in high-resolutionage dating all demanded close attention with data integrationand specialists chapters.The vast progress in Precambrian and Phanerozoic stratig-raphy achievedduring the last decade requiredintense involve-ment of many more geoscientists than initially envisioned. Al-though the more ambitious scope and bigger team did pushback completion deadlines, we are condent it has enhancedthe consensus value of the new geologic time scale, namedGTS2004. Had we known beforehand that a total of 18 seniorand 22 contributing authors, for a total of 40 geoscience spe-cialists from 15 different countries, would work on the bookand deal with the new time scale, we might have had secondthoughts about our undertaking. The number of e-mails sentcriss-cross over the globe as part of GTS2004 is in the tensof thousands. A fundamental difference between multidisci-plinary studies and geologic time scale studies is that all chap-ters must align along the arrow of time. To put it simply: theCarboniferous cannot end at 291 Ma with the Permian start-ing at 299 Ma. Close agreement on type of data and standardsadmitted in actual time scale building is also vital. Hence, theactual data standardization and time scale calculation for eachchapter was kept to a small team in which Mike Villeneuve,Frits Agterberg, F.M.G., andJ.G.O. playedkeyroles, withothersenior authors as advisors. The new Neogene time scale wasdeveloped by Luc Lourens and his team of tuning specialists.The fascination in creating a new geologic time scale isthat it evokes images of creating a beautiful carpet, using manyskilled hands. All stitches must conform to a pre-determinedpattern, in this case the pattern of physical, chemical, and bio-logical events on Earth aligned along the arrow of time. It isthus that this newscale is a tribute to the truly close cooperationachieved by this slate of outstanding co-authors. We also con-sider the new time scale a tribute to the scientic competenceharbored and fostered by ICS.We are deeply grateful to all co-authors who without reser-vation accepted the challenge to be part of this dedicated team,slowly (!) stitching and weaving this carpet of time and itsevents that are Earths unique and splendid history.xvxvi PrefaceIt is withdeepregret that we learnedinmid2002 that one ofour most valuable scholars in Paleozoic stratigraphy, ProfessorMichael House, had passed away, very shortly after submissionof his draft chapter on Devonian stratigraphy. It has been anhonor to complete the task he set himself to create this eruditechapter of expansive and dramatic Earth history between 416and 359 Ma. Vascular plants and forests established on Earth,exceptional high global sea level occurred, ice caps formedin the south polar region in late Devonian time, and presentcontinents and shelves assembled on one hemisphere. Old RedSandstone is one of the Devonians great continental remnants.Through the NUNAConference in Canada in March 2003onNewFrontiers inthe FourthDimension: Generation, Cal-ibration and Application of Geologic Timescales the essayToward a natural Precambrian time scale by Wouter Bleekercame to this book. Hence, this period of over 88% of Earthhistory is getting some more urgent attention. We thank MikeVilleneuve and his teamfor organizing this timely geochronol-ogy conference.We are pleased to acknowledge the nancial contributionof ExxonMobil, Statoil, ChevronTexaco, and BP. With thesevital donations the elaborate graphics became possible. J. G.O. acknowledges partial support by the US National ScienceFoundation under Grant No. 0313524. The Geological Surveyof Canada and the Network of Offshore Records in Geologyand Stratigraphy (NORGES) project at the Geology Museumof the University of Oslo assisted with design and printing ofthe time scale wall chart.Cambridge University Press patiently awaited the fruitsof our labor, and we are much obliged to Matt Lloyd, SallyThomas, andLesleyThomas for their thorougheditorial adviceand assistance.Figure 1.4 in the Introduction chapter illustrates the 1960geologic time scale by its pioneer, Arthur Holmes, who intro-duced period scaling from observed maximum thickness. Theappearance datum of this new opus in mid 2004 is nearly 90years after Arthur Holmess rst humble geologic time scalein 1913 in search of the age of the Earth and its remarkablehistoric components.This publication on the International Geologic Time Scalewas producedunder the auspices of the International Commis-sion on Stratigraphy (ICS). Information on ICS, its organiza-tion, its mandate, andits wide-ranging geoscience programcanbe obtained from www.stratigraphy.org.Felix M. GradsteinJames G. OggAlan G. SmithAcknowledgmentsThe authors and co-authors of the 23 chapters in this bookare very pleased to acknowledge the many geoscientists thatactively and generously gave of their time to assist with andadvise on GTS2004.F.M.G. and J.G.O. thank Frits Agterberg, Sam Bowring,DavidBruton, Pierre Bultynck, Cinzia Cervato, Roger Cooper,Ferdinand Corfu, Sorin Filipescu, Stan Finney, Rich Lane,Luc Lourens, John McArthur, Ed de Mulder, Jurgen Remane,Paul Renne, Otto vanBremen, andMike Villeneuve for generaladvice and/or help over the many years of book gestation.John McArthur also checked every section dealing withgeochemistry and reviewed and helped update all relevant g-ures; Luc Lourens assisted with the task to link the complexand detailed Neogene and Pleistocene chapters; Heiko P alikeoffered insight in future trends with regard to orbital tuningof the steadily improving Paleogene time scale.GTS2004wouldnot have beenpossible without Gabi Oggstireless dedicationandtruly hardandhighly creative work withscientic design, gure and table drafting, and error checking.Virtually all of the 156 gures came from her drawing table.We also thank Jane Dolven and Gisli Sigtryggsson for theircontributions with drafting, revising, and printing of somegures.F.M.G. and J.G.O. thank the International Union of Geo-logical Sciences (IUGS) and the Commission for the Geolog-ical Map of the World (CGMW) for advice and support withthe GTS project.Linda Hinnov is very grateful to Marie-France Loutre andTimothy Herbert for taking the time to comment on specicissues discussed in her chapter on Earths orbital parametersand cycle stratigraphy.Mike Villeneuve thanks Richard Stern and Otto vanBremen for assistance with Chapter 6, Radiogenic isotopegeochronolology.Frits P. Agterberg, author of Chapter 8, Geomathemat-ics, expresses thanks to Roger Cooper, Felix Gradstein, JimOgg, Peter Sadler, and Alan Smith for helpful discussionsand information; to Mike Villeneuve for discussions regard-ing geochronological aspects; to Graeme Bonham-Carter andDanny Wright for help with software; and to Gabi Ogg for helpwith constructing the many diagrams.Wouter Bleeker reviewed Chapter 9, The Precambrian: theArchean and Proterozoic Eons; his valuable comments im-proved the manuscript. In Chapter 10, Toward a NaturalPrecambrian time scale, W. Bleeker discussed some of the ideasexpressed by Euan Nisbet, who together with Preston Cloudhas been a vocal critic of the current Precambrian time scale.Discussions withYuri AmelinandRichardSternhelpedclarifythe magnitude of the limitations due to uncertainties in the de-cay constant. He thanks Felix Gradstein and Mike Villeneuvefor their encouragement to formulate this study, and for theircritical comments on an early draft of the manuscript. Chap-ter 10 is Geological Survey of Canada contribution 2 003 066.Chapter 11, The Cambrian Period, was reviewed by SorenJensen, Graham Shields, and David Bruton; senior chapterauthor John Shergold thanks these contributors.The authors of Chapter 12, The Ordovician Period,Roger Cooper and Peter Sadler, thank Mike Villeneuve, SamBowring, and Bill Compston for their comments on radioiso-topic dating methods. However, the choice of dates used forcalibration of the time scale was theirs alone. Frits Agterbergand Felix Gradstein performed the time scale calculations.Chapter 13, The Silurian Period, beneted from help-ful comments and review by David Bruton, and discussionswith Alfred Lenz, Tatiana Koren, David Loydell, GodfreyNowlan, Gary Mullins, and John Beck. Mike Villeneuve dis-cussed radiometric dating methodology with Roger Cooper,and Frits Agterberg and Felix Gradstein performed age datastandardization and time scale and uncertainty calculations.Mike Melchin, who is senior author of the chapter, also grate-fullyacknowledges nancial support fromthe Natural Sciencesand Engineering Research Council of Canada.Completion of Chapter 14, The Devonian Period, under-taken by F.M.G. after Professor Michael House sadly passedaway, would not have been possible without the vital helpof Pierre Bultynck, Bernd Kaufmann, and David Bruton,all of whom provided valuable advice and information; theircomments improvedthe text andgures. F.M.G. thanks Dr Jimxviixviii AcknowledgmentsHouse for his assistance in obtaining gures from the estateof Professor House, and thanks E. A. Williams for permis-sion to utilize his illustration of discrepancies between datingmethods. Frits P. Agterberg kindly executed the statisticalmathematical routines.The Carboniferous Period, Chapter 15, was undertaken bysenior author Vladimir Davydov. He thanks many individualsfor their input toward this chapter on this complex subject. Inparticular, Peter Jones, John Groves, Paul Brenckle, and PhilHeckel reviewed the text and made several signicant sug-gestions that improved both style and content. The authorsare grateful to E. Trapp and B. Kaufmann for providing newTIMSdates fromGermany, and thank SamBowring and MikeVilleneuve for advice on radiometric dates, Chris Klootwijk foradvice on magnetostratigraphy, Frits Agterberg for help withthe numerical analysis, and Gabi Ogg for drafting the gures.The senior author of Chapter 16, The Permian Period,Bruce Wardlaw, thanks Sam Bowring and Frits Agterberg foradvice and assistance, and Gabi Ogg for drafting of the manygures.Chapter 17, The Triassic Period, was authored by J.G.O.He is most grateful to Tim Tozer, Linda Hinnov, Nereo Preto,John McArthur, and, especially, Mike Orchard for providingvaluable insights into the intricate history and current debateson Triassic subdivisions and stratigraphy and/or reviewingearlier versions of this summary. The reviewers cautiously ad-vise that the Triassic time scale may evolve in unforeseen waysin the future as GSSPs are formalized and better global corre-lations are achieved. Josef P alfy, SamBowring, RolandMundil,and Paul Renne enlightened the author about the constraintsand disagreements with radioisotopic ages. Gabi Ogg draftedthe nal versions of the diagrams.As author of The Jurassic Period, Chapter 18, J.G.O.acknowledges the countless Jurassic experts who have con-tributed their expertise on the intricacies of the Jurassic world.In particular, he thanks Josef P alfy, Nicol Morton, Angela Coe,Bill Wimbledon, and Simon Kelly for reviewing drafts of thischapter; and John McArthur for updating the geochemistry.Gabi Ogg drafted the nal versions of the diagrams.Drafts of The Cretaceous Period, Chapter 19, were re-viewed by Peter Rawson, Stephane Reboullet, Jorg Mutter-lose, Jurgen Thurow, and Linda Hinnov; and John McArthurupdated the geochemistry. Gabi Ogg drafted the nal versionsof the detailed diagrams. Senior author J.G.O. thanks all thesecontributors for their vital assistance in elucidating the stratig-raphy of the longest period (80 myr) in the Phanerozoic.The Paleogene Period, Chapter 20, was authored by Hans-Peter Luterbacher and many contributors. Heiko P alike re-viewed and advised on the actual time scale, Jan Hardenboladvised on an early draft of the manuscript, and Frits Agter-berg performed the scaling and uncertainty calculations.The Neogene Period, Chapter 21, came to light under thesenior authorship of Luc Lourens; he and his co-authors liketo express their thanks to Hayfaa Abdul-Aziz, Torsten Bickert,Katharina Billups, Henk Brinkhuis, Peter DeMenocal, WoutKrijgsman, Klaudia Kuiper, Alan Mix, Heiko P alike, JamiePowell, Isabella Raf, Javier Sierro, Ralf Tiedemann, ElenaTurco, Jan Van Dam, Erwin Van der Laan, and Jan-WillemZachariasse for their indispensable contributions. The authorsacknowledge Bill Berggren, whose studies inspired them inwriting this chapter. Davide Castradori andLinda Hinnov pro-vided very helpful and constructive comments on an earlierversion of this manuscript.A fruitful discussion on the nettlesome Quaternary issuebetween F.M.G. and Phil Gibbard during the First Conferenceon Future Directions in Stratigraphy in 2002 in Urbino, Italy,led to Chapter 22, The Pleistocene and Holocene Epochs. LucLourens is thanked for a careful review, and Brad Pillans foradvice. The authors thank Steve Boreham and Gabi Ogg forhelp in the compilation of the chart.Wendy Green and Lawrence Rush helped with getting thereferences intoEndNote. The GSSPmaps were drawnwiththeATLASprogramof Cambridge Paleomap Services, written byLawrence Rush.Abbreviations and acronymsORGANI ZATI ONSCGMW Commission for the Geological Mapof the WorldDNAG Decade of North American GeologyDSDP Deep Sea Drilling ProjectGSC Geological Survey of CanadaICS International Commission of StratigraphyIGC International Geological CongressIGCP International Geological Correlation ProjectINQUA International Quaternary AssociationIUGS International Union of Geological SciencesODP Ocean Drilling ProjectSNS Subcommision (of ICS) on NeogeneStratigraphySQS Subcommission (of ICS) on QuaternaryStratigraphySOS Subcommission (of ICS) on OrdovicianStratigraphyUNESCO United Nations Education, Scientic,and Cultural OrganizationUSGS United States Geological SurveyTI ME SCALE PUBLI CATI ONS ( see Referencesfor details)NDS82 Numerical Dating in Stratigraphy (Odinet al., 1982)GTS82 A Geologic Time Scale (Harland et al.,1982)DNAG83 Geologic Time Scale, Decade of NorthAmerican Geology (Palmer, 1983)KG85 Kent and Gradstein (1985)EX88 Exxon 1988 (Haq et al., 1988)GTS89 A Geologic Time Scale 1989 (Harlandet al., 1990)OB93 Obradovich (1993)JGR94 Journal of Geophysical Research 1994(Gradstein et al., 1994)SEPM95 Society for Sedimentary Geology 1995(Gradstein et al., 1995)GO96 Gradstein and Ogg (1996)GEOSCI ENTI FI C CONCEPTSFAD First appearance datumFCT Fish Canyon Tuff sanidine monitor standard(in ArAr dating)GPTS Geomagnetic polarity time scaleGSSP Global Stratotype Section and PointGSSA Global Standard Stratigraphic Age(in Precambrian)HRSIMS High-resolution secondary ion massspectrometry (in UPb dating)LAD Last appearance datumLA2003 Laskar 2003 numerical solution of orbitalperiodicitiesMMhb-1 McClure Mountain hornblende monitorstandard (in ArAr dating)SL13 Sri Lanka 13 monitor zircon standard(in HRSIMS dating)TIMS Thermal ionization mass spectrometry(in UPb dating)TCR Taylor Creek Rhyolite Sanidine monitorstandard (in ArAr dating)SYMBOLSka 103years ago (kilo annum)kyr 103years durationGa 109years ago (giga annum)Ma 106years ago (mega annum)myr 106years durationSI Syst` eme Internationale dUnit esyr year durationxixPart I Introduction1 Introductionr. . oa+ns+rr The development of newdating methods andthe extensionof existingmethods has stimulated the need for a comprehensive review of thegeologic time scale. The construction of geologic time scales evolvedas a result of applying new ideas, methods, and data.1. 1 A GEOLOGI C TI ME SCALE 2004The geologic time scale is the framework for deciphering thehistory of the Earth. Since the time scale is the tool par ex-cellence of the geological trade, insight in its construction,strengths, and limitations greatly enhances its function and itsutility. All earth scientists should understand howthe evolvingtime scales are constructed and calibrated, rather than merelyusing the numbers in them.This calibration to linear time of the succession of eventsrecorded in the rock record has three components:1. the international stratigraphic divisions and their correla-tion in the global rock record,2. the means of measuring linear time or elapsed durationsfrom the rock record, and3. the methods of effectively joining the two scales.For convenience in international communication, the rockrecord of Earths history is subdivided into a chronostrati-graphic scale of standardized global stratigraphic units, suchas Jurassic, Eocene, Harpoceras falciferum ammonitezone, or polarity Chron C24r. Unlike the continuous tick-ing clock of the chronometric scale (measuredinyears beforepresent), the chronostratigraphic scale is basedonrelative timeunits, in which global reference points at boundary stratotypesdene the limits of the main formalized units, such as Devo-nian. The chronostratigraphic scale is an agreed convention,whereas its calibration to linear time is a matter for discoveryor estimation (Fig. 1.1).A Geologic Time Scale 2004, eds. Felix M. Gradstein, James G. Ogg, and AlanG. Smith. Published by Cambridge University Press. c F. M. Gradstein,J. G. Ogg, and A. G. Smith 2004.ChronostratigraphicscaleStageNorianCarnianLadinianAnisianChronometricscaleArArUPbAbsoluteagesAstronomicalcyclesGeologic time scalee.g.GTS2004CalibrationFigure 1.1 The construction of a geologic time scale is the mergerof a chronometric scale (measured in years) and achronostratigraphic scale (formalized denitions of geologic stages,biostratigraphic zonation units, magnetic polarity zones, and othersubdivisions of the rock record).By contrast, Precambrian stratigraphy is formally classi-ed chronometrically (see Chapter 9), i.e. the base of eachPrecambrian eon, era, and period is assigned a numerical age(Table 1.1).Continual improvements in data coverage, methodology,and standardization of chronostratigraphic units imply thatno geologic time scale can be nal. A Geologic Time Scale 2004(GTS2004) provides an overviewof the status of the geologicaltime scale and is the successor to GTS1989 (Harland et al.,1990), whichinturnwas precededbyGTS1982(Harlandet al.,1982).34 PART I I NTRODUCTI ONTable 1.1 Current framework for subdividing terrestrial stratigraphyEon Era Denition of base Age in (Ma)Phanerozoic CenozoicMesozoicPaleozoicBoundaries dened in rock(chronostratigraphically) byGSSPsTo be discovered by correlation fromGSSPs and dating. Base ofPhanerozoic dated at 542 MaProterozoic NeoproterozoicMesoproterozoicPaleoproterozoicBoundaries dened in time(chronometrically) by arbitraryassignment of numerical ageAge of basal Proterozoic denedas 2500 MaArchean NeoarcheanMesoarcheanPaleoarcheanEoarcheanBoundaries dened in time(chronometrically) by arbitraryassignment of numerical ageAge of basal Archean not denedSince 1989, there have been several major developments:1. Stratigraphic standardization through the work of the In-ternational Commission on Stratigraphy (ICS) has greatlyrened the international chronostratigraphic scale. In somecases, traditional European-based geological stages havebeen replaced with new subdivisions that allow global cor-relation.2. Newor enhancedmethods of extracting linear time fromtherock record have enabled high-precision age assignments.An abundance of high-resolution radiometric dates havebeen generated that has led to improved age assignments ofkey geologic stage boundaries. The use of global geochem-ical variations, Milankovitch climate cycles, and magneticreversals have become important calibration tools.3. Statistical techniques of extrapolating ages and associateduncertainties to stratigraphic events have evolved to meetthe challenge of more accurate age dates and more pre-cise zonal assignments. Fossil event databases with multiplestratigraphic sections through the globe can be integratedinto composite standards.The compilation of GTS2004 has involved a large num-ber of specialists, including contributions by past and presentchairs of different subcommissions of ICS, geochemists work-ing with radiogenic and stable isotopes, stratigraphers usingdiverse tools from traditional fossils to astronomical cycles todatabase programming, and geomathematicians.The set of chronostratigraphic units (stages, eras) andtheir computed ages which constitute the main framework forA Geologic Time Scale 2004 are summarized as Fig. 1.2, withdetailed descriptions and stratigraphic scales in appropriatechapters.1. 2 HOW THI S BOOK I S ARRANGEDThe foundation of the geologic time scale is the standard-ized system of international stratigraphic units. Chapter 2summarizes the philosophy of the construction of this inter-national standard, gives selected examples of dening bound-aries, and reviews the origin of the main divisions of eons anderas.Biostratigraphy, or the use of fossils in the rock recordfor assigning relative ages, has merged with mathematical andstatistical methods to enable scaled composites of global suc-cession of events. Chapter 3 on biostratigraphy summarizesthese quantitative methods, which were used to construct theprimary standard for most of the Paleozoic time scale (from542 to 251 Ma).Periodic multi-thousand-year oscillations in the Earths or-bit and tilt relative to the Sun produce cyclic environmentalchanges that are recorded in sediments. Chapter 4 summa-rizes how these astronomical signals are extracted from thesediments and used to construct a very high-resolution timescale that can be tied to the present orbital condition (lineartime) or to measure actual elapsed time. Cycle stratigraphy hascalibrated the time scales for most of the Neogene Period (i.e.for the past 23 million years), and for portions of the Paleo-gene Period (from 65 to 23 Ma) and Mesozoic Era (from 251to 65 Ma).Reversals of the Earths geomagnetic eld are recordedby sediments, by volcanic rocks, and by the oceanic crust.Chapter 5 explains how the oceanic magnetic anomalies arecalibrated with spreading models to produce a powerful corre-lation tool for sediments deposited during the past 160 millionyears. These calibrated C-sequence and M-sequence polaritytime scales enable assignment of ages to stage boundaries andtoIntroduction 5PHANEROZOIC PRECAMBRIANGEOLOGIC TIME SCALEAGE(Ma) AGE(Ma)Stage EpochMESOZOICPeriod707580859510010511011512012513013514090145150155165170175180185190195200205210160215220225235240245250230CretaceousCampanianSantonianTuronianConiacianAlbianAptianBerriasianTithonianKimmeridgianOxfordianCallovianBathonianBajocianAalenianToarcianPliensbachianSinemurianHettangianRhaetianNorianCarnianLadinianAnisianBarremianValanginianHauterivianOlenekianMaastrichtianCenomanianJurassicTriassicLateLateLateMiddleMiddleEarlyEarlyEarly65.570.683.585.889.393.599.6112.0125.0130.0136.4140.2145.5150.8155.7161.2164.7167.7171.6175.6183.0189.6196.5199.6203.6216.5228.0237.0245.0249.7251.0Induan50101520AGE(Ma) AGE(Ma)StageMioceneHolocenePlioceneEEMLLAquitanianBurdigalianLanghianSerravallianTortonianMessinianZancleanPiacenzianPleisto-ceneEpochGelasianNeogeneCENOZOICPeriod253035404550556065EocenePaleoceneOligoceneDanianSelandianThanetianYpresianLutetianBartonianPriabonianRupelian EEEMMLLLChattianPaleogene65.558.755.837.233.928.461.748.640.423.0320.4315.9713.655.333.607.2511.611.812.59255535540260265275280285290295300305310315320270325330335345350355360365370375380385390340395400405415420425430435440445450455460410465470475485490495500505510515520525530480Penn-sylvanianMississippianCarboniferousDevonianSilurianOrdovicianPermianCambrianLateLateMiddleEarlyMiddleEarlyPALEOZOICAGE(Ma) AGE(Ma)PridoliLudlowWenlockLlandoveryLopingianGuada-lupianCisuralianFurongianMiddleEarlyLateMiddleEarly LateMiddleEarlyEpoch PeriodFamennianGivetianEifelianTremadocianLochkovianPragianDarriwilianChanghsingianWuchiapingianCapitanianWordianRoadianKungurianArtinskianSakmarianAsselianGzhelianKasimovianMoscovianBashkirianSerpukhovianViseanTournaisianFrasnianEmsianHomerianTelychianAeronianRhuddanianPaibianStageSheinwoodianLudfordianGorstian411.2416.0418.7422.9428.2501318.1345.3326.4374.5385.3391.8397.5407.0443.7445.6260.4265.8275.6284.4294.6303.9306.5311.7421.3426.2436.0439.0455.8460.9468.1471.8478.6488.3513542.0268.0270.6299.0251.0253.8359.260070080010001100120013001400150016001700180019009002000210022002400250026002700280029003000310032003300230034003500360010001200140016001800250028003200230036003800390040004100420043004400450046003700AGE(Ma) AGE(Ma)Era Eon PeriodEdiacaranCryogenianTonianStenianEctasianCalymmianStatherianOrosirianRhyacianSiderianNeo-proterozoicNeo-archeanMeso-archeanPaleo-archeanEoarcheanMeso-proterozoicPaleo-proterozoicArcheanProterozoicLower limitis notdefined2050850~630542HirnantianQuaternaryFigure 1.2 Summary of A Geologic Time Scale 2004.6 PART I I NTRODUCTI ONbiostratigraphic and other stratigraphic events through muchof that interval.Chapter 6 on radiogenic isotopes summarizes the evolvingtechniques used to acquire high-precision ages from the rockrecord. However, high precision does not always imply accu-racy, and this chapter explains some of the pitfalls induced bygeological distortions or laboratory standards.Stable isotopes of strontium reveal a wealth of informationabout past environmental conditions and geochemical cycling.Chapter 7 explains the use of trends in the strontium isotoperatios of past seawater for global correlation and for relativescaling of stratigraphic events, and presents these trends forthe past 600 million years.Assembling the array of radiometric, biostratigraphic, cy-cles, magnetic, and other data into a unied geologic timescale, and extrapolating the ages and uncertainties on strati-graphic boundaries is the topic of Chapter 8, Geomathematics.This chapter also details construction methods and results forGTS2004.The Precambrian encompasses the 4 billion years from theformation of the Earth to the evolution of multicellular life.In addition to summarizing major geologic and geochemicaltrends, the two chapters on the Precambrian highlight thephilosophical difference in establishing chronostratigraphicsubdivisions based on pure linear age versus identifying sig-nicant global events.The Phanerozoic (the past 542 million years of Earthhistory) is subdivided into 11 periods. Each of the periodchapters has three principal parts: an explanation of the for-mal subdivision into stages using global boundary stratotypesassociated with primary and secondary correlation markers; asummary of the biostratigraphy, cycle stratigraphy, magneticstratigraphy, and geochemical stratigraphy features that areapplied to construct high-resolution relative time scales; andthe methods of calibration to a linear time framework. Eachperiod chapter includes a detailed graphic presentation of itsintegrated geologic time scale, and these are drawn at a uni-form scale among all chapters and in the color plates sectionto allow visual comparison of rates.The summary of GTS2004 (Fig. 1.2) inChapter 23 reviewsthe entire geologic time scale, summarizes its construction anduncertainties, and outlines potentially rewarding directions forfuture time scale research.1. 3 CONVENTI ONS AND STANDARDSAges are given in years before Present (BP). To avoid a con-stantly changing datum, Present was xed as AD1950 (as in14Cdeterminations), the point intime at whichmodernisotopedating research began in laboratories around the world. Formost geologists, this offset of ofcial Present from todayis not important. However, for archeologists and researchersinto events during the Holocene (the past 11500 years), thecurrent offset (50 years) between the BP convention fromradiometric laboratories and actual total elapsed calendar yearsbecomes signicant.For clarity, the linear age in years is abbreviated as a(for annum), and ages are generally measured in ka or Ma, forthousands, millions, or billions of years before present. Theelapsed time or duration is abbreviated as yr (for year), anddurations are generally in kyr or myr. Therefore, the Cenozoicbegan at 65.5 Ma, and spans 65.5 myr (to the present day).The uncertainties on computed ages or durations are ex-pressed as standard deviation (1-sigma or 68% condence) or2-sigma (95% condence). The uncertainty is indicated by and will have implied units of thousands or millions ofyears as appropriate to the magnitude of the age. Therefore,an age cited as 124.6 0.3 Ma implies a 0.3 myr uncertainty(1-sigma, unless specied as 2-sigma) on the 124.6 Ma date.We present the uncertainties () on summary graphics of thegeologic time scale as 2-sigma (95% condence) values.Geologic time is measured in years, but the standard unitfor time is the second s. Because the Earths rotation is not uni-form, this second is not dened as a fraction (1/86 400) of asolar day, but as the atomic second. The basic principle of theatomic clock is that electromagnetic waves of a particular fre-quency are emitted when an atomic transition occurs. In 1967,the Thirteenth General Conference on Weights and Measuresdened the atomic second as the duration of 9 192 631 770 pe-riods of the radiation corresponding to the transition betweentwo hyperne levels of the ground state of cesium-133. Thisvalue was established to agree as closely as possible with thesolar-day second. The frequency of 9 192 631 770 hertz (Hz),which the denition assigns to cesium radiation was carefullychosen to make it impossible, by any existing experimental ev-idence, to distinguish the atomic second from the ephemerissecond based on the Earths motion. The advantage of havingthe atomic second as the unit of time in the International Sys-tem of Units is the relative ease, in theory, for anyone to buildand calibrate an atomic clock with a precision of 1 part per 1011(or better). In practice, clocks are calibrated against broadcasttime signals, with the frequency oscillations in hertz being thependulum of the atomic time keeping device.1 year is approximately 31.56 mega seconds (1 a =31.56Ms).Introduction 7The Syst` eme Internationale dUnit es (SI) conventions at103intervals that are relevant for spans of geologic time throughsizes of microfossils are:109giga G106mega M103kilo k100unity 1103milli m106micro 109nanno nAlthoughdates assignedinthe geologic time scale are measuredin multiples of the atomic second as unit of time (year), thereare two other types of seconds: mean solar second and ephemerissecond.1.3.1 Universal timeUniversal time is utilized in the application of astronomy tonavigation. Measurement of universal time is made directlyfrom observing the times of transits of stars; since the Earthsrotationis not uniform, corrections are appliedto obtaina moreuniform time system. In essence, universal time is the meansolar time on the Greenwich meridian, reckoned in days of 24mean solar hours beginning with zero hour at midnight, andderives from the average rate of the daily motion of the Sunrelative to the meridian of Greenwich. The mean solar second is1/86 400 of the mean solar day, but because of non-uniformitythis unit is no longer the standard of international time.1.3.2 Ephemeris timeEphemeris time (ET) is uniform and obtained from observa-tion by directly comparing positions of the Sun, Moon, and theplanets with calculated ephemerides of their coordinates. Web-sters dictionary denes ephemeris as any tabular statement ofthe assigned places of a celestial body for regular intervals.Ephemeris time is based on the ephemeris second dened as1/31 556 925.9447 of the tropical year for 1900 January 0 day12 hour ET. The ephemeris day is 86 400 ephemeris seconds,which unit in 1957 was adopted by the International Astro-nomical Union as the fundamental invariable unit of time.1. 4 HI STORI CAL OVERVI EW OF GEOLOGI CTI ME SCALESStitching together the many data points on the loom oftime requires an elaborate combination of Earth science andmathematical/statistical methods. Hence, the time and ef-fort involved in constructing a new geologic time scale andassembling all relevant information is considerable. Because ofthis, and because continuous updating in small measure withnew information is not advantageous to the stability of anycommon standard, new geologic time scales spanning the en-tire Phanerozoic tend to come out sparsely (e.g. Harland et al.,1982, 1990; Gradstein and Ogg, 1996; Remane, 2000).In the absence of accepted accurate dates at each stageboundary, extrapolating the ages of geologic stages is a majorchallenge in time scale building and various methods have beenemployed by different compilations, including this GTS2004version. A major challenge in itself is to try to understand theprecision of radiometric ages, including calibrations betweendifferent radiometric methods, now that analytical errors aregreatly reduced.Figure 1.3 summarizes those and some others in terms of12 methods applied since 1937. Radiometric age dating, strati-graphic reasoning, and biostratigraphic/geomagnetic calibra-tions are three corner stones of time scale building. Strati-graphic reasoning, although fuzzy, evaluates the complex webof correlations around stage boundaries or other key levels, andis paramount in the science of stratigraphy. Geomathematicalmethods involve mathematical/statistical routines and inter-polations that can estimate margins of error on limits of strati-graphic units; such errors are of two main types, stratigraphicand analytical (see Chapter 8). Tuning of cyclic sequences toorbital time scales, either counting back from an anchor levelsuch as the present, or tuning individual cyclic segmentswith orbital periodicities (oating time scale), has the potentialto be the most accurate calibration of the geologic time scale(see Chapters 4 and 21). Such an orbitally tuned time scale canalso calibrate the standards and decay constants of radiometricmethods.1.4.1 Arthur Holmes and agethickness interpolationsArthur Holmes (18901965) was the rst tocombine radiomet-ric ages with geologic formations in order to create a geologictime scale. His book, The Age of the Earth (1913, 2nd edition1937), writtenwhenhewas only22, hada major impact onthoseinterested in geochronology. For his pioneering scale, Holmescarefully plotted four radiometric dates, one in the Eoceneand three in the Paleozoic from radiogenic helium and lead inuranium minerals, against estimates of the accumulated maxi-mum thickness of Phanerozoic sediments. If we ignore sizableerror margins, the base of Cambrian interpolates at 600 Ma,curiously close to modern estimates. The new approach was8 PART I I NTRODUCTI ONHolmes (1937)Holmes (1960)Kulp (1961)Funnell (1964)Berggren (1972)Hardenbol and Berggren (1978)NDS - Odin (1982)GTS82 - Harlandet al.(1982)DNAG - Palmer (1983)EX88 - Haq etal. (1987)GTS89 - Harlandetal. (1990)Odin (1994)Obradovich (1993)McArthuretal. (1994)SEPM95 - Gradsteinet al. (1995)Tucker and McKerrow (1995)Berggrenet al. (1995a)Lourensetal. (1996a,b)Weedon and Jenkyns (1999)Rhlet al. (2001)Shackleton et al. (1999)Young and Laurie (1996)Remane (2000)GTS2004Time scale method / Author referencesMaximum thickness of sediments per time periodEqual duration of stagesRate of radioactive decay of elementsTuning of cyclic sequences to orbital time scaleStratigraphic reasoningBiostratigraphic/geomagnetic calibrationEqual duration of (sub) zone to scale stagesZone duration is proportional to zone thicknessConstancy of spreading in ocean-floor segmentsTrends in Sr/Sr stable isotope scaleGeomathematical/statistical interpolationsBest-fit line of age dates versus strat. assignmentX XXX X X X X X X X XX X XXX X X X X X X XX X XXXX X XX XX X X XX X X X X X XX X X X XX X X X X X X X X X X X X X X X X X X X X X XX X X X X X X X X X X X X X X X X X X XFigure 1.3 Twelve methods in geological time scale building applied since 1937.a major improvement over a previous hour-glass methodthat tried to estimate maximum thickness of strata per periodto determine their relative duration, but had no way of esti-mating rates of sedimentation independently. As late as 1960,Holmes, being well aware of limitations, elegantly phrased itthus (p. 184):The [now obsolete] 1947 scale was tied to the ve dateslisted. . . . In order to estimate dates for the beginning andend of each period by interpolation, I adopted a modi-cation of Samuel Haughtons celebrated principle of 1878that the proper relative measure of geological periods isthe maximum thickness of strata formed during those pe-riods, and plotted the ve dates against the cumulativesums of the maximum thicknesses in what were thought tobe their most probable positions. I am fully aware that thismethod of interpolation has obvious weaknesses, but at leastit provides an objective standard, and so far as I know, noone has suggested a better one.In 1960, Holmes compiled a revised version of the age-versus-thickness scale (Fig. 1.4). Compared with the initial1913 scale, the projected durations of the Jurassic and Permianare more or less doubled, the Triassic and Carboniferous areextended about 50%, and the Cambrian gains 20 myr at theexpense of the Ordovician.1.4.2 Phanerozoic radiometric databases, statisticalscales, and compilationsW. B. Harland and E. H. Francis as part of a Phanerozoic timescale symposium coordinated a systematic, numbered radio-metric database with critical evaluations. Items 1337 in ThePhanerozoic Time-Scale: A Symposium (Harland et al., 1964)were listed in the order as received by the editors. Supple-ments of items 338366 were assembled by the Geological So-cietys Phanerozoic Time-scale Sub-Committee frompublica-tions omitted from the previous volume or published between1964 and 1968, and of items 367404 relating specically to thePleistocene most were provided by N. J. Shackleton. The com-pilation of these additional items with critical evaluationswas included in The Phanerozoic Time-Scale: A Supplement(Harland and Francis, 1971). In 1978, R. L. Armstrong pub-lished a re-evalution and continuation of The PhanerozoicTime-Scale database (Armstrong, 1978). This publication didnot include abstracting and critical commentary. These cata-logs of items 1404 and of Armstrongs continuation of itemsIntroduction 902142609011016120523525427430033037241245225406070 135 180 225 270 305 350 400 440 500 600 116CambrianOrdovicianSilurianDevonianCarboniferousLowerUpperPermianTriassicJurassicCretaceousEocenePaleoceneOligoceneMiocenePliocenePrecambrianThousands of FeetThousands of Feet500100150200250300350400450Millions of YearsMillions of Years4760 100 200 300 400 500 600 700675Figure 1.4 Scaling concept employed by Arthur Holmes in the rsthalf of the previous century to construct the geologic time scale. Thecumulative sum of maximum thicknesses of strata in thousands offeet per stratigraphic unit is plotted along the vertical axis andselected radiometric dates from volcanic tuffs, glauconites, andmagmatic intrusives along the horizontal linear axis. This version(Holmes, 1960) incorporated an uncertainty envelope from theerrors on the radiometric age constraints.404522 were denoted PTS and A, respectively, in laterpublications.In 1976, the Subcommision on Geochronology recom-mended an intercalibrated set of decay constants and isotopicabundances for the UThPb, RbSr, and KAr systemswith the uranium decay constants by Jaffey et al. (1971) asthe mainstay for the standard set (Steiger and Jaeger, 1978).This new set of decay constants necessitated systematic up-ward or downward revisions of previous radiometric ages by12%.In A Geological Time Scale (Cambridge University Press,1982), Harland et al. standardized the MesozoicPaleozoicportion of the previous PTSA series to the new decay con-stants and included a few additional ages published in Con-tributions to the Geological Time Scale (Cohee et al., 1978) andby McKerrow et al. (1985). Simultaneously, G. S. Odin super-viseda major compilationandcritical reviewof 251 radiometricdating studies as Part II of Numerical Dating in Stratigraphy(Odin, 1982). This NDScompilationalsore-evaluatedmanyof the dates includedinthe previous PTSA series. Avolumeof papers on The Chronology of the Geological Record (Snelling,1985) from a 1982 symposium included re-assessments of thecombined PTSNDS database with additional data for differ-ent time intervals.After applying rigorous selection criteria to the PTSAand NDS databases and incorporating many additional stud-ies (mainly between 1981 and 1988) in a statistical evaluation,Harland and co-workers presented A Geological Time Scale1989 (Cambridge University Press, 1990).The statistical method of time scale building employed byGTS82 and rened by GTS89 derived from the marriage ofthe chronogramconcept withthe chronconcept, bothof whichrepresented an original path to a more reproducible and ob-jective scale. Having created a high-temperature radiometricage data set, the chronogram method was applied that mini-mizes the mist of stratigraphically inconsistent radiometricage dates around trial boundary ages to arrive at an estimatedage of stage boundaries. From the error functions a set ofage/stage plots was created(Appendix 4 inGTS89) that depictthe best age estimate for Paleozoic, Mesozoic, and Cenozoicstage boundaries. Because of wide errors, particularly in Paleo-zoic and Mesozoic dates, GTS89 plotted the chronogramagesfor stage boundaries against the same stages with relative dura-tions scaled proportionally to their component chrons. Forconvenience, chrons were equated with biostratigraphic zones.The chron concept in GTS89 implied equal duration of zonesin prominent biozonal schemes, such as a conodont schemefor the Devonian. In Chapter 8, the chronogram method is10 PART I I NTRODUCTI ONdiscussed in more detail and compared with the maximumlikelihood method of interpolations using all radiometric ages,not only chronostratigraphically inconsistent ones.The Bureau de R echerches G eologiques et Mini` eres andthe Soci et e G eologique de France published a stratigraphicscale and time scale compiled by Odin and Odin (1990). Ofthe more than 90 Phanerozoic stage boundaries, 20 lacked ad-equate radiometric constraints, the majority of which were inthe Paleozoic.Three compilations spanning the entire Phanerozoic werepublished in the late 1990s. A comprehensive review of thegeologic time scale by Young and Laurie (1996) was orientedtoward correlating Australian strata to international standards,and is rich in detail, graphics, and zonal charts. Gradsteinand Ogg (1996) assembled a composite Phanerozoic scale fromvarious published sources, including McKerrow et al. (1985),Berggren et al. (1995a), Gradstein et al. (1995), Roberts et al.(1995a), and Tucker and McKerrow (1995). The Interna-tional Stratigraphic Chart (Remane, 2000) is an importantdocument for stratigraphic nomenclature (including Precam-brian), andincludeda contrast of age estimates for stratigraphicboundaries modied fromOdin and Odin (1990), Odin (1994),Berggren et al. (1995a), and individual ICS subcommissions.During the 1990s, a series of developments in integratedstratigraphyandisotopic methodologyenabledrelative andlin-ear geochronology at unprecedented high resolution. Magne-tostratigraphy providedcorrelationof biostratigraphic datumsto marine magnetic anomalies for the Late Jurassic throughCenozoic. Argonargon dating of sanidine crystals and newtechniques of uraniumleaddatingof individual zirconcrystalsyielded ages for sediment-hosted volcanic ashes with analyti-cal precessions less than 1%. Comparison of volcanic-derivedages to those obtained from glauconite grains in sediments in-dicated that the majority of glauconite grains yielded system-atically younger ages (e.g. Obradovich, 1998; Gradstein et al.,1994a), thereby removing a former method of obtaining directages on stratigraphic levels. Pelagic sediments record featuresfrom the regular climate oscillations produced by changes inthe Earths orbit, and recognition of these Milankovitch cy-cles allowed precise tuning of the associated stratigraphy toastronomical constants.Aspects of the GTS89 compilation began a trend in whichdifferent portions of the geologic time scale were calibratedby different methods. The Paleozoic and early Mesozoicportions continued to be dominated by renement of inte-grating biostratigraphy with radiometric tie points, whereasthe late Mesozoic and Cenozoic also utilized oceanic magneticanomaly patterns and astronomical tuning.A listing of the radiometric dates and discussion of specicmethods employed in building GTS 2004 can be found withinindividual chapters relating to specic geological periods.1.4.3 Paleozoic scalesThe Paleozoic spans 291 myr between 542 and 251 Ma. Itsestimated duration has decreased about 60 myr since the scalesof Holmes (1960) and Kulp (1961). Selected key Paleozoictime scales are compared to GTS2004 in Fig. 1.5a,b; historicchanges stand out best when comparing the time scale at theperiod level in Fig. 1.5a.Differences in relative estimated durations of componentperiod and stages are substantial (e.g. the Ludlow Stage in theSilurian, or the Emsian Stage in the Devonian). Whereas mostof the Cenozoic and Mesozoic have had relatively stable stagenomenclature for some decades (Figs. 1.6 and 1.7), the priorlackof anagreednomenclature for the Permian, Carboniferous,Ordovician, and Cambrian periods complicates comparison oftime scales (Fig. 1.5a; see also Chapters 11, 12, 15, and 16).The 570 through 245 Ma Paleozoic time scale in GTS89derived fromthe marriage of the chronogrammethod with thechron concept. The chron concept in GTS89 implied equalduration of zones in prominent biozonal schemes, such as aconodont scheme for the Devonian, etc. The two-way graphsfor each period in the Paleozoic were interpolated by hand,weighting tie points subjectively. Error bars on stage bound-aries calculated with the chronogram method were lost in theprocess of drawing the best-t line. The fact that the Paleozoicsuffered both froma lack of data points and relatively large un-certainties led to poorly constrained age estimates for stages;this uncertainty is readily noticeable in the chronogram/chrongures of GTS89.The 545 through 248 Ma Paleozoic part of the Phanerozoictime scale of Gradstein and Ogg (1996) is a composite fromvarious sources, includingthe well-knownscales byMcKerrowet al. (1985), Harland et al. (1990), Roberts et al. (1995a), andTucker and McKerrow (1995).The International Stratigraphic Chart (Remane, 2000) pro-vides two different sets of ages for part of the Paleozoic stageboundaries. The column that has ages for most stages appearsto slightly update Odin and Odin (1990), and Odin (1994) andis shown here.Modern radiometric techniques that are having signicantimpact for Paleozoic dates include high-precision UPb datesfrom magmatic zircon crystals in tuffs (K-bentonites) wedgedin marine strata that supercede older schemes with 40Ar39Ar,RbSr, and KAr dates on minerals like glauconite, and onHolmes(1937)Holmes(1960)Kulp(1961)GTS82Harland et al.(1982)NDS82Odin(1982)DevonianSilurianSilurianOrdovicianPermianCarboniferousDevonianOrdovicianCambrianCambrian2753133313922705350104001044010280345405425(227) (225) 2304705001560020500600245529010360104001041854951053010248590??250300350400450500550600225Ma GTS89Harland et al.(1990)GTS2004 Odin(1994)Gradstein &Ogg (1996)Young andLaurie (1996)Remane (2000)245290.0362.5408.5439.0510570245295360410435500540410435500540248.2290354417443495545 545251293354410434490250295355250300350400450500550600225Ma416444488542251299359Carboniferous286360408438505Figure 1.5(a) Comparison at the period level of selected Paleozoic time scales with GTS2004.11Holmes(1937)Holmes(1960)Kulp(1961)GTS82Harland et al.(1982)NDS82Odin(1982)GTS89Harland et al.(1990)GTS2004 Odin(1994)Gradstein &Ogg (1996)Young and Laurie(1996)Remane (2000)CarboniferousCarboniferousCarboniferousCarboniferousDevonianDevonian DevonianDevonianSilurianSilurianSilurianSilurianOrdovicianOrdovicianOrdovicianOrdovicianPermianPermianPermianCambrianCarboniferousDevonianSilurianOrdovicianPermianCambrianCambrian CambrianCambrian2753133313922705350104001044010280345405425(227) (225) 2304705001560020500600245529010/536010/540010/54185/1049510/553010Tatarian-UfimianKungurianArtinskianSakmarian-AsselianStephanianStephanian StephanianWestphalianWestphalian WestphalianNamurianNamurian NamurianVisean ViseanVisean ViseanViseanViseanTournaisian Tournaisian TournaisianTournaisian TournaisianTournaisianFamennianFamennian FamennianFamennianFamennianFrasnianFrasnianFrasnianFrasnianFrasnianGivetian Givetian GivetianGivetianGivetianEifelianEifelianEifelianEifelianEifelianEmsianEmsianEmsianEmsianEmsianSiegenianGedinnianPridoliPridoli Pridoli PridoliPridoliPridoliPridoliLudlow LudlowLudlowLudlowLudlowLudlow LudlowWenlockWenlockWenlock Wenlock Wenlock 425Wenlock WenlockLlandovery Llandovery Llandovery LlandoveryLlandoveryLlandoveryLlandoveryAshgillAshgill AshgillAshgillAshgillCaradoc Caradoc CaradocCaradocCaradocLlandeilo LlandeiloLlandeiloLlandeiloLlanvirnLlanvirnLlanvirnLlanvirnLlanvirnArenigArenigArenigArenig ArenigTremadocTremadoc TremadocTremadocTremadocTremadocMerioneth MerionethSt. David'sSt. David'sCaerfaiCaerfaiLochkovianLochkovianLochkovianLochkovianPragianPragian PragianPragianZechsteinRotliegendesGzhelian GzhelianKasimovian KasimovianMoscovian MoscovianBashkirian BashkirianSerpukhovian Serpukhov.248258263268286296315333352360367374380387394401408414421428438448458468478488505523540590245256.1290.0295.1303.0311.3322.8332.9349.5362.5367377381386390396408.5410.7424.1430.4439.0443.1463.9468.6476.1493510517.2536.0570TatarianTatarianTatarianKazanianKazanianKazanianKungurianKungurianKungurianArtinskianArtinskianArtinskianSakmarianSakmarianSakmarianAsselianAsselianAsselianEMLE E E E EMMMMMMLL LLLLLEMLLELELLE245250258265275285295305315325350360375385410415425430435445455470485500540410415430435500540-248.2252.1256260269282290296.5303311323327342354364370380391400412417419423420428443449458464470485486495500518545 545251263267270274285293298305314325344354364.5369378384399.5Pragian404.5410414425434443459467490497.5509Ufimian250272295320355345325370375380390395455465520LopingianGuadalupianCisuralianPenn-sylvanianMississippianDarriwilian-??250300350400450500550600225Ma250300350400450500550600225MaFamennianGivetianEifelianTremadocianLochkov.PragianDarriwilianChanghsing.WuchiapingianCapitanianWordianRoadianKungurianArtinskianSakmarianAsselianGzhelianKasimovianMoscovianBashkir.SerpukhovianViseanTournaisianFrasnianFrasnianEmsianPaibian411416419423428501318345326375385392398407444260266276284295304307312461468472479488513542268271299251254359MEEFigure 1.5(b) Comparison at the stage level of selected Paleozoictime scales with GTS2004. In some columns, epochs and stages arestacked together; scales 1, 2, and 3 are more detailed thanshown.12Introduction 13Holmes(1937)60708090100110120130140150160170180190200210220230240250JurassicTriassic19314510870Ma Holmes(1960)22518013570EX88Haq et al.(1987)66.5131210250DNAG83Kent & Grad-stein (1983)66.4144208245GTS82Harlandt etal. (1982)65144213248Kulp(1961)Triassic23018113563GTS2004651.5130320442455NDS82Odin(1982) 60708090100110120130140150160170180190200210220230240250Ma65.5145.5199.6251.0GTS89Harland etal. (1990)65145.6208245Odin(1994)SEPM95Gradstein etal. (1995)65 0.1144.22.6205.74.0248.24.813520524565CretaceousJurassicFigure 1.6(a) Comparison at the period level of selected Mesozoic time scales with GTS2004.Holmes(1937)60708090100110120130140150160170180190200210220230240250CretaceousJurassicTriassic19314510870MaConiacianHolmes(1960)CretaceousJurassicTriassic22518013570EX88Haq et al.(1987)Maastricht.CampanianSantonianTuronianAlbianAptianKim.OxfordianCallovianBathonianBajocianAalenianToarcianSinemurianHettangianRhaetianNorianCarnianLadinianAnisianBarremianValanginianHauterivianRyazanianVolgianTithonianBerr.OlenekianInduan66.5108113116.5121128131136145152157165171179186194201210215/217223231236240245250Pliensbach.Cenoman.DNAG83Kent and Grad-stein (1983)Maastricht.CampanianSantonianTuronianConiacianAlbianAptianTithonianKim.OxfordianCallovianBathonianBajocianAalenianToarcianPliensbach.SinemurianHettangianNorianCarnianLadinianAnisianScythianBarremianValanginianBerriasianHauterivian88.566.474.58487.59197.5113119124131138144152156163169176183187193198204208225230235240245Cenoman.GTS82Harland etal. (1982)Maastricht.CampanianSantonianConiacianTuronianAlbianAptianBarremianValanginianBerriasianTithonianKim.OxfordianCallovianBathonianBajocianAalenianToarcianPliensbach.Sinemur.HettangianRhaetianNorianCarnianLadinianAnisianSpathian-ScythianHauterivian65738387.588.59197.5113119125131138144150156163169175181188194200206213219225231238243248Cenoman.Kulp(1961)JurassicTriassicMaastricht.Campanian-SantonianConiacian-TuronianAlbian-AptianNeocomian23018113511012090847263166Bath.Bajoc.Cenoman.ValanginianGTS2004651.59511303150317842044229523952455NDS82Odin(1982)EarlyTriassicMiddleTriassicLateTriassicEarlyJurassicMiddleJurassicLateJurassicEarlyCretaceousLateCretaceous60708090100110120130140150160170180190200210220230240250MaCampanianSanton.TuronianConiac.AlbianAptianBerriasianTithonianKimmeridg.OxfordianCallovianBathonianBajocianAalenianToarcianPliensbach.SinemurianHettangianRhaetianNorianCarnianLadinianAnisianBarremianHauterivianOlenekianCenoman.65.570.683.585.889.393.599.6112.0125.0130.0136.4140.2145.5150.8155.7161.2164.7167.7171.6175.6183.0189.6196.5199.6203.6216.5228.0237.0245.0249.7InduanMaastricht.251.0GTS89Harland etal. (1990)CampanianSantonianAlbianAptianBerriasianTithonianKim.Oxford.CallovianBathonianBajocianAalenianToarcianSinemurianHettangianRhaet.NorianCarnianLadinianAnisianBarremianValanginianHauterivian74GriesbachianNammalianSpathian658386.697112124.5131.8135140.7145.6152.1154.7157.1161.3166.1173.5178187194.5203.5208209.5223.4235239.5241.1241.9243.4245Pliensbach.Maastricht.Turon.Coniac. 88.590.4Cenoman.Odin(1994)CampanianTuronianConiacianAlbianBerriasianKim.SinemurianHettangianRhaetianValanginianHauterivianSEPM95Gradstein etal. (1995)65 0.183.50.585.80.5890.593.50.298.90.6112.21.11211.41271.61321.91372.2144.22.6150.73.0154.13.2159.43.6164.43.8169.24.0176.54.0180.14.0189.64.0195.33.9201.93.9205.74.0209.64.1220.74.4227.44.5234.34.6241.74.7244.84.8248.24.8Pliensbach.Cenoman.Maastricht.154CampanianSantonianTuronianConiacianAlbianAptianBerriasianTithonianKim.OxfordianCallovianBathonianBajocianAalenianToarcianSinemurianHettangianRhaet.NorianCarnianLadinianAnisianScythianBarrem.ValanginianHauterivian967283878891108114116122130135141146160167176180187194201205220230235240245Pliensbach.Cenoman.AptianTithonianOxfordianCallovianBathonianBajocianAalenianToarcianNorianCarnianLadinianAnisianBarremianOlenekianInduan71.30.5Sant.748488899296Maastricht.65Figure 1.6(b) Comparison at the stage level of selected Mesozoic time scales with GTS2004.14Introduction 15whole-rock samples. A good review in this respect for the De-vonianis foundinWilliams et al. (2000), whose studypoints outthat it is clearly desirable to combine high analytical precisionwithnarrowbiostratigraphic control to provide the most usefulpoints for time scale calibration. These authors make a case thatthe CarboniferousDevonian boundary is near 362 Ma insteadof near 354 Ma or even younger, as shown in more recent scalesof Fig. 1.5. The same authors point out the considerable varia-tion in the estimated age for the SilurianDevonian boundaryfrom418 to 410 Ma, andsome exceptional short estimates forstage durations, such as 1 myr for the Pridoli Stage (Tuckeret al., 1998) and 0.9 myr for the Pragian Stage (Compston,2000b). The latter conicts with the analysis of cyclicity inthe limestones in the classical Devonian sections of the Bar-randian (Czech Republic), which suggests the Pragian Stageis not much shorter than the underlying Lochkovian Stage(Chlup ac, 2000).Because of the relative scarcity of reliable dates withhighstratigraphic precision, geomathematical/statistical tech-niques for direct estimation of stage boundaries are not eas-ily applicable in the Paleozoic, and various best-t line tech-niques are utilized. Tucker and McKerrow (1995, their Fig. 1)plotted selected age dates for CambrianDevonian from well-established stratigraphic levels against their fossil age in aniterative manner, juggling radiometric dates of selected sam-ples against their stratigraphic age determined by fossils suchthat a straight t was created relative to the adjusted stageboundaries.An improved version of this graphical method was em-ployed by Tucker et al. (1998) to arrive at a time line for theDevonian. First, they usedgraphical correlationplus biostrati-graphic intuition to scale the seven Devonian stages. Then, asuite of UPb zircon ages using the TIMS method for six vol-canic ashes closely tied to biostratigraphic zones were usedto adjust and calibrate this scaling. The Devonian scale inGTS2004 uses a modied version of their biostratigraphicscaling with a calibration from additional age dates (see Chap-ter 14). A similar technique is applied to the Carboniferousand Permian in GTS2004.Cooper and Sadler added a new tool to the arsenal of timescale methodology, as applied to the Early Paleozoic time scale(see Chapters 12 and 13). Using detailed graptolite sequencesfrom over 200 sections from oceanic and slope environmentbasins, a robust composite fossil sequence was calculated usingthe constrained optimization method of compositing. The Or-dovician is taken to be 44.6 myr in duration, and lasted from488.3 to 443.7 Ma; the Silurian lasted for 27.7 myr from 443.7to 416 Ma. Calculated uncertainties are relatively small.1.4.4 Mesozoic scalesThe Mesozoic time scale spans an interval of 186 myr, from251 to 65.5 Ma, which is a decrease of 60 myr since Holmes(1937) andof 35 myr comparedto the scales of Holmes (1960)and Kulp (1961). Selected key Mesozoic time scales are com-pared to GTS2004 in Figs. 1.6a,b. The geologic time scalefor the Mesozoic has undergone various improvements duringthe last two decades. The Larson and Hildes (1975) marinemagnetic anomaly prole displayed by the Hawaiian spreadinglineation was adapted for scaling of the Oxfordian through Ap-tian Stages in KG85 and SEPM95 to compensate for a paucityof isotope dates. Databases of radiometric ages have been sta-tistically analyzed with various best-t methods to estimateages of stage boundaries (GTS89 and SEPM95). Neverthe-less, there have been substantial differences in the estimatedages and durations of stages and periods among scales con-structed in the last two decades. For example, GTS1989 andSEPM95 estimated the Barremian Stage to be over 6 myr long,whereas EX88 and Odin and Odin (1993) suggested a durationof 2 myr.Age differences are particularly obvious for the JurassicCretaceous transition: the TithonianBerriasian boundary(which lacks an international denition) is 130 Ma in NDS82,135 Ma in Remane (2000), but 145 Ma in GTS89 andSEPM95, both of which excluded glauconite dates.The Jurassic scales of van Hinte (1976), NDS82, KG85,EX88, Westermann(1988), andSEPM95reliedonbiochronol-ogy to interpolate the duration of stages. As a rst approxima-tion, it was assumedthat the numerous ammonite zones and/orsubzones of the Jurassic have approximately equal mean du-ration between adjacent stages. Toarcian and Bajocian Stageshave double the number of ammonite subzones compared tothe Aalenian, so are assumed to span twice as much time. Thelimited age control on the duration of the entire Jurassic in-dicates that the average duration of each zone is 1 myr andeach subzone is 0.45 myr (e.g. Westermann, 1988). KG85and SEPM95 also took into account some intra-Jurassic agecontrol points to constrain the proportional scaling of thecomponent stages. A smoothing spline t was applied by F. P.Agterberg in SEPM95 that incorporates the error limits of theisotope age dates. At the individual subzone or zonal level,this equal-duration assumption is known to be incorrect. Forexample, McArthur et al. (2000) observed a dramatic vari-ability in Pliensbachian and Toarcian ammonite zones whenscaled to a linear trend in the 87Sr/86Sr ratio of the oceans(see Chapter 18). However, the average of the durations is notmuch off. Westermanns (1988) estimate, and application of16 PART I I NTRODUCTI ONa combined strontium trend and cycle stratigraphy to LowerJurassic stages (Weedon et al., 1999) yielded relative durationsfor the Hettangian, Sinemurian, and Pliensbachian that arewithin error limits of those of SEPM95.The advent of 40Ar/39Ar radiometric age dates on ben-tonites in local ammonite zones in a large part of the US West-ern Interior Cretaceous was a signicant improvement for LateCretaceous chronology. With this method Obradovich (1993)calibrated a Late Cretaceous time scale. He rejected all agesderived from biotites in bentonites as too young, and consid-ered all his previous KAr ages on sanidines to be obsolete.The monitor standards for 40Ar/39Ar dating have undergonerevisions during the late 1990s (see detailed discussion inChapter 6). The text of Obradovich (1993) implies that allages were normalized to a value of 520.4 Ma for the McLureMountain hornblende monitor MMhb-1, thereby requiringsignicant recalculation to the current recommendation of523.1 (0.5 myr older for Late Cretaceous ages). But in fact,Obradovich used the Taylor Creek (TC) rhyolite as an internalmonitor standard with a value of 28.32 Ma ( J. Obradovich,pers. comm. 1999), hence recalculation to the currently rec-ommended TC monitor value of 28.34 Ma is only on theorder of 0.05 myr. Correlation of the North American am-monite zonation and Obradovichs associated linear scale toUpper Cretaceous European stages and zones was partiallyachieved through rare interchanges of ammonite and othermarine macrofauna (reviewed in Cobban, 1993) and stron-tium isotope curves for portions of the Campanian and Maas-trichtian (e.g. McArthur et al., 1993, 1994). Gradstein et al.(1994a, 1995) incorporated the high-precision 40Ar/39Ar dataof Obradovich (1993); the authors applied a cubic-spline t tothe data set. An even more rened version of this analysis is thebasis for the GTS2004 scale for Late Cretaceous (see Chap-ters 8 and 19). Unfortunately, except for the basal-Turonian,it is difcult to associate the ammonite zones calibrated byObradovich (1993) with the international denitions of LateCretaceous stage boundaries.In 2000, P alfy et al. summarized 14 UPb TIMS datesfrom the Lower and Middle Jurassic of Western Canada, cal-ibrated to regional ammonites stratigraphy. Complex UPbsystematics made it difcult to obtain precise ages for someof the samples, and additional uncertainties enter when cali-brating the regional biostratigraphy to the European standardammonite zonation, but this data set provides the most impor-tant constraint on the basal-Jurassic through Toarcian stages(see Chapter 18).Cycle stratigraphy, which has become the primary methodof scaling the Cenozoic time scale, has been applied to portionsof the Triassic, Jurassic, andCretaceous time scales (reviewedinChapters 1719). As an example, Herbert et al. (1995) summa-rized orbitally tuned cycle counts using geochemical and colordata fromoutcropandcore studies innorthernandcentral Italyto estimate the duration of the Cenomanian as 6.0 0.5 myr,the Albian as 11.92 0.2 myr, and Aptian as 10.6 0.2 myr.The Cenomanian and Albian cycle-scaling results have beenveried by additional studies in Italy by Fiet et al. (2001) andGrippoet al. (2004) usingother proxies andmethods of spectralanalysis, and are within the error bars of results derived fromstatistical ts to the limited radiometric data (e.g. SEPM95).The main differences seem to be in the choice of the pin agefor hanging the cycle series from the base-Turonian or base-Cenomanian, the selected marker for the yet-to-be-denedstage and substage boundaries within the Albian and Aptian,and which orbital frequency is for tuning. This cycle scalingof the Albian events, but incorporating a potential nannofossilmarker for the AlbianAptian boundary, is used in GTS2004(see Chapter 19).1.4.5 Cenozoic scalesThe Cenozoic time scale, from65 Ma to Recent contains stagesthat vary in duration fromalmost 8 myr for the Lutetian to lessthan 1 myr for the Gelasian, and with the Holocene Epoch ofonly 11 500 yr.Although the Cenozoic Era is known in most detail, stan-dardization of stage boundaries with consensus denitions andGSSPs has been slow. In the Paleocene Period, only the epochboundaries are formally dened: base-Paleocene, base-Eocene,andbase-Oligocene. All Cenozoic standardstages are originallybased on European stratotypes, with the Neogene Mediter-ranean ones more difcult to correlate world-wide as a func-tion of increasing provincialismand diachronismin faunal andoral events in the face of higher latitude climatic cooling. Se-lected key Cenozoic time scales are compared to GTS2004 inFigs. 1.7a,b.Since 1964, whenB. F. Funnel presentedthe rst, relativelydetailed and accurate Cenozoic time scale with radiometric ageestimates, many marine time scales have been erected with aprogressive enhancement of scaling methods. Berggren (1972)and NDS82 combined radiometric age dating, stratigraphicreasoning, and biostratigraphic/geomagnetic calibrations.Hardenbol andBerggren(1978), GTS82, DNAG83, andEX88added marine magnetic reversal calibrations.Whereas the Paleozoic and Mesozoic time scales gener-ally lack a unifying interpolation method, the marine mag-netic reversals prole provides a powerful interpolator for theIntroduction 171.8522.537.553.5 53.565 6537Hardenbol &Berggren(1978)GTS82Harland etal. (1982)NDS1982Odin(1982)DNAG83Berggren etal. (1985a)EX88Haq et al.(1987)GTS89Harland etal. (1990)Berggren etal. (1995a)GTS200425.124.6243854.965 65 652334531.6 1.65.323.636.557.766.51.6525.2365466.55.2 5.223.335.456.51.855.3223.833.754.565Berggren(1972)PlioceneMioceneOligoceneEocenePaleocene1 1 116111325 254840366058636870 7032Holmes(1937)Holmes(1960)Kulp(1961)0510152025303540455055606570Ma0510152025303540455055606570Ma65.555.833.923.035.331.81PleistocenePlioceneFigure 1.7(a) Comparison at the period level of selected Cenozoic time scales with GTS2004.18 PART I I NTRODUCTI ONQuaternaryPliocenePliocenePlioceneMioceneMioceneMiocenePlioceneMioceneOligoceneOligoceneOligoceneEoceneEoceneEocenePaleocenePaleocenePaleoceneOligoceneEocenePaleocene1 1 116111325 254840366058636870 7032Pleist.Calabrian CalabrianAstianPia-cenzianZanclean Zanclean ZancleanMessinianTortonianSerravallianLanghianBurdigalianAquitanianChattianRupelianLattorfianLattorfianYpresianThanetianDanian11.845~710.515.516.52022.5303537.553.560YpresianThanetianDanian53.56065 65L.M.E.4552ChattianRupelian3237BartonianBarton.PriabonianLutetian4349PriabonianLutetian404944Holmes(1937)Holmes(1960)Kulp(1961)Hardenbol &Berggren(1978)GTS82Harland etal. (1982)NDS82Odin(1982)DNAG83Berggren etal. (1985a)EX88Haq et al.(1987)GTS89Harland etal. (1990)Berggren etal. (1995a)GTS2004Messinian Messin. Messin.Tortonian Tortonian TortonianSerravallianSerravallian SerravallianLanghianLanghian Langh.Burdigalian BurdigalianBurdigalianAquitanianAquitanian AquitanianYpresianYpresianThanetianDanianChattianChattian ChattianChattianRupelianRupelian RupelianRupelianBartonianPriabonianPriabonianPriabonianLutetianPiacenzian PiacenzianPiacenzian-HolocenePleistocene25.114.424.62432.8384250.554.960.265DanianDanianDanian65 65ThanetianThanetianYpresianYpresianBartonianBartonianBartonianLutetianLutetianLutetian232734394553591.63.4Piacenzian1.63.45.36.510.61516.62223.63036.539.843.55257.762.466.5Selandian Selandian1.653.56.310.215.216.22025.2303639.442495460.266.5Zanclean5.2Zanclean5.2MessinianMessinianTortonianTortonianSerravallianSerravallianLanghianLanghianBurdigalian BurdigalianAquitanian AquitanianYpresianYpresianChattian ChattianRupelianRupelianPriabonianPriabonianDanian DanianThanetianThanetianBartonianBartonianLutetianLutetian6.710.414.216.320.523.329.335.438.642.15056.560.51.852.63.55.327.1211.214.816.420.523.828.533.73741.34954.557.960.965Piacenz.ZancleanGelasian0510152025303540455055606570Ma0510152025303540455055606570MaDanianSelandianThanetianYpresianLutetianBartonianPriabonianRupelianChattianAquitanianBurdigalianLanghianSerravallianTortonianMessinianZancleanPiacenz.Gelas.65.558.755.837.233.928.461.748.640.423.0320.4315.9713.655.333.607.2511.601.812.59Berggren(1972)Figure 1.7(b) Comparison at the stage level of selected Cenozoic time scales with GTS2004.Introduction 19Cenozoic time scale. The large number of geomagnetic eldreversals since Late Santonian time, coupled with a wealth ofseaoor magnetic proles, and detailed knowledge of the ra-diometric age of selected magnetic polarity reversals in lavasand sediments provide a nely spaced scale. These are com-bined with a line t or cubic spline to produce spreading-ratemodels for ocean basins and an associated magnetic polaritytime scale (see Chapter 5). An excellent account of the methodand its early applications is given by A. V. Cox in Harland et al.(1982).The method itself dates back to Heirtzler et al. (1968),who selected a detailed prole in the Southern Atlantic fromanomalies 2 through 32. The only calibrated tie point was mag-netic anomaly 2Aat 3.4 Ma, based on the radiometrically datedmagnetic reversal scale of Cox et al. (1964) in Pliocene throughPleistocene lavas. Assuming that ocean-oor spreading had aninvariant spreading rate of 1.9 cm/103yr through the Cam-panian (80 Ma), ages were assigned to the main Campanianthrough Pleistocene polarity chrons. This ambitious extrap-olation has turned out to be within 10% of later interpo-lations using a more detailed composite seaoor prole, andan improved array of age-calibrated tie points (Hardenbol andBerggren, 1978, DNAG83, EX88, and GTS89).Cande and Kent (1992a,b, 1995) constructed a new ge-omagnetic reversals time scale using a composite of marinemagnetic anomalies from the South Atlantic with short splicesfrom fast-spreading Pacic and Indian Ocean segments, betterestimates of anomaly width, nine age tie points, and a cubic-spline smoothing. Using an array of biomagnetostratigraphiccorrelations with the Cande and Kent spreading model,Berggren et al. (1995a) compiled a comprehensive Cenozoictime scale.Orbital tuning has become the dominant method for con-structing detailed Neogene time scales (e.g. Shackleton et al.,1990, 1999, 2000; Hilgen, 1991a; Hilgen et al., 1995, 1997),and is making inroads in the Paleogene. These Milankovitchcycles of climate oscillations are recorded in nearly all oceanicand continental deposits, and have become a requirement forplacement of stage-boundary stratotypes within the Neogene(see Chapter 21). Among a long list of differences we men-tion that the OligoceneMiocene boundary appears 800 kyryounger, but the TortonianMessinian boundary is 120 kyrolder than in Berggren et al. (1995a). In general, the Candeand Kent (1995) geomagnetic polarity time scale for the LateNeogene is slightly too young.Cycle tuning relative to the well-dated base-Paleogene hasenabledscalingof Paleocene magnetic chrons (R ohl et al., 2001)and rened estimates of spreading rates for the South Atlanticprole (see Chapter 5). If the current pace of cycle stratig-raphy applications continues, it is quite likely that tuning toastronomical cycles will enable detailed scaling of many moresegments of the geologic time scale within the next decade.2 Chronostratigraphy: linking time and rockr. . oa+ns+rr , J. o. ooo, +n +. o. sr +nGeologic stages and other international subdivisions of the Phanero-zoic portion of the geologic scale are dened by their lower boundariesat Global Stratotype Sections and Points (GSSPs). The main crite-ria for a GSSP are that primary and secondary markers provide themeans for global correlation. GSSP theory and criteria are outlined,the status of ratied GSSPs provided, and three examples discussedof prominent GSSPs. Subdivisions of the International StratigraphicChart are summarized and illustrated.2. 1 TI ME AND ROCKGeologic time and the observed rock record are separatebut related concepts. A geologic time unit (geochronologicunit) is an abst