preglacial regolith has been identified on uplands
and in glacial sediments preserved on Baffin Island
in arctic Canada (26).
Other means of generating high meteoric 10Be
concentrations in sediment extracted from the
GISP2 silty ice zone do not appear plausible. It is
unlikely that the 10Be we measured was concentrated from basal meltwater. Ice in the GISP2
core contains on average 1.4 × 104 atoms meteoric 10Be g−1 (25), four orders of magnitude
less than measured in the silty ice. Stable water
isotope data from basal ice of the nearby Greenland
Ice Core Project (GRIP) core (19) fall on the meteoric water line and are inconsistent with large-scale melting that would be needed to concentrate
10Be. The positive, linear correlation between
OC and 10Be (Fig. 2) is also inconsistent with
delivery of meteoric 10Be from meltwater. Long
and/or warm interglacials (MIS 5e, 9, and 11) are
unlikely to have melted the entire GIS, and even
if they did, the peak of each interglacial is so short
(thousands of years) that insufficient meteoric
10Be would be delivered to any exposed soil at
Summit to build up even a few percent of the
meteoric 10Be inventory we infer. On the basis of
the marine isotope record (27), other interglacials
were less intense than MIS 5e, 9, and 11 and thus
even less likely to melt the entire GIS. In situ
production of 10Be by the interaction of cosmic
rays (specifically muons) with sediment under the
ice sheet, which is several kilometers thick, generates an inventory of only a few atoms 10Be g−1
over a period of 2.7 million years (28).
Previous but imprecise measurements of in
situ 10Be, 26Al, and 36Cl in rock collected below
the silty ice zone of the GISP2 core are consistent
with short exposure of that rock (at most a few
thousand years) around 0.5 T 0.2 Ma, possibly
during the long MIS 11 interglaciation (29). If
there were a short period of exposure during MIS
11, then the pre-Pleistocene soil below GISP2
was not fully eroded during that exposure. Any in
situ produced 10Be, 26Al, and 36Cl accumulated
in rock during brief exposure at MIS 11 could
have been produced under a cover of uneroded,
pre-Pleistocene soil, or the soil we sampled could
have been transported from nearby. The 10Be mea-
surements presented here demand a much longer
period of subaerial exposure than the data of
Nishiizumi et al. (29), many tens to a few hun-
dreds of thousands of years, a duration of exposure
only possible at Summit before the expansion of
the GIS at ~2.7 Ma.
The continued presence of soil in GISP2 ice
for several million years after formation of the
GIS indicates extremely low rates of sub-ice ero-
sion and slow rates of horizontal ice advection
away from Summit. Such low shear rates are con-
sistent with both long-term stability of the GIS
ice divide location and a frozen bed for most, if
not all, of the past 2.7 million years. The existence
of ancient soil implies low sub-ice erosion rates;
however, the 13-m thickness of the silty ice zone
suggests erosion, entrainment, and mixing of the
preglacial soil. Although persistence of this ancient
soil precludes substantial erosion at the bed (16),
it is possible that cold-based ice incorporated
some basal debris (30). The flux of silty ice past
the GISP2 site is likely dominated by deforma-
tion within the ice rather than by sliding (11). If
warm-based ice was present at Summit for short
periods of time in the past, it did little more than
mix sediment from the bed into the overlying ice.
Meteoric 10Be data constrain both the pregla-
cial landscape history and the Quaternary history
of Summit, Greenland. 10Be and OC data are con-
sistent with inferences, based on the stable oxygen
isotopic composition of ice and the concentration
and isotopic composition of gases in the silty ice
zone, that the GIS overran a local snowfield or
ice that itself overlay a soil (13). Since the GIS
formed, the soil has been preserved and only slow-
ly eroded, implying that an ancient landscape
underlies 3000 m of ice at Summit. These new
data are most consistent with continuous cover of
Summit by ice for the entire Quaternary, with at
most brief exposure and minimal surface erosion
during the warmest or longest interglacials.
References and Notes
1. J. S. Eldrett, I. C. Harding, P. A. Wilson, E. Butler,
A. P. Roberts, Nature 446, 176–179 (2007).
2. H. C. Larsen et al., Science 264, 952–955 (1994).
3. S. Funder, N. Abrahamsen, O. Bennike, R. W. Feyling-Hanssen,
Geology 13, 542 (1985).
4. D. Winkelmann, W. Jokat, L. Jensen, H.-W. Schenke,
Quat. Sci. Rev. 29, 1069–1077 (2010).
5. L. Håkansson, J. P. Briner, A. Aldahan, G. Possnert,
Quat. Res. 76, 452–459 (2011).
6. A. E. Carlson, J. S. Stoner, J. P. Donnelly, C. Hillaire-Marcel,
Geology 36, 359 (2008).
7. A. J. Long et al., Earth Planet. Sci. Lett. 272, 8–18
8. H. Cremer, O. Bennike, B. Wagner, Quat. Sci. Rev. 27,
9. R. Greve, Ann. Glaciol. 42, 424–432 (2005).
10. E. P. Laine, Quat. Res. 14, 188–198 (1980).
11. M. L. Bender, E. Burgess, R. B. Alley, B. Barnett,
G. D. Clow, Earth Planet. Sci. Lett. 299, 466–473 (2010).
12. E. Willerslev et al., Science 317, 111–114 (2007).
13. R. Souchez et al., Geophys. Res. Lett. 33, L24503
14. M. Stuiver, P. M. Grootes, Quat. Res. 53, 277–284
15. A. J. Gow et al., J. Geophys. Res. 102, 26559 (1997).
16. A. J. Gow, D. A. Meese, Ann. Glaciol. 22, 134 (1996).
17. D. Weis, D. Demaiffe, R. Souchez, A. J. Gow, D. A. Meese,
Earth Planet. Sci. Lett. 150, 161–169 (1997).
18. M. Suwa, J. C. von Fischer, M. L. Bender, A. Landais,
E. J. Brook, J. Geophys. Res. 111, D02101 (2006).
19. R. Souchez et al., Geophys. Res. Lett. 21, 693–696
20. P. Huybrechts, Ann. Glaciol. 23, 226 (1995).
21. J. A. Graly, P. R. Bierman, L. J. Reusser, M. J. Pavich,
Geochim. Cosmochim. Acta 74, 6814–6829 (2010).
22. T. D. Hamilton, “Surficial geology of the Dalton Highway
(Itkillik-Sagavanirktok rivers) area, southern Arctic
foothills, Alaska” (Alaska Division of Geological &
Geophysical Surveys Professional Report 121, 2003);
23. I. Callesen, K. Raulund-Rasmussen, C. J. Westman,
L. Tau-Strand, Boreal Environ. Res. 12, 681–692 (2007);
24. G. Hugelius, P. Kuhry, C. Tarnocai, T. Virtanen,
Permafr. Periglac. Process. 21, 16–29 (2010).
25. R. C. Finkel, K. Nishiizumi, J. Geophys. Res. 102,
26. K. A. Refsnider, G. H. Miller, Quat. Sci. Rev. 67, 176–189
27. L. Lisiecki, M. Raymo, Paleoceanography 20, PA1003
28. B. Heisinger et al., Geochim. Cosmochim. Acta 66, A558
29. K. Nishiizumi et al., EOS 77, OS41B (1996).
30. K. M. Cuffey et al., Geology 28, 351 (2000).
Acknowledgments: Supported by NSF grants ARC1023191
(P.R.B.) and ARC0713956 (P.R.B. and T.A.N.). Alaska soil
collection and preparation were supported by NSF grants
ARC0806394 and ARC0806399. P.R.B. analyzed the data
and wrote the first draft of the manuscript; L.B.C. extracted
the 10Be and prepared final drafts of the figures; J.A.G.
interpreted the 10Be data; T.A.N. provided glaciological
expertise; B. T.C. provided the Alaskan samples and context
for their interpretation; D.H.R. made the accelerator mass
spectrometry measurements; A.L. made and interpreted the
OC and TN measurements; and all authors read and edited
multiple versions of the manuscript. The data reported in
this paper are tabulated in the supplementary materials.
Materials and Methods
Tables S1 to S3
27 November 2013; accepted 28 March 2014
Published online 17 April 2014;
Fig. 4. Depth profile of
meteoric 10Be measured
in a stable soil profile de-velopedonglacialsediment
deposited >150,000 years
ago, North Slope, Alaska.
Prominent subsurface bulge
is indicative of a stable soil
profile (21). Maximum measured and decay-corrected
concentrations of meteoric
10Be measured in GISP2 silt
are overlain. Data in table S2. Dept
Meteoric 10Be Concentration (108 atoms g-1)
4 6 8 10 18 12 14
Alaskan Soil Permafrost Profile