Icarus 181 (2006) 388–407
www.elsevier.com/locate/icarus
Debris-covered piedmont glaciers along the northwest flank
of the Olympus Mons scarp: Evidence for low-latitude ice
accumulation during the Late Amazonian of Mars
Sarah M. Milkovich a,∗ , James W. Head a , David R. Marchant b
a Department of Geological Sciences, Brown University, Box 1846, Providence, RI 02912, USA
b Department of Earth Sciences, Boston University, Boston, MA 02215, USA
Received 31 August 2005; revised 29 November 2005
Available online 24 January 2006
Abstract
We use Viking and new MGS and Odyssey data to characterize the lobate deposits superimposed on aureole deposits along the west and
northwest flanks of Olympus Mons, Mars. These features have previously been interpreted variously as landslide, pyroclastic, lava flow or glacial
features on the basis of Viking images. The advent of multiple high-resolution image and topography data sets from recent spacecraft missions
allow us to revisit these features and assess their origins. On the basis of these new data, we interpret these features as glacial deposits and the
remnants of cold-based debris-covered glaciers that underwent multiple episodes of advance and retreat, occasionally interacting with extrusive
volcanism from higher on the slopes of Olympus Mons. We subdivide the deposits into fifteen distinctive lobes. Typical lobes begin at a theater-like
alcove in the escarpment at the base of Olympus Mons, interpreted to be former ice-accumulation zones, and extend outward as a tongue-shaped
or fan-shaped deposit. The surface of a typical lobe contains (moving outward from the basal escarpment): a chaotic facies of disorganized
hillocks, interpreted as sublimation till in the accumulation zone; arcuate-ridged facies characterized by regular, subparallel ridges and interpreted
as the ridges of surface debris formed by the flow of underlying ice; and marginal ridges interpreted as local terminal moraines. Several lobes
also contain a hummocky facies toward their margins that is interpreted as a distinctive type of sublimation till shaped by structural dislocations
and preferential loss of ice. Blocky units are found extending from the escarpment onto several lobes; these units are interpreted as evidence of
lava–ice interaction and imply that ice was present at a time of eruptive volcanic activity higher on the slopes of Olympus Mons. Other than
minor channel-like features in association with lava–ice interactions, we find no evidence for the flow of liquid water in association with these
lobate features that might suggest: (1) near-surface groundwater as a source for ice in the alcoves in the lobe source region at the base of the
scarp, or (2) basal melting and drainage emanating from the lobes that might indicate wet-based glacial conditions. Instead, the array of features
is consistent with cold-based glacial processes. The glacial interpretations outlined here are consistent with recent geological evidence for lowlatitude ice-rich features at similar positions on the Tharsis Montes as well as with orbital dynamic and climate models indicating extensive snow
and ice accumulation associated with episodes of increased obliquity during the Late Amazonian period of the history of Mars.
 2005 Elsevier Inc. All rights reserved.
Keywords: Mars, surface; Mars, climate; Mars, atmosphere
1. Introduction and background
Fan-shaped lobate deposits are found on the west and northwest flanks of the Tharsis Montes and Olympus Mons and those
on Olympus Mons are superposed upon the surrounding aure* Corresponding author. Currently at the Jet Propulsion Laboratory, Mail Stop
183-501, 4800 Oak Grove Drive, Pasadena, CA 91109, USA.
E-mail address: sarah.m.milkovich@jpl.nasa.gov (S.M. Milkovich).
0019-1035/$ – see front matter  2005 Elsevier Inc. All rights reserved.
doi:10.1016/j.icarus.2005.12.006
ole deposits (Fig. 1). The Olympus Mons deposits have been
variously interpreted to be landslide deposits, explosive and effusive volcanic deposits (Carr et al., 1977; Scott and Tanaka,
1986; Morris and Tanaka, 1994) and rock glaciers (Lucchitta,
1981). With the recent availability of multiple, high-resolution
datasets and in light of recent analyses of proposed glacial features at the Tharsis Montes (e.g., Head and Marchant, 2003;
Shean et al., 2005; Parsons and Head, 2005), we have undertaken a re-examination of these intriguing deposits. We begin
Olympus Mons Amazonian debris-covered glaciers
389
Fig. 1. Location and setting of the lobate deposits. (A) Location of lobate deposits (yellow; As, slide material) on the NW margin of Olympus Mons and the Tharsis
Montes (from Scott and Tanaka, 1986). See Scott and Tanaka (1986) for further description of additional units. (B) MOLA gradient map of Olympus Mons and the
surrounding aureole. A–A shows the location of the profile in (D). (C) Distribution of the mapped unit along the base of the Olympus Mons scarp (from Morris and
Tanaka, 1994). The apron materials are shown in shades of yellow and consist of Aar (ridged apron material) and Aah (hummocky apron material). See Morris and
Tanaka (1994) for more detailed description of additional units. (D) MOLA topographic profile across Olympus Mons (derived from 128 pixel/deg gridded data).
Location shown in (B).
by reviewing the characteristics of the aureole deposits, which
contrast greatly with the lobate deposits on top of them. We next
summarize the previous studies of the lobate deposits themselves and review the morphological characteristics of four pos-
sible emplacement mechanisms—landslide, glacial, pyroclastic, and lava flows—as observed on Earth. We then examine
the lobate deposits in recent Mars Global Surveyor and Mars
Odyssey images and topographic data to determine the most
390
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
plausible emplacement mechanism. Finally, we discuss the implications of our interpretations for martian climate during the
Late Amazonian.
Olympus Mons, a massive martian volcano located in the
equatorial regions, rises over 25 km and is surrounded by a
basal escarpment up to 8 km high (Fig. 1). Extending as much
as 1000 km beyond this scarp are lobes of ridged materials
known as the aureole deposits (Carr et al., 1977; MouginisMark et al., 1992). The aureole deposits form an asymmetric ring of material around Olympus Mons and are made up
of at least seven distinct lobes that were deposited in at least
four separate intervals. The surfaces of these lobes are characterized by a corrugated texture, formed from curvilinear,
flat-topped ridges and troughs with heights of several hundred
meters to a few kilometers (Tanaka, 1985). The corrugations
are generally parallel to the margins of aureole lobes (Tanaka,
1985). The lobes are up to 4 km thick (Francis and Wadge,
1983) and in places they are cut by graben (Morris, 1982;
Francis and Wadge, 1983). Linear faults and fractures in the aureoles do not cut through multiple overlying lobes nor do they
intersect the basement; this implies that the aureole deposits are
mechanically detached from the underlying material (Francis
and Wadge, 1983). Mechanisms favored for the emplacement of
the aureole deposits include large catastrophic landslides originating at the basal escarpment (Lopes et al., 1982), or gravitydriven landslides of lubricated material, possibly ice (Tanaka,
1985).
Superimposed upon the west and northwest aureole deposits
at the base of the escarpment (Fig. 1) are smaller lobate features
described by Scott and Tanaka (1986) as “fan-like corrugated
sheets as wide as 600 km that appear to override topographic
obstacles without deflection of internal structure; source areas
(are) hummocky, (and) contain small hills and circular depressions” (Scott and Tanaka, 1986). Carr et al. (1977) observed
that steeper portions of the basal scarp are associated with lobes
containing larger marginal zones of arcuate ridges.
Many mechanisms have been proposed for the emplacement
of the lobate deposits. Scott and Tanaka (1986) interpreted the
lobate deposits to be avalanches of volcanic debris from either slope failure or explosive volcanism. Other initial hypotheses for the emplacement mechanism of these deposits include
landslide debris aprons lubricated during emplacement by compressed gas (Carr et al., 1977). Lucchitta (1981) demonstrated
that the morphology of these lobate deposits as revealed in
Viking imagery is very similar to that of terrestrial glaciers.
Ridges associated with these deposits are superposed on the
surrounding aureole and are not deflected by underlying topography; thus Lucchitta (1981) interpreted them to be draped over
the aureole surface as might be the case for ridges produced
by downwasting of a debris-covered glacier or rock glacier.
Lucchitta (1981) also stressed the morphological similarities
of the lobate deposits with those associated with terrestrial
piedmont glaciers such as the Malaspina Glacier (Fig. 2) (and
Fig. 5b in Lucchitta, 1981). The lobate deposits on Olympus
Mons and the supraglacial debris on the Malaspina Glacier both
show long, even, curvilinear ridges that are subparallel to deposit margins.
Fig. 2. Terrestrial piedmont glacier deposit: Malaspina Glacier, Alaska.
(A) False color Landsat 7 image. Note subarcuate, parallel ridge structure towards margins. Red indicates vegetation. This image was taken on 8/31/00 and
is from Landsat 7 WRS Path 63, Row 18, center: 60.10◦ , −141.78◦ . (B) Perspective view created from false color Landsat image in (A) and Shuttle Radar
Topography Mission elevation model with a vertical exaggeration of 2. Glacial
ice is light blue, snow is white, vegetation is green, bare rock is gray and tan.
The view is oriented to look north. Images courtesy NASA/JPL/NIMA.
Subsequent to the initial Lucchitta (1981) Viking-era study
several new datasets have been acquired: Mars Global Surveyor’s high-resolution Mars Orbiter Camera (MOC) images
(with meters/pixel resolution) and Mars Orbiter Laser Altimeter (MOLA) topographic data, as well as Mars Odyssey’s Thermal Emission Imaging System (THEMIS) near-IR images (with
100 m/pixel resolution). In this study we employ these new
datasets to assess four previously proposed mechanisms for emplacement of the lobate deposits: landslide, glacial, pyroclastic,
or lava flow. We then place these deposits in the context of recent research into martian climate history.
What criteria can be used to distinguish among the multiple suggested origins for these deposits? Here we examine the
characteristics of terrestrial landslide, volcanic and glacial deposits in order to provide a framework in which to interpret the
morphology of the lobate deposits.
Landslide deposits Two major categories of terrestrial
landslide deposits, slumps and debris avalanches, have fundamentally different morphologies that reflect differences in their
Olympus Mons Amazonian debris-covered glaciers
mode of origin. The debris apron of a slump deposit is wide
relative to its length, with common length-to-width ratios of
approximately 1:2. Such a debris apron is also often cut by
transverse faults to form large blocks and ridges, and is generally steep along curvilinear fronts. Slump deposits may also
lack a well-defined source region (Moore et al., 1989).
In contrast, the morphology of the apron of a debris avalanche commonly shows trough structures near source regions
and hummocky terrain near distal ends (Moore et al., 1989;
Ablay and Hürlimann, 2000). The surface of the apron is characterized by longitudinal grooves and transverse ridges, giving rise to an overall corrugated texture (Moore et al., 1989;
Ablay and Hürlimann, 2000; Laberg and Vorren, 2000). Debris
avalanches originate at horseshoe-shaped depressions or theaters (Moore et al., 1989) that commonly widen in the direction
of avalanche movement (Moon and Simpson, 2002).
Pyroclastic flow deposits Pyroclastic surge and flow deposits are emplaced by laterally moving density currents containing ash particles and gas. The morphology of pyroclastic
deposits is dependent on several factors, including the proportion of ash and gas in the flow. Pyroclastic flows tend to extend
radially away from a source vent and are controlled by local
topography. They typically pond in depressions and thin out on
topographic highs (Freundt et al., 2000).
Lava flow deposits Lava flows tend to be hundreds to thousands of times longer than they are wide and are made up of
multiple overlapping sequences of lobes. The fronts and margins may have complex budding tongues and toes. A lava flow
can have a surface characterized by irregular crustal fragments
or be smooth and continuous depending on the composition of
the flowing material (Kilburn, 2000).
Glacial deposits The morphology of glacial deposits is
controlled to a large extent by the thermal regime of glacier ice
(wet-based or cold-based), the extent of surface melting, the origin and abundance of supraglacial, englacial, and basal debris,
and the geometry of underlying terrain (e.g., Benn and Evans,
1998).
Valley glaciers For typical valley glaciers that achieve wetbased conditions as well as surface melting in ablation zones,
debris that is entrained by surface rockfall and/or basal plucking is ultimately deposited as basal till, ablation till, stratified
ice-contact drift, or proglacial fluvial and lacustrine sediment.
Typical landforms associated with these glaciers include eskers;
terminal, lateral, and recessional moraines; kames and kame
terraces; ice-contact deltas; and outwash. Ice-margin advances
associated with these wet-based glaciers tend to erode previous deposits, resulting in a single flow trace at the ground surface that represents active flow during ice retreat. Multiple flow
traces, showing a palimpsest landscape, may be preserved in regions overrun by polythermal ice (e.g., Kleman, 1994). In this
case, dissection and erosion of earlier deposits is incomplete
and two or more flow traces may be preserved on a regional
scale. Complete preservation of unconsolidated landforms is
possible beneath cold-based ice. In such glaciers, basal ice remains below the pressure-melting point; and flow thus occurs
391
without basal sliding, accommodated entirely by internal deformation. Such cold-based glaciers are common in the Antarctic
Dry Valleys (e.g., Waller, 2001) and may also occur on Mars
(Head and Marchant, 2003).
Debris-covered glaciers As the name implies, debriscovered glaciers contain a core of glacier ice that is overlain
by a veneer of surface debris. This veneer of surface debris,
which may be meters thick, most commonly originates through
a combination of rockfall onto glacier ice and/or sublimation
of debris-rich ice. The veneer protects the underlying ice and
slows melting and/or ice sublimation; as a result, ice in terrestrial debris-covered glaciers may be considerably older than
typical valley glaciers. Flow of the buried glacier ice creates a
ridge-and-furrow topography in the surface debris (e.g., Ackert,
1998; Kääb and Weber, 2004).
Non-glacial ice-rich lobate deposits Some lobate deposits
that superficially resemble those produced by glaciers, particularly debris-covered glaciers, may in fact be related to flow
and deposition associated with non-glacial, icy regolith (for a
review, see Martin and Whalley, 1987; Whalley and Martin,
1992). These non-glacial deposits include rock glaciers (with a
core of non-glacier ice), protalus ramparts, and protalus lobes.
Rock glaciers Rock glaciers are a mass of fine-to-coarsegrained material that contains interstitial ice and/or an ice core;
in active rock glaciers, surface debris commonly shows ridges
and furrows suggestive of downslope flow. Rock glaciers may
have a variety of morphologies depending on the surrounding topography. Confined within a valley, the lobate body is
commonly termed a tongue-shaped rock glacier; if distal margins are not confined topographically, lateral flow may yield a
spatulate shape, culminating in a piedmont rock glacier. Rock
glaciers may be tens of meters to a few kilometers long (Martin
and Whalley, 1987; Whalley and Martin, 1992). Confusion relating to the origin of terrestrial rock glaciers arises from the
fact that some rock glaciers are cored wholly or in part with
glacier ice; in such cases, we prefer to use the term debriscovered glacier; we reserve rock glaciers for those deposits
cored by segregation ice or where ice origins are unknown.
Protalus ramparts and lobes These lobate deposits generally form at distal boundaries of perennial snowbanks. The
snow surface provides a rampart on which rockfall debris slides
and accumulates at distal margins. Periodic snowmelt infiltrates
rockfall debris, occasionally forming pore ice that may be of
sufficient concentration to yield creep deformation and downslope flow. Protalus lobes commonly occur where ice-rich, flowing deposits, are frequently stacked one on another. Protalus
lobes and ramparts are commonly found in association with
rock glaciers (Whalley and Azizi, 2003).
2. Detailed characteristics of the lobate deposits
at Olympus Mons
Using the range of basic characteristics and recognition criteria outlined above, we now turn to a detailed description of
392
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
(A)
Fig. 3. (A) Mosaic of THEMIS IR images of lobate deposits. Individual lobes are numbered. Boxes indicate regions examined in more detail in the following figures;
Region 1 in Fig. 4, Region 2 in Fig. 5, Region 3 in Fig. 6. (B) Sketch map. Units are as follows: red indicates the basal escarpment, characterized by alcoves and
a scarp lip with variable morphology; gray indicates the chaotic facies, topographic lows characterized by disorganized hillocks; light purple indicates the blocky
facies, characterized by knobs oriented in a linear pattern; light blue indicates the arcuate-ridged facies, lobes extending from the basal escarpment and characterized
by arcuate, subparallel ridges; dark blue indicates the hummocky facies characterized by a series of pits and knobs; these units are described further in Section 2.
Orange indicates the Olympus Mons flank; yellow indicates the material surrounding Olympus Mons and the lobate deposits, which is made up of aureole deposits,
volcanic flows, and aeolian deposits.
the deposits themselves. In terms of their general, broad distribution, the lobate deposits rest unconformably on older, aureole
deposits and lava plains derived from Olympus Mons eruptions,
and extend for over 340 km from an azimuth of N 30◦ W to
S 70◦ W along the western Olympus Mons scarp (Figs. 1, 3–6)
covering an area of ∼15,000 km2 . The deposits are made up of
Olympus Mons Amazonian debris-covered glaciers
393
(B)
Fig. 3. (continued)
fifteen lobes, individually extending tens to one hundred kilometers from the base of the scarp; lobes have a spatulate form,
increasing in width as they extend outward from the base of
Olympus Mons. We examined the lobate deposits at a variety
of scales in THEMIS, MOC, and MOLA data. An overview
of topography is presented in Figs. 8–10. In the figures and
descriptions that follow, the region of the lobate deposits is subdivided into three regions identified in Fig. 3; individual lobes
are numbered in this figure as well. A THEMIS IR mosaic and
a map of the lobate deposits are found in Fig. 3 and serve as
an overview for the more detailed figures of each region that
follow. THEMIS mosaics, geologic sketch maps, and MOLA
394
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
Fig. 4. Region 1. (A) THEMIS IR mosaic. (B) Sketch map of THEMIS mosaic. Labeled units follow the definitions of Scott and Tanaka (1986); Aos is Olympus Mons
shield member, Aoa4 is Olympus Mons aureole member 4 (the uppermost aureole unit), Ae is aeolian deposits. Numbers indicate individual lobes. (C) Geologic
sketch map of region; orange unit is Olympus Mons flank, red is the basal escarpment, gray is the chaotic facies, light purple is the blocky facies, light blue is
the arcuate-ridged facies, dark blue is the hummocky facies, and yellow is surrounding aureole, volcanic, and aeolian deposits. (D) MOLA topography map of the
region; contour interval is 125 m.
data of each region (Figs. 4–6), and MOLA profiles (Fig. 7)
provide the basis for our characterization. Each facies and/or
feature on the map in Fig. 3B is described in the sections that
follow, beginning at the base of Olympus Mons and working
radially outward. These include the basal escarpment surrounding Olympus Mons (Figs. 1B and 1D), four facies (chaotic,
blocky, arcuate-ridges, and hummocky), and a distal marginal
ridge.
The basal escarpment, indicated in red in Fig. 3B, is found
at the base of Olympus Mons (Fig. 1) and rises to heights
of 8 km. The lip of the scarp has variable morphology along
its extent (e.g., Basilevsky et al., 2005); throughout Region 1
(Fig. 4A) and in several locations in Region 2 (Fig. 5A) and
Region 3 (Fig. 6A), the lip displays an arcuate, sharp-crested
ridge. Arcuate, outward-facing, theater-like alcoves (Fig. 8)
are found throughout the escarpment. Layered outcrops made
up primarily of bedded lava flows of the flank of Olympus
Mons dominate the upper part of the scarp and form spurs
and ridges (see Basilevsky et al., 2005 for more details), giving way to extensive talus deposits in the lower half of the
scarp. The talus piles and aprons clearly embay and postdate the hillocks and knobs of the hummocky facies (Figs. 8B
and 8C). In portions of Regions 2 and 3, the lip at the top
of the escarpment is gentle, rather than sharp-crested, consis-
Olympus Mons Amazonian debris-covered glaciers
395
Fig. 5. Region 2. (A) THEMIS IR mosaic. (B) Sketch map of THEMIS mosaic. Labeled units follow the definitions of Scott and Tanaka (1986); Aos is Olympus
Mons shield member, Aoa4 is Olympus Mons aureole member 4 (the uppermost aureole unit). Numbers indicate individual lobes. (C) Geologic sketch map of
region; orange unit is Olympus Mons flank, red is the basal escarpment, gray is the chaotic facies, light purple is the blocky facies, light blue is the arcuate-ridged
facies, and yellow is surrounding aureole, volcanic, and aeolian deposits. (D) MOLA topography map of the region; contour interval is 125 m.
tent with a mantle of superposed lava flows emanating from
the flanks of Olympus Mons. The four mapped facies (chaotic,
blocky, arcuate-ridged, and hummocky) all extend outward
from the base of the scarp in approximately the sequence mentioned.
The chaotic facies, shown in gray in Fig. 3B, is found immediately adjacent to the scarp, usually defining a topographic
low in the core of an individual lobe (Figs. 4D and 7). It is characterized by disorganized hillocks, each ranging in size from
hundreds of meters to tens of kilometers (Fig. 9). Collectively
the facies extends outward tens of kilometers from the basal
escarpment (Fig. 3). The proximal contact is obscured by superposed talus from the scarp (Figs. 8 and 9), while the distal
contact is transitional to the next outermost arcuate-ridged facies (Figs. 8C and 9).
The blocky facies, indicated in light purple in Fig. 3B, is
found in areas where the lip of the adjacent escarpment is
smooth rather than sharp-crested, and the slopes of the scarp
396
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
Fig. 6. Region 3. (A) THEMIS IR mosaic. (B) Sketch map of THEMIS mosaic. Labeled units follow the definitions of Scott and Tanaka (1986); Aos is Olympus
Mons shield member. Numbers indicate individual lobes. (C) Geologic sketch map of region; orange unit is Olympus Mons flank, red is the basal escarpment,
gray is the chaotic facies, light purple is the blocky facies, light blue is the arcuate-ridged facies, and yellow is surrounding aureole, volcanic, and aeolian deposits.
(D) MOLA topography map of the region; contour interval is 125 m.
are much lower (compare topographic maps in Figs. 4D, 5D,
and 6D). This facies (Fig. 10) commonly extends 40–70 km
from the base of the scarp; it thus forms the majority of the lobate deposit in these regions. The surface of the blocky terrain
is characterized by oriented knobs that are arrayed in a linear, generally radial pattern outward from the base of the scarp
(Fig. 10). The individual knobs in this facies can be as small as
a few hundred meters, or thousands of meters long and hundreds of meters wide. The closer to the basal escarpment, the
longer and more distinctly linear are the features (right-hand
parts of Figs. 10A–10C). The linear aspects of the proximal
parts of the deposits are generally contiguous with radial lava
flows streaming down the sides of Olympus Mons. Basilevsky
et al. (2005) described the blocky facies as part of their type 2
slopes.
Indicated in light blue in Fig. 3B, the arcuate-ridged facies
forms the majority of the lobate features in areas where the lip
of the basal escarpment is sharp-crested and characterized by
numerous alcoves (Figs. 4A, 4B, 5A, 5B, 6A, 6B). Individual
lobes of the ridged facies are characterized by regular, arcu-
Olympus Mons Amazonian debris-covered glaciers
397
Fig. 7. (A) Contour map of northwest fan-shaped deposits and surrounding region. Lines and letters show locations of topographic profiles in (B). Contour interval
is 125 m. (B) MOLA topographic profiles across the lobate deposits.
ate, subparallel ridges up to 60 km long (Fig. 11). In many
locations, these ridges are draped over preexisting topography
such as knobs or other ridges (Fig. 11B, left side) rather than
curving around them. Many depressions are found in this facies
(Fig. 11A, right; Fig. 11C, upper right; Fig. 11D, right); several
are circular and are interpreted to be small impact craters; most
others are irregularly shaped. Depressions tend to be hundreds
of meters wide and thousands of meters long with depths on
the order of tens of meters. This facies extends up to 100 km
from the base of Olympus Mons (Fig. 3). Topographic profiles show that individual lobes thin toward their distal margins,
commonly have convex upward profiles (Fig. 7), and reach as
much as 800–1200 m in thickness.
The hummocky facies, indicated by dark blue in Fig. 3B, is
found at the distal margin of several lobes in Region 1 (Fig. 4).
This unit is 1–7 km wide and is characterized by a series
of pits and knobs hundreds of meters in diameter (Fig. 12).
The facies often forms outward-facing indentations in adjacent terrain, which follow the local to regional topography (see
Fig. 12B).
The marginal ridge facies is comprised of a continuous ridge
found along the outer margin of each lobe independent of the
interior morphology of the lobe; Fig. 13A illustrates this ridge
at the margin of the arcuate-ridged facies and Fig. 13B shows it
at the margin of the blocky facies. At locations where multiple
lobes meet, the marginal ridge facies is broken into individual
linear segments along the boundary between the lobes; here
several marginal ridges appear to be superposed (Figs. 13C–
13E). The distal marginal ridge is typically less than 100 m in
height, but can be up to 125 m.
Although facies exposure varies among lobes, the relative positioning of facies in each lobe is consistent with respect to the basal escarpment. For example, the sequence for
lobe 2 (Fig. 4) shows, extending outward from the basal scarp:
chaotic facies, arcuate-ridged facies, hummocky facies, and the
marginal-ridge facies. Similarly, the sequence at lobe 3 (Fig. 4)
shows the basal escarpment, a small amount of chaotic terrain, arcuate-ridged facies, hummocky facies, and the marginal
ridge. Although lobe 9 (Fig. 5) only exhibits the blocky facies
and the marginal ridge, these are arrayed in the sequence observed in other lobes.
The material that surrounds the lobate deposits has been defined by Scott and Tanaka (1986) as Aos, the Olympus Mons
shield member found on the flanks of the volcano and on
the smooth surrounding plains as well. It is made of complex, interfingering lobes of lava flows and contains channels
and levees of volcanic origin as well as collapse pits. In our
geologic sketch maps (Figs. 3–6), we have differentiated between Aos on Olympus Mons itself and Aos on the surrounding plains in the interest of highlighting the location of the
basal escarpment. Corrugated material cut by deep troughs and
graben, found in the northwest corner of Region 1 (Fig. 4), is
identified by Scott and Tanaka (1986) as Aoa4 , the youngest
member of the aureole unit. North of this material are broad
398
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
Fig. 9. Examples of exposures of the chaotic facies. (A) Chaotic facies in
lobe 15. (B) Chaotic facies in lobe 2. Note disorganized hillocks.
Fig. 8. Examples of exposures of the basal escarpment. (A) Basal scarp associated with lobe 7. Note arcuate, cirque-like feature. (B) Basal scarp associated
with lobe 5. Note alcove-like feature that has been partially filled in by talus
aprons. (C) Basal scarp associated with lobe 3. Note possible alcove-like feature that has been partially filled in by talus aprons.
level plains that appear locally etched and striated in the direction of prevailing winds; these are interpreted as aeolian
deposits.
A survey of MOC images found throughout the western
side of Olympus Mons reveals that aeolian features such as
dunes cover the surfaces of all units. Fig. 14 shows several
examples of dunes in the lobate deposits; Figs. 14A and 14B
contain dunes between individual arcuate ridges of lobe 10
while Fig. 14C shows dunes on the surrounding plains beyond
the marginal scarp as well as dunes within the lobe 10 itself.
Fig. 14D shows dunes between individual ridges of the arcuateridged facies and the marginal ridge between lobes 3 and 4.
The dunes are also found within crater floors in this area. This
is consistent with Thermal Emission Spectrometer (TES) thermal inertia and albedo studies that indicate the region is covered
by fine-grained dust (e.g., Putzig et al., 2005).
We found no evidence of fluvial channels in association with
the lobate deposits, either in their source regions (the scarp areas) or at their lateral or distal margins, or generally within the
ridges and materials of the lobate deposits themselves. The one
exception to this is the presence of several small channel-like
features seen in the blocky facies near the edge of the scarp
(Fig. 5).
Olympus Mons Amazonian debris-covered glaciers
Fig. 10. Examples of exposures of the blocky facies. (A) Blocky facies in lobe 8.
Blocks are arranged in a linear pattern radial from the basal escarpment (located
to the right); features are more distinctly linear towards the basal escarpment.
(B) Blocky facies in lobe 9. Note radial pattern from the basal escarpment (located to the right). (C) Blocky facies in lobe 13. Note extensive individual
knobs.
3. Interpretation of the lobate deposits
We now assess the most plausible emplacement mechanisms
for the lobate deposits. Because the deposits lack (1) the corrugated texture of terrestrial debris avalanches, (2) the topographic profile of typical terrestrial slump deposits, and (3) the
radial grooves of martian landslides such as those found in
Valles Marineris (Fig. 15A; Lucchitta, 1978; Lucchitta et al.,
1992), we do not favor a landslide origin for the Olympus Mons
lobate deposits. While portions of the lobate deposits appear
to be influenced by lava flows (e.g., lobe 9 and the associated
region of the basal escarpment), many features are inconsistent with emplacement by lava or pyroclastic flow, including
the chaotic facies near the basal escarpment and the hummocky
facies at the distal margin of several lobes, as well as the ubiquitous arcuate-ridged facies. Furthermore, the margin of a typical
martian lava flow (Fig. 15B) is characterized by many small
lobes and tongues rather than the continuous, regular ridges
(Fig. 13) found at the margin of the Olympus Mons lobate deposits (Fig. 15C). This implies that the lobate deposits are not
lava or pyroclastic flow features. There is no evidence of infilling of topographic lows to a relatively level surface, which
399
would distinguish pyroclastic flows from lava flows (see profiles in Fig. 7). Candidate volcanic source vents have also not
been identified.
Among the various candidate origins for the lobate deposits
previously mentioned, we conclude that a glacial origin is most
consistent with the data, as originally proposed by Lucchitta
(1981) on the basis of Viking data. The multiple lines of evidence originally cited by Lucchitta are considerably strengthened by our analysis of the new data. Furthermore, among the
strengths of the glacial model is that it explains the relative
positioning of the facies; although each facies considered in
isolation might not be entirely diagnostic of a glacial origin,
the observed facies sequence is strongly suggestive of glacial
processes.
Following this line of reasoning, we interpret the theaterlike alcoves in the basal escarpment as former ice-accumulation
zones, many of which show the types of widening and overdeepening typical of terrestrial cirques. The main body of the lobate deposits could be interpreted as a combination of debriscovered glaciers and/or rock glaciers. We conclude that these
deposits originated as multiple debris-covered glaciers: in this
scenario, snow and ice that accumulated in alcoves was covered with rockfall emanating from the high-cliffs as the ice
flowed outward from the basal scarp. Progressive sublimation
in the advancing exposed ice led to concentration of debris
on the surface and the formation of an overlying sublimation
till. If the rockfall and resulting sublimation till were of sufficient thickness and extent, they could potentially significantly
retard the loss of underlying ice (e.g., Marchant et al., 2002;
Helbert and Benkhoff, 2005). On the other hand, if rockfall
was sparse and discontinuous and the ice contained minimal
englacial debris, then sublimation could remove all ice leaving
only a “residue” of sublimation till. The presence of distinctive
talus cones derived from the upper cliffs subsequent to glaciation (Figs. 8 and 9), suggests that there was a plentiful source
of supraglacial debris to partially cover the accumulation zone
and contribute to the evolving sublimation till. We interpret
the chaotic facies at the base of the escarpment to be largely
composed of this type of sublimation till. The dominance of
exposed ice in the proximal regions would result in preferential sublimation of ice in the accumulation zone, producing a
depression, less sublimation in the thin till distal to the accumulation zone, and production there of chaotic hummocks. The
relatively low topographic depression at the site of the chaotic
facies (Fig. 7) suggests that the chaotic facies lacks a uniform
core of ice, but some of the individual larger hills could be underlain by an ice core protected by debris.
The continuous, curvilinear ridges characterizing the arcuate-ridged facies are very similar to those found on rock glaciers
and debris-covered glaciers and are interpreted as ridges of surface debris formed by the flow of underlying ice (e.g., Martin
and Whalley, 1987; Adamson and Colhoun, 1992). Comparison
of the two different views of the Malaspina Glacier (Fig. 2) and
the Olympus Mons lobes (Figs. 4–6, 15C) shows a remarkable
similarity in terms of scale and morphology. The limited available data do not allow us to confirm the presence or absence
of ice in this facies, but we speculate that if sufficient debris
400
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
Fig. 11. Examples of exposures of the arcuate-ridged facies. (A) Arcuate-ridged facies in lobe 3. Note distinctive regular, subparallel ridges. (B) Arcuate-ridged
facies in lobe 7. Note ridges draped over preexisting topography (left side). (C) Arcuate-ridged facies in lobe 15. (D) Arcuate-ridged facies in lobe 10. Many
depressions can be observed in this lobe.
Fig. 12. Examples of exposures of the hummocky facies. (A) Hummocky facies in lobe 1. (B) Hummocky facies in lobe 3.
covered underlying ice, then remnant ice may still core this facies; such ice, if present, would account for the convex-upward
profile of lobes 1 and 3 (Fig. 7). Radar experiments on Mars
Express and Mars Reconnaissance Orbiter may be able to test
for the presence of buried ice.
The hummocky facies at the edge of several lobes (Fig. 13)
is interpreted as sublimation till locally enhanced by structural
deformation of the ice. The location of the facies is coincident with a rise in the topography toward a plateau and the
individual parts of the facies are developed in outward-facing
depressions in this topography (Fig. 12B). The preferential development of the hummocky facies where the margin of the
glacial deposit would interact with the background topography
is typically where thrusting and fracturing of the protective debris cover occurs in terrestrial glaciers (e.g., Benn and Evans,
1998). This process results in exposure of the underlying ice,
causing local preferential sublimation and leaving behind a sublimation till.
The marginal ridges (Fig. 13) occur at the distal and lateral
portions of the lobes and thus are interpreted to be local terminal moraines. The distal ridges represent the furthest extent of
glacier ice in each lobe. The segments of marginal ridges between individual lobes, especially lobes 4, 5, and 6 (Figs. 13C–
13E), are interpreted as lateral moraines. Superposed and crosscutting lateral moraines suggest multiple phases of ice advance
and retreat (Figs. 13A, 13D, 13E). The unusual prominence of
Olympus Mons Amazonian debris-covered glaciers
401
Fig. 13. Examples of exposures of the marginal ridge (indicated by arrows). (A) Marginal ridge along the margins of lobes 5, 6, and 7. Note transition from
continuous ridge to multiple linear segments where these three lobes meet. (B) Marginal ridge along the margin of blocky facies in lobe 10. (C) Marginal ridge
between lobes 2 and 3. Note individual linear segments. (D) Marginal ridge between lobes 4 and 5. Note individual linear segments. (E) Marginal ridge between
lobes 5, 6, and 7. Note individual linear segments.
these ridges is attributed to the stabilization of the glacier at this
position for an extended period of time; in this case, forward
advance of the ice continues to deliver debris to the margin,
but sublimation maintains a balance with advance, resulting in
the deposition of the debris in one location as a large marginal
ridge.
The blocky facies lacks a close analog with terrestrial glacial
deposits. Its association with inferred lava flows descending the
flanks of Olympus Mons, however, lead us to conclude that it
represents ice–lava interactions (e.g., Larsen, 2002). Normally,
lava flows descending the flanks of Olympus Mons encounter
the scarp and drape over it, build up lava aprons on the flanks
402
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
Fig. 14. Aeolian features on the lobate deposits in MOC images. (A) Subframe of M12-00197; dunes between individual subparallel ridges of the arcuate-ridged
facies within lobe 10. (B) Subframe of M12-00197; dunes within the arcuate-ridged facies of lobe 10. (C) Subframe of M12-00197; dunes within the arcuate-ridged
facies of lobe 10 (lower) and the surrounding plains (upper) with the marginal ridge in between. (D) Subframe of E03-02540; dunes between individual ridges of
the arcuate-ridged facies and the marginal ridge between lobes 3 and 4. Note dunes within crater interiors.
and the flats below (see the lower part of region 3, Figs. 3
and 6). In lava flow-glacier scenario, however, lava flows from
the upper slopes of Olympus Mons flow radially downslope
and cross the escarpment, encountering glacial ice banked up
against the escarpment. On the basis of known terrestrial experience (e.g., Larsen, 2002), some melting and interaction would
take place and the flows would break up into smaller, linear
segments (interpreted to represent segments of flow margins).
Olympus Mons Amazonian debris-covered glaciers
403
Fig. 15. Examples of martian lobate deposits representing different candidate emplacement mechanisms. (A) Landslide in Ganges Chasma, Valles Marineris (e.g.,
Lucchitta, 1978; Lucchitta et al., 1992). Note radial grooved texture of debris apron. Viking Orbiter image 14A30, located approximately 10◦ S, 45◦ W. (B) Lava
flow in Daedalia Planum. Note blocky texture of flow surface and many tongues and toes along flow margin. Subframe of MOC image R22/00474, located
approximately 28◦ S, 135◦ W. (C) Debris-covered glacial lobe at the base of the Olympus Mons scarp. Note subparallel arcuate ridges in debris apron. THEMIS IR
mosaic. Lobes 2, 3, 4, and 5 as identified in Fig. 3A.
Such relationships are observed in Figs. 5 and 10. Flows descending the scarp appear to have done so in regions of lobate
glacial deposits (note the shapes and terminal moraine deposits;
Fig. 5), and undergone significant disruption in contrast to those
in areas without lobate deposits (area in the southern part of
Fig. 6). If this interpretation is correct, the blocky facies implies that in selected areas along the flanks of Olympus Mons,
volcanic and glacial activity must have overlapped in time. The
likely extended period of time in which lava flows have erupted
on the flanks of Olympus suggests that glacial ice may have
been present for geologically extended periods. Basilevsky et
al. (2005) describe similar relationships, distinguishing type 1
slopes (steep scarps with no lava flows), type 2 slopes (less
steep scarps with disrupted flows) and type 3 slopes (lowest
slopes with lava aprons without significant disruption of the
flows). They point out the existence of a micro-chaos texture
that they attribute to melting of underlying glacial ice by superposed lava flows.
Debris-covered glaciers or rock glaciers? As outlined above,
differentiation of terrestrial debris-covered glaciers from rock
404
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
glaciers in older deposits is often difficult due to the loss of subsurface ice with time. In the Olympus Mons lobes, the convex
profile of many of the lobate deposits is suggestive of a large
core of remaining ice, and hence is consistent with an origin
as debris-covered glaciers. Furthermore, the huge scale of the
Olympus Mons lobes argues against secondary ice-related rock
glaciers. The low-lying terrain associated with the chaotic facies and the hummocky facies suggests considerable ice loss,
more than that typically associated with pore ice and interspersed ice lenses of terrestrial rock glaciers. Furthermore, there
is no positive evidence of channels in the source regions that
might have been a source of supply of groundwater to form ice
for any non-glacial (rock glacier) deposits. The lack of associated meltwater features indicates that the ice, whether of glacial
origin or otherwise, failed to reach the basal pressure-melting
point and that any near-surface ice remained frozen. On the
basis of these numerous observations, we thus envision the following scenario for the emplacement of the lobes (Fig. 16): In
Phase 1, snow and ice build up on the NW slope of the basal
escarpment and are preferentially concentrated and preserved
in alcoves. Mass wasting, thermal cycling and glacial-assisted
debris eroded from the steep escarpment is deposited on top of
the ice. In Phase 2 as ice deposits build up and reach a critical thickness, they begin to flow away from the escarpment
and form a debris-covered glacier. Progressive concentration of
rockfall and debris by sublimation of ice produces an increasingly thick cover of sublimation till with increasing distance
from the source region. The glacier flows out away from the
scarp, spreading onto the flat plains to develop the spatulate
shape observed today (Fig. 3). Several periods of advance and
retreat are recorded in multiple cross-cutting lateral moraines.
At its furthest extent, each glacier deposited a large terminal
moraine (Fig. 13). During Phase 3, at some times and places
during these phases of advance and retreat, lava flows from the
Olympus Mons upper flanks erupted, flowed down and covered
portions of the basal escarpment and interacted with the ice
collected there. In Phase 4, climatic conditions changed such
that snow and ice were no longer accumulating, and the glacial
deposits were subsequently degraded by the sublimation of internal ice and modification by aeolian erosion. In some regions
debris seems sufficiently thick to protect a portion of underlying ice from sublimation (Fig. 7). Locally, sublimation of
debris-rich ice likely produced the hummocky terrain at the
distal margins of several lobes. The low-lying chaotic material
at the head of individual lobes and at the base of the basal escarpment formed from sublimation of relatively clean (debris
poor) ice in the accumulation zone. The open, curvilinear depressions at the base of the alcoves represent the remains of the
very ice-rich accumulation zones. No evidence has been found
for either groundwater contributions to the accumulation zones
Fig. 16. Proposed sequence of the lobate debris-covered glaciers. Phase 1: Snow, ice, and debris accumulate in the alcoves within the basal escarpment. Phase 2:
Enough material builds up to allow debris-covered glaciers to move across the surrounding plains surface. Several episodes of glacial advance and retreat may occur.
Phase 3: At some point during the history of glaciation, lava flows cascade off the Olympus Mons flank and interact locally with some glaciers. Phase 4: Following
climate change, unprotected and thinly buried ice sublimates away, leaving behind the assemblage of facies observed today.
Olympus Mons Amazonian debris-covered glaciers
or for basal melting, strongly suggesting conditions of a thick
cryosphere and cold-based glacial conditions. Basilevsky et al.
(2005) have described evidence that ice and some small glaciers
may have covered portions of flanks of Olympus Mons above
the scarp, and that some ice may still be present, covered by
sublimation till. This interpretation suggests that snow and ice
accumulation may have been more widespread and not limited
to the alcoves at the base of the scarp.
4. Discussion
In addition to the lobate deposits at the base of the scarp
on the northwest flanks of Olympus Mons, Lucchitta (1981)
also identified glacier-like features along the NW flanks of the
nearby Tharsis Montes. Recent analysis of MOLA topography
data confirms this result and concludes that cold-based glaciers
once occupied the west–northwest flanks of Arsia Mons (Head
and Marchant, 2003), Pavonis Mons (Shean et al., 2005), and
Ascraeus Mons (Parsons and Head, 2004). The glacial features found around the Tharsis Montes are far more extensive
than those at Olympus Mons; the Arsia Mons deposit covers
∼180,000 km2 (Head and Marchant, 2003), the Pavonis Mons
deposit covers ∼75,000 km2 (Shean et al., 2005), and the Ascraeus Mons covers ∼14,000 km2 (Parsons and Head, 2004)
compared to ∼15,000 km2 for the total Olympus Mons lobate deposits. Each of the fan shaped deposits on the Tharsis
Montes contain three characteristic facies. From distal to proximal regions, these include a ridged facies, interpreted to be
distal drop moraines resulting from lateral glacial advance and
retreat; a knobby facies, interpreted to be sublimation till from
vertical downwasting of the glaciers; and an interior smooth facies, characterized by broad lobes hundreds of meters thick with
a series of concentric ridges tens of meters high on their surfaces, and interpreted to be debris-covered alpine-like glaciers.
Collectively, these features are interpreted to be formed by coldbased glaciers that advanced westward out onto the surrounding
plains of the Tharsis Montes, with the smooth facies comprised of debris-covered glaciers from the most recent phases of
glaciation (Head and Marchant, 2003; Parsons and Head, 2004;
Shean et al., 2005). Thus, the smooth facies at the Tharsis
Montes is most analogous to the lobate deposits on Olympus
Mons. Apparently, a greater quantity of ice was present during
glacial periods at the Tharsis Montes than at Olympus Mons.
There is growing evidence for the existence of martian ice
ages, periods when water ice moved away from the poles and
towards the equator (e.g., Head et al., 2003). Models of volatile
behavior at high obliquity have demonstrated that H2 O ice is
more stable toward the equator than at the poles (Jakosky and
Carr, 1985; Mellon and Jakosky, 1993). Climate models including water vapor in the atmosphere and run at obliquities
of 45◦ , such as those by Richardson and Wilson (2002), indicate
that H2 O ice will sublimate from the polar caps and be transported towards the equatorial regions under these conditions.
Geological evidence for ice-rich deposits at a range of latitudes around the planet (e.g., Milliken et al., 2003; Head et al.,
2005) combined with the identification of a possible lag deposit
within the polar layered deposits of the north cap (Milkovich
405
and Head, 2005) support the hypothesis that ice has been transported from the polar regions to lower latitudes during periods
of climate change in recent geological history. What are the
ages of these deposits? Previous geological mapping studies assigned them to the Amazonian Period (Scott and Tanaka, 1986;
Morris and Tanaka, 1994). Neukum et al. (2004) recently performed crater counts on Mars Express High-Resolution Stereo
Camera (HRSC) images of the lobate deposits and found a
range of surface exposure ages for individual lobes. Major lobes
date to 130–280 Ma while some lobes are 20–60 Ma (Neukum
et al., 2004) and others may be as young as 4 Ma (e.g., Head et
al., 2005). These data thus indicate a range of ages and strongly
suggest several episodes of glacial deposit formation in recent
(Late Amazonian) history, consistent with our interpretation of
the morphological characteristics of the deposit and evidence
for multiple, overlapping lobe formation.
Under what conditions might these deposits have formed?
General circulation models under various orbital conditions
on Mars indicate that at obliquities of ∼45◦ , water is transported from the north pole to the equatorial regions and can be
deposited there (e.g., Richardson and Wilson, 2002; Mischna
et al., 2003; Levrard et al., 2004). Furthermore, Forget et al.
(2005) have completed climate simulations with a numerical
model designed to simulate the current water cycle, at an obliquity of 45◦ , and they predict accumulation of water ice specifically on the northwest flanks of Olympus Mons and the Tharsis
Montes. Recent analysis of martian obliquity implies that obliquity has exceeded 45◦ multiple times in the last 20 Myr and
may have frequently done so in the last 250 Myr (Laskar et al.,
2004). Therefore, climate and orbital parameter models and calculations predict that it is physically plausible to deposit snow
and ice to form glacial features on the flanks of Olympus Mons
episodically in the last several hundred million years.
5. Summary and conclusions
In summary, under current climatic conditions on Mars, significant water ice deposition is not favored on the surface in
equatorial and mid-latitude regions. Several workers, however,
have found evidence for glacial-like deposits at these latitudes,
interpreted to represent previous climatic conditions in which
the deposition and accumulation of ice were favored. We have
used MGS and Odyssey data to characterize the extensive lobate deposits along the northwest flank of Olympus Mons at and
adjacent to the bounding scarp as follows: (1) At least twelve
individual lobes extend from the scarp base for distances ranging from 15 to 140 km and averaging ∼45 km, and have widths
typically 10–30 km but up to ∼80 km. (2) These lobes form
a deposit that extends along an arcuate segment of the western Olympus Mons scarp for over 340 km, from an azimuth of
N 30◦ W to S 70◦ W from the summit caldera center, defining
an angular segment of 80◦ centered on N 70◦ W, and covering an area of ∼15,000 km2 . (3) The individual lobes typically
consist of five facies and other associated features, including,
from distal to proximal in relation to the scarp: (a) an outer
bounding ridge; (b) a hummocky zone; (c) a series of arcuate
ridges convex outward from the scarp and generally parallel
406
S.M. Milkovich et al. / Icarus 181 (2006) 388–407
to the lobe boundaries; (d) a chaotic terrain, usually within a
topographic low, at the core of the lobe adjacent to the scarp;
(e) an arcuate, theater-like alcove forming a portion of the major Olympus Mons scarp; (f) the scarp itself, which can consist
of talus slopes (lower) and exposed layered outcrops (upper)
in the scarp wall, or lobate flow-like features mantling the
scarp to varying degrees; (g) a blocky unit characterized by
oriented knobs that are arrayed in a linear, though radial pattern outward from the base of the scarp; and (h) the lip of the
scarp, which has variable morphology including sharp-crested
and gentle. (4) Topographic profiles show that these lobes are
thin distally, thicken toward the base of the scarp, often have
concave-downward profiles and depressions at the base of the
scarp, and reach as much as 800–1200 m in thickness. The
distal marginal ridges are typically less than 100 m in height,
but can be up to 125 m. (5) Low thermal inertia and high
albedo data suggest that the surfaces of these deposits are covered by fine-grained dust (e.g., Putzig et al., 2005). (6) Very
high-resolution MOC images show that the deposits are currently undergoing aeolian modification and erosion. (7) Counts
of superposed impact craters indicate that these deposits formed
in the last 300 Myr, and that they may have been locally active in the last few million years (e.g., Neukum et al., 2004;
Head et al., 2005). The lobate deposits give a range of ages,
suggesting that they formed episodically over an extended period of time.
We considered several possible origins for these deposits:
landslides, lava flows, pyroclastic flows and glaciers. Detailed
assessment of each possible origin and comparison to terrestrial
analogs reveals strong similarities to debris-covered piedmont
glaciers on Earth. On the basis of the characteristics discussed
above, we conclude that the lobate deposits were emplaced as
debris-covered glaciers from snow and ice accumulating on the
western margins of Olympus Mons. From distal to proximal facies, the outer ridge is interpreted to be a terminal moraine; the
hummocky facies to be marginal ice-rich portions of the glacier
that have undergone sublimation and collapse; the arcuate-ridge
facies to be drift ridges typical of those that have formed over
flowing ice; the chaotic facies as the collapsed core of glacier
ice via sublimation; the talus cones represent mass wasting in
recent times subsequent to the period of glacier formation. On
the basis of the lack of meltwater-related and basal scour features, we interpret these deposits to be formed by cold-based
glaciation. Superposed marginal ridges are interpreted as lateral
moraines and suggest that there were several phases of advance
and retreat as ice stability zones fluctuated with changing martian climate. Readvance of glaciers occurred in some places
along the Olympus Mons escarpment and not in others, perhaps
due to local microclimate conditions. Another possible explanation is that glacier advance occurred along the entire region
of the basal escarpment but that some glaciers had insufficient
debris cover to preserve the glacial record.
Taken together, these characteristics suggest that during periods in the Late Amazonian, snow and ice accumulated in the
vicinity of the western flanks of Olympus Mons, centered on
alcoves in the marginal scarp, and flowed outward, downslope
into adjacent terrains over distances as great as 140 km. These
glaciers were debris-covered, primarily by talus eroded from
the scarp outcrops and deposited on the accumulating ice, and
transported distally to form a sublimation till. The proximal pitted terrain (chaotic facies) and related depressions are located
at the base of the theaters in a region that plausibly corresponds
to former accumulation zones. The present negative topography
and morphology of this facies strongly suggest that sublimation
has dominated over accumulation in the recent geological past
and that ice, if present, is no longer active. Relationships along
the upper parts of the scarp, the scarp itself, and the flanks of
the volcano suggest that (1) ice and snow accumulated on the
flanks of the volcano above the scarp in several places, stopping
and diverting flank lava flows, (2) in some cases, lava flows extended over the margins and flowed over the glacial deposits,
and (3) in some cases glacial/volcanic interactions may have
formed significant edifice rim topography (e.g., Basilevsky et
al., 2005).
Recent general circulation models for conditions unlike today (obliquity at 45◦ and water vapor in the atmosphere) predict
snow and ice accumulation at Olympus Mons and the Tharsis
Montes (e.g., Forget et al., 2005). This is consistent with our
interpretation that these deposits date from times earlier in the
Amazonian when obliquity was higher, and polar ice was mobilized and redeposited in equatorial regions. Some small lobes
near the base of the scarp have concentric structures and may
be evidence for local debris-covered glaciers in more recent
periods (Head et al., 2005). The thickness of the large lobate
deposits strongly suggests that very ancient ice currently lies
buried below the surface in these regions. Radar sounding instruments on Mars Express and Mars Reconnaissance Orbiter
may be able to detect such near-surface buried ice. The possible
presence of similar ice-rich deposits on the flanks of Olympus
Mons earlier in the Amazonian may have been a major factor
in initiating the large-scale slope failures that formed the more
extensive Olympus Mons aureole deposits.
Acknowledgments
The authors thank David Shean, Marshall Agnew, Jay Dickson, and Rebecca Parsons for helpful discussions and assistance. Constructive reviews by James Zimbelman and Baerbel
Lucchitta improved the manuscript. Thanks are extended to
Anne Côté and Peter Neivert for assistance in manuscript preparation. J.W.H. and S.M.M. were partially supported by a NASA
Mars Data Analysis Program Grant NNG04GJ99G, which is
gratefully acknowledged.
References
Ablay, G., Hürlimann, M., 2000. Evolution of the north flank of Tenerife by
recurrent giant landslides. J. Volc. Geotherm. Res. 103, 135–159.
Ackert Jr., R.P., 1998. A rock glacier/debris-covered glacier system at Galena
Creek, Absaroka Mountains, Wyoming. Geogr. Ann. A 80, 267–276.
Adamson, D.A., Colhoun, E.A., 1992. Late Quaternary glaciation and deglaciation of the Bunger Hills, Antarctica. Antarctic Sci. 4, 435–446.
Basilevsky, A.T., and 10 colleagues, 2005. Morphology and geological structure of the western part of the Olympus Mons volcano on Mars from the
analysis of the Mars Express HRSC imagery. Solar Syst. Res. 39 (2), 85–
101.
Olympus Mons Amazonian debris-covered glaciers
Benn, D.I., Evans, D.J.A., 1998. Glaciers and Glaciation. Arnold, London, UK.
Carr, M.H., Greeley, R., Blasius, K.R., Guest, J.E., Murray, J.B., 1977. Some
martian volcanic features as viewed from the Viking Orbiters. J. Geophys.
Res. 82, 3985–4015.
Forget, F., Haberle, R.M., Montmessin, F., Levrard, B., Head, J.W., 2005. Formation of tropical to mid-latitude glaciers on Mars by atmospheric precipitation at high obliquity. Science. In press.
Francis, P.W., Wadge, G., 1983. The Olympus Mons aureole: Formation by
gravitational spreading. J. Geophys. Res. 88, 8333–8344.
Freundt, A., Wilson, C.N.J., Carey, S.N., 2000. Ignimbrites and block-and-ash
flow deposits. In: Sigurdsson, H., Houghton, B., McNutt, S.R., Rymer, H.,
Stix, J. (Eds.), Encyclopedia of Volcanoes. Academic Press, New York,
pp. 581–599.
Head, J.W., Marchant, D.R., 2003. Cold-based mountain glaciers on Mars:
Western Arsia Mons. Geology 31, 641–644.
Head, J.W., Mustard, J.F., Kreslavsky, M.A., Milliken, R.E., Marchant, D.R.,
2003. Recent ice ages on Mars. Nature 426, 797–802.
Head, J.W., and 13 colleagues, 2005. Tropical to mid-latitude snow and ice
accumulation, flow and glaciation on Mars. Nature 434, 346–351.
Helbert, J., Benkhoff, J., 2005. Beyond the equilibrium paradigm—Glacial deposits in the equatorial region of Mars. Lunar Planet. Sci. XXXVI. Abstract
1352.
Jakosky, B.M., Carr, M.H., 1985. Possible precipitation of ice at low latitudes
of Mars during periods of high obliquity. Nature 315, 559–561.
Kääb, A., Weber, M., 2004. Development of transverse ridges on rock glaciers:
Field measurements and laboratory experiments. Permafrost Periglac.
Process. 15, 379–391.
Kilburn, C.R.J., 2000. Lava flow and flow fields. In: Sigurdsson, H., Houghton,
B., McNutt, S.R., Rymer, H., Stix, J. (Eds.), Encyclopedia of Volcanoes.
Academic Press, New York, pp. 291–305.
Kleman, J., 1994. Preservation of landforms under ice sheets and ice caps. Geomorphology 9, 19–32.
Laberg, J.S., Vorren, T.O., 2000. The Trænadjupet Slide, offshore Norway—
Morphology, evacuation and triggering mechanisms. Mar. Geol. 171, 95–
114.
Larsen, G., 2002. A brief overview of eruptions from ice-covered and icecapped volcanic systems in Iceland during the past 11 centuries: Frequency,
periodicity and implications. In: Smellie, J.L., Chapman, M.G. (Eds.), Volcano–Ice Interactions on Earth and Mars. Cromwell Press, Trowbridge, UK,
pp. 81–90. Geological Society Special Publication.
Laskar, J., Correia, A.C.M., Gastineau, M., Joutel, F., Levrard, B., Robutel, P.,
2004. Long-term evolution and chaotic diffusion of the insolation quantities
of Mars. Icarus 170, 343–364.
Levrard, B., Forget, F., Montmessin, F., Laskar, F., 2004. Recent ice-rich deposits formed at high latitudes on Mars by sublimation of unstable equatorial ice during low obliquity. Nature 431, 1072–1075.
Lopes, R., Guest, J.E., Hiller, K., Neukum, G., 1982. Further evidence for a
mass movement origin of the Olympus Mons aureole. J. Geophys. Res. 87,
9917–9928.
Lucchitta, B.K., 1978. A large landslide on Mars. Geol. Soc. Am. Bull. 89 (11),
1601–1609.
Lucchitta, B.K., 1981. Mars and Earth: Comparison of cold-climate features.
Icarus 45, 264–303.
Lucchitta, B.K., Clow, G.D., Geissler, P.E., Singer, R.B., Schultz, R.A.,
Squyres, S.W., 1992. The canyon system on Mars. In: Kieffer, H.H.,
Jakosky, B.M., Snyder, C.W., Mathews, M.S. (Eds.), Mars. Univ. of Arizona Press, Tucson, pp. 453–492.
Marchant, D.R., Lewis, A.R., Phillips, W.M., Moore, E.J., Souchez, R.A.,
Denton, G.H., Sugden, D.E., Potter, N., Landis, G.P., 2002. Formation of
407
patterned ground and sublimation till over Miocene glacier ice in Beacon
Valley, southern Victoria Land, Antarctica. Geol. Soc. Am. Bull. 114 (6),
718–730.
Martin, H.E., Whalley, W.B., 1987. Rock glaciers. I. Rock glacier morphology:
Classification and distribution. Prog. Phys. Geogr. 11, 260–282.
Mellon, M.T., Jakosky, B.M., 1993. Geographic variations in the thermal and
diffusive stability of ground ice on Mars. J. Geophys. Res. 98, 3345–3364.
Milkovich, S.M., Head, J.W., 2005. North polar cap of Mars: Polar layered
deposit characterization and identification of a fundamental climate signal.
J. Geophys. Res. 110 (E1), doi:10.1029/2004JE002349.
Milliken, R.E., Mustard, J.F., Goldsby, D.L., 2003. Viscous flow features on the
surface of Mars: Observations from high-resolution Mars Orbiter Camera
(MOC) images. J. Geophys. Res. 108, doi:10.1029/2002JE002005.
Mischna, M.A., Richardson, M.I., Wilson, R.J., McCleese, D.J., 2003. On
the orbital forcing of martian water and CO2 cycles: A general circulation model study with simplified volatile schemes. J. Geophys. Res. 108,
doi:10.1029/2003JE002051.
Moon, V., Simpson, C.J., 2002. Large-scale mass wasting in ancient volcanic
materials. Eng. Geol. 64, 41–64.
Moore, J.G., Clague, D.A., Holcomb, R.T., Lipman, P.W., Normark, W.R., Torresan, M., 1989. Prodigious submarine landslides on the Hawaiian ridge.
J. Geophys. Res. 94, 17465–17484.
Morris, E.C., 1982. Aureole deposits of the martian volcano Olympus Mons.
J. Geophys. Res. 87, 1164–1178.
Morris, E.C., Tanaka, K.L., 1994. Geologic maps of the Olympus Mons region
of Mars. U.S. Geol. Surv. Misc. Invest. Ser. Map I-2327.
Mouginis-Mark, P.J., Wilson, L., Zuber, M.T., 1992. The physical volcanology
of Mars. In: Kieffer, H.H., Jakosky, B.M., Snyder, C.W., Mathews, M.S.
(Eds.), Mars. Univ. of Arizona Press, Tucson, pp. 424–452.
Neukum, G., and 11 colleagues, 2004. Recent and episodic volcanic and glacial
activity on Mars revealed by the High-Resolution Stereo Camera. Nature 432, 971–979.
Parsons, R.L., Head, J.W., 2004. Ascraeus Mons, Mars: Characterization and
interpretation of the fan-shaped deposit on its western flank. Lunar Planet.
Sci. XXXV. Abstract 1776.
Parsons, R.L., Head, J.W., 2005. Ascraeus Mons fan-shaped deposit, Mars: Geological history and volcano–ice interactions of a cold-based glacier. Lunar
Planet. Sci. XXXVI. Abstract 1139.
Putzig, N.E., Mellon, M.T., Kretke, K.A., Arvidson, R.E., 2005. Global thermal inertia and surface properties of Mars from the MGS mapping mission.
Icarus 173, 325–341.
Richardson, M.I., Wilson, R.J., 2002. Investigation of the nature and stability
of the martian seasonal water cycle with a general circulation model. J.
Geophys. Res. 107 (E5), doi:10.1029/2001JE001536.
Scott, D.H., Tanaka, K.L., 1986. Geologic map of the western equatorial region
of Mars. U.S. Geol. Surv. Misc. Invest. Ser. Map I-1802-A.
Shean, D.E., Head, J.W., Marchant, D.R., 2005. Origin and evolution of a coldbased tropical mountain glacier on Mars: The Pavonis Mons fan-shaped
deposit. J. Geophys. Res. 110 (E5), doi:10.1029/2004JE002360.
Tanaka, K.L., 1985. Ice-lubricated gravity spreading of the Olympus Mons aureole deposits. Icarus 62, 191–206.
Waller, R.I., 2001. The influence of basal processes on the dynamic behavior of
cold-based glaciers. Quatern. Int. 86, 117–128.
Whalley, W.B., Azizi, F., 2003. Rock glaciers and protalus landforms: Analogous forms and ice sources on Earth and Mars. J. Geophys. Res. 108,
doi:10.1029/2002JE001864.
Whalley, W.B., Martin, H.E., 1992. Rock glaciers. II. Models and mechanisms.
Prog. Phys. Geogr. 16, 127–186.