East Antarctic Ice Sheet Sensitivity to Pliocene Climatic Change from a Dry Valleys
Perspective
Author(s): George H. Denton, David E. Sugden, David R. Marchant, Brenda L. Hall, Thomas I.
Wilch
Source: Geografiska Annaler. Series A, Physical Geography, Vol. 75, No. 4, A Special Volume
Arising from the Vega Symposium: The Case for a Stable East Antarctic Ice Sheet (1993), pp.
155-204
Published by: Blackwell Publishing on behalf of the Swedish Society for Anthropology and
Geography
Stable URL: http://www.jstor.org/stable/521200
Accessed: 08/05/2009 11:12
Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless
you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you
may use content in the JSTOR archive only for your personal, non-commercial use.
Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at
http://www.jstor.org/action/showPublisher?publisherCode=black.
Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed
page of such transmission.
JSTOR is a not-for-profit organization founded in 1995 to build trusted digital archives for scholarship. We work with the
scholarly community to preserve their work and the materials they rely upon, and to build a common research platform that
promotes the discovery and use of these resources. For more information about JSTOR, please contact support@jstor.org.
Swedish Society for Anthropology and Geography and Blackwell Publishing are collaborating with JSTOR to
digitize, preserve and extend access to Geografiska Annaler. Series A, Physical Geography.
http://www.jstor.org
EAST ANTARCTIC ICE SHEET SENSITIVITY TO
PLIOCENE CLIMATIC CHANGE FROM A DRY
VALLEYS PERSPECTIVE
BY
GEORGE H. DENTON1 2, DAVID E. SUGDEN3, DAVID R. MARCHANT2 3, BRENDA L. HALL',
and THOMAS I. WILCH2 4
of Geological Sciences and 2Institute for Quaternary Studies,
University of Maine, Orono, Maine, USA.
3Department of Geography, University of Edinburgh, Edinburgh, Scotland.
4Department of Geoscience, New Mexico Institute of Mining and Technology,
Socorro, New Mexico, USA.
'Department
Denton, GeorgeH., SugdenDavid, E., Marchant,David R.,
Hall, BrendaL. and Wilch,ThomasI., 1993:East Antarctic
Ice Sheet sensitivityto Pliocene climaticchange from a Dry
Valleysperspective. Geogr.Ann. 75 A (4): 155-204.
ABSTRACT.A case is made for the stabilityof the EastAntarctic Ice Sheet during Pliocene time from landscape development and surficialsediments in the Dry Valleyssector
of the TransantarcticMountains.The alternatehypothesisof
Pliocene meltdownrequiresatmospherictemperatures20?C
above present values, late Pliocene ice-sheet overridingof
the TransantarcticMountains, and possible rapid late
Pliocene mountain uplift of 1000-3000 m. The geomorphologicalresultssuggestthat these conditionswere not met
in the Dry Valleysregion. Rather,Pliocene mean annualatmospherictemperatureswere at most only 3? to 8?C above
present values; ice-sheet overriding occurred in Miocene
time (>13.6 Ma); Pliocene glacier expansion was limited;
and Pliocene surface uplift was only about 250 to 300 m.
These conclusions are based on field studies in Taylorand
WrightValleys, in the western Asgard Range, and in the
Quartermain Mountains. The chronology comes from
numerous 40Ar/39Ardates on in-situ volcanic ashes that
occur in stratigraphic association with unconsolidated
diamictonsin the western Dry Valleys,basalticlava flows interbedded with widespread tills in Taylor Valley, and reworked basaltic clasts in alpine moraines in east-central
WrightValley.The combined evidence from the Dry Valleys
region indicates that slope evolution was severely restricted
throughoutPliocene time, and has been so since at least the
middle Miocene. The implicationis that most of the DryValleys landscape is relict and that it reflects ancient erosion,
possiblyundersemi-aridclimateconditions,priorto middleMiocene time.
The Antarctic Ice Sheet and Pliocene Global
Change
The modern Antarctic Ice Sheet consists of about
30 x 106km3 of ice spread over an area of 13.6 x 106
km2 (Drewry et al. 1982; Drewry, 1983; Oerlemans
and van der Veen, 1984). It is divided by the Trans-
Geografiska Annaler *75 A (1993) ?4
antarctic Mountains into separate components in
West and East Antarctica (Fig. 1).
The marine West Antarctic Ice Sheet (3.3 x 106
km3; 6 m sea-level equivalent) rests on a rugged
bedrock floor (Fig. 2), much of it well below sea
level and likely to remain submerged if deglaciation and isostatic compensation were completed
(Drewry, 1983). Ice thickness is greatest in central
West Antarctica, where subglacial troughs exceed
-2000 m. In the interior, there is a complex of icesurface divides, domes, and saddles of generally
low elevation. Mountains that fringe central West
Antarctica protect the grounded marine ice. Most
surface topographic features dovetail into ice
streams, which drain 90 percent of the interior and
are the most dynamic components of the ice sheet
(Hughes, 1977; Bindshadler, 1991).
The terrestrial East Antarctic Ice Sheet (26 x 106
km3; 60 m sea-level equivalent) rests on a base that
is predominately above sea level (Fig. 2). Some
subglacial basins extend below sea level, although
many would rise above sea level if the ice sheet
were removed and isostatic compensation completed (Fig. 3a) (Drewry, 1983). Of particular importance to our discussion are the Wilkes and Pensacola subglacial basins, both situated inland of
the Transantarctic Mountains (Fig. 3b). On one
flank the East Antarctic Ice Sheet is dammed behind the TransantarcticMountains, but elsewhere
it terminates either in the ocean or in floating ice
shelves. It is higher and has a smoother surface topography than the West Antarctic Ice Sheet; central domes reach 3200-4000 m elevation (Fig. 1)
(Drewry, 1983). Inland flow discharges into the
Filchner-Ronne Ice Shelf or into major outlet
155
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND TI. WILCH
ontourline
'30?0_
on surfaceof Ice sh
Contour line (m) on surface of Ice sheet
Fig. 1. Present-dayAntarcticIce Sheet.
glaciers that pass through the Transantarctic
Mountains, whereas coastal flow discharges locally by ice streams into fringing ice shelves or into
the Southern Ocean. Between some of the major
outlet glaciers, local domes and ice ridges feed
glaciers that pass through the mountains or terminate in the western Dry Valleys. One dome that is
important to our discussion is the local Taylor
Dome, situated just inland of the Dry Valleys sector of the TransantarcticMountains (Fig. 1). The
Taylor Dome (formerly the McMurdo Dome)
reaches 2450 m elevation on the periphery of the
East Antarctic Ice Sheet. It feeds Taylor and
Wright outlet glaciers, which terminate in the Dry
Valleys (Drewry, 1982).
Ice shelves fringe much of coastal Antarctica.
Most occur in protected embayments or are anchored at pinning points. The two largest, the Ross
156
(536 x 103 km2) and Filchner-Ronne (532 x 103
km2) Ice Shelves, occur in deep embayments and
are fed by ice flow from both the West and East
Antarctic Ice Sheets.
Today the Antarctic Ice Sheet gains mass by accumulation of about 2.2-2.7 x 1015kg a-1 (Giovinetto and Bentley, 1985; Doake, 1985). Unlike
the Greenland Ice Sheet (and most former Northern Hemisphere ice sheets), the Antarctic Ice
Sheet does not now have surface melting ablation
zones. As a result, it currently loses mass almost
entirely by basal melting beneath ice shelves and
by calving of icebergs. In the absence of summertemperature control of surface melting margins,
the primary factors controlling the area and volume of the ice sheet are thought to be eustatic sealevel oscillations (which cause grounding-line advance and retreat; Denton and Hughes, 1981;
Geografiska Annaler ?75 A (1993) ?4
EAST ANTARCTIC ICE SHEET SENSITIVITY
Fig. 2. Subglacialtopographybeneath present-dayAntarcticIce Sheet, without isostaticadjustmentsfor the removalof the
ice sheet. Adapted from Denton et al. (1991) and based on Drewry(1983).
Stuiver et al. 1981), changes in interior accumulation (which today apparently varies with mean atmospheric temperature above the ground inversion; Robin, 1977), changes in ice shelves (and
hence in possible backstress on grounded ice), and
changes in ice dispersal into the surrounding ocean
from ice streams (Bindshadler, 1991).
A basic problem of global change involves the
response of the Antarctic Ice Sheet to warmerthan-present polar climates, including that of a
greenhouse world. This problem can be attacked
by developing numerical models based on the
dynamics of the present ice sheet (Huybrechts,
1993), and then subjecting these models to the
forcing expected in a greenhouse world. An understanding of ice-sheet behavior can also come from
examining the response of the Antarctic Ice Sheet
during past intervals of warmer-than-present cliGeografiska Annaler ?75 A (1993) *4
mate. In this regard, it has long been assumed that
the East Antarctic Ice Sheet has been a stable feature on the surface of the Earth since middle-Miocene time (Shackleton and Kennett, 1975; Savin et
al. 1975; Kennett, 1982; Miller et al. 1987). But this
paradigm has been severely challenged by the discovery of reworked marine diatoms in outcrops of
the Sirius Group along much of the Transantarctic
Mountains (Fig. 4). These deposits are taken to indicate that the East Antarctic Ice Sheet underwent
dramatic deglaciation during Pliocene warm intervals (Fig. 3b). The problem is important because it
involves nothing less than the evolution of polar
ice sheets and the global climate system.
Our purpose here is to examine the proposition
that the East Antarctic Ice Sheet indeed collapsed
during intervals of excess Pliocene warmth. We
start by reviewing the hypothesis of Pliocene de157
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
Fig. 3. (a) Map adaptedfrom Burkleand Pokras(1991)showingtopographyof Antarcticawith the ice sheet removedand the
bedrocksurfaceisostaticallyadjusted(after Drewry,1983). Contourintervalis 500 m; bold line indicates0 m contour.Note
that the PensacolaSubglacialBasin and muchof the WilkesSubglacialBasin lie above sea level. As shownin Figure4, several
SiriusGroupoutcropsoccureast of the PensacolaSubglacialBasin.The sourcefor marinediatomsin these SiriusGroupoutcrops has not been adequatelyaddressed.See also the figuresin Huybrechts(this volume), which show that removalof ice
from the PensacolaSubglacialBasin requiresan atmospherictemperaturerise of 20?C.
glaciation. This hypothesis yields certain predictions concerning Pliocene paleoclimate and icesheet history. We then test these predictions from
an examination of landscape development and surficial sediments in the Dry Valleys sector of the
TransantarcticMountains.
Pliocene Deglaciation Hypothesis
A major shift in the climate system of the Earth occurred in the Pliocene Epoch between 2.9 and 2.2
Ma ago (Shackleton, 1993). Late Cenozoic ice
ages began on a global scale during this decline,
first with a 41,000-yr and then a 100,000-yr
158
pulsebeat. Prior to 2.9 Ma ago, repeated intervals
of warmer-than-present climate marked Pliocene
time, and large ice sheets were limited to Antarctica. The most recent of these warm intervals occurred about 3.0 Ma ago, when high-latitude air
and sea-surface temperatures in the Northern
Hemisphere were elevated due to increased
meridional oceanic heat transfer (Dowsett et al.
1992). Other evidence for Pliocene warmth during
this and earlier intervals comes from such diverse
paleoclimatic indicators as European flora (van
der Hammen et al. 1971; Grube et al. 1986); the
lack of significant ice-rafted detritus in North Atlantic Ocean cores (Ruddiman and Raymo, 1988);
Geografiska Annaler *75 A (1993) ?4
EASTANTARCTICICE SHEETSENSITIVITY
(b) Sketch of Antarctic Ice Sheet duringinterval of extensive deglaciationin late Pliocene time postulated by Webb et al.
(1984) (adapted from Denton et al. 1991).A similarfigureis shown in Sugdenet al. (this volume). These sketches are based
on subglacialtopographywithout isostatic adjustment(after Drewry,1983).
terrestrial plant fossils in the Arctic (Funder et al.
1985; Matthews, 1989); African pollen assemblages (Bonnefille, 1976, 1983), micromammals (Wesselman, 1985), and Bovids (Vrba,
1988a, 1988b); molluscs along the western margin
of the North Atlantic (Stanley, 1985); diatoms,
radiolaria, and silicoflagellates from the Pacific
and Southern Oceans (Hays and Opdyke, 1967;
Ciesielski and Weaver, 1974); Andean flora
(Hooghiemstra, 1993); and some interpretations
of the marine oxygen-isotope record (Prentice and
Denton, 1988; Shackleton, 1993). Warmer-thanpresent Antarctic conditions are inferred from
early-to-mid Pliocene benthic foraminifers in
Geografiska Annaler ?75 A (1993) ?4
cores from DVDP sites 10 and 11 in eastern Taylor
Valley (Ishman and Rieck, 1992); from early
Pliocene diatoms and silicoflagellates in the subantarctic Southern Ocean (Ciesielski and Weaver,
1974; Abelman et al. 1990); and from microfossils
in raised marine deposits in the Vestfold Hills (Pickard et al. 1988).
The deglaciation hypothesis calls for Pliocene
meltdown of the East Antarctic Ice Sheet at 3.0 to
2.5 Ma ago (Fig. 3b) (Harwood, 1983, 1986;Webb
and Harwood, 1987; Barrett et al. 1992) and also
during earlier Pliocene warm intervals (Webb et al.
1984; Webb and Harwood, 1991). This hypothesis
carries with it the implication that the East Antarc-
159
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
30? 0? 30 ?
600
60?
85?
900
900 \
Wisconsin Plateau
(1)t.
Mtn.(t)
\Metaolcanic
,
<
~\9
\
^^
\^
Wisting (3)
\
/
ll(1)s
--ua
'.RobeQuartz
0
Mt. -Blackburn (2)
1200? WscoM'^^nsin@
Rngs*e
i_rMt.
Vl
\ \
\
Saltonstall
33i
/
(2)
/
>1200
<\\
\
\
'<
/
\
Roberts Massif (5,6)
Half CenturyNunatak(3)
Buttress (5)
'/ ?^ /Dismal
'
'T.
-i--....:~--:-";'t''
.[Co
~i;/:; ...t:: ^
/
6)
Mt.Wisting(3)
1
r,fSAS',.,
/ ;I ~,..:\~Z7(_~("~e2./
:..
\
\
Massif (
/J
.,
(/
~ 02/
I,.
<120?
\
O..Otway.Massitf(7)
^+~ DominionRange(7,8.10,11)
iley Point(10,11)
:. :.::
.
:
:..:.::::.:::::: ::.:^&
The Cloudmaker(10,11)
;Roth:(:3:~~
'i:iiMt~
~
,~ -C/~~
Mt. Sirius (8,9)
~1P~
/ Miller Range(10,11)
Ii;
!0FIa
.t To..p(I10.-II):
/
.!iii;;;;.:;::!:..
:!:..
..
::::.'..
.-
.'.t.-.- M t.Krkpatrick(0
a
Falls/a (10 11)
.
.M
. . . .
.
Markham Plateau (10,11)
1)
.
Mackellar(101....
Mt.
Mt. Egerton(12) r e
|.:::..'.'
.'
. . . .&.:.Mt.Tuatara(11,13
.
.
ton
::::::::nor:i:i::Hritanniaathe
?
* ? :*
800
:
Approximate Scae
800
*b
***'abe
km""l;:::::::::
*
Outcrop(s) of semi-lithifled Sirius drift
Location and literature reference
X
Erratic(s) of semi-lithified Sirius drift
Location and literature reference
Mtn.(14)
T
:::::::::::::::::::180?
Mt.'::'':'''::':
::::::::::::: :::d
.
'/:*:*:::::::::::Ros land:... -.
:::;:;:;:;:,.:.:::::
0I
/eCarapace
Taylor Valley /
WrightVlll
Nunatak.(16)
Allan
Nunatak (16)
2A'
/.,
y CoombsHills(17)
1800
Fig. 4. Location of Sirius Group outcrops and erratics in the Transantarctic Mountains. References are as follows: (1) Mercer
(1968); (2) Doumani and Minshew (1965); (3) Mayewski (1975); (4) LaPrade (1984); (5) McGregor (1965); (6) Claridge and
Campbell (1968); (7) Elliot et al. (1974); (8) Mercer (1972); (9) Barrett and Elliot (1973); (10) Prentice et al. (1986); (11) Denton et al. 1991; (12) J. Anderson, unpublished; (13) Faure and Taylor (1981); (14) Barrett and Powell (1982); (15) Brady and
McKelvey (1979); (16) H.W Borns Jr., unpublished; (17) Brady and McKelvey (1983). Fossil Nothofagus wood occurs in the
northern Dominion Range deposit. Adapted from Denton et al. (1991).
160
Geografiska Annaler - 75 A (1993) - 4
EAST ANTARCTIC ICE SHEET SENSITIVITY
Scale
0
20
40
60
km
Fig. 5. The Dry Valleysregion, is bounded by the MackayGlacierto the north and FerrarGlacierto the south. Map is from
U.S. Geological Survey,Mc MurdoSound 1:1,000,000sheet, 1974.
Annaler 75 A (1993)*4
Geografiska
161
SP^r'^
IV^&?
Fr
Shapeless^O
Mountain
163'00'
162'00'
161'00'
160'00'
(ZN
t-i
J-4
et
Mc
.
S
~~~~~~~~~q
*v
*^
~~~~~~~~~~~~~~~km
'
(
0~~~~~~~~~~~0C
i~)
it
F
r,
r,
Flom
ngFyl
19
-,,0
L
Labyrinth
!.
Dais
South
orih Fork
Fork
^
^Mt.
Ii
A G 0
DA
Taylor
1
~~~~~~~~~~
3
-Dome
El
7
t,
11
9 )Caventtish
C)
^^i1^
0
(a
S)
S.'i^^^^
Mt.
~~~ ^fii,fi
~ ~ ~
W
SD
SD
-4
(0.
1610'0'
1600UUU'
lOU
uu
16
101
UU
u
'
(uts~~~~~~t2e,~~~~~~5?
~ ~ ~~
~
Feather~
~ ~~~~~
~ ~~~~~~~~~~ a~'~~~~
~~~~~~~,Location
in~~~~~~~
Z1
vu
Fig. 6. (a) Sketchmap showinglocation of drillcores, place names, and stratigraphic sections used in text. Place names are numbered as follows: 1, Conrow
Glacier;2, BartleyGlacier;3, Meserve Glacier; 4, Hart Glacier; 5, Goodspeed
Glacier;6, Rhone Glacier;7, MatterhornGlacier; 8, LacroixGlacier; 9, Seuss
Al~r
-4
0.100W
U1J
Glacier; 10, Canada Glacier; 11, Com
13,Thomson moraine;14, LakeVanda
17, Hart Ash; 18, Prospect Mesa; 19, W
Glacier;21, Arena ValleyAsh.
Co
Mite
m^^'is, ^^W^?)
'% ^h
G)
00
(a
^~~ng
\~9
'/S,1
"Wev
3
'-
'
Fork
40,
100
~
~8~~~~~~~~~~~
~~~~
~ ~ ~ ~ ~ ~~km
,
CD
C4
:>
CoV
McI(etlSi 4
Mountain
.'"~
Taylorthr
Di
9ma Labyrinth
Co
26~~~~~~~~~~2
420
S Gast
1
Taylor
'Dome
\
^-
ii
FS
Mr^f\y,
p^
Fork
30
160'00'
'.~~
7*'"'"''
W
':^
1610-0'
^
^
^
^'i^
%
^&^^
^^^
t-S
^
.
^^^^^T-
Th'
Location
1b'00'
(b) Sketch map showing locations of figures and plates used in this paper and other papers
in this volume. Arrows indicate direction of view for each photograph. Numbers refer to
figuresas listed below. 1, is Fig. 9 (top) of Denton et al. (1993); 2, is Fig. 10 (top) of Denton
et al. (1993); 3, is Fig. 9 (bottom) of Denton et al. (1993); 4, is Fig. 23 (top) of Denton et
al. (1993); 5. is Fig. 11 (top) of Denton et al. (1993); 6, is Fig. 6 of Denton etal. (1993); 7,
is Fig. 16of Denton et al. (1993); 8,. is Fig. 17 of Denton et al. (1993); 9, is Fig. 13 (top) of
Denton et al. (1993); 10, is Fig. 13of Denton etal. (1993); 11, is Fig. 2 ofWilch etal. (1993b)
and Fig. 20 of Denton et al. (1993); 12, is Fig. 25 of Denton et al. (1993); 13, is Fig. 10 (bottoCiC)of Denton et tl. (1993); 14, is Fig. 11(bottom) of Denton et al. (1993); 15, is Figs. 6
and 7 of Wilch et al. (1993b) and Fig. 26 of Denton et al. (1993); 16, is Fig. 7 of Marchant
'../.___5.~
:
160 uu'
et al. (1993b) and Fig. 27 of Denton et al. (1
28 of Denton et al. (1993); 18, is Fig. 29 of D
al. (1993); 20, is Fig. 24 (top) of Denton et
al. (1993); 22, is Fig. 11 of Hall etal. (1993
of Hall et al. (1993); 25. is Fig. 10 of Hall
(1993b); 27, is Fig. 11 of Marchantet al. (1
29, is Fig. 2 of Marchantet al. (1993a); 30, i
of Marchantet al. (1993a); 32, is Fig. 4 of M
et al. (1993a); 34, is Fig. 8 of Marchantetal
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
tic Ice Sheet is susceptible to climatic warming of
the magnitude last reached in the Pliocene. By
analogy, Barrett et al. (1992) implied that the ice
sheet could again nearly disappear in a greenhouse
world a few degrees warmer than present
The deglaciation hypothesis is critically dependent on the inferred age and origin of reworked
marine diatoms in Sirius Group outcrops along the
Ross Sea sector of the Transantarctic Mountains
(Fig. 4) (Webb et al. 1984; Harwood, 1986). The
implication is that the diatoms originated in
marine basins, and that subsequently the East
Antarctic Ice Sheet expanded to cover the basins,
override the Transantarctic Mountains, and rework marine microfossils into the Sirius Group.
The only realistic location for the basins is in the interior of East Antarctica-hence the argument
that the ice sheet must have been sufficiently small
to expose the Wilkes and Pensacola Subglacial Basins to the sea. Diatom flora that include elements
indicative of warm (2?-5?C) interior East Antarctic seas are assigned ages varying from early to late
Pliocene on the basis of stratigraphicranges in subantarctic deep-sea cores (Harwood, 1986). Isotopic dating of a volcanic ash layer in the CIROS-2
core in the Ferrar Glacier trough (Figs. 5 and 6)
confirms the biostratigraphic age control for certain critical diatoms in the age range of 2.5 to 3.0
Ma ago (Barrett et al. 1992).
The deglaciation hypothesis would be strengthened on two counts if the Pliocene paleotopography was greatly different from the present. The
first is that the present-day Wilkes and Pensacola
Subglacial Basins must have been much deeper
than they are today to allow marine flooding (and
hence marine diatom growth) consequent on Antarctic deglaciation. Huybrechts (1993) calculates
that deglaciation must go nearly to completion
with far more recession than shown in Figure 3b
before the Pensacola basin near the South Pole becomes ice free; under these conditions isostatic
compensation would have lifted the Pensacola and
most of the Wilkes Subglacial Basins above sea
level.
The deglaciation hypothesis also would be
strengthened if the Pliocene-age Transantarctic
Mountains were considerably lower than now.This
would make it much easier to explain overriding
by an East Antarctic Ice Sheet that emplaced
marine microfossils into the Sirius Group at or
after 2.5 Ma ago (Elliot et al. 1991). Overriding of
lower mountains by a relatively thin temperate ice
sheet would not require polar ice shelves in the
164
Ross Sea, nor the existence of a West Antarctic Ice
Sheet (Huybrechts, 1993). Because Nothofagus
fossil wood occurs along with marine diatoms in
Sirius Group deposits now at 1800 m elevation
near Beardmore Glacier, the overriding ice sheet
is inferred to have advanced under temperate conditions into scrub vegetation with Nothofagus
(Webb et al. 1987; Webb and Harwood, 1987;
McKelvey et al. 1987, 1991; Carlquist, 1987;Webb
and Harwood, 1991; Hill et al. 1991). If it is assumed that Pliocene mountain elevations were indeed lower so that the Nothofagus scrub grew near
sea level, then the postulated late-Pliocene/early
Pleistocene Nothofagus growth would imply atmospheric warming of 20?-25?C above present
temperatures (Barrett, 1991, p. 46), a value also
suggested for early Pliocene time (Webb and Harwood, 1991, p. 215). Without uplift, the Pliocene
temperature in Antarctica necessary to allow
Nothofagus growth at the current high elevation of
the fossil wood-bearing beds would have been at
least 30?-35?C warmer than at present. In this regard, Pliocene-Pleistocene mountain uplift of
1000-3000 m has been postulated from faulting of
Sirius Group deposits near Beardmore Glacier
and from the current high elevation of the
Nothofagus wood beds (Webb and Andreasen,
1986; Webb and Harwood, 1987), as well as from
the "youthful-appearing"rift shoulder scarp along
the mountain front (Behrendt and Cooper, 1991).
Pliocene uplift, but of much less magnitude, also is
inferred from benthic foraminifers in DVDP cores
10 and 11 in eastern Taylor Valley (Ishman and
Rieck, 1992). A mountain barrier created by tectonic uplift is credited with forcing late Pliocene
climate change and transforming the character of
the Antarctic Ice Sheet from temperate to polar
(Behrendt and Cooper, 1991).
The possible Pliocene collapse of the East Antarctic Ice Sheet has been taken up by other
branches of earth science. High sea levels (35 ? 18
m) of mid-Pliocene age (- 3.0-3.5 Ma ago) on the
Atlantic coastal plain (Orangeburg, Chippenham,
and Thornburg Scarps) have been related to melting of the East Antarctic Ice Sheet (Dowsett and
Cronin, 1990; Krantz, 1991). Wide oscillations of a
dynamic East Antarctic Ice Sheet are used to explain fluctuations in the marine-oxygen isotope record and in biogenic productivity noted in some
deep-sea cores (Abelman et al. 1990; Krantz, 1991;
Ishman and Rieck, 1992). A multi-disciplinary
study (PRISM) designed to assess the Pliocene
warm peak at 3.0 Ma ago (Cronin and Dowsett,
Geografiska Annaler *75 A (1993) *4
EAST ANTARCTIC ICE SHEET SENSITIVITY
1991) has incorporated East Antarctic ice collapse
into an emerging global reconstruction (Dowsett
et al. 1992).
The Pliocene deglaciation hypothesis has testable implications. First, the East Antarctic Ice
Sheet must have overtopped (or nearly overtopped) the Transantarctic Mountains to emplace
Sirius Group outcrops on mountain surfaces now
at high elevations. The amount of thickening required depends critically on the surface elevation
of the mountains at the time of overriding. For
reasons already described, the deglaciation
hypothesis is more tenable if the Transantarctic
Mountains underwent extensive Pliocene-Pleistocene surface uplift. Second, the deglaciation
hypothesis requires that mountain overriding occurred after the oldest part of the stratigraphic
range of the youngest diatom included in Sirius deposits. This means that overriding was younger
than 3.0 Ma ago (Barrett et al. 1992) and perhaps
coincided with the Pliocene climatic deterioration
and consequent growth of Northern Hemisphere
ice sheets centered at 2.5 Ma ago (Elliot et al.
1991).
The third implication is of climatic warming so
substantial that Pliocene environments as far
south as 86?S latitude were at least subantarctic
and perhaps even southern Patagonian in character (Mercer, 1986). Subantarctic diatoms in Sirius
rocks at Reedy Glacier imply sea-surface temperatures of 2-6?C in the Pensacola Subglacial Basin
near the South Pole, and Nothofagus fossil wood
in Sirius deposits at Beardmore Glacier implies air
temperatures at least 20?C above present-day values if the mountains were much lower than now.
Further, the only plausible mechanism to cause deglaciation is to melt the ice sheet down from the
surface; this requires extensive surface melting
ablation zones that in turn demand a climatic
warming of 17?-20?C (Huybrechts, 1993). It has
been suggested that surging collapse of the East
Antarctic Ice Sheet may have caused deglaciation
of the inland basins, leading to warming (Harwood, 1986). However, Huybrechts (1993) concludes that alternate instability mechanisms
(marine ice-sheet instability, large-scale surging,
creep instability) are unlikely to have caused East
Antarctic deglaciation.
The Dry Valleys
The Dry Valleys sector of the TransantarcticMountains (Figs. 5 and 6) is particularly suitable for testGeografiska Annaler ?75 A (1993) ?4
ing predictions of the deglaciation hypothesis for
several reasons. First, it contains high-elevation
Sirius outcrops with enclosed marine Pliocene
diatoms, including a deposit on Mt. Feather, one
of the original sites used to postulate the deglaciation hypothesis (Webb et al. 1984). Second, it includes the CIROS-2 core, where the presence of
critical diatoms that also occur in Sirius deposits is
tied to an 40Ar/39Ardate of a volcanic ash layer
(Barrett et al. 1992). Third, the Dry Valleys feature
other cores (DVDP 10 and 11 in Taylor Valley,
CIROS-1 at the mouth of Ferrar Valley) and sections (Prospect Mesa in central WrightValley) that
contain microfossils that have been used to infer
warmer-than-present Pliocene marine paleoenvironments (Webb, 1972, 1974; Barrett, 1992;
Ishman and Rieck, 1992; Prentice et al. 1993).
Fourth, widespread surficial Pliocene and Miocene drifts and volcanic ashes are still well exposed
in the Dry Valleys because Quaternary glacier expansions were so small. These factors make it possible to address issues of Pliocene paleoclimate
and ice-sheet overriding of the mountains. A
cautionary note is necessary, however, because the
TransantarcticMountains potentially have a complex history of denudation and surface uplift.
Therefore, results from individual tectonic blocks,
including the Dry Valleys, do not necessarily apply
to the mountains as a whole.
The Dry Valleys feature the largest tract of icefree terrain in the TransantarcticMountains. Long
trunk valleys now largely free of ice (Taylor,
Wright, and the Victoria system) cross the mountains from the inland ice sheet to the Ross Sea.
These valleys are flanked by the Quartermain
Mountains and by the Asgard and Olympus
Ranges. The Wilson Piedmont Glacier blocks the
mouths of Wright and Victoria Valleys. The Taylor
Dome, a peripheral dome of the East Antarctic Ice
Sheet, feeds local East Antarctic ice into Taylor
and Wright outlet glaciers, which flow into the
western valleys. Inland ice flow is drained north
around the Dry Valleys by Mackay Glacier and
south by Mulock Glacier.
Today the Dry Valleys are cold, windy, and
hyper-arid. An excess of sublimation over precipitation at all elevations yields a distinct precipitation deficit (Chinn, 1980). For example, beside
Lake Vanda at 123 m elevation on the floor of
WrightValley, mean annual temperature is -19.8?C
(Schwerdtfeger, 1984), and mean annual precipitation is 10 mm water equivalent (Keys, 1980). Snowfalls are light; they can occur at any time of year
165
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
but are concentrated in the summer. Alpine
glaciers occur where wind-blown snow is concentrated into pre-existing theater-shaped embayments. Alpine glaciers are largest near the coast
where snowfall (and therefore wind-concentrated
accumulation) is greatest. Local glaciers are small,
or even nonexistent, in alcoves close to the polar
plateau. Because they are small and occur in a very
cold environment, these alpine glaciers are frozen
to underlying beds and are nearly free of debris
(Meserve Glacier in Wright Valley has a basal
temperature of -18?C; Bull and Carnein, 1968). In
contrast, the larger outlet glaciers can attain basal
melting conditions; for example, the deepest ice of
Taylor Glacier is wet based (Robinson, 1984).
Today, 90% of glacier ablation in the Dry Valleys is
by sublimation and only 10% by melting (Chinn,
1980). Small meltwater streams, all at relatively
low elevations, feed enclosed lakes in Taylor and
Victoria Valleys. In WrightValley, the Onyx River
flows westward along the valley floor from Lake
Brownworth (dammed and fed by Lower Wright
Glacier) to Lake Vanda (enclosed in the lowest
part of the valley). Meltwater from alpine glaciers
on the south valley wall also feeds the Onyx River.
In the remainder of this paper we discuss those
aspects of the Dry Valleys glacial and tectonic record that reflect on the problem of Pliocene icesheet stability. We start with landscape analysis of
the Dry Valleys morphology, then move on to a
consideration of late Tertiary landscape submergence and uplift, and conclude with a discussion of the surficial volcanic and glacial deposits
that rest on the landscape.
LandscapeAnalysis
General Statement
Landscape analysis of the Transantarctic Mountains is the foundation for deciphering the late Tertiary behavior of the East Antarctic Ice Sheet dammed along the inland mountain flank. This is because late Tertiary glacial deposits resting on the
landscape surface can not be interpreted in terms
of ice-sheet history without a knowledge of the topographic evolution of the mountains. Evaluating
the evidence for ice-sheet overriding of the Dry
Valleys sector of the Transantarctic Mountains is
particularly important, because the deglaciation
hypothesis requires such overriding after 3.0 Ma
(Barrett et al. 1992) to emplace high-elevation
Sirius deposits. Further, extensive Pliocene/Pleistocene uplift invoiced to explain both the high pre166
sent-day elevation of many Sirius outcrops, as well
as the high shoulder scarp of the Transantarctic
Mountains, should be evident in the sequence of
landscape development.
The Transantarctic Mountains comprise gently
tilted blocks of sedimentary rocks (DevonianTriassic Beacon Supergroup) which overlie Precambrian-Devonian basement and are intruded
and capped by Jurassic dolerites and basalts (Ferrar Supergroup). They form the faulted, uplifted,
and tilted shoulder of the highly asymmetric West
Antarctic intracontinental rift system (Fig. 7). In
southern Victoria Land they rise steeply from the
coast to elevations of 2000 m (Dry Valleys block)
to 4000 m (Royal Society block). The offshore Victoria Land basin is 10-14 km deep (Cooper and
Davey, 1987a, 1987b), giving a vertical displacement of 15-20 km.
The timing of tectonic events is not clear, because the mountains themselves contain only
sparse post-Jurassic deposits and because sedimentary units in the offshore Victoria Land basin
are poorly dated. It is thought that there were at
least two phases of rifting activity in the Ross Sea
(Tessensohn and Worner, 1991; Cooper et al. 1991;
LeMasurier and Rex, 1991). The first occurred in
the Mesozoic (possibly coincident with the separation of Antarctica and Australia about 90 Ma ago)
and was marked by diffuse crustal thinning. The
second began in the Eocene (Fitzgerald etal. 1986)
and continued to the present, although the details
are poorly known. The second phase involved uplift and tilting of the Transantarctic Mountains,
subsidence of the rift system in the Ross Embayment, and alkali magmatism beginning about 25
Ma ago (Kyle, 1990).
The spatial-temporal patterns of denudation
(rock uplift relative to the landscape surface) and
the resulting flexural isostatic response (along
with the interactions of morphotectonics, denudation, and ice-sheet behavior) are potentially very
important in the evolution of the Transantarctic
Mountains. New results from fission-track thermochronology show that denudation pulses along the
3000-km length of the Transantarctic Mountains
were complex (Fitzgerald, 1992). The implication
is that individual blocks within the Transantarctic
Mountains may have had differing denudational
histories. Here we consider the denudational and
surface uplift history of the Dry Valleys block, one
which is bounded by transform faults beneath
Mackay Glacier on the north and Ferrar Glacier
on the south (Fig 5). Identical fission-track denGeografiska Annaler - 75 A (1993) *4
EASTANTARCTICICE SHEETSENSITIVITY
Fig. 7. Interpretative map showing the modern plate boundaries and the West Antarctic Rift system (adapted from
LeMasurier,1990). Locationsof fault basins in the Ross Sea are from Behrendtet al. (1987) and Cooper et al. (1987a, 1987b).
VLB, VictoriaLand Basin;TR, TerrorRift;, CT, CentralTrough;EB, EasternBasin; RI, Roosevelt Island;BST,Bently SubglacialTrench;BSB, Byrd SubglacialBasin; and EF, EllsworthFault. For comparison,Rio GrandRift (RGR) and Basin and
Range Provinceshown at same scale in inset as Antarcticmap. MV indicatesMonumentValleyon the ColoradoPlateau and
shows location of photographsin the frontispieceand in Fig. 8.
udation profiles for the north walls of Wright and
Ferrar Valleys (Fitzgerald et al. 1986) suggest that
the Dry Valleys sector of the TransantarcticMountains has behaved as a single tectonic block since
the early Tertiary.
The Dry Valleys Landscape
Landscape Antiquity. Traditionally, the Dry Valleys landscape has been attributed to glacier erosion combined with salt weathering and wind deflation. In this scenario, the major east-west trending valleys were cut by outlet glaciers (Selby, 1971,
1990; Denton et al. 1984). Slow backweathering of
the initial steep glacial walls by salt weathering and
wind deflation in a frigid, arid environment produced rectilinear slopes below free faces, tent-like
ridges bounded by rectilinear slopes, or rectilinear
Geografiska Annaler ?75 A (1993) ?4
slopes beneath plateaus (Selby, 1971, 1974). Theater-shaped basins in intervening mountain ranges
between east-west trending valleys are attributed
to ancient cirque glaciation (Selby, 1971, 1974;Wilson, 1973; Denton et al. 1984) followed by prolonged backwearing of rectilinear slopes and free
faces, again by salt weathering and wind deflation
in a cold and rainless desert (Selby and Wilson,
1971; Wilson, 1973). In the high sandstone-anddolerite mountain ranges of the western Dry Valleys, this backwearing left residual buttes and
mesas.
As an alternate hypothesis, we suggest here that
the Dry Valleys landscape is largely inherited from
a climatic regime that existed prior to the imposition of the present cold, hyper-arid desert environment. This new hypothesis is based on the enormous antiquity of Dry Valleys landscapes and the
167
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
realization that significant slope evolution has not
occurred since near the end of the middle Miocene.
For some time it has been evident that the eastwest trending trunk valleys are ancient. Microfossils at the base of Dry Valleys Drilling Project
(DVDP) core 11 in eastern Taylor Valley are late
Miocene in age (Ishman and Rieck, 1992) (Figs. 5
and 6). In central Wright Valley the diatoms in
DVDP core 4A in Lake Vanda are middle-to-late
Miocene in age (Brady, 1979, 1982; Prentice et al.
1993). From 87Sr/86Sr ratios, reworked shells in
Jason glaciomarine diamicton beside Lake Vanda
are estimated to be 9?+1.5 Ma old, and in-situ
shells (Chlamys tuftsensis) in the Prospect Mesa
gravels on the floor of central Wright Valley are
thought to be as much as 5.5 ?0.4 Ma old (Prentice
et al. 1993). The lowermost sediments in the
CIROS-1 core taken from a deltaic environment
off-shore of Ferrar Valley are early Oligocene in
age (Barrett, 1989); the implication is that a protoFerrarValley existed by that time.
But the key breakthrough for our reevaluation
was the realization of the enormous antiquity of
Dry Valleys slopes and high-elevation planation
surfaces. The benches of middle TaylorValley have
remained essentially unmodified since at least
early-to-mid Pliocene time, as demonstrated by
numerous 4?Ar/39Ar whole-rock ages of surface
basanite volcanic deposits (Wilch et al. 1993). In
the western Dry Valleys, the evidence comes from
the undisturbed nature of old till sheets, moraines,
and colluvial deposits on rectilinear valley-wall
slopes that rise above planation surfaces or flattish
valley floors. New laser-fusion 4?Ar/39Ar dates of
single feldspar crystals from volcanic ashes in the
Asgard and Quartermain Mountains point to remarkable slope stability since middle-Miocene
time. Details are given elsewhere in this volume
(Marchant et al. 1993 a,b). Little-to-no slope
evolution has occurred in the high western Asgard
Range during at least the last 13.6 Ma. In Arena
Valley in the Quartermain Mountains, the preservation of Miocene- and Pliocene-age ashes on
steep valley slopes shows that the major bedrock
landforms are relict and that little slope evolution/
colluviation has occurred during at least the last
11.3 Ma. Likewise, the morphology of alpineglacier lateral moraines on the south wall of eastcentral WrightValley is preserved nearly intact, despite the fact that some of these moraines date to
>3.7 Ma in age (Hall et al. 1993).
The implication of these data is that the rectili168
near slopes and much of the overlying colluvium in
the Dry Valleys region are relict. Overall, we infer
that most Dry Valleys slopes are close to moribund
under the present hyper-arid cold-desert environment and that valley floors have undergone only
minor erosional modification since middle-to-late
Miocene time.
The current processes that shape the Dry Valleys landscape, particularly at elevations above
1500 m, feature salt weathering and wind deflation
(Selby, 1971, 1974). These processes have modified
pre-existing bedrock forms only modestly during
at least the last 11.3-13.6 Ma (Marchant et al. 1993
a, b). Salt weathering and wind deflation are particularly evident on exposed sandstone bedrock;
comparison with adjacent colluvium-mantled
slopes suggests that 0.5-3 m of bare sandstone
bedrock commonly has been removed by these
processes since late Miocene time (Marchant et al.
1993 a, b). Glacial erosion is now minimal. Outlet
glaciers may scour their beds where they are wet
based (Robinson, 1984), but cold-based alpine
glaciers are protective rather than erosional
(Holdsworth, 1970). Small meltwater streams
originate beside some glaciers below 1500 m elevation at the height of summer in the central Dry Valleys region and feed enclosed lakes on valley floors
(Chinn, 1980). The largest stream is the Onyx,
which flows inland seasonally along the floor of
Wright Valley from Lake Brownworth near the
Wilson Piedmont Glacier to Lake Vanda. Some
meltwater streams cut ravines and form small and
thin alluvial fans, but the discharges are rarely
high enough to move more than sand-sized particles. In general, only wind deflation, salt weathering, rockfall, minor colluviation, and minor erosion by small streams are now actively altering the
landscape.
We thus postulate that the Dry Valleys landscape is largely inherited from a previous climatic
regime. But what was that regime? To answer this
question, we compare the relict Dry Valleys landscape with possible counterparts elsewhere on the
planet. The semi-arid platform deserts of the Colorado Plateau (Oberlander, 1989) and southwestern Jordan adjacent to the Dead Sea Rift (Osborn,
1985) exhibit landscapes nearly identical with the
upland surfaces of the Dry Valleys region. Denudation of the near-horizontal strata of platform deserts yields a tabular landscape with rock plains
that head in escarpments and support isolated
mesas and buttes. The cuesta is the dominant component of platform deserts, which are therefore
Geografiska Annaler ?75 A (1993) 4
EAST ANTARCTIC ICE SHEET SENSITIVITY
Fig. 8. Detached buttes rising
above a flat-lying pediment in
Monument Valley on the Colorado Plateau. We postulate
that such landscapesas these on
the Colorado Plateau are very
similar,and perhaps formed by
the same mechanisms, as the
Dry Valleys upland landscapes
shownin Figures9 and 10.
termed arid or semi-arid cuestaform landscapes
(Oberlander, 1989). We now describe cuestaform
landscapes as a basis for comparison with Dry Valleys morphology.
Platform Desert Cuestaform Landscapes. A semiarid cuestaform landscape develops by backwearing and parallel recession of the cuesta scarp, accompanied by expansion of canyon systems and
headward growth of pediment rock plains (Oberlander, 1977; Doehring, 1977; Cooke et al. 1993).
The active element of the scarp is the caprock substrata, which erodes back to cause successive collapses of unsupported caprock. The substrata are
generally (but not always) less resistant to weatherGeografiska Annaler ?75 A (1993) ?4
ing than the caprock. They form a rampart leading
from the base of the cap rock (commonly a rectilinear slope) to the rock plain. This rampart can be
rilled or smooth, talus-covered or bare. The substrata rampart can begin at the base of the caprock
face or, more commonly, the break in slope at the
top of the rampart can occur in the substrata below
the caprock face. Denudation agents on the rampart include slope wash, creep, spring sapping
from groundwater, salt weathering, and granulation. Flash floods transport the material eroded
from the scarps along canyon floors and across
pediments. On the Colorado Plateau, Cenozoic
scarp retreat ranges from 0.5 to 7 km/myr'l
169
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
Fig. 9. Top: oblique aerial view
looking WNW of the upper planation surface with the inselberg of Shapeless Mountain
(2739 m) in the background.
The main escarpmentbounding
the surfaceis 600 to 900 m high.
The scallopedfront of this scarp
is the result of backwearingby
box canyons in the Olympus
Range, each bounded by a rectilinear slope topped by a free
face. The irregularitieson the
upper planation surface are
likely to be the result of areal
scouringbeneathoverridingice.
Bottom: an oblique aerial view
looking WNW across the western Asgard Range toward
Shapeless Mountain. In the
background, detached buttes
and mesas, remnants of the
upper planation surface, occur
at the western edge of main escarpment.In the foregroundare
similar buttes, together with
cols (wind gaps) that occur
along the mountaindividein the
westernAsgard Range.
(Schmidt, 1989). Embayments and headlands develop because of varying resistance of the cuesta
scarp to backwearing (for example, due to
lithologic or joint-density). Retreating headlands
can leave detached mesas and buttes, or inselbergs, that rise above an expanding pediment rock
plain (Frontispiece, Fig. 8). Rapidly retreating embayments may develop headward to form canyons
with stubby tributaries. The ramparts of cuesta
scarps can change form during recession as new
erodible substrates are exposed at different levels
(or old erodible substrates terminate) during general scarp recession. Moreover, climate changes
can cause alternating talus development and strip170
ping on ramparts (Gerson, 1982). This can produce talus flatiron relicts with intertalus gaps along
the front of retreating ramparts. Talus formation
or stripping has differing climatic thresholds, depending on rock types and many other factors.
Therefore, under some climatic conditions ramparts can be free of talus, whereas under other
climatic conditions the ramparts can be covered
with talus. Finally, climatic variations can alter the
shape of the cap rocks from domed to cliffed, if
there are variations in the importance of
slopewash erosion.
The role of groundwater sapping of the substrata may be of primary importance in scarp reGeografiska Annaler ?75 A (1993) 4
EAST ANTARCTIC ICE SHEET SENSITIVITY
Fig. 10.Top:oblique aerial view
looking south across the western OlympusRange towardthe
western Asgard Range. The
upper free face and lower rectilinear slope rise above the surrounding surfaces and flatfloored valleysthat make up the
intermediatesurface of Fig. 14.
In the right backgroundis the
upper surface of Fig. 14 in the
western Quartermain Mountains.
Bottom: oblique aerial view
looking northward across detached buttes in the Olympus
Range towardthe Insel Range.
The buttes are erosional remnants of the upper surface that
now rise above the intermediate
surfaceof Fig. 14.
cession and canyon development (Ahnert, 1960;
Laity and Malin, 1985; Howard and Kochel, 1988;
Laity, 1988). It has been suggested that groundwater flowing through the cap rock can converge to
cause increased substrata sapping, thus enhancing
headward migration. In fact, Laity and Malin
(1985) proposed that the erosional process of
groundwater sapping (and hence undercutting of
cap rock) is instrumental in producing networks of
theater-headed valleys on the Colorado Plateau.
Where sapping by laterally flowing groundwater is
dominant, the canyons have flat floors, hanging
valleys, steep walls, theater-shaped heads, and
constant width; drainage extension by groundwaGeografiska Annaler ?75 A (1993) ?4
ter sapping exploits joint patterns and is particularly important on the dipslope side of a cuesta
where groundwater is channelled preferentially. In
contrast, where overland flow predominates, the
valley heads are tapered, and the networks are
more arborescent than those cut by the groundwater sapping. There can be markedly asymmetric
erosional patterns on opposing sides of a divide.
The scarp along the backslope can be deeply
indented by prominent canyons that extend headward because of forward spring sapping. The scarp
along the updip slope can be either relatively
straight or moderately indented. When headwardcutting valleys breach the divide, the cap rock can
171
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
Fig. II11.
Top: oblique aerial view
looking eastward down Taylor
Valley. In the right foreground is
the winding inner fluvial valley,
with the two lobes of presentday Lake Bonney occupying the
sinuous floor and separated by
interlocking spurs. In the right
background is an intermediate
valley bench with remnant fluvial spurs that have been subglacially smoothed to form riegels.
In the left foreground is the
Rhone platform, a high valley
bench. The Rhone platform features the Rhone volcanic cone
(nearest the viewer) and the
West Matterhorn volcanic complex. TheThomson moraine and
drift, which covers the platform,
is older than the Rhone cone
and younger than the west Matterhorn volcanic complex.
Bottom: oblique aerial view
looking westward up Taylor Valley. In the lower center is the
inner winding valley. On the left
foreground is an intermediate
valley bench, with surface volcanic outcrops (dark).
be removed to form low cols (wind gaps) through
the divide.
Dry Valleys Escarpment Landscapes. The detailed
morphologic forms of the relict upland surfaces of
the Dry Valleys bear a remarkable resemblance to
morphologic forms in semi-arid cuestaform landscapes of platform deserts (Frontispiece; Figs. 8,
9, and 10). The similarities start with the fact that
the high-elevation, western Dry Valleys bedrock is
composed of near-horizontal strata (Beacon sedimentary formations, Ferrar dolerite sills, and
Kirkpatrick extrusions) resting on a crystalline
basement complex. The denuded landscape cut
172
into this bedrock is tabular. It shows several extensive rock plains separated by a prominent scarp,
with isolated mesas and buttes. The escarpments,
which are the marker features of the landscape,
show resistant cap rocks over ramparts (rectilinear
slopes). Flat-floored valleys with stubby tributaries have cut toward drainage divides. The result
ing indentations are asymmetric, both in the western Asgard and Olympus Ranges and in the Quartermain Mountains. North-facing valley networks
are longer, with more frequent stubby tributaries,
than south-facing networks. In the Asgard Range,
headward cutting has in places removed cap rock,
leaving cols (wind gaps). In places in the Olympus
Geografiska Annaler ?75 A (1993) 4
EAST ANTARCTIC ICE SHEET SENSITIVITY
Fig. 12. Oblique aerial views of
the Miller Glacier trough. Top:
view northwest up the trough.
The trough is cut by glacier ice,
and featuressteep sidewallsthat
arecut into an old landscapesurface of the Dry Valleysregion.
Bottom: view west towards the
steep west wall of the glacier
trough. Again this wall is cut
into an old Dry Valleys landscape.
Range, headward cutting has removed cap rock all
along the divide, and side-wall expansion has removed intervalley ridges to leave residual buttes
and mesas. These features closely resemble the
cuestaform landscapes of semi-arid platform deserts, except that the Dry Valleys uplands exhibit
cuesta scarps more prominently than the gentle
dip slopes, which are largely buried by inland ice.
Therefore, it is preferable to refer to these uplands
as escarpment landscapes.
Because of the remarkable similarities in the detailed morphology of the Dry Valleys uplands with
the platform terrain of the Colorado Plateau
(Oberlander, 1989) and of southwestern Jordan
Geografiska Annaler ?75 A (1993) *4
near the Dead Sea Rift (Osborn, 1985), it is tempting to suggest that the upper-level platform surfaces of the Dry Valleys also reflect denudation
under semi-arid conditions. This would be consistent with the suggestion of Selby (1974) that the
straight slope segments in the Dry Valleys represent a Richter-slope development which is most
likely in an arid or semi-arid environment with little vegetation.
However, an arid or semi-arid climate is not prerequisite for the evolution of platform structure,
because it is possible for escarpments to develop in
other climatic environments as long as accumulating debris is removed efficiently. For example, the
173
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
Fig. 13. Top: the Labyrinth in
upper Wright Valley, viewed
eastward down Wright Valley.
We postulate that the Labyrinth
was cut by subglacialmeltwater
streams.
escarpment terrains cut in horizontally bedded
sandstone in both southeastern and northwestern
Australia show canyons, scarps, and theater-shaped embayments; yet they probably developed in a
humid climate (Young, 1983, 1987). Also, Moon
and Selby (1983) argued that the form of the great
Namib escarpments and the western edges of the
southern Africa plateau, both of which probably
developed largely in semi-arid conditions, are controlled by rock mass strength. This implies a mode
of origin which is not dependent on climate. In this
regard, Augustinus and Selby (1990) argued that
characteristic Dry Valleys slopes are composed of
cliffs in strength equilibrium above Richter denudation slopes.
Bottom: meltwater channel and pothole system cut along
the western flank of a sandstonebutte in the westernAsgard
Range, viewed southward. Cliffs alongside tributarychannels in foregroundare about 30 to 40 m high. We postulated
that these channelswere cut by subglacialmeltwaterstreams
beneathnorthward-flowingice that overrodethe westernAsgard Range.
174
Valley Systems. Deep transverse valleys graded to
near sea level are cut into the upper-level landform
assemblage of the Dry Valleys region. The overall
valley systems, when reconstructed beneath adjacent glaciers, are broadly arborescent (Fig. 14).
They feature long trunk valleys that extend inland
to near the Transantarctic Mountains crest, where
the inland ice sheet now submerges the western
mountains. Some valleys, such as the middle part
of Wright Valley (Figs. 15 and 16), have flattish
floors and are nearly straight. Others, such as central Taylor Valley (Figs. 11, 17) and east-central
Wright Valley (Fig. 16), are sinuous and narrow,
with truncated or interlocking spurs. The mouths
of Ferrar, Taylor, Debenham, and Mackay Valleys
are below sea level, while the central portions of
Wright and Taylor Valleys are close to sea level.
Several of the deeper valleys show elevated valley
Geografiska Annaler ?75 A (1993) ?4
EAST ANTARCTIC ICE SHEET SENSITIVITY
(Note that for technicalreasons Figures23 and 24 are out of
sequence).
Fig. 23. Top: oblique aerial view looking southwardup the
Arena Valley, Quartermain Mountains. Moraine loops in
foregroundrepresent incursionsof southward-flowingPlioPleistocene lobes of TaylorGlacier into lower Arena Valley.
Well-developedrectilinearslopes occur on the valley walls.
Farin the backgroundis Mt. Feather,which at 2985 m elevation is the highest peak in the QuartermainMountains.The
Sirius Group occurs at 2650 m elevation on the northeast
flankof this mountain.
Bottom: oblique aerial view of
part of the Rhone platform on
the north wall of centralTaylor
Valley,lookingnortheast.In the
center of the photograph, the
Thomson moraine overlies a
volcanic complex showing two
eruptive events, one isotopically dated at about 3.5 Ma and
the other at about3.0 Ma (Wilch
et al. 1993b).
benches. The rock floors of central and eastern
portions of Wright and Victoria Valleys today have
a reverse inland slope.
The western valley floors rise inland in theater-
shaped steps over resistant strata. Some steps have
steep risers; some have degraded risers. The heads
Geografiska Annaler ?75 A (1993) 4
of all the major valleys are steep walled, semi-circular indentations cut into high planation surfaces. For example, the Taylor-Ferrarvalley system has a major step at Cavendish Rocks; other
steps occur beneath upper Taylor Glacier. The
deep portion of WrightValley extends far inland to
175
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
Fig. 24. Top:oblique view of the
Arena Valley Ash deposit. The
in-situash overlies a buriedventifact desert pavementat 1410m
elevation in central Arena Valley. Laser fusion 4?Ar/39Ar
analysesof 18individualcrystals
yield an age of 4.343?0.108 for
the ArenaValleyAsh (Marchant
et al. 1993b,c, d).
Bottom: in-situ volcanic lapilli
on top of a well-preserveddesert pavement on the Rhone
platform near the Thomson
moraine in Fig. 23. 4"Ar/39Ar
analysesindicatethat this lapilli
(and hence the underlying desert pavement) is 2.97?0.14 Ma
(Wilchetal. 1993b).
semi-circular steps that rise abruptly up to the
level of the Labyrinth at the head of North and
South Forks; another steep semi-circular step at
Airdevronsix Icefalls separates the Labyrinth surface from an upper planation surface. Victoria Valley has a steep, theater-shaped head now covered
with Victoria Upper Glacier.
All the deep transverse trunk valleys have small,
high-level tributaries that are cut into an upperlevel planation surface. These tributaries grade to
an intermediate planation surface or to valley
benches and are now left hanging above the major
trunk valleys. These hanging tributaries are stubby
and broad; some have steep, theater-shaped
heads. Only a few major tributaries feed directly
into the deep transverse valleys. Taylor Valley has
176
two major tributaries (Beacon and Arena); both
enter upper Taylor Valley from the south. Wright
Valley has only one major tributary (Bull Pass),
which hangs above the central north wall. Victoria
Valley has several major tributaries. One is Barwick Valley, which has a steep semi-circular head
cut into an upper planation surface. Others are
Balham and McKelvey Valleys, which have tapered heads and fluvial notches at the inland escarpment.
Under what climatic regime were the valley systems dissected into the upper platform surfaces? It
is obvious from the interlocked and truncated
spurs, as well as from the winding valley segments,
that fluvial erosion was important in eroding the
middle and eastern reaches of Wright, Taylor, and
Geografiska Annaler ?75 A (1993) ?4
EASTANTARCTICICE SHEETSENSITIVITY
Fig. 14. Map showing the distributionof the main landscape
types and the relationshipto reconstructed river valleys. The
upper surface in the west and
frontingscarpgive wayto dissected mountain landscapes towards the coast. The sites of
Sirius deposits on the old surfaces are indicated.
Ferrar Valleys (Victoria Valley does not preserve
spurs). These valley segments are cut in basement
rocks and lie well below the high planation surfaces. The valley walls here are generally long,
rectilinear, and free of extensive colluvium. This is
consistent with slope formation in a weathering-limited environment. The inland reaches of the valleys approach the high-elevation planation surfaces and are cut into slightly dipping Beacon sandstone and Ferrar Dolerite. Here the valley slopes
and patterns closely resemble those on the upper
planation surfaces that we suggest were cut in a
semi-arid environment. It is thus possible that the
valley systems were cut through the Transantarctic
Mountains under a semi-arid climatic regime, at
least in part. Both groundwater and surface flow
may have been important in the backwasting necessary to cut the tributarybox canyons and the abrupt steps of the western valley floors. But the
drainage basins (both surface and groundwater) of
these valley systems were sufficiently extensive to
support overland stream flow on the basement
Geografiska Annaler ?75 A (1993) ?4
rocks in the deep lower valleys. The formation of
weathering-limited rectilinear slopes in the crystalline and metamorphic rocks of the lower valleys is
also consistent with a semi-arid environment.
However, it is also possible that the valleys were
cut under a more humid climate that allowed development of weathering-limited slopes. This
would be consistent with the fossil impression of a
Nothofagus leaf in the CIROS-1 drill core that
penetrated Oligocene deltaic and glacial sediments at the mouth of Ferrar Glacier trough (Barrett, 1989; Mildenhall, 1989).
Glacial Erosion. Although we suggest that escarpment landscapes are widespread in the Dry Valleys, there also is evidence that glaciers were important in landscape evolution. Middle-Miocene
tills rest on the floors of theater-headed valleys in
the western Asgard Range and in the Quartermain
Mountains (Marchant et al. 1993 a, b), and perhaps on the floors and walls of WrightValley (Hall
et al. 1993). Spurs in Taylor and Wright Valley are
177
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
Fig. 15. Oblique aerial photographof upper WrightValley and adjacent ice-free areas in the Asgard (right) and Olympus
Ranges (left). View is eastwardtoward the Ross Sea. In the foreground,WrightUpper Glacier extends about 7 km down
WrightValley,where it terminatesin the Labyrinth.Cliffedwalls in upperWrightValleygive way to smooth rectilinearslopes
in centraland lowerWrightValley.Inselbergsand buttes rise above the intermediateplanationsurfacein the OlympusRange.
The buttes are progressivelysmallerand more restrictedwith increasingdistancefromthe mainscarpfront. US Navy,VXE-6
photograph,TMA 2448, 244.
truncated. Victoria and central Wright Valleys are
overdeepened and straightened. In Taylor Valley,
it is possible that the relict valley represented by
sidewall benches was straightened by axial glaciers
before fluvial erosion of the inner winding valley
(itself with truncated spurs). Cliffed troughs occupy some of the main valleys, such as those of the
Mackay and Miller Glaciers (Fig. 12). Other cliffed troughs truncate valley benches in Wright, Victoria, and Taylor Valleys. All are marked by
178
straightened valley sections and the sharp truncation of the gentle upper slopes. Areal scouring is
another glacial feature. Typically, landscapes of
areal scouring have a relief of tens of meters and
bear evidence of roches moutonn6es or streamlined forms. Areal scouring occurs on the dolerite
of the upper planation surface, the valley benches,
and the rolling hills and lowlands of the coast. In
several places areal scouring is associated with
major subglacial meltwater channel systems, for
Geografiska Annaler *75 A (1993) ?4
EAST ANTARCTIC ICE SHEET SENSITIVITY
Fig. 16. Oblique aerial view of central and lower Wright Valley, showing straightened valley section in foreground and truncated spurs near the valley mouth. A prominent valley bench occurs in the lower left corner; compare with similar benches
in Fig. 17. US Navy, VXE-6 photograph, TMA 1564, f-31, 162.
example, at the head of WrightValley (Labyrinth)
(Fig. 13) and Miller Glacier Valley. Details of such
channels in the Asgard Range are described by
Sugden et al. (1991) (Fig. 13). A final glacial feature concerns the small cirques which are scattered
through the Dry Valleys region. The best examples
are cliffed, arcuate basins excavated into rectilinear slopes in the Kukri Hills and on the northern
walls of the Ferrar Glacier trough.
The CIROS-1 core taken from the seafloor
offshore of Ferrar Glacier trough shows intermitGeografiska Annaler *75 A (1993) ?4
tent glaciation as early as early Oligocene time
(Barrett, 1989). We thus infer that glaciers filled
the developing Dry Valleys drainage systems intermittently since early Oligocene time.
TransantarcticMountains Evolution
The hypothesis that landscape evolution of the
Dry Valleys block extends far back into the Tertiary under warmer-than-present climate that allowed fluvial planation and downcutting carries
179
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
Fig. 17. Oblique aerial view of middle and lowerTaylorValley,looking to the east. Mount Erebus and McMurdoSound are
in the background. Prominent benches and truncated spurs occur in the center of the photograph. Remnant fluvial spurs separate two lobes of Lake Bonney, which fill the old fluvial valley. US Navy, VXE-6 photograph, TMA 540, f-31, 006.
with it the implication that the pattern of denudation and its effects can be tied into the tectonic
evolution of the TransantarcticMountains. This is
because denudation pulses of backwearing and
fluvial erosion would reflect changing base levels
at the mountain front. It should be pointed out
that such linkage would not occur under the traditional hypothesis of landscape development of the
Dry Valleys because salt weathering, wind deflation, and glacial erosion in a rainless, frigid desert
have no obvious linkage to base-level changes induced by tectonism. In fact, in this environment
180
uplift could actually slow erosion by elevating the
landscape surface into a colder regime. It also follows that under our interpretation of paleoclimate, the linkage between erosion and tectonism
would cease with the imposition of a hyper-arid,
cold-desert climate by the end of middle-Miocene
time. Overall, a knowledge of elevation and shape
changes of the land surface is of essential importance in interpreting late Cenozoic ice-sheet behavior (particularly ice-sheet overriding) from glacial drifts that rest on the Dry Valleys landscape.
Geografiska Annaler ?75 A (1993) 4
EASTANTARCTICICE SHEETSENSITIVITY
E
W
Fig. 19. Schematiccross-section
showing the relationship between the various landscape
types from the inlandupperplanation surfaceto the sea.
Obelisk
Mt.Newall
Faults
Geomorphological Map. To exploit the postulated
linkage between denudation and tectonic evolution, we constructed a new geomorphological map
at a scale of 1:250,000 (Fig. 18) (see Map 1, fold out
on back cover). This map, which covers some 7000
km2 of the Dry Valleys region, ties together the relict erosional features into overall patterns that
show the sequencing of landscape development.
The Dry Valleys landscape was classified initially
on the basis of form, with further subdivision on the
basis of inferred process. Thus the primary classification on the morphogenetic map is into categories
such as planation surfaces, cliffs, and rectilinear slopes. The secondary classification distinguishes the
types of cliffs, for example, those caused by glacial
activity and those formed by subaerial processes.
Inevitably, there are many difficulties involved in
such an approach, the most important of which is
that similar forms may result from different processes. Nevertheless, the results are internally consistent and, even with the inherent limitations, produce new insights. The mapped landforms and their
relationship to the tectonic history of the Dry Valleys block of the TransantarcticMountains are discussed in detail elsewhere (Sugden et al. 1994).
Here we present only the main findings.
Three main landform associations are highlighted on the geomorphological map. The first comprises the planation surfaces and erosional remnants which form the highest tablelands, summits,
and coastal lowlands. The surfaces rise in steps
from the coast; three levels are separated by clear
escarpments (Fig. 19). The highest surface, commonly underlain by Ferrar dolerite, occurs at elevations of 2000-2400 m on the inland flank of the
Dry Valleys and is bounded on its coastal side by a
scarp 500-1200 m high. An intermediate surface
occurs at an elevation of 1700-1850 m on interfluves between the Dry Valleys and is increasingly dissected towards the coast until it is termi-
Sea Level
nated by the abrupt 1600-m downfaulted mountain front above the coastal lowland. This lowland,
which is generally below 200 m elevation and
could be a downfaulted part of the intermediate
surface, slopes gently seaward.
Taken together, the pattern of the surfaces, escarpments, and associated landforms demonstrates that the landscape originated by scarp retreat
from the coast. Scarp retreat is well illustrated by
the way the buttes and inselbergs in the Asgard and
Olympus Mountains are progressively smaller and
more restricted with increasing distance from the
main scarp front. Such backwearing with a series of
erosional scarps parallel to the coast is duplicated
in southern Africa and in Australia, where it reflects pulses of erosion penetrating inland from the
coast due to base-level changes. Given suitable cap
rocks, more than one pediplain surface can result
from an initial base-level change.
The second landform association consists of the
valley systems. The major trunk valleys extend
through the faulted mountain front. The sides of
these valleys are largely rectilinear in form, although usually they have a concave slope at the
foot rather than an abrupt break of slope. These
rectilinear valley sides pass uniformly across the
mountain front, suggesting that faulting occurred
prior to the onset of the present cold-desert climate that terminated all but minor landscape modification. The frontal fault zone truncates an inner
bench of TaylorValley, implying that some faulting
postdated initial valley cutting. A final point is that
the bedrock floors of Victoria and Wright Valleys
now have a reverse, inland slope. One explanation
is that tectonic backtilting occurred after the rivers
ceased to flow so that they could not excavate new
channels to the changing base level.
The third landform assemblage is glacial. The
significance of the glacial landforms in the context
of the geomorphological map is that glaciers have
See MAP 1,fold-out on back cover
Fig. 18. Geomorphologicalmap showing the distributionof the landscapetypes and the relationshipto reconstructedriver
valleys.
Geografiska Annaler ?75 A (1993) *4
181
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
not contributed the main features of the landscape
in the Dry Valleys region, except for a few troughs.
However, glacial erosion, although relatively
minor, has affected all surfaces from the highest
plateaus to the valley bottoms and coastal lowlands. This implies that glacial modification either
accompanied or followed the main stages of landscape dissection.
Landscape Subsidence and Uplift. A major implication of this landscape analysis and the limiting
ages for the old surfaces is that the Dry Valleys
morphology was essentially formed prior to the
end of middle-Miocene time and that many of the
erosion surfaces are probably much older (Marchant et al. 1993 a, b). We postulate that the imposition of a cold-desert environment halted major
landscape development in the Dry Valleys by the
end of middle-Miocene time. Thus any subsequent
elevation changes must be inferred from sediments superimposed on the preexisting landscape
surface.
We recognize substantial subsidence (>400 m)
of the ancient landscape surface after denudation
had effectively ceased, followed by surface uplift
of about 300 m since 3.5 Ma ago. The evidence for
landscape subsidence comes from fjord sediments
and from the present elevation of fluvial valley segments. The inner, lower fluvial segment of Taylor
Valley is now close to (or even partly below) sea
level. The floor of the Lake Bonney basins, Which
are fluvial features of this inner valley (Fig. 17), is
less than 30 m above sea level (C.H. Hendy, written communication, 1993). Between Canada and
Suess Glaciers, the floor of this inner valley is 90 m
below sea level (Mudrey and McGinnis, 1975).
The preservation of spurs and sinuosity suggests
that this inner fluvial valley has not been modified
significantly by glacial erosion. Hence the implication is of subsidence of Taylor Valley since fluvial
erosion of the narrow, inner valley. Likewise, Ferrar Valley shows truncated and smoothed spurs
that extend to near sea level. The bedrock floor of
lower Ferrar Valley is now 165 m below sea level.
Certainly Ferrar Valley has experienced much
more glacier erosion and straightening than has
Taylor Valley. Nevertheless, fluvial features are
preserved, and the situation is compatible with
submergence.
Perhaps the best evidence for landscape subsidence comes from WrightValley. Fjord sediments
in central Wright Valley date to 9?1.5 Ma (Jason
glaciomarine diamicton now at 3-250 m elevation)
182
and mid-to-early Pliocene (Prospect Mesa gravels
now at 165 m elevation) (Webb, 1972, 1974; Prentice et al. 1993). The fjord(s) were about 100 m
deep, and hence reached to about the present-day
300-m contour (Prentice et al. 1993). Any fjord in
central Wright Valley must have flooded east-central Wright Valley, simply because it affords the
only access for seawater. The floor of east-central
WrightValley, now at 190-270 m elevation, shows
evidence of fluvial erosion in the form of interlocking valley spurs. Hence the fjord(s) flooded a fluvial segment of WrightValley to a present-day elevation of about 300 m. The suggestion is of valley
submergence of as much as 300 m, with an unknown but comparatively small adjustment for
sea-level change.
If it is assumed that the Dry Valleys formed a
coherent tectonic block from north to south, then
the combined data from Wright and TaylorValleys
suggest a total subsidence in excess of 400 m. The
evidence from WrightValley fjord sediments indicates that this subsidence had occurred prior to
9+ 1.5 Ma ago, and that it persisted until sometime
after 3.5 Ma ago (Hall et al. 1993). This is consistent with interpretations of benthic foraminifers in
eastern Taylor Valley that a deep but shallowing
fjord existed between at least -6.0 Ma and 3.4 Ma
ago (Ishman and Rieck, 1992).
Following substantial subsidence, the Dry Valleys landscape surface was uplifted about 300 m in
the last 3.5 Ma without reactivating the frontal
faults now exposed above sea level. The evidence
comes from surface deposits and sediment cores.
In Wright Valley, the Pliocene fjord surface, represented in part by the Prospect Mesa Gravels,
reached to the present-day 300-m contour, marked
by the truncated ends of lateral moraines of small
alpine glaciers that flowed down the south valley
wall and calved into the fjord (Hall et al. 1993).
The youngest of these moraines dates to < 3.5 Ma
ago (Hall et al. 1993); this limiting fjord age is consistent with the date based on foraminifers from
the C. tuftsensisshell beds at Prospect Mesa in central Wright Valley (Webb, 1972, 1974). Thus, we
conclude that subsequent to 3.5 Ma ago, the fjord
drained and the valley-floor threshold was lifted to
its current elevation of 270 m (Calkin, 1974). The
magnitude of the uplift is such that it cannot be
explained by Pliocene sea level changes (Haq et al.
1988; Hall et al. 1993).
In Taylor Valley, 40Ar/39Aranalyses of numerous in-situ subaerial cinder-cone deposits of
Strombolian-style eruptions provide limits on surGeografiska Annaler *75 A (1993) ?4
EASTANTARCTICICE SHEETSENSITIVITY
"
~"
/
....
'
\
m
815.7
1V
(
)
Marr
Fig. 20. McMurdoVolcanicGroup outcropsin TaylorValley.Volcanicoutcrops(shaded) includeboth in-situand remobilized
volcanicrocks.Age estimatesof volcanicevents are listed at outcroplocalitiesalong with lowest in-situoutcropelevation. See
Table1 for descriptionof age and elevation data (adapted fromWilchet al. 1993a).
face uplift inland of the frontal faults (Wilch et al.
1993a, b). Twenty-seven 4?Ar/39Ar incremental
heating analyses on whole-rock samples from outcrops of known elevations define 14 volcanic eruptions ranging in age from 3.89 to 1.50 Ma ago (Fig.
20; Table 1). Because Taylor Valley opens directly
onto the Ross Sea, the results show that any surface uplift during the past 2.57 Ma was less than
300 m (elevation data corrected for a potential extreme Pliocene sea-level lowstand; Haq et al.
1988). If more uplift had occurred, the volcanic deposits would exhibit submarine rather than subaerial characteristics. Seaward of a prominent frontal fault, fjord sediments in eastern Taylor Valley
contain benthic foraminifers which indicate that
some uplift has indeed occurred after 3.4 Ma ago
(Ishman and Rieck, 1992); however, the ages of
the subaerially erupted volcanic deposits suggest
that this uplift was limited to a few hundred meters, which is consistent with WrightValley data.
Tectonicand StructuralRelations. Two lines of evidence suggest that denudation of the Dry Valleys
region was associated with rock uplift that was
greatest near the coast, declining inland. One is
the tilt of the basement surface (Kukri peneplain),
which is assumed to have been approximately horizontal at the time of its formation before being coGeografiska Annaler *75 A (1993) ?4
vered by a uniform layer of Beacon sediments
(Gunn and Warren, 1962). In the inland parts of
Taylor and WrightValleys, this basement surface is
exposed at an elevation of 700 and 900 m, respectively. It rises toward the coast at an angle of 2?-3?,
reaching elevations of about 1800 m near the
mountain front. The basement is downfaulted at
the mountain front, as shown by its relationship
with particular dolerite sills in a series of fault
blocks (Fitzgerald, 1992). This pattern of maximum uplift near the faulted mountain front, declining inland, is mirrored by results of apatite fission-track analysis. Gleadow and Fitzgerald
(1987) plotted the dislocation of a pre-uplift, approximately horizontal isochron, which reveals
the same faulting with a downthrow of at least 1650
m. As mentioned above, the fact that rectilinear
valley sides cut across the faults suggests that much
of the faulting was completed before the valleys
were vacated by rivers and before the imposition
of a hyper-arid, cold-desert environment.
These structural relationships give some insight
into the depth of denudation. Making certain assumptions, which are discussed elsewhere (Sugden et al. 1994), we suggest that a wedge of rock
some 4-5 km thick has been eroded from the
mountain front, declining to about 1 km at the
western rim of the Dry Valleys, 75 km inland. Apa183
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
Table 1.
Summary of 40Ar/39Ar Isotopic Age, Paleomagnetic,
Sample
TotalGas
Age (Ma)
Plateau
Age (Ma)
/%39
and Elevation Data from McMurdo Volcanic Group rocks in Taylor Valley
40Ar/36Ar %39 Pol. Elev. Elev.
Isotope
lowest highest
Correlation
(m)
(m)
Ages (Ma)
East RhoneSite: Twoevents at about2.7 and 1.5 Ma.
TWV87019 1.49 ?0.14 1.50 ?0.05 93.7 1.50 ?0.05
295.5
86K-20
301.3?14.9 100.0
2.88 ?0.35
2.65 ?0.31
61.7 2.71 ?0.24
West Matterhom
Site: Twoevents at about3.5 and 3.0 Ma.
3.23 ?0.47 3.08 ?0.30 59.1 2.97 ?0.14
86K-06
TWV87048 3.03 ?0.45 3.12 ?0.15 70.7 2.98 ?0.14
TWV87047 3.23 ?0.43 3.18 ?0.33 88.8 3.05 ?0.08
TWV87037 2.87 ?0.61 3.03 ?0.21 69.2 3.26 ?0.08
TWV87037 3.03 ?0.45 3.08 ?0.28 96.4 3.15 ?0.16
3.25 ?0.42
86K-021
TWV87042 3.38 ?0.85
TWV87032 3.48 ?0.22
3.47 ?0.10
3.52 ?0.20
3.47 ?0.17
64.1 3.47 ?0.05
72.5 3.49 ?0.07
61.8 3.56 ?0.09
93.7
296.9 ?2.2
296.9 ?10.4
297.5 ?3.5
290.2 ?3.7
290.0 ?13.8
100.0
94.6
100.0
100.0
69.2
292.9 ?3.6
299.3 ?5.1
292.8 ?2.6
72.3
82.3
83.5
R
R
Notes
K/Arage
(polarity)(from
Armstrong,1978)
Intactcone, withproximalpyrociasticanddistal
lavaflowdeposits. Twovolcanicevents established
on the basis of distinct40Ar/39Ar
ages. Paleomagnetk
datasuggest the oldereventwas <2.60 Ma.
671
1108
1.84 + 0.26 (R)
671
1108
2.05 ? 0.36
602
1250
3.42 ? 0.20
3.47 ? 0.28 (R)
Erodedpyroclasticandlavaflowdeposits. Two
volcanicevents establishedon the basis of
distinct40Ar/39Ar
ages. Paleomagneticdatasuggest
the oldereventwas >3.55Maandthe youngerevent
was between3.05 and 3.13 Ma.
602
1250
East Matterhorn
Site: One eventat about3.7 Ma.
TWV87016A3.80 ?0.16 3.59 ?0.13 63.8 3.74 ?0.25
296.4 ?19.1 100.0
R
828
1013
no data
Erodedpyroclasticand lavaflowdeposits.
East La CroixSite: One eventat about2.6 Ma.
TWV87070 2.58 ?0.17 2.55 ?0.21 55.3 2.57 ?0.09
294.4?7.4
100.0
R
209
> 369
no data
Erodedweldedpyroclasticdeposits.
East BornsSite: Twoevents at about1.6 and2.5 Ma.
1.62 ?0.15
TWV88015 1.68 ?0.62 1.73 ?0.32
TWv87071 1.61 ?0.14 1.65 ?0.10 87.1 1.70 ?0.06
TWV88014 1.81 ?0.27 1.71 ?0.18 55.9 1.79 ?0.06
TWV87027 1.81 ?0.17 1.82 ?0.12 79.6 1.84 ?0.06
297.6 ?5.5
291.6 ?4.5
294.6 ?1.5
292.6 ?7.1
85.4
100.0
69.2
100.0
R
382
1677
1.99 ? 0.24 (R)
1.89 ? 0.22
1.57 ?0.12 (R)
2.05 ? 0.12
ventsites Including
partially
Muitiple
erodedcone at 964 m elevation.
Erodedpyroclasticand lava-flowdeposits.
Twovolcanicevents establishedon
the basis of distinct40Ar/39Ar
ages.
53.6 2.53 ?0.13
64.5 2.61 ?0.04
292.7 ?9.6
296.2 ?2.0
93.9
100.0
R
587
994
Sollas Site (west): One eventat about2.5 Ma.
TWV87066 2.46 ?0.28 2.53 ?0.23 90.1 2.51 ?0.04
295.6 ?1.7
86.7
R
416
1116
3.19 ? 0.18 (N)
3.03 ? 0.14 (R)
Feederdikesand sills withspatter-fedlava.
Outcropsimmediatelyadjacentto SollasGlacier.
SollasSite (north):One eventat about1.9 Ma.
TWV87014 2.14 ?0.34 2.04 ?0.30 61.9 1.89 ?0.13
300.5 ?10.5 94.9
R
376 no data
R
324
TWV88017 2.40 ?0.18
TWV88021 2.65 ?0.19
2.54 ?0.17
2.63 ?0.06
Sollas Site (northand east): Threeevents at about3.9, 3.6, and 2.2 Ma.
295.6 ?1.9 85.3
TWV87012N2.13 ?0.26 2.20 ?0.08 77.9 2.19 ?0.04
295.0 ?14.5 87.9
TWV87007 2.09 ?0.25 2.28 ?0.07 53.2 2.20 ?0.18
291.4 ?4.8 100.0
2.23 ?0.07
TWV87008 2.10 ?0.14 No Plateau
TWV87010 3.58 ?0.27
No Plateau
TWV87009 3.91 ?0.25
3.87 ?0.15
ManrSite: One eventat about2.5 Ma.
TWV87004 2.67 ?0.59 2.64 ?0.39
TWV87102F 2.54 ?0.63 2.50 ?0.35
754
Erodedlavaflow. Eventestablishedon the basisof
a distinctpaleomagneticsignature.
3.08 ? 0.20 (R)
2.73 ? 0.12
Erodedpyroclasticand lava-flowdeposits.
296.6 7.5
100.0
R
672 no data
4.76 ? 0.24 (R)
Ventpondlava deposit.
85.4 3.890.06
295.9 ?2.0
100.0
N
816 no data
4.62 ?1.40
data
feederdike. Paleomagnetic
Columnar-jointed
suggest this eruptionwas > 4.03 Ma.
74.2 2.54 ?0.17
51.0 2.56 ?0.13
297.1 ?10.9 100.0
292.2 ?8.0 82.7
R
423
2.95 ? 0.30 (R)
2.97 ? 0.20
ventsites, withproximalpyroclasticand
Multiple
distallavaflowdeposits. One eventestablishedon
3.57 ?0.14
850
A onf
e2
;.U1
t. u.Zu
4uArriAr Dasis U
t4
ana paliomagrlelii anaiysya.
Note: Locationsare describedrelativeto nearbyalpineglaciers(cross referencewithFig.2). Listedtotalgas, plateau,and isotopecorrelationages withassociatedanalytical
uncertainties
(2 standarddeviationunits). Ages calculatedusingdecayconstantsrecommendedby SteigerandJaeger(1978).
tite fission-track analysis also suggests that an
eroded wedge of comparable size has been removed in the last 55-60 Ma (Fitzgerald, 1992).
On the basis of landscape analysis, buttressed
by tectonic relationships and dating, we suggest a
hypothesis of landscape evolution tied to the tectonic history of the Dry Valleys block of the Transantarctic Mountains (Fig. 21). The initial stage is
184
rifting and the creation of a new base level along
the line of the rift. At this time, the flattish interior
of the pre-rifted continent supported surfaceerupted Kirkpatrick Basalt, an equivalent of
which today comprises inselbergs rising above the
upper planation surface at Allan Nunatak. The
second stage is planation that extends inland from
the new coastline and subsequently from the axis
Geografiska Annaler *75 A (1993) . 4
EASTANTARCTICICE SHEETSENSITIVITY
W
E
M Kirkpatrick
20001 basalt
----level
1000
0
(a)
-upper sill
-1000Kukri'peneplain'
-2000
-3000-
3000|
2000(b) 10000-1000'
Sea
originalsurface
upper.. _^
| __Jintermediate"
coastal
0
Obelisk
Newall
stal
intermediate
-coastal
Fleming
upper
2000
(c) 1000
-1000-
Al;
f
Obelisk
Fleming
Newall
(d)2000
1000-
~
-10000
lb
df
2b
0
30
Km
4b
50
60
70
Kukripeneplain
;,,,,,
_-^*s
trasts. The third stage is the dissection of the planation surfaces by the downcutting of valleys to near
sea level. Presumably much of this downcutting accompanied faulting and rock uplift of the mountain front. The fourth stage is submergence of the
lower reaches of the main valleys by the sea, followed by the subsequent uplift of the landscape
surface to present-day elevations.
Our reconstruction supports the view that the
Transantarctic Mountains are associated with
asymmetric rifting and resulted from the separation of the last fragments of Gondwana about 5055 Ma ago (Fitzgerald, 1992). The Ross Sea basin
was thinned by extension and formed the lower
plate, while the TransantarcticMountains formed
the upper plate.
The model of landscape evolution is at an early
stage and is discussed elsewhere (Sugden et al.
1994). Nevertheless, the initial conclusions are important. Denudation removed a wedge of rock
from the coastal zone and the process was accompanied by rock uplift, mainly along the mountain
front. Almost all the denudation was completed by
the middle Miocene. This reconstruction agrees
with structural and tectonic evidence, as well as
with the wider plate tectonic processes that would
be expected following the final breakup of Gondwana.
sill / basalt
Fig. 21. Model of landscapeevolution in the Dry Valleysregion showing the situation (a) at continental break-up, (b)
duringplanationundersemi-aridconditions, (c) after valley
cuttingby rivers,and (d) followingfurtherminorupliftat the
mountain front. Apatite fission-trackanalyses suggest surface uplift of about 5 km since mountain formation
(Gleadow and Fitzgerald,1987). Our analyses(based on differentialuplift of the KukriPeneplain)suggest rockupliftof
4 km near the mountainfront and 2 km inland. Downcutting
and backwearingby rivers and glaciers kept pace with tectonic uplift, and valleys were graded to sea level. Denudation since riftinghas been 4 to 4.5 km at the coast and 1 km
75 km inland.
of major river valleys. Scarp retreat may have
been stimulated by successive changes in base
level to form the upper and then the intermediate
surface. Alternatively, these surfaces may have resulted from the single base-level change, with different escarpments reflecting lithological conGeografiska Annaler ?75 A (1993) - 4
Miocene Ice-Sheet Overriding of the Dry Valleys
Region
Denton et al. (1984) argued that a thick ice sheet
overrode the Dry Valleys sector of the Transantarctic Mountains, flowing northeastward across major pre-existing valleys. Two overriding episodes
were postulated. Denton et al. (1984) further
suggested that the Peleus and Asgard tills in
Wright Valley and the western Asgard Range, respectively, represent early phases of East Antarctic ice-sheet expansion that led to overriding.
Based on the assumption that the Peleus till postdated Prospect Mesa gravels in central WrightValley, Denton et al. (1984) placed the younger overriding episode in the Pliocene.
Here we again argue for ice-sheet overriding
from a suite of glacial erosional features that are
superimposed on the Dry Valleys landscape. Most
of these features are of relatively small scale, and
therefore the resulting imprint on the landscape is
minor. The features include zones of areal scouring that are particularly evident on dolerite of the
upper plateau surface, on rolling topography near
185
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
the coast, and on valley benches. They also include
stoss-and-lee features (with lee-side potholes) on
sandstone cols in the western Asgard Range and
Quartermain Mountains. They include channel
systems on areally scoured terrain and across
mountain divides (Sugden et al. 1991). Finally,
they encompass widespread stripping of surficial
sediments from the landscape that left large areas
of bare bedrock and remnant patches of drift and
colluvium. From our studies in the western Asgard
Range and in the Quartermain Mountains we now
recognize only one episode of massive overriding
(Marchant et al. 1993 a). We concur that Asgard
and Peleus tills represent an outlet glacier that filled Wright Valley during the initial phase of overriding. Ice passing across the western Asgard
Range from the Taylor Valley drainage system
merged with Asgard/Peleus ice. During maximum
overriding, coalesced ice masses submerged the
Dry Valleys. The major adjustment that we have
made in the earlier concept involves the chronology of overriding. On the basis of our new volcanic
ash chronology in the western Asgard Range and
in the Quartermain Mountains, we now believe
that overriding occurred in the middle Miocene,
sometime between 14.8/15.2 Ma and 13.6 Ma
(Marchant et al. 1993 a). This age is based on 40Ar/
39Ar analyses of individual volcanic crystals removed from in-situ ash deposits that occur in stratigraphic association with Asgard till. It is in accord
with our conclusion that Peleus till in east-central
WrightValley is >3.8 Ma (Hall et al. 1993).
Because the relative timing of glacial events and
landscape subsidence is unknown, there remain
several possible explanations for ice-sheet overriding of the Dry Valleys region. One extreme is that
a large and thick ice sheet, similar to that reconstructed by Denton et al. (1984), overrode the
Transantarctic Mountains with surface elevations
similar to present values. This reconstruction
came from the fact that the younger of two postulated overriding episodes was assumed to be Pliocene in age by Denton et al. (1984) from the inferred stratigraphicposition of Peleus till at Prospect
Mesa. An argument was then made that Pliocene
landscape elevations were similar to present values, and the reconstruction was made on this
basis.
The concept of massive Pliocene overriding of
the Dry Valleys region needs revision on several
counts. The first is that our new data show that the
overriding was, in fact, middle Miocene in age
(Marchant et al. 1993 a, b); this renders irrelevant
186
any use of the Pliocene landscape surface as the
basis for reconstructing an overriding ice sheet.
The second is that we now recognize an episode of
considerable landscape subsidence in the evolution of the TransantarcticMountains, which occurred at least as early as late Miocene time and probably considerably earlier. This opens the possibility that ice-sheet overriding may be due, at least in
part, to lowering of the western thresholds, which
would promote greatly increased ice flow into the
Dry Valleys. There is as yet no geologic evidence
that a subsequent rise of these western thresholds
cut off overriding ice. Rather, overriding ice had
dissipated from the Dry Valleys prior to 13.6 Ma
ago (Marchant et al. 1993 a), whereas the documented surface uplift that drained the Wrightfjord
and shallowed the Taylor fjord was Pliocene/Pleistocene in age. This surface uplift began after 3.5
Ma ago in Wright Valley (Hall et al. 1993) and
about 3.4 Ma ago in eastern TaylorValley (Ishman
and Rieck, 1992). However, the geologic record is
fragmentary, and there may well have been an
early phase of surface uplift prior to 3.5 Ma that
has not yet been recognized.
A final point is that increased accumulation, although it may have played a minor part in initiating overriding, probably was not a dominant factor for two reasons. One is that the environment
during expansion was a cold desert as shown by the
paleoclimatic data from the western Asgard
Range and the Quartermain Mountains (Marchant et al. 1993 a, b). Another is that, although
the larger alpine glaciers in eastern Wright Valley
expanded to flow into the Asgard trunk glacier
during the initial phase of overriding, the small
snowbanks and glacierets in the western Asgard
Range did not. Rather, Asgard ice lobes from the
trunk glacier in WrightValley expanded southward
into these valleys (just as today, lobes from Taylor
Glacier project southward into Beacon and Arena
Valleys in the Quartermain Mountains). The implication is that accumulation during the early
phase of overriding was not sufficiently heavy to
cause alpine glaciers to expand out of these basins
and to merge with the trunk glacier. This puts severe limits on the degree to which increased accumulation on the nearby ice sheet caused overriding.
Huybrechts (1993) points out the difficulties
from a glaciological viewpoint for a massive icesheet overriding of the present Transantarctic
Mountains. Such a reconstruction requires mutually contradictory paleo-environmental condiGeografiska Annaler *75 A (1993) 4
EAST ANTARCTIC ICE SHEET SENSITIVITY
tions. Namely, polar conditions are required for
expansion of grounded ice across the Ross Sea
floor, yet warmer-than-present conditions are required for the greatly increased accumulation
needed for interior thickening. Moreover, increased accumulation rates would not cause interior ice-surface elevations to rise significantly,
because of changed ice temperatures and reduced
areal extent. The implication of these modelling
experiments is that overriding is difficult to
achieve with the present topography, unless the climate system operated in a completely differentway (which our paleoclimatic data suggest was not
the case). It follows that any evidence for mountain overriding may, in fact, reflect landscape subsidence and be in itself a clue to the evolution of
the TransantarcticMountains.
Within these constraints, there are at least two
possible explanations for overriding. The first is
simply that landscape subsidence alone was sufficient to submerge the Dry Valleys region by inland
ice. Another explanation for overriding is that a
polar climate allowed ice shelves and then a
grounded ice sheet to form in a shallower-thanpresent Ross Sea. Low portions of theTransantarctic Mountains (such as the subsided Dry Valleys
sector) were submerged by an expanded Antarctic
Ice Sheet that resulted from such grounding. This
ice sheet need not have been nearly as high as that
reconstructed by Denton et al. (1984) in order to
override a Dry Valleys land surface that was considerably lower than now, and thus would not have
required greatly increased interior accumulation.
The dissipation of overriding ice from the Dry
Valleys that led to the present ice-free conditions
occurred in cold-desert environmental conditions
and was accomplished largely by sublimation prior
to 13.6 Ma ago (Marchant et al. 1993 a, b). One
possible reason for such recession is that the overriding ice sheet deepened its bed by stripping sediments in the Ross Sea and by cutting major
troughs through the Transantarctic Mountains.
Once the ice sheet was then perturbed, it retreated
by grounding-line recession from this deeper bed
back to its present configuration. Another possible reason is that there may have been an early
phase of surface uplift prior to 3.5 Ma that has not
yet been recognized in the geologic record. Any
such uplift of the western rim of the Dry Valleys
could have greatly reduced the inflow of East Antarctic ice and initiated the Taylor Dome on the inland flank of the Dry Valleys.
We are not able to resolve these issues with the
Geografiska Annaler 75 A (1993) ?4
data at hand from the Dry Valleys. Answers may
come from investigations in the Royal Society
Range to the south (where areally scoured bedrock from overriding ice extends over a considerable elevation range) and in the Convoy Range to
the north (which is marked by extensive areal
scouring and channel systems from overriding
ice). Meanwhile, we can draw one conclusion that
is central to the present discussion. Namely, icesheet overriding of the Dry Valleys region, no matter what its cause, antedated 13.6 Ma ago and
hence occurred in the Miocene, not in the Pliocene
as required by the deglaciation hypothesis.
Limited Pliocene Glacier Expansion
The extent of Pliocene glacier expansion in the
Dry Valleys is a key to evaluating the deglaciation
hypothesis. An important requirement of this
hypothesis is that ice-sheet expansion across the
Transantarctic Mountains occurred after 3.0 Ma
ago (Barrett et al. 1992), in order to emplace late
Pliocene marine diatoms stripped from interior
marine basins into Sirius Group outcrops, some of
which occur on the high inland erosion surfaces of
the Dry Valleys. We have direct glacial geologic
control, coupled with numerical dates, that places
limits on the Pliocene expansion of Taylor Glacier
(Wilch et al. 1993; Marchant et al. 1993 b; Marchant et al. 1994) and of alpine glaciers on the
south wall of central Wright Valley (Hall et al.
1993). In addition, we have complementary evidence that large tracts of the Dry Valleys remained
free of glacier ice throughout the Pliocene Epoch
(Marchant et al. 1993 a,b,c,d).
Taylor Glacier
Taylor Glacier originates at Taylor Dome and extends for nearly 50 km into Taylor Valley, where it
terminates in Lake Bonney at about 50-60 m elevation. A Pliocene longitudinal profile of Taylor
Glacier (Fig. 22) comes from two key moraine
sequences (Figs 23, page 175, and 25). The first is
on the north valley bench in Taylor Valley below
Mt. J.J. Thomson (Fig. 26) (Wilch etal. 1993). This
bench is referred to as the Rhone platform. The
second is in Arena Valley in the Quartermain
Mountains on the south side of upper Taylor
Glacier (Fig. 23) (Marchant et al. 1993 b).
Figure 26 shows the distribution of glacial drift
sheets and McMurdo Volcanic basanites on the
Rhone platform. The key outcrop that limits Plio187
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
Ice surface profiles for Taylor Glacier
2500
2000
1500
z
1000
500
0
Fig. 22. Numerical
20
reconstructions
40
of the Taylor Glacier
60
Distance(km)
assuming
80
plastic flow and a flow
100
120
GLACIERnson,
1984).
ig. 22. Numericalreconstructionsof the TaylorGlacier assumingplastic flow and a flow law with n=4.2 (Robinson, 1984).
Reconstructionsthat assume plasticflow closely approximatethe presentice-surfaceprofile of TaylorGlacierand show that
a 325 m thickeningat ArenaValley(TaylorIVa, Quaternarymaximum)yields a correspondingice-surfacerise atTaylorDome
of about 160m (Marchantet al. 1994). Similarly,the maximumPliocene-agedrift, representedbyTaylorIVb in ArenaValley
and theThomsonmorainein TaylorValley,yields a correspondingice-surfacerise atTaylorDome of about250 m. Thisimplies
limited Pliocene East AntarcticIce Sheet expansionin this sector of Antarctica.Location of cross sections shown in map insert.
cene ice extent occurs on this bench near the Matterhorn Glacier. Here a complex of pyroclastic and
lava-flow deposits of the McMurdo Volcanic
Group extends from 1250 m to 602 m elevation
(Wilch et al. 1993). Glacial drift(s) on the valley
bench can be traced westward along the north valley wall. The configuration of this drift(s) shows
that it was deposited by an expanded Taylor
Glacier. The Thomson moraine, which marks the
upper margin of this drift sheet, passes across the
surface of the volcanic complex at 1082 m elevation. Based on the whole-rock 40Ar/39Arages of
underlying volcanics, the Thomson moraine has a
maximum age of 2.97?0.14 Ma. It represents the
greatest expansion of Taylor Glacier in the last
3.47 Ma, because numerous 40Ar/39Arages show
that the volcanic complex on which the moraine
rests extends at least this far back into Pliocene
time. A minimum age for the Thomson moraine is
afforded by a 40Ar/39Ar whole-rock date of
2.71?0.14 Ma ago for the portion of Rhone volcanic cone on the valley bench near Rhone Glacier
above 900 m that remains pristine and has not
been overridden by glacier ice. This pristine portion of the Rhone cone is lower than, and up
glacier from, the upper limit of the Thomson
Figs 23 and 24, see pp. 175-176.
188
Geografiska Annaler ?75 A (1993) 4
EASTANTARCTICICE SHEETSENSITIVITY
Fig. 25. Oblique aerial view looking eastwardacross lower Arena Valley and down Taylorand FerrarValleys. Morainesin
lowerArenaValleyrepresentincursionof southwardflowing Plio-PleistoceneTaylorGlacier lobes. The maximumelevation
of Quaternary-ageTaylorIVa moraines in Arena Valley is markedin white; the maximumelevation of Pliocene-ageTaylor
IVb morainesis markedin black. The latter morainesare correlatedwith the Thomson moraineon the north wall of central
TaylorValley.Collectively,the dashed black lines and the Thomson moraine delineate the maximumthickeningof Taylor
GlacierduringPliocene time. US Navy,VXE-6 photograph,TMA 2448, 224.
moraine. Hence, deposition of the Thomson
moraine must have antedated the cone, which
began to erupt at about 2.71 Ma. Therefore, we
conclude that the Thomson moraine and associated drift is between 2.7 and 2.97 Ma in age.
We note, also, that the maximum Quaternary expansion of Taylor Glacier is marked by a line of erratic boulders that crosses the base of the Rhone
Geografiska Annaler ?75 A (1993) ?4
cone at 900 m elevation. This represents the
greatest ice extent since the eruption of the Rhone
cone ceased about 1.50?0.05 Ma ago (Wilch et al.
1993).
Figure 27 (see Map 2, fold out on back cover)
shows a series of drift sheets, most with boulderbelt moraines, in lower Arena Valley in the Quartermain Mountains of the western Dry Valleys re189
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
Fig. 26. Map of surficial deposits on Rhone platform from Wilch et al. (1993b). Light gray areas delineate Thomson drift.
gion. These drift sheets were deposited by a southward-flowing marginal lobe of Taylor Glacier
(Marchant et al. 1994). A drift chronology comes
from surface-exposure ages of boulders from
moraine crests, using in-situ cosmogenic 3He and
"'Be. The results show that the drifts range from
>4.4 Ma old (Quartermain I) to about 100,000 yr
old (Taylor II). Taylor IVb drift has a minimum age
of about 2.1 Ma. Taylor IVa drift predates 1.3 Ma.
Table 2 shows the tentative correlation among
drifts deposited by Taylor Glacier in upper and
lower Taylor Valley. Taylor IVb and Thomson
drifts are correlated on the basis of limiting numerical ages and of relative position within drift sequences. Figure 22 shows the longitudinal profile of
Taylor Glacier when it stood at the Taylor IVb and
Thomson ice limits. This profile represents the
maximum expansion of Taylor Glacier in at least
the last 3.47 Ma (Rhone platform) or 4.4 Ma
(lower Arena Valley). The results show that the
rise of the ice surface during the Taylor IVb/Thomson expansion was <250 m at the Taylor Dome and
about 475 m near Arena Valley; the ice surface
reached to 1082 m elevation at the Thomson
moraine. To place this Pliocene expansion in perspective, we compare it in Figure 22 with the
documented Quaternary advances of Taylor
Glacier. The earliest and greatest such advance
(Taylor IVa in Arena Valley and the Rhone cone erratic line near Mt. J. J. Thomson; Marchant et al.
1993 b; Wilch et al. 1993b) occurred in early
Quaternary time, after 1.5 Ma ago. It is evident
from Figure 22 that the maximum Pliocene expansion of Taylor Glacier was only marginally greater
than the largest Quaternary advance. It is not
clear whether this minor difference represents a
real but small paleoclimatic difference or whether
it is somehow related to the Pliocene/Pleistocene
surface uplift of about 300 m documented for the
Dry Valleys sector of the Transantarctic Mountains.
The Pliocene reconstruction of Taylor Glacier
has a bearing on the interpretation of the Sirius
Group. Sirius Group sediments crop out in the
Dry Valleys region at Table Mountain (Barrett and
Powell, 1982), Mt. Feather (Brady and McKelvey,
1979, 1983), just south of Mt. Fleming, and just
north of Shapeless Mtn. (McKelvey, 1991) (Fig.
18). The deposit at 2650 m elevation on the north
flank of Mt. Feather was one of the original Sirius
outcrops used to establish the deglaciation
hypothesis (Webb et al. 1984; Harwood, 1986), because it contains a diverse diatom flora that includes mid-to-late Pliocene elements. This deposit
See MAP 2, fold-out on back cover.
Fig. 27. Glacial geologic and surficial geomorphic map of Arena Valley. From Marchant et al. (1993b).
190
Geografiska Annaler ?75 A (1993) ?4
EAST ANTARCTIC ICE SHEET SENSITIVITY
Table 2. Suggested Correlations of Pliocene-Pleistocene and Older Drifts in Lower Arena, Central Taylor, and Eastern
Wright Valleys and in the Western Asgard Range
Lower Arena Valley'
Central Taylor Valley2
Eastern Wright Valley3
Taylor II drift
Bonney Drift
Alpine II
Taylor III drift
Taylor III drift
Taylor IVa drift
Taylor IVa drift
Taylor IVb drift
Thomson moraine
Western Asgard Range4
Alpine III
QuartermainI till
Alpine IV
QuartermainII till
Peleus till
Asgard till
'From Denton et al. (1971, 1989) and Marchant et al. (1994, 1993b). 2From Wilch et al. (1993b) and Denton et al.
(1971, 1989). 'From Hall et al. (1993) and Calkin and Bull (1972). 4From Marchant et al. (1993a).
is perched on an old geomorphologic surface at the
top edge of the steep Beacon Valley headwall (Fig.
18) (Brady and McKelvey, 1979, 1983). It consists
of local rocks and rests on striated sandstone bedrock. The implication drawn by Webb et al. (1984)
from the enclosed marine diatoms is that the
Wilkes Subglacial Basin inland of the Dry Valleys
was flooded with marine waters in mid-to-late
Pliocene time. Subsequent overriding of the Transantarctic Mountains by a wet-based ice sheet
would be required to emplace the Sirius outcrop
on Mt. Feather. Barrett et al. (1992) implied that
overriding occurred after 2.8?0.3 Ma ago.
The Pliocene glacial-geologic reconstruction of
Taylor Glacier affords an independent test of
whether mid-to-late Pliocene marine diatoms
could actually have been emplaced in the Mt.
Feather Sirius outcrop at 2650 m elevation by the
East Antarctic Ice Sheet (and hence whether
Wilkes Basin was flooded in mid-to-late Pliocene
time). Mt. Feather is the highest peak of the Quartermain Mountains, which lie between the upper
reaches of Taylor and Ferrar Glaciers. Taylor
Glacier originates at Taylor Dome, which rises 150
m above the East Antarctic ice-sheet surface and
merges with a broad ice divide that extends far inland to Dome Circe (Drewry, 1982). Such an ice
configuration suggests that the interconnected
Taylor and Ferrar Glaciers monitor local and regional East Antarctic ice-sheet fluctuations.
The independent test involves a comparison between the maximum Pliocene-Pleistocene ice-surface elevation of the Taylor and Ferrar Glaciers
and the elevation of the Sirius outcrop on Mt.
Geografiska Annalei, 75 A (1993) ?4
Feather. If the maximum Pliocene-Pleistocene icesurface elevation lies below 2650 m elevation, then
the Sirius Group outcrop could not have been deposited during Pliocene-Pleistocene time by East
Antarctic ice. If, on the other hand, upper Taylor
and Ferrar Glaciers thickened to at least 2650 m,
then this would allow emplacement of the Sirius
Group outcrop on Mt. Feather by East Antarctic
ice that submerged the Quartermain Mountains
during Pliocene/Pleistocene time.
Figures 23 and 25 show the location of the Arena
Valley moraines relative to the Mt. Feather Sirius
outcrop and to the Thomson moraine in central
Taylor Valley. Figure 22 shows the reconstructed
Pliocene longitudinal profiles for Taylor Glacier
based largely on the Thomson moraine and Taylor
IVb drift, from near Mt. J. J. Thomson and in
Arena Valley, respectively. The Thomson/Taylor
IVb longitudinal profile represents the surface of
Taylor Glacier sometime between 2.71 and 2.97
Ma ago, a position not exceeded in at least the last
3.47-4.4 Ma. Because it is physically connected
with Taylor Glacier, Ferrar Glacier almost surely
showed the same behavior.
The key point of Figures 22, 23, and 25 is that
the upper reaches of Taylor Glacier did not thicken
nearly enough in the last 3.47-4.4 Ma to engulf
Mt. Feather. Rather, ice lobes from Taylor Glacier
pushed only into the mouths of Arena (and
Beacon) Valley. This conclusion is consistent with
the implication of surface volcanic ash deposits
that most of Arena and Beacon Valleys remained
free of ice in Pliocene time (Marchant et al.
1993a,b,c,d). For example, in Arena Valley, an in191
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
situ ashfall layer (Arena Valley Ash) dated to 4.34
Ma ago rests on a desert pavement (Marchant et
al. 1993c); in Beacon Valley, a volcanic ashfall deposit dated to 10.4 Ma ago fills a relict thermal contraction crack that marks the same surface as
today (Marchant et al. 1993d,e).
Thus we conclude that the mid-to-late Pliocene
diatom flora in the Mt. Feather Sirius outcrop
could not have been emplaced by an expansion of
the East Antarctic Ice Sheet that engulfed the
Quartermain Mountains.
WrightValleyAlpine Glaciers
The Goodspeed, Hart, Meserve, Bartley, and
Conrow Glaciers head in short theater-shaped valleys in the eastern Asgard Range and flow down
the south wall of east-central Wright Valley. The
areal distribution, stratigraphy,and limiting 40Ar/
39Ar dates of lateral drift sheets delineate the
maximum mid-to-late Pliocene areal extent of
these glaciers.
Figure 28 (see Map 3, fold out on back cover)
shows the distribution of surficial deposits in eastcentral WrightValley. Peleus till, at >3.8 Ma old, is
stratigraphically lowest and represents the last expansion of East Antarctic ice through Wright Valley. Drift units (Alpine I, Alpine II, Alpine III, and
Alpine IV) flanking the alpine glaciers on the
south valley wall are younger than Peleus till and
record subsequent alpine-glacier fluctuations. On
the basis of 4?Ar/39Ardates of reworked basalt
clasts enclosed in these drifts (Hall et al. 1993), Alpine IV drift (the oldest alpine drift) is >3.7 Ma
old, and Alpine III drift is <3.5 Ma old. Advanced
soil development (weathering Stage 5 of Campbell
and Claridge, 1987) of Alpine III drift suggests a
probable Pliocene or early Quaternary age (Hall et
al. 1993). Finally, Wright, Onyx, and Loop drifts,
as well as Alpine I and II drifts, are Quaternary in
age (Hall et al. 1993).
These data point to the striking conclusion that
Wright Valley alpine glacier tongues have expanded by only several hundred meters, to the Alpine III drift limits, in the past 3.5 Ma. Scaled for
glacier size, this result is compatible with that for
Taylor Glacier. Taylor Glacier (100 km in length)
advanced only through lower Taylor Valley and
into lower Arena Valley within the last 3.47 Ma
years; the maximum Pliocene extent barely exceeded the maximum Quaternary advance.
Wright Valley alpine glaciers (4-6 km in length)
advanced only several hundred meters in the past
3.5 Ma years; again the maximum position
achieved in the past 3.5 Ma (Alpine III drift)
barely exceeded the Quaternary maximum extent
(Alpine II drift).
Ice-free TerrainFrom Surficial Ash Deposits
40Ar/39Ardated, in-situ volcanic ashfall deposits
in contraction cracks, avalanche deposits, or on
well-preserved desert pavements show that large
tracts of the Dry Valleys remained free of ice during the Pliocene Epoch (Figs 29 and 30). The
Arena Valley Ash, which rests on a desert pavement at 1410 m elevation in central Arena Valley,
indicates that ice-free conditions existed at this site
4.343?0.108 Ma ago (Fig. 24) (Marchant et al.
1993c). Similarly,the Hart Ash, which overlies colluvium on the south wall of lower WrightValley between the Hart and Goodspeed Glaciers (Hall et
al. 1993), indicates that ice-free conditions existed
at this site at 3.9?0.3 Ma ago. The areal distribution of these ashfall deposits shows that alpine
glaciers in lower Wright and Arena Valleys could
not have been much more extensive at times of
Pliocene ashfall than they are now. In addition, the
preservation of in-situ ashfall deposits on valley
floors and walls indicates that erosive wet-based
glaciers could not have extended across ashfall
sites since at least mid-Pliocene time, otherwise
ash deposits would show evidence of significant reworking or else lie buried beneath Pliocene-age
till, which is not the case (Marchant et al. 1993c).
In particular, the Arena Valley Ash limits Pliocene
expansion of upper Ferrar Glacier during the last
4.34 Ma. The Arena Valley Ash is located about 4
km north of Ferrar Glacier. A low bedrock
threshold (1600 m elevation) at the head of Arena
Valley now prevents Ferrar Glacier from spilling
into upper Arena Valley and covering the Arena
Valley Ash. In places, this bedrock threshold lies
only 30 m above the present ice surface of Ferrar
Glacier (Marchant et al. 1993 b). Ancient tills and
glacial erosional features that occur along this
threshold indicate that wet-based ice from Ferrar
Glacier advanced northward into Arena Valley at
See MAP 3, fold-out on back cover.
Fig. 28. Glacial geologic and surficialgeomorphicmap of east-centralWrightValley.FromHall et al. (1993).
192
Geografiska Annaler *75 A (1993) ?4
EASTANTARCTICICE SHEETSENSITIVITY
Fig. 29. Oblique aerial view looking north across the westernAsgard Range, the Labyrinth,and the OlympusRange. Numbers show locationsof dated surficialvolcanicash sites. 40Ar/39Aranalysesof individualvolcaniccrystalsremovedfromeach
ash deposityield the following:DMS-91-41,13.6Ma; DMS-91-46,12.5Ma; DMS-89-132B, 10.5Ma; DMS-89-211B, 14.5Ma;
DMS-89-143B, 13.5 Ma; DMS-90-38 B, 12.0 Ma; DMS-90-36-B, 10.0 Ma; DMS-90-124B, 14.8 Ma; and DMS-91-22, 15.0
Ma. For details, see Marchantet al. (1993 a). US Navy,VXE-6 photograph.TMA 2303 f-31, 165.
times prior to 7.4 Ma ago (Marchant et al. 1993 b).
However, the preservation of the in-situ Arena Valley Ash suggests that no wet-based ice advanced
across this threshold and into central Arena Valley
during at least the last 4.34 Ma ago. These data are
consistent with limited glacier expansion during
Pliocene time.
Geografiska Annaler ?75 A (1993) ?4
Summary
Our results from Taylor and WrightValleys are consistent in showing relatively restricted Pliocene
glacier expansion, relative to current glacier size.
The large Taylor Glacier expanded into lower Taylor and Arena Valleys whereas the much smaller
Wright Valley alpine glaciers simply advanced
down the valley wall. In both cases these advances
were only slightly more extensive than the maximum Pleistocene advance. Thus a noticeable re193
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
Fig. 30. Oblique aerial view
looking south across the western Asgard Range towards the
QuartermainMountains. 40Ar/
39Aranalyses of individualvolcanic crystalsremovedfrom surficial ash deposits in Arena Valley show that surficial ash deposits in Arena Valley are of
mid-Pliocene to middle Miocene in age. (DMS-87-26, 11.2
Ma; DMS-87-27, 11.5 Ma,
DMS-86-86 B, 4.3 Ma; DMS86-131, 7.4 Ma; DMF-86-141,
6.4 Ma; DMS-86-110C, 8.4 Ma;
DMS-86-113, 11.3 Ma.) For details, see Marchantet al. (1993
b). US Navy, VXE-6 photograph.TMA 2301,f-33, 108.
sult is that there was little discrepancy between
Pliocene and Pleistocene glacier behavior. The implication is that any Pliocene thickening of sectors
of the East Antarctic Ice Sheet adjacent to the Dry
Valleys was minimal. This is consistent with the
volcanic ash data, which shows that wide tracts of
the western Dry Valleys adjacent to the inland ice
sheet remained free of glacier ice throughout the
Pliocene Epoch (Marchant et al. 1993 a, b).
Cold-desert Pliocene Paleoclimate
General Statement
One important requirement of the deglaciation
hypothesis is that Pliocene air temperatures were
sufficiently warmer than at present (20?-25?C
above present-day values if the Beardmore Sirius
deposits were near sea level) to allow growth of
Nothofagus trees or shrubs at 85?S latitude (Barrett, 1991;Webb and Harwood, 1991), and also to
allow surface melting zones to be superimposed on
the East Antarctic Ice Sheet (Huybrechts, 1993).
In Antarctica, an atmospheric warming of 20?C
194
would place the 0?C isotherm high on the Dry Valleys landscape. Since the 0?C isotherm occurs at
approximately the position of the snowline on
mountain glaciers north of the Antarctic Convergence, such warming would introduce melting
ablation zones on Dry Valleys glaciers, which
therefore would have left a temperate glacier landform assemblage, including extensive outwash
plains (with coarse gravel), kame terraces, and
massive moraines (with enclosed lodgement and
meltout till). In addition, one would expect evidence of water activity on steep hillslopes, including debris flows and stream gullies. With this
background, we now examine the landform assemblage and landscape surfaces of the Dry Valleys region for evidence of Pliocene paleoclimate.
Dry Valleys Hillslopes
The rectilinear slopes in the Dry Valleys are largely
relict and had basically achieved their current
form by the end of Miocene time. There is little
evidence of Pliocene development of these rectiGeografiska Annaler ?75 A (1993) ?4
EAST ANTARCTIC ICE SHEET SENSITIVITY
linear slopes beyond the minor weathering that is
occurring today. Thus, in our view, the evolution
Dry Valleys hillslopes shows no evidence of enhanced Pliocene warmth accompanied by increased slope development. Our evidence, which
comes from the high western Asgard Range and
Quartermain Mountains and from eastern Wright
Valley, is reviewed by Marchant et al. (1993 a, b)
and Hall et al. (1993). In summary, Peleus till,
which is >3.8 Ma old, rests on the south rectilinear
wall of central Wright Valley to 1150 m elevation
with little disturbance. Pliocene alpine moraines
also rest on this rectilinear slope with little modification. In the western Asgard Range, in-situ Miocene and Pliocene glacial sediments and colluvium
cover portions of rectilinear slopes. In Arena Valley in the Quartermain Mountains, Pliocene and
Miocene drifts and ash-avalanche deposits rest on
essentially unmodified rectilinear slopes. A comparison of adjacent exposed and colluvium-covered segments of rectilinear slopes suggests that
current weathering processes have generally removed no more than 0.5-3.0 m of exposed bedrock since late Miocene time.
Terrestrial-MarineComparisons
Background. Our purpose here is to compare inferred paleo-marine conditions in the Dry Valleys
with our paleoclimatic interpretations derived
from the terrestrial stratigraphic record. We wish
to determine if there was substantial atmospheric
warming compatible with growth of Nothofagus
600 km to the south at Beardmore Glacier. Or did
cold-desert environments persist, indicating only
modest warming at best? The necessary terrestrial-marine comparisons to address these issues
can be made for crucial Pliocene intervals in
Wright Valley and also in Ferrar and Taylor Valleys.
Pliocene marine water bodies occupied eastern
Ferrar Glacier trough, eastern Taylor Valley, and
Wright Valley. Foraminiferal and diatom biostratigraphy has been carried out for the following
marine cores: CIROS-1 and CIROS-2 (Ferrar
Glacier trough; Harwood, 1989; Webb, 1989),
DVDP-10 and DVDP-11 (eastern Taylor Valley;
Ishman and Rieck, 1992), and DVDP-4A (Wright
Valley; Brady, 1982) (see Fig. 6). In addition,
there is biostratigraphic control for the Prospect
Mesa section that exposes Pliocene marine sediments (Prospect Mesa gravels) in central Wright
Valley (Webb, 1972, 1974; Prentice et al. 1993).
Geografiska Annaler ?75 A (1993) 4
The biostratigraphic chronologies permit valleyto-valley correlation of marine sediments, and
equally important, paleo-water temperatures can
be inferred from the microfossil assemblages in
the cores and the Prospect Mesa gravels.
One example comes from central WrightValley.
Webb (1972, 1974) concluded from the benthic
foraminiferal assemblage in Prospect Mesa gravels
(Pecten gravel) that bottom-water temperatures in
a shallow WrightValley fjord (- 100 m paleo-water
depth) were within the range of -2?C to +5?C, and
perhaps as much as + 10?C. The associated diatom
assemblage, along with the absence of coccolithophores, is taken to indicate water temperature of 0?C to <3?C (Burckle and Pokras, 1991;
Prentice et al. 1993). Prospect Mesa gravels contain the mid-Pliocene foraminifer marker species
Ammoelphidiella antarctica (Trochoelphidiella
onyxi of Webb, 1972, 1974), which in TaylorValley
core DVDP-10 is taken to occur at 3.4-3.8 Ma ago
(Ishman and Rieck, 1992). Burckle et al. (1986)
found late Pliocene diatoms in Prospect Mesa
gravels now taken to suggest an age of 2.5 to 3.0
Ma from diatom biostratigraphy (Prentice et al.
1993). However, the 87Sr/86Srratios of two shells
(C. tuftsensis) assumed to be unaltered from Prospect Mesa gravels suggest a date of 5.5?0.4 Ma
(Prentice et al. 1993), which is considerably older
than the age estimates from biostratigraphic
ranges of the enclosed microfossils.
Another example comes from the CIROS-2
marine core from FerrarGlacier trough (Barrett et
al. 1992). This core reveals the Thalassiosira insignalT. vulnifica and T vulnifica diatom zones. In
Southern Ocean subantarctic cores, T insigna and
T vulnifica occur together at 2.5-3.1 Ma; T vulnifica spans the range 2.2-3.1 Ma. A volcanic ash
in the CIROS-2 core interval containing T.vulnifica yielded single crystal and glass laser-fusion
4?Ar/39Ardates of about 2.8?0.3 Ma in age. This
is taken as the age of the T vulnifica zone in Ferrar
Glacier trough. Most important, however, the age
is also taken to date marine flooding of interior
East Antarctic marine basins, because Tinsigna
and T vulnifica occur together in Sirius Group outcrops in the TransantarcticMountains. From other
diatoms also enclosed in Sirius outcrops, these interior marine water bodies are inferred to have
had surface temperatures of 2?-5? C (Harwood,
1986). The implication of such relatively warm interior water bodies is that similarly warm water
must have occurred in McMurdo Sound and the
Dry Valleys fjords. By this scenario, the Dry Val195
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
leys region would be nearly surrounded by relatively warm marine water. Not only would these
water bodies include McMurdo Sound and its extension into Ferrar,Taylor and WrightValleys, but
also a flooded interior Wilkes Basin on the western
mountain flank, replacing part of the present-day
East Antarctic Ice Sheet. Such a paleogeographic
reconstruction is consistent with the growth of
Nothofagus in the TransantarcticMountains as far
south as 85?S latitude alongside Beardmore
Glacier.
Wright Valley. Our terrestrial-marine comparison in WrightValley depends on the relationship of
the Alpine III and IV moraines, discussed in detail
by Hall et al. (1993), to the Prospect Mesa gravels
of Pliocene age. The stratigraphic position of
Peleus till (which is widespread on the floor of central and east-central Wright Valley) relative to
Prospect Mesa gravels is crucial to the interpretation of WrightValley glacial and marine history. At
Prospect Mesa in central Wright Valley where a
piece of Peleus till rests on Prospect Mesa Gravel,
Denton etal. (1984, 1991) assumed that this overlying Peleus till was in-situ and therefore postdated
Prospect Mesa gravels. By this interpretation, the
marine flooding of Wright Valley (represented by
Prospect Mesa gravels) antedated through-valley
Peleus glaciation. But it is now thought that the
piece of Peleus till that rests on the Prospect Mesa
gravels at Prospect Mesa almost surely slumped
down the north valley wall (Prentice et al. 1993).
The implication is that the widespread Peleus till
on the valley floor is, in fact, older than the subsequent fan deposits below Bull Pass (including
the Prospect Mesa gravels). This reinterpretation
is based on irregular fabrics and platy structure in
the retransported Peleus till, taken together with
the discovery of two new fossil pecten localities on
portions of adjacent fan surface not covered with
Peleus till.
If correct, this new interpretation of the stratigraphy at Prospect Mesa means that marine flooding of WrightValley represented by Prospect Mesa
gravels occurred after Peleus glaciation, not before. It also means that the Prospect Mesa gravels
and the Alpine IV/III drifts occupy the same relative stratigraphic position in Wright Valley; both
are younger than Peleus till and both antedate extensive soil development (weathering stage 6 of
Campbell and Claridge, 1987). The Prospect Mesa
gravels are mid-to-late Pliocene in age from
benthic foraminifers and from diatoms, and 5.4
Ma in age from 87Sr/86Srratios of two C. tuftsensis
196
valves. From 4?Ar/39Ardates of reworked basalt
clasts, Alpine IV drift is >3.7 Ma old, and Alpine
III drift <3.5 Ma old (Hall et al. 1993).
As evident in Figure 28, there is a striking difference in the areal pattern of Alpine III and IV drifts
as compared to Alpine II drift alongside Conrow,
Bartley, Meserve, Hart, and Goodspeed Glaciers
on the south wall of WrightValley. In all cases, Alpine II drift forms complete terminal loops as well
as lateral moraines. In no case except for
Goodspeed Glacier does either Alpine III or Alpine IV drift form terminal loops. In fact, Alpine
III and IV drifts are strikingly absent on the valley
floor and comprise prominent lateral moraine
complexes with lower ends that terminate at 250300 m elevation on the south valley wall. In two
cases (Conrow and Bartley left lateral), flat and irregular tongues of colluviated drift extend
downslope from the lateral moraines to near the
valley floor.
Given the available chronologic and stratigraphic data, our favored explanation for the lack of
terminal loops is that the Alpine III and IV glaciers
calved into a fjord that flooded WrightValley to an
elevation of about 300 m (Hall et al. 1993), and was
drained by the time Alpine II drifts were deposited. The downslope termination of Alpine III
and IV lateral moraines approximately mark the
former shoreline on the south valley wall, where
the alpine glaciers had calving termini. The flat
and irregular drift tongues below these lateral
moraines represent downslope slumps of drift.
The chronologic implication of this reconstruction
is that a marine water body existed in Wright Valley both >3.7 Ma (Alpine IV drift) and <3.5 Ma
(Alpine III drift). This explanation is consistent
with the presence of an early-to-mid-Pliocene
fjord in Wright Valley inferred from the Prospect
Mesa gravels.
If our explanation of the lack of terminal Alpine
III/IV loops is correct, then we can reconstruct the
approximate extent and depths of the associated
marine water body in Wright Valley. If we assume
little tilting of Wright Valley from Pliocene/Pleistocene surface uplift, then marine waters extended westward from McMurdo Sound into
North and South Forks. Near Prospect Mesa the
marine water depth was about 135 m deep, a value
consistent with the 100 m paleo-depth estimated
from foraminiferal fauna (Webb, 1972, 1974). The
marine water body was narrow and shallow in eastern WrightValley where truncated fluvial spurs are
preserved on both valley walls. At the valley
Geografiska Annaler *75 A (1993) ?4
EAST ANTARCTIC ICE SHEET SENSITIVITY
threshold near present-day Lake Brownworth,
marine water depth was only 20-30 m.
The evidence that comes from the physical
character of the alpine drifts suggests that the Alpine III/IV glaciers were cold-based and existed in
a cold-desert environment. There are no temperate glacier features, such as basal till, striated bedrock (even where the bedrock has a protective
sediment cover), kame terraces, bulky moraines
(with enclosed till layers), waterlain till at former
grounding lines, or coarse gravel outwash above
the former marine shoreline. Moreover, there is
remarkably little glaciomarine sediment on the
valley floor. Rather, the complex of alpine glacial
deposits is typical of the present cold-desert regime. The drift sheets are composed of coarse
gravel with numerous ventifacts, almost no
striated stones, and no enclosed lodgement or
meltout till layers. Alluvial fans deposited by the
meltwater streams are similar to modern fans. The
existence of cold-based ice is consistent with the
fact that the alpine glacier tongues have not carved
troughs on the rectilinear valley wall. This overall
argument for cold-desert conditions from former
alpine glaciers also is in accord with little-to-no
slope development in central Wright Valley since
deposition of Peleus till and Hart Ash (Hall et al.
1993). We argue, then, that temperate ice tongues
with surface melting zones did not exist even at sea
level, during deposition of Alpine III and IV drifts
when a fjord occupied WrightValley.
Ferrarand TaylorValleys. From 4?Ar/39Archronologies, we can link the key T vulnifica zone in the
CIROS-2 core in Ferrar fjord (Barrett et al. 1992)
with our terrestrial sequence (Wilch et al. 1993).
The ash layer in the T vulnifica zone is about 3.0
Ma old (2.8 ? 0.3 Ma) (Barrett et al. 1992). From
the 40Ar/39Archronology of the Thomson moraine
at 1082 m elevation near Mt. J. J. Thomson in
lower Taylor Valley (Wilch et al. 1993), along with
the surface-exposure chronology of a drift sequence in lower Arena Valley (Brook et al. 1993), we
have reconstructed the longitudinal profile of Taylor Glacier at 2.71-2.97 Ma ago. The results show
that Taylor Glacier, which drains the peripheral
Taylor Dome of the East Antarctic Ice Sheet, was
then at its maximum extent in the past 3.47 Ma,
rather than in a collapsed state (Marchant et al.
1993b; Wilch et al. 1993b).
What was the paleoenvironment in Taylor Valley at -3.0 Ma ago? We can answer this question
again with data from near Mt. J. J. Thomson and
Geografiska Annaler ?75 A (1993) ?4
in Arena Valley.The upper moraine limit of Thomson drift at 1082 m elevation rests on a 1-1.5 mthick deposit of lapilli with included volcanic
bombs (Fig. 24, p. 176) (Wilch et al. 1993b). The
preservation of delicate spines on all lapilli pieces
shows that the deposit is in-situ. This lapilli
deposit makes up part of a large volcanic complex
of pyroclastic and lava-flow deposits. This lapilli
rests on a well-developed, in-situ desert pavement
formed of an interlocking mosaic of gravel-sized
ventifacts of dolerite and granite. Volcanic bombs
encased within the lapilli deposit yielded an 4?Ar/
39Ar date 2.97?0.14 Ma. We take this to be the
age of burial of the desert pavement, because insitu lapilli rests on the pavement surface and infills cavities between ventifacts (Wilch et al.
1993b).
The existence of this desert pavement, similar in
all respects to modern desert pavements, indicates
that the Dry Valleys were arid at -3.0 Ma ago.
Our analysis of the landforms in Arena Valley
alongside upper Taylor Glacier indicates that the
environment was also polar as well as arid (Marchant et al. 1993 b). The evidence comes from the
physical characteristics of Taylor IVb drift (the
basis of the longitudinal profile of Taylor Glacier
at -3.0 Ma). This drift is composed predominately
of boulder moraines with a matrix of loose gravel
with numerous ventifacts and no striated clasts. In
all respects except surface weathering, these Taylor IVb moraines are identical to moraines of Taylor II and III drifts deposited in Arena Valley during late Quaternary time (Marchant et al. 1994;
1993 b). We thus infer that Taylor IVb moraines
were deposited by a cold-based lobe of the Pliocene Taylor Glacier (Marchant et al. 1993 b). This
inference is supported by the lack of meltwater
channels, fans, or lacustrine sediments associated
with Taylor IVb drift. It is also in accord with the
conclusion of Marchant et al. (1993c) that polar desert conditions have persisted in Arena Valley for
at least the past 4.34 Ma from an in-situ ashfall
layer resting on an desert pavement in Arena Valley. Taken together, these data indicate that the average temperatures during the Pliocene in Arena
Valley were no more than 3?C above present values (Marchant etal. 1993b,c,d).
Surficial Ash Deposits
Under the present hyper-arid cold-desert climate,
the surface of unconsolidated deposits in the Dry
Valleys region shows well-developed ventifact
197
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT, B.L. HALL AND T.I. WILCH
pavements, interlocking contraction cracks (sand
wedges of Marchant et al. 1993d), and coarsegrained gravel lags. Sand wedges form only in cold
continental climates, where mean annual temperatures are below 0?C (Pewe, 1959, 1966; Berg and
Black, 1966; Black, 1976) and ventifact pavements
with associated gravel lags generally form only in
arid climates. The discovery of Miocene- and Pliocene-age volcanic ash in stratigraphic association
with these morphologic forms allows construction
of detailed paleoclimate records. We have identified and dated over 50 different surficial ash deposits in the Dry Valleys region (Marchant et al.
1993a,b,c,d). Most of these ash deposits occur
within relict sand wedges and are of Miocene and
Pliocene age. 40Ar/39Aranalyses of single volcanic
crystals and glass shards removed from ashfall deposits within relict sand wedges in the western Asgard Range indicate cold climate conditions at
15.0 Ma (DMS-91-21), 13.6 Ma (DME-91-41),
13.5 Ma (DMS-89-143), 12.0 Ma (DMS-90-38 B),
10.5 Ma (DMS-89-132 B), and 10.0 (DMS-90-36
B). Numbers in parenthesis refer to ash sites discussed in Marchant et al. (1993 a, b). Likewise,
dated ashfall deposits that rest on desert pavements indicate arid conditions in the western Asgard Range at 14.8/15.2 Ma (DMS-90-124 B) and
in Arena Valley, Quartermain Mountains, at 4.34
Ma (DMS-86-86 B) (Marchant et al. 1993a, b).
The lack of numerous clay-sized grains within
Dry Valleys ash deposits is consistent with persistent cold-desert conditions throughout Pliocene
time. In-situ volcanic ash deposits in the Dry Valleys region contain less than 10% clay-sized grains,
and volcanic crystals lack evidence of chemical
weathering (Marchant et al. 1993a,b,c,d). This is
because where suitable climate conditions cause
glass to be unstable at the ground surface it quickly
alters to clay. The rate at which surficial ash deposits weather to clay minerals depends on atmospheric temperature and the abundance of pore
water (rates are increased at high atmospheric
temperatures and high pore-water pressures;
Lowe and Nelson, 1983; Lowe, 1986). For example, under humid temperate conditions in New
Zealand, which are compatible with growth of
Nothofagus, volcanic ash deposits older than
about 50,000 years are >60% clay (Birrell and Pullar, 1973; Lowe and Nelson, 1983; Lowe, 1986).
The absence of clay-sized grains in Miocene and
Pliocene surficial ash deposits suggests that the
warm climate conditions required for Nothofagus
growth in the TransantarcticMountains could not
198
have existed in the western Dry Valleys region during Pliocene time. In fact, the lack of significant
weathering within over 50 Miocene- and Plioceneage ashes in the Dry Valleys region suggests that
cold-desert conditions have persisted during the
last 15.0 Ma (Marchant et al. 1993a).
QuaternaryAnalogue
A striking result of our reconstructions is the remarkable resemblance between Pliocene and
Quaternary glacier behavior in the Dry Valleys.
This is illustrated by the drift chronologies of Taylor Glacier and WrightValley alpine glaciers. In all
cases, Pliocene drifts occur just outside (or just
above) the outermost Quaternary moraine. These
results, then, do not show a striking dissimilarity
that might be expected if the Pliocene climatic regime were markedly different from the Quaternary regime. Not only were the advances of similar
magnitude, but the drift characteristics were similar.
The youngest Quaternary drifts (Taylor II, Taylor III, Alpine II) all fall within, or close to, the
time span covered by the Vostok ice core in central
East Antarctica, which suggests that the mean annual temperature was only about 2?C warmer in
the penultimate interglaciation than it was during
the Holocene (Lorius et al. 1985). This is consistent with the notion that the Dry Valleys remained
a cold desert in late Quaternary time. The similarities in areal distribution and physical characteristics of Pliocene and late Quaternary drifts and
moraines imply similar cold-desert paleoenvironments in the Pliocene. This is consistent with other
terrestrial indicators (slope stability, ash-covered
desert pavements, ash-filled contraction cracks,
and lack of temperate ice landforms and sedimentary units). Given this situation, the Quaternary
conditions may afford insights into Pliocene paleoclimate and glacier behavior. One recurring
suggestion is that under the late Quaternary
hyper-aridcold-desert conditions, the fluctuations
of Dry Valleys polar glaciers were controlled
primarily by changing accumulation rates (Denton
et al. 1989). The same fundamental control may
have applied to the Pliocene glacier system in the
Dry Valleys if it also existed in a hyper-arid colddesert environment.
Overview
From terrestrial-marine correlations, we conclude
Geografiska Annaler *75 A (1993) ?4
EASTANTARCTICICE SHEETSENSITIVITY
that the geomorphic evidence argues strongly for
cold and dry climatic conditions in the Dry Valleys
adjacent to the marine water bodies with microfossils indicative of warmer-than-present temperatures. The persistence of cold-desert conditions
suggests that paleo-water temperatures of the Pliocene Wright Valley fjord were at the low extreme
of the ranges estimated by Webb (1972, 1974) and
Prentice et al. (1993) from the Prospect Mesa
gravels. There is no evidence from the terrestrial
record at -3.0 Ma for paleoclimate compatible
with an adjacent inland marine water body with
surface temperatures of 2?-5?C. Pliocene expansion of the Dry Valleys glacier system was of limited extent. Reconstructed ice-surface profiles
show that the maximum Pliocene thickening of
Taylor Glacier was only slightly larger than the
modest Quaternary thickening. Finally, the terrestrial record inferred from volcanic ashfall deposits and the physical characteristics of drift
sheets gives no hint of temperatures elevated
enough to produce the melting ablation surfaces
necessary for ice-sheet collapse (Huybrechts,
1993). On the contrary, Taylor Glacier, which
drains Taylor Dome on the periphery of the East
Antarctic Ice Sheet, reached its maximum Pliocene extent in the last 3.47 Ma at close to 3.0 Ma
ago.
Conclusions
We applied a geomorphologic test in the Dry Valleys of whether the East Antarctic Ice Sheet remained robust or melted down during intervals of
excess Pliocene warmth. The geomorphologic results do not meet the predictions of the meltdown
hypothesis. Ice-sheet overriding of the Dry Valleys
region occurred in the middle Miocene, not in the
late Pliocene/early Pleistocene. Pliocene expansion of the Dry Valleys glacier system was modest
and large tracts of the western Dry Valleys remained free of ice; such limited Pliocene expansion means that the East Antarctic ice could not
have emplaced the mid-to-late Pliocene marine
diatoms in the high-elevation Mt. Feather Sirius
outcrop (one of the original sites used to propose
meltdown; Webb et al. 1984). Pliocene surface uplift of the Dry Valleys block of the Transantarctic
Mountains was minimal (about 300 m) rather than
extensive (1000-3000 m). Hyper-arid cold-desert
environmental conditions have persisted since at
least the end of middle Miocene time. Any Pliocene warming was modest and did not approach
Geografiska Annaler ?75 A (1993) *4
the values deemed necessary for meltdown (Huybrechts, 1993). The implication of these geomorphologic data is that the East Antarctic Ice Sheet
remained robust throughout Pliocene-Pleistocene
time. But a distinct weakness of the geomorphologic approach is that a well documented alternative mechanism of emplacing marine Pliocene
diatoms in high-elevation Sirius Group outcrops
has yet to emerge.
The geomorphologic history of the Dry Valleys
is interlocked with the tectonic evolution of the
TransantarcticMountains. Denudation and backwearing under a semi-arid or highly seasonal
humid environment accompanied renewed rifting
and rock uplift that began about 55 Ma ago. The
resulting Transantarctic topography features planation surfaces with erosional remnants that together form an escarpment landscape. These surfaces rise in steps from the coast, with three levels
separated by scarps. The intermediate surface is
terminated by the downfaulted mountain front.
The high-level planation surfaces are dissected by
valley systems graded to near sea level. Much of
this downcutting, which was accomplished by fluvial and glacial erosion, was accompanied by faulting and rock uplift of the mountain front. Following imposition of a hyper-arid, cold-desert climate
by the end of middle-Miocene time, which effectively halted denudation, there occurred subsidence of the landscape, followed by about 300 m
of Pliocene-Pleistocene surface uplift. One
noteworthy conclusion is that all Sirius Group outcrops in the Dry Valleys now occur as remnants on
the oldest elements of the landscape. The geomorphologic implication is that they represent early
phases of glacial activity in the denudation of the
Transantarctic Mountains. Another conclusion is
that ice-sheet overriding of the Dry Valleys region
might have been a consequence of mountain subsidence.
Acknowledgements
The field research in Antarctica was funded and
supported by the Division of Polar Programs of the
United States National Science Foundation. Data
reduction and analysis was supported both by the
National Science Foundation and by the National
Oceanographic and Atmospheric Administration.
D. E. Sugden was partially funded by the Royal
Society and the Royal Society of Edinburgh.
Richard Kelly drafted many of the geologic maps
and figures. We thank the many colleagues who
199
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
have given us invaluable help and advice during
the course of our field work. This manuscript was
improved greatly by comments from L.A.
Burckle, N. Potter, Jr., and J. Splettstoesser. We
are very grateful to M.J. Selby for detailed comments on the geomorphology of the Dry Valleys.
and geophysics, Madison,Wisconsin,USA, August 2227, 1977 (C. Craddock, ed.), 1059-1067. Univ. of Wisconsin Press, Madison.
Barrett, PJ., Adams, C.J., Mclntosh, W.C., Swisher III,
C.C. and Wilson, G.S., 1992: Geochronological evidence supportingAntarctic deglaciation three million
years ago. Nature359: 816-818.
Behrendt,J.C. and Cooper,A.K., 1991:Evidence of rapid
Cenozoic uplift of the shoulder escarpment of the
George H. Denton, Department of Geological SciCenozoic WestAntarcticRift system and a speculation
on possible climate forcing. Geology 19:315-319.
ences and Institutefor Quaternary Studies, UniverBerg, T.E. and Black, R. E, 1966:Preliminarymeasurements
sity of Maine, Orono, Maine 04469, USA.
of growthof nonsortedpolygons,VictoriaLand,Antarctica. In: Antarctic soils and soil forming processes
David E. Sugden, Department of Geography, Uni(J.C.F. Tedrow, Ed.) American Geophysical Union,
Ant. Res. Ser. No. 8, Washington,DC., 61-108.
versity of Edinburgh, Edinburgh, Scotland EH8
Bindschadler,R.A., 1991:West Antarctic Ice Sheet Initia9XP
tive. In: NASA Conference Publication3115,vols. 1 &
2.
David R. Marchant, Institute for Quaternary Birrell,K.S. and Pullar,W.A., 1973:Weatheringof paleosols
in Holocene and late Pleistocene tephras in central
Studies, Universityof Maine, Orono, Maine 04469
North Island,New Zealand. New ZealandJour.of Geol.
USA and Department of Geography, Universityof
and Geophys. 16:687-702.
Edinburgh, Edinburgh, Scotland EH8 9XP
Black, R.E, 1976: Periglacialfeatures indicative of permafrost:ice and soil wedges. Quat. Res. 6, 3-26.
Brenda I. Hall, Department of Geological Sci- BonnefilleR., 1976:Palynologicalevidencefor an important
change in the vegetation of the Omo Basin between 2.5
ences, University of Maine, Orono, Maine 04469,
and2.0 millionyearsago. In: "Earliestman and environUSA.
ments in the Lake Rudolph Basin." (Y Coopens, EC.
Howell, G.L. Isaac and R.E.E Leakey, eds.): 421-431.
Thomas I. Wilch, Institutefor Quaternary Studies,
Universityof ChicagoPress, Chicago.
1983:Evidencefor a cooler and drierclimatein Ethiopia
University of Maine, Orono, Maine 04469 USA
uplandstowards2.5 Myr ago. Nature303, 487-491.
and Department of Geoscience, New Mexico Insti- Brady,
H., 1979: A diatom report on DVDP cores 3, 4A,
tute of Mining and Technology, Socorro, New
12,14,15,and other relatedsurfacesections. In: Proceedings of the Seminar III on the Dry Valley Drilling ProMexico, 87801 USA.
ject, 1978.Memoirsof the NationalInstituteof PolarResearch ( T Nagata, Ed.): 165-175.Toyko, Japan.
1982:Late Cenozoic historyof TaylorandWrightValleys
and McMurdoSound inferredfrom diatoms in Dry ValReferences
ley Drilling Projectcores. In: AntarcticGeoscience ( C.
Abelmann,A., Gersonde,R. and Spiess, V, 1990:PlioceneCraddock, ed.): 1123-1131. Univ. of Wisconsin Press,
Pleistocene paleooceanography in the Weddell SeaMadison.
siliceous microfossilevidence. In: Geological Historyof
Brady, H. and McKelvey,B., 1979:The interpretationof a
the PolarOceans:ArcticVersusAntarctic, (U. Bleil and
Tertiary tillite at Mount Feather, southern Victoria
J. Theide eds.), KluwerAcademic Publishers,729-759.
Land, Antarctica.Jour.of Glac. 22(86): 189-193.
1983:Some aspects of the Cenozoic glaciationof southAhnert, E, 1960:The influence of Pleistocene climatesupon
the morphology of cuesta scarps on the Colorado
ern Victoria Land, Antarctica. Jour. of Glac. 29(102):
Plateau. Assoc. of Amer. Geograph.Ann., 50: 139-56.
343-349.
Brook, E.J., Kurz, M.D., Ackert, R.P Jr, Denton. G.H.,
Augustinus,P C. and Selby,M.J., 1990:Rock slope development in McMurdoOasis, Antarctica, and implications
Brown, E. T, RaisbeckG.M. and Yiou, E 1993:Chronofor interpretationsof glacial history.Geogr.Ann. 72 A:
logy of TaylorGlacier advancesin ArenaValley,Antarc55-62
tica, using in-situcosmogenic3He and '(Be. Quaternary
Res. 39, 11-23.
Barrett,PJ. (ed.) 1989.AntarcticCenozoic historyfrom the
CIROS-1 drillhole, McMurdo Sound. DSIR Bulletin
Bull, C. and Carnein,C.R., 1968:The massbalanceof a cold
245: 1-254.
glacier:MerserveGlacier,SouthernVictoriaLand, Ant1991.Antarcticaand globalclimaticchange:a geological
arctica.Journalof Glac. 13, 415-29.
Burckle, L.H., Prentice, M.L. and Denton, G.H., 1986:
perspective. In: Antarcticaand Global ClimaticChange
(C. Harrisand B. Stonehouse eds.): 35-50.
Neogene Antarctic glacial history: new evidence from
marine diatoms in continental deposits. EOS Tran.67:
Barrett,P J. and Elliot, D.H., 1973:Reconnaissancegeologic
295. AmericanGeophysicalUnion, Washington,D.C.
map of the Buckley Island quadrangle, Transantarctic
Mountains, Antarctica, Antarctic geologic map A-3.
Burckle, L.H. and Pokras, E.M., 1991: Implicationsof a
Pliocene stand of Nothofagus (southern beech) within
U.S.G.S., Washington,D.C.
500 kilometersof the South Pole. AntarcticScience3(4):
Barrett,PJ. and Powell, R.D., 1982: Middle Cenozoic glacial beds atTableMountain,SouthernVictoriaLand. In:
389-403.
Antarctic Geoscience-symposiumon Antarcticgeology
Calkin, PE., 1974:Subglacialgeomorphologysurrounding
200
Geografiska Annaler *75 A (1993) *4
EASTANTARCTICICE SHEETSENSITIVITY
the ice-free valleys of SouthernVictoria Land, Antarctica. Jour. of Glac. 13(69):415-429.
Campbell, I.B. and Claridge, G.G.C., 1987: Antarctica:
soils, weatheringprocesses, and environment.Developmentsin Soil Science 16. Elsevier, New York,368 p.
Carlquist, S., 1987: Upper Pliocene-lower Pleistocene
Nothofagus wood from the TransantarcticMountains.
Aliso 11:571-583.
Chinn, TJ., 1980:Glacier balances in the Dry Valleysarea,
Victoria Land, Antarctica. In: Proceedings of the
Riederalp WorkshopIAHS-AISH Publicationno. 125,
237-247.
Ciesielski,PE and Weaver,EM., 1974:EarlyPliocene temperature changes in the Antarctic Seas. Geology 2: 51115.
Claridge, G.G.C. and Campbell, I.B., 1968: Soils of the
ShackletonGlacierregion, Queen MaudRange,Antarctica. N.Z. Jour.Sci. 11, 11-15.
Cooke, R.U., Warren,A. and Goudie, A.S., 1993: Desert
Geomorphology. University College London Press.
London.
Cooper,A.K. and Davey, EJ., 1987a:Seismic Stratigraphy
and Structureof the VictoriaLand Basin, WesternRoss
Sea, Antarctica. In: The AntarcticContinentalMargin:
Geology and Geophysics of the Western Ross Sea,
CPCEMR Earth Science Series 5B, 27-65: Houston,
Texas, Circum-PacificCouncil for Energy and Mineral
Resources.
Cooper,A.K. and Davey, EJ. (eds.) 1987b:The Antarctic
Continental Margin: Geology and Geophysics of the
Western Ross Sea. Earth Science Series 5B, 93-118.
Houston, Texas; CircumpacificCouncil for Energy and
MineralResources.
Cooper,A.K., Davey, EJ. and Behrendt,J.C., 1991:Structural and depositional controls on Cenozoic and
(?)Mesozoic strata beneath the western Ross Sea. In:
Geological Evolution of Antarctica(M.R.A. Thomson,
J.A. Crame, and J.W.Thomson, eds.), CambridgeUniversity Press, Cambridge:279-283.
Cronin, T.M. and Dowsett, H.J. (eds.) 1991: Pliocene Climates. Quat. Sci. Rev. 10(2/3).
Denton, G.H. and Hughes, TJ., 1981: The Last Great Ice
Sheets.Wiley-Interscience,New York.
Denton, G.H., Prentice,M.L., Kellogg, D.E. and Kellogg,
TB., 1984: Late Tertiaryhistory of the Antarctic Ice
Sheet: Evidence fromthe DryValleys. Geology 12:263267.
Denton, G.H., Bockheim, J.G., Wilson, S.C. and Stuiver,
M., 1989:Late Wisconsinand early Holocene glacialhistory,Inner Ross Embayment,Antarctica.Quat. Res. 31:
151-182.
Denton, G.H., Prentice, M.L. and Burckle, L.H., 1991:
Cainozoic history of the Antarctic Ice Sheet. In: The
Geology of Antarctica( R.J. Tingey,ed.): 365-433, Oxford UniversityPress.
Doake, C.S.M., 1985:Antarcticmass balance:glaciological
evidence fromAntarcticPeninsulaand WeddellSea sector. In: Glaciers, ice sheets, and sea level: effect of a
CO2-induced climatic change: 317-30, United States
Departmentof Energy,Washington,D.C.
Doehring, D. O. (ed.) 1977:Geomorphologyin arid regions.
Boston: Allen & Unwin.
Doumani, G.A. and Minshew,V.H., 1965:General geology
of the Mt. Weaverarea, Queen Maud Mountains,Antarctica. In: Geology and paleontology of the Antarctic,
Geografiska Annaler *75 A (1993) ?4
Ant. Res. Ser., v. 6 (J.B. Hadley ed.), 127-139. American GeophysicalUnion, Washington,D.C.
Dowsett, H.J. and Cronin, TM., 1990: High eustatic sea
level during the middle Pliocene: Evidence from the
southeastern U.S. Atlantic Coastal Plain. Geology 18:
435-38.
Dowsett, H.J., Cronin,T.M., Poore, R.Z., Thompson,R.S.,
Whately,R.C. and Wood, A.M., 1992: Micropaleontological evidence for increased meridional heat transport in the North Atlantic Ocean during the Pliocene.
Science258: 1133-35.
Drewry, D.J., 1982: Ice flow, bedrock, and geothermal
studies from radio-echo sounding inland of McMurdo
Sound, Antarctica. In: Antarctic Geoscience (C. Craddock, ed.): 977-983. University of Wisconsin Press,
Madison.
- 1983:Antarctica:Glaciological and Geophysical Folio.
Scott Polar ResearchInstitute, Cambridge.
Drewry,D.J., Jordan, S.R. and Jankowski, E., 1982:Measuredpropertiesof the AntarcticIce Sheet: surfaceconfiguration, ice thickness, volume, and bedrock characteristics.Annals of Glaciology3: 83-91.
Elliot, D., Barrett,PJ. and Mayewski, PA., 1974:Reconnaissance geologic map of the Plunket Point Quadrangle,TransantarcticMountains,Antarctica. U.S.G.S.
AntarcticMap No. 4.
Elliot, D.H., Bromwich, D.H., Tzeng Ren-Yow,Harwood,
D.M. and Webb,PN., 1991:The SiriusGroup: a possible test of climatemodelingresultsforAntarctica.In: International Conference on the Role of the Southern
Ocean andAntarcticain GlobalChange:An Ocean Drilling Perspective,abs.
Faure,G. and Taylor,K.S., 1981:Provenanceof some glacial
Mountainsbased on Rb-Sr
depositsin the Transantarctic
datingof feldspars. ChemicalGeology 32(3/4), 271-90.
Fitzgerald, P G., 1992: The TransantarcticMountains of
southernVictoriaLand:the applicationof apatitefission
trackanalysisto a rift shoulderuplift. Tectonics11,634662.
- 1986:Asymmetricextension associated with uplift and
subsidence in the TransantarcticMountains and Ross
Embayment.Earthand Plan. Sci. Let. 81: 67-78.
Funder, S., Abrahamsen, N., Bennike, 0. and FeylingHanssen, R.W., 1985: Forested Artic: Evidence from
North Greenland. Geology, 13:542-46.
Gerson, R., 1982:Talus relicts in deserts: a key to major
climatic fluctuations. IsraelJour. of EarthSci. 31, 123132.
Giovinetto, M.B. and Bentley,C.R., 1985:Surface balance
in ice drainage systems of Antarctica.Ant. Jour. of the
U.S. 20, 6-13.
Gleadow,A.J.W. and Fitzgerald,P.G., 1987: Uplift history
of the TransantarcticMountains:new evidence from fission trackdatingof basementapatitesin the Dry Valleys
area, southernVictoriaLand. Earthand Plan. Sci. Let.
82: 1-14.
Grube, E, Christensen,S. and Vollmer,T, 1986:Glaciations
in NorthwestGermany.Quat. Sci. Rev. 5: 347-358.
Gunn, B.M. and Warren,G., 1962:Geology of VictoriaLand
between the Mawsonand MulockGlaciers,Antarctica.
New Zeal. Geol. Surv.Bull. 71.
Hall, B.L., Denton, G.H., Lux, D.R. and Bockheim, J.G.
1993:Late Tertiaryantarcticpaleoclimateand ice-sheet
dynamicsinferredfrom surficialdeposits in WrightValley. Geogr.Ann. 75 A, 239-267.
201
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
Haq, B. U., Hardenbol,J. and Vail.PR., 1988:Mesozoic and
Cenozoic chronostratigraphy
and eustaticcycles. In: Sea
level changes:An integratedapproach(Wilgus, C.S. et
al. eds.). Tulsa Oklahoma, Society of Economic Paleontologists and Mineralogists:71-108.
Harwood, D.M., 1983:Diatoms from the SiriusFormation,
TransantarcticMountains.Ant. Jour. of the U.S. 18(5):
98-100.
- 1986: Diatom biostratigraphyand paleoecology and a
Cenozoic history of Antarcticice sheets . PhD dissertation, Ohio State University,ColumbusOhio.
- 1989:Siliceous microfossils.In: AntarcticCenozoic history from the CIROS-1 drillhole, McMurdo Sound.
DSIR Bull. 245: 67-97.
Hays, J.D. and Opdyke, N.D., 1967:Antarctic radiolaria,
magnetic reversal and climate change. Science 158,
1001-11.
Hill, R.S., Harwood, D.M. and Webb,P.N., 1991:Last remnant of Antarctica's Cenozoic flora: Pliocene
Nothofagus of the Sirius Group, TransantarcticMountains. In: Eighth Gondwana Subcommission Symposium, Hobart, Australia.(Abstract).
Holdsworth,G. and Bull, C., 1970:The flow law of cold ice:
investigationsof Meserve Glacier,Antarctica.In: International Symposiumon AntarcticGlaciologic Exploration (Hanover, New Hampshire,USA, 3-7 September
1968):204-216: IAHS Publ. no. 86.
Hooghiemstra,H., 1993: Pliocene-QuaternaryPaleoclimatic Changeand Evolution of NorthernAndean Montane
Forests and Paramo Ecosystems (ext. ab.). In: Conference on Palaeoclimateand Evolution with Emphasison
Human Origins, Airlie Conference Center, Virginia
(May 17- May 21, 1993).
Howard, A.D. and Kochel, R.C., 1988: Introduction to
cuesta landformsand sappingprocesseson the Colorado
Plateau. In: Sappingfeaturesof the ColoradoPlateau:a
comparativeplanetarygeology field guide (A.D. Howard, C. Kocheland H.E. Holt, eds.), NASA specialpublication SP-49.
Huybrechts, P 1993: Glaciological and climatological aspects of a stabilist versus a dynamic view of the late
Cenozoic glacialhistoryof East Antarctica.Geogr.Ann.
75 A, 221-238.
Hughes, T.J., 1977:West Antarctic ice streams. Reviewsof
Geophysicsand Space Physics 15: 1-46.
Ishman, S.E. and Rieck, H.J., 1992:A Late Neogene AntarcticGlacio-eustaticRecord, VictoriaLand Basin Margin, Antarctica.In: The AntarcticPaleoenvironment:A
perspective on Global Change, Part One (J.P. Kennett
and D.A. Warnke,eds.), American GeophysicalUnion,
Ant. Res. Ser. 56, WashingtonD.C., p. 327-347.
Kennett, J.P., 1982: Marine Geology. Prentice-Hall, Englewood Cliffs, New Jersey.
Keys, J.R., 1980:Air temperature,wind, precipitationand
atmospherichumidityin the McMurdoregion.Antarctic
Data Seriesno. 9, VictoriaUniversityof Wellington,No.
17.
Krantz,D.E., 1991:A chronologyof Pliocene sea-level fluctuations:The U.S. MiddleAtlanticcoastal plain record.
QuaternaryScienceReviews10: 163-174.
Kyle, R., 1990: McMurdoVolcanic Group-western Ross
Embayment. In: Volcanoes of the Antarctic plate and
southernoceans. (WE. LeMasurierand J.W.Thomson,
eds.) American Geophysical Union, Ant. Res. Ser. 48,
Washington,D.C., p. 19-25.
Laity,J.E., 1988:The Role of GroundwaterSappingin Val202
ley Evolution on the ColoradoPlateau. In: Sappingfeatures of the Colorado Plateau: a comparativeplanetary
geology field guide (A.D. Howard,C. Kochel and H.E.
Holt, eds.), NASA special publicationSP-49:63-70.
Laity, J.E. and Malin, M.C., 1985: Sappingprocesses and
the development of theater-headedvalley networks on
the ColoradoPlateau. Geol. Soc. ofAmer. Bull. 96: 203217.
Laprade, K.E., 1984:Climate, geomorphology,and glaciology of the Shackleton Glacier Area, Queen Maud
Mountains, Antarctica. In: Ant. Res. Ser. 36 ( M.D.
Turnerand J.F. Splettstoesser), 163-96. AmericanGeophysicalUnion, Washington,D.C.
LeMasurier,WE., 1990: Late Cenozoic Volcanismon the
AntarcticPlate: an overview.In: Volcanoes of the AntarcticPlate and SouthernOceans (WE. LeMasurierand
J.W. Thomson, eds.). American Geophysical Union,
Washington,Ant. Res. Ser. 48, WashingtonD.C., p. 118.
LeMasurier,WE. and Rex, D.C., 1991:The Marie Byrd
Land volcanic provinceand its relationto the Cainozoic
West Antarcticrift system. In: The Geology of Antarctica ( R.J. Tingey, ed.): 249-84. Oxford University
Press.
Lorius, C., Jouzel, J., Ritz, C., Merlivat,L., Barkov,N.I.,
Korotkevich,YS. and Kotlyakov,VM., 1985:A 150,000
year climaticrecordfromAntarcticice. Nature316, 591596.
Lowe, D.J., 1986: Controlson the rates of weatheringand
clay mineralgenesis in airfalltephras:a review and New
Zealand case study.In: Rates of chemicalweatheringof
rocks and minerals (S.M. Colman and D.P. Dethier
eds.): 265-330. Academic Press, Inc., New York.
Lowe, D.J. and Nelson, C.S., 1983:Occas. Rep. 11. Department of Earth Sciences, University of Waikato,Hamilton, New Zealand.
Marchant,D. R., Denton, G.H., Sugden,D.E. and Swisher
III, C.C. 1993a:Miocene glacial stratigraphyand landscape evolution of the western Asgard Range, Antarctica, Geogr.Ann. 75 A, 303-330.
Marchant, D.R., Denton, G.H. and Swisher III, C.C.,
1993b: Miocene-Pliocene-Pleistoceneglacial history of
Arena Valley, Quartermain Mountains, Antarctica,
Geogr.Ann. 75 A, 269-302.
Marchant, D.R., Swisher III, C.C., Lux, D.R., West,Jr.,
D.P and Denton, G.H., 1993c: Pliocene Paleoclimate
and EastAntarcticice-sheet historyfromsurficialash deposits. Science260: 667-670.
Marchant,D.R., SwisherIII, C.C., PotterJr., N.P. and Denton, G.H., 1993d:Antarcticpaleoclimate and ice-sheet
dynamicsreconstructedfrom volcanic ashes in the Dry
Valleysregion. Palaeo. Palaeo. Palaeo. submitted.
Marchant,D.R., Denton, G.H., Bockheim, J.G., Wilson,
S. C. and Kerr,A.R., 1994:Quaternaryice-level changes
of upper Taylor Glacier, Antarctica: Implications for
paleoclimateand ice-sheet dynamics.Boreas, in press.
Matthews,J. V, 1989:Late TertiaryArctic Environments:a
Vision of the Future. GEOS, 18(3): 14-18.
Mayewski, PA., 1975: Glacial geology and late Cenozoic
History of the Transantarctic
Mountains,Antarctica,Institute of Polar Studies report, No. 56. Ohio State University,Columbus.
McGregor,VR., 1965:Notes on the geology of the area between the heads of the Beardmore and Shackleton
Glaciers,Antarctica.N.Z. Jour. Geol. Geophys.8,27891.
Geografiska Annaler *75 A (1993) ?4
EASTANTARCTICICE SHEETSENSITIVITY
McKelvey,B.C., 1991:The Cainozoicglacialrecordin south
VictoriaLand: a geological evaluationof the McMurdo
Sound drilling projects. In: The Geology of Antarctica
(R.J. Tingey,ed.): 434-449, Oxford UniversityPress.
McKelvey,B.C., Webb, P.N., Harwood, D.M. and Mabin,
M.C.G., 1987: The Dominion Range Sirius Group: a
late Pliocene-early Pleistocene record of the ancestral
Beardmore Glacier. In: Fifth InternationalSymposium
on AntarcticEarthSciences. Cambridge.(Abstract97).
McKelvey,B.C., Webb,P.N., Harwood, D.M. and Mabin,
M.C.G., 1991:The Dominion Range SiriusGroup:a record of the late Pliocene-early Pleistocene Beardmore
Glacier. In: Geological Evolution of Antarctica
(M.R.A. Thomson, J.A. Crame and J.W. Thomson,
eds.): 675-682. Cambridge University Press, Cambridge.
Mercer,J.H. 1968: Glacial geology of the Reedy Glacier
area, Antarctica. Geol. Soc. of Amer. Bull. 79: 471-86.
- 1972: Some observations on the glacial geology of the
BeardmoreGlacierarea. In:Antarcticgeology and Geophysics-symposiumon Antarcticgeology and solid earth
geophysics, Oslo, 6-15 August 1970 (R.J. Adie ed.),
427-33. Universitetsforlaget,Oslo.
- 1986: Southernmost Chile: A modern analog of the
southernshoresof the Ross EmbaymentduringPliocene
warm intervals. AntarcticJournal of the United States
21(5): 103-5.
Mildenhall,D. C., 1989:Terrestrialpalynology.In: Antarctic
Cenozoic historyfromthe CIROS-1drillhole,McMurdo
Sound (P.J. Barretted.), DSIR Bulletin 245: 119-27.
Miller, K.G., Fairbanks,R.G. and Mountain, G.S., 1987:
Tertiaryoxygen isotope synthesis, sea level history,and
continental marginerosion. Paleoceanography2 (1), 119.
Moon, B.P and Selby, M.J., 1983:Rock mass strengthand
scarp forms in southernAfrica. Geogr.Ann. 65A: 135145.
Mudrey,M.G. and McGinnis, L.D. (eds.) 1975:Dry Valley
Drilling Project, Bulletin No. 5. Northern Illinois University.
Oberlander,T.M., 1977: Origin of segmented cliffs in massive sandstonesof southeasternUtah. In: Geomorphology in Arid Regions (D.O. Doehring, ed.): 79-114.
- 1989:Slope and pediment systems. In: Arid Zone Geomorphology,Belhaven Press, London: 56-86.
Oerlemans,J. and van der Veen,C.J. 1984:Ice sheetsand climate. Reidel, Dordrecht.
Osborn, G., 1985:Evolution of the Late Cenozoic Inselberg
Landscape of Southwestern Jordan. Palaeo., Palaeo.,
Palaeo. 49: 1-23.
Pewe, T.L., 1959:Sand-wedgepolygons (tesselations) in the
McMurdoSound region, Antarctica-a progressreport.
Amer.Jour. Sci. 257, No. 8, 545-552.
- 1966: Paleoclimatic significance of fossil ice wedges.
BiuletynPeryglacjalny15, 65-73.
Pickard,J., Adamson, D.A., Harwood, D.M., Miller,G.H.,
Quilty,P G. and Dell, R.K., 1988:EarlyPliocene marine
sediments, coastline and climate of East Antarctica.
Geology 16: 158-161.
Prentice, M.L., Denton, G.H., Lowell, TV, Conway, H.
and Heusser,I.E., 1986:Pre-late Quaternaryglaciation
of the BeardmoreGlacierregion, Antarctica.Ant. Jour.
U.S. 21(5), 95-98.
Prentice,M.L. and Denton, G.H., 1988:The deep-sea oxygen isotope record, the global ice sheet system and
hominid evolution. In: Evolutionary History of the
Geografiska Annaler ?75 A (1993) - 4
Robust Australopithecines.(E Grine, ed.), Aldine de
Gruyter,New York:383-403.
Prentice, M.L., Bockheim, J.C., Wilson, S.C., Burckle,
L.H., Hodell, D.A., Schluchter,C. and Kellogg, D.E.,
1993:Late Neogene AntarcticGlacialHistory:Evidence
from Central Wright Valley. American Geophysical
Union, Ant. Res. Ser. In press.
Robin, G. de Q., 1977: Ice cores and climatic change.
Philosophical Transactionsof the Royal Society of London B280: 143-68.
- 1988:The Antarcticice sheet, its historyand responseto
sea level and climaticchanges over the past 100 million
years. Palaeo., Palaeo., Palaeo. 67: 31-50.
Robinson, PL., 1984: Ice dynamicsand thermal regime of
TaylorGlacier,SouthVictoriaLand,Antarctica.Journal
of Glac. 30: 133-160.
Ruddiman,W.FEand Raymo, M.E., 1988: Northernhemisphere climateregimesduringthe last 3 myr:Possibletectonic connections. Phil. Trans.R. Soc. ***
Savin, S.M., Douglas, R.G. and Stehli, EG., 1975:Tertiary
marine paleotemperatures.Geol. Soc. Amer. Bull. 86:
1499-1510.
Schmidt, K.H., 1989:The significanceof scarp retreat for
Cenozoic landformevolution on the Colorado Plateau,
USA. EarthSurf. Proc. Land 14:93-105.
Schwerdtfeger,W, 1984:Weatherand climate of the Antarctic. In: Developments in Atmospheric Science 15.
Elsevier PublishingCompany,Amsterdam.
Selby, M.J., 1971: Slopes and their development in an icefree, arid area of Antarctica. Geogr.Ann. 53(a): 235245.
- 1974:Slope evolution in an Antarcticoasis. New Zeal.
Geogr. 30: 18-34.
Selby, M.J and Wilson, A. T, 1971: The origin of the
Labyrinth,WrightValley,Antarctica. Geol. Soc. Amer.
Bull. 8: 471-476.
Shackleton,N.J., 1993:New Data on the Evolution of Pliocene Climatic Variability (abs.). In: Conference on
Palaeoclimateand Evolution with Emphasison Human
Origins, Airlie Conference Center, Virginia (May 17May 21, 1993).
Shackleton,N.J. and Kennett,J. P, 1975:Paleotemperature
history of the Cainozoic and the initiation of Antarctic
glaciation: oxygen and carbon analysis in DSDP sites
277, 279 and 281. In: Initial Reports of the Deep Sea
Drilling Project (Kennett, J.P. and Houtz, R., eds.) 29:
743-755.
Stanley,S.M., 1985: Climatic cooling and Plio-Pleistocene
mass extinction of molluscs around the marginsof the
Atlantic. SouthAfricanJournalof Science81(5): 266.
Stuiver,M., Denton, G.H., Hughes, T.J. and Fastook,J.L.,
1981:History of the marineice sheet in WestAntarctica
duringthe last glaciation:A workinghypothesis.In:The
Last Great Ice Sheets (G.H. Denton and T.J. Hughes,
eds.): 319-346. Wiley-Interscience,New York.
Sugden, D.E., Denton, G.H. and Marchant, D.R., 1991:
Subglacialmeltwaterchannelsystemsand ice sheet overriding, Asgard Range, Antarctica. Geogr. Ann. 73A:
109-121.
- 1994:Landscapeevolution of the Dry Valleys,TransantarcticMountains.Jour. of Geophy.Res. , submitted.
Tessensohn,F and Worner,G., 1991:The Ross Sea rift system, Antarctica:Structure,evolution and analogues.In:
Geological evolution of Antarctica(M.R.A. Thomson,
J.A. Crame and J.W.Thomson, eds.), CambridgeUniversity Press, Cambridge:273-277.
203
G.H. DENTON, D.E. SUGDEN, D.R. MARCHANT,B.L. HALL AND T.I. WILCH
van der Hammen, T., Wijmstra,T.A. and Zagwijn, WH.,
1971:The floral recordof the late Cenozoic of Europe.
In: Late Cenozoicglacialages (K.K. Turekian,ed.), Yale
UniversityPress:391-424.
Vrba,E.S., 1988a:The environmentalcontext of the evolution of earlyhominidsand theirculture.In: Bone modification ( R. Bonnichsenand M.H. Sorg, eds.), Centerfor
the Study of Early Man :27-42.
- 1988b:Late Pliocene climaticeventsand hominidevolution. In: Evolutionary history of the Robust Australopithecines(E Grine, ed.), Aldine de Gruyter,NY:
383-403.
Webb,PN., 1972:Pliocene marineinvasionof an Antarctic
dry valley.Ant. Jour. of the U.S. 7: 227-224.
- 1974:Micropaleontology,paleoecology, and correlation
of the Pecten gravels,WrightValley,Antarctica.Jour.of
Foram.Res. 4: 185-89.
- 1989:Benthic foraminifera.In: AntarcticCenozoic history from the CIROS-1 drillhole, McMurdo Sound.
DSIR Bull. 245: 99-118.
Webb,P.N., Harwood, D.M., McKelvey,B.C., Mercer,J.H.
and Stott, L.D., 1984: Cenozoic marine sedimentation
and ice-volume variation on the East Antarcticcraton.
Geology 12:287-291.
Webb,P.N. andAndreasen,J.E., 1986:Potassium/argondating of volcanicmaterialassociatedwith the Pliocene Pecten Conglomerate (Cockburn Island) and Scallop Hill
Formation(McMurdoSound), Ant. Jour.of the U.S. 21:
59.
Webb,P.N. and Harwood, D.M., 1987:Terrestrialflora of
the Sirius Formation:Its significancefor late Cenozoic
glacial history.Ant. Jour. of the U.S. 22(4): 7-11.
204
Webb, P.N., McKelvey, B.C., Harwood, D.M., Mabin,
M.C.G. and Mercer,J.H., 1987:SiriusFormationof the
BeardmoreGlacier region. Ant. Jour.of the U.S. 22: 813.
Webb,P.N. and Harwood, D.M., 1991:Late Cenozoic glacial history of the Ross Embayment,Antarctica. Quat.
Sci. Revs. 10:215-23.
Wesselman,H. B., 1985:Fossil micromammalsas indicators
of climaticchange about 2.4 Myrago in the Omo Valley,
Ethiopia. SouthAfr.Jour.of Sci. 81: 260-261.
Wilch,T.I., Lux, D.R., Denton, G.H., and McIntosh,W.C.,
1993a: Minimal Pliocene-Pleistocene uplift of the dry
Mountains:A key parvalley sector of the Transantarctic
ameterin ice-sheet reconstructions.Geology21(9): 84144.
Wilch, T.I., Denton, G.H., Lux, D.R. and Mcintosh, W.C.
1993bLimitedPliocene glacierextent and surfaceuplift
in middle TaylorValley,Antarctica. Geogr.Ann. 75 A,
331-351
Wilson,A. T, 1973:The great antiquityof some Antarctic
landforms-Evidence for an Eocene temperate glaciation in the McMurdo region. In: Palaeoecology of
Africa, and the surroundingislandsof Antarctica(E.M.
van Zinderen Bakker Sr., ed.), VIII, Balkema, Cape
Town.
Young,R. W., 1983:The tempo of geomorphologicalchange:
Evidence fromsoutheasternAustralia.Jour.of Geol. 91,
221-230.
- 1987:Sandstonelandformsof the tropicaleast Kimberly
region, northwesternAustralia.Jour. of Geol. 95: 205218.
Geografiska Annaler *75 A (1993) ?4
GL A
C IE FI
Harbour
~~~~~~~~~~~
1810
m
,A,~~~~,
(n~~~~~~~~~~~~~~~~~~~'
%-,6
Pea
Mt.
P,
,.
Skew,,
".170,,'..
Dearborn.,,~,
~~~~~~~~06%
~,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1
:~_.~~Oft,o42
>
z~~~~~~~~~~~~~~
'
.'~.'."
~~~~~~~~~~t
J
le
a
0~~~~~~~~~~
ei
jl~g.,~V
.
Mt...........
.....
...
"t':
!i...
0~~~~~~~~~~~~~~~~~~~~~~~~:iiiii::i
.
..._..
.~~~~~~~~~~~~~~~~~~~~~~t
MtFlemin
w
L.Vanda
....
~,~~~~~~~~~~~~A
v~~~~~~~~t\
"~':'~::!'~~~~~~~~~~
'.tD
"
::~~~~~~~~
?,
-~~~~
0~~~
bi
-
AN...
2
.
m~~~~~~~~~~~~~:::':.E
as
atterh ~~~~~.
~'~<~~~~~~~~~~~~~~~~~~~~~~0
onney
V
~~~~~~~~~~~~~~~~~~~~~Y
(n 2700'_~..."..
24b
On~~~~~~~~~~~~~~~~~~-:~
0.1~~~~~~~~~~~~~~~~~~~~~
x
I
Im
I
.i^^^y^"
,
m~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,,
------
N
A
Harbour
rt
CIER
0
-Z
A.
-2o1
0
m
34816
Majormeltwaterchannel
Summit inselbergs
....
................
p.,
*. , .
..........
bounded
slopes
Upper surface
Intermediatesurface
Rollingtopography
Valley benches
MAP1.
Fig. 18. Denton et al. p. 181.
Geomorphologicalmap showing the distributionof the landscapetypes and the relationshipto
reconstructedriver valleys. The main landscape elements start with a high originalsurface that
supportslarge, isolated mountainremnants.Intermediatesurfacesare cut into the highrockplatform and separatedfrom it by a prominentcuesta, which is the markerfeatureof the landscape.
The isolated buttes, needles, and mesas of the intermediatesurfacesform a spectacularinselberg
landscape.The present valley system is incised in this landscapeand forms a dendriticnetwork
with a sinuousplanprofile.All majorvalleys have high-leveltributaries,some gradedto the inter-
mediate surface and others to valley benches. The
landscapeis minor.Small cirque basins occur incise
and, in places, valleys have been widened and straig
mittentlybetween episodes of streamincision and v
occursin centralTaylorValley,where truncatedspuI
on the valley walls and sinuous streamchannelsocc
bounded
slopes
It"'"'.'"'
.....
::
:'::::::
j^^%i
ig (glacial)
.realscouring
Q(v
ear slope
itte
._
*
phy
Rollingtopograp
Valleybenches
nt
Coastalpiedmor
,nt
Valleyconcorda
withvalleybenc:hor
surface
planation
Contour(m)
Spotheight (m)
nches. The imprintof glacial erosion on this semi-arid
occurincised sporadicallyalong some rectilinearslopes
d and straightened.Glacial modificationoccurredinter;isionand valley downcutting.The best evidence for this
ncated spursand straightenedvalley sections occurhigh
iannels occur at the valley bottom (see Figs 11 and 17).
0.
f
('~'-'-~""'~~~~~ff
"N
ff.
th
ff-
>'~....i
?
.~"
o ~.,
~...J ? . .b
........~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~f
'!i~
"
'-'""
o~
t~~~~~~'R~~~~~~~~
t..'':- ^0-rIA
'~,'. ~..k.^?,.i
i
SlumpMountain
...
*n
|_
|ni,<
Ripples ~
- ?,..'.:~
,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Wle
/ .....:.:'...
..........~~~~~~DkesMeaure
:.~~~~~~~~~~~~~a
.....X.-:..LA'.'NATI"."O...N
~.._~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~erc
M:e....h:~.
.....
.;ll-te
Sts'o
___.'_
_____^^
i.,,r~/~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~rhourzt
opgah
Unifretae
Divid
,~(~~~~~~~~~~~~~~~~~
~Bahtill
(saprdg)
,i~~~~~~~~~~~~~~~~~~~~~~~~~olvu
...D_-;OL
)
ruddrd
Quartermain~
I'ilDvd
~ ~ ~Unifretae
colluvium
(SS)
0~~,:
Slm
onai.oFrorDlrt
0~~~~~~~,,
IItil
':'~~~~~~~~~~~~~~~~~uremi
roas /'~~~~
diam...cton
~~
~ :"~.~
~~~~
Unifrnitd,,
~
~~
A,~
S^'''R
\. /^ g-B^^
eO
)t
?
0
>
t\
/
l
S
" /'
'*'*:::
''''^
'
* ^^^\
*^ \
s\
g
|v-
C?<S?
*
BeacoE
XPLANATION
^
EI
/
J}R
I Brawhm ill
| A|e]
r
'[~
Ripples
aa
E<
^ Dikes
T errace edge
v
^
:
Arena till
1 Altartill
Divide (sharp ridge)
Undifferentiated|
Quartermain
II till
Ecl
tt"
'
F7
F
Stoss/Lee topography anco
^s^^
^^
]
X
i^J^^^^^
* * ^^
~
^
|
b
^[ =
Eas/t
^
*'?'V^t
6!.
//
rj^
.
^ -.
<
A
<
s
e e'L^'(
Ri
~~~~~~~~
Ilnimehia
S
1A
-
Ridge
l@
diamictonL
Colluvium chute
diamicton
Snow and ice
1
'Depositcontactre)
._ Monasterycolluvium -r Patternedground
|n
,|-*-|
Channel
_
14
-
-
EXPLA
NATION
.
Im w snd ow
IFH.]
Hummocky
dposits
i
[I
Alpin I drm
sslo@
| Prt-lpin
|-
Indlcatesone deosit
"o
overlying
anothrr
|-:L
11drift
Rock glacier
-
Un
alpine
& --| UndOffiontlot.d
"s IorMt ? i s H
P
e
ountaln'
P..k
W
:erv~;,Goepooll,Llneatlons
~
-
Mlorin rst
f--
till
PGlalus
Alpine
m broI
morhln
Wgy
g
[
Colluvilted drift
morphologycovd Lo
by Alpine
Z
Debris
flw
[11
drift
(4n m,
channel
oLdrft
drif
3
p
!1
t
Hart1, IIa, lib drift
;
Trilogydrf
Br'tly
,l Ia., Ilb drif
3Good.pd
M
(realict)
el
Bouder.s
H
Undlffferntlatd drlff
Gelfluctlonlohbe
I
**
(:v!)
/
[Scarp
[
Colluvlum
1, 11,-II dt"ft
Moserve I, la, &lb drift
Wrightdrift
| Undifferentiated
alpine III/IVdrift
,iLLi7
Excavation
("Ar/mArsample)
\
chute
u~~~~
~ ~
\\
A'
0,..
1
Surf
Geom
6C
Eastern
A
0
40
MAP3.
Fig. 28. Denton et al. p. 192.
Fig. 5. Hall et al. p. 245.
Glacialgeologic and surficialgeomorphicmap of east-centralWrightValley.From Hall et
al. (1993).
Surficialgeologic map of east-centralWrightValleyshowingthe
alpine glacierdrifts, and Ross Sea drifts.