Icarus The Ascraeus Mons fan-shaped deposit: Volcano–ice interactions and the climatic

advertisement
Icarus 197 (2008) 84–109
Contents lists available at ScienceDirect
Icarus
www.elsevier.com/locate/icarus
The Ascraeus Mons fan-shaped deposit: Volcano–ice interactions and the climatic
implications of cold-based tropical mountain glaciation
Seth J. Kadish a , James W. Head a,∗ , Rebecca L. Parsons b , David R. Marchant c
a
b
c
Department of Geological Sciences, Brown University, Providence, RI 02912, USA
Department of Geosciences, Stony Brook University, New York, NY 11794, USA
Department of Earth Sciences, Boston University, Boston, MA 02215, USA
a r t i c l e
i n f o
a b s t r a c t
Article history:
Received 24 October 2007
Revised 13 February 2008
Available online 7 April 2008
Amazonian-aged fan-shaped deposits extending to the northwest of each of the Tharsis Montes in the
Tharsis region on Mars have been interpreted to have originated from mass-wasting, volcanic, tectonic
and/or glacial processes. We use new data from MRO, MGS, and Odyssey to characterize these deposits.
Building on recent evidence for cold-based glacial activity at Pavonis Mons and Arsia Mons, we interpret
the smaller Ascraeus fan-shaped deposit to be of glacial origin. Our geomorphological assessment reveals
a number of characteristics indicative of glacial growth and retreat, including: (1) a ridged facies,
interpreted to be composed of drop moraines emplaced during episodic glacial advance and retreat,
(2) a knobby facies, interpreted to represent vertical downwasting of the ice sheet, and (3) complex
ridges showing a cusp-like structure. We also see evidence of volcano–ice interactions in the form of:
(1) an arcuate inward-facing scarp, interpreted to have formed by the chilling of lava flows against the
glacial margin, (2) a plateau feature, interpreted to represent a subglacial eruption, and (3) knobby facies
superimposed on flat-topped flows with leveed channels, interpreted to be subglacial inflated lava flows
that subsequently drained and are covered by glacial till. We discuss the formation mechanisms of these
morphologies during cold-based glacial activity and concurrent volcanism. On the basis of a Mid- to
Late-Amazonian age (250–380 Ma) established from crater size-frequency distribution data, we explore
the climatic implications of recent glaciation at low latitudes on Mars. GCM results show that increased
insolation to the poles at high obliquities (>45◦ ) forces sublimation of polar ice, which is transported
to lower latitudes and deposited on the flanks of the Tharsis Montes. We assess how local orographic
effects, the mass balance of the glacier, and the position of equilibrium line altitudes, all played a role
in producing the observed geomorphologies. In doing so, we outline a glacial history for the evolution of
the Ascraeus Mons fan-shaped deposit and compare its initiation, growth and demise with those of Arsia
Mons and Pavonis Mons.
© 2008 Elsevier Inc. All rights reserved.
Keywords:
Ices
Volcanism
Mars, climate
Geological processes
1. Introduction
Ascraeus Mons is the northernmost of the three Tharsis Montes
shield volcanoes (Arsia Mons, Pavonis Mons, and Ascraeus Mons),
which are aligned along a N40◦ E trend on the crest of the broad
Tharsis Rise (Fig. 1). Previous studies have included investigations into their caldera morphologies as indicators of magmatic
styles and activity (Crumpler et al., 1991, 1996; Scott and Wilson,
2000), investigations into the evolutionary history of the volcanoes based on analyses of structural and morphologic features on
and around them (Carr et al., 1977; Crumpler and Aubele, 1978;
Mouginis-Mark et al., 1982; Zimbelman and Edgett, 1992), and
combined morphologic studies and crater counting to improve understanding of the overall stratigraphies of the volcanoes and sur-
*
Corresponding author. Fax: +1 401 863 3978.
E-mail address: James_Head@brown.edu (J.W. Head).
0019-1035/$ – see front matter
doi:10.1016/j.icarus.2008.03.019
©
2008 Elsevier Inc. All rights reserved.
rounding terrains (Crumpler and Aubele, 1978; Scott and Tanaka,
1981). Such studies have shown that Arsia, Pavonis and Ascraeus
Mons exhibit similar structural histories, but also display differences, which imply that each records important variations on the
overall theme of volcano evolution.
One particularly interesting commonality between the three
Tharsis Montes shield volcanoes is the occurrence of a distinctive
and unusual fan-shaped deposit (FSD) extending approximately
northwest on the western flank of each volcano (Fig. 1). On the
basis of the density of superposed impact craters and stratigraphic
relationships, these deposits are thought to be among the youngest
in the region, forming during the Late Amazonian, concurrent with
late-stage minor volcanism, most likely in the form of fissure eruptions on the volcano flanks (Scott and Tanaka, 1986). Three major
facies are generally contained within the FSDs (Carr et al., 1977;
Zimbelman and Edgett, 1992; Scott and Zimbelman, 1995; Scott
et al., 1998; Head and Marchant, 2003; Shean et al., 2005): (1) a
Ascraeus Mons tropical mountain glaciation
Fig. 1. A shaded relief map of the Tharsis Montes overlain by MOLA topography
data from Shean et al. (2005). The fan-shaped deposits, which extend to the westnorthwest of each of the Tharsis Montes, are outlined in white. Note the distinctly
smaller size of the Ascraeus Mons fan-shaped deposit.
ridged facies consisting of numerous ridges that trace the distal
margin of the lobe, commonly superposed on lava flow features
without any obvious deflection, (2) a knobby facies consisting of
a chaotic assemblage of similarly-sized knobs, and (3) a smooth
facies consisting of concentric ridges superposed on broad lobes
(Fig. 2).
For some years there has been a debate over the emplacement
of the FSDs associated with each of the Tharsis Montes. Early ob-
85
servations of these features from Mariner 9 and Viking Orbiter
images led to the proposal of various interpretations for their origin including gravity-driven sliding (Carr, 1975; Carr et al., 1977;
Scott and Tanaka, 1981; Edgett, 1989; Zimbelman and Edgett, 1992;
Edgett et al., 1997), glaciation (Williams, 1978; Lucchitta, 1981;
Helgason, 1999), a combination of catastrophic sliding and subsequent pyroclastic activity (Zimbelman and Edgett, 1992; Edgett
et al., 1997), and a combination of sliding, volcanism and groundice activity (Scott and Zimbelman, 1995). More recently, various
authors (Head and Marchant, 2003; Milkovich and Head, 2003;
Shean et al., 2005) have used MOLA data combined with MOC and
THEMIS images to build on early comparisons of the FSDs with
terrestrial glacial deposits, such as the Malaspina glacier in southeastern Alaska (Lucchitta, 1981). These later studies have used depositional frameworks of polar glaciers in the Mars-like Antarctic
Dry Valleys to demonstrate the consistency of the FSDs with coldbased glacial deposits (Marchant and Head, 2007).
This study focuses on the Ascraeus Mons FSD, which is centered
at 12◦ N and 108◦ W within the Tharsis province. In their study of
the Tharsis Montes, Zimbelman and Edgett (1992) identified a significantly smaller deposit to the west-northwest of Ascraeus Mons
(Fig. 3) that contains both ridged and knobby facies, but appears to
lack any unit comparable to the smooth facies observed within the
other Tharsis Montes deposits. The report of a smaller feature containing fewer interior deposits at Ascraeus Mons illustrates that,
despite morphologic, relative stratigraphic, and geographic similarities between the FSDs that suggest a similar formation mechanism, each individual deposit displays unique complexities; these
differences must be addressed in order to understand fully the origin of these features. Our goal is a re-examination of the Ascraeus
Mons FSD using higher resolution data than have been previously
available in order to assess the plausibility of an origin from coldbased glaciation.
Fig. 2. The three predominant facies of the fan-shaped deposits with possible terrestrial analogs from Head and Marchant (2003). The facies are depicted by Viking Orbiter
data of the Arsia Mons deposit: A—ridged facies; B—knobby facies; C—smooth facies. Terrestrial images of the Antarctic Dry Valleys in D–F are selections from US Geological
Survey aerial photographs taken on 21 November, 1993. These are: D—drop moraines in Arena Valley (serial TMA 3079/303); E—sublimation till in Beacon Valley (serial TMA
3078/006); F—Debris-covered rock glacier in Mullins Valley (serial TMA 3080/275).
86
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Fig. 3. THEMIS IR (100 m/pixel) mosaic of the Ascraeus Mons fan-shaped deposit.
The deposit is situated between subaerial lava flows to the west and the Ascraeus
Mons flank to the east. Note the position of the ridges that define the distal rim of
the deposit, the arcuate east-facing scarp with curvature similar to the ridges, the
widespread distribution of the knobby facies, and the absence of a smooth facies.
2. Data and methodology
Photogeologic interpretation of the Ascraeus Mons FSD was performed using Mars Orbital Camera (MOC) (Malin et al., 1998),
High Resolution Stereo Camera (HRSC) (Neukum et al., 2004), Thermal Emission Imaging Spectrometer (THEMIS) (Christensen et al.,
2003) and High Resolution Imaging Science Experiment (HiRISE)
(McEwen et al., 2007) data. Quantitative geomorphological assessments were made using topographic analyses of Mars Orbital Laser
Altimeter (MOLA) (Smith et al., 2001) data in conjunction with the
images, allowing for the comparison not only of the geographic
extent of certain facies and features, but also their profiles and
elevations. Conclusions regarding the evolution of the glacial deposit and the relative sequence of events were based on (1) stratigraphic relationships, (2) quantitative geomorphology, (3) previous work on climate and glacial models and interpretations (Head
and Marchant, 2003; Shean et al., 2005, 2007; Milkovich et al.,
2006; Fastook et al., 2005, 2006; Forget et al., 2006), (4) terrestrial
analogs (Marchant and Head, 2007), and (5) crater size-frequency
distribution analyses performed on THEMIS VIS and IR mosaics.
3. Geomorphology and interpretation of the fan-shaped deposit
3.1. Regional setting and general description
The Tharsis province is a vast region of tectonism and volcanism extending approximately 5000 km across the martian surface
from ∼220◦ to 300◦ E and from ∼50◦ S to 60◦ N. Tharsis likely
formed by a combination of structural uplift and volcanism resulting from prolonged and extensive mantle upwelling (e.g. Head and
Fig. 4. Perspective 3D depictions of the Ascraeus Mons fan-shaped deposit, using
THEMIS VIS and IR images. The viewing geometries of the four images are shown
in Fig. 10. These reveal the regional and local topography, most notably the downslope trend from south to north (right to left in A). The textures of the facies and
features are easily observed, including the distinct roughness of the knobby facies
and smoothness of the degraded flank unit. The vertical exaggeration is 5×.
Solomon, 1981; Smith et al., 1999; Zuber et al., 2000). The highest portion of the rise is dominated by the Tharsis Montes—Arsia
Mons, Pavonis Mons, and Ascraeus Mons—a series of three young
shield volcanoes (Fig. 1). These volcanoes are aligned at approximately equidistant intervals of 800 km and are thought to be
genetically linked based on their morphologic and structural similarities as well as their spatial association (Crumpler and Aubele,
1978).
The Ascraeus FSD extends up to 100 km west of the base of the
shield along a N82◦ W trend (Figs. 1, 3). The outer arc of the fan
is concentric to the volcano base. The width of the deposit, measured along the base of the shield, is approximately 190 km and
it covers an area of ∼14,000 km2 , equivalent to 5% of the total
area of the Tharsis Montes FSDs (Fig. 1). The Ascraeus FSD attains
elevations of >4.5 km above Mars datum in its southern portion
(Figs. 5a, 7, 8; see Fig. 6 for profile tracks in Figs. 7, 8). Overall,
elevations show a gradual decline from the southern to the northern edge of the FSD (Figs. 5a, 8). The lowest observed elevations
of around 1.2 km above Mars datum occur where the deposit is
draped over lobate-fronted lava flows emplaced within a large to-
Ascraeus Mons tropical mountain glaciation
(a)
87
(b)
Fig. 5. (a) Topography and slope maps of Ascraeus Mons and its fan-shaped deposit. The topographic maps show MOLA data over a THEMIS IR mosaic, with 1000 m contours
on the image of Ascraeus Mons and 500 m contours on the image of the deposit. The slope maps, derived from MOLA data, reveal the extremely flat context surrounding
the fan-shaped deposit. These highlight the steepness of the arcuate scarp, the plateau feature, and the margins of the flat-topped ridges. Note also that the degraded flank
unit—a unit which has a slope of ∼5◦ bordering a >10◦ sloped escarpment on the side of the volcano—is unique to the western perimeter of Ascraeus Mons. (b) A hillshade
representation, derived from MOLA data, of Ascraeus Mons with the fan-shaped deposit extending to the west. Illumination is from the north (top of the image).
pographic depression approximately 50 km west of the shield base
of Ascraeus Mons (Figs. 8N, 8O). This lower section is separated
from the distal margins of the deposit by an arcuate scarp with a
relief of up to 300 m (Figs. 5, 7D–7G, 8K). The southern end of this
scarp appears to have been disrupted by a lobate flow that can be
traced from a source near the southwest flank of Ascraeus Mons
(visible in the bottom of Fig. 3, right side of Fig. 4A, top of Fig. 4B,
bottom of Fig. 4C; see Fig. 10 for context of images in Figs. 4, 11,
12, 16, 17). This flow appears to follow the rim of the topographic
depression rather than flow into it as did earlier lava flows (visible in the bottom of Fig. 4A, right side of Fig. 4B). The majority of
the area within the depression is extremely flat, with most slopes
under 0.6◦ (Fig. 5).
3.2. Ridged facies
The ridged facies (Figs. 11A, 12) consists of at least seven nearcontinuous, concentric ridges that trace the distal margin of the
FSD over a distance of ∼185 km (Fig. 9, red lines). These ridges
extend approximately parallel to the outer scarp, which is located
between ∼4 and 20 km to the east of the innermost ridge. Although the ridges are generally parallel to each other, they frequently converge (Figs. 11A, 12A, 12H), and appear to cross in some
cases (Figs. 12A, 12E). Detailed inspection of the ridged facies using MOC and THEMIS VIS images also indicates some degree of
variation in the morphology of individual ridges, generally ranging
from linear to beaded (Figs. 12D–12H). MOLA data reveal general
ridge heights ranging between 10 m and 40 m (Figs. 13, 14), with
the exception of the outermost ridge (Fig. 12G), which is the most
prominent, reaching heights of ∼80 m (Fig. 14) and extending for
the entire length of the facies. MOLA data also show the ridged facies to be located at typical elevations of between 1500 m above
Mars datum to the north and 2200 m above Mars datum to the
south (Fig. 5a).
Two distinct characteristics of the ridged facies are the variations in the frequency of ridges—the number of ridges across the
FSD for a given azimuth—and the spacing intervals between individual ridges across the deposit. The frequency of the ridges is
highest in the central part of the facies, where up to 7 ridges are
observed. In this central region, most ridges can be traced over a
lateral distance of ∼50 km. Laterally, these ridges either end or
gradually merge as they trend toward the north and south ends
of the deposit. Ridges eventually converge to form two and three
distinct ridges to the north and south respectively. THEMIS and
Viking images show that, to the north, the ridges diminish gradually as they curve to the northeast, ending ∼40 km west of the
break in slope at the base of Ascraeus Mons (Figs. 9, 12H). In contrast, the southern part of the ridged facies ends more abruptly
∼80 km west of the Ascraeus Mons scarp (Figs. 9, 12A). Although
there are fewer ridges at the northern and southern ends of the
ridged facies, the ridges are consistently wider within these regions (>1200 m) than near the most distal portion of the facies
(<800 m) (Fig. 14). There is no consistent spacing between the
ridges; this variation is best observed by comparing the distance
between two prominent ridges at several locations (Fig. 14). The
88
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Fig. 6. Context map for the latitudinal (A–I) and longitudinal (J–P) MOLA profiles in
Figs. 7 and 8 on a THEMIS mosaic.
separation reaches 1600 m in the central and southern parts of the
ridged facies, but decreases to less than 250 m at the northern end.
This is in contrast to the record-groove-like ridges at Arsia, which
tend to remain parallel with equal spacing (Head and Marchant,
2003; Fig. 12I).
A fundamental characteristic of the outer ridged facies at Ascraeus is the changing radius of curvature from south to north,
making it distinct from the more symmetric ridges at Pavonis and
Arsia. At Ascraeus, the ridged facies has a much smaller radius of
curvature in the south than in the north (Fig. 9). This is probably
due to the distinctive regional topography, forcing constriction of
the southern edge of the glacier as it encountered the slope, and
letting it expand in the downhill northern direction (Fig. 5).
In addition to the main outer ridged facies, smaller but morphologically similar linear ridges occur near the base of the outer
scarp (Fig. 12F). These ridges appear to be concentric to the outer
ridged facies despite being located to the east of the arcuate scarp
at an elevation of ∼250 m lower than the outermost ridge. They
are much less prominent than those constituting the distal margin of the deposit, extending only ∼25 km in length and having
heights of only a few meters. These ridges terminate at their northern end as they approach rough terrain associated with the arcuate
scarp, only to reappear briefly ∼15 km to the northeast (Fig. 9). At
least three other sets of ridges are visible in the northern half of
the deposit. One set, which is ∼55 km west of the shield base,
contains only four ridges, the longest of which is less than 10 km
in length; these ridges differ in orientation from all other ridges,
trending N22◦ W, making them almost perpendicular to the distal
ridges for that latitude (Fig. 9). On the basis of their distinct radius of curvature, this may signify that they were produced by a
distinct lobe of the glacier, a separate glaciation, or that they represent the edge of a lobate facies formed during late-stage glaciation.
Farther north, a pair of discontinuous concentric ridges runs parallel to the distal ridges, ∼10 to 15 km inside the outermost distal
ridge. A single ∼35 km long ridge is present only ∼20 km from
the Ascraeus Mons flank and is oriented ∼N–S. Although the ridge
is somewhat sinuous, it is neither concave nor convex to the deposit. On the basis of its linearity and orientation, it is possible that
this feature is not a moraine, but rather is a dike intrusion into
the glacier. Similar features have been observed at Pavonis Mons
(Shean et al., 2005). We also identified a single ridge in a HiRISE
image only ∼10 km from the shield base (Figs. 9, 12B, 12C). This
ridge has a convex curvature which mimics the shield, and is covered by dunes.
The morphology, spatial distribution and superposition relationships of the arcuate ridges support their interpretation as drop
moraines formed at the margins of cold-based glaciers during periods of glacial stability, similar to those observed at Arsia and
Pavonis Mons (Head and Marchant, 2003; Shean et al., 2005, 2007).
Drop moraines, as opposed to englacial or supraglacial moraines,
are a type of terminal or recessional moraine formed via deposition of sediment at the margin of the glacier. The observation of a prominent outermost ridge enclosing a series of less
prominent, lower relief ridges is common in many terrestrial icemarginal moraine sequences (Benn and Evans, 1998). The outermost ridge at the distal edge of the Ascraeus FSD is thus interpreted to represent the terminal moraine marking maximum ice
extent. The change—usually reduction—in ridge size inward of the
outermost ridge is explained as a function of the length of the
period of standstill during which the ridges formed, which is commonly seen in terrestrial examples (e.g. Koch and Kilian, 2005;
Benn and Evans, 1998). In this scenario, the outermost ridge developed at a stable ice margin during a period of prolonged
standstill, while subsequent ridges developed during shorter intervals of standstill that punctuated an overall period of ice recession. The overlapping and undisruptive nature of the ridges is
consistent both with the protective, rather than erosional, style
of debris deposition that is characteristic of cold-based glaciers
(Holdsworth and Bull, 1970) and with the nature of ridge deposition at Arsia and Pavonis Mons (Head and Marchant, 2003;
Shean et al., 2005). The occurrence of overlaps and crossovers
(Figs. 12A, 12E) also suggests that ice recession was punctuated by
periods of re-advance. The younger generation of ridges could represent a separate advance; if so, the ridged facies as a whole could
be the result of deposition during at least two separate glacial periods.
3.3. Arcuate scarp
A unique component of the Ascraeus FSD, not present at Pavonis or Arsia, is the presence of an east-facing arcuate scarp proximal to its distal margins (Fig. 11B; labeled in Figs. 7D–7G, 9). This
scarp extends parallel to the entire ridged facies (Fig. 9, green line).
MOLA profiles across this scarp (Figs. 7B, 7D–7G) reveal that the
ridged facies is separated from the remainder of the FSD by a relief that usually ranges between 180–300 m. This outer scarp forms
the boundary of a topographic depression adjacent to the western
shield base of Ascraeus Mons (Fig. 5), which contains the majority
of the interior deposits constituting the Ascraeus FSD. The southern portion of this scarp has been disrupted by lobate lava flows
from the south. These flows were prevented from flowing directly
north—along the line of steepest descent—due to the presence of
the extant glacier.
It has been suggested that the outer scarp contained within
the Ascraeus FSD could be either erosional (Zimbelman, 1984) or
tectonic (Zimbelman and Edgett, 1992) in origin. Escarpment formation as a result of erosion of lobate flow features that border
the south and southeast margins of the topographic depression is
Ascraeus Mons tropical mountain glaciation
89
Fig. 7. Latitudinal profiles extracted from 128 pixel/degree gridded MOLA data. See Fig. 6 for tracks of individual profiles. Vertical exaggeration is ∼30×.
based on the observation of truncated channels at the southeast
end of these lobate flow features (Zimbelman, 1984). Zimbelman
and Edgett (1992) also describe the lower elevation part of the
FSD as a “downdropped block” bounded by faults that are parallel
to the distal margin of the deposit. In this case, tectonic faulting
may have been associated with emplacement of the FSD, possibly
by landslide of the western flank material (Zimbelman and Edgett,
1992).
In an alternative explanation, we have interpreted this arcuate scarp to be the result of lava flow–glacier interactions. In this
scenario, the presence of a large glacier served as a barrier, deflecting lava flows around its margins (Fig. 15). The presence of
a glacier >300 m in height that extended ∼80 km west from
the base of the western shield of Ascraeus Mons to the position
now marked by the scarp would provide a barrier for effusive
lava flows from the flanking eruptive centers to the south. Plains-
forming lava flows originating from Ascraeus and/or Pavonis to the
south would then have traveled around the western flanks of Ascraeus, being deflected around and chilling against the ice margins.
In this model, the ridged facies must have been deposited subsequent to scarp formation when ice later overrode cooled lava
flows and advanced to a position at least ∼100 km west of the
western base of the Ascraeus shield. This model of scarp formation
by lava flow–glacier interactions is consistent with that proposed
for similar arcuate scarps (Wilson and Head, 2007a), particularly
those observed within the Pavonis Mons FSD (Shean et al., 2005;
Head and Wilson, 2007).
An additional, considerably less prominent arcuate scarp with
a relief of up to 50 m is present approximately 10 km east of the
outer scarp, and faces west (illuminated in Figs. 4A, 4B). This inner
scarp begins at the southern boundary of the topographic depression and extends parallel to a small portion of the outer scarp for
90
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Fig. 8. Longitudinal profiles extracted from 128 pixel/degree gridded MOLA data. See Fig. 6 for track of individual profiles. Vertical exaggeration is ∼30×.
a distance of ∼50 km. The coincidence of the inner scarp with the
western margin of a lobate-fronted lava flow that has embayed
the topographic depression implies that it may represent a flow
margin and has since become more pronounced as a result of deposition of material onto the lava flow surface. This scarp appears
to be most prominent in its southern reaches, where it forms a
ridge-like bank of material 30 km long and 2 km wide (Fig. 11E).
Although they do not represent a large area within the FSD,
the scarps are a key component of glacial history in the region,
providing tangible evidence for the occurrence of lava flows concurrent with an extant glacier. The largest arcuate scarp not only
delineates an instance of the distal rim of the glacier, but it also
provides constraints on the height of the ice. The asymmetry of
its radius of curvature, matching that of the outer ridges, provides further evidence for a topographically-controlled shape of the
glacier.
3.4. Knobby facies
The most spatially extensive terrain in the Ascraeus Mons FSD
is an assemblage of rounded to subrounded knobs or hummocks
(Fig. 16; labeled in Figs. 7F, 8L, 8M, 9, unit Kf), which cover a total
area of approximately 4800 km2 . These knobs typically range from
<100 to 800 m in diameter and are 10 to 60 m in height. Knobs
have simple conical shapes, although some appear to be elongated
in a northwest to north-northwest direction (Fig. 16A). Summits
are usually rounded and slope–foot boundaries appear to be grad-
ual. These boundaries are sometimes marked, particularly on the
north-facing flanks, by a rim of eolian material (Fig. 16E). Flank
slopes of the knobs are very low (<10◦ ).
The knobby facies is contained within a few distinct areas
(Fig. 9), the most extensive and continuous of which occurs in the
southern portion of the deposit. Here, the knobby facies occurs as
a near-continuous sheet. A smaller, patchier area of the knobby
facies occurs in the northern portion of the deposit. The highest
elevation observed for the knobby facies is ∼4.5 km (Figs. 5a, 9)
and occurs where the southern knobby facies is draped over lobate flow features in the central and southern parts of the deposit (Fig. 16B). Along its northwestern margin, this portion of the
knobby facies abuts a unique terrain type consisting of a chaotic
field of hills and ridge-like features (Figs. 9, unit Ch; 17). The
boundary between these two terrain types is extremely abrupt relative to other margins of the knobby facies (Figs. 17A, 17D). More
diffuse margins tend to be characterized by a gradual reduction in
the density of knobs and an increase in the spacing between them
(Fig. 16A).
Analysis of the knobby facies contained within the Arsia and
Pavonis Mons FSDs has led to the proposal of various origins for
the conical hills. Scott et al. (1998) noted that certain concentrations of the knobby facies in the northern regions of the Pavonis FSD appear elongated in plan view, which they interpreted to
be characteristic of drumlin fields or drift deposits formed during disintegration of warm basal ice. Streamlined and coalesced
groups of hills observed within the knobby facies were inter-
Ascraeus Mons tropical mountain glaciation
91
Fig. 10. Context map for figure locations on a THEMIS IR mosaic. White boxes represent the borders of the images in Figs. 11, 12, 16 and 17, although those figures
are not necessarily THEMIS IR images. The arrows indicate the viewing angle of the
3D perspectives in Fig. 4.
Fig. 9. Geologic map of the Ascraeus Mons fan-shaped deposit showing the geographic location and extent of significant units and features. The ridged facies
(maroon lines) is primarily located along the distal margin of the deposit, but sets
of ridges are observed in more central locations with various orientations and curvatures. The arcuate scarp (green line) is located just inside the outer ridged facies,
with a similar curvature. Lobate scarps (blue lines) exist in the northern half of the
deposit, showing no preferred orientation. Numerous graben (yellow lines) are emplaced on the western flank of Ascraeus Mons with a general N–S orientation. One
graben extends from Alba Patera down to the fan-shaped deposit, cutting across
the arcuate scarp to a northern region of the knobby facies. A radial ridge (pink line
with hash marks) is present in the northern half of the deposit, which we interpret
to be a subglacial dike. The knobby facies (light blue—Kf) is concentrated predominantly in the southern half of the deposit, but exists in isolated patches near the
outer ridged facies and in the northern half near the lobate scarps. The complex
hummocks (light green—Ch) border the largest region of knobby facies to the northwest. Just south of the complex hummocks are a set of raised ridges which consist
of a 2 km wide bank of material (beige—Rr). The flat-topped flows (brown—Ff) extend northward from the SW flank of Ascraeus Mons. These lobate features are
largely covered by knobs. The plateau (light red—P) in the central part of the deposit
is also partially covered by knobs. The mountainous terrain (purple—Mt) borders the
arcuate scarp to the west, and is present in two distinct sections. Knobs are superimposed on a small part of this unit as well. The degraded flank unit (orange—Df)
borders the flank of Ascraeus Mons, extending almost the entire length of the deposit from north to south. (For interpretation of the references to color in this figure
legend, the reader is referred to the web version of this article.)
preted as eskers, which are also products of wet-based or warm
ice (Scott et al., 1998). As an alternative, the knobby facies may
be aligned hummocks on a lava or debris flow, or possibly windfaceted yardangs in friable material formed by katabatic winds
descending from the summit of Pavonis Mons (Scott et al., 1998).
Conversely, Shean et al. (2005) noted an apparent lack of additional evidence of wet-based activity within the knobby facies of
the Pavonis deposit. Furthermore, they describe the knobby facies at Pavonis as an inhomogeneous distribution of hills lacking both a preferred orientation and a confinement to any one
particular unit. These observations imply that a more likely origin for the knobby facies may be from in situ down-wasting of
cold-based ice (Head and Marchant, 2003) in a process analo-
gous to sublimation till formation in terrestrial glacial environments such as the Antarctic Dry Valleys (Marchant et al., 2002;
Shean et al., 2005). From their analysis of the FSD west of Arsia
Mons, Head and Marchant (2003) describe the process of till formation as “deflation” and internal settling of the deposit as ice is
progressively removed by sublimation.
The knobby facies contained within the Ascraeus Mons FSD,
like those at Arsia and Pavonis Mons, consists of numerous randomly orientated and irregularly shaped mounds that appear to
mantle various underlying units. Furthermore, the knobby facies
also appears to continue uninterrupted over several high relief
features including lobate flow features that are characterized by
thicknesses of hundreds of meters. The deposition of this facies as
a series of mounds is a common characteristic of tills that form
from differential ablation or sublimation of debris-mantled ice.
A sublimation till forms when ice underlying a granular medium
sublimates while any englacial debris that is exposed at the new
ice surface as a result of this ice deflation contributes to a growing
supraglacial lag (e.g. Benn and Evans, 1998; Marchant et al., 2002;
Shaw, 1977). Growth and final deposition of this lag deposit as
a series of irregular mounds occurs via several cycles of topographic inversion and debris reworking. During sublimation, mass
movement causes debris to be transferred away from topographic
highs on the glacier surface, exposing ice surfaces. These newly exposed surfaces are subsequently removed by sublimation and new
depressions are created in former high points on the glacier surface.
Based on the interpretation of the knobby terrain as sublimation till, the specific position of the knobs within the FSD suggests
a strongly inhomogeneous distribution of debris in the glacier. Sublimation of the glacier left significantly more debris in the southern portion of the deposit than in the north. The position of the
lobate lava flows, extending from the southwest flank of Ascraeus,
supports the notion that the predominant source of debris is volcanic, in the form of ash or broken up lava flows.
92
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Fig. 11. A gallery of images highlighting some of the main facies and features in the Ascraeus fan-shaped deposit at THEMIS IR resolution (D, F—100 m/pixel), THEMIS VIS
resolution (A, B, C, D, H, K—18 m/pixel), and MOC narrow-angle resolution (E, G, I, J—6 m/pixel). (A) The outer ridged facies showing merging of ridges (THEMIS V11712006).
(B) A section of the east-facing arcuate scarp (THEMIS V11712006). (C) Flat-topped flows with leveed channels (THEMIS V17515015). (D) The plateau feature showing a mound
on the southeast end, covered by knobs (THEMIS VIS and IR mosaic). The largest crater in the fan-shaped deposit is in the center of the plateau. (E) The bank of material
composing the raised ridges (MOC M07-02694). (F) A lobate scarp (THEMIS IR mosaic). The black box outlines the area shown in image G. (G) A higher resolution image of
the lobate scarp (MOC E05-00055). (H) A small lobate scarp (THEMIS V14445008). (I) The degraded flank unit showing dunes and slope streaks extending down small hills
(MOC M08-06313). (J) The escarpment at the base of the Ascraeus flank, descending to the degraded flank unit (MOC M04-03254). The smooth degraded flank is visible,
covered by dunes. (K) The mountainous terrain (THEMIS V11712006). The terrain is partially covered by knobs.
3.5. Complex curved ridges and chaotic hills (hummocky-like terrain)
Immediately adjacent to the western margin of the southern
knobby terrain is an area of ∼325 km2 that contains a heterogeneous covering of ridge-like and knob-like features (Figs. 9, unit
Ch; 17). These features appear to display a gradational sequence
forming two units: (1) curved and elongated ridges running ap-
proximately parallel to a N60◦ E trend (Fig. 17C), and (2) a chaotic
assemblage of irregularly aligned elongated hills (Fig. 17D). The
elongated ridges cover the western portion of this terrain (Figs. 9,
17A). The morphology of these ridges differs substantially from
those characteristic of the ridged facies (compare Figs. 12 and 17).
They are markedly discontinuous, consisting of many distinct segments 500–2000 m long, 150–300 m wide, and 25–75 m high.
Ascraeus Mons tropical mountain glaciation
93
Fig. 12. A gallery of images of the ridged facies at THEMIS IR resolution (F—100 m/pixel), THEMIS VIS resolution (A, B, I—18 m/pixel), MOC narrow-angle resolution (D, E,
G, H—6 m/pixel), and Hi-RISE resolution (C—50 cm/pixel). (A) The outer ridged facies merging at the southern end of the deposit (THEMIS V11712006, V18763010). (B) The
degraded flank showing the location of image C (THEMIS V17203012). (C) A Hi-RISE image of a narrow discontinuous ridge covered by dunes within the degraded flank
unit (PSP_002196_1920). (D) The outer ridged facies showing parallel, non-overlapping ridges (M02-02580). (E) The same ridges at a section farther north where two of
them cross (M02-02580). (F) Narrow ridges with a beaded texture just east of the arcuate scarp (THEMIS IR mosaic). (G) A MOC image of the outermost ridge of the distal
ridged facies (MOC M09-03537). (H) The distal ridged facies merging at the northern end of the deposit (MOC E15-01295). (I) The ridged facies at Arsia Mons for comparison
(THEMIS V13336001). Note the striking difference in the number of ridges and the spacing between them.
They are irregularly spaced at <100–400 m intervals. In planform
these sections are distinctly curved or cusp-shaped, having the opposite curvature of the ridged facies. The southern ends of the
curved ridges appear to retain a N60◦ to N40◦ E trend, but the
northern ends point almost directly north, and some ridges are
randomly oriented (Fig. 17B). The summits of these ridges appear
to be relatively sharp and the slope–foot junctions are sharp relative to the transitional bases of the ridged facies. The flanks of
these ridges have gentle slopes, varying between 2◦ and 5◦ . The
transition between the complex ridges and the field of randomly
orientated and elongated hills is abrupt (Figs. 17A, 17B). These
hills are variable in size, but are comparable to knobs constituting the knobby facies. Typically, hills exhibit lengths of 50–600 m,
widths of 30–200 m, and heights of 30–50 m. These features appear to be significantly degraded relative to the complex ridges,
with rounded summits and a less consistent, more variable form
94
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Fig. 13. A THEMIS context image for the MOLA profiles shown in Fig. 14. Note the
six distinguishable ridges. The outermost ridge, 6, is the most prominent. Ridges 2
and 3 merge in profile C.
(compare Figs. 17C–17D). Flank slopes are similarly shallow ranging between 2◦ and 3◦ , and the boundaries at the base of these
flanks are relatively sharp.
As discussed in the previous section, the boundary between this
terrain and the knobby terrain that lies to the east is abrupt with
no evidence of a gradual transition from one terrain type to the
next. The hummocky-like terrain is better defined than the adjacent knobby facies; the knobby facies appears subdued, as though
it originally appeared similar to the hummocky-like terrain until it
was covered by dust, flows, or till. This may suggest a genetic link
between the knobby facies and the hummocky-terrain.
On the western margin of the complex ridges and hummockylike terrain lies a relatively smooth surface covered sparsely by the
ridged facies, again with an apparent lack of any gradual transition between these two very distinct surfaces (Figs. 17A, 17B). As
previously mentioned, a sharp, low-relief, west-facing scarp marks
this margin in places where the hummocky-like terrain is raised
up to ∼50 m above adjacent smooth surfaces. This scarp is most
prominent at the southwestern boundary of the hummocky-like
terrain, where the aforementioned 2 km wide bank of material extends approximately northwards (Fig. 11E). The curved nature of
this feature is coincident with the arcuate western margin of the
hummocky-like terrain and also extends approximately parallel to
the outer scarp and ridged facies that trace the distal margin of
the Ascraeus FSD.
The gradational nature and spatial relationships of the units
characterizing the hummocky-like terrain suggests that they are
genetically linked. The sharp boundaries and peaks of the knob
and ridge-like features constituting this terrain result in an apparent “fresh” appearance. However, the overall chaotic distribution
and dissected character of these features suggest that the deposit
is modified and may thus have undergone a more complex history
relative to the ridged and knobby facies. In terrestrial glacial environments, post-depositional deformation or relocation of proglacial
material occurs as a direct result of glacigenic processes, such as
Fig. 14. MOLA profiles of the ridged facies seen in Fig. 13, where north is on the
right. The profiles have been corrected for the angle at which the MOLA track
crosses the ridges; they have been projected as though the tracks were orthogonal
to the ridges in order to show the actual spacing between and widths of the individual ridges. The spacing between ridges is quite consistent. Although the heights of
the ridges vary, the outermost ridge is always the tallest. Note the merge of ridges
2 and 3 in profile C. Vertical exaggeration is ∼5×.
squeezing and pushing, associated with the overriding and/or overburden pressure of ice (Benn and Evans, 1998). The squeezing of
saturated sediment from beneath ice margins has been observed
in front of several terrestrial glaciers (e.g. Price, 1970) and tends
to occur coeval with the bulldozing or pushing of this material by
minor advances of the glacier snout (e.g. Boulton and Eyles, 1979).
Push and squeeze moraines resulting from these processes are reported to exhibit relatively minor relief (<10 m in height), steep
or vertical sides, saw-tooth and curvilinear planforms, and may degrade rapidly as they are commonly subjected to repeated reworking by ice push (Benn and Evans, 1998). This notion of increasing
deformation with continued minor re-advance of ice during a general recession may explain the increasingly degraded appearance
of the hummocky-like terrain southwards. The more degraded and
chaotic appearance of the hills relative to the complex ridges may
Ascraeus Mons tropical mountain glaciation
95
Morphologically similar features have been described as a component of the knobby facies contained within the Pavonis Mons
FSD (Shean et al., 2005). It has been suggested that these features
may represent an alternative class of terrestrial glacial features,
namely “ribbed” or “Rogen moraines” (Scott and Zimbelman, 1995;
Shean et al., 2005), produced from glaciotectonic processes. The
morphology of these features shares several consistencies with the
Ascraeus hummocky-like terrain; they consist of numerous, closely
spaced “ridges” that typically develop transverse, and sometimes
parallel, to ice flow and display straight and arcuate planforms,
often with “horns” pointing in a downflow direction (Benn and
Evans, 1998). One hypothesis for the formation of Rogen moraines
involves the deformation of debris-rich ice due to the compressive
forces acting at the glacier base resulting in moraine formation as
a stack of thrust slices of basal ice (Hättestrand and Kleman, 1999;
Kleman and Hättestrand, 1999). This formation mechanism for Rogen moraines is considered to be more applicable to knobby features in the FSDs relative to alternative hypotheses of formation
that invoke basal sliding of ice. Ice flow by basal sliding would
effectively reduce the preservation potential of previous glacial
landforms, which is inconsistent with the proximity of the Pavonis knobby facies with the well-preserved ridged facies (Shean et
al., 2005).
3.6. Lobate flows (flat-topped ridge region) and plateau feature
Fig. 15. A conceptual model showing map and side views of the formation of the
east-facing arcuate scarp after Shean et al. (2005). The four major steps are: (1) The
steady-state glacier extends to the current location of the scarp. (2) Lava flows from
the southwest flank of Ascraeus Mons. The flows circumvent and embay the glacier,
chilling at the contact to form the arcuate scarp. (3) The glacier advances onto the
solidified lava flows, reaching a new steady-state at its maximum extent. (4) The
glacier retreats, but a series of periods of standstill leaves ridges superimposed on
the lava flow, as well as some knobs.
be a reflection of the degree of deformation each unit has undergone.
An interpretation as post-depositional deformation of debris by
pushing and squeezing is not a unique origin for the hummockylike terrain. For example, we do not observe a saw-tooth pattern
of moraines as is common with terrestrial push moraines (e.g.
Matthews et al., 1979). Furthermore this hypothesis suggests that
sediment underlying the ice is water-rich, which contradicts earlier evidence and the interpretation that ice in the regions of the
FSDs is below the pressure melting point throughout based on lack
of additional signs of meltwater. However, deformation of debris
by such glacigenic processes should not be ruled out as a formative mechanism for the hummocky-like terrain because of the
potential for meltwater production by lava–ice interactions or at
points where basal ice temperatures exceed the pressure melting
point.
“Lobate flows” (Zimbelman and Edgett, 1992) covering an area
of approximately 2300 km2 form a distinct feature extending
northwards—following regional and local slopes—from a position
proximal to the southwest flank of the Ascraeus Mons shield base
(Figs. 9, unit Ff; 18). This lobate feature consists of at least three
broad, elevated plateaus with steep margins that rise up to 500 m
above the surrounding surfaces (Figs. 7G–7I, 8L–8N). Plateau surfaces are marked in parts by segments of wide, leveed channels,
and the two largest lobes have medial channels (Figs. 11C, 18).
At the distal end of the medial channels, smaller lobate features
emerge and follow local topography. These characteristics make
the lobate flows distinct from the adjacent lava flows to the south
and west, which lack steep margins and wide levees. Flank slopes
vary between 2.8◦ and 4.5◦ for lower flanks and between 8◦ and
23.5◦ for upper flanks. The upslope (southern) end of the feature
is truncated by lava flows originating from troughs parallel to the
shield base of Ascraeus (Zimbelman and Edgett, 1992), while the
downslope (northern) end terminates abruptly at the scarp. There
is, however, a small fraction of this northern end of the flow feature that appears to head in a westward direction, connecting to
the southern end of the ridged facies (visible in Figs. 3, 12A).
In addition, a relatively flat-topped, elevated plateau can be observed ∼7 km west of the shield base in the central portion of
the deposit (Figs. 9, unit P; 11D). The mesa-like feature, which is
34 km long and 19 km wide, is stepped in nature with two terraces, both lobate and elongated to the north, generally aligned
with the regional northward slope. Analysis of the topography reveals the presence of a small mound, a few tens of meters high,
on the southeast end of the plateau. A narrow ridge extends from
the base of the plateau at the southern end of the mound to the
south for ∼20 km and disappears under the flat-topped flows. The
ridge can be seen extending along strike to the north, where it
emerges from beneath the plateau and extends for several tens of
km within the deposit, discontinuously and in a somewhat sinuous
manner. Unlike the lobate flows farther south, the upper surface of
this flow-like feature lacks any evidence of leveed channels. Surfaces do appear to be degraded, particularly on the western side
where the terraced surfaces are observed, both of which are covered by knobs constituting the southern portion of knobby facies
(Figs. 9, 16D). The asymmetry between the western and eastern
96
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Fig. 16. A gallery of images of the knobby facies at THEMIS VIS resolution (A, B, C, D—18 m/pixel) and MOC narrow-angle resolution (E—6 m/pixel). (A) Sparsely distributed
knobs in the northern end of the fan-shaped deposit showing elongation along a NW-SE trend (THEMIS V18451009). (B) Knobby facies superimposed on the flat-topped flows
(THEMIS V18139015, V17827031). (C) Randomly distributed knobs near the center of the deposit (THEMIS mosaic). The black box shows the location of the MOC image shown
in E. (D) The southern end of the plateau feature, with the knobby facies extending continuously over both the plateau and the surrounding terrain (THEMIS V17827031).
(E) The knobby facies in the center of the fan-shaped deposit showing no preferred orientation (MOC M08-04401). Debris is present on the northern faces of the knobs.
edges extends to flank slopes, which are 9.1◦ and 6.1◦ on average
(Fig. 7F) and which reach maxima of 15.1◦ and 9.5◦ on the western
and eastern flanks, respectively.
Flow-like features morphologically similar to those described
here have been identified within the Pavonis FSD (Zimbelman and
Edgett, 1992; Scott et al., 1998; Shean et al., 2005). Based on Viking
data, Scott et al. (1998) described these lobate features as elongate
sinuous ridges and suggested that they might be eskers formed
beneath the wastage zone of a disintegrating ice sheet. These features, as discussed previously, are not simply lobate ridges, but are
broad, elevated plateaus, which is not consistent with typical terrestrial esker morphology (e.g. Huddart et al., 1999; Price, 1966).
Furthermore, eskers are characteristic of wet-based glaciers and
would therefore not be expected at the site of a cold-based glacier.
Examination of Arsia and Pavonis has revealed no evidence, such
as sinuous channels, lake deposits, and/or braided streams, for extensive wet-based glacial activity (e.g. Head and Marchant, 2003;
Shean et al., 2005). More consistent with the lobate nature of the
features, their elevated topography, and their apparent relationship with fissures on the southwest flank of Ascraeus (Zimbelman
and Edgett, 1992) is the alternative hypothesis that these features
are distinctive types of lava flows (Scott et al., 1998). Zimbelman
and Edgett (1992) interpret these features as pyroclastic flows that
originated from troughs in the lower flanks of the volcanoes. They
suggest that the effusive activity that produced these post-dated
the emplacement of the FSD and was triggered by the removal of
a lithostatic load accompanying a large landslide event that would
have formed the FSD. However, such a proposition is inconsistent
with the observations that show superposition of the knobby facies
over both lobate flow features at Ascraeus Mons and over comparable flow features at Pavonis Mons (Shean et al., 2005).
On the basis of the stratigraphic relationships, the geomorphology, and the topography, we interpret these features to be due to
lava–ice interactions. We prefer the interpretation of these lobate,
steep-sided flow features as subglacial lava flows (non-pyroclastic),
and suggest that they formed concurrent with the presence of a
glacier west of Ascraeus Mons (Fig. 18). These subglacial flows superpose older subaerial flows which were deflected around the
glacier, forming the outer arcuate scarp. In this model, subaerial
flows cooled against the ice margin without the production of
large quantities of meltwater; lava accumulating on ice is reported
to “rapidly form a lower chilled margin that insulates the hot
flow interior and prevents further melting of ice” (Helgason, 1999,
p. 234; Wilson and Head, 2007b). During a subsequent glacial advance which extended farther south, additional flows, this time
subglacial, formed the steep-sided, flat-topped features. The obser-
Ascraeus Mons tropical mountain glaciation
Fig. 17. Images of the complex curved ridges and hummocky terrain at THEMIS VIS
resolution (A—18 m/pixel) and MOC narrow-angle resolution (B, C, D—6 m/pixel).
(A) Sharp transitions between the chaotic hills, hummocky terrain, and knobby facies (THEMIS V18451009). (B) A MOC image of the same facies (MOC E10-04849).
Note the accumulation of debris on the northern faces of complex curved ridges.
The black boxes show the locations of images C and D. (C) An enlargement of the
top box in (B). The cusp-like structures show a preferred SW-NE orientation. (D) An
enlargement of the bottom box in (B). The transition between the hummocky terrain and the knobby facies is abrupt. The knobby facies appears subdued compared
to the hummocky terrain, as if it once looked like the hummocky terrain but has
since been covered by dust or a lava flow.
vation of the knobby facies superposed on this lobate flow feature
is consistent with our interpretation, supporting sublimation of the
glacier after the flows were emplaced. Based on these relationships
and the observation that the knobby facies also covers a portion of
the more centrally located plateau feature, it is likely that both
flow features formed prior to the knobby facies and subsequent
to the ridged facies. The topographic separation of the ridged and
knobby facies suggests that they were formed during two separate
periods of ice recession and perhaps also two distinct glaciations.
Volcano–ice interactions on Earth and Mars have been extensively studied (e.g. Smellie and Chapman, 2002). Although the
occurrence of volcanism during the presence of the glaciers may
seem like an unlikely coincidence, the evidence for lava–ice interactions is well-established on Mars in general (e.g. Head and
Wilson, 2007; Head et al., 2003), and in the Tharsis region in particular (e.g. Shean et al., 2005; Wilson and Head, 2007b). Wilson
97
and Head (2002) developed the theory of magma–ice interactions on the Earth and Mars, showing that dikes could penetrate
into overlying glacial ice, forming ridges, and that lava flows at
the ground surface–ice interface (sills) could also occur, forming
thick subglacial “flows” with steep sides. Head and Wilson (2007)
also assessed the relationship of surface flows and glacial ice, and
showed how chilled margins and bases inhibited the development
of meltwater in the glacial ice.
The morphologies of the two lobate flow features at Ascraeus
Mons appear distinctly different, which may be indicative of varying formative mechanisms. The mesa-like lobate feature in the
central portion of the FSD lacks the leveed channels interpreted
to reflect flowing material and is significantly more tabular in
planform relative to the elongate, lobate morphology of the southern flow features. The tabular structure of this plateau feature is
reminiscent of tuyas or table-top mountains found in Iceland and
north central British Columbia, which formed from subglacial volcanic activity (Fig. 19a). We interpret the sinuous ridge leading to
the plateau as the remnant of a dike intruding into the ice, and
the mound at the SE end of the plateau is the eruptive center
(Figs. 19a, 19b). The terraces thus represent subglacial eruptions
that formed marginally-chilled flows which headed north, downhill.
A subglacial eruptive mechanism for the formation of lobate
flow features was one of the explanations proposed for comparable
features observed within the Pavonis FSD (Shean et al., 2005). In
addition, they identified three arcuate scarps which likely formed
by lava flows banking against the rim of the extant glacier (Shean
et al., 2005). Similar proposals were made for plateau-like and
ridged deposits in the Arsia Mons FSD (Wilson and Head, 2007b),
where evidence for associated meltwater distribution is seen (e.g.
local eskers, glacial surging and distal fluvial channels). The meltwater was generated through subglacial and englacial eruptions, as
well as from dike-related synglacial eruptions. Wilson and Head
(2007b) found evidence for sill-like subglacial eruptions at both
Arsia and Pavonis, which produced enough heat to force a local
transition from cold-based to wet-based conditions. The drainage
of meltwater formed local fluvial zones on a variety of scales, from
short subglacial channels to jokulhlaups (Wilson and Head, 2007b).
When subglacial eruptions occur, the mass displaced by the
lavas is accommodated by (1) melting and loss from beneath the
glacier by basal outflow, (2) melting and local refreezing, (3) deformation and displacement of the adjacent and overlying ice, and/or
(4) penetration of the ice and venting of tephra to the surface (e.g.
Wilson and Head, 2002, 2007a). Once the overlying ice is gone,
it is difficult to assess the relative roles of these accommodation
mechanisms. Meltwater generated by subglacial eruptions at Ascraeus Mons would likely have ponded locally behind the outer,
edifice-facing, arcuate scarp. The lack of identified fluvial features,
however, suggests that the role of meltwater flow and discharge
was minimal, which can occur in cold-based glacial environments.
As such, it is probable that significant volumes of ice were displaced during subglacial eruptions, and lava flows breaking out
onto the surface of the glacier solidified, were broken up, and
were deposited as sublimation till. The geomorphologies indicative
of lava–ice interactions which are present strengthen the interpretations made here for the Ascraeus FSD. Based on the relative
positions of the facies and subsequent necessary timing of events,
we propose a time-step glacial history which marks the most significant stages of the glacial growth, retreat, and interaction with
volcanism (Fig. 20).
3.7. Mountainous terrain
The mountainous terrain consists of two isolated regions at
11.9◦ and 12.3◦ latitude that transcend the outer scarp proximal to
98
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Fig. 18. On the left is a THEMIS VIS and IR mosaic of the flat-topped ridges extending from the southwest flank of Ascraeus Mons. The lobate flow features display wide
leveed channels and steep margins, and are largely covered by knobs. Their morphology suggests that they formed via subglacial lava flows. On the right is a sketch map of
the region, showing the probable flow paths of the lava as it traveled northward in the downslope direction. The edge of a subaerial flow can be seen in the bottom center
of the sketch, as well as in the bottom center and bottom left of the THEMIS mosaic. An arcuate ridge is present near the top left.
(a)
(b)
Fig. 19. (a) A comparison of the plateau in the fan-shaped deposit (left) to the tuya Herdubreid, located in northeast Iceland (right). Herdubreid formed primarily beneath
Vatnajökull, the largest glacier in Iceland, although the top formed subaerially when the lava penetrated the top of the glacier. We believe that the plateau in the deposit
is a table mountain, having also formed via a subglacial eruption. The images of the plateau feature are from a THEMIS VIS and IR mosaic. MOLA data is overlain on the
map view. The vertical exaggeration of the perspective view is 10×. The map view of Herdubreid is from a Landsat 7 ETM+ scene (band 7, 30 m/pixel, WRS-2, Path 217,
Row 014, taken on 23 September, 2000). The perspective view is courtesy of Wikimedia Commons. (b) A sketch map of the plateau in the Ascraeus FSD as seen in (a). Note
the arcuate scarp which ends abruptly where it encounters the southeast margin of the plateau. The ridge leads toward the eruptive center—the highest topographic point
of the plateau. Curvilinear arrows indicate the direction of the proposed subglacial lava flows. The knobby facies is superimposed over most of the southern portion of the
plateau, as well as some of the northeast edge.
Ascraeus Mons tropical mountain glaciation
99
The mountainous regions specified here represent only a fraction
of the features that these authors described as isolated mountainous areas. We find that a debris flow origin for these areas is not
consistent with the high relief of individual mountains and hills
that transect the outer scarp and there is no indication of a source
for the mountainous terrain on either the shield flank or in the
outer reaches of the FSD. The well-preserved nature of the lobate
terrain in its lower elevation, northern extent argues against an
erosional origin for the margins of elevated sections of this terrain
and thus for the escarpment that separates this elevated terrain
from the rest of the FSD. As an alternative, we interpret the mountainous regions to be lava flows that, as discussed concerning the
scarp formation, flowed around and embayed glacial ice. This hypothesis is supported by the apparent genetic link between the
mountainous regions and lobate flow features that exhibit channels suggestive of flowing lava and appear closely associated with
fissures on the flanks of Ascraeus Mons.
3.8. Degraded flank material
Fig. 20. We propose a nine step history of the glacial activity necessary to form the
Ascraeus Mons fan-shaped deposit. Based on the geomorphologies and stratigraphic
relationships, we find this history to be the most plausible, but other potential histories can account for the range of features in the fan-shaped deposit. Additional
evidence is necessary to constrain further the steps that occurred in the formation of the deposit. Our history is as follows: (1) The flank of Ascraeus Mons was
originally free of ice. (2) During the Mid- to Late-Amazonian, accumulation of ice
and snow led to rapid glacial growth. The expansion of this glacier responded to
regional topography, resulting in an asymmetrical shape. (3) Subaerial lava flows
from the southwest flank of Ascraeus Mons circumvented the distal rim of the
glacier and chilled at the margin, creating the arcuate scarp. (4) The glacier advanced past the scarp, reaching its maximum extent. (5) Slow retreat with periods
of standstill created drop moraines, yielding the concentric ridged facies at the distal rim of the deposit. (6) Continued rapid retreat may have created some of the
observed knobby facies in the northern end of the deposit. It is not clear whether
the glacier retreated for a brief period, or disappeared completely. (7) A re-advance
of the glacier, or perhaps a distinct second glaciation occurred. This time, much
more accumulation occurred near the southern end of the deposit, creating a more
symmetrically-shaped glacier. During this period of glaciation, subglacial flows occurred, forming the flat-topped ridges in at the southern end of the deposit and
the plateau feature. The glacier remained in relative equilibrium in this position
with minor fluctuations, forming the ridges east of the arcuate scarp. (8) Gradual
recession led to the formation of the complex curved ridges. Sublimation then accelerated, due perhaps to a sudden stoppage of accumulation. This led to the formation
of the largest region of knobby facies, which were emplaced over the flat-topped
flows and plateau. (9) Eventually, the glacier retreated completely, leaving the deposit presently observed.
the distal margins of the Ascraeus FSD. A complex assemblage of
irregularly shaped, kilometer-wide hills and mountains constitutes
these areas of mountainous terrain (Figs. 9, unit Mt; 11K). Heights
of these features are variable ranging from a few meters to over
200 m. The northern of the two mountainous regions covers an
area of 100 km2 and has an elevation range of ∼1500 to 1750 m.
The southern region covers an area of approximately 190 km2 and
extends at least 38 km with elevations of ∼1670 to 1990 m. This
section of mountainous terrain appears to truncate or mark the
flow front of the lobate feature that is interpreted to have flowed
coincident with the southern end of the scarp along an ice margin. The continuity of the lobate feature with the southern region
of mountainous terrain is consistent with geomorphic maps produced by Zimbelman (1984) and Zimbelman and Edgett (1992).
The mountainous terrains have been interpreted as debris flows
(Zimbelman, 1984) and as material which predates the Ascraeus
Mons shield and surrounding plains (Carr, 1975; Zimbelman, 1984).
A distinct unit exists immediately adjacent to the scarp of the
western Ascraeus flank (Figs. 9, unit Df; 11I, 11J), extending ∼15–
20 km west of the scarp. Slope maps and topographic profiles
(Figs. 5a, 7A–7G, 8O, 8P) show that this unit is present only on the
western flank of Ascraeus Mons and thus likely has some genetic
link to glaciation; the unit terminates in the south between the
shield base and the lobate lava flows, and in the north where we
would expect the northern margin of the glacier to reach (based
on projecting the ridged facies to the shield base). The unit, which
has a slightly higher albedo in the THEMIS images than the rest
of the FSD floor, is not observed on the flank, nor can it be found
anywhere else within the deposit. MOLA topography shows the degraded flank material to be extremely smooth at the 50 m contour
scale, whereas the immediately adjacent surface to the east on the
Ascraeus flank is quite rough, likely due to lava flowing down the
edifice. The difference in texture occurs abruptly at the scarp. The
unit tapers along its western margin into the flat floor of the FSD.
A HiRISE image of the region (Fig. 12C) shows it is largely covered
by dunes. These dunes may be due to the accumulation of dust—
and perhaps dust-nucleated ice—carried across the Tharsis region
by a westerly wind; the dust would be trapped when it encounters
the steep escarpment at the base of the Ascraeus flank.
We believe that the degraded flank unit may represent an accumulation zone or a region where the pressure from the overlying
ice has mechanically degraded the flank material. With the glacier
emplaced, subsequent lava flows descending the flank of Ascraeus
would have chilled and stopped upon encountering the glacier,
building up the scarp. The scarp may then have served to enhance
preservation of snow and ice in alcoves, although no flow features
were observed in the region. As the glacier thinned during its waning stages, lava flows reaching the scarp may have extended over
the ice. These flows would have been deposited as till upon complete sublimation of the glacier.
4. Comparisons among the FSDs at Ascraeus, Pavonis and Arsia
4.1. The role of topography
Throughout the previous section, we have made various contrasts between the geomorphologies in, and the size of, the FSDs
at Ascraeus, Pavonis and Arsia. Although these variations have a
probable climatic cause, which will be discussed later, they are
also governed by local topography, debris supply, and the timing of
glacial advance and volcanism. While the FSDs do exhibit a number of similarities, including concentric ridges, arcuate scarps, lobate flow features, and knobby terrain, there are variations among
100
S.J. Kadish et al. / Icarus 197 (2008) 84–109
the units, as well as general differences among the FSDs. As discussed, the incline at the southern end of the Ascraeus FSD, which
slopes downhill in the northward direction, contributed to the formation of an asymmetrical glacier, and is thus the cause of the
change in the radius of curvature of the ridged facies from south
to north. In general, the orientations of the slopes west of Ascraeus
Mons form a depression near the side of the mountain; although
the southern end of the FSD is topographically higher than the
northern end, the surrounding slopes are primarily inclined uphill
moving away from the deposit. This topography would have generally inhibited expansion of the glacier. This is not the case at Pavonis or Arsia, where gentle and uniform slopes underlie the FSDs,
which allowed for more symmetrical growth of these glaciers.
4.2. Ridged facies
The ridged facies contained within the Ascraeus Mons FSD resembles the ridges that trace the distal margins of equivalent
deposits west of Arsia and Pavonis Mons in terms of both morphology and areal distribution. Analogous to Ascraeus, the ridged
facies associated with Arsia and Pavonis Mons encompasses almost
the entire length of the deposits. The margin of each facies is defined by a taller 50–100 m ridge with smaller ridges inwards of
and concentric to this outer margin (Head and Marchant, 2003;
Shean et al., 2005). These ridges usually exhibit a much higher degree of regularity than those observed at Ascraeus Mons, with a
typical spacing of ∼1 km. There are sections, however, where the
Ascraeus Mons ridges approach this uniformity (Figs. 13, 14). A particularly distinctive characteristic reported from the ridged facies
of both Arsia Mons (Carr et al., 1977; Head and Marchant, 2003)
and Pavonis Mons (Edgett et al., 1987; Shean et al., 2005) is its superposition on underlying topography, including an impact crater
and lava flows, without either deflection of the ridges or disruption to the underlying terrain. Such examples of the ‘blanket-like
nature’ (Head and Marchant, 2003) of the ridged facies are similarly observed at Ascraeus Mons, where ridge superposition occurs,
and support the interpretation of the ridges as drop moraines from
cold-based glaciers (e.g. Head and Marchant, 2003).
The most distinctive difference between the ridged facies of Ascraeus Mons and the other Tharsis Montes is the overall frequency
of the ridges and lateral extent covered by the facies. The ridged
facies of Arsia and Pavonis Mons consist of hundreds of concentric
ridges that extend for a lateral distance of ∼80 and ∼180 km, respectively, at their widest points. The lower frequency and shorter
extent of the ridged facies at Ascraeus Mons may be linked to the
smaller size of the FSD, and could suggest a limited supply of debris relative to other FSDs. Regardless of their size and frequency,
our interpretation of the ridged facies as terminal drop moraines is
well-supported. Their relative curvatures and positions at the distal
rim of the FSD and in smaller concentric sets within the deposit
indicate multiple periods of glacial stability and perhaps distinct
glaciations.
4.3. Knobby facies
The knobby facies in the Ascraeus FSD appears similar to that
at Pavonis and Arsia. Shean et al. (2005) note that, at Pavonis
Mons, the knobby facies consists of kilometer to subkilometer hills
which overlie the topography and are superposed on the ridged
facies. They also display localized anisotropic distribution throughout the deposit (Shean et al., 2005). All of these characteristics
are also true of the knobby facies at Ascraeus Mons. Although
we did notice fewer dunes between knobs at Ascraeus than at
Pavonis and Arsia, we believe this is a result of post-modification
from the accumulation of dust and not related to excess finegrained debris entrained in the glacier. We find that the forma-
tion mechanism for the knobby facies at Ascraeus is the same
as that proposed for the knobby facies at Pavonis and Arsia, requiring additional phases of glacial advance in order to superpose
the knobs on the ridged facies (Shean et al., 2005). In all three
cases, it is expected that the knobs were produced by sublimation and down-wasting of glacial ice (Head and Marchant, 2003;
Shean et al., 2005).
4.4. Smooth facies
One of the most striking differences among the geomorphology
of the FSDs is the lack of a smooth facies at Ascraeus Mons. On the
basis of MOLA topography and its association with the other facies,
the smooth facies, which was originally interpreted to be pyroclastic flows (Scott and Zimbelman, 1995), is now considered to be
debris-covered ice (Head and Marchant, 2003; Shean et al., 2005).
Some differences are apparent between the smooth facies at Arsia
and Pavonis, most notably the presence of closely spaced, concentric ridges, which are more common at Arsia than Pavonis (Head
and Marchant, 2003; Shean et al., 2005). In both cases, however,
the presence of the lobate, high relief deposits suggests the presence of ice within the FSDs. The absence of this debris-rich ice in
the Ascraeus deposit is expected based on the age of the Ascraeus
deposit, and on climatic and topographic factors. The Ascraeus deposit has the lowest elevation of the three FSDs, and thus a larger
proportion of its glacier was below the lower of two equilibrium
line altitudes, having negative net mass balance—this will be discussed in detail in a later section. The Ascraeus glacier, which has
the oldest age based on crater counting, was also the smallest and
likely the first to sublimate. Thus, while debris covered glacial remnants still exist in the other two deposits, debris-rich ice has been
removed from the Ascraeus FSD.
5. Chronology
The ages of the Tharsis Montes and Olympus Mons have been
established in previous studies using both stratigraphic relationships and crater counting. These stratigraphic relationships had
been previously assessed and, in accordance with Viking crater
counting data, Tanaka (1986) assigned each of the Tharsis Montes a
Late-Amazonian age. Based on their observations and the resulting
palaeostratigraphic reconstruction, Scott and Tanaka (1981, 1986)
argued that the Tharsis Montes’ FSDs are approximately the same
age as the youngest flows coming from each of the volcanoes. Subsequent regional mapping has been conducted (e.g. Scott and Zimbelman, 1995; Scott et al., 1998) which verifies that the FSDs were
emplaced throughout the Mid- to Late-Amazonian (Fig. 21). Scott
and Zimbelman (1995) discern separate Amazonian ages for the
distinct units within the Arsia Mons FSD. Scott et al. (1998) identify ages for lava flows on the shield of Pavonis Mons, arguing that
the majority are Early Amazonian in origin while Late-Amazonian
flows are limited in extent; the FSD formed concurrently with
these flows.
Our efforts to date the Ascraeus FSD, as well as those made by
Shean et al. (2005, 2006) to date the Pavonis and Arsia FSDs, help
to place the glaciations into the context of the volcanic history
of the Tharsis Montes. While previous work (e.g. Scott and Zimbelman, 1995; Scott et al., 1998) shows a brief hiatus in volcanic
activity, during which some of the FSD units formed (Fig. 21), we
see clear evidence that volcanism occurred during the emplacement of the facies. The concurrent volcanism is related to units
At6 and At5 —Mid- to Late-Amazonian lava flows—which is consistent with the setting for the deposits over the time interval of the
glaciations. Additionally, the units of the FSDs were interpreted
to be somewhat sequential (Fig. 21), with the smooth facies (As)
forming subsequent to the ridged and knobby facies (Ar and Ak).
Ascraeus Mons tropical mountain glaciation
101
Table 1
THEMIS IR crater-counting statistics for N(0.35)–N(2) over the coverage of IR data
(99% of the FSD)
Fig. 21. A compilation of the geologic histories composed by Scott and Tanaka
(1986), Scott et al. (1998), and Scott and Zimbelman (1995) based on their geologic
mapping of the Tharsis region. The relative ages of the units were determined using
stratigraphic relationships and crater counts. In their geologic map of the western
equatorial region of Mars, Scott and Tanaka (1986) offer a history of the large volcanic shields and associated lava flows (AHt3 , At4 , At5 , At6 ). This history shows
continuous formation of lava flows, with no gap between At5 and At6 . Scott et al.
(1998) provide a map of Pavonis Mons, its lava flows, and the associated units in
the fan-shaped deposit (As, Ar, Ake, Ak). They suggest a time gap between At5 and
At6 , during which time the ridged facies (Ar), elongate ridged material (Ake), and
knobby facies (Ak) formed. The smooth facies (As) then formed concurrently with
smooth, fresh-appearing lava flows (At6 ). Scott and Zimbelman (1995) also show a
time gap between At5 and At6 . They note, however, that the flows which compose
At5 overlie the knobby facies but underlie the ridged facies in places. They also suggest that the smooth facies is roughly the same age as the flows of At6 .
Our analysis, as well as the analyses made by Shean et al. (2005,
2006), show the formation of the units to be largely integrated,
resulting from multiple glacial advances and recessions.
Crater counting using high resolution images from MOC and
HRSC has further confirmed Amazonian ages for the Tharsis
Montes (e.g. Hartmann et al., 1999; Hartmann and Neukum, 2001)
and their respective FSDs (Sheen et al., 2005, 2006), as well as
for Olympus Mons (e.g. Basilevsky et al., 2005; Grier et al., 2001;
Hartmann and Neukum, 2001; Head et al., 2005). Shean et al.
(2005) used THEMIS IR and VIS data to establish a Late Amazonian age for the Pavonis Mons FSD, between 10 and 200 Ma
based on isochrons from Hartmann and Neukum (2001). They
note, however, that limited coverage of the area and a poor statistical distribution make this absolute age somewhat uncertain;
a young crater-retention age is expected due to (1) modification
of the units of the FSD since their deposition and (2) erosion and
Crater diameter bin (m)
# of craters, N
Area (km2 )
N/km2
Error
350–500
500–700
700–1000
1000–1400
1400–2000
14
6
4
2
1
13363
13363
13363
13363
13363
1.05E−03
4.49E−04
2.99E−04
1.50E−04
7.48E−05
2.80E−04
1.83E−04
1.50E−04
1.06E−04
7.48E−05
Fig. 22. A size-frequency plot of the crater-counting data for N(0.063)–N(2) based
on THEMIS IR and VIS mosaics of the Ascraeus Mons FSD. The data are plotted on
isochrons from Hartmann (2005). The dotted line shows the best fit age, 380 Ma,
based on data from N(0.7)–N(2). Including smaller craters, which are being modified
and eroded at a rate comparable to their production, lowers the age of the deposit.
This makes it difficult to establish accurately an absolute age. Based on the largest
crater data, we find an age of a few hundred Myr appropriate, placing its formation
in the Mid- to Late-Amazonian.
infilling of smaller craters (Shean et al., 2005). Additionally, some
craters observed may be impacts into the Tharsis lava plains, which
underlie the FSD units; counting these craters would result in an
overestimate of the age (Shean et al., 2005). Shean et al. (2006)
conclude that a Mid- to Late-Amazonian age for the Arsia Mons
FSD is the most likely. Using THEMIS VIS and HRSC images, they
calculated a lower boundary for the absolute age of >25 Ma (90%
confidence) by counting only the fresh craters on the deposit and
an upper boundary of <650 Ma (90% confidence) by counting all
craters on the deposit. This upper boundary population included
craters that had experienced infilling and modification, indicating
that they were either older or emplaced on an underlying flow
(Shean et al., 2006).
We undertook an analysis of the impact crater size-frequency
distribution over a near-complete (>99% coverage) THEMIS IR mosaic (100 m/pixel) of the Ascraeus Mons FSD. Our results yield a
Mid- to Late-Amazonian age. Craters were counted down to 350 m
in the IR mosaic. The largest crater in the deposit is 1.8 km in diameter, and only 27 craters exist over 350 m in diameter (Table 1).
Using the data from N(0.7)–N(2) gives a best fit age of 380 Myr
based on the Hartmann (2005) isochrons (Fig. 22). However, if we
expand that range, using N(0.5)–N(2), the best fit age lowers to
250 Myr.
102
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Table 2
THEMIS VIS crater-counting statistics for N(0.063)–N(0.35) over the coverage of VIS
data (83% of the FSD)
Crater diameter bin (m)
# of craters, N
Area (km2 )
N/km2
Error
63–88
88–125
125–177
177–250
250–350
14
16
15
15
18
11076
11076
11076
11076
11076
1.26E−03
1.44E−03
1.35E−03
1.35E−03
1.63E−03
3.38E−04
3.61E−04
3.50E−04
3.50E−04
3.83E−04
The THEMIS VIS mosaic (18 m/pixel) covers 83% of the deposit,
although the images are of varying quality. Only craters that were
too small to be seen in the IR data were counted using the VIS
data, although the diameters of the larger craters were checked
using the VIS data. Craters smaller than 65 m were not counted,
but 78 were observed between 65m and 350 m (Table 2). Plotting the VIS craters on the same isochrons shows that the data
flatten out for diameters less than 350 m (Fig. 22), crossing the
isochrons (Hartmann, 2005). The aforementioned arguments concerning the crater retention of the FSD used by Shean et al. (2005)
can be applied here. The deposit has undergone modification and
possible resurfacing since its deposition; the flattening of the plot
at low diameters (Fig. 22) suggests that craters <350 m in diameter are being eroded or infilled at the same rate that they are
being produced. Craters between 350 m and 700 m are also affected by this modification, and show the initial signs of a rollover
in the data as it crosses the 100 Myr isochron (Fig. 22). Because
of this, we prefer to rely on the larger craters in order to assign
an absolute age. However, with only 7 craters larger than 700 m,
the statistics of small numbers make determining an absolute age
with high levels of confidence difficult. Nonetheless, we find the
Mid- to Late-Amazonian age consistent with the ages of the Arsia
and Pavonis FSDs, which, as we will show, has important implications for the both climate and ice-sheet models.
6. Glacial mass balance, ice-sheet models, and equilibrium line
altitudes
A detailed understanding of the geomorphology and age of
the Tharsis Montes FSDs can provide crucial constraints on martian ice-sheet models, the history of martian obliquity, and climate
models. While there is much progress to be made in understanding
how glacial dynamics operate on Mars, key advancements permit
us to model possible histories for the growth and retreat of the
Tharsis Montes glaciers that match with relative accuracy the size
and position of glaciers responsible for the FSDs (Fastook et al.,
2005, 2006). Because the terminal moraines at the distal rim of
the FSDs indicate periods of relative stability, their positions should
represent a significant climate signal. We can more clearly see this
connection by looking at the following chain of factors relating climate to the surface morphology (modified from Meier, 1965):
Obliquity → accumulation and ablation → mass budget and
equilibrium line altitude → dynamic response and debris transport
→ position of moraines.
The obliquity of Mars dictates the amount of incident insolation for a given latitude; direct insolation on ice or ice-rich soil can
lead to sublimation, increasing the concentration of volatiles in the
atmosphere. The obliquity also affects the atmospheric dynamics,
altering the magnitude and directionality of regional winds. Accumulation and ablation rates depend both on the saturation of
volatiles in the atmosphere and on the regional winds. The mass
budget of a glacier and the position of the equilibrium line altitude (ELA) are controlled largely by the accumulation and ablation
rates. The position of the ELA, as will be discussed later, is also
dependent on some key atmospheric properties. The mass budget
and ELA are necessary to establish a glacier’s dynamic response
to changes in climate, and therefore influence how debris will be
transported in and/or on a glacier under different climate regimes,
and the timing of debris deposition. The deposition of debris and
the size of a glacier at equilibrium, as achieved via dynamic response to climate changes and mass loading, determine the position of the moraines. Thus, by working backwards through the
above chain of factors, the positions of the drop moraines may offer clues as to which of the various obliquity histories, as proposed
by Laskar et al. (2004), is most probable.
The key to decoding this relationship is an understanding of
the mass balance of glaciers on Mars, which would involve solving the larger problem of how glacial dynamics operate under
the martian climate. This relationship between glacial dynamics
and climate is well established on Earth; the mass balance and
ELA of a glacier can readily be measured using a number of
methodologies based on direct observations, remote sensing, hydrological evaluations and/or climatic calculations. Numerous studies have used models, an understanding of paleoclimatic conditions, and geomorphological observations to reconstruct ELAs of
former terrestrial glaciers (e.g. Benn et al., 2005; Meierding, 1982;
Porter, 2001). Although the problem is much more complex on
Mars, some efforts have been undertaken to investigate martian
mass balance (e.g. Fastook et al., 2005; Greve, 2000; Schmidt,
2004). Such studies and reconstructions have a number of poorly
constrained parameters including: (1) the water vapor saturation
of the atmosphere, (2) the atmospheric pressure, (3) the atmospheric dust content, (4) variations in obliquity and insolation,
and (5) the historic lapse rate. Some of our understanding of the
mass balance of former glaciers on Mars has come from studies
of the perennial north polar ice cap on Mars (e.g. Greve, 2000;
Greve and Mahajan, 2005; Hvidberg, 2006; Ivanov and Muhleman,
2000; Schmidt, 2004). Even these studies, however, indicate that
the modern mass balance is poorly constrained, and that more accurate meteorological data are needed, including continuous measurements of the water vapor distribution, dust content and winds
over the ice cap.
Preliminary mass balance assessments for tropical mountain
glaciers on the Tharsis rise require knowledge of both glacier
extent (known) and past climate conditions (unknown). Despite
the latter unknown, modern GCMs applied to Mars’ history have
shown that, during periods of high obliquity (>45◦ ), snow and ice
likely accumulate on the NW flanks of the Tharsis Montes (Forget
et al., 2006; Fig. 23). With estimates of past ice accumulation and
ablation as inputs in glaciological flow models, Fastook et al. have
been able to simulate glacial histories that match the spatial extent of the FSDs (Fastook et al., 2005, 2006, 2007). The Fastook
et al. model uses bed topography, surface temperature, geothermal heat flux, and the accumulation rate distribution (ARD) as the
primary inputs. The bed topography is quite well known, and the
surface temperatures have been estimated based on climate models. The geothermal heat flux and the ARD (Fastook et al., 2004,
2005) are less well known.
With respect to the mass balance problem on Mars, Fastook
et al. (2006) argue that ice accumulation depends on the saturation
vapor pressure of the atmosphere, which has an exponential temperature dependence. Ablation for martian glaciers is due primarily
to sublimation, not melting. The rate of sublimation depends on
saturation vapor density, and thus has the same exponential temperature dependence (Fastook et al., 2006). The vertical lapse rate
on Mars is less than half of what it is on Earth; current estimates suggest that the dry adiabatic lapse rate on Mars is around
4.3 K/km, while on Earth it is 9.8 K/km (Leovy, 2001). Fastook
et al. (2006) note that the gradual reduction in temperature at high
elevations means that both sublimation and accumulation rates
will decrease. Sublimation, however, has an inverse dependence on
Ascraeus Mons tropical mountain glaciation
Fig. 23. A map of the Tharsis region from Forget et al. (2006) with 2000 m contours
from MOLA data showing net surface water ice accumulation. The martian global
climate model used to simulate the accumulations uses a 45◦ obliquity and assumes
that surface water ice is present on the northern polar cap.
atmospheric pressure, and will thus decline less rapidly than accumulation as elevation increases, creating negative net mass balance
at high elevations. Similar to Earth, we would also expect negative net mass balance at low elevations where there is declining
relative humidity, increasing ventilation, and possibly some melting/sublimation. Based on this framework, Mars should have two
ELAs—a high and a low elevation ELA, with positive mass balance
in between, and negative mass balance above the upper ELA and
below the lower ELA (Fastook et al., 2005). The glacier at Ascraeus
Mons, being lower in elevation than the glaciers at Pavonis and
Arsia, would likely have had the highest percentage of its surface
area below the lower ELA.
As is the case with many mountain glaciers, a sharp precipitation gradient exists due to local orographic effects. We expect that
the Tharsis Montes glaciers also displayed an orographic effect, accumulating precipitation on the NW flanks due to a predominant
westerly wind direction (Forget et al., 2006). Thus, the ELA gradients need further refinement as they have important implications
for the mass balance and configuration of glacier ice (Fig. 24). The
thickness of the glacier cannot significantly exceed the elevation of
the upper ELA because ice sublimates rapidly above this elevation.
As a result, if the glacier were to reach this maximum thickness
at its accumulation zone, additional accumulation would yield significant outward growth, extending the snout of the glacier. The
gradient of the upper ELA determines the maximum thicknesses
of regions downslope of the accumulation zone—if the ELA were
perfectly horizontal, the maximum thickness would be constant as
a function of distance away from the accumulation zone (Fig. 24).
The gradient of the lower ELA determines the stability of the ice
at the snout. Thus, the elevations and gradients of the ELAs can be
manifested as significant changes in mass balance and in resulting
ice configurations.
Reconstructed ice sheet profiles for the Tharsis Montes glaciers
were created by Shean et al. (2005) based on a model proposed
by Denton and Hughes (1981). The reconstructions use MOLA topography for the base of the ice sheet along flow lines from the
calderas to the most distal moraines; because the profiles use the
maximum elongation of the glaciers, they provide estimates for the
maximum ice sheet thicknesses. At Ascraeus Mons, the iterative
steady state profile for perfectly plastic ice with a yield stress of
1.0 bars has a maximum thickness of 2.1 km and an average of
1.6 km; the maximum thicknesses at Pavonis and Arsia are 3.1 and
103
Fig. 24. A conceptual diagram showing how the position of horizontal ELAs can
affect the shape of a glacier, as well as how it responds to changes in accumulation and sublimation. Above: A glacier which has an accumulation zone comfortably
between the two ELAs can remain in steady-state. Increases or decreases in accumulation, however, have positive feedback; a slight increase in accumulation increases
the ratio of surface area at positive mass balance elevations to negative mass balance elevations, encouraging further growth. Alternatively, a small decrease in accumulation will decrease the ratio of surface area at positive mass balance elevations
to negative mass balance elevations. This in turn lowers the ratio of accumulation
to sublimation, leading to further glacial recession and a potential collapse of the
glacier. Neither of these deviations from steady-state would necessarily change the
shape or slope of the glacier. If, however, the glacier grew until its accumulation
zone reached the elevation of the upper ELA, further growth would likely lead to
a change in the shape and slope of the glacier. Because accumulation above the
upper ELA would rapidly sublimate, the highest elevation at which the glacier contacts the mountain cannot change significantly. Instead, additional accumulation will
occur away from the flank of the mountain, causing the glacier to increase preferentially its length rather than its thickness. Again, the ratio of the surface area at
positive mass balance elevations to negative mass balance elevations will increase,
yielding additional growth. As this occurs, the shape of the glacier will change: it
will become much longer, and minimally thicker, decreasing its overall slope as it
gets larger.
3.2 km respectively (Shean et al., 2005). On the basis of the topographic profiles (Figs. 7A–7G), we see that the degraded flank unit
exceeds elevations of 3 km above Mars datum in some places. This
elevation matches the maximum elevations achieved by a glacier
with a parabolic shape, a lowest elevation of ∼1.2 km, and an extremely flat base. This supports the degraded flank unit region as a
candidate for the accumulation zone. It does not, however, strictly
define the ELA elevations or gradients; it suggests that at Ascraeus,
3 km above Mars datum had positive mass balance and was thus
between the upper and lower ELA. Future work in this area may
provide the crucial information necessary to establish the ELAs and
mass balance of these tropical mountain glaciers, and subsequently
decode the climate signal left by the terminal moraines.
7. Hawaii as a terrestrial analog
7.1. Mesoscale models
Results from mesoscale atmospheric models may help limit
uncertainty in the ARD and constrain former ELAs. A mesoscale
model of atmospheric circulation at Arsia Mons by Rafkin et al.
(2002) shows dust transport using a simulation of the thermal circulation. The simulation, which reflects some seasonal variations,
predicts strong adiabatic cooling of the air as it rises up the western flank of the mountain, with a small vortex showing the wind
changing directions as it rises over ∼13 km in elevation. The more
dominant flow, however, becomes horizontally divergent at the top
104
S.J. Kadish et al. / Icarus 197 (2008) 84–109
Fig. 25. Mesoscale models of the atmospheric circulation at Hawaii (modified from Garrett, 1980) and Arsia Mons (modified from Rafkin et al., 2002). In the Hawaii model,
differential heating of the land and sea during the day drive combined anabatic and sea breeze winds. The air cools adiabatically as it carries moisture up the eastern flank
of Hawaii where it eventually encounters the trade wind inversion at an elevation of ∼2200 m. The majority of orographic clouds cannot rise above this level, and thus
precipitation is heaviest at this elevation. The inversion also causes a return flow, which is strong enough to overcome weak trade winds. At higher elevations, between 3000
and 4000 m, wind shear occurs from westerlies overlying the trade winds. The Arsia model, which was created to show dust mixing and transportation, reveals some similar
atmospheric behaviors. The arrows in the figure indicate wind velocity vectors in the plane of the page. Below an elevation of 10 km, a strong westerly wind blows up the
flank of the mountain. The air in this rising branch is cooled adiabatically. Between 10 and 15 km, a return flow can be seen which overcomes the westerly wind. Just above
this, at ∼20 km, the westerly wind again dominates. This is overlain by easterly winds at ∼25 km, causing wind shear between the strata. Although the Rafkin et al. (2002)
model is run without water vapor in the air, they note that it would, “. . . not be unexpected for a water-ice cloud to develop at the top of the thermal circulation” (Rafkin
et al., 2002), and orographic clouds have been observed in MOC images on the Tharsis Montes and Olympus Mons (see MOC release No. MOC2-144).
of the thermal circulation, cooling to 135 K at 30 km in elevation. Rafkin et al. (2002) note that, at this height, water-ice clouds
would be expected to form. A high wind shear zone would also result from easterlies overlying westerlies at 40–50 km in elevation
(Rafkin et al., 2002).
The composition, structure, and general circulation patterns of
the atmosphere of Mars differ substantially from those of Earth
(Zurek et al., 1992) and thus much caution must be exercised in
the analysis of terrestrial examples and applications to Mars. Terrestrial analogues may, however, help to improve mesoscale models on Mars, providing empirical data that could constrain better
the range of possible local circulation patterns on Mars. Numerous
mesoscale models of the atmospheric circulation at Hawaii (e.g.
Chen and Feng, 2001; Feng and Chen, 1998; Wang et al., 1998;
Garrett, 1980) have a number of features in common with the
Rafkin et al. model for Mars, including adiabatic cooling of the
air ascending the windward side of the mountain (Chen and Feng,
2001), orographic cloud formation (Chen and Feng, 2001; Feng and
Chen, 1998; Wang et al., 1998; Garrett, 1980), vertical vortices with
an overturning of the air and a return flow, associated with an inversion layer (Chen and Feng, 1995, 2001; Feng and Chen, 1998;
Garrett, 1980), and wind shear at high elevations due to opposing
wind directions overlying each other (Garrett, 1980; Fig. 25).
7.2. Microclimate zonation
Terrestrial examples can also be informative in the context of
the influence of topography on broad circulation patterns and how
these might map out into lateral and vertical microclimate zones
that affect geological processes. Microclimates on Mars, like on
Earth, would result from variations in local climatic and geographic
factors such as precipitation, insolation, elevation, atmospheric circulation, topography and temperature. Martian microclimates have
been suggested to exist within gullies and other depressions which
would allow for summer melting of frost accumulated during the
winter (Hecht, 2002). Marchant and Head (2004) have described
the microclimate zones in the Antarctic Dry Valleys (ADV). Using
the ADV as a terrestrial analog for Mars, they suggest that polygon
morphology may help define martian climate zones, and that the
presence of microclimates on Mars could have left a geomorphic
record of climate change on Mars which would aid in explaining
the evolution of periglacial features we have observed (Marchant
and Head, 2007).
Efforts to identify microclimates on and around the Tharsis
Montes will benefit greatly from an understanding of microclimate
formation on large terrestrial volcanoes. Although considerably
smaller than the Tharsis Montes, the volcanic island of Hawaii contains the largest concentration of microclimates on Earth, including both arid and periglacial environments (Fig. 26). A number of
factors, including Hawaii’s geographical setting, topographic relief,
geology, and local atmospheric circulation, make it a strong candidate for studying microclimates on the Tharsis Montes (Kadish and
Head, 2007).
First, Hawaii experiences a predominant wind direction from
the northeast trade winds. This prevailing wind is present 80–95%
of the summer months, but can diminish to only 50% in the winter
(Juvik and Nullet, 1994; Leopold, 1949). This mimics the seasonal
variability of atmospheric models for the Tharsis region (Forget
et al., 2006). Second, the unidirectional movement of atmospheric
moisture from northeast to southwest, coupled with a substantial
rise in elevation due to Mauna Kea (4205 m) and Mauna Loa (4169
m), create a pronounced orographic effect. The elevation change
on Hawaii is also responsible for steep gradients in temperature,
atmospheric humidity, and insolation, each of which contributes
to an altitudinal zonation of climate (Juvik and Nullet, 1994;
Loope and Giambelluca, 1998). Anabatic winds carry moisture up
the eastern flank of the mountain until they encounter the trade
wind inversion. Clouds ascending the flank are unable to rise above
the inversion layer at an elevation of ∼2200 m, forcing precipitation on the windward side. The leeward side is dominated by
katabatic winds coupled with weakened dry trade winds (Chen
and Feng, 1995; Leopold, 1949; Mendonca and Iwaoka, 1969). On
Ascraeus Mons tropical mountain glaciation
Fig. 26. A map of the distribution of microclimates in Hawaii modified from
Giambelluca and Sanderson (1993). Periglacial conditions exist at the top of Mauna
Loa and Mauna Kea. Arid climates are restricted to the northwest side of the mountain, which is the leeward side due to the prevailing NE trade winds.
Mars, the orographic effect occurred as a volatile-rich air mass
moved from west to east and rose upon encountering Ascraeus
Mons (18,209 m), Pavonis Mons (14,037 m) and Arsia Mons (17,779
m). Martian atmospheric studies have identified a temperature
inversion at an elevation of ∼10 km (e.g. Haberle et al., 1999;
Schofield et al., 1997); this inversion may mark the altitude of
clouds observed in Pathfinder images (Smith et al., 1997). The role
of katabatic winds on Mars has also been firmly established (e.g.
Howard, 2000; Magalhaes and Gierasch, 1982; Malin et al., 1998)
and the diurnal behavior of these has been modeled in the Tharsis
region (Rafkin et al., 2001). Third, like the Tharsis Montes, Hawaii
is capable of supporting localized glaciations; during the middle
and late Pleistocene, Hawaii experienced three or four sequential glaciations over a span of 160 kyr. The most recent of these
glaciers, lasting from 40,000 to 13,000 BP, had a maximum area
of 70 km2 and was 100 m thick (Loope and Giambelluca, 1998;
Porter, 1979; Wolfe et al., 1997). As a result, terminal, lateral, and
ground moraines are all observed on Hawaii, as well as morphologies suggestive of subglacial volcanism (Gregory and Wentworth,
1937; Porter, 1979; Wentworth and Powers, 1941).
These similarities between Hawaii and the Tharsis Montes, particularly in topography and local atmospheric circulation—two of
the most important factors that contribute to the zonation of
microclimates—support the notion that the abundance of microclimates on Hawaii also existed on the Tharsis Montes. Microclimates,
which have received minimal attention on Mars (e.g. Hecht, 2002;
Kadish and Head, 2007; Mangold, 2003; Marchant and Head, 2004,
2007), are an important aspect of the history of the martian tropical mountain glaciers; an understanding of the altitudinal zonation of climate on the Tharsis Montes would aid our investigation
of how the glaciers grew and receded in response to local envi-
105
Fig. 27. The N001 and P001 obliquity histories from Laskar et al. (2004). The age of
the Ascraeus Mons fan-shaped deposit, near the boundary between the Mid- and
Late-Amazonian, is more consistent with scenario P001, which puts Mars at high
obliquity (>45◦ ) at 250 Myr ago.
ronmental conditions. Future attempts to delineate microclimates
on Mars should take advantage of Hawaii and other terrestrial
analogs in considering: (1) local accumulations of frost, ice, and
snow, which may provide information on ELA elevations, (2) juxtaposition of desert/non-desert conditions such as those seen at
the perimeter of the Ka’u Desert on Kilauea Volcano, Hawaii,
(3) changes in the shape and orientation of eolian landforms such
as dunes, which provide information about local wind directions,
(4) spatial and temporal variations in surface moisture and temperature, which would produce distinct surface weathering and
spectral characteristics, (5) differences in alteration and weathering near calderas and vents due to vog, and (6) geometries which
greatly alter the insolation incident on the surface on diurnal and
seasonal timescales.
8. Climate change
The position of the ridged facies and the age of the deposit provide important information about the spin-axis obliquity and the
transportation of volatiles to low latitudes. The obliquity histories
modeled by Laskar et al. (2004) are extremely sensitive to initial
conditions and, while the solutions are robust from 20 Ma until the
present, they vary significantly between 250 and 20 Ma due to the
chaotic nature of the solutions. If the age of the Ascraeus Mons
FSD is less than 250 Myr, then the presence of tropical mountain glaciers means that Mars would have been at high obliquity
for some time during this period. Fastook et al. (2006) suggested
that two of the Laskar et al. obliquity histories are most likely:
P000 and N001 (Fig. 27). However, only one of these, P000, has
Mars at high obliquity at the beginning of the Late-Amazonian.
Fastook et al. (2006) point out that the most significant difference between P000 and N001 is that N001 has only 4 distinct ice
sheet growth events, whereas P000 has numerous growth events,
106
S.J. Kadish et al. / Icarus 197 (2008) 84–109
which may correspond to the high frequency of moraines seen at
Pavonis and Arsia. The age of the Ascraeus FSD, near the Mid- to
Late-Amazonian boundary, suggests that ice sheet growth occurred
between 380 and 250 Ma, which is more consistent with scenario
P000 than N001.
The small size of the Ascraeus FSD relative to those at Arsia
and Pavonis, and the more limited number of moraines at Ascraeus
Mons, can be explained in the context of the model by Forget et al.
(2006). These GCMs predict a strong northwesterly wind direction
in the Tharsis region during the northern hemisphere summer at
45◦ obliquity. In their model, enhanced polar ice sublimation allows volatiles to accumulate in specific regions at low latitudes.
In the Tharsis region, the atmosphere, carrying H2 O, moved from
west to east, and rose upon encountering the slopes of the volcanoes, inducing adiabatic cooling. This caused H2 O to precipitate as
snow and ice—from 30 to 70 mm/yr—on the western flanks, allowing glaciers several kilometers thick to form on thousand-year
timescales (Forget et al., 2006). Forget et al. (2006) also show that
Ascraeus and Olympus only receive precipitation during the northern summer, whereas accumulation of ice can occur throughout
the year at Pavonis and Arsia. During other seasons, Ascraeus and
Olympus are exposed to weaker winds while Pavonis and Arsia
experience precipitation from a symmetrical southern hemisphere
monsoon circulation which occurs throughout the southern hemisphere spring and summer (Forget et al., 2006).
In light of this, we interpret the limited number of moraines
and small size of the FSD at Ascraeus Mons as the result of any
combination of the following effects: (1) Moraines can only form
when the glacier is able to remain at equilibrium for long periods of time. However, even during periods of high obliquity,
Ascraeus Mons only receives precipitation for half of the year—it
experiences only ablation with no accumulation during that time.
This may hinder maintaining a net mass balance of zero for extended timescales, therefore preventing the Ascraeus Mons glacier
from creating numerous ridges. (2) The elevation of the Ascraeus
FSD, being lower than those of Pavonis and Arsia, means that
the glacier at Ascraeus Mons had a larger percentage of its surface area below the lower ELA. This suggests that a fractionally
larger proportion experienced negative mass balance, which could
hinder its ability to stay at equilibrium. (3) The transportation
of volatiles to Ascraeus Mons is somewhat inhibited by Olympus Mons. This can be understood by looking at epithermal neutron data collected by the Mars Odyssey gamma-ray spectrometer
(Fig. 28). These data indicate the presence of near-surface hydrogen, which has been interpreted to represent water-equivalent hydrogen (WEH) and the likely existence of H2 O ground ice and/or
hydrated minerals (e.g. Boynton et al., 2002; Feldman et al., 2002;
Mellon et al., 2004). An analysis of these data for the Tharsis region has shown a WEH range of 2–8 wt% (Elphic et al., 2005).
The epithermal neutron data reveal a higher concentration of
hydrogen on the western slopes of the Tharsis Montes and Olympus Mons than on the eastern slopes, consistent with ground
ice/hydrated minerals being concentrated on the western sides of
the Tharsis Montes (Fig. 28). This is interpreted to be caused by the
presence of an orographic effect—the snow and ice is precipitated
on the western flanks as it adiabatically cools and rises, leaving the
atmosphere dry as it passes over the summits of the volcanoes.
From this, we can see that the westerly wind will first encounter
Olympus Mons as it approaches the Tharsis Montes. Being the farthest north, Ascraeus Mons lies directly in the ‘rain shadow’ of
Olympus Mons. The effect is readily apparent in the epithermal
neutron data, showing a tail of low hydrogen extending east of
Olympus Mons. Thus, there is a paucity of atmospheric volatiles
when the westerly wind reaches Ascraeus Mons. Pavonis and Arsia Mons, south of Ascraeus, are consequently unaffected by the
Olympus Mons ‘rain shadow.’
Fig. 28. A map of the distribution of epithermal neutrons in the Tharsis region, as
measured by the gamma ray and neutron spectrometer on Mars Odyssey. The data
are superimposed on a hillshade map generated by MOLA data (128 pixel/degree).
A predominant westerly wind at an elevation of 2 km established by Forget et al.
(2006) during the northern summer (Ls = 125◦ to 155◦ ) at periods of high obliquity carried volatiles to the western flanks of the Tharsis Montes. A pronounced
orographic effect prevented much of the volatiles from reaching the eastern sides
of the volcanoes. As a result, we observe high epithermal neutron counts east of
the Tharsis Montes, which correspond to low hydrogen concentration at the surface. The western sides, however, show low epithermal neutron counts, possibly
suggesting high water-equivalent hydrogen concentrations. We also see the effect of
a potential orographic effect east of Olympus Mons. This may be partially responsible for the small size of the Ascraeus Mons fan-shaped deposit; by the time the
airmass reached the Tharsis Montes, much of the volatiles in the north had already
been precipitated, resulting in lower accumulation rates at Ascraeus compared to
Pavonis and Arsia.
9. Conclusions
9.1. Cold-based glacial geomorphologies
We interpret the following features to have formed via deposition from glacial activity: (1) Multiple sets of ridged facies, the
largest of which defines the distal margin of the Ascraeus FSD,
(2) a knobby facies which covers the majority of the southern half
of the deposit as well as some portions of the northern half, and
(3) a hummocky-like terrain containing complex ridges. These geomorphologies are similar to those observed in the FSDs at Pavonis
and Arsia, and have been previously interpreted as glacial in origin. The concentric ridges composing the ridged facies, which we
interpret as drop moraines, formed during periods of glacial stability, and as such, mark the positions of paleoglacial standstills.
This provides insight into the climatic history of Mars during the
Amazonian. The numerous sets of ridges with their various placements and orientations, and the overlap, merging, and crossing of
ridges within individual sets suggests a complex history with multiple re-advances, and possibly multiple glaciations. The changing
radius of curvature of the outer ridged facies is a reflection of the
broad topography of the region, which generally slopes downhill
heading north. The knobby facies, interpreted here as a sublimation till, shows variations in knob distribution and density. We
argue that inhomogeneities in the appearance of the knobby facies and the non-uniform distribution of knobs over the FSD result
from an unequal supply of debris within the glacier, and the absence or subdued appearance of knobs in regions is due to burial
by subsequent lava flows or dust.
Ascraeus Mons tropical mountain glaciation
9.2. Lava–ice interactions
During periods of glaciation, volcanic activity led to lava–ice
interactions responsible for producing a number of the observed
geomorphologies including: (1) An arcuate scarp which runs parallel to the outer ridged facies, (2) mountainous terrain adjacent
to the arcuate scarp, (3) flat-topped ridges with leveed channels
extending from the southwest flank of Ascraeus Mons, and (4) a
mesa-like plateau near the center of the deposit. The inward-facing
arcuate scarp has a curvature which mimics that of the outer
ridges, supporting the notion that it formed via lava flows which
circumvented and embayed the distal rim of the extant glacier.
These same lava flows were responsible for producing the mountainous terrain on the rim of the scarp. The flows are likely genetically linked to the flat-topped ridges extending from the source of
the lava: the southwest flank of Ascraeus. The flat-topped ridges
and leveed channels end abruptly at the southern end of the depression that contains the majority of the FSD. The termination of
the flat-topped ridges, which resulted from the lava flows encountering the massive glacier, creates a somewhat chaotic extension
of the arcuate scarp. The presence of the outer ridged facies located beyond the arcuate scarp and the extent of the knobby facies
covering the flat-topped ridges is an indication of glacial advance
and/or larger distinct glaciations subsequent to the lava flows. The
plateau feature is interpreted to have formed via subglacial volcanism which, like the flat-topped ridges, is also partially covered by
the knobby facies. This feature appears similar to terrestrial table
mountains.
9.3. Accumulation of ice and snow
The results of Forget et al. (2006) illustrate that global climate
models during periods of high obliquity (>45◦ ) show a strong
westerly wind across Tharsis. This wind carried volatiles that were
transported equatorward from the polar caps. As the air encountered the steep slopes of Ascraeus and the other Tharsis Montes,
it rose and cooled adiabatically, forcing the volatiles to precipitate on the western flanks of the volcanoes (Forget et al., 2006).
The accumulation of ice and snow predominantly on the western
flanks of the Tharsis Montes is consistent with regional epithermal
neutron measurements, which have identified a noticeably higher
concentration of hydrogen—likely water equivalent hydrogen—on
the western flanks than on the eastern flanks. The degraded flank
unit is the most probable candidate for the accumulation zone of
ice and snow at Ascraeus Mons. Further work on mesoscale atmospheric models will be crucial for establishing former accumulation
zones with greater certainty. These models will help predict a variety of factors including cloud level formation, local wind circulation patterns around the flanks of the volcanoes, and precipitation
distribution.
9.4. Glacial models
Adaptations of a terrestrial ice-sheet model by Fastook et al.
(2005) are consistent with repeated growth and decay of Tharsis Montes glaciers, including periods of relative standstill, during
which time the outer ridged facies could have formed. New iterations of the model have improved the accuracy of the output
(Fastook et al., 2007), better constraining the maximum growth of
the glaciers to the size and shape of the FSDs. A number of the
model’s parameters remain poorly constrained, and as these values become more accurately defined, the output of the models will
hopefully become more robust. Perhaps the most important elements which would benefit from revision are the ARD and the elevation and gradient of the ELAs. An accurate assessment of these
would require more knowledge of the martian paleoclimate than
107
we currently have, but once obtained, these would vastly improve
our knowledge of the mass balance of the glacier, the elevation at
which they accumulated, how they responded to changes in precipitation and sublimation, and how they grew and changed shape
in response to topography.
9.5. The Tharsis tropical mountain glaciers and martian climate change
We know, based on surface morphology and mineralogy, that
the martian climate has been notably different in the past. Models are helping us to define better the significance of the various
surface features by recreating the paleoconditions that may have
existed since the Noachian. Understanding the evolution of the
Tharsis tropical mountain glaciers is a small but crucial step toward reconstructing the history of martian climate change. The
Mid- to Late-Amazonian age of the Ascraeus FSD determined in
this paper, in concurrence with the similar ages of the Pavonis and
Arsia FSDs, confirms the presence of stable surficial ice near the
equator within the last few hundred Myr. This notion is extremely
informative when attempting to reveal the obliquity and climate
history of Mars; the obliquity must have been above 45◦ at some
point during the Mid- to Late-Amazonian in order to sublimate
volatiles at the poles and transport them to the equator. Such a
dramatic climate change in the geologically recent past is an indication of just how rapidly the conditions can change, welcoming
substantially different environments throughout martian history.
Continued analysis of the martian surface, as well as atmospheric
studies, will ultimately contribute to establishing the complete history of martian climate change.
9.6. Future work
Although we have described the vast majority of the geomorphologies within the Ascraeus FSD, there remain some outstanding
issues concerning the appearance and relationships between the
features that require further analysis. The first of these concerns
the presence of a number of lobate scarps in the northern end of
the deposit which appear to be randomly oriented (Figs. 9, blue
lines; 11F–11H); in some cases the features form closed loops,
while in others they appear as curvilinear segments that suddenly
taper off at both ends. These scarps are distinct from the arcuate
scarp, having a smaller relief (∼30–70 m) and do not mimic the
shape of the distal rim of a glacier. They are all located east of
the arcuate scarp, and the knobby facies superposes them in some
cases. The second issue is the set of ridges which are oriented almost orthogonal to what we would expect based on their location
within the deposit if they are, in fact, drop moraines. Although
they may, as discussed, represent the advance of a distinct glaciation or a separate lobe of the glacier, they may also indicate that
the concentric ridges can form via more than one process. The final
issue which we would like to understand better is the relationship between the ridged facies and the complex curved ridges and
the relationship between the hummocky terrain and the knobby
facies. How are the complex curved ridges related to the ridged facies? Are they genetically related? Is the knobby facies simply a
vast, partially buried section of the hummocky terrain? We have
made arguments above that address these relationships and the
various formation mechanisms, but as we experienced throughout
this study, the acquisition of higher resolution data constantly improves and strengthens our interpretations. These relationships are
complex and future analysis may either reinforce our conclusions
or alter our understanding.
Acknowledgments
We gratefully acknowledge financial support from the NASA
Mars Data Analysis Program (NNG04GJ99G), the NASA Mars Ex-
108
S.J. Kadish et al. / Icarus 197 (2008) 84–109
press Participating Scientist Program (JPL1237163) and the NASA
AISR program (NNG05GA61G) (to J.W.H.). We appreciate the work
done by the science and engineering teams of all the instruments
and missions from which data were incorporated into this study.
Additionally, we thank Caleb Fassett for significant help with data
preparation, processing and analysis, James Dickson for assistance
with data preparation, and James Fastook, David Shean, and Francois Forget for productive discussions on glacial processes and martian climate change.
References
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, 85–101.
Benn, D.I., Evans, D.J.A., 1998. Glaciers and Glaciation. Arnold, London, UK.
Benn, D.I., Owen, L.A., Osmaston, H.A., Seltzer, G.O., Porter, S.C., Mark, B., 2005. Reconstruction of equilibrium-line altitudes for tropical and sub-tropical glaciers.
Quatern. Int. 138–139, 8–21.
Boulton, G.S., Eyles, N., 1979. Sedimentation by valley glaciers: A model and genetic
classification. In: Schluchter, C. (Ed.), Moraines and Varves. A.A. Balkema, Brookfield, Vermont.
Boynton, W.V., and 24 colleagues, 2002. Distribution of hydrogen in the near surface
of Mars: Evidence for subsurface ice deposits. Science 297, 81–85.
Carr, M.H., 1975. Geologic map of the Tharsis Quadrangle of Mars. US Geol. Surv.
Misc. Invest. Ser. Map I-893.
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.
Chen, Y., Feng, J., 1995. The influences of inversion height on precipitation and airflow over the Island of Hawaii. Mon. Weather Rev. 123, 1660–1676.
Chen, Y., Feng, J., 2001. Numerical simulations of airflow and cloud distributions
over the windward side of the Island of Hawaii. Part I. The effects of trade wind
inversion. Mon. Weather Rev. 129, 1117–1134.
Christensen, P.R., and 21 colleagues, 2003. Morphology and composition of the surface of Mars: Mars Odyssey THEMIS results. Science 300, 2056–2061.
Crumpler, L.S., Aubele, J.C., 1978. Structural evolution of Arsia Mons, Pavonis Mons,
and Ascraeus Mons Tharsis region of Mars. Icarus 34, 496–511.
Crumpler, L.S., Aubele, J.C., Head, J.W., 1991. Calderas on Mars: Models of formation
for the Arsia-type. Lunar Planet. Sci. XXII. Abstract 269.
Crumpler, L.S., Head, J.W., Aubele, J.C., 1996. Calderas on Mars: Characteristics, structure, and associated flank deformation. Geol. Soc. London Spec. Publ. 110, 307–
348.
Denton, G.H., Hughes, T.J., 1981. The Last Great Ice Sheets. John Wiley, Hoboken, NJ.
Edgett, K.S., 1989. The lobate features west of each of the Tharsis Montes, Mars:
A re-evaluation of their origins. Lunar Planet. Sci. XX. Abstract 256.
Edgett, K.S., Zimbelman, J.R., Branstrator, J.W., 1987. The geology of Pavonis Mons,
Mars. Lunar Planet. Sci. XVIII. Abstract 254.
Edgett, K.S., Butler, B.J., Zimbelman, J.R., Hamilton, V.E., 1997. Geologic context of
the Mars radar “Stealth” region in southwestern Tharsis. J. Geophys. Res. 102,
21545–21568.
Elphic, R.C., Feldman, W.C., Prettyman, T.H., Tokar, R.L., Lawrence, D.J., Head, J.W.,
Maurice, S., 2005. Mars Odyssey neutron spectrometer water-equivalent hydrogen: Comparison with glacial landforms on Tharsis. Lunar Planet. Sci. XXXVI.
Abstract 1805.
Fastook, J.L., Head, J.W., Marchant, D.R., Shean, D.E., 2004. Ice sheet modeling: Terrestrial background and application to Arsia Mons lobate deposit, Mars. Lunar
Planet. Sci. XXXV. Abstract 1452.
Fastook, J.L., Head, J.W., Marchant, D.R., Shean, D.E., 2005. Ice sheet modeling: Mass
balance relationships for map-plane ice sheet reconstruction: Application to
Tharsis Montes glaciation. Lunar Planet. Sci. XXXVI. Abstract 1212.
Fastook, J.L., Shean, D.E., Head, J.W., Marchant, D.R., 2006. Ice sheet modeling during
high-obliquity climates on Mars: Application to Tharsis Montes tropical mountain glaciation. Lunar Planet. Sci. XXXVII. Abstract 1794.
Fastook, J.L., Head, J.W., Marchant, D.R., 2007. Tharsis Montes ice sheet models at
high obliquity driven by GCM results. Lunar Planet. Sci. XXXVIII. Abstract 1119.
Feldman, W.C., and 12 colleagues, 2002. Global distribution of neutrons from Mars:
Results from Mars Odyssey. Science 297, 75–78.
Feng, J., Chen, Y., 1998. Evolution of katabatic flow on the island of Hawaii on August
1990. Mon. Weather Rev. 126, 2185–2199.
Forget, F., Haberle, R.M., Montmessin, F., Levrard, B., Head, J.W., 2006. Formation
of glaciers on Mars by atmospheric precipitation at high obliquity. Science 311,
368–371.
Garrett, A.J., 1980. Orographic cloud over the eastern slopes of Mauna Loa Volcano,
Hawaii, related to insolation and wind. Mon. Weather Rev. 108, 931–941.
Giambelluca, T., Sanderson, M., 1993. The water balance and climatic classification.
In: Sanderson, M. (Ed.), Prevailing Trade Winds, Weather and Climate in Hawai’i.
Univ. of Hawaii Press, Hawai’i, Honolulu, pp. 56–72.
Gregory, H.E., Wentworth, C.K., 1937. General features and glacial geology of Mauna
Kea, Hawaii. Geol. Soc. Am. Bull. 48, 1719–1742.
Greve, R., 2000. Waxing and waning of the perennial north polar H2 O ice cap of
Mars over obliquity cycles. Icarus 144, 419–431.
Greve, R., Mahajan, R.A., 2005. Influence of ice rheology and dust content on the
dynamics of the north-polar cap of Mars. Icarus 174, 475–485.
Grier, J.A., Bottke, W.F., Hartmann, W.K., Berman, D.C., 2001. Mars: Chronological
studies of the large volcanoes in Tharsis. Lunar Planet. Sci. XXXII. Abstract 1823.
Haberle, R.M., Joshi, M.M., Murphy, J.R., Barnes, J.R., Schofield, J.T., Wilson, G., LopezValverde, M., Hollingsworth, J.L., Bridger, A.F.C., Schaeffer, J., 1999. General circulation model simulations of the Mars Pathfinder atmospheric structure investigation/meteorology data. J. Geophys. Res. 104, 8957–8974.
Hartmann, W.K., 2005. Martian cratering. 8. Isochron refinement and the chronology
of Mars. Icarus 174, 294–320.
Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the evolution of Mars.
Space Sci. Rev. 96, 165–194.
Hartmann, W.K., Malin, M., McEwen, A., Carr, M., Soderblom, L., Thomas, P., Danielson, E., James, P., Veverka, J., 1999. Evidence for recent volcanism on Mars from
crater counts. Nature 397, 586–589.
Hättestrand, C., Kleman, J., 1999. Ribbed moraine formation. Quaternary Sci. Rev. 18,
43–61.
Head, J.W., Marchant, D.R., 2003. Cold-based mountain glaciers on Mars: Western
Arsia Mons. Geology 31, 641–644.
Head, J.W., Solomon, S.C., 1981. Tectonic evolution of the terrestrial planets. Science 213, 62–76.
Head, J.W., Wilson, L., 2007. Heat transfer in volcano–ice interactions on Mars:
Synthesis of environments and implications for processes and landforms. Ann.
Glaciol. 45, 1–13.
Head, J.W., Wilson, L., Mitchell, K.L., 2003. Generation of recent massive water floods at Cerberus Fossae, Mars by dike emplacement, cryospheric cracking, and confined aquifer groundwater release. Geophys. Res. Lett. 30 (11),
doi:10.1029/2003GL017135. 1577.
Head, J.W., and 13 colleagues, 2005. Tropical to mid-latitude snow and ice accumulation, flow and glaciation on Mars. Nature 434, 346–351.
Hecht, M.H., 2002. Metastability of liquid water on Mars. Icarus 156, 373–386.
Helgason, J., 1999. Formation of Olympus Mons and the aureole-escarpment problem on Mars. Geology 27 (3), 231–234.
Holdsworth, G., Bull, C., 1970. The flow law of cold ice: Investigations on Meserve
Glacier, Antarctica. IASH Publ. 86, 204–216.
Howard, A.D., 2000. The role of Eolian processes in forming surface features of the
martian polar layered deposits. Icarus 144, 267–288.
Huddart, D., Bennett, M.R., Glasser, N.F., 1999. Morphology and sedimentology of a
high-arctic esker system: Vegbreen, Svalbard. Boreas 28, 253–273.
Hvidberg, C.S., 2006. The north polar layered deposits, Mars: Topography, flow and
implications for timescales. In: 4th Mars Polar Sci. Conf. Abstract 8101.
Ivanov, A.B., Muhleman, D.O., 2000. The role of sublimation for the formation of
the northern ice cap: Results from the Mars orbited laser altimeter. Icarus 144,
436–448.
Juvik, J.O., Nullet, D., 1994. A climate transect through tropical montane rain forest
in Hawaii. J. Appl. Meteorol. 33, 1304–1312.
Kadish, S.J., Head, J.W., 2007. Microclimates on the Tharsis Rise and volcanoes: Orographic effects and geological processes. Lunar Planet. Sci. XXXVIII. Abstract
1120.
Kleman, J., Hättestrand, C., 1999. Frozen-bed Fennoscandian and Laurentide ice
sheets during the Last Glacial Maximum. Nature 402, 63–66.
Koch, J., Kilian, R., 2005. ‘Little Ice Age’ glacier fluctuations, Gran Campo Nevado,
southernmost Chile. Holocene 15, 20–28.
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.
Leopold, L.B., 1949. The interaction of trade wind and Sea breeze, Hawaii. J. Atmos.
Sci. 6, 312–320.
Leovy, C., 2001. Weather and climate on Mars. Nature 412, 245–249.
Loope, L.L., Giambelluca, T.W., 1998. Vulnerability of island tropical montane cloud
forests to climate change, with special reference to East Maui, Hawaii. Climatic
Change 39, 363–377.
Lucchitta, B.K., 1981. Mars and Earth: Comparison of cold-climate features.
Icarus 45, 264–303.
Magalhaes, J., Gierasch, P., 1982. A model of martian slope winds: Implications for
eolian transport. J. Geophys. Res. 87, 9976–9984.
Malin, M.C., and 15 colleagues, 1998. Early Views of the martian surface from the
Mars orbiter camera of Mars Global Surveyor. Science 279, 1681–1685.
Mangold, N., 2003. Geomorphic analysis of lobate debris aprons on Mars at Mars
Orbiter Camera scale: Evidence for ice sublimation initiated by fractures. J. Geophys. Res. 108, doi:10.1029/2002JE001885.
Marchant, D.R., Head, J.W., 2004. Microclimate zones in the dry valleys of Antarctica: Implications for landscape evolution and climate change. Lunar Planet.
Sci. XXXV. Abstract 1405.
Marchant, D.R., Head, J.W., 2007. Antarctic dry valleys: Microclimate zonation, variable geomorphic processes, and implication for assessing climate change on
Mars. Icarus 192, 187–222.
Ascraeus Mons tropical mountain glaciation
Marchant, D.R., Lewis, A.R., Phillips, W.M., Moore, E.J., Souchez, R.A., Denton, G.H.,
Sugden, D.E., Potter Jr., N., Landis, G.P., 2002. Formation of patterned ground
and sublimation till over Miocene glacier ice in Beacon Valley, southern Victoria
Land, Antarctica. Geol. Soc. Am. Bull. 114, 718–730.
Matthews, J.A., Cornish, R., Shakesby, R.A., 1979. Saw-tooth moraines in front of Bödalsbreen, southern Norway. J. Glaciol. 22, 535–546.
McEwen, A.S., and 14 colleagues, 2007. Mars Reconnaissance Orbiter’s High Resolution Imaging Science Experiment (HiRISE). J. Geophys. Res. 112, doi:10.1029/
2005JE002605. E05S02.
Meier, M.F., 1965. Glaciers and climate. In: Wright, H., Frey, D. (Eds.), The Quaternary
of the United States. Princeton University Press, Princeton, NJ, pp. 795–805.
Meierding, T.C., 1982. Late Pleistocene equilibrium-line altitudes in the Colorado
Front Range: A comparison of methods. Quaternary Res. 18, 289–310.
Mellon, M.T., Feldman, W.C., Prettyman, T.H., 2004. The presence and stability of
ground ice in the southern hemisphere of Mars. Icarus 169, 324–340.
Mendonca, B.G., Iwaoka, W.T., 1969. The trade wind inversion at the slopes of Mauna
Loa, Hawaii. J. Appl. Meteorol. 8, 213–219.
Milkovich, S.M., Head, J.W., 2003. Olympus Mons fan shaped deposit morphology:
Evidence for debris glaciers. In: 6th Int. Mars Conf. Abstract 3149.
Milkovich, S.M., Head, J.W., Marchant, D.R., 2006. 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. Icarus 181 (2), 388–407.
Mouginis-Mark, P.J., Zisk, S.H., Downs, G.S., 1982. Ancient and modern slopes in the
Tharsis region of Mars. Nature 297 (5867), 546–550.
Neukum, G., and 10 colleagues, 2004. Recent and episodic volcanic and glacial activity on Mars revealed by the High Resolution Stereo Camera. Nature 432,
971–979.
Porter, S.C., 1979. Hawaiian glacial ages. Quaternary Res. 12, 161–187.
Porter, S.C., 2001. Snowline depression in the tropics during the last glaciation. Quaternary Sci. Rev. 20, 1067–1091.
Price, R.J., 1966. Eskers near the Casement Glacier, Alaska. Geogr. Ann. A 48, 111–
125.
Price, R.J., 1970. Moraines at Fjallsjökull, Iceland. Arctic Alpine Res. 2, 27–42.
Rafkin, S.C.R., Haberle, R.M., Michaels, T.I., 2001. The Mars regional atmospheric
modeling system: Model description and selected simulations. Icarus 151, 228–
256.
Rafkin, S.C.R., Maria, M.R.V., Michaels, T.I., 2002. Simulation of the atmospheric
thermal circulation of a martian volcano using a mesoscale numerical model.
Nature 419, 697–699.
Schmidt, K.G., 2004. Modeling the mass balance of the north polar ice cap on Mars.
Lunar Planet. Sci. XXXV. Abstract 1554.
Schofield, J.T., Barnes, J.R., Crisp, D., Haberle, R.M., Larsen, S., Magalhaes, J.A., Murphy, J.R., Seiff, A., Wilson, G., 1997. The Mars Pathfinder Atmospheric Structure
Investigation/Meteorology (ASI/MET) Experiment. Science 278, 1752–1758.
Scott, D.H., Tanaka, K.L., 1981. Mars—Paleostratigraphic restoration of buried surfaces
in Tharsis montes. Icarus 45, 304–319.
Scott, D.H., Tanaka, K.L., 1986. Geological map of the western equatorial region of
Mars. US Geol. Surv. Misc. Invest. Ser. Map I-1802-A.
Scott, E.D., Wilson, L., 2000. Cyclical summit collapse events at Ascraeus Mons, Mars.
J. Geol. Soc. London 157, 1101–1106.
Scott, D.H., Zimbelman, J.R., 1995. Geological map of Arsia Mons Volcano, Mars. US
Geol. Surv. Misc. Invest. Ser. Map I-2480.
Scott, D.H., Dohm, J.M., Zimbelman, J.R., 1998. Geologic map of Pavonis Mons Volcano, Mars. US Geol. Surv. Misc. Invest. Ser. Map I-2561.
109
Shaw, J., 1977. Tills deposited in arid polar environments. Can. J. Earth Sci. 14, 1239–
1245.
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, doi:10.1029/2004JE002360. E05001.
Shean, D.E., Head, J.W., Kreslavsky, M., Neukum, G., HRSC Co-I Team, 2006. When
were glaciers present in Tharsis? Constraining age estimates for the Tharsis
Montes fan-shaped deposits. Lunar Planet. Sci. XXXVII. Abstract 2092.
Shean, D.E., Head, J.W., Fastook, J.L., Marchant, D.R., 2007. Recent glaciation at high
elevations on Arsia Mons, Mars: Implications for the formation and evolution
large tropical mountain glaciers. J. Geophys. Res. 112, doi:10.1029/2006JE002761.
E03004.
Smellie, J.L., Chapman, M.G., 2002. Introduction: Volcano–ice interaction on Earth
and Mars. Geol. Soc. London Spec. Publ. 202, 1–3.
Smith, P.H., and 28 colleagues, 1997. Results from the Mars Pathfinder Camera. Science 278, 1758–1765.
Smith, D.E., and 18 colleagues, 1999. The global topography of Mars and implication
for surface evolution. Science 284, 1495–1503.
Smith, D.E., and 23 colleagues, 2001. Mars Orbital Lased Altimeter: Experiment
summary after the first year of global mapping of Mars. J. Geophys. Res. 106,
23689–23722.
Tanaka, K.L., 1986. The stratigraphy of Mars. J. Geophys. Res. 91 (B13), E139–E158.
Wang, J., Juang, H.H., Kodama, K., Businger, S., Chen, Y., Partain, J., 1998. Application
of the NCEP regional spectral model to improve mesoscale weather forecasts in
Hawaii. Weather Forecast. 13, 560–575.
Wentworth, C.K., Powers, W.E., 1941. Multiple glaciation of Mauna Kea, Hawaii. Geol.
Soc. Am. Bull. 52, 1193–1218.
Williams, R.S., 1978. Geomorphic processes in Iceland and on Mars: A comparative
appraisal from orbital images. Geol. Soc. Am. Abstr. Progr. 10, 517.
Wilson, L., Head, J.W., 2002. Heat transfer and melting in subglacial basaltic volcanic eruptions: Implications for volcanic deposit morphology and meltwater
volumes. In: Smellie, J.L., Chapman, M.G. (Eds.), Volcano–Ice Interaction on Earth
and Mars, Geol. Soc. Spec. Publ. 202, pp. 5–26.
Wilson, L., Head, J.W., 2007a. Heat transfer in volcano–ice interactions on Earth.
Ann. Glaciol. 45, 83–86.
Wilson, L., Head, J.W., 2007b. Tropical mountain glaciers on Mars: Volcano–ice interactions and influences on glacial evolution. In: Seventh International Conference
on Mars. Abstract 3123.
Wolfe, E.W., Wise, W.S., Dalrymple, G.B., 1997. Geology and petrology of Mauna Kea
volcano, Hawaii: A study of postshield volcanism. USGS Professional Paper 1557,
pp. 1–129.
Zimbelman, J.R., 1984. Geologic interpretation of remote sensing data for the martian volcano Ascraeus Mons. Ph.D. dissertation, Arizona State University, Tempe,
AZ.
Zimbelman, J.R., Edgett, K.S., 1992. The Tharsis Montes, Mars—Comparison of volcanic and modified landforms. Lunar Planet. Sci. 22, 31–44.
Zuber, M.T., and 14 colleagues, 2000. Internal structure and early thermal evolution
of Mars from Mars Global Surveyor topography and gravity. Science 287, 1788–
1793.
Zurek, R.W., Barnes, J.R., Haberle, R.M., Pollack, J.B., Tillman, J.E., Leovy, C.B., 1992.
Dynamics of the atmosphere of Mars. In: Kieffer, H.H., Jakosky, B.M., Snyder,
C.W., Matthews, M.S. (Eds.), Mars. Univ. of Arizona Press, Tucson, AZ, pp. 835–
933.
Download