Stratigraphic simulations of the shelf of the Gulf of Lions: testing subsidence rates and sea-level curves during the Pliocene and Quaternary

Please download to get full document.

View again

of 9
2 views
PDF
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Document Description
Stratigraphic simulations of the shelf of the Gulf of Lions: testing subsidence rates and sea-level curves during the Pliocene and Quaternary
Document Share
Document Tags
Document Transcript
  Stratigraphic simulations of the shelf of the Gulf of Lions:testing subsidence rates and sea-level curves during thePliocene and Quaternary Estelle Leroux, 1,2 Marina Rabineau, 1 Daniel Aslanian, 2 Didier Granjeon, 3 Laurence Droz 1 andChristian Gorini 4 1 CNRS, UMR6538, Domaines Oc   eaniques, IUEM, UBO, 29280, Plouzan   e, France;  2 IFREMER, Centre de Brest, GM, BP 70, 29280,Plouzan   e, France;  3 IFP-Energies Nouvelles, 4 Avenue du Bois Pr   eau, 92852, Rueil-Malmaison, France;  4 UPMC, Univ. Paris 06, UMR7193, ISTEP, F-75005, Paris ABSTRACT Determining the relative importance of factors interacting tocontrol stratigraphic organization is a key issue in sedimentol-ogy. The Pliocene-Quaternary chronostratigraphy on the Gulfof Lions platform is still poorly constrained, giving rise to dif-ferent interpretations of the evolution of its subsidencethrough time. This paper examines the Pliocene-Quaternarysedimentary filling of the Gulf of Lion’s shelf with Dionisos, anumerical stratigraphic model. Our results show that a con-stant subsidence rate accurately reproduces the observedgeometries, whereas a varying subsidence rate reproducesthem only if the acceleration of subsidence is limited. At thistime-scale, a third-order eustatic curve is also reappraised: ahigher resolution curve (built using  d 18 O measurements) givesa more realistic restitution of our stratigraphic markers.Finally, the constant subsidence rate and sediment fluxesimplied in these modellings are discussed relative to climateand local factors of sedimentation. Terra Nova, 0, 1–9, 2014 Introduction Geological setting The Gulf of Lions (Fig. 1) passivemargin was formed by the anticlock-wise rotation of the Corsica-Sardiniablock at the Oligocene/Mioceneboundary ( ~  25 Ma) and the simulta-neous opening of a micro-ocean, theLiguro-Provenc  al basin (e.g. Zijderveljd et al. , 1970; Alvarez, 1974; Malinvernoand Ryan, 1986; Gorini, 1993). ThisWestern Mediterranean basin is anatural laboratory for climatic andstratigraphic research because of itsconnection with the Atlantic Ocean(through the Gilbraltar Strait), its largesedimentation and subsidence rates,and the exceptional dataset availablefor this basin. Moreover, its sedimen-tological record shows a very clearregional palaeobathymetric and chro-nostratigraphic marker: the MessinianMargin Erosional Surface (MES) (e.g.Ryan, 1978; Guennoc  et al. , 2000; Lofi et al. , 2011). The closure of theGibraltar Strait (Hodell  et al. , 1989;Krijgsman  et al. , 1999), and a highevaporation rate (enhanced by achange in climate), led to a drasticMediterranean sea-level fall at the endof the Miocene (Barr and Walker,1973; Chumakov, 1973; Clauzon,1973; Hs € u  et al. , 1973; Ryan, 1973),together with intensive erosion of themargin. The MES was then fossilizedduring the Zanclean transgression andsea-level high stand. This well-datedMES surface (5.33 Ma) constitutes thedeepest marker in our study. Seismic stratigraphy and boreholecorrelation This study is based on the correla-tion between an industrial seismicsurvey (LRM) and detailed strati-graphic analyses from petroleumboreholes (Cravatte  et al. , 1974) andacademic boreholes PRGL1 & 2 onthe shelf (Bern  e  et al. , 2004) (Fig. 1).The overall geometry of Pliocene-Quaternary strata shows progradingclinoforms (or prisms) with a clearchange in stacking pattern, fromLate Pliocene prograding clinoformsto prograding  –  aggrading Quaternaryclinoforms (after horizon p11,Fig. 2). Earlier studies in the areashowed similarities, but also conflict-ing seismic interpretations (Fig. 3).While the Zanclean/Piacenzian tran-sition around 3.6  –  3.8 Ma is identifiedin the same way and at the sameposition by Rabineau (2001), Lofi et al.  (2003) and Duvail  et al.  (2005)(blue p7 and red p6 horizons onFig. 3), these authors disagreedabout the stratigraphic position of the Pliocene/Pleistocene transition (at2.58 Ma after Gibbard  et al. , 2010)(blue p11 and red p7 horizons onFig. 3). Rabineau (2001) and Duvail et al.  (2005) also picked a differentreflector for the foresets. Lofi  et al. (2003) suggest attributing an ageyounger than 2.7 Ma to the reflectorslightly above blue horizon p11.The dating of Q5 at 434 000 a(MIS12) as proposed by Rabineau et al.  (2006) is now fully verifiedbased on results from the PROMESSdrillsite (Bassetti  et al. , 2008; Sierro et al. , 2009).Detailed analyses of planktonicforaminifera were performed on wells(AU, MI, TR and SI) by Cravatte et al.  (1974) and further examined byRyan (1976). The first appearance of  Globorotalia truncatulinoides  in theAutan well indicates an age youngerthan 1.86 Ma for the Q10 disconti-nuity. The stratigraphic position of the P11 discontinuity corresponds tothe appearance of   Neogloboquadrinaatlantica  in the Autan well. This Correspondence: Estelle Leroux, CNRS,IUEM, Place Nicolas Copernic, 29280Plouzan  e, France. Tel.: +33 6 82 60 71 63;e-mail: stll.leroux@gmail.com ©  2014 John Wiley & Sons Ltd  1 doi: 10.1111/ter.12091  planktonic foraminifera speciesappeared around 11 Ma for a brief period with large specimens. Thesame species, but much smaller insize, reappears between 2.72 Ma and2.41 Ma. This species is used todetect the base of the Gelasian(2.588 Ma) in marine environments(Suc  et al. , 1992; Lourens  et al. ,2004), and P11 is therefore dated atabout 2.6 Ma. The occurence of   Glo-borotalia margaritae  up to P7 pro-vides an age younger than 3.8 Mafor sediments below this discontinu-ity. Above P7, this species is nolonger observed in any wells (Crav-atte  et al. , 1974). The disappearanceof   Globorotalia margaritae  around P7therefore suggests that this strati-graphic marker can be dated toapproximately 3.7  –  3.8 Ma, the topof the Lower Pliocene (Cravatte et al. ,  1974 ); the interval P1  –  P7 isthus interpreted as an Early Plioceneinterval (5.33  –  3.8 Ma).The most significant differencebetween earlier interpretations of thesubsidence history of the Gulf of Lions surrounds the origin of thechange from prograding to prograd-ing  –  aggrading geometry, for whichdifferent scenarios concerning Plio-Quaternary subsidence on the shelf have been proposed. By measuringthe topset slopes of the Plio-Quater-nary prisms and the accommodation,Rabineau (2001) and Rabineau  et al  .(2013) determined a constant sub-sidence rate since 5.33 Ma (a sea-ward tilt of the margin reaching255 m Ma  1 at 70 km from the coast,with a rotation point 13 km upstreamfrom the present-day coastline),whereas the spatial organization of of-flap-break led Duvail  et al.  (2005) toconsider an increasing subsidence rateafter their p6 marker (red in Fig. 3),estimated at 2.6 Ma. Lofi  et al.  (2003)did not evaluate subsidence.The purpose of this study was totest these different subsidence-ratehypotheses using numerical strati-graphic modelling. Methods Dionisos is a process-based model-ling tool developed at IFP-EnergiesNouvelles (Granjeon, 1997; Granjeonand Joseph, 1999), which reproduces2D and 3D geometry with multiplelithologies in siliciclastic deltaic tooffshore environments, (Rabineau,2001; Rabineau  et al. , 2005; Jouet,2007; Csato  et al. , 2012), on theslope and in the deep-sea (Euzen et al. , 2004), and in carbonate envi-ronments (Burgess  et al. , 2006); ithas also been used to test the impactof icehouse  vs.  greenhouse-relatedsea-level curves (Somme  et al. , 2009).Applications of this model are lim-ited to large time- (thousands to mil-lions of years) and space-scales (tensto thousands of kilometres) and toprocesses that can be described by adiffusion equation with a constantsize meshing. 1˚1˚2˚2˚3˚3˚4˚4˚5˚5˚6˚6˚7˚7˚8˚8˚9˚9˚10˚10˚38˚ 38˚39˚ 39˚40˚ 40˚41˚ 41˚42˚ 42˚43˚ 43˚44˚ 44˚−3000 −2000 −1000 0 1000 2000 30001˚1˚2˚2˚3˚3˚4˚4˚5˚5˚6˚6˚7˚7˚8˚8˚9˚9˚10˚10˚38˚ 38˚39˚ 39˚40˚ 40˚41˚ 41˚42˚ 42˚43˚ 43˚44˚ 44˚ Depths relief (m) SardiniaCorsicaProvençal BasinMinorcaMajorcaMercator ProjectionNorthBalearic basin    L  i  g  u  r  i  a  n    B  a  s  i  n    P  r  o  v  e  n  c  e Algerian BasinGulf of LionsPyreneesAlpsMassif CentralFRANCESPAIN Fig. 1  Location of the study area and dataset on the bathymetric map of the Gulf of Lions. Red line: LRM18 seismic profile used in this study; black triangles: petro-leum wells. 2  ©  2014 John Wiley & Sons Ltd Plio-Quaternary stratigraphic simulations  •  E. Leroux  et al.  Terra Nova, Vol  0 , No. 0, 1–9 .............................................................................................................................................................  In this study, we present the resultsfrom two sets of runs with Dionisos: 1  We test two hypotheses for thesubsidence rate history 2  We test the influence of using a low- vs.  high-resolution sea-level curveIn both cases, sedimentary fluxesare calculated and tuned by iterationto give the best possible fit betweensimulated and real (interpreted) of-flap-breaks. Outputs of the modelare therefore always compared toseismic geometries.Further details on Dionisos andthe quantification of input parame-ters used in this study are describedin Data S1. Simulation results Subsidence The simulation using a constant rateof subsidence (250 m Ma  1 at70 km) yields a good reproduction of prograding, then prograding  –  aggrad-ing geometries (Fig. 4A). Both theheight of clinoforms and the offlap-break positions are reproduced. Thechange observed in the stacking pat-tern (progradation to pro-aggrada-tion) is not linked to a suddenchange in subsidence, but to thecombination of ongoing tilt of themargin together with variations insediment fluxes and sea-level duringthe rebuilding of the shelf.Very small or negligible subsidenceduring the Pliocene did not repro-duce the observed height of theclinoforms (see Fig. 4B2  –  B3). Astrong increase in subsidence duringthe Quaternary cannot restituteclinoform geometries (the position of Q5, for example, is too deep). Theonly good restitution of observedgeometries (Fig. B1) is reached witha Pliocene (5.32  –  2.6 Ma) subsidence 1.4001.1001.2000.900 1.000 0.7000.8000.4000.5000.6000.100 2.000 1.7001.8001.5001.600 0.000 Offset: SP: 70000 m 1000.0 2000.0 3000.0 20000 m 50000 m 60000 m30000 m 40000 m10000 m 0.2000.3001.3001.900 MES (Messinian Erosional Surface) MIOCENE PLIO-PLEISTOCENE LRM18     T    i   m   e    (   m   s   t   w   t    ) Vertical Exageration = X5 P7 (3-6-3.8 My)P11 (2.6 My)Seismic chronostratigraphic markers :  6000 m p1 ProgradationProgradation + aggradation Vertical Exageration = X5 Fig. 2  Interpretated LRM18 seismic reflection profile. The offlap-breaks are represented by white dots; the first Pliocene prism(p1) is highlighted (yellow). Inset: the overall geometry of Pliocene  –  Quaternary strata shows prograding clinoforms (or prisms)with a clear geometrical change in the Late Pliocene to Quaternary clinoforms (after yellow horizon p11) from essentially pro-grading (dark green) to prograding  –  aggrading (light green). 2500240023002200210020001900180017001600150014001300120011001000900800700600500400300200100 0    D  e  p   t   h  s   (  m   ) 0 500 1000 1500 2000 2500 3000 3500 4000 Shot points ChronohorizonsRabineau (2001) with offlap-breaks ( )Duvail et al. (2005) with offlap-breaks ( )Lofi et al. (2003) 3    . 6    -  3    . 8     M    a    3 .8  M  a  3    . 8     M    a    2     . 6      M      a     y o ung e r  t ha n 2 .7  M a  2  .7   M   a   0.45 M a1 M a p4p7p11q10q5p1p2p3p4p7p3p5p6 Offset 10 km 30 km 40 km 50 km 60 km 70 km0 km 20 km Fig. 3  Synthesis of LRM18 seismic interpretations on the shelf of the Gulf of Lionsby different authors. Interpretation from Rabineau (2001) in blue; interpretationfrom Lofi (2002) and Lofi  et al.  (2003) in green; interpretation from Duvail  et al. (2005) and Duvail (2008) in red. The interpreted chronostratigraphic markers havebeen digitized, time-depth converted and superimposed on the same vertical sectionwith respective age estimates. ©  2014 John Wiley & Sons Ltd  3 Terra Nova, Vol  0 , No. 0, 1–9 E. Leroux  et al.  •  Plio-Quaternary stratigraphic simulations ............................................................................................................................................................  22002100200019001800170016001500140013001200110010009008007006005004003002001000    D  e  p   t   h  s   (  m   ) 0 500 1000 1500 2000 2500 3000 3500 4000 02500240023002200210020001900180017001600150014001300120011001000900800700600500400300200100    D  e  p   t   h  s   (  m   ) 0 500 1000 1500 2000 2500 3000 3500 4000 Fictive wells p4, p7, p11, q10 are located on the vertical of the offlap-breaks corresponding to the major surfaces identified by Rabineau, 2001. They show the sand proportions. 02500240023002200210020001900180017001600150014001300120011001000900800700600500400300200100    D  e  p   t   h  s   (  m   ) 0    . 0    M   a  0    . 5   M   a   1 Ma1.5 Ma 2  Ma 2  . 2  M  a  2.7 Ma2.9 Ma 3  . 3  0   M   a   4  . 5  M  a  5             .  3             M             a           5 Ma 2   . 6     M    a    2.5 Ma 3  . 8  M  a  3 Ma 3   . 6   M   a  3   . 7    M   a   02500240023002200210020001900180017001600150014001300120011001000900800700600500400300200100    D  e  p   t   h  s   (  m   ) 0 .0  M a  0  . 5  M  a  3  . 3  0   M   a   4  . 5  M  a  5 Ma 2   . 6    M    a    3  . 8  M  a  3   . 6   M   a  3   . 7    M   a   2 .9  M a  2 M a  1  M  a   2  . 2  M  a  1 . 8  M  a 1 . 2  M  a  1 . 5 M a  Fictive wells p3, p4, p5, p6 and p7 are located on the vertical of the offlap-breaks corresponding to the major surfaces identified by Duvail et al.,  2005. They show the sand proportions.Depths of seismic reflectors (grey) are conpared to simulated depths of chronostratigraphic markers (coloured) on each vertical sections (on theright). Vertical exageration = X20 for all figures hereSand proportion in the fictive wells(100% = 1) Legend: Eustatic curve and time markers Shot points 250 m/Ma at 70 km from the present-day coast line 200 m/Ma from 5.3 Ma to 2.6 Ma,300 m/Ma from 2.6 Ma to Present-Day) (B2)  = 100 m/Ma from 5.3 Ma to 2.6 Ma,400 m/Ma from 2.6 Ma to Present-Day) (B3)  = null from 5.3 Ma to 2.6 Ma,300 m/Ma from 2.6 Ma to 1.5 Ma665 m/Ma after 1.5 Ma Shot pointsShot points the height of simulated and observed seismic clinoforms  Deposits paleobathymetry (A)  CONSTANT SUBSIDENCE RATE (B)  VARYING SUBSIDENCE RATE Offset (km) 10 20 30 40 50 60Offset (km)7000 10 20 30 40 50 60 70Offset (km) 10 20 30 40 50 60 700 Offset (km) 10 20 30 40 50 60 700 0 500 1000 1500 2000 2500 3000 3500 4000 (B1)  = Shot points 0 500 1000 1500 2000 2500 3000 3500 4000 Fig. 4  Simulations testing subsidence rates with a 100 000-year time-step (modified from Leroux, 2008, 2012). Comparison of LRM18 seismic profile interpretations (grey lines) with simulated geometries and deposit bathymetries predicted by Dionisos(coloured solid lines). The simulations are run with a third-order sea-level curve (Haq  et al. , 1987). (A) Simulation with a con-stant subsidence rate of 250 m Ma  1 at 70 km from the coast, then (B1, B2, B3) using increasing subsidence rates. Note thatthe change to an overall prograding  –  aggrading pattern occurs without a change in the subsidence rate. The observations couldnot be reproduced when the increase in subsidence is too strong or with no subsidence between 5.3 Ma and 2.6 Ma (B2 andB3). Note also that, in both cases, the observed height of the Early Pliocene clinoforms could not be reproduced. Once the hor-izontal positions of the offlap-breaks along the seismic profile have been adjusted, their simulated depths are too deep (see thehighlighted yellow area on the right). In such a subsidence scenario, the accommodation created after 2.6 Ma increases toomuch the topset slopes of the clinoforms already deposited. Note on the fictive wells that sand is deposited at the shelf-break,whereas finer sediments (silt and shale) are deposited on the shelf and on the slope (with a downstream decreasing granulome-try). 4  ©  2014 John Wiley & Sons Ltd Plio-Quaternary stratigraphic simulations  •  E. Leroux  et al.  Terra Nova, Vol  0 , No. 0, 1–9 .............................................................................................................................................................  rate higher than 200 m Ma  1 , fol-lowed by a subsidence rate lowerthan 300 m Ma  1 . Eustatism The sea-level curve used as an inputparameter is also an important factorin stratigraphic simulation. Figure 5shows how the frequency and ampli-tude of eustatic cycles strongly influ-ence the geometries of the simulateddeposits. In both cases, the subsi-dence rate and sedimentary fluxremain constant.The first case (A) represents a simu-lation using the sea-level curve of Haq et al.  (1987). The offlap-break trajec-tory line appears segmented, withalternating seawards and landwardsmigrations (progradation/retrograda-tion) following the five major eustaticcycles of the curve. The third-ordereustatic curve of Haq  et al.  (1987)emphasizes large-scale amplitudes of sea-level fluctuation and favours ero-sion of prior transgressive systemtracks during major regressions. Thesimulation does not reproduce theobserved geometries (compareFigs 5A and 2).The second case (B) represents asimulation using a higher resolution d 18 O-derived sea-level curve (Lisieckiand Raymo, 2005). The offlap-breaktrajectory appears more continuous,with smaller shifts of the offlap-breakdue to weaker amplitude eustaticvariations and a clearer progradingto prograding  –  aggrading stackingpattern as observed in the seismicdata (compare with Fig. 2). The highfrequency  d 18 O data also distinguishindividual regressive/transgressivecycles (and erosional surfaces) similarto those observed in the Quaternary(Rabineau  et al. , 2005; Jouet, 2007). Sediment flux and water discharge Figure 6 presents the sediment fluxesand water discharge used in the sim-ulations for the two sea-level curvestested. In both cases, Early Pliocenefluxes are relatively small (18  –  25 km 3 Ma  1 ), whereas Late Plio-cene and Quaternary fluxes arehigher (95  –  100 km 3 Ma  1 ), whichimplies a considerable increase insediment supply around 3.8 Ma. Thewater discharge evolves in the sameway: after 3.8 Ma, a three-foldincrease is observed (from 100 to300 m 3 s  1 ). Discussion The first result of our simulations isthat a constant subsidence rate dur-ing the entire Pliocene  –  Quaternaryreproduces the geometry of sedimen-tary filling, in particular the positionsof Pliocene offlap-breaks (p1  –  p7),which follow a more or less straightline, as well as the change in stackingpattern after p7, with prograding  –  ag-grading clinoforms. The secondresult is that a strong increase in sub-sidence (during the Quaternary) can-not reproduce the clinoform’sgeometries, which tends to falsify thishypothesis.In fact, the amount of space cre-ated on the inner shelf (at the posi-tion of p1  –  p7) is smaller than thatcreated on the outer shelf because of the global seaward tilt of the margin.There is no need to infer a variationin subsidence rate through time, butthe subsidence does vary spatially.Therefore, a regular ongoing process(the tilt of the margin) in conjunctionwith sea-level and sediment-supplyfluctuations can produce an impor-tant change in the resulting geome-tries of the margin (pro to pro-aggrading pattern).Note that our subsidence calcula-tion concerns only ‘total’ subsidence,and we did not aim to decipher therelative roles of tectonics, thermalsubsidence and loading of sediments.Although the load of sediments is animportant factor controlling the sub-sidence on margins (e.g. Watts andRyan, 1976), it is not the primaryfactor creating space for sedimenta-tion.Finally, independent higher resolu-tion studies of the Upper Quaternarymeasured a similar subsidence ratefor the last 434 000 a of 240 m Ma  1 at 70 km (Rabineau et al. , 2005, 2006, 2007). This valueis fully confirmed by data from thePromess-1 drilling sites PRGL2 (Bas-setti  et al. , 2008) and PRGL1 (Sierro et al. , 2009).Our best-fit simulation implies anincrease in sediment and water fluxesat 3.8 Ma. The implied evolution of water discharge is consistent with cli-mate simulations for the Piacenzianinterval indicating high precipitationin north-western Europe (Jost  et al. ,2009). The  d 18 O record for the last5 Ma indicates a step-by-step cli-matic deterioration that testifies tothe transition from a warm Messini-an to a colder Pleistocene climate(Shackleton  et al. , 1995; Lisiecki andRaymo, 2005). Major cooling eventshave occurred since the Lower Plio-cene, as attested by isotopic and pal-ynologic signals (Suc  et al. , 1995;Popescu  et al. , 2010) with short-termand low-amplitude temperature fluc-tuations. The first cooling event,recorded around 3.5 Ma, corre-sponds to a change in the extent of continental ice in the NorthernHemisphere (Lear  et al. , 2000; Za-chos  et al. , 2001), and is also associ-ated with the extinction of planktonic foraminiferas (Rio  et al. ,1990) and a major sea-level fallnamed the TB 3.4  –  3.5 cycle by Haq et al.  (1987). This cooling episode atthe Zanclean/Piacenzian transitionmay imply increased and/or moreefficient glacial erosion on land andincreased terrigenous sedimentationin the basin. As Clauzon (1987)showed, during the Early Pliocene,detrital sediments were trapped on-land in the form of Gilbert deltas.We infer that these Gilbert deltaswere first completely filled onshore,which then allowed downstreamdelivery of sediments to the shelf (with an increase in offshore fluxes).This is consistent with the well-known marine  –  continental transitionand the Pliocene abandonment sur-face preceding the earliest glaciationat 2.6 Ma, observed on present-dayonshore outcrops (Clauzon, 1996;Clauzon  et al. , 1996).Other relevant climatic eventsoccurred during the Quaternary,notably around 2.6 Ma, with theemergence of glacial/interglacialcycles in the Northern Hemisphere(Shackleton  et al. , 1995; Zachos et al. , 2001) and increased floatingice in the North Atlantic (Backman,1979). More rapid climatic variationsand higher frequency sea-level fluctu-ations occured after that period. Aworldwide transition, called the ‘MidPleistocene Revolution’, reflects afundamental change around 0.8  –  0.9 Ma in climatic cyclicity, from adominant 41 ka cycle to a dominant100 ka cycle, with increased ice vol-ume (Head and Gibbard, 2005). ©  2014 John Wiley & Sons Ltd  5 Terra Nova, Vol  0 , No. 0, 1–9 E. Leroux  et al.  •  Plio-Quaternary stratigraphic simulations ............................................................................................................................................................
Similar documents
View more...
Search Related
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks