Word Search

What drove sea-level fluctuations during the mid-Cretaceous greenhouse climate?

Description
The major states, in which Earth's climate operates, i.e., icehouse, greenhouse and hothouse, are epochs of tens of millions of years. These states set long-termboundary conditions that need to be considered for climate and sea level
Categories
Published
of 8
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.
Related Documents
Share
Transcript
  What drove sea-level  fl uctuations during the mid-Cretaceousgreenhouse climate?  Jens E. Wendler ⁎ , I. Wendler Department of Geosciences, University of Bremen, 28334 Bremen, Germany a b s t r a c ta r t i c l e i n f o  Article history: Received 6 March 2015Received in revised form 28 July 2015Accepted 21 August 2015Available online xxxx Keywords: Climate modesSequence stratigraphyAquifer-eustasyGlacio-eustasyHydrological cycleStable carbon and oxygen isotopes Themajorstates,inwhichEarth'sclimateoperates,i.e.,icehouse,greenhouseandhothouse,areepochsoftensof millionsofyears.Thesestatessetlong-termboundaryconditionsthatneedtobeconsideredforclimateandsea-level interpretations. This paper summarizes the conceptual models for hydrological cycling derived from thecharacteristicsofthesethreeclimatestates.Whileglacio-eustaticforcingofsea-levelchangesundericehousecli-mate conditions is fairly well understood, the drivers of eustatic sea-level  fl uctuations under greenhouse condi-tionsremainenigmatic.Thislackofunderstandingmayberelatedtoincoherenciesinthecurrentideasabouttheimpact of accelerated hydrological cycling on sea level under greenhouse climate conditions.As an example for a greenhouse climate, we review evidences that link proxies for climate and sea level for theintenselystudied,butcontroversiallydiscussed,mid-Cretaceoussea-levelhistory.Basedonsequencestratigraphyand a recently published high-precision timescale, we demonstrate that the late Middle Turonian Pewsey  δ 13 Cisotope maximum represents a major transgression, not a regression as previously stated, which con fl icts withthe interpretation of a co-occurring  δ 18 O maximum to re fl ect a short glacial episode. This contradiction can besolved by the concept, presented here, that dominance of aquifer-eustasy characterized sea-level forcing duringthe Turonian greenhouse climate, despite a possible, though contentious, sporadic presence of minor ice sheets.The effects of temperature and ice volume both lead to a pronounced  δ 18 O carb  maximum during glacio-eustaticregressions. In contrast, the opposing effects of temperature and groundwater volume on oxygen-isotope frac-tionation lead to a  δ 18 O carb  maximum during aquifer-eustatic transgressions. We suggest that, throughout Earthhistory,bothaquifer-eustaticand glacio-eustaticforcingformeda combinedsea-levelresponse, withdominanceofaquifer-eustasybeingtypicalforthegreenhouseclimatemode.Duringtheicehousemode,aquifer-eustasyap-parently remains active as a background process, but is outpaced by the glacio-eustatic effect.© 2015 Elsevier B.V. All rights reserved. 1. Introduction Earthexperiencesgeneralclimatestateswithanapproximatecyclicityof 150 Myr(Frakesetal.,1992). Theconceptofclimatestates (ormodes,epochs) mainly srcinates from Fischer (1982) and has been re fi nedrecently by Kidder and Worsley (2010, 2012). Given the highly dynamic natureofclimateprocesses,itisfascinatingthatsuchlong-periodgeneralcharacteristicscanactuallybefoundintheclimateproxydata.Clearly,theclimate modes cannot be taken as quasi-constant warmth- or cold-periods of tens of millions of years; there must be  fl uctuations, but inwhich range? One question is, to what extent the main climate statesmayswitchfromonemodetoanotherwithinshorttimeranges,particu-larly at the ~1-Myr scale climate beat that seems responsible for 3rd-ordersea-level fl uctuations(withmagnitudesof20 – 75m)andfortheas-sociated sequence stratigraphic cycles? Or shouldn't these sea-level  fl uc-tuations be possible within either greenhouse or icehouse mode,without rapidly switching between these climate modes?Obviously, sea-level  fl uctuations at large magnitude are possiblewithin the icehouse mode. With two polar ice caps in the Holocene,ice expansion during glacial times can create apparently suf  fi cientwater storage to explain over 100 m of sea-level change. This resultedin the tendency to attribute any signi fi cant sea-level change of 3rd- orhigher order (100 kyr – 2 Myr) to the waxing and waning of ice sheetsand to assume that this is the only possible mechanism. The ice-centeredviewcomeswiththecaveatthatothersea-level-forcingmech-anisms are underestimated, because their impact cannot be distin-guished from today's large ice-sheet effect. However, the latter hasinherent uncertainties in mass calculation, particularly for the glacialice extent (exact ice-sheet area and thickness), that are in the samerange of the water volume actually needed to drive tens of meters of sea-level change. Therefore, it would be dif  fi cult to constrain properlyanywater-storage processother thaninice sheetsthat mightaddition-ally have been in action during the Holocene.There are two drivers that apparently have equal potential to force3rd order sea-level changes: glacio-eustasy (e.g., Miller et al., 2005)and aquifer-eustasy (Hay and Leslie, 1990). Today groundwater andice each equally yield ~1 – 2% (20 – 30 × 10 6 km 3 ) of the Earth's total Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx – xxx ⁎  Corresponding author. E-mail address:  wendler@uni-bremen.de (J.E. Wendler). PALAEO-07428; No of Pages 8 http://dx.doi.org/10.1016/j.palaeo.2015.08.0290031-0182/© 2015 Elsevier B.V. All rights reserved. Contents lists available at ScienceDirect Palaeogeography, Palaeoclimatology, Palaeoecology  journal homepage: www.elsevier.com/locate/palaeo Please cite this article as: Wendler, J.E., Wendler, I., What drove sea-level  fl uctuations during the mid-Cretaceous greenhouse climate?,Palaeogeogr. Palaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.2015.08.029  (free) water (Trenberth et al., 2007). Considering that the active porevolume of groundwater aquifers during the Cretaceous is estimated to~40 × 10 6 km 3 (Hay and Leslie, 1990), the amount of water(10 – 20 × 10 6 km 3 ) that is equivalent to 10 – 40 m of sea-level changecould easily be contained within aquifers (for more detail seeSection 2 in Wendler et al., this volume-a). The similar potential of  aquifer- and glacio-eustasy to drive 3rd order sea-level changes raisesthe question: Do both processes act simultaneously and if so, wouldtheyadd toeachother's effectonsea-levelchangeor would they coun-teract?Thisquestionismuchthesameforicehouseandgreenhousecli-matestatesalike,butforwarm greenhouseclimateperiods,suchasthemid-Cretaceous,icesheetsareconsideredtohavebeenephemeral(e.g.,Miller et al., 2008) or non-existent (Moriya et al., 2007; MacLeod et al.,2013).Accordingly,analternativedriverisessentiallyneededtounder-stand the continuous, cyclic sea-level  fl uctuations that are evidencedfrom the sedimentary record (e.g., Gale et al., 2002; Haq, 2014;Wendler et al., 2014; Wendler et al., this volume-b).Interpretation of the causes for sea-level changes during the mid-Cretaceous is particularly challenging, because this period was tempo-rallycharacterizedbyhothouseconditionsthatappeartobecompletelyincompatible with polar ice (Hay and Floegel, 2012; Kidder andWorsley, 2012; MacLeod et al., 2013). Nevertheless, this period hasbeen associated with sporadic glacio-eustatic events. However, if theywere presentatall, sporadic  “ cold snaps ” areinsuf  fi cient tohaveforcedcyclic 3rd-order sea-level  fl uctuations of several tens of meters, forwhich continuous and cyclic build-up and -decay of ice sheets wouldhavebeenrequired.Sincesuchsigni fi cantamountsofice(halfoftoday'sicesheetonAntarctica:~12×10 6 km 3 )areunlikelyunderwarmgreen-house conditions, aquifer-eustasy (Wendler et al.,2011; Wendler et al.,thisvolume-b)wasinvokedasanalternativedriver(withfocusonlakelevels, the process was also called limno-eustasy by Wagreich et al.,2014).Apotentialinterplayofglacio-andaquifer-eustasyintheLateCreta-ceous should be expressed in temperature and sea-level records. Withthisrespect,thequestionneedstobesolvedwhether δ 18 Omaximadur-ing greenhouse conditions really re fl ect cooling that coincides with asea-levelfall or if the isotopedata and sedimentary patternscan beun-derstoodfromanalternativeperspective?Inthepresentpaper,we fi rstsummarizetheknowncharacteristicsofEarth's climatemodes anddis-cuss glacio-eustasy and aquifer-eustasy in the context of the boundaryconditions of these modes. We then go on to focus on the Middle toLate Turonian sea-level  fl uctuations and temperature records and at-tempt to disentangle the existing controversy in their interpretation inthe light of an interplay of the two possible drivers of eustasy. 2. The climate modes Theconceptoflong-termclimatemodesofEarthhasreceivedrecentcomprehensiveattention by KidderandWorsley(2010,2012) whode- fi ned the three basic conditions of the climate system, icehouse, green-house and hothouse.The greenhouse, divided into cool and warm greenhouse, can beconsideredthedefaultstatethatcharacterizesover70%ofthePhanero-zoic.Thepole – equator thermal contrast and planetarywindbelt veloc-ity are reduced, but tropical cyclones strengthen and extend to highlatitudes.Relevantforsealevel,theincreasedpolewardmoisturetrans-portduringgreenhouseconditions(KidderandWorsley,2010)isrelat-ed to cloud cover that promotes heat-trapping, thus preventing iceformationbutdeliveringincreasedpolarrainfall(Fig.1A).Inturn,runoff ishighandpotentialpolaraquiferstoragescanbecharged,becausetheyare not covered by permanent ice or locked due to permafrost. Forma-tionofseasonalicedoesnotaffectsealevelbecauseitdoesnotaccumu-late ice storage. The cool greenhouse can have small polar ice caps andAlpine glaciers while the warm greenhouse may only have seasonalice. Cool greenhouse polar ice and a latitudinal temperature gradient,that is somewhat larger than in thewarm greenhouse,could be relatedtocertainorbitalconditions(obliquitynodes)whensummermeltingisnotcomplete,sothaticesheetscanbuildupovertime.Thus,regressionsthatcoincidewithmaximain δ 18 Ovaluesmaybeexplainedbybuild-upof ice sheets under cool greenhouse conditions, e.g., during part of theLower Cretaceous. However, build-up of ice sheets during obliquitynodes was hypothesized to explain extreme sea-level lowstands alsoduring the warm greenhouse climate of the mid-Cretaceous (Kuhntetal.,2009).Moreover,somepositive δ 18 Oshifts,e.g.,duringtheMiddleTuronian,donotcorrespondtoasea-levelfallandmeritfurtherdiscus-sion, as explained in Section 3.The icehouse mode requires forcing. It has extensive polar ice capsandcirculationisinastrongthermalmode(modern-stylethermohalinecirculation). The icehouse mode is triggered by continent – continentcollision orogeny, associated with strong silicate weathering thatabsorbs CO 2  and stimulates cooling (Kidder and Worsley, 2010). Ahigh pole – equator temperature contrast causes increased planetarywind belt velocity and tropical cyclones are restricted to low latitudes.Relevantforsealevel,theseconditionscauselessstratosphericwaterin- jection, decreased poleward water transport and reduced heat transfer,ascomparedtothegreenhousemode.Arealicecoverageandpermafrostare expanded and eliminate potential polar aquifer storage (Fig. 1B).Therefore,theeffectofwaterstorageincontinentalicesheetsdominatessea-level  fl uctuations during the icehouse mode.With regard to hot climate extremes, the greenhouse mode can beforced towards a hothouse, also termed HEATT (Haline Euxinic AcidicThermal Transgression) by Kidder and Worsley (2010). Large igneousprovince(LIP)volcanismisconsideredthemainforcingofthehothousemode. Voluminous CO 2  emission causes a series of events that initiatepronounced haline-mode circulation with warm, saline bottom watersthat are produced at expanded arid subtropical and middle latitudes.With respect to sea level, ocean warming results in thermal expansionof ocean water and rapid transgression associated with the onset of aHEATT. The Oceanic Anoxic Event 2 (OAE2, ~94 Ma) is considered torepresent such a period. Hothouse modes are typically not sustainedfor longer than about 1 million years.Besidetheboundaryconditionsandtrigger-prerequisitesinternaltothe Earth's climate system, the climate modes have been shown to fol-low long-term cyclicity of ~135 ± 9 Myr that appears to correlatewith cosmic ray  fl ux of a similar periodicity of 143 ± 10 Myr (Shavivand Veizer, 2003; Wendler, 2004). From this observation, long-termcosmic ray  fl ux was hypothesized to control  fl uctuations of cloudcover on the scale of changes between climate modes, thereby lendingthe hydrological cycle a potential role as a thermostat. Such stabilizingnegative feedback would also suggest that switching between climatemodes is limited, so that short-term occurrences of an icehouse statewithin a greenhouse mode, and vice versa, seem to be impossible.In summary, the warm greenhouse mode that Earth experiencedduring the Albian – Santonian had little or no ice. Once in a greenhousemode, the climate system can divert from it only within certainamplitudinalandtemporallimits.WhileLIPvolcanism(astheprerequi-sitetoforceahothouse)hasarelativelyabruptonsetandterminationinthemillion-yearscale(orless),collision-orogenyasaprerequisiteoftheicehousemodeisalong-termprocessthatcannotbeswitchedonandoff atarelativelyshort-termscaleoffewmillionyears.Thus,sea-level fl uc-tuationswithinthegreenhouse climatemode mayeither re fl ectswingsbetween the cool and warm greenhouse states or represent processesother than glacio-eustasy. In the following, we compare the climatemode characteristics and sea-level reconstructions for an exempli fi edstratigraphic level during the Turonian greenhouse climate, in order todisentangledivergingconceptsontherelationbetweenclimateproxies,reconstructed oceanographic parameters and sea-level response. 3. Sequence stratigraphy andstable-isotope data fromthe Turonian Two Middle to Late Turonian maxima in  δ 18 O records, coeval withthe positive Pewsey and Hitch Wood carbon isotope excursions (CIEs), 2  J.E. Wendler, I. Wendler / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx –  xxx Please cite this article as: Wendler, J.E., Wendler, I., What drove sea-level  fl uctuations during the mid-Cretaceous greenhouse climate?,Palaeogeogr. Palaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.2015.08.029  have been related to sea-level fall and potential build-up of ice sheets(Stoll and Schrag, 2000; Voigt and Wiese, 2000; Bornemann et al.,2008). However, improved stratigraphic constraints ( Jarvis et al.,2006; Wendler et al., 2010; Richardt and Wilmsen, 2012; Janetschkeand Wilmsen, 2014; Uli č ný et al., 2014; Wendler et al., 2014) supportthe opposite interpretation of the same intervals as sea-level maxima(Fig. 2), as was srcinally proposed by Gale (1996). The controversy on the existence of a greenhouse glacial episode centers on thePewseyCIE interval that is positioned between two major Turonian sequenceboundaries (SB), termed Tu2 and Tu3. These SBs were srcinally datedwith Tu2 at 91.2 Ma and Tu3 at 90.9 Ma (Hardenbol et al., 1998).Detailed sequence stratigraphic research indicates that Tu3 (KTu4 of Haq, 2014) is a signi fi cant and globally recognized 3rd-order SB nearthe Middle/Late Turonian boundary (Bauer et al., 2003; Buchbinderet al., 2000; Frijia et al., 2015; Gale, 1996; Hardenbol et al., 1998;Niebuhr et al., 2011; Niebuhr et al., 2014; Richardt and Wilmsen, 2012;Robaszynski et al., 1990; Sageman et al., 1997; Schulze et al., 2003,2004, 2005; Simmons et al., 2007; Uli č ný et al., 2014; Voigt andHilbrecht, 1997; Walaszczyk et al., 2013; Wendler et al., 2010; Wendleret al., 2014; Wiese, 2009; Wilmsen and Nagm, 2013). According toWilmsen and Nagm (2013), SB Tu2 of  Hardenbol et al. (1998) is more than 300 kyr older than SB Tu3, and it corresponds to a major MiddleTuronian SB that is recorded globally ~1 Myr before Tu3 (Bauer et al.,2003; Buchbinder et al., 2000; Gale, 1996; Niebuhr et al., 2011; Richardtand Wilmsen, 2012; Schulze et al., 2003, 2005; Wendler et al., 2010;Wendler et al., 2014; Wiese, 2009; Wilmsen and Nagm, 2013).ThePewsey δ 13 Cmaximumisrelatedtoasea-levelmaximumintheEnglish Chalk sections (Gale, 1996; Jarvis et al., 2006) and its eustaticcharacter was con fi rmed by corresponding  fl ooding of the RussianPlatform (Sahagian et al., 1996; Jarvis et al., 2006). A relationshipbetween long-term trends in sea level and  δ 13 C values has been sug-gested repeatedly (Arthur et al., 1987; Berger and Vincent, 1986; Jarvis et al., 2002, 2006; Jenkyns, 1996; Scholle and Arthur, 1980;Voigt and Hilbrecht, 1997; Weissert et al., 1998; Wendler, 2013).This positive correlation is thought to srcinate from  12 C depletionof seawater through enhanced C org  burial on expanded shelf seaareas during long-term sea-level highstand and  12 C release fromC org  exhumation during regression. It must be kept in mind thatfeedback loops in the global carbon cycle cause the relationshipbetween sea-level and  δ 13 C records to be more complex for high-frequency cycles, and a clear correlation with  δ 13 C records was dem-onstrated only for sea-level cycles  N  400 kyr (Uli č ný et al., 2014).Additionalevidenceforpredominantlyhigheustaticsealevelduringthe Pewsey interval comes from sequence stratigraphic investigationsof sections in Germany (Richardt et al., 2013) and in Jordan (Wendler et al., 2014) that show a major transgression, positioned betweentheir SB Tu2 – SB Tu3 and TuJo2 – TuJo3 (=sequence S7 of  Schulze et al.,2003), respectively (Fig. 2). Furthermore, in the Western Interior Sea- way, a major transgression and highstand span the whole MiddleTuronian (Sageman et al., 1997). Likewise, this stratigraphic intervalin Site TDP 31from Tanzania (Wendler et al., this volume-a) is thoughtto represent a sea-level highstand. Thus, a 3rd-order high in global sea Fig. 1. Schematicviewof greenhouseandicehouse climatemodes (basedonKidderandWorsley,2010,2012) inrelation to sea-level change.The scheme illustratesthemaincharacter-isticsofthehydrologicalcycleundergreenhouseandicehouseclimateconditionsandtherespectiveconsequencesforwaterstorageonlandandforsea-levelchanges.Aquifer:dynamicstorage ofliquidwaterin terrestrial surface and subsurfaceaquifers( fl uvial back fl ow to ocean not illustrated; for balance of water fl ows see Fig. 3 in Wendler etal.,this volume-b); Ice:water storage in continental ice sheets. Note that the glacio-eustatic and aquifer-eustatic effects counteract, and that aquifer-eustasy dominates during the greenhouse climate mode,while glacio-eustasy dominates during the icehouse climate mode. The inferred net effect on seawater  δ 18 O values is indicated: upward arrow = increasing values, downwardarrow = decreasing values (for explanation see Section 4). A) Greenhouse mode. Aquifer storage potential is high due to 1) reduced coverage with continental ice and permafrost inhigh latitudes, creating larger areas with active aquifers, 2) enhanced water transfer from low to high latitudes, and 3) latitudinally expanded cyclonic activity. Seasonal ice can occur,but minor ice sheets may only build up upon lacking summer melting associated with nodes in long-term orbital cycles (low seasonality). B) Icehouse mode. Aquifer storage potentialismuchlowerthaninthegreenhousemodedueto1)decreasedhydrologicalcycling,2)coverageofaquifersinhighlatitudesbyiceandpermafrost,and3)restrictedlatitudinalexpandof cyclonic activity. Abbreviations: ITCZ = intertropical convergence zone; AB = arid belts; HZ = humid zones (warm and temperate); S/N = south/north.3  J.E. Wendler, I. Wendler / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx –  xxx Please cite this article as: Wendler, J.E., Wendler, I., What drove sea-level  fl uctuations during the mid-Cretaceous greenhouse climate?,Palaeogeogr. Palaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.2015.08.029  level can be assumed for the Pewsey CIE interval. Higher-frequent sed-imentological changes (e.g., chalk-marl cycles) are characteristic formany sections and most likely represent high-frequency sea-level fl uc-tuations (400 kyr and less), as e.g., demonstrated for the BohemianBasin (Uli č ný et al., 2014). Sea-level reconstructions from the latterstudy further indicate that short-term sea-level highs were more pro-nounced during long-term (~1-Myr scale) sea-level highs.Itisintriguingthatthesea-levelhighstandofthePewseyCIEintervalapparentlycoincideswithamaximuminforaminiferalandbulkcarbon-ate δ 18 Ovalues.InODPSite1259,this δ 18 Opeaksuggestsatemperaturedecrease for a 200-kyr period that comprises the Pewsey CIE and thathas been interpreted to represent  16 O sequestration in ice sheets,basedoncomparisonwithTEX 86 data(Bornemannetal.,2008).Howev-er, the build-up of signi fi cant amounts of polar ice should have causedglobal sea level to fall. The  δ 18 O maximum at the Pewsey CIE has beenrelated previously to a sea-level lowstand (Voigt and Wiese, 2000;Bornemannet al., 2008), however without fi rm support from sequencestratigraphy that, instead, indicates this  δ 18 O maximum coincides withthe sea-level maximum of a ~1 Myr long (3rd-order) sequence, asdiscussed above. While this 3rd-order sequence comprises higher-frequency (4th- and 5th-order) sea-level falls (Uli č ný et al., 2014;Wendleretal.,2014),itclearlydoesnotre fl ectanoutstandingmajorre-gressionassuggestedbytheinterpretationofaglacialepisodefromthedata of ODP Site 1259 by Bornemann et al. (2008).The discrepancy between glacio-eustatic interpretation of the  δ 18 OmaximumandsequencestratigraphyatthePewseyCIEcantheoretical-ly have thefollowingreasons: (a) the δ 18 O could carry a diagenetic sig-nalinsteadofre fl ectingtemperature,(b)thesedimentaryindicatorsforsea-level in the Pewsey CIE interval could be misinterpreted, (c) the δ 18 O maximum re fl ects cooler temperatures and the build-up of icesheets, but is miscorrelated and represents the regression at SB Tu3above thePewsey event, or (d) the δ 18 O maximum re fl ects cooler tem-peraturesandpossiblythebuild-upofminoricesheets,butwasindeedaccompanied by sea-level transgression,indicatingthata processotherthan glacio-eustasy controlled sea level.Reason (a) could be true for the bulk rock  δ 18 O values (Stoll andSchrag, 2000; Voigt and Wiese, 2000), whereby similarity of   δ 18 O re-cords among several sections could potentially mean that sea-levelchanges lead to globally synchronized, enhanced formation of syndepositional cements (cf. Schrag et al., 2013; Wendler, 2013). Rea-son (b) is indicated, e.g., by con fl icting interpretation of the Pewseyhardground in the English Chalk as a transgressive (Gale, 1996) or re-gressive (Voigt and Wiese, 2000) hardground. However, as discussedabove, sequence stratigraphic reconstructions show converging evi-dence for a sea-level highstand during the Pewsey CIE (Fig. 2). Reason(c)canberuledout,becauseTu3isgloballyassociatedwiththemarkedminimum in  δ 13 C values around the Southerham to Bridgewick CIEs(plateau of  Jarvis et al., 2006) that are clearly positioned above the δ 18 O and  δ 13 C maxima referred to by Bornemann et al. (2008) and byVoigt and Wiese (2000) (Fig. 2). It is interestingto explore reason (d): The coolinginferred from thepositiveshiftin δ 18 OvaluesatthePewseyCIEisthoughttoberelatedtodryer conditions that followed on a phase of accelerated hydrologicalcycling during the preceding  δ 13 C minimum (Bornemann et al., 2008). Thislong-termintervaloflow δ 13 Cvalues includes thenegativeGlyndeCIE, the  δ 13 C minimum at the New Pit marl level ( “ NP ”  in Fig. 2) and aremarkable Middle Turonian 3rd-order SB that is globally recognized,as mentioned above (Tu2 in Germany  “ Weiße Grenzbank ” , TuJo2 in Jordan and possibly TuTz2 in Tanzania). Accelerated hydrological cy-cling has been hypothesized to be linked with sea-level fall by the pro-cessofgroundwaterstorage(Föllmi,2012;HayandLeslie,1990;Jacobsand Sahagian, 1993, 1995; Wagreich et al., 2014; Wendler et al., 2014;Wendler et al., 2011; Wendler et al., this volume-b). This aquifer-eustatic model predicts sea-level rise due to aquifer discharge underdryerconditions,asinferredforthetransgressiveintervalofthePewseyCIE (Bornemann et al., 2008; Wendler et al., this volume-b). Fig. 1 summarizes the characteristics of the two main climate modes, green-house and icehouse, and illustrates the pathways of water for theaquifer- and glacio-eustatic process.Newevidenceinfavoroftheaquifer-eustaticideacomesfromadrillsite in Tanzania (TDP 31; Wendler et al., this volume-a). This recordincludes the stratigraphic interval of the Pewsey CIE (uppermost H. helvetica  Zone) and provides single-species  δ 18 O data from translu-cent foraminifera. In TDP 31, the falling-stage and lowstand depositsshow minor  δ 18 O minima in benthic and planktic foraminifera, indica-tive of warmer temperature and/or slightly decreased salinity. Bothinterpretations are consistent with the concept of aquifer-eustasy thatpredicts warmer temperatures and enhanced  fl uvial run-off throughan intensi fi ed hydrological cycle during the regressive phases(Fig. 1A). Is it possible that, during the Turonian greenhouse climate,the transfer of water from the ocean to continental aquifers outpacedtheglacio-eustatic process of water storage in hypothesized ephemeralAntarctic ice sheets that have been inferred from the  δ 18 O and TEX 86 data?Inotherwords:Didcooling,assuggestedfromthesedata,indeedtriggertheformationofsmallpolaricesheets,yetsealevelrose?Finally,if aquifer-eustasy was the dominant driver of sea-level changes undergreenhouse conditions, what are the implications for interpretation of  δ 18 O data? 4. Oxygen-isotope fractionation: comparison of glacio- andaquifer-eustasy  Oxygen-isotopefractionationoccursthroughpreferentialremovalof  16 O from seawater during evaporation and preferential removal of   18 Ofromthewatervaporduringcondensation.Accordingly, theprecipitate(rainorsnow)hasahigher δ 18 Ovaluethanthewatervaporfromwhichit formed, but it has a lower  δ 18 O value than the seawater, from whichthat vapor had evaporated. As atmospheric vapor moves from the siteofevaporationinthewarmtropicstowardscolderregionsathigherlat-itudes, it experiences successive condensation, whereby its  δ 18 O valuecontinues to decrease through repeated removal of   18 O during eachstep of precipitation. This process causes continental freshwater orsnow/icetohavealower δ 18 Ovaluethanseawater.Therefore,theevap-orative removal of water from the oceans and its storage on land, bothascontinentaliceandasliquidwaterinaquifers,sequesters 16 Oandin-creases the  δ 18 O value of the remaining water in the oceans (called theeffect of ice volume and groundwater volume). However, the magni-tude of this increase is much larger for the build-up of ice sheets thanfor the  fi lling of surface and sub-surface aquifers, because oxygen-isotope fractionation is temperature-dependent and glacio-eustatic re-gressions are forced by cooling, while aquifer-eustatic regressions areforced by warming (Fig. 1). The resulting difference in the net effecton seawater  δ 18 O values is related to three main aspects that are ex-plained in the following and summarized in Table 1.(1) The strength of the fractionation effect depends strongly on thenumberofprecipitationsteps.Thisnumber,inturn,dependsonthelat-itudinaltemperaturegradient,sothatthenumberof precipitation(andfractionation) steps is reduced when latitudinal temperature gradientsarelowduringawarmgreenhouseclimate,ascomparedtoanicehouseclimate. Because temperature has an opposing effect on the strength of the glacio-eustatic and aquifer-eustatic process, it follows that theamount of water accumulated in aquifers is negatively related to thenumber of precipitation steps (and thus to oxygen-isotope fraction-ation), while build-up of ice sheets is positively related to the numberof precipitation (andfractionation) steps.Therefore, duringregressionscausedbyglacio-eustasy(cooling),theincreasein δ 18 Ovaluesofseawa-ter,andofcarbonateformedfromit(i.e.,microfossilsandauthigenicce-ments), gets stronger with decreasing temperatures and increasingnumber of precipitation steps. In contrast, during regressions causedby aquifer-eustasy (warming), the increase in  δ 18 O values of seawaterbecomes less important with increasing temperatures and decreasingnumber of precipitation (and fractionation) steps. 4  J.E. Wendler, I. Wendler / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx –  xxx Please cite this article as: Wendler, J.E., Wendler, I., What drove sea-level  fl uctuations during the mid-Cretaceous greenhouse climate?,Palaeogeogr. Palaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.2015.08.029  (2)Thestorageofwaterinaquifersisadynamicprocess,i.e.,thereisaconstantback fl owoffreshwater(and 16 O)totheoceans,andthestateof  fi lledaquifersneedstobemaintainedbycontinuousre- fi llthroughastrong hydrological cycle. In contrast to liquid water, the  fl ow of ice-sheets is much slower and the process of such water storage has amore static character. This means that, for glacio-eustasy, the fraction-ation process is cumulative and leads to a progressive and non-linearincreasein δ 18 Ovaluesofseawater,asopposedtowaterstorageinaqui-fers,wheretheamountof  16 Oremovedfromtheoceandependslinearlyon the amount of water that is (dynamically) stored in aquifers.(3) During both evaporation of water and calcite crystallization, thestrengthofoxygen-isotopefractionationdeclineswithincreasingtemper-ature, because the higher-energetic conditions of warmer temperaturesallow more water molecules with  18 O to evaporate, or carbonate ionswith 18 Otoremaininsolutioninsteadofbeingpreferentiallycrystallized.This means that, for glacio-eustatic regressions (cooling), the effects of temperatureandevaporation(ice-volumeeffect)onoxygen-isotopefrac-tionation add to each other, both leading to  18 O enrichment of seawaterand of carbonate formed from it. In contrast, for aquifer-eustatic regres-sions(warming),thesetwoeffectscounteractandprogressivelydampenthe evaporative  18 O enrichment of seawater (groundwater-volume ef-fect) as temperature rises. It is possible that, under warm greenhouseclimate conditions, the temperature effect dominates over thegroundwater-volume effect, given that the latter is much reduced com-pared to theice-volume effect, as discussed above inthe fi rsttwo points.If the  δ 18 O carb  values would mainly re fl ect the temperature effect, theywould be lower (indicating warmer temperatures) during aquifer-eustatic regressions than during transgressions, and vice versa, which isin line with the Middle and Late Turonian foraminiferal  δ 18 O data fromTanzania (MacLeod et al., 2013; Wendler et al., this volume-a) and withthe TEX 86  data from ODP Site 1269 (Bornemann et al., 2008).Anadditionalcomplicationforinterpreting δ 18 Orecordsfromgreen-houseclimateperiodsarisesfromthefactthatwarmtemperaturesleadtobothenhancedevaporationandstrongerprecipitationthroughanin-tensi fi ed hydrological cycle (Fig. 1A). The associated increase in  fl uvialrun-off, together with stronger rates of evaporation, is expected to cre-atelargelateralvariationsinsurface-watersalinity.Bulkcarbonatefrompelagic sections typically consists to a large part of shells from surface-and thermocline-dwellingorganisms(i.e., nannofossils and planktic fo-raminifera), and their  δ 18 O values may have been in fl uenced by lowersalinity if the site was located near the mouth of a large river, or byhigher salinity if evaporation exceeded the freshwater input in thisarea.Additionally,formationofwarm,salinebottomwatermayhavelo-cally in fl uenced the  δ 18 O values of shells from benthic organisms.  Table 1 Comparisonoftheeffectsfromtemperatureandcontinentalwater-volumeonoxygen-isotopefractionationundericehouse-versusgreenhouseconditions.ForexplanationseeSection4.Aspect Glacio-eustasy (dominant in icehouse) Aquifer-eustasy (dominant in greenhouse)Character Oxygen-isotopefractionation δ 18 O of seawaterCharacter Oxygen-isotopefractionation δ 18 O of seawaterTemperature Low Strong High High Weak LowTemperature trend during regression Cooling Increasing Increasing Warming Decreasing DecreasingLatitudinal temperature gradient and number of precipitation steps High Strong Strong rise Low Weak Weak riseDynamics of terrestrial water storage Low Cumulative Non-linear rise High Non-cumulative Linear rise Fig. 2.  ComparisonofMiddle – LateTuronian δ 18 Orecordsandsequencestratigraphy.Notecorrelationofthe δ 18 Ominimawith sequenceboundariesandlowstandwedgeintervalsandcor-respondence of the Pewsey CIE(yellow dashed line) and the associated maximum in  δ 18 O values with sea-level rise. The 3rd-order sequence boundaries (red) as de fi ned from sedimento-logicaldataareconstrainedchronostratigraphicallybytheirpositionsrelativetothe δ 13 Crecords.Abbreviations:SB=sequenceboundary,TS=transgressivesurface;CIEs:B=Bridgewick;1=LowerSoutherham,2=UpperSoutherham,3=Caburn;P=Pewsey,G=Glynde;NP=NewPitMarlsandcorrespondinglevel;LW=LowWoollgari;RD=RoundDown.Agemodelbased on Laurin et al. (2015), for astrochronological calibration of sequence stratigraphy and discussion of sequence boundary ages see Wendler et al. (2014). 5  J.E. Wendler, I. Wendler / Palaeogeography, Palaeoclimatology, Palaeoecology xxx (2015) xxx –  xxx Please cite this article as: Wendler, J.E., Wendler, I., What drove sea-level  fl uctuations during the mid-Cretaceous greenhouse climate?,Palaeogeogr. Palaeoclimatol. Palaeoecol. (2015), http://dx.doi.org/10.1016/j.palaeo.2015.08.029
Search
Similar documents
View more...
Related Search
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