Reconstructing Sea Ice Conditions in the Arctic and Sub-Arctic Prior to Human Observations - Arctic Sea Ice Decline: Observations, Projections, Mechanisms, and Implications - page 26

the ocean and the atmosphere [e.g., Walsh , 1983]. Arctic sea
ice is also an important parameter with respect to the ther-
mohaline circulation since it constitutes a reservoir of fresh
water that is eventually exported through surface currents
to the northern North Atlantic through Fram Strait and the
Canadian Arctic Archipelago [e.g., Barry et al. , 1993; Car-
mack , 2000], whereas brines resulting from sea ice growth
and evaporation contribute to enhance the salinity of the
Deep North Atlantic Water masses [ Meincke et al. , 1997].
Therefore, sea ice is a critical component of the earth system
which needs to be well constrained before any assessment
of sea ice extent in the future based on coupled atmosphere-
ocean modeling. Sea ice is, however, a complex parameter
with respect to its dynamics and thermodynamics, and its pa-
rameterization may yield to diverse responses despite identi-
cal forcing. This is illustrated, for example, by the very large
range of Arctic sea ice changes that are predicted for the end
of the 21st century (see Intergovermental Panel for Climate
Change Fourth Assessment Report) [cf. also Arzel et al. ,
2006; Stroeve et al. , 2007]. Moreover, as shown by Stroeve
et al. [2007], almost all models simulate trends for the last
decades that differ from observations and generally underes-
timate the actual decline in summer sea ice extent. Thus, sea
ice modeling still remains a challenge for the scientific com-
munity. The examination of intervals characterized by the
extreme climate of the recent geological past may provide
a means for evaluating model capabilities for extrapolating
sea ice under various boundary conditions [e.g., Hewitt et
al. , 2001; Smith et al. , 2003; Vavrus and Harrison , 2003;
Renssen et al. , 2005; Goosse et al. , 2007]. Near analogues of
sea ice coverage for the next decades might well be found in
the geological records that contain the archives for extreme
sea ice cover at the surface of the earth from total ice cover-
age to ice-free conditions. In this perspective, it is relevant to
examine the recent geological history that has been marked
by large-amplitude variations of the sea ice cover under natu-
ral forcing. However, sea ice is a sensitive parameter, which
is characterized by a high variability in space and time that
increases the difficulty of its reconstruction from paleocli-
matological archives. The most direct indications of past sea
ice cover are found in marine sediments, which contain vari-
ous tracers or proxies of environments characterized by sea
ice. They include sedimentary tracers of particles entrained
and dispersed by sea ice, biogenic remains associated with
production under sea ice or with ice-free conditions, in addi-
tion to geochemical and isotopic tracers of brine formation
[ Hillaire-Marcel and de Vernal , 2008; Haley et al. , 2008].
Here, we provide an overview of the approaches that can be
used for the reconstruction of sea ice cover in the past, with
emphasis given to proxies allowing quantification of sea ice
coverage. Special attention is thus paid to the assemblages of

organic-walled dinoflagellate cysts (or dinocysts) that were
used to develop transfer function for the reconstruction of sea
ice cover [ de Vernal and Hillaire-Marcel , 2000; de Vernal et
al. , 2000, 2001, 2005a, 2005b]. We also refer to the isotopic
composition (δ 18 O) of foraminifera that may provide com-
plementary information on brine production related to sea
ice formation [ Hillaire-Marcel and de Vernal , 2008]. On the
basis of these approaches, we provide examples of estimated
changes in sea ice cover at the high latitudes of the Northern
Hemisphere during the early Holocene, which is an interval
characterized by higher summer insolation at high northern
latitudes and by warmer climate than at present.

2. TRACeRS AND PROXIeS OF SeA ICe

2.1. Sedimentological Tracers

In marine sediments, which result from detrital terrigenous
input and pelagic fluxes, there are particles related directly
or indirectly to sea ice. Among tracers of sea ice that are
independent from the biogenic production and purely relate
to sedimentary processes, the grain size is most commonly
used. The coarse mineral material (coarse silt and fine sand
fractions) that is too large and too dense to be carried by
wind or marine currents has to be associated with ice-rafted
debris (IRD). Coarse material can be entrained by sea ice
during ice formation on the shelf before being transported
within the drifting ice that is eventually released by melting,
notably at the sea ice margin [e.g., Pfirman et al. , 1990; Heb-
beln and Wefer , 1991]. Coarse materials recovered in marine
cores thus provide direct evidence for sea ice in the overly-
ing surface water. However, the coarse material can result
from drifted multiyear ice or icebergs having remote origin,
as well as from annual ice having a more proximal source.
The mineralogical or geochemical characteristics of IRD
may help identifying the source rocks of detrital particles
incorporated into the ice on the shallow shelves and are thus
useful for interpretations in terms of ice drift patterns [e.g.,
Bischof and Darby , 1997; Darby , 2003; Andrews and Eberl ,
2007].
In general, IRD accumulate where sea ice is melting and
therefore characterize the areas close to distal margins of
sea ice. Therefore, IRD usually correspond to seasonal sea
ice, whereas the deep-sea sediment under perennial sea ice
is generally deprived of coarse debris. The interpretation of
IRD is not straightforward depending upon the context. For
example, high-IRD content in sediments from the Arctic
Ocean may indicate seasonal melt of sea ice and thus re-
duced ice concentration [e.g., Darby et al ., 2001; de Vernal
et al. , 2005a], whereas IRD occurrences in North Atlantic
sediments would rather reflect peaks in seasonal spreadings