QuaternaryInternational,Vol. 45/46, pp. 3-18, 1998.
Pergamon
PII: S 1040-6182(97)00002--5
Copyright © 1997INQUA/ElsevierScienceLtd
Printedin GreatBritain.All rights reserved.
1040-6182/98 $19.00
LATE-WEICHSELIAN ICE SHEETS IN ARCTIC AND PACIFIC SIBERIA
M i k h a i l G. G r o s s w a l d
Institute of Geography, Russian Academy of Sciences, 29 Staromonetniy Street, 109017 Moscow, Russia
Interpretationof Space images coupled with field data on glacial geomorphologyand glacial geology of Northern Eurasia have provided
compelling evidence for a continuousLate Weichselianice sheet covering the entire Arctic margin of the continent. In addition to the
Scandinavianice sheet, three ice sheets of the same order - - the Karan, East Siberian and Beringian- - are identified witin the major
Eurasian ice sheet. Also, there is evidence and other considerationssuggesting a former ice sheet in the Sea of Okhotsk.
Ice-spreading centers of the ice sheets were situated on the continentalshelves off the Siberian coasts, thus we can term the ice sheets
'Siberian'. So far, the conventionalstratigraphicapproach to the problemof Siberianglaciationhas provenfruitlessand yieldednothingbut
uncertainty, while geomorphologicalstudies have discovered a wealth of glacial landformsin Arctic Siberia. It was these landformsand
their spacing that made the existenceof former Siberian ice sheets evident.
The East Siberianice sheet rested on the shelves of the eastern Laptev and East SiberianSeas, and its flow was directed radially,including
up the Yana and IndigirkaRivers. The Beringianice sheet was groundedon the Chukchi-,Bering-,and Beaufort Sea shelves, overridingthe
Chukchi and Seward Peninsulas from the north and flowing through the Bering Strait. It was continued by a thick floating ice shelf
buttressed by the submarineAleutian-CommanderRidge. The Okhotsk ice sheet was groundedon the Okhotsk shelf. Specific geomorphic
signatures of marine ice sheets - - the 'glacially cut comers' - - have been identifiedon all promontoriesjutting out into the Bering and
Okhotsk Seas. The Beringianand Okhotsk ice sheets are consistentwith the results of deep-sea drilling in the adjacentNorth Pacific during
Leg 145 of D/V JOIDES Resolution. © 1997 INQUA/Elsevier Science Ltd
INTRODUCTION
This paper discusses the problem of former Arctic
Siberian ice sheets. In my view, the term 'Siberian'
should be applied to the ice sheets which had their icespreading centers on the mainland or continental shelf of
Siberia. On this principle, three of the five major
constituents of the complex Eurasian ice sheet can be
referred to as Siberian - - the Kara ice sheet centered on
the Kara-Sea shelf (though its vast western outskirts
spread out into Arctic Europe), the East Siberian ice sheet
centered on the shelf surrounding New Siberian Islands,
and the Beringian ice sheet centered on the Chukchi Sea
shelf (although it impinged upon the Alaskan North Slope
and invaded the Bering Sea). Also, there is paleoclimatic
and geologic evidence for an Okhotsk ice sheet located in
the Sea of Okhotsk.
Former glaciations in Arctic Siberia are a subject of
long controversy. To characterize them, a number of
different hypotheses have been advanced which are
virtually incompatible. Each of them implies a certain
specific set of Pleistocene climates and paleoenvironments, in each case different, which compounds the
existing uncertainty and misleads Quaternary researchers.
As a result, much effort and resources are wasted on
developing advanced models and theories based on
erroneous premise.
The hypotheses fall into three groups. One of them, still
exceptionally strong in Russia, includes a variety of
antiglacialistic, or 'diluvialistic' concepts; their followers
deny ice-sheet glaciations of the high-latitude lowlands
which suggests uniformity of the Arctic paleo-climates.
The basic premise of these concepts is the conviction that
recent crustal movements and tectonically induced marine
transgressions, not glaciations, played a leading role in
past global changes in the Arctic. Another group, with the
belief commonly referred to as the 'concept of restricted
glaciation', admits some, but only minor, polar glaciations. According to that school of thought, the ice-age
Arctic was dominated by ice-free environments, while ice
caps were small and confined to the western peri-Atlantic
part of the region. The third concept, which is defended in
this paper, proposes a continuous system of marine ice
sheets grounded on the Arctic continental margins, with a
floating ice shelf over the deep Arctic Basin, and a chain
of proglacial lakes integrated into a trans-Eurasian
meltwater-drainage system alongside the southern margin
of the ice sheets.
The hypothesis of vast Arctic ice sheets was proposed
in the 1960s when evidence for a grounded marine ice
sheet was discovered in the northwestern Barents Sea
(Schytt et al., 1968). Later it developed into the theory of
a marine Eurasian ice sheet. Initially, the latter appeared
as a complex of two ice domes grounded on the BarentsKara continental shelf (Grosswald, 1980Grosswald,
1983). As more data were gathered, the limits of the
reconstructed ice sheet were extended eastward to
encompass the East Siberian shelf (Grosswald, 1988),
and even further east- and southeastward to cover the
Chukchi, Beaufort and Bering Seas (Hughes and Hughes,
1994; Grosswald and Hughes, 1995).
A 1980 version of this theory was adopted by Denton
and Hughes (1981) and CLIMAP Project Members
(1981), and it gained support from certain glaciological
4
M.G. Grosswald
modeling (Lindstrom, 1990; Lindstrom and MacAyeal,
1989). That version was repeatedly used by paleoclimatologists as data on boundary conditions for climate
sensitivity experiments. Since the early 1980s, a comprehensive program of geological studies has been undertaken by Norwegian scientists in the western Barents Sea.
In the process, the glacialistic concept defended here was
confirmed and substantiated by a wealth of new evidence
of marine and terrestrial geology (Salvigsen, 1981;
ElverhOi and Solheim, 1983; Vorren and Kristoffersen,
1986; Vorren et al., 1988; Mangerud et al., 1992;
Salvigsen et al., 1992; S~ettem et al., 1992, and many
others). These results turned the Arctic ice sheet
hypothesis into a piece of common wisdom as far as the
western Barents Sea was concerned. However, in Russia,
it has not been acknowledged and accepted by the
prevailing number of specialists (see: Biryukov et al.,
1988; Velichko, 1993).
Up to now, it is generally believed that the available
evidence for and against the ice-sheet glaciations in the
rest of Arctic Eurasia is scant and incomplete, thus
permitting a range of alternative reconstructions. This is
no longer true. Today we are in a position to make a
sensible selection from the existing alternatives, based on
the facts of glacial geomorphology and geology. There
are sufficient data to permit a single reconstruction.
Moreover, the hypotheses which purport to pose as
'alternative concepts' do not stand even the elementary
test: they fail to conform with the evidence provided by
geological surveys and space image interpretations.
THE KARA ICE SHEET
During the last, Late Weichselian, glacial hemicycle,
the Barents-Kara continental shelf and adjacent land were
glaciated by the Kara ice sheet. Judging from end
moraines, glacial striae, grooves and other ice-motion
directional indicators, the ice sheet was centered on the
western Kara Sea, merged with the Scandinavian ice
sheet and spread out in all directions, invading the
Barents Sea and northern coastal areas of the Russian
Plain and Siberian mainland.
Despite the constant growth of corroborating evidence,
the existence of the Kara ice sheet is still debated. In a
number of recent reconstructions, the Kara continental
shelf is indicated either as generally ice-free (e.g.:
Biryukov et al., 1988), or as ice-free during the last
glacial maximum (LGM) (Astakhov, 1992). Meanwhile,
the bulk of evidence for the Late Weichselian Kara ice
sheet has become impressive. It includes a single
concentric end-moraine system on the northeastern
Russian Plain, West Siberia and Taimyr Peninsula,
centered in the Kara Sea; a number of prominent iceshoved features; many sites with drumlins, flutes, glacial
stri~e and grooves, all having the radial orientation
compatible with ice spreading from a Kara-Sea center;
similarly oriented through-valleys and boulder trains, as
well as the traces of ice-dammed lakes and meltwater
channels beyond the southern limit of ice.
Fig. 1 presents the most important geomorphological
evidence for the west-central part of the last Kara ice sheet.
Its components are adopted from such sources as
' G e o m o r p h o l o g i c a l m a p o f the USSR. S c a l e
1:2,500,000' (1981), reports and maps of the State
Geological Survey (Arslanov et al., 1987; Astakhov,
1979; Kind and Leonov, 1982; Lavrov, 1977, 1981;
Voronov, 1951), and results of research projects of the
Academy of Sciences (Arkhipov et al., 1980, 1986;
Grosswald, 1979, 1980, 1988, 1993, 1994). Nearly all the
elements comprising the map have been published
elswhere, in particular, by this author (Grosswald,
1993). The only difference is in areal coverage, the scale
and, what is important, in the new, more easterly, position
of the LGM-moraine in the basin of Northern Dvina. The
map also shows a more precise relationship between the
newly discovered end moraines of the Kola Peninsula
(Grosswald, 1993, 1996b) and ice marginal formations of
the North-Dvina and Mezen Basins.
Positions of the ice margins of both the Scandinavian
and Kara ice sheets on the Russian Plain at the last glacial
maximum-LGM-Sc and LGM-K-have been established
by Lavrov (Arslanov et al., 1987; Lavrov and Potapenko,
1993). The margins are shown as coincident with
prominent systems of large end moraines which constrain-limit and confine-the areal extent of distal lacustrine terraces in the Mezen (145 m a.s.1.), Vychegda
(130 m a.s.1.), and Pechora (150 m a.s.1.) basins. Neither
the moraines nor the high terraces are directly dated.
Nevertheless, for topographic reasons, only ice-dammed
lakes related to those terraces were able to discharge
meltwater southward to the Caspian Sea, accounting, at
least partly, for its high stand at about 20 ka. The parts of
the same basins lying upglacier of the moraines contain
only terraces at 100-110 m and lower altitudes, which
attest to the paleolakes draining to the west and southwest, not to the south. This fact should not be overlooked
when trying different ice-margin positions for the LGM
Kara ice sheet.
The ages of the younger moraines related to the
Scandinavian ice are relatively well documented; they are
marked on the map. As for ages of the moraines related to
the Kara ice sheet, they are still viewed as uncertain. Not
only the dates of moraines, obtained by Lavrov and others
(Arslanov et al., 1987; Arkhipov et al., 1980), but also the
very fact of their ice-marginal origin, are sometimes
questioned (Mangerud et al., 1994; Zubakov and
Borzenkova, 1990).
However, in this author's opinion, the ~4C-dates
obtained on organic sediments contained in the low
lacustrine terraces of the Pechora, Mezen and Ob river
basins appear convincing. The dates span the intervals
of 12.5-10.5 ka and 9.5-8.5 ka and thus suggest,
broadly, the Late-glacial age of the second (from the
south) major end-moraine belt and the Holocene age
of the third belt. Corroborating evidence for these ages
came from the Kola Peninsula-White Sea area, where
Late-glacial collision of the two great ice sheets
occurred, and a direct, physical relationship between the
dated Scandinavian ice-marginal forms and the 'Kara'
Late-Weichselian ice sheets in Arctic and Pacific Siberia
60'
66'
Kara
arents
Sea
5
72"
78"
7'2-
--
Se
Sea
80
£7-"
~B"
5q"
'
'
%'~"
p
Ot
160km
80
t
i
~
i
66"
FIG. I. Selected Late Wiechselian glacial features in northern Russian Plain and adjacent parts of West Siberia, the Barents and Kara Seas
(compiled from the sources given in the text). Note the evidence for a Kara-Sea ice-spreading center in the Novaya Zemlya-Vaygach-PaiKhoi area, and the 'Kara' moraines overlapping the 'Scandinavian' ones in the west: 1 - - end-moraines, established and inferred; 2 - prominent ice-shoved features; 3 - - former LGM ice-dammed lakes (not shown in West Siberia); 4 - - horizontal pressure direction in
ice-shoved features; 5 - - glacial lineations (drumlins, flutes, giant grooves, striae, crag-and-tails); 6 - - inferred directions of iceflow; 7 - meltwater overflow channels (spillways); 8 - - glacial breaches (through valleys) in mountains; 9 - - meaningful finds of Late-glacial
erratics (Kola Peninsula - - of the Novaya Zemlya provenance, east of the White Sea - - of the Mezen basin); 10 - - approximate boundary
of the 'Kara' and 'Scandinavian' ice during LGM. Lettering: R - - Rogovskaya lobe; S - - Salekhardskiye-Uvaly lobe; spillways - - MV
- - Mezen-Vychegdan, M - - Mylvan, and K - - Keltman.
m o r a i n e s of p r o b l e m a t i c age c a n be o b s e r v e d a n d
established.
Maps of glacial features of the Kola P e n i n s u l a based on
putting together the data from older work (e.g. R a i n i o e t
a l . , 1995), bear little r e s e m b l a n c e to reality. O n the other
hand, recent geomorphic analysis of air photos and space
images, c o n d u c t e d by this author, clearly indicated that
the m a i n glacial landforms of the critical area, presented
in Fig. 2, are: (1) two ' l o n g i t u d i n a l ' e n d - m o r a i n e belts,
both facing southwestward, (2) the Tersky Keivy lateralmoraines, attesting, by their long profiles and the type of
e n e c h e l o n a l i g n m e n t , to the northeast-to-southwest ice
flow along the adjacent Gorge of the W h i t e Sea, and (3)
lateral moraines and meltwater channels related to a
K a n d a l a k s h a lobe of the Y o u n g e r Dryas S c a n d i n a v i a n ice
sheet.
Of these, the ' l o n g i t u d i n a l ' end moraines deserve
particular attention. The structure of the older (southern)
one, with p r o n o u n c e d glaciotectonic slices thrusted upon
each other to form a pattern of fish scales, u n m i s t a k a b l y
attests to a vigorous ice-sheet a d v a n c e from the northeast
(Fig. 3a). The second ' l o n g i t u d i n a l ' moraine, h a v i n g lower
relief, is conspicuous by t o n g u e - s h a p e d lobes coincident
with topographic saddles and projecting southwestward.
Both the moraines are absent on published maps though
they are easily distinguishable on space images. W h y ? I
think this is because the moraines do not fit into existing
concepts. It was for this reason that e v e n the first moraine,
which is a really huge ice marginal complex, perhaps the
largest in Europe, was left u n n o t i c e d or w h e n it was
noticed it was misinterpreted - - taken for an exposure of
tectonically distorted bedrock (see, e.g. Punkari, 1993).
Fig. 3b makes evident the cross-cutting relationship
b e t w e e n (2) and (3). This implies that the Tersky Keivy
6
M.G. G r o s s w a l d
N
S
o
100km
50
SW
Kola Peninsula
"°''°" "°°°Q
•
:~,,,-NE
-6
-4 E
-1 "~
0 =
<
-1
FIG. 2. Glacial featutes of the eastern Kola Peninsula related to a Late-glacial (probably Younger Dryas) re-advance of the Kara ice-sheet.
A schematic ground-plan and profile looking NW: I - - terrain elevations; 2 - - "Kara' moraines, frontal and lateral; 3 - - system of
parallel ridges, probably older lateral moraines; 4 - - ' K a r a ' ice-shoved moraines: 5 - - lateral moraines and meltwater channels formed by
Kandalaksha lobe of the Scandinavian ice sheet; 6 - - inferred ice-sheet profiles: 7 - - bottom-profile, Gorge of the White Sea; 8 - - terrain
profile; 9 - - inferred profiles of ice lobes; 10 - - inferred directions of iceflow: 1 I
find-sites, erratics of Novaya Z e m l y a provenance•
Moraines are younger than the lateral forms of the
Kandalaksha lobe. Indeed, the subbasins of Lake
Babozero are quite definitely contained in the icemarginally eroded furrows of (3), and at the same time
are as obviously impounded by the Tersky Keivy. Further,
as the Kandalaksha lobe apparently is correlated to the
Salpausselk~i I Moraine of Finland and dates to 10.8 ka
BP, the Tersky Keivy should be considered to have been
formed after that date. This conclusion is consistent with
and supported by 14C dates from the youngest lacustrine
beds of the North Dvina and Vychegda valleys (Arslanov
et al., 1984, 1987). The beds formed in a meltwater lake
which was dammed by the ice lobe that intruded into the
White Sea Gorge and transgressed up the valleys
(Grosswald, 1993, 1996b).
On a broader scope, the Late-glacial (most probably,
the late-to-post Younger Dryas) age of the 'Kara' ice
encroachment upon the Kola Peninsula is clear from
Figure 1, showing wide glacier lobes overlapping and
eclipsing the 'Scandinavian' moraines dating back to 15
to 10.8 ka BP. Also, a readvance of the Kara ice sheet,
having taken place at the time of Younger Dryas-to-
Preboreal transition, appears consistent with the chronology of the glaciomarine-sedimentation pulses in the
eastern Barents Sea, discovered by Polyak et al. (1995).
These workers demonstrated that sedimentation there
occurred in two pulses which started at 12.7 and 10.5 ka
BP, and were separated by an interval of nondeposition
that lasted about 1.5 ka.
The resulting paleogeography is portrayed in the
sketch-map (Fig. 4). The latter makes it manifest that a
broad lobate margin of the Kara ice sheet, including the
Gorge lobe, covered the Arctic coastal zone of the
Russian Plain at the time of Late-glacial to Holocene
transition. The areal extent of this readvance was among
the largest on Earth, which appears consistent with
specific ice-dynamic and glacio-climatic conditions in the
Barents Sea at the end of the Younger Dryas time. The
conditions were charactierized by abrupt climate warming
and increase in snowfalls (suggested by ice-core analyses
from Summit Station, Greenland) and by dramatic
destabilization of the Kara Sea ice sheet, as the latter,
about 3000 m high, was left with no buttressing after its
Barents Sea part collapsed between 14 and 11 ka. In style,
Late-Weichselian ice sheets in Arctic and Pacific Siberia
7
(a)
~q
s
0
I0
I
I
20 km
I
(b)
I
50 km
1
FIG. 3. (a) Fragment of the first (southern) 'longitudinal' end moraine c~f the Kola Peninsula, attesting to a ~igorous ice-sheet advance
from the northeast. A quadruplet of air photos. Note, that this landscape was repeatedly interpreted as an exposure of tectonically distorted
bedrock, although a wealth of end- and side-morainic ridges, glaciotectonic slices, dramlms, giant flutes and dead-ice forms can be
identified, which are assembled into a regular system of glacially thrusted features ('fish scales' ). (b) Intersection of the lateral drainage
channels of Kandalaksha ice lobe and Tersky Keivy Moraine. A stereogram of airphotos. Note that the Lake Babozero is contained in the
channels and impounded by the moraine.
the discussed ice-sheet readvance resembled the Younger
Dryas glacial event in southwestern Scandinavia (Bj6rck
and Digerfeldt, 1991), but occurred a few centuries later
and was much more extensive, which appears consistent
with its instability context.
Two consequences of the reedvance are worth
mentioning. First, that ice-dammed lakes formed along
the ice front and were forced to drain into the Baltic Ice
Lake, thus strongly increasing its catchment area. Second,
the weight of evidence uncovered and presented by
Lavrov and this author, as well as by Prokopchuk (1985),
attest to advancing of the Late-glacial ice from the
northeast, from the Kara ice spreading center. In this
context, for physical reasons, the 'Kara' ice, being able to
reach the Kola Peninsula and the White Sea, had to
encroach upon the basins of Pechora and Mezen. Thus no
reason can be conceived for arguing against the 'young'
glaciation of the Pechora Basin, despite all attempts to
uncover evidence for the contrary (Mangerud et al., 1994:
Tveranger et al., 1995).
Motions to question the glacial, ice-marginal, nature of
the landforms in question also appear groundless. These
landforms were painstakingly surveyed in the field and
carefully mapped by means of airphoto interpretation.
The resulting map (Fig. 1), displaying the end-moraine
belts with a variety of larger and smaller details, such as
morainic lobes, ice-shoved features, fluted and drumlinized ice-tongue basins, is an objective portrait of a
glaciated landscape (Lavrov, 1977, 1981 ). From my own
experience, the discussed 'young" end moraines are, as a
rule, easily discernible and identifiable on airphoto and
space images. The pattern of moraines, with the spacing
and character of the entire geomorphic complexes,
provide convincing evidence for the glacial origin of
both the landscape in general and of the smallest of its
details in particular. Among other things, this concerns
M.G. Grosswald
~
i2 i <--,-~
I
(TD
is
I
.
],~
"~C~l
d~.J
. . . . Barents-Sea part of \
"~.."
/"i'×-"
~'A~C._,-~.~,7\.,-I
.F
x.."
20*
9.00
400 l<m
.
I I
--
30 °
."
/\,/
•
O
.-
I I
60 °
I
/
i'~'"
I[
\ .
~i"l'SYg-I.-l~A
/"~.
/I i
,
~.,tvL ~
lt'~/~l(-f
l
~
I
"\
\
f
--
50 °
FIG. 4. Scandinavian and Kara ice sheets in northeastern Europe about 10.5 ka, their collision in the White Sea-Kolan area: 1 - - ice
margin and ice flowlines; 2 - - j u n c t i o n of the two ice sheets during the LGM; 3 - - ice-dammed lakes; 4 - - forms of glacial lineation; 5 - transport of erratics from the basin of Mezen; 6 - - sites where the lake-sediments were ~4C-dated.
the Late-glacial and Holocene, or 'Markhida', moraines
of the region.
Some additional questions should be addressed in the
context of the morainic patterns. First; what interpretation
of ice-sheet dynamics may be derived from it? The
patterns suggest that the LGM moraine, which looks quite
'massive', was formed by thick and probably cold-based
ice. By contrast, the younger moraines, appearing much
more lobate (see also: Lavrov, 1981), and often associated
with drumlins and flutings, could have been produced by
much thinner and faster ice which was probably warmb a s e d . . I n other words, the explanation for similar
phenomena in southwestern Minnesota given by Patterson
(1993) seems applicable to the ice-marginal features of
the northeastern Russian Plain. Indeed, the Late-glacial
and Holocene morainic lobes are typically long and
narrow, such as the Rogovaya lobe, 120 km long and
25 km wide (marked by 'R' in Fig. 1), or fluted lobes
entering the Mezen mouth from the north and the Pechora
from the northeast. It is indeed possible that the ice was
moving there on a deformable bed of water-saturated
glacio-lacustrine deposits. The lobes commonly correspond to up-glacier passages, either to saddles of the PaiKhoi Mountains or to inter-island channels of the Novaya
Zemlya and Vaygach area. As well, the lobes stretch
parallel to ice-sheet flowlines. This implies that their
dynamics originated from somewhere in the ice sheet
interior, which rules out the possibility that the lobes have
been produced by collapses of dead-ice bodies. Also,
given the nearly flat surface of the plain, the kind of
lobation displayed by the map suggests that the ice-sheet
margins were thin during Late-glacial and Holocene
times. The probable age of the Late-glacial readvance was
close to the end of Younger Dryas time, when air
temperature and snowfall abruptly increased. This type of
advance, which was suggestive of ice-sheet surging, is
consistent with these climatic changes.
Second, where was the center of ice spreading? The
ice-front position during the LGM and the orientation of
younger ice lobes on the Russian Plain, as well as all
known ice-motion indicators, strongly suggest that the ice
flow there was invariably directed from the northeast. On
the adjacent floor of the western and eastern Barents Sea,
that direction turned to become east to west (Salvigsen et
al., 1992; Gataullin et al., 1993). Judging from the
evidence of giant glacial grooves, drumlins, flutes and
glacial breaches from Novaya Zemlya, Vaygach Island
and the Pai-Khoi Mountains, recently summarized by this
author (Grosswald, 1994), the center of ice-spreading was
located in the southern Kara sea. Figures 5, 6 and 7 show
some signatures of ice flow across Novaya Zemlya,
Vaygach Island and Yugorsky Peninsula. Given this
evidence, all attempts to argue, mostly based on farfetched considerations, in favor of a major ice-spreading
center in the Barents Sea or on Novaya Zemlya, rather
than on the Kara-Sea floor (Astakhov, 1992; Forman et
al., 1995, and many others), appear dubious and hardly
tenable.
Third, what does this imply for the ice-sheet reconstructions? The answer is obvious: continuous, south-
Late-Weichselian ice sheets in Arctic and Pacific Siberia
9
Ice flow d i r e c t i o n /
hr
/
5 km
I
I
FIG. 5. Giant glacial grooves intersecting the outcrops of Paleozoic limestones and sandstones. Southern Novaya Zemlya, vertical
airphograph.
facing end-moraine belts of Late Weichselian age
conclusively indicate that a marine ice sheet of that age
was grounded on the entire adjacent continental shelf.
Furthermore, this shows that ice-sheet thicknesses and
geometry can be resolved only by glaciological modeling,
not by manipulating the Holocene shoreline altitudes
which are virtually irrelevant. It is unfortunate that the
major evidence for the Kara ice sheet - - its ice-marginal
formations - - is too often ignored, while the pattern of
Holocene land-emergence is placed into focus (e.g.
Forman et al., 1995). Evidence of that pattern played a
role during the 1960s, when it provided the earliest hints of
marine glaciation. At the same time, it is fairly clear that
the pattern of residual isostatic uplift mainly reflects the
mode, chronology, and pattern of deglaciation, not the
LGM-distribution of ice thicknesses (Grosswald, 1980,
1983). As far as the absence of Late-glacial shorelines on
the Arctic coasts of Eurasia is concerned, this phenomenon
appears predictable and does not attest to the absence of
glacio-isostatic effects. The coasts of Siberia, being
landward of the marine Eurasian ice sheet, were protected
from the ocean. Thus their vertical movements were not
recorded by flights of shorelines. By contrast, in a majority
of glaciated regions, as in Scandinavia or Svalbard, the
coasts happened to lie seaward of a former terrestrial ice
sheet. These coasts were between ice margins and the
ocean, and were therefore easily sculptured by the latter
(Grosswald, 1983; Grosswald and Hughes, 1995).
Fourth, which end-moraine belts of West Siberia
correlate with the LGM, with Late-glacial, and with
Holocene moraines of the Russian Plain? A suggestion is
provided by Fig. 1. One can see, in the Late-glacial and
Holocene scenarios, an ice sheet spreading from the Kara
Sea divided into two broad lobes around the northern part
of the Polar Ural (about 1500 m a.s.1.), which played the
role of an ice-sunderer. For glacio-dynamic reasons, the
lobes had to be of commensurate lengths; thus, the
Salekhardsky-Uvaly Moraine of West Siberia (marked by
'S'), lying just south of the Gulf of Ob (taken by Arkhipov
et al. (1980) for a LGM-boundary), should be correlated
with the Rogovskaya lobe (see above) and assigned a Lateglacial age. All Siberian moraines lying north of
Salekhardsky Uvaly appear to correlate with the Holocene
(Markhida) belt. In this context it is only natural to expect
that the Siberian LGM-boundary is located on the eastern
continuation of its European correlative. Presently, the
most likely candidate for this role is the limit suggested by
Goncharov (1986) - - the limit which lies on that
continuation and is constrained by the evidence of
paleo-lake Yelogui. Further east, across the Yenisei into
Central Siberia, the LGM-boundary of the Kara ice sheet
probably extends along the end-moraine systems mapped
by Andreyeva and Isayeva (1988). Finally, in the north, the
ice sheet had to reach the outer edge of the Barents-Kara
shell which is clear from both data obtained by the
Norwegian marine geological surveys and glaciological
considerations by Weertman (1974).
This reconstruction is not yet fully substantiated by
geological evidence, but it is glacio-dynamically justified. It is consistent with the modeling experiments
by Fastook and Hughes (1991), and conforms well with
certain 14C-dates obtained in West Siberia and the
Taimyr Lowland from the beds underlying the upper till
sheet (Arkhipov et al., 1980). Paradoxically, it is
inconsistent with many other dates derived from sediments overlying the same till. Those dates, suggesting
ages 'older than 35-40 ka BP', are being pointedly used
as an argument against the concept of a Late Weichselian Kara ice sheet; virtually, they are the main and
often the only argument on which minimum reconstructions base.
However, in my view, the evidence of 'old' dates is
unconvincing. As a rule it is at variance with the facts of
glacial geomorphology. Also, in many instances, the dates
10
M.G. Grosswald
FIG. 6. Field of drumlins. SoutheasternVaygachIsland,
vertical airphotograph.
themselves are suspicious, coming from recycled and
displaced materials, and of being contaminated by 'old'
carbon. This recycling and contamination are predictable
consequences of the Siberian paleogeography, since the
areas in question were recurrently overridden by ice or
flooded by proglacial lakes transgressing southward from
the Arctic continental shelf, so that massive spreading of
sediments originating from distal environments had taken
place.
The same should be said of mammoth bones coming
from glaciated areas and dated at 14-24 ka BP. They
do not present an unsurmountable obstacle to the
concept of East Siberian glaciation, however strong the
'mammoth argument' may seem (Sher, 1995). The
bones could be recycled, redeposited, and contaminated,
too. After re-processing, they may yield quite different
ages, as happened years ago in Scandinavia: there,
several 19-24 ka dates, incompatible with the Scandinavian ice sheet, were also obtained on mammoth bones.
However, after re-processing the samples, those dates
turned out either much younger or much older. The
same may happen to the ages of the 'Russian'
mammoth. It wouldn't be surprising because many of
them are in striking contrast with the Arctic paleoenvironments. Could the mammoth fauna flourish in Severnaya Zemlya at the LGM, as suggested by the dates?
Despite the dates, mammoth occupation is out of the
question for an archipelago which is heavily (50% of
its area) glacierized today. Quite obviously, a lot of
additional work should be done concerning mammoth
finds before employing their ages as a piece of evidence
for or against the Siberian glaciations.
Considerable misunderstanding has sprung up from
some erroneous interpretations. An example of such
'errors of consequence', leading to far-fetched inferences
on 'retarded deglaciation' and 'older-than-it-seems'
geomorphology of Siberia, has been demonstrated by a
discussion of the Siberian site 'Ice Hill' (Astakhov and
Isayeva, 1988). At this, special importance was assigned to
a couple of 14C dates on wood fragments, namely
43.1±1 ka BP and more than 50 ka BP, originated from
a debris bed overlying a thick tabular body of buried
glacier ice. Judging by the authors' description (and my
own memories of the site), the overlying debris is typical
ablation till reworked by the processes of slumping and redeposition in meltwater pools. Nevertheless, when
inferring the age of the buried ice, Astakhov and Isayeva
used the assumption that the wood incorporated into that
ablation till is geologically younger than the underlying
ice. But why? The 'Ice Hill' section is not a marine or
lacustrine sequence, it is a remnant of a kilometer-thick ice
sheet, the latter's near-bottom part. Such parts are typically
rich in debris produced by glacial scour of underlying
rocks and incorporated into ice from below, along ice
shear-plains or other near-bottom faults. The faults reach
up to several tens of meters above glacier bed, and so do
the masses of debris. And it was melting-out of that debris
that first stopped the Late-glacial ice-sheet down-wasting
by fbrming a protective cover of ablation till. Thus the
datable materials contained in that till are typically derived
from beneath the ice sheets, and are older, not younger,
that the latter. In this context, Astakhov's inference as to
the Early Weichselian age of the buried glacier ice in the
'Ice Hill' locality (and elsewhere) appears groundless.
Moreover, given the wide occurrence of buried glacier ice
in Siberia, the above mechanism of how the older materials
turned up on the surface of the younger ones may prove to
be of more universal application, providing us with a clue
to the long-debated problem of supra-till 'old' dates.
THE EAST SIBERIAN AND BERINGIAN ICE
SHEETS
Until recently, all evidence for ice-sheet glaciation of
the East Siberian shelves and coastal lowlands was sparse,
indirect, and equivocal. The glaciation itself was
considered highly problematic --especially since the
ice-age climates of that part of Siberia were, and still are,
broadly believed to have been too dry to enable initiation
and growth of large ice sheets.
Conversely and quite paradoxically, recent computer
simulations based on advanced climate models suggest
the opposite. Specifically, they yield vast ice-sheet
glaciation of the Siberian North East, which appears
more extensive than the Scandinavian ice sheet (Verbitsky and Oglesby, 1992; Marsiat, 1994; Huybrechts and
T'siobbel, 1995).
What can be learned from recent studies of the region?
Judging from current publications (e.g. Biryukov et al.,
1988; Velichko, 1993), only cirque and valley glaciers
occurred in the north of East Siberia during the
Quaternary. Even Hopkins (1972), who envisaged larger
glaciers there, related them to the penultimate, not to the
last glaciation. What is more important, he maintained
that they were basically terrestrial features, restricted to
highlands, and not marine ice sheets transgressing from
continental shelves.
The first convincing geological signatures of vast East
Siberian glaciations were uncovered during 1987-1990.
Initially they comprised a complex of ice-shoved ridges,
tunnel valleys, outwash plains and other glacial landforms
of the New Siberian Islands and surrounding shelf. Their
Late-Weichselian ice sheets in Arctic and Pacific Siberia
I
6 km
drumlin-tails pointed to the southwest. Thus, there is little
doubt that the geomorphology of the Tiksi area also
implied the northeast-to-southwest direction of former ice
flow (Grosswald and Spektor, 1993).
The Tiksi hill-and-hole pairs were AMS-L4C dated. To
this end, Wibj6rn Karlen and his team visited the area in
the spring, and drilled into the lake bottoms through the
seasonal ice. The sediment samples were processed and
dated at Uppsala University, Sweden, and yielded a Late
Weichselian age of the whole Tiksi glacio-geomorphic
complex (Grosswald et al., 1992).
Based on the evidence from the New Siberian Islands
and Tiksi, a 400-km long ice flowband was reconstructed
offering the first clue to solving the mystery of Siberian
ice sheets. That flowband extended from the New
Siberian Islands southwestward, reaching the foothills
of the Verkhoyansky Range. With this accomplished, it
was possible to take the following step and hypothesize
the tentative southern ice boundary (Grosswald. 1988).
Further progress in establishing the ice sheet's southern
limits was achieved by mapping the oriented tundra
landforms of the Yana-Indigirka, Lena delta and Bol.
Lyakhovsky Island areas, including so called 'oriented
lakes ~. Orientation of these tundra landforms within the
entire region proved independent of wind directions or
tectonic structures, but strikingly consistent with the flow
pattern determined. The whole geomorphic continuum of
the Yana-lndigirka Lowland and adjacent areas, including
its ridge-and-lake complex, is sketched on Fig. 9. There.
the ridges and lakes are clearly aligned along submeridional directions, and they radiate from the environ of
the New Siberian Islands to form a fan-like pattern,
diverging southward.
Another characteristic feature of the complex is the
occurrence of a second, transversal, system of ridges and
1
FIG. 7. Giant flutes, northwestern Yugorsky Peninsula
(northern promontory of the Pal Khoi Mountains),
vertical airphotograph.
geometry was indicative of former ice motion from the
northeast. The tusks of Mammoth primigenius were found
within ice-distorted beds of the northern Bunge Land
Island, implying a late Weichselian age for the whole
complex (Grosswald, 1988, 1990).
Another assemblage of ice-shoved features was found
and mapped in the Tiksi area of North Yakutia a few
years later. It is represented by well preserved ice-thrust
forms (stacking orders of imbricate slices) coupled with
glaciotectonically excavated basins (hill-and-hole pairs),
as well of parallel rock-drumlin clusters carved from
Paleozoic shales (Fig. 8). The glaciotectonic slices were
found to be displaced southwestward, shale plates in
drumlins were overturned in the same direction, and the
Lake Sevastian
"i
iii
ii ¸¸¸
~i~i~~
!
1
1I
~i!i~ ¸
l km
j
FIG. 8. Rock-drumlins carved from Paleozoic shales. South of Lake Sevastian, Tiksi area, vertical airphotograph.
12
M.G. Grosswald
t20"
76"
i
Lapte
i
v/Sea
//
/
/
/
I
/
1
/
/
/
/
/
/
/
/
//
I
i
/
/
/I
0
t:
777,
~100
I
I
ii
/
/
/
/
l
/
/
/
------~N e w
f'~
/
Is
S .ri bJ e r i a n
ands
!
ol. Lyakhovsky
/
~
~2hromaInlet
!
20Okra
I
FN-I
FIG. 9. G e o m o r p h o l o g i c a l c o m p l e x e s of the Y a n a - I n d i g i r k a L o w l a n d and adjacent areas: 1 - - large ice-shoved features; 2 - - direction o f
horizontal glacial pressure; 3 - - drumlins; 4 - - direction o f long axes o f the tundra oriented forms; 5 - - transverse tundra ridges; 6 - relic valleys o f m e l t w a t e r streams; 7 - - inferred ice flowlines; 8 - - m o u n t a i n s and highlands; 9 - - areas discussed in the text; 10 - glacial and m e l t w a t e r breaches. Areas: A - - ice-shoved features, the N e w Siberian Islands, B - - the same, Tiksi area; C - - orientred
forms, northwestern L e n a delta, D ~ r n m l i n s o f the Bol. L y a k h o v s k y Island, E - - field o f oriented lbrms, Y a n a - I n d i g i r k a Lowland, F - site of w a s h b o a r d moraines, G - - p u s h - m o r a i n e of the Allaikha valley, J - - O g u s t a k h drumlin field.
valleys, oriented perpendicularly to the lake-and-ridge
alignments (Fig. 10). Similar sets of intersecting tundra
features are known from Ayon Island (see below) and
Alaska (e.g. Rosenfeld and Hussey, 1958).
Both in Alaska and Siberia, the complexes occur on
ice-rich sediments and are being interpreted as thermokarst features, while their elongation and regular orientation are traditionally accounted for by the effects of
prevailing summer winds. However, we argued (Grosswald and Spektor, 1993; Grosswald and Hughes, 1995)
that this explanation is inapplicable to Arctic Siberia.
Instead, we suggested that the oriented tundra landforms
had acquired their orientation and alignment from glacial
drumlinization and large-scale fluting, produced by a
marine ice sheet which transgressed from the adjacent
Arctic shelf. Among other arguments supporting this
hypothesis, Grosswald (1996a) pointed out prima facie
drumlins occuring on continuations of the oriented tundra
complexes (Fig. 11).
As far as the systems of transverse ridges and valleys
are concerned, we interpreted them as ice-marginal
features, akin to Urstromtdler, formed by meltwater
streams and marking consecutive positions of a retreating
ice margin. These features have been eroded into the thick
blanket of specific ice-rich silts and sands, so-called
'yedoma', accumulated in proglacial (not subglacial, as
Sher (1995) queried) environments during, mainly, Lateglacial and early Holocene times. Ice-sheet impoundment
in the north was a vital prerequisite of their formation. At
a number of sites the yedoma was 14C-dated to the Lateglacial (Kaplina and Lozhkin, 1982), and thus provided
chronological limits for the oriented complexes.
All in all, we suggested that the landforms in question
represent Late Weichselian glacial features that may be
extensively used for ice-sheet reconstructions. Being
carved from deeply frozen and ice rich sediments, the
forms have been distorted and disfigured by following
thermokarst, solifluction and other periglacial processes
which are gathering speed under current climate warming. This disfiguring accounts for the systematic misinterpretation of the oriented landforms.
Based on the flowlines derived from both the iceshoved features and oriented tundra forms (Fig. 9), a new,
more advanced reconstruction of an East Siberian ice
sheet was accomplished (Grosswald and Hughes, 1995).
The spreading center of the ice sheet proved to be on the
Arctic shelf, in the vicinity of the New Siberian Islands. It
was coalescent with the Kara ice sheet in the west and
F---I
50 km
13
I
FIG. 10. Field of oriented tundra forms and transverse ridges. Yano-lndigirka Lowland, vertical space image.
t
5 km
I
FIG. 11. Ogustakh drumlin field, Indigirka basin. Vertical airphotograph.
with another great ice sheet in the east, for which the
name of the 'Beringian ice sheet' is proposed.
Evidence for the latter is currently being gathered
and tested. A strikingly p r o n o u n c e d and large
(12,000 km 2) field of push moraines of the Kholercha
tundra, the lower Kolyma basin, and oriented tundra
landforms of the Karchyk Peninsula and~adjacent Ayon
Island, west of the Chaun Guba, are among the
evidence. Being identified on space images by Grosswald (1996a), both geomorphic complexes clearly attest
to the iceflow directed southwestward, thus suggesting
another ice-spreading center situated northeast of the
Kolyma delta. Additional pieces of coroborative evidence include the Upper- and Mid-Pleistocene till sheets
of the Vankarem Lowland, Chukchi Peninsula (Laukhin
et al., 1989); accumulations of glacial erratics on Cape
Serdtse Kamen; as well as abundant glacial erratics
and Lappland-style geomorphology of Wrangel Island
(observed by this author, 1991-1993).
Geomorphic evidence for ice flow through the Bering
Strait and across the Chukchi Range of the Chukchi
Peninsula, all directed from north to south, is of particular
importance. Cape Dezhnev, bounding the strait from the
west, turned out to be a 700-m high and steep glaciated
wall with a truncated spur at its foot and a huge mass of
coarse debris accumulated on its lee (southern) side. In
turn, the Chukchi Range, on its entire 600-kin-long part
stretching between Bering Strait and the Pekulney Range,
14
M.G. Grosswald
is breached by dozens of U-shaped valleys oriented in a
north-to-south direction. The same direction is implied by
local oriented lakes and by large end moraines in the
lower Anadyr valley. Ice moved there from the north and
northeast, across the Anadyr Lowland and the adjacent
shelf. Tabular bodies of relic glacier ice occur on the
Anadyr coastal lowland, which contain marine fossils
from that shelf (Kaplyanskaya and Tarnogradsky, 1978).
Compelling evidence for a continuous glaciation of the
Bering Sea, including its deep-sea basin, comes from the
geomorphology of the southeastern Chukchi Peninsula
and of some promontories jutting out into the sea. The
first has become a fjordland (Sinyavino Maze of Fjords)
by ice moving from the northeast, cutting and dissecting
the land comer. As to the promontories, in particular,
Cape Navarin and Cape Olyutorsky peninsulas, they also
display intersection by glacial troughs (Fig. 12).
The phenomenon of 'ice-cut comers' seems ubiquitous
on the shores abutting the former marine ice sheets. The
Irish Sea ice sheet cut the promontories on both sides of
St. George's Channel, between Ireland and Britain (Eyles
and McCabe, 1989). Another marine ice sheet, that of the
Kane Basin between Ellesmere Island and Greenland,
overrode and dissected Cape Herschel Peninsula (Blake,
1992). The Central Arctic ice shelf overrode and eroded
the Yermak Plateau 'comer' (Vogt et al., 1994). As to our
example, it also strongly suggests that a marine ice sheet,
this time confined within the Bering Sea, transgressed
across the bounding headlands, such as Cape Dezhnev,
the Sinyavino fjordland, and the above-mentioned capes.
Incidentally, one of them, Cape Olyutorsky, lies quite
close to the deep southern Bering Sea and, judging from
this setting, was overridden by the edge of an ice shelf
floating in this part of the Bering Sea. Furthermore, a
triangle-shaped channel incised into the Shirshov submarine ridge just offshore of the cape looks similar to that
of the Yermak Plateau. Based on this, it may be predicted
that iceberg plowmarks of the kind described by Vogt et
al. (1994) will be spotted offshore of Cape Olyutorsky in
the years to come.
As it is, the 'hanging' troughs of Cape Olyutorsky are
consistent with the ice shelf inferred earlier by Grosswald
and Vozovik (1984) and Hughes and Hughes (1994).
Additional piece of evidence for the ice shelf is provided
by glaciated troughs crossing the Commander-Aleutian
Ridge and its islands. In particular, Bering Island has been
breached by several through valleys with U-shaped cross
sections and striations on their slopes (Erlich and
Melekestsev, 1974; Y. Muravyov, Petropavlovsk-Kamchatsky, oral commun.).
The hypothesis of a Beringian ice sheet consisting of
grounded parts and of a floating ice shelf appears
consistent with the evidence for a recent, post-Sangamon,
episode of ice-overriding St. Lawrence Island (Benson,
1993). Also, it provides a first reasonable explanation for
the giant submarine troughs deeply incised into the shelf
margin-so called 'canyons' of Bering, Bristol, Pribylov,
Pervenets, Zhemchug and others-which were described,
but virtually not accounted for, by Scholl and others
(Scholl et al., 1970).
Moreover, the hypothesis is consistent with the results
of deep-sea drilling in the North Pacific, undertaken in the
course of Leg 145 of D/V JOIDES Resolution (Anonymous, 1992; Leg 145 scientific party, 1993). Analysis of
the core sediments yielded by the drilling strongly
suggested extensive glaciation of the ocean's coasts and
shelves between 2.6 Ma and the very end of the
Pleistocene. Variations in density of the sediments have
FIG. 12. Cape Olyutorskypeninsulawith intersectingglacial troughs.Vertical airphoto.
Late-Weichselian ice sheets in Arctic and Pacific Siberia
led to the inference that the last 95 ka of that period,
including the Isotope Stage 2, was punctuated by a
sequence of abrupt paleoceanographic changes similar to
and simultaneous with Heinrich events and DansgaardOeschger cooling cycles of the North Atlantic (Kotilainen
and Shackleton, 1995). Pronounced traces of similar
changes have been spotted in the Chinese loesses, also. In
my view, they must have reflected glacial and climatic
events in the 'nearby' North Pacific, although Porter and
Zhisheng (1995), who discovered the traces, chose to
ascribe them to atmospheric teleconnections with the
North Atlantic.
The evidence for Beringian glaciation is constantly
mounting. By now, it has gathered critical mass to
warrant tentative reconstruction of an ice sheet. We may
envisage a paleo-ice cover resting on the Chukchi and
Bering shelves and periodically, in synchrony with
Heinrich events of the North Atlantic, ejecting armadas
of icebergs into the North Pacific. It has been this kind of
scenario that was earlier suggested for the North Atlantic;
now we are trying it on the North Pacific. The North
Atlantic is known to have been bounded by huge ice
sheets and ice shelves. Isn't its paleoceanographic
similarity to the North Pacific suggestive that the latter
was characterized by the same paleo-glacial environment? The sketch-map in Fig. 13 portrays one of the
possible versions of a Beringian ice sheet at the LGM.
AN ICE SHEET IN THE SEA OF OKHOTSK?
There is some evidence for Pleistocene glaciation of
the Sea of Okhotsk, too. It comes from the data provided
by paleoclimatic, geomorphological and marine geological studies.
Climate and paleoclimate of the Sea of Okhotsk and
adjacent land is, and was, as cold as that of the Kara Sea,
and rich in snowfall, thus favorable for initiation and
growth of glaciers. Judging from recent modeling
experiments, mountains and uplands boardering the sea
were most susceptible to heavy glacierization during
glacial hemi-cycles (Marsiat, 1994; Huybrechts and
T'siobbel, 1995). Among other things, the experiments
predicted that the Kolymsky, Suntar Khayata and Koryak
Ranges, adjacent to the discussed sea, had been subject to
extensive ice-sheet glaciation. This seems to come true as
topographic maps and space images of the ranges display
clear evidence for former glacier complexes of Cordilleran type.
Hence, the Sea of Okhotsk was semi-surrounded by
large and continuous glacier complexes, which were most
probably characterized by very low, close to the sea-level,
equilibrium lines. This suggested that seaward margins of
the complexes extended into the sea and invad its
northern periphery. Actually, that event - - past glacial
invasion - - has been well documented by the sea's
coastal and bottom geomorphology (Udintsev, 1957).
Even small-scale maps make it clear that big glaciated
valleys enter the Penzhina and Gizhiga inlets and
continue seawards at least as far as southern limit of the
15
Shelikhov Bay. The southern exit of this bay is bounded,
on both sides, by 'ice-cut corners' of the Utkholok and
Piyagina peninsulas, thus suggesting a wide marine ice
lobe moving outwards, in southwestern direction. At this,
its left margin pushed sideways and impinged onto the
western coast of Kamchatka Peninsula, where its position
is marked by a chain of parallel end moraines, eskers and
drainage channels (Grosswald, unpublished).
Further south, Shelikhov Bay opens up into the Tinro
Basin of the Sea of Okhotsk, which is about 1000 m deep
and U-shaped; its bottom topography is dominated by
large longitudinal ridges of enigmatic nature (Udintsev,
1957; Volnev, 1983). A variety of explanations were
advanced for the ridges, but, in my view, the ridges
cannot be accounted for other than by impact of grounded
ice and subglacial meltwater activity.
The rest of the Sea of Okhotsk falls into two different
parts, of which one lies north of 47-49°N, where I
envisage a shelf break. The other is confined between that
break and the Kuril submarine ridge. The second part, no
doubt, is a piece of deep ocean akin to the deep Bering
Sea basins, while the first, though extraordinary deep,
looks to me very much as a continuation of a continental
shelf. The basins and sills of that part, including Deryugin
Basin with its depth of about 2 kin, thus are intra-shelf
formations, which strongly suggest, at least to me, that
they are of glacial origin. In other words, I hypothesize
that the basins in question have been excavated by a
marine ice sheet, and the submarine ridges bearing the
names of Academy of Sciences and of Institute of
Oceanology, as well as other similar forms are glacial
sills, or riegels.
I am fully aware that, according to the currently
prevailing view, the edge of the shelf lies farther north,
and the area of the above basins and ridges is underlain by
the crust of transitional, not continental, type. Not only
those big submarine forms, but also many smaller troughs
and sills, are believed to have been produced by faulting
and folding of the crust (Udintsev, 1957). However, this
concept has not been proven, either, and also needs to be
tested.
I did not mention that the floor of the Sea of Okhotsk is
commonly covered, even far from the coasts, by coarse
sediments, including boulders (Bezrukov, 1960). A field
of lodgement till, a sheet of unsorted solid boulder clays,
3 to 8 m thick, was found under 15 m of late- and
postglacial sediments by drilling off the coast of southwestern Kamchatka (Kuzmina and Yeremeeva, 1990). On
the northern and western sides of the sea, all 'corners'
seem to have also been glacially cut, and the lower
reaches of the Amur River exhibit clear indications of
former impoundment.
We cannot say how far south the Okhotsk ice sheet
extended, and what its morphology was like. But it hardly
was just an icy apron attached to the foot of the coastal
.mountains, though the ice sheet could be shaped like this
at the beginning of its growth. Later its ice spreading
center had to shift to the shell in the southeasterly,
windward direction. At this stage, the ice sheet would
move centrifugally, in all directions, including landwards,
16
M.G. Grosswald
impinging upon some of the intra-montane saddles from
the sea. As of now, we only know of fairly clear traces of
ice-sheet pressure against Taigonos Peninsula, between
the Penzhina and Gizhiga inlets, directed to the northeast,
from the sea landwards.
The southern margin of the ice sheet most probably
graded into a floating ice shelf which, in the same way as
the Bering ice shelf, was buttressed by a submarine ridge,
in this case by the Kuril one. This ridge is also breached
by a number of deep U-shaped straits. Some of the latter,
like the Bussol, are quite deep, up to 2.5 km. So it is not
unlikely that the straits played the role, in concert with the
Commander-Aleutian straits, of conduits for the icebergs
supplying into the North Pacific.
All in all, judging from paleoclimate of the Sea of
Okhotsk, the modeling experiments, and some geomorphic and marine geological data, this sea may turn
out to be an additional center of former Siberian
glaciation. The marine Okhotsk ice sheet, if confirmed,
would further strengthen the stance of the Leg 145
scientific party (Leg 145 scientific party, 1993; Anonymous, 1992) who argued in favour of similarity, as
opposed to antagonism, of the ice-age North Atlantic and
North Pacific. A few decades ago, a visionary geologist
proclamed this similarity. It was Okamoto (1972) who
used his till finds in Japan to argue that '...glaciations on
the eastern side of Asia were large and comparable on
size to those in Atlantic North America, due to maritime
climate with strong precipitation, the low mean yearly
temperature at about the same latitudes and the wide
distribution of alpine regions'. This quotation sounds
quite up to date. And Okamoto's inference looks sensible
and consistent with the latest developments.
CONCLUDING REMARKS
About 20 years ago Hughes et al. (1977) advanced a
hypothesis of a huge and continuous ice sheet of the
Arctic, behaving as a single dynamic system. A grand
problem was posed by the title of the paper: 'Was there a
late-Wiirm Arctic ice sheet?'. Today, we can answer that
question: yes, there was. Moreover, it was larger than was
suggested by our 'maximum version' of 1977.
A continuous Eurasian ice sheet impounded northward-flowing rivers, causing the formation of large
proglacial lakes and their integration into transcontinental meltwater drainage system. The lakes experienced
drastic changes in their extent, outlines, and levels,
resulting from retreat and re-advances of the ice sheet,
from consequent isostatic rebound, and from erosional
deepening or sediment infilling of spillways and overflow channels. So far, the drainage systems are not
properly studied; however, the very fact of their
existence, documented by a number of spillways, paleolake terraces and other features, constitutes a strong
and independent argument confirming a continuous
ice sheet in the Arctic.
Here, although I am not discussing the lakes and their
history, a couple of remarks seem appropriate. The first
one concerns the erosional incision crossing the Laptev
Sea shelf just east of Taimyr Peninsula. As hypothesized
in a report edited by Ftitterer (1994), the feature was
eroded on emerged shelf by the ice-age Lena River,
which implied that the last East Siberian ice sheet had not
existed. In our concept, however, the incision in question
and an ice sheet in the north were quite compatible: the
erosional form was produced by a Late-glacial flood, or
floods, resulted from outbursts of a meltwater Lena paleolake, dammed by the East Siberian ice sheet.
The second remark concerns the Turgay Valley of
southern West Siberia. The role of this valley as a huge
late Weichselian spillway, draining Siberian meltwater
into the Aralo-Caspian basin, was questioned by Astakhov (1992). However, an analysis of the newest space
images by V. Baker, coupled with older information on
the geology and age of the spillway (Grosswald, 1983),
left little doubt that it has been a morphologically young
formation eroded by powerful and turbulent streamflow
from West Siberian ice-dammed lakes. The same was
found to be true for another major spillway, the Manych
Valley, which connected the ancient Caspian Sea, or the
Khvalyn meltwater basin, with the Black Sea (V. R.
Baker, pers. commun.). These young spillways are hardly
compatible with Astakhov's position in the dispute over
the Eurasian proglacial lakes.
I am not discussing here the problem of a giant CentralArctic floating ice shell either. However, it is hard not to
mention that new discoveries in the Arctic Ocean have
provided compelling evidence for such an ice shelf.
Specifically, they are huge iceberg (or crushed ice shelf)
plowmarks detected on the Yermak Plateau at the depths
of 500-850 m, perhaps even at 2 km (Vogt et al., 1994);
and also, facts providing new insights about ecological
conditions of the ice-age Arctic Ocean. In particular,
when crossing the ocean aboard the US and Canadian
icebreakers, and '...studying the box core sediments for
evidence of biological life, researchers found nothing in
the layers dated to between 13,000 and 26,000 years ago.
This abiotic zone suggests that during that glacial period
the Arctic Ocean was covered by an ice shelf hundreds of
meters thick that killed off all life in the waters below...
Thus the Central Arctic was locked up' (Travis, 1994).
This discovery is particularly remarkable, as some of our
opponents, e.g. Zubakov and Borzenkova (1990), repeatedly emphasized that the 'fact (fact?) of uninterrupted,
continuous evolution of organic life in the Arctic Ocean
presented strong evidence against the ice shelf and the
whole Arctic ice sheet'.
The paper by Hughes et al. (1977) provoked a surge of
critical papers targeted at refuting the concept of the
Arctic ice sheet, and presented similar 'facts' and
considerations appearing to disprove it. However, the
elapsed time interval has been sufficiently long to confirm
Bacon's observation that time is the greatest innovator.
New trends in the thinking of ice-age researchers become
discernible, revealing a .slow but steady progress towards
adopting our concept. Virtually all new discoveries in the
region, let alone modeling experiments, bring increasing
support for the Arctic ice sheet. So, virtually, we no
Late-Weichselian ice sheets in Arctic and Pacific Siberia
longer need to defend the case: rapid development of
Arctic science is doing the work.
ACKNOWLEDGEMENTS
My participation in the Stockholm workshop on the Siberian ice
sheets was sponsored by Marcus Wallenberg's Foundation for International Scientific Cooperation, as well as by the Royal Swedish Academy
of Sciences and the Stockholm University. This paper was written while
I was visiting the Scott Polar Research Institute (SPRI), Cambridge,
U.K., and I thank the Institute and its Director John Heap for provision
of resources. My thanks are also due to Robert Headland, SPRI, and
Chalmers Clapperton, University of Aberdeen, who critically read the
manuscript. This research was supported in part by Grants N4Y000 and
N4Y300 from the International Science Foundation.
REFERENCES
Andreyeva, S.M. and Isayeva, L.L. (1988). The dynamics of the ice
sheet on the northeastern part of the Central Siberian Plateau in the
Late Pleistocene. Polar Geography and Geology, 12, 3 212-220.
Anonymous (1992). Ocean Drilling Program tracks Northern Pacific
Ocean's climate history. Ocean Drilling Program Public Information.
Texas A and M University Research Park, Texas.
Arkhipov, S.A., Astakhov, V.I., Volkov, I.A., Volkova, V.S. and
Panychev, V.A. (1980). Paleogeografiya Zapadno-Sibirskoi Ravniny
v Epokhu Pozdne-Zyryanskogo Lednikovogo Maksimuma [Paleogeography of the West Siberian Plain during the Late Zyryan Glacial
Maximum]. Nauka, Novosibirsk.
Arkhipov, S.A., Isayeva, L.L., Bespaly, V.G. and Glushkova, O. Yu
(1986). Glaciation in Siberia and the North East USSR. In: Sibrava,
V., Bowen, D.Q. and Richmond, G.M. (eds), Quaternary Glaciations
in the Northern Hemisphere. Oxford et al., Pergamon Press, pp. 463474.
Arslanov, Kh.A., Levina, N.B., Ostanin, V.E., Baranova, V.N.,
Smirnova, V.M. and Tertychnaya, T.V. (1984). On the geochronology and stratigraphy of the Late Pleistocene and Early Holocene in
the Severnaya Dvina basin. Vestnik Leningr Univer, 12, 57-66.
Arslanov, Kh.A., Lavrov, A.S., Potapenko, L.M., Tertychnaya, T.V. and
Chernov, S.B. (1987). New data on the late Pleistocene and early
Holocene geochronology in the northern Pechora lowland. In:
Punning, Y.-M.K. et al. (eds), Novyye Dannyye po Geokhronologiy
Chetvertichnogo Perioda [New Data on Geochronology of the
Quaternary], pp. 101-111. Nauka, Moscow.
Astakhov, V.I. (1979). New data on the latest activity of the Kara
glaciers in West Siberia. In: Sibrava, V. (ed.), Project 73/1/24
Quaternary Glaciations in the Northern Hemisphere, Vol. 5, pp. 2231. Prague.
Astakhov, V.I. (1992). The last glaciation in West Siberia. In:
Robertsson, A.-M. et al. (eds), Quaternary Stratigraphy, Glacial
Morphology and Environmental Changes (Research Papers, SGU
Ser. Ca 81), pp. 21-30. Uppsala.
Astakhov, V. and Isayeva, L.L. (1988). The 'Ice Hill': an example of
'retarded deglaciation' in Siberia. Quaternary Science Reviews, 7,
29-40.
Benson, S.L. (1993). Glacially deformed sediments of Lavrentiya Bay,
Chukotka Peninsula, Far Eastern Russia, and the north shore of
St.Lawrence Island, Alaska. In: Aber, J.S. (ed.), Glaciotectonics and
Mapping Glacial Deposits, pp. 1-8. Hignell Printing Ltd., Winnipeg.
Bezrnkov, P.L. (1960). Bottom sediments in the Sea of Okhotsk. In:
Bezrukov, P.L. (ed.), Geologicheskiye issledovaniya v Dalnevostochnykh moriakh, 15-95 (Proc., Inst. of Oceanol., Vol. 32). Academy of
Science Publishers, Moscow.
Biryukov, V.Y., Faustova, M.A., Kaplin, P.A., Pavlidis, Y.A.,
Romanova, E.A. and Velichko, A.A. (1988). The paleogeography
of Arctic shelf and coastal zone of Eurasia at the time of the last
glaciation (18,000 yr. B.P.). Palaeogeography, Palaeoclimatology,
Palaeoecology, 68, 117-125.
Bj6rck, S. and Digerfeldt, G. (1991). Aller0d-Younger Dryas sea level
changes in southwestern Sweden and their relations to the Baltic Ice
Lake development. Boreas, 20, 115-133.
Jr.Blake, W. (1992). Holocene emergence at Cape Herschel, east-central
Ellesmere Island, Arctic Canada: implications for ice sheet configuration. Canadian Journal of Earth Sciences, 29, 9 1958-1980.
Denton, G.H. and Hughes, T.J. (eds) (1981). The Last Great Ice Sheets.
Wiley-Interscience, New York.
17
Elverh¢i, A. and Solheim, A. (1983). The Barents Sea Ice Sheet - - a
sedimentological discussion. Polar Research, 1, 23-42.
Erlich, E.N. and Melekestsev, I.V. (1974). Commander Islands. In:
Luchitsky, I.V. (ed.), Kamchatka, Kurilskiye i Komandorskiye
ostrova, pp. 327-337. Nauka Publishing, Moscow.
Eyles, N. and McCabe, A.M. (1989). The late Devensian (<22,000 BP)
Irish Sea Basin: the sedimentary record of a collapsed ice sheet
margin. Quaternary Science Reviews, 8, 307-351.
Fastook, J.L. and Hughes, TJ. (1991). Changing ice loads on the Earth's
surface during the last glaciation cycle. In: Sabadini, R., Lambeck, K.
and Boschi, E. (eds), Glacial lsostasy, Sea-Level and Mantle
Rheology, pp. 165-200. Kluwer, Dordrecht.
Forman, S.L., Lubinski, D., Miller, G.H., Snyder, J., Matishov, G.,
Korsun, S. and Myslivets, V. (1995). Postglacial emergence and
distribtion of late Waichselian ice-sheet loads in the northern Barents
and Kara seas, Russia. Geology, 23, 113-116.
Fiitterer, D.K. (ed.). (1994). The Expedition ARCTIC '93, Leg ARK-IX/4
of RV 'Polarstern' 1993. (Berichte zur Polarforschung 149 '94).
AWI, Bremerhaven.
Gataullin, V., Polyak, L., Epstein, O. and Romanyuk, B. (1993).
Glacigenic deposits of the Central Deep: a key to the Late Quaternary
evolution of the eastern Barents Sea. Boreas, 22, 47-57.
Goncharov, S.V. (1986). Limit of the last ice sheet in the middle reaches
of the Yenisei River: position and age. Doklady AN SSSR, 290, 6
1436-1439.
Grosswald, M.G. (1979). The last Eurasian ice sheet. Polar Geography
and Geology, 3, 169-188.
Grosswald, M.G. (1980). Late Weichselian ice sheet of northern Eurasia.
Quaternary Research, 13, 1-32.
Grosswald, M.G. (1983). Pokrovn3:ve ledniki kontinentalnykh shelfav
[h'e sheets ~f the continental shelves]. Nauka, Moscow.
Grosswald, M.G. (1988). Antarctic-style ice sheet in the Northern
Hemisphere: toward the New global glacial theory. Polar Geography
and Geology, 12, 239-267.
Grosswald, M.G. (1990). An ice sheet on the East Siberian shelf in the
Late Pleistocene. Polar Geography and Geology, 14, 4 294-304.
Grosswald, M.G. (1993). Extent and melting history of the Late
Weichselian ice sheet, the Barents-Kara continental margin. In:
Peltier, W.R. (ed.), Ice in the Climate System, 1-20 (NATO ASI
Series 1 'Global Environmental Change' 12). Springer, Berlin.
Grosswald, M.G. (1994). The drumlin fields of the Novaya ZemlyaUrals region and the Kara Sea center of glaciation. Polar Geography
and Geology, 18, 15-32.
Grosswald, M.G. (1996). Evidence for an ice-sheet advance from the
Arctic Shelf upon the coast of Northeastern Siberia. Doklagy
Academmi Nauk, 350, 4 535-540.
Grosswatd, M.G. (1996). The Post-Younger Dryas (ca. 10 ka B.P.) icesheet transgression from the Barents Shelf toward Northeastern
Europe. Doklady Akademii Nauk, 350, 5 687-691.
Grosswald, M.G., Karldn, W., Shishorina, Z. and Bodin, A. (1992).
Glacial landforms and the age of deglaciation in the Tiksi area, East
Siberia. Geogr. Annaler, 74A, 295-304.
Grosswald, M.G. and Hughes, T.J. (1995). Paleoglaciology's grand
unsolved problem. Journal of Glaciology, 41, 313-332.
Grosswald, M.G. and Spektor, V.B. (1993). The glacial relief of the
Tiksi region (west shore of Buor-Khaya Inlet, Northern Yakutia).
Polar Geography and Geology, 17, 154-166.
Grosswatd, M.G. and Vozovik, Y.I. (1984). A marine ice cap in South
Beringia (A working hypothesis). Polar Geography and Geology, 8,
128-146.
Hopkins, D.M. (1972). The paleogeography and climatic history of
Beringia during Late Cenozoic time. lnter-Nord, 12, 121-150.
Hughes, TJ., Denton, G.H. and Grosswald, M.G. (1977). Was there a
late-Wiirm Arctic ice sheet?. Nature, 266, 596-602.
Hughes, B.A. and Hughes, T.J. (1994). Transgressions: rethinking
Beringian glaciations. Palaeogeography. Palaeoclimatology, Palaeoecology, 110, 275-294.
Huybrechts, P. and T'siobbek S. (1995). Thermomechanical modelling
of Northern Hemisphere ice sheets with a two-level mass-balance
parameterization. Annals of Glaciology, 21, 111-116.
Kaplina, T.N. and Lozhkin, A.V. (1982). On the age of the 'icy complex'
of the Arctic lowlands of Yakutia. lzvestiya AN SSSR, Ser. geogr.,
N2, 84-95.
Kaplyanskaya, F.A. and Tarnogradsky, V.D. (1978). Relic glacier ice
and its role in formation of the Quaternary deposits and land forms
within the permafrost area. In: Quaternary Geology and Geomorphology (Transactions, VSEGEI, ns., 297), pp. 65-76. Leningrad.
Kind, N.V. and Leonov, B.N. (eds) (1982). Antropogen Taimyra
[Antropogene of the Taimyr Peninsula]. Moscow, Nauka.
18
M.G. Grosswald
Kotilainen, A.T. and Shackleton, N.J. (1995). Rapid climate variability
in the North Pacific Ocean during the past 95,000 years. Nature, 377,
323-326.
Kuzmina, N.N. and Yeremeeva, G.P. (1990). New data on the origin and
age of the boulder clay from the shelf off West Kamchatka. Doklady
Akad. Nauk SSSR, 310, 1425-1428.
Laukhin, S.A., Drozdov, N.I., Panychev, V.A. and Velichko, S.V.
(1989). The age of last glaciation in northern East Chukotka. Izvestiya
AN SSSR, Ser. geol., N3, 136-140.
Lavrov, A.S. (1977). Kola-Mezen, Barents-Pechoran and Novaya
Zemlya-Kolvan paleo-ice streams. In: Chebotariova, N.S. (ed.),
Structura i Dinamika Poslednego Evropeiskogo Lednikovogo Pokrova [Structure and Dynamics of the Last European Ice Sheet], pp. 81100. Moscow, Nauka.
Lavrov, A.S. (1981). Subglacial deformation moraines produced by the
Barents Sea ice sheet. Bull. Komis. Izuch. Chetvertich. Perioda, 51,
51-60.
Lavrov, A.S. and Potapenko, L.M. (1993). [Quaternary geology: a map
and explanation] Map-sheet P-38, 39-Syktyvkar. New Series.
1:1,000,000. (Manuscript, VSEGEI, St.Petersburg).
Leg 145 scientific party (1993). Paleoceanographic record of North
Pacific quantified. EOS, 74, 406, 411.
Lindstrom, D.R. (1990). The Eurasian ice sheet: formation and collapse
resulting from natural atmospheric CO2 concentration variations.
Paleoceanography, 5, 2 207-227.
Lindstrom, D.R. and MacAyeal, D.R. (1989). Scandinavian, Siberian,
and Arctic Ocean glaciation: effect of Holocene atmospheric CO2
variations. Science, 245, 628-631.
Mangerud, J., Bolstad, M., Elgersma, A., Helliksen, D., Landvik, J.Y.,
Lonne, I., Lykke, A.K., Salvigsen, O., Sandahl, T. and Svendsen, J.I.
(1992). The last glacial maximum on Spitsbergen, Svalbard.
Quaternary Research, 38, 1-31.
Mangerud, J., ~,stakhov, V., Svendsen, J. and Tveranger, J. (1994).
PECHORA: Paleo environment and climatic history of the Russian
Arctic. In: Perdahl, J.-A. (ed.), 21 Nordiska Geol. Vintermrtet ....
Lule~. Abstracts of lectures and posters, 132.
Marsiat, I. (1994). Simulation of the Northern Hemisphere continental
ice sheets over the last glacial-interglacial cycle: experiments with a
latitude-longitude vertically integrated ice-sheet model coupled to a
zonally averaged climate model. Palaeoclimates-Data and Modelling, 1, 59-98.
Okamoto, Y. (1972). Piedmont glaciation in the taiga forests of ice ages
in Japan and Northern Italy similar to those now present in Southern
Alaska. In: Quaternary Geology. Section 12, Proc. of 24th lnt. Geol.
Congress, Montreal, pp. 175-186.
Patterson, C.J. (1993). Mapping glacial terrain, southwestern Minnesota,
USA. In: Aber, J.S. (ed.), Glaciotectonics and Mapping Glacial
Deposits, pp. 155-176. Hignell Printing Ltd., Winnipeg.
Polyak, L., Lehman, S.J., Gataullin, V. and Jull, A.J.T. (1995). Two-step
deglaciation of the southeastern Barents Sea. Geology, 23, 567-571.
Porter, S.C. and Zhisheng, An (1995). Correlation between climate
events in the North Atlantic and China during the last glaciation.
Nature, 375, 305-308.
Prokopchuk, B.I. (1985). On the direction of ice flow in the northern
East European Platform. Doklady AN SSSR, 283, 446-447.
Punkari, M. (1993). Modelling of the dynamics of the Scandinavian ice
sheet using remote sensing and GIS methods. In: Aber, J.S. (ed.),
Glaciotectonics and Mapping Glacial Deposits, pp. 232-250.
University of Regina.
Rainio, H., Saamisto, M. and Ekman, I. (1995). Younger Dryas end
moraines in Finland and NW Russia. Quaternary International, 28,
179-192.
Rosenfeld, G.A. and Hussey, K.M. (1958). A consideration of the
problem of oriented lakes. Proceedings of the Iowa Academy of
Science, 65, 279-287.
Salvigsen, O. (1981). Radiocarbon dated raised beaches in Kong Karls
Land, Svalbard, and their consequences for the glacial history of the
Barents Sea area. Geografiska Annaler, 63A, 283-291.
Salvigsen, O., Adrielsson, L., Hjort, C., Johansson, K., Kelly, M.,
Landvik, J.Y. and Ronnert, L. (1992). Ice movements in eastern
Svalbard. In: Mrller, P., Hjort, C. and Ingrlfsson, O. (eds),
LUNDQUA Report 35, pp. 9-16.
Sa~ttem, J., Poole, D.A.R., Ellingsen, K.L. and Sejrup, H.P. (1992).
Glacial geology of outer BjCmCyrenna, southwestem Barents Sea.
Marine Geology, 103, 15-51.
Scholl, D.W., Buffington, E.C., Hopkins, D.M. and Alpha, T.R. (1970).
The structure and origin of the large submarine canyons of the Bering
Sea. Marine Geology, 8, 187-210.
Schytt V., Hoppe G., Blake W., Jr and Grosswald, M.G. (1968). The
extent of the WOrm glaciation in the European Arctic. In: lnternat.
Assoc. Sci. Hydrology IUGG, Publ. 79, pp. 207-216. Gentbrugge.
Sher, A. (1995). Is there any real evidence for a huge shelf ice sheet in
East Siberia?. Quaternary International, 28, 39-40.
Travis, J. (1994). Taking a bottom-to-sky 'slice' of the Arctic Ocean.
Science, 266, 1947-1948.
Tveranger, J., Astakhov, V. and Mangerud, J. (1995). The margin of the
last Barents-Kara ice sheet at Markhida, northern Russia. Quaternary
Research, 44, 328-340.
Udintsev, G.B. (1957). Geomorphology of the Sea of Okhotsk floor. In:
Bezrukov, P.L. (ed.), Geologichtskiye issledovaniya v Dalnevostochnykh moryakh, 3-76. (Proc., Inst. of Oceanol., Vol. 22). Acad. Sci.
Publ., Moscow.
Velichko, A.A. (ed.). (1993). Evolution of Landscapes and Climates of
Northern Eurasia. Late Pleistocene-Holocene. l - - Regional
Paleogeography. Nauka, Moscow.
Verbitsky, M.Y. and Oglesby, R.J. (1992). The effect of atmospheric
carbon dioxide concentration on continental glaciation of the
Northern Hemisphere. Journal of Geophysical Research, 97,
5895-5909.
Vogt, P.R., Crane, K. and Sundvor, E. (1994). Deep Pleistocene iceberg
plowmarks on the Yermak Plateau: sidescan and 3.5 kHz evidence
for thick calving ice fronts and a possible marine ice sheet in the
Arctic Ocean. Geology, 22, 403-408.
Volnev, V.M. (1983). On the origin of the ridgy topography in the Tinro
Basin, the Sea of Okhotsk. Geomorfologiya, N3, 66-75.
Voronov, P.S. (1951). New data on glaciation and Quaternary deposits
of the Central Pai-Khoi. In: Collected Papers on the Arctic Geology,
Vol. 2, pp. 84-92.
Vorren, T., Hald, M. and Lebesbye, E. (1988). Late Cenozoic
environments in the Barents Sea. Paleoceanography, 3, 601-612.
Vorren, T.O. and Kristoffersen, Y. (1986). Late Quaternary glaciation in
the south-western Barents Sea. Boreas, 15, 51-59.
Weertman, J. (1974). Stability of the junction of an ice sheet and an ice
shelf. Journal of Glaciology, 13, 3-11.
Zubakov, V.A. and Borzenkova, I.I. (1990). Global Palaeoclimate of the
Late Cenozoic. Elsevier, Amsterdam.