evidence_rapid_sea-level_rise

8/7/2019 evidence_rapid_sea-level_rise

1/4

GEOLOGY, October 2007 89Geology, October 2007; v. 35; no. 10; p. 891894; doi: 10.1130/G23859A.1; 3 fi gures. 2007 The Geological Society of America. For permission to copy, contact Copyright Permissions, GSA, or [email protected].

*Current address: Large Lakes Observatory, Uni-versity of MinnesotaDuluth, Duluth, Minnesota55812, USA; E-mail: [email protected].

INTRODUCTION

Dating the accretion of coral reefs revealsa smooth pattern of postglacial sea-level rise(Fairbanks, 1989; Chappell and Polach, 1991;Bard et al., 1996) as major continental ice sheetsin the Northern Hemisphere disappeared. How-ever, in the Caribbean-Atlantic region, a reefback stepping ca. 7500 calibrated (cal) yr B.P.has been interpreted as a rapid sea-level rise of6.5 m (Blanchon and Shaw, 1995). The validityof this event has been subject to debate, because

no rapid sea-level rise corresponding to thisevent has been observed from coral records inthe neighboring areas (e.g., Bard et al., 1996;Montaggioni et al., 1997; Toscano and Lund-berg, 1998). Recent work on dating salt-marshdrowning on the eastern U.S. coast (Brattonet al., 2003), together with data from north-west Europe (Christensen, 1995, 1997; Lemke,2004; Behre, 2007), has renewed interest in thisquestion. All of these records indicate the sameamplitude of sea-level rise, and thus imply aglobal expression of this event. If it is global,and because it occurred under climate bound-

ary conditions similar to the present, determin-ing its magnitude and rate may provide a betterunderstanding on how future sea level respondsto global warming and the potential melting ofpolar ice sheets (Alley et al., 2005), particularlythe land-based Greenland Ice Sheet.

The accuracy of coral records depends on theknowledge of growth position of coral communi-

ties to mean sea level and the temporal variabil-ity of marine radiocarbon reservoir age (Clark,1995). Sea-level reconstructions through datingsalt-marsh basal peat in open marine settings aresubject to the uncertainties of tidal amplitude.Therefore, these data sets should be tested withother independent evidence, ideally from areasof tectonic stability and small tidal range.

STUDY AREA AND METHODS

In previously glaciated areas, local sea-level

changes are primarily governed by the relativerates of crustal rebound and global sea-levelrise. We reconstruct Holocene sea-level historyon the southeastern Swedish Baltic coast bysystematically dating the lacustrine to marineand marine to lacustrine transitions in sedimentsequences from eight closed basins with a sillof known elevation (Fig. 1). In addition to thesesites, a beach ridge at Olsng (Fig. 1A), situ-ated 8.0 m above sea level (asl) and representingthe highest Holocene sea level in this area, wasdated. A sample from peat immediately belowthe beach ridge yields a radiocarbon date of

5690 70 yr B.P., a maximum age for the beachridge. Due to the narrow connection with theAtlantic, the Baltic Sea is a brackish-water bodywith an extremely low amplitude tide of


2/4


3/4

GEOLOGY, October 2007 893

data set (Lambeck et al., 1998), is illustratedin Figure 3B and includes the upper and lower

limits of the predicted change as the uncertaintyin earth-model parameter estimates. We havetherefore applied differential isostatic correc-tions to reduce the isostatic uplift data to a ref-erence site (Kalvviken Bay). Figure 3B alsoillustrates the observed sea levels reduced to thereference site and the original features of theuncorrected record (Fig. 2); in particular, therapid sea-level rise ca. 7600 cal yr B.P. is stillpreserved. These corrections are model depen-dent, but the isostatic models are also consis-tent with a range of instrumental observations(Lambeck et al., 1998; Milne et al., 2001), all of

which indicate that the differential isostatic cor-rections between the sites are small and that theobserved rapid rise is not an artifact of differentisostatic uplift rates at the sites.

DISCUSSION

We reconstruct a high-resolution local meansea-level record in southeastern Sweden bydating transgression into and regression out ofcoastal basins with different sill elevations. Inparticular, our record reveals an ~4.5 m rapidsea-level rise ca. 7600 cal yr B.P. with a dating

uncertainty of ~400 yr. We endeavored to dateterrigenous plant macrofossils. Only three sites

were dated on bulk organic matter, which maycontain pre-aged carbon. However, the strati-graphic consistency of radiocarbon dates alongthese cores suggests that the reworking of pre-aged organic carbon is not important. It hasbeen suggested that seismic subsidence mayhave been important for the apparent rise oflocal sea level (Mrner, 2004). However, thereis no evidence in this region to support large-scale faulting during the Holocene. Thus weconclude that the rapid sea-level-rise event isrepresentative for the entire region and broadlyconsistent with the evidence for global sea-

level rise during this period.A similar rapid sea-level rise event has beenidentifi ed elsewhere, suggesting that it is globalin extent. For example, a sea-level rise of ~4 mduring the time span 79007600 cal yr B.P. hasbeen documented offshore along the GermanNorth Sea coast (Behre, 2007); the rise is relatedto a similar rise in southern Kattegat Sea datedto 77007400 cal yr B.P. (Christensen, 1997), aswell as to a ~6 m rise in the southwest Baltic Seadated as 80007700 cal yr B.P. (Lemke, 2004).This rapid sea-level rise also can be correlated

with beach ridges along the Danish Kattegaand resund coasts (Christensen, 1995). In theCaribbean Sea, submarine terraces and drownedstrandlines point to a stepped rise of sea leveca. 7500 cal yr B.P. (Blanchon and Jones1995; Blanchon et al., 2002), corresponding tothe rapid sea-level rise in the Red Sea (Siddalet al., 2003) and on the Sunda Shelf (Bird et al.2007). The spatial variation in the magnitude othis event is most likely due to localized tectonic

conditions or the differential isostatic responseof the mantle to water loading.

We propose a eustatic origin for this event. AnAntarctic source of meltwater (e.g., Blanchonand Shaw, 1995) is not convincing, althoughseemingly supported by a spike of Antarcticaderived ice-rafted debris (IRD) blanketing theSouth Atlantic between 8000 and 7500 cal yrB.P. (Hodell et al., 2001). However, this IRDrecord is poorly defi ned because of the lackof reliable radiocarbon reservoir ages for theSouthern Ocean. In addition, both 10Be (Stoneet al., 2003) and radiocarbon (Conway et al.

1999) dates show that the thinning of the WesAntarctic Ice Sheet did not accelerate until afte7000 yr B.P. Carlson et al. (2007) proposed aNorthern Hemispheric origin: their 10Be datereveal an ~600 km abrupt decay of the Labradorsector of the Laurentide Ice Sheet between 7400and 6800 yr B.P., further supported by a suddendepletion of surface-water 18O isotopes in theLabrador Sea (Hillaire-Marcel et al., 2001).

The greatest uncertainty in predicting futuresea-level changes is in the contribution of thepolar ice sheets. In contrast to the AntarcticIce Sheet, the land-based Greenland Ice Sheeappears to be the major source for the ongoing

eustatic sea-level rise (Shepherd and Wingham2007). Our record reveals a rapid sea-level riseca. 7600 cal yr B.P. with an amplitude similarto those identifi ed elsewhere. Within the datinguncertainties, the rate of this rapid sea-level riseis estimated to be ~10 mm/yr, much higher thanthat observed over the last century. Our fi ndingshows that sea level can rise rapidly withoularge continental ice sheets under climate boundary conditions similar to the present. If this risecame from the remnant Laurentide Ice Sheetas originally proposed by Carlson et al. (2007)then a land-based ice sheet can decay relatively

fast, boding ill for the Greenland Ice Sheet.

CONCLUSION

Dating the transgression and subsequenregression in marginal basins of the southeasternSwedish Baltic Sea provides a high-resolutionlocal mean sea-level record. The local sea-levechanges, reduced to the reference site, can bedivided into two phases. (1) From ca. 8600 cal yB.P., local mean sea level rose progressively andculminated ca. 6500 cal yr B.P., refl ecting thedominance of the ice-volume component due to

Age (cal yr B.P.)

Re

lativesealevel(m)

LITTORINA SEA LLSELS

Smygen

Hunnemara

RyssjnInlngan

Hallarum

Spjlk

Frsksjn

Olsng

Siretorp

10000 8000 6000 4000 2000 0

5

0

5

Transgression

Regression

Beach ridge

Figure 2. Holocene meansea-level changes alongsoutheastern SwedishBaltic coast. Horizontalbars represent 2 cali-brated age range, andvertical bars are defi nedby uncertainties of sillelevation. ELSearlyLittorina Sea; LLSlate

Littorina Sea. Dashedlines indicate durationof open brackish-water(OB) facies at the sites.Gray heavy line high-lights trend of local meansea-level changes.

Predictedisos

taticrebound(m)

Relative

sealevel(m)

Age (cal yr B.P.) Age (cal yr B.P.)

A B

10000 8000 6000 4000 2000 0

0

5

10

15

20

25

30

10000 8000 6000 4000 2000 0

0

5

10Siretorp

Kalvviken

Olsng

Figure 3. A: Predicted Holocene isostatic uplift for three sites on southeastern SwedishBaltic coast. B: Predicted relative sea level at Kalvviken Bay plotted against observed sea-level data after being corrected for the differential isostatic uplift effect.


4/4

evidence_rapid_sea-level_rise

Documents

Transcript of evidence_rapid_sea-level_rise