We use cookies to ensure that we give you the best experience on our website. You can change your cookie settings at any time. Otherwise, we'll assume you're OK to continue.

Durham University

Department of Geography

Departmental Research Projects

Publication details

Clark, P.U, Mitrovica, J.X, Milne, G.A & Tamisiea, M Sea-Level Fingerprinting as a Direct Test for the Source of Global Meltwater Pulse IA. Science. 2002;295:2438-2441.

Author(s) from Durham


The ice reservoir that served as the source for the meltwater pulse IA remains enigmatic and controversial. We show that each of the melting scenarios that have been proposed for the event produces a distinct variation, or fingerprint, in the global distribution of meltwater. We compare sea-level fingerprints associated with various melting scenarios to existing sea-level records from Barbados and the Sunda Shelf and conclude that the southern Laurentide Ice Sheet could not have been the sole source of the meltwater pulse, whereas a substantial contribution from the Antarctic Ice Sheet is consistent with these records.


1. B. D. Douglas, J. Geophys. Res. 96, 6981 (1991).
2. E. Bard, B. Hamelin, R. G. Fairbanks, A. Zindler, Nature
345, 405 (1990).
3. See (34) for a comprehensive review of observations
and arguments related to this question.
4. W. R. Peltier, Science 265, 195 (1994).
5. J. P. Kennett, N. J. Shackleton, Science 188, 147
6. A. Leventer, D. F. Williams, J. P. Kennett, Earth Planet.
Sci. Lett. 59, 11 (1982).
7. L. D. Keigwin, G. A. Jones, S. J. Lehman, E. A. Boyle, J.
Geophys. Res. 96, 16811 (1991).
8. R. G. Fairbanks, C. D. Charles, J. D. Wright, in Radiocarbon
After Four Decades, R. E. Taylor, A. Long, R. S.
Kra, Eds. (Springer-Verlag, New York, 1992), pp. 473Ѝ
9. J. T. Andrews, H. Erlenkeuser, K. Tedesco, A. E. Aksu,
A. J. T. Jull, Quat. Res. 41, 26 (1994).
10. L. D. Keigwin, G. A. Jones, Paleoceanography 10, 973
11. Some of these objections are as follows: (i) The size
of the meltwater pulse would have required that the
entire southern sector become an ablation zone and
melt away at unprecedented rates. The geological
evidence, however, clearly shows that ice remained
throughout most of this region at the end of the
event (35). (ii) It is difޣult to explain how those
sectors of the Laurentide Ice Sheet that drained to
the Hudson Strait and the Gulf of St. Lawrence
remained immune to a climate forcing that induced a
nearby catastrophic melting event. (iii) Most ocean
models (36), although not all (37), suggest that a
freshwater ߵx of this magnitude derived from the
Mississippi River would substantially reduce the Atlantic
thermohaline circulation ( THC) without delay.
Proxy records of changes in the THC, however, show
that mwp-IA occurred more than 1000 years before
the next signiޣant change in the THC associated
with the Younger Dryas cold interval. (iv) The d18O
anomaly in the Gulf of Mexico record shows that
excursions of similar magnitude and sign occurred
before mwp-IA, indicating that the anomalies may
reߥct relatively minor ߵctuations in meltwater discharge
down the Mississippi River (34).
12. D. R. Lindstrom, D. R. MacAyeal, Nature 365, 214
13. G. E. Birchޥld, H. Wang, J. J. Rich, J. Geophys. Res.
99, 12459 (1994).
14. Oxygen isotope records from both the Norwegian and
Greenland Seas and reconstructed salinities for the
northeastern North Atlantic indicate little, if any, meltwater
discharge from either the Barents Sea or Fennoscandian
Ice Sheets during mwp-IA (34). In addition,
neither ice sheet appears to have been capable of
delivering volumes sufޣient to account for mwp-IA.
15. T. J. Hughes, in The Geology of North America, Vol.
K-3, North America and Adjacent Oceans During the
Last Deglaciation, W. F. Ruddiman, H. E. Wright Jr.,
Eds. (Geological Society of America, Boulder, CO,
1987), pp. 183в20.
16. E. Bard et al., in The Geological History of the Polar
Oceans: Arctic Versus Antarctic, U. Bleil, J. Thiede, Eds.
(Kluwer Academic, Norwell, MA, 1990), pp. 405д15.
17. A. Shemesh, L. H. Burckle, J. D. Hays, Paleoceanography
10, 179 (1995).
18. R. S. Woodward, U.S. Geol. Surv. Bull. 48, 87 (1888).
19. W. E. Farrell, J. A. Clark, Geophys. J. R. Astron. Soc. 46,
647 (1976).
20. J. X. Mitrovica, M. E. Tamisiea, J. L. Davis, G. A. Milne,
Nature 409, 1026 (2001).
21. M. Tushingham, W. R. Peltier, J. Geophys. Res. 96,
4497 (1991).
22. G. A. Milne, J. X. Mitrovica, J. L. Davis, Geophys. J. Int.
139, 464 (1999).
23. J. X. Mitrovica, W. R. Peltier, J. Geophys. Res. 96,
20053 (1991).
24. For this purpose, we have adopted the seismic model
PREM ( preliminary reference Earth model) described
in (38).
25. The neglect of thermal expansion effects in the
present study contrasts with analyses of the geographic
variation in present-day sea-level change
driven by ongoing melting of polar ice complexes (20,
39д2). In this case, thermal expansion on the order
of 1 mm/year (43) is comparable in amplitude to the
observed rate of present-day global sea-level rise (1).
26. This is not to say that far-ޥld data are immune to
contamination from GIA. We quantify this contamination
in (32) for the case of the differential sea-level
change between Barbados and the Sunda Shelf across
the mwp-IA event.
27. E. Bard et al., Nature 382, 241 (1996).
28. T. Hanebuth, K. Stattegger, P. M. Grootes, Science
288, 1033 (2000).
29. M. Stuiver et al., Radiocarbon 19, 355 (1998).
30. The marker dated at 13,500 yr B.P. (solid red squares
in Fig. 3), which deޮes the end of our analysis time
window, is a wood sample collected more than 100
km from other Sunda Shelf data. Hence, our estimate
of a sea-level rise of 25 m from 14,200 yr B.P. to
13,500 yr B.P. at this site is dependent on the veracity
of a composite sea-level curve. Furthermore, the
sea-level variation just before mwp-IA is not well
resolved on the Sunda Shelf curve.
31. In the case of a southern Laurentian source for mwp-
IA, a sea-level rise of 25 m at Barbados would also be
accompanied by a sea-level rise of 43 m at Tahiti.
From Fig. 3, sea level at Tahiti subsequent to mwp-IA
was ; и0 m (27). For the southern Laurentian
scenario, we would thus infer a sea level at Tahiti, just
before the onset of mwp-IA, of ; б23 m. The large
discrepancy between this value and the observed
contemporaneous sea level at other far-ޥld sites
(e.g., Fig. 3) is also an argument, albeit indirect,
against this speciޣ scenario. The same argument
applies to the scenarios involving all Northern Hemisphere
ice sheets (North-ICE3G) and all global ice
sheets (All-ICE3G).
32. This conclusion is relatively insensitive to the sea-level
signal as a result of ongoing GIA across the mwp-IA
time window. We performed GIA calculations in which
we predicted sea-level changes during a 1000-year period
beginning at 14,500 yr B.P. that were driven by late
Pleistocene glaciation/deglaciation events before the
meltwater pulse. GIA calculations are commonly based
on spherically symmetric, viscoelastic Earth models that
are speciޥd by values for the upper and lower mantle
viscosity (where the boundary between the two regions
is taken to be 670 km depth) and the thickness of the
elastic lithosphere. We varied each of these three
parameters over a broad range of possible values and
found that the differential sea-level rise (Barbados
minus Sunda Shelf ) was no greater than 1.8 m. This
bound takes into account that the sea-level marker
dated to 13,500 calendar yr B.P. at the Sunda Shelf is
signiޣantly displaced geographically relative to older
markers on the same curve (30).
33. There are several important implications if the meltwater
pulse originated largely from Antarctica: (i) It indicates
that substantial deglaciation occurred several
thousand years before the onset of Antarctic deglaciation
suggested from ice sheet and GIA models [see
discussion in (34)]. (ii) If substantial ice is required in
Antarctica after mwp-IA to account for the younger
deglaciation phase predicted by these models, then the
excess, preЭwp-IA ice may resolve the discrepancy
between total ice volumes during the last glacial maximum
reconstructed from ice sheet models and from
some GIA models (44). (iii) An Antarctic source would
resolve the apparent conundrum of having a large freshwater
forcing in the North Atlantic with a 1000-year
delay in the response of the Atlantic THC (11).
34. P. U. Clark et al., Paleoceanography 11, 563 (1996).
35. A. S. Dyke, V. K. Prest, Geogr. Phys. Quat. 41, 237
36. S. Manabe, R. J. Stouffer, Paleoceanography 12, 321
37. K. Sakai, W. R. Peltier, J. Geophys. Res. 100, 13455
38. A. M. Dziewonski, D. L. Anderson, Phys. Earth Planet.
Inter. 25, 297 (1981).
39. J. A. Clark, J. A. Primus, in Sea-Level Changes, M. J.
Tooley, I. Shennan, Eds. (Institute of British Geographers,
London, 1987), pp. 356г70.
40. S. M. Nakiboglu, K. Lambeck, in Glacial Isostasy, Sea-
Level and Mantle Rheology, R. Sabadini, K. Lambeck,
E. Boschi, Eds. (Kluwer Academic, Dordrecht, Netherlands,
1990), pp. 237в58.
41. C. Conrad, B. H. Hager, Geophys. Res. Lett. 24, 1053
42. H.-P. Plag, H.-U. Ju�ttner, Mem. Natl. Inst. Polar Res.
(special issue 54), 301 (2001).
43. C. Cabanes, A. Cazenave, C. Le Provost, Science 294,
840 (2001).
44. P. U. Clark, A. C. Mix, Quat. Sci. Rev. 21, 1 (2002).
45. E. Bard, M. Arnold, R. G. Fairbanks, B. Hamelin, Radiocarbon
35, 191 (1993).
46. R. G. Fairbanks, Nature 342, 637 (1989).
47. Supported by the NSF Earth System History program
(P.U.C.), the Canadian Institute for Advanced Research
( J.X.M.), PREA funding from the Government
of Ontario ( J.X.M., M.E.T.), the Natural Sciences and
Engineering Research Council of Canada ( J.X.M.), and
a McLean Research Award from the University of
Toronto ( J.X.M.). We thank M. Ishii for assistance in
preparing Fig. 2.

Department of Geography