The volcanic history of Olympus Mons from paleo

Earth and Planetary Science Letters 363 (2013) 88–96
Contents lists available at SciVerse ScienceDirect
Earth and Planetary Science Letters
journal homepage: www.elsevier.com/locate/epsl
The volcanic history of Olympus Mons from paleo-topography
and flexural modeling
Ryan J. Isherwood, Lauren M. Jozwiak 1, Johanna C. Jansen, Jeffrey C. Andrews-Hanna n
Department of Geophysics and Center for Space Resources, Colorado School of Mines, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 25 July 2012
Received in revised form
11 December 2012
Accepted 13 December 2012
Editor: T. Spohn
Paleotopography and flexural modeling are here used to constrain the formation history and eruption
rates of Olympus Mons, the tallest shield volcano on Mars. The timing of the initiation of significant
edifice construction is constrained using lava flows whose paths deviate significantly from the downslope direction of the present-day flexural trough, and thus are classified as topographically discordant.
Flexural models are used to place limits on the fraction of Olympus Mons that could have been present
at the time of emplacement of one strongly discordant flow. Comparison of the predicted flexural
response with the paleotopography indicates that no more than 29–51% of the volume of Olympus
Mons could have been present at the time the discordant flow was emplaced. The end of the primary
edifice construction stage is constrained by the formation of the aureole deposits, which are inferred to
þ 0:05
þ 0:55
Ga for the discordant flow and 2:540:69
Ga for
post-date the bulk of the volcano. The ages of 3:670:10
the aureole deposit span the period during which the majority of Olympus Mons formed, a period of
þ 0:74
Gyr. The resulting eruption rate of 0.003–0.015 km3/yr is similar to that
approximately 1:130:65
observed in terrestrial hot-spot volcanism, supporting a similar geodynamic mechanism driving shieldforming volcanism on Earth and Mars. After this period, the rate of volcanic resurfacing dropped off
considerably, but low levels of volcanic activity have been maintained through the last several hundred
million years.
& 2013 Elsevier B.V. All rights reserved.
Keywords:
Mars
volcanism
paleotopography
flexure
1. Introduction
Olympus Mons is the largest known shield volcano in the Solar
System, standing an average of 21 km above the Martian datum and
up to 24 km above the surrounding plains. Olympus Mons is located
to the northwest of the Tharsis rise and the somewhat smaller
Tharsis Montes shields. As the largest single volcano on Mars, the
volcanic history of Olympus Mons has important implications for
the geodynamic history of Tharsis and Mars as a whole. Previous
studies have shown that the age for the majority of the volcanic
surface is 200 Ma (Basilevsky et al., 2006; Neukum et al., 2004;
Robbins et al., 2011; Werner, 2009). However, the ages of isolated
exposures of the surface date back to the Noachian and late
Hesperian (Neukum et al., 2004). Although this suggests a long
history of active volcanism, the specific volcanic history of Olympus
Mons is difficult to work out. The fundamental problem is that
crater retention ages date only the surface of the edifice, which is
dominated by the youngest flows and may have little bearing on the
age of the bulk of Olympus Mons.
n
Corresponding author. Tel.: þ1 303 273 3500.
E-mail address: jcahanna@mines.edu (J.C. Andrews-Hanna).
1
Now at: The Department of Geological Sciences, Brown University, USA.
0012-821X/$ - see front matter & 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.epsl.2012.12.020
In this study, a combination of paleo-topography, flexural
modeling and crater retention ages are used to investigate the
volcanic history of Olympus Mons. To constrain the onset of
volcanic loading, we focus on the prominent flexural trough
surrounding the edifice, resulting from the deformation of the
lithosphere by the volcanic load (Fig. 1). In Section 2, we identify
lava flows on the outer margins of the flexural trough that deviate
from the modern down-slope direction and thus pre-date the
flexural trough. Because these topographically discordant flows
formed prior to the topography of the flexural trough they must
predate the bulk of the edifice volume. In Section 3, thin-shell
spherical harmonic flexural models are used to evaluate what
fraction of the edifice volume would have been needed to redirect
one strongly discordant flow.
To constrain the end of the main edifice-construction phase,
we use the aureole deposits that are thought to have formed
when Olympus was similar to its present-day size (McGovern
et al., 2004a). The ages of the topographically discordant lava flow
and the aureole deposits thus bracket the primary construction
period of Olympus Mons. Section 4 uses crater size-frequency
distributions to estimate the ages of these features. Using these
constraints it is possible to constrain the eruption rate during the
time in which the majority of the Olympus edifice formed and
compare this with terrestrial volcanoes.
R.J. Isherwood et al. / Earth and Planetary Science Letters 363 (2013) 88–96
89
Fig. 1. (a) MOLA topography (Smith et al., 2001) context map of Olympus Mons and the surrounding flexural trough. (b) MOLA topography and contour map showing
the topographically discordant lava flow (arrows). (c) THEMIS daytime infrared image mosaic (Christensen et al., 2001) over the region shown in (b).
2. Paleo-topography analyses from lava flows on the flexural
trough
The concept of paleo-topography has been used in terrestrial
geodynamics to reconstruct the vertical motions of the lithosphere (e.g., Liu and Gurnis, 2010), but has seen less use in
planetary applications (Phillips et al., 2001). Our paleotopography
reconstructions are predicated on the fact that fluids (e.g., lava)
flow along the path of the steepest descent. Thus, the down-flow
direction of a lava flow should match the down-slope direction at the
time of its formation. Olympus Mons is surrounded by a large flexural
trough that has been partially infilled by concurrent volcanic eruptions (Fig. 1a). We surveyed the inward-facing flanks of the flexural
trough in Mars Orbiter Laser Altimeter (MOLA) topography (Smith
et al., 2001) to identify topographically discordant flows whose paths
deviate significantly from the down-slope direction. An analysis of six
90
R.J. Isherwood et al. / Earth and Planetary Science Letters 363 (2013) 88–96
typical flows in the region that did not display strong discordance
with the local topography, revealed that they follow sinuous paths
(Fig. 2), with a mean local deviation across all flows of 187121 from
the regional down-slope direction, as determined by fitting a plane to
the local topography. This value was adopted as a threshold angle
between the down-flow and down-slope directions in order for a flow
to be classified as topographically discordant. Flows exceeding the
mean deviance of 181 are potentially topographically discordant,
while those exceeding the mean plus 1-r deviance of 301 are strongly
discordant. Without precise knowledge of the paleotopography at the
time the non-discordant flows formed, this threshold deviance for
classifying a flow as discordant may be biased to higher values by the
potential inclusion of weakly discordant flows in the non-discordant
category. With respect to the conclusions of this work, this bias would
be conservative in allowing a greater fraction of Olympus Mons to
have formed before the primary discordant flow that is the focus of
this study.
Several topographically discordant flows were identified on
the slopes of the flexural trough that deviated from the downslope direction by more than this natural variability. One flow on
the eastern part of the trough, originating near 22.01N, 242.51E,
deviates from the down-slope direction by 42 7221 in the upper
reaches, 197121 in its middle section, and 377231 in the lower
reaches. Another flow originating near 34.01N 117.21E deviates
from the down-slope direction by 26 7191, while a nearby
flow originating near 31.81N 246.51E deviates by 217101.
These latter two flows are crossed by arcuate graben that are
related to the flexural stresses arising from Olympus Mons
loading, providing further support for the interpretation that they
pre-date a significant fraction of the edifice. However, these
flows are near the limit to be considered significantly discordant,
and are located sufficiently far from the edifice as to warrant
concern over competing influences on the present-day topography (e.g., intrusive uplift or flexural loading associated with Alba
Patera).
In the northeast quadrant of the Olympus Mons flexural trough,
one lava flow is oriented at an angle of 787341 away from the
current down-slope direction (Fig. 1b). The slope on the surface of
the flow in the down-flow direction is 0.11%, while the slope in the
local down-slope direction is 0.33% towards the trough. This flow is
located an average of 175725 km from the boundary where the
trough fill meets the flexural slope, in a location where the presentday topography conforms to the expectation for the flexural slope
on the outer flank of the trough. The implication is that this lava
flow formed before the flexural trough altered the region’s topography. This flow will be the focus of the subsequent analyses,
because of its location and strongly deviant path. Through flexural
modeling and crater age dating, it will now be possible to constrain
what fraction of Olympus Mons could have been in place at the time
of formation of this flow without redirecting the flow of lava toward
the present-day down-slope direction.
displacements of the lithosphere (wlm):
rc
gh
rm l lm
1
3rm
1
3rm
gl ¼ 1
ð2lþ 1Þr al ð2l þ 1Þr
wlm ¼ al ¼
t¼
s¼
D¼
½lðlþ 1Þð1nÞ
s½l3 ðl þ 1Þ3 þ4l2 ðl þ 1Þ2 4lðl þ 1Þ þ t½lðl þ 1Þ þ2 þ ½lðl þ 1Þ þ ð1nÞ
ET e
R2 g rm
D
R4 g rm
ET e 3
12 1n2
ð1Þ
where rc is the crustal load density, rm is the mantle density
(assumed to be 3400 kg/m3), r is the mean planetary density
(assumed to be 3940 kg/m3), n is Poisson’s ratio (assumed to be
0.25), E is Young’s modulus (assumed to be 100 GPa), Te is the
lithosphere thickness, R is the mean planetary radius, g is the
gravitational acceleration, r and t are the dimensionless constants, and D is the flexural rigidity (Johnson et al., 2000). The
final topography is then simply the sum of the thickness of the
applied load and the flexural deformation in response to that load.
In this model a volcanic edifice of a fractional Olympus Mons was
used initially, with a height of 2 km while keeping the flank slopes
equal to the present-day values of 51. After the initial volcanic
edifice emplacement, additional loading events were used to represent the volcanic infilling of the flexural trough resulting from the
centralized loading, until a flat trough-fill surface was achieved with
an elevation consistent with that of the present-day trough. Additional volcanic events were then modeled so as to bring the
topography of the edifice up to cones of successively greater elevation, with slopes matching the observed values, flattened tops
representing the caldera, and cut-offs to approximate the basal scarp.
Each incremental loading event began with the load thickness from
the previous iteration, and increased it so as to increase the preflexural height of the volcanic edifice by a prescribed amount (5 km
in the early stages of construction, decreasing in the later stages), or
to bring the pre-flexural surface of the trough fill up to the presentday level. Each edifice construction event increased the height by up
to 4 km, including the resulting flexural subsidence. Multiple trough
filling events were computed between each edifice construction
event so that the trough fill would reach its present day level relative
to the surrounding terrain. In this manner, each intermediate step in
the modeled construction of Olympus Mons resulted in a volcano
with flank slopes similar to the present-day slopes but a lesser total
3. Modeling of Olympus Mons loading and flexure
3.1. Methods
In order to quantify the fraction of the volume of Olympus
Mons that would be required in order to deviate the flow from the
down-slope direction, thin-shell spherical harmonic models were
used to examine the flexural response due to volcanic loading
(Evans et al., 2010; Johnson et al., 2000; Willemann and Turcotte,
1981). This approach relates the spherical harmonic coefficients
of the thickness of the load of degree l and order m (hlm) to the
spherical harmonic coefficients of the membrane-flexural
Fig. 2. MOLA topographic shaded relief map of a set of lava flows classified as nondiscordant.
R.J. Isherwood et al. / Earth and Planetary Science Letters 363 (2013) 88–96
3.2. Flexural modeling results and implications for the paleotopography
height (km)
The models were first evaluated against the observed width of
the flexural trough of 607742 km, measured from the center of
the edifice to the outer edge of the trough fill. This value
represents the average over the southeastern half of the volcano,
as the trough and outer flexural bulge to the northwest are
completely buried by lavas originating from within the flexural
trough, possibly as a result of flood basalt eruptions associated
with the early stages of a mantle plume (Fuller and Head, 2009).
Given the range in the observed trough width, a broad range of
lithosphere thicknesses and load densities are permissible (Fig. 4).
Thicker lithospheres favor lower densities, while thinner lithospheres favor higher densities. For a given load density, a thinner
lithosphere will result in a narrower flexural trough. For example,
with a load density of 3050 kg/m3, a lithosphere thickness of
60 km results in a trough width of 578 km, while a lithosphere
thickness of 80 km results in a trough width of 617 km. This
approach does not result in a unique solution for the lithosphere
thickness and load density, thus we adopt the best-fit load density
of 3150 kg/m3 from McGovern et al. (2002, 2004b) in subsequent
analyses. For this load density, the best-fit lithosphere thickness
for the mean trough width is between 60 and 70 km. The trough
width in the direction of the discordant flow is slightly greater
25
20
15
10
5
0
-5
-10
-15
-20
-25
-30
( 640 km) favoring an 80 km-thick lithosphere. These results are
in reasonable agreement with the independent analysis of
McGovern et al. (2002, 2004b), that favored a lithosphere thickness 470 km.
As our primary interest is in reproducing the lithospheric flexure
at the location of the discordant flow, the flexural profiles predicted
by the model for a load density of 3150 kg/m3 and a range of
lithosphere thicknesses were compared with the observed topographic profile through the discordant flow oriented perpendicular
to the trough (Fig. 5). Because of asymmetries in the shape of the
flexural trough, the position of the discordant flow is measured
relative to the boundary between the inwards-facing flexural slope
and the outer edge of the trough fill for comparison with the flexural
models. The current downhill slope at the location of the discordant
flow is 0.33% at a distance of 175 km from the edge of the trough fill.
The mean value of the level of the trough fill adopted generates
complications in the area of the discordant flow, where the elevation
difference between the outer flexural bulge and the trough fill reaches
a local maximum. This local variability would have only a modest
effect on the total flexural response to the load, because the flexural
response of a thick lithosphere effectively filters out the short
wavelength variability in the load thickness. However, local variability
in the observed level of the trough fill must be considered when
comparing the flexural topography with the observations. This is
accounted for by following the flexural profile below the surface of
the model trough fill (dashed line in Fig. 5) to account for the locally
shallower trough fill. The observed topography was then aligned with
and compared to the modeled flexural profiles to match the slope and
concavity (Fig. 5a). It is then possible to identify the lithosphere
thickness that provides the best fit to the observed flexural profile at
the discordant flow.
3300
3250
3200
ρ (kg/m3)
height, and a flexural trough that is volcanically filled to the presentday level (Fig. 3). The trough-filling events contributed significantly to
both the total volume of Olympus volcanics and the resulting flexural
response of the lithosphere.
The models were designed to match the present-day height of the
edifice of 21 km relative to the trough fill on the highest side. The
level of the trough fill relative to the surroundings is affected by a
regional westward slope due to the location of Olympus on the edge
of Tharsis. As a result, the level of the trough fill relative to the outer
flexural bulge ranges from 0 km in the west, where the trough has
been completely flooded by lava flows, to in excess of 2 km in the
northeast. An average value of 1 km is adopted here as representative of the typical fill height of the trough volcanics relative to the
surroundings.
The model was used to investigate a range of lithosphere
thicknesses and load densities. McGovern et al. (2002, 2004b) showed
through analysis of gravity and topography admittance that Olympus
Mons is supported by a lithosphere thickness of at least 70 km, with a
best-fit load density of 3150 kg/m3. In this study, lithosphere thicknesses from 60 to 90 km, and load densities from 3000 to 3200 kg/m3
were considered.
91
3150
3100
Te=60 km
Te=70 km
Te=80 km
Te=90 km
3050
3000
2950
550
600
650
trough width (km)
700
Fig. 4. The flexural trough width as a function of load density for lithosphere
thicknesses of 60–90 km. The vertical dotted line represents the observed mean
trough width of 607 km.
T e =70 km
T e =80 km
trough width
0
300
600
900 1200 1500 1800 2100 0
distance (km)
300
600
900 1200 1500 1800 2100
distance (km)
Fig. 3. Profiles of the surface and base of the Olympus Mons edifice and trough fill during incremental stages of its construction, for lithosphere thicknesses of 70 km (a)
and 80 km (b). Note that the intermediate stages of construction (gray) represent the elevation of the surface and base of the edifice as they were at those times, and do not
correspond to the present-day elevations of those now-buried surfaces.
92
R.J. Isherwood et al. / Earth and Planetary Science Letters 363 (2013) 88–96
Te=60 km
Te=70 km
Te=80 km
Te=90 km
2.0
1.5
discordant flow
elevation (km)
2.5
1.0
0.5
0
slope at discordant flow (%)
0
0.4
50
100 150
200
distance (km)
250
300
Te=70 km
Te=80 km
0.3
0.2
0.1
0
−0.1
0
20
40
60
80
100
percent by volume of Olympus Mons
Fig. 5. (a) Comparison of the observed topographic profile of the flexural trough
through the discordant flow (red) to the flexural model predictions at the end of
edifice construction for lithosphere thicknesses of 60–90 km (elevation is relative
to the trough fill level). (b) The slope at the discordant lava flow as a function
of the percent of Olympus Mons emplaced during incremental stages of its
construction. The dashed lines indicate the observed slope at the discordant flow
and the resulting constraints on the maximum volume of Olympus Mons in place
at that time. Small dots represent incremental steps of edifice construction and
trough in-filling, and large dots represent the final condition at each stage in the
construction. (For interpretation of the references to color in this figure legend, the
reader is referred to the web version of this article.)
The results show that lithosphere thicknesses of 70 and 80 km
both provide good matches to the observed topographic profile,
again in agreement with the results of McGovern et al. (2002,
2004b). These models provide the best match to the overall slope
and concavity of the profile, including that at the location of the
flow, but slightly underestimate the flexure near the edge of the
trough fill. Note that this approach implicitly assumes flat
topography prior to loading by Olympus Mons, which is a reasonable approximation given the location of Olympus in the smooth
volcanic plains outside of Tharsis. If the surface at the location of
the profile analyzed here initially exhibited a significant westward dip due to a regional gradient away from Tharsis, then the
reduced mean slope for the corrected topographic profile would
provide a better match to thicker lithospheres. However, because
this flow is nearly perpendicular to the flexurally generated
down-slope direction, and the paleo-slope perpendicular to the
flow direction should be zero, we infer that there is little
contribution to the observed flexural profile from pre-existing
slopes. Nevertheless, even allowing for changes in overall slope,
the concavity of the observed profile still provides the best match
to the 70–80 km thick lithosphere models. A variable lithosphere
thickness surrounding Olympus Mons is suggested by the noncircular outline of the flexural trough, so values of both 70 and
80 km are considered in the following analyses.
The 70 km lithosphere results in 26.8 km of downward flexure
beneath the center of the edifice (Fig. 3a). The slope at the
discordant flow is 0.28% and the trough width is 623 km, which
falls within the observed range of 607742 km. The total volume
of volcanics is 1.13 107 km3, 19% of which is contained within
the edifice above the level of its base, while 81% is contained
within the trough fill and the levels of the edifice below its base.
A similar effect of lesser magnitude is observed in the volume
distribution between the edifice and trough fill of smaller terrestrial shield volcanoes (Robinson and Eakins, 2006). A larger
fraction of the total volcanic volume is contained within the filled
troughs surrounding Venusian volcanoes for which the flexural
troughs have been completely infilled (McGovern and Solomon,
1997). The 80 km lithosphere results in a slope of 0.40% at the
discordant flow, a trough width of 643 km, and total downward
flexure beneath the center of the edifice of 21.2 km (Fig. 3b).
The total volume of volcanics is 1.01 107 km3, 79% of which is
contained within the trough fill and 21% within the edifice above
the level of its base. Note that the surface representing the
summit during the early to intermediate stages of the construction history would now be buried to a level below the surrounding trough-filling volcanics because of the accumulated flexural
subsidence during continued edifice growth. This continuing
burial of the earlier volcanic surfaces and calderas is an expected
outcome of shield construction (McGovern and Solomon, 1997).
The predicted flexural slope at the location of the discordant
flow was calculated throughout the modeled construction of
Olympus Mons (Fig. 5b). It is now possible to use the predicted
flexural response to Olympus Mons to determine what fraction of
the rise could have been present at the time of formation of the
discordant flow, without causing it to be oriented such that it
flows down the flexural slope toward the trough. To do this, we
find the point during the modeled construction of Olympus Mons
at which the added flexural slope would be sufficient to first
classify this flow as topographically discordant, using the criterion from Section 2 of an angle between the down-slope and
down-flow directions exceeding 18.51. The application of this
criterion is conservative, in that this flow maintains a consistent
orientation over a much greater distance than the typical natural
deviations from the down-slope direction observed in nondiscordant flows (Fig. 2). Taking the down-flow slope of 0.11%
to represent the original pre-Olympus slope, then an added
perpendicular flexural slope of 0.037% would be sufficient to
re-orient the flow by 18.51. This added slope corresponds to the
flexural response of a proto-Olympus Mons of 29% of its
present-day volume for an 80 km lithosphere and 51% for a
70 km lithosphere.
4. Volcanic history
4.1. Constraints on the beginning and end of edifice construction
The combination of the topographically discordant flow as an
indicator of paleotopography and flexural modeling suggest that
Olympus Mons had reached no more than 29–51% of its current
volume at the time that this flow formed. This estimate is
conservative, as the near-orthogonality of the flow with the slope
of the flexural trough suggests that the flow likely formed prior to
any flexural response to Olympus Mons loading. Thus, the age of
the flow provides a constraint on the age of the early stages of
significant volcanic construction.
In order to constrain the end of the primary construction
phase, we focus on the aureole deposits, which surround the main
edifice of the volcano and extend up to 750 km away from the
edge of the edifice. McGovern et al. (2004a) argued that the
aureole deposits formed due to flank failure-induced landslides
during the late stage of Olympus Mons growth. They showed that
the volume of one of the aureole deposits matched the void
volume within a concave embayment in the basal scarp surrounding the edifice. Significant continued buildup of the edifice
R.J. Isherwood et al. / Earth and Planetary Science Letters 363 (2013) 88–96
following the formation of the aureole would have modified or
even completely buried the basal scarp. Thus, the aureole deposits
likely formed when Olympus Mons was similar in size to its
present-day state. These aureole deposits have escaped the
volcanic resurfacing that has dominated the Olympus Mons flanks
and thus provide a better constraint on the age at which the bulk
of the edifice was in place.
4.2. Crater retention ages
We now have two features bracketing the bulk of Olympus Mons
formation: the discordant flow northwest of the rise which formed
when Olympus Mons was no more than 29–51% of its present-day
volume, and the aureole deposits which formed after the edifice was
essentially in its present-day form (McGovern et al., 2004a). Crater
size-frequency distributions can now be used to determine model
ages for these surfaces. Craters were counted over the discordant flow
using THEMIS visible image data (Christensen et al., 2004), and the
size-frequency distribution was matched to theoretical isochrons
using the Craterstats program (Michael and Neukum, 2010).
The crater retention age for the discordant flow was found to be
þ 0:05
3:670:10
Ga (Fig. 6a), corresponding to the early Hesperian epoch. At
this time, Olympus Mons had reached no more than 29–51% of its
present-day volume.
The aureole deposits preserve an imperfect record of craters due
to the rough and blocky nature of the terrain. Craters formed in
this rough surface will tend to be irregular in outline. The abundant
steep slopes have a poor potential for crater preservation, while the
smooth surfaces between the aureole blocks have experienced
apparent resurfacing. Circular to quasi-circular depressions may be
either true impact craters or may have formed through a different
mechanism (e.g., collapse pits overlying fractures, or coincidental
arrangement of aureole scarps in a quasi-circular geometry). Thus,
crater retention ages of the aureoles may over- or under-estimate
the true age. Nevertheless, the aureoles are useful as features that
likely correspond to the late stages of Olympus Mons construction,
while having escaped the continued volcanic resurfacing that has
affected the flanks and caldera up through the present era. Hiller et al.
(1982) determined ages of 3.63, 3.51, 3.68, and 3.67 Ga for Lycus
Sulci, an unnamed aureole lobe due north of Olympus (hereafter the
north aureole), Cyane Sulci, and Gigas Sulci, respectively. The north
aureole corresponds to the study site of McGovern et al. (2004a), in
which the volume of the aureole corresponds to the void volume
within a concave embayment in the basal scarp. This aureole
thus presents the strongest case for post-dating the bulk of the
present-day edifice. We have recalibrated the age of the north aureole
to the Hartmann and Neukum (2001) chronology, arriving at an age
of 2.95 Ga.
We also performed our own independent crater counts for the
north aureole using a THEMIS daytime infrared image mosaic
(Christensen et al., 2001). Candidate craters were assigned a
subjective classification based on the confidence of their identification (possible, likely, and definite impact craters). Including all of
þ 0:07
the potential craters, we obtained an age of 3:470:16
(Fig. 6b).
Considering only the likely and definite craters, we find an age of
þ 0:55
2:540:69
Ga, while the definite crater population alone results in
þ 0:51
an age of 0:980:61
Ga. The 3.47 Ga age is likely an overestimate
due to the inclusion of non-crater depressions, while the 0.98 Ga
age is likely an underestimate due to the exclusion of imperfectly
preserved craters, the poor preservation of craters on the steep
slopes of the aureole blocks and the resurfacing in the spaces
between blocks. The age from the likely and definite craters is
adopted as the best estimate of the aureole age, which also
provides the best match to the adjusted age from Hiller et al.
(1982) and is consistent with the Early Amazonian classification of
the aureoles by Scott and Tanaka (1986). This age range indicates
an end to major edifice construction by the Early Amazonian.
Several studies have examined the crater chronology of portions of the edifice itself. Neukum et al. (2004) examined the
western basal scarp using HRSC data. The oldest age in that study
of 3.8 Ga suggests that the edifice is older than the discordant
flow, in apparent conflict with this study. However, that age is
based on only 4 craters, each greater than 1 km in diameter.
Examination of these craters reveals that all four are members of
chains of 2–3 craters oriented approximately NW–SE in alignment with nearby tectonic features. These craters have a muted
expression and rounded rims, with aspect ratios of 1.05–1.09.
These observations all indicate that these are collapse pits rather
than impact craters, and thus we do not give further consideration to the possibility of a 3.8 Ga age for the edifice. The next two
oldest ages from that study of 2.97 and 3.34 Ga are consistent
with the age of the aureole deposit, as well as with the interpretation of this study that the bulk of the edifice formed post3.67 Ga. Thus, although there is significant uncertainty in the age
of the aureole in this study, it is supported by the broad
agreement between all age constraints for the end of primary
edifice construction from multiple aureole deposits as well as
exposures of older surfaces on the edifice itself. Much younger
ages in the range of 25–700 Ma have been found for the flanks
and calderas, but these are representative only of the late-stage
resurfacing of the edifice (Basilevsky et al., 2006; Neukum et al.,
2004; Robbins et al., 2011; Werner, 2009). We adopt the age of
north aureole deposit
+0.05
3.67 - 0.10 Ga
cum. crater density (km-2)
cum. crater density (km-2)
discordant trough flow
10-2
10-3
10-4
10-1
93
3.47 +0.07
-0.16 Ga
10-3
2.54+0.55
-0.69 Ga
0.98+0.51
-0.61 Ga
10-4
10-5
100
crater diameter (km)
101
10-1
100
crater diameter (km)
101
Fig. 6. Cumulative crater density plots for the discordant flow (a) and the north aureole deposit (b). The aureole isochrons show fits to all possible craters (top gray line),
the definite and likely crater populations (black line), and the definite craters only (bottom gray line).
94
R.J. Isherwood et al. / Earth and Planetary Science Letters 363 (2013) 88–96
þ 0:55
2:540:69
Ga from the north aureole as a constraint on the end of
the primary constructional phase of Olympus Mons, with ages at
the upper end of this range supported by the crater retention ages
from the aureole by Hiller et al. (1982) and from the western
basal scarp by Neukum et al. (2004).
4.3. Volcanic construction history
Combining the results from the paleotopography and flexural
analysis with the crater retention ages, we find that at least
49–71% of Olympus Mons must have formed in the time period
þ 0:05
þ 0:55
between 3:670:10
and 2:540:69
Ga. This corresponds to a conþ 0:74
struction period of less than 1:130:65
Gyr, and a volumetric
eruption rate of greater than 0.003–0.015 km3/yr. These rates are
conservative lower bounds, as the duration of volcanism was likely
at the short end of the range, and we cannot exclude the possibility
that 100% of Olympus formed during this interval. Thus, volumetric
eruption rates during the primary period of Olympus Mons construction greater than 0.015 km3/yr appear likely.
After the primary shield building phase, the eruption rate must
have decreased dramatically. However, volcanic resurfacing has
clearly continued to the present day as indicated by the young
crater retention ages of the majority of the caldera and edifice
surface (Basilevsky et al., 2006; Neukum et al., 2004; Robbins et al.,
2011; Werner, 2009). This late-stage volcanism must have been
sufficient to completely obscure the 2.5 Ga crater population on
the surface of the edifice. Based on the surface area of the edifice,
one would expect 15 craters greater than 3 km in diameter for an
age of 2.5 Ga. Complete burial of the rim of a 3 km diameter crater
requires volcanic resurfacing to a depth of 120 m, equating to a
volume of 3.2 104 km3 over the entire edifice. This amount of
resurfacing over a period of 2.5 Ga places a lower bound on the
late-stage resurfacing rate of 1.3 10 5 km3/yr. An upper bound
on the volume of late-stage eruptions can be arrived at by noting
that although younger lavas over-spilled the basal scarp in places
(Neukum et al., 2004), they have not significantly obscured the
scarp or changed its topography adjacent to the north aureole. The
volume of lava sufficient to fill this scarp in and restore the edifice
to a conical shape was calculated to be 2.95 105 km3, using a
truncated conical volume with an outer slope equal to the flank
slope and a vertical inner edge of height equal to the basal scarp
height. This volume provides an upper bound on the late-stage
eruption rate of 1.2 10 4 km3/yr since 2.5 Ga.
5. Discussion and conclusions
Through use of paleotopography indicators, flexural modeling,
and crater retention ages of the aureole deposits and discordant
lava flow, we have shown that a minimum of 49–71% of the
volume of Olympus Mons was emplaced during a period of less
þ 0:74
þ 0:05
þ 0:55
Gyr between 3:670:10
and 2:540:69
Ga. The bestthan 1:130:65
fit models were determined to have lithosphere thicknesses of
70–80 km for a density of 3150 kg/m3, resulting in a total volume
of 1.13 107 km3 for a lithosphere thickness of 70 km and
1.01 107 km3 for a lithosphere thickness of 80 km, with 80%
of the volume of erupted material contained within the flexural
depression below the surface of the volcanic trough fill. Combining the constraints on the volume and duration of volcanism, we
arrive at a lower bound on the eruption rate during the primary
shield-building phase of 0.003–0.015 km3/yr. The assumptions
made throughout this analysis were conservative in the sense of
favoring lower eruption rates, and thus rates at or exceeding the
higher end of this range are preferred. Comparatively, terrestrial
hot-spot volcanism in Hawaii and the Emperor seamounts has led
to long-term average eruption rates of 0.010–0.017 km3/yr
(Bargar and Jackson, 1974; Robinson and Eakins, 2006). The total
eruption volume for the Hawaii–Emperor chain over the past
80 Myr was calculated to be 1.36 106 km3, compared with 1.01–
1.13 107 km3 for Olympus Mons.
Despite significant uncertainties in the timing and rates of
volcanism at Olympus Mons, the constraints provided by this
study suggest strikingly similar long-term average eruption rates
for Olympus Mons on Mars and hotspot volcanism on Earth.
Although this result is not surprising given the similar shield
morphologies on Earth and Mars, this is the first quantitative
constraint to confirm this similarity in eruption rate. The similarity in eruption rates is suggestive of similar underlying geodynamic mechanisms responsible for the magma supply during the
primary construction of Olympus Mons and the terrestrial
hotspot volcanoes, often explained through the action of mantle
plumes (Lei and Zhao, 2006; Ribe and Christensen, 1999; Watson
and McKenzie, 1990; Wilson, 1963). By analogy with Hawaii,
which experienced a 10-fold increase in the eruption rate in the
past 1 Myr (Bargar and Jackson, 1974), Olympus Mons also likely
experienced a more complicated history of volcanism during the
primary shield-building phase. The longer duration and greater
total volume of volcanism at Olympus Mons is consistent with the
fact that Hawaii remains an active hotspot today, and the
beginning of the hotspot track has been lost to subduction.
The results also indicate that the primary shield-building
phase was followed by a long period of continued volcanism at
significantly lower rates, as also suggested by Hiller et al. (1982).
This result addresses the problem identified by Wilson et al.
(2001), that the estimated mean rate of volcanism over the
history of the shields of 0.0015 km3/yr falls significantly short
of the rate of 0.03–0.3 km3/yr required to offset the conductive
cooling of the magma chamber. The preferred lower bound of
0.015 km3/yr on the mean eruption rate during the shieldbuilding phase falls within a factor of two of the required longterm rate to maintain a continuously active magma chamber in
the subsurface. In contrast, the late-stage volcanism at low rates
during the Amazonian would require episodically active magma
chambers.
The decrease in eruption rates by two to three orders of
magnitude after the end of the shield-building phase suggests a
different geodynamic mechanism causing melt generation and
eruption. This progression from the primary construction of the
edifice, to the later extended period of reduced eruption rates is also
observed in terrestrial hotspot volcanism, with the transition from
the shield-building phase to the post-shield phase. Only 80% of the
eruption volume along the Emperor–Hawaiian seamount chain is
emplaced during the shield-building phase (DePaolo and Stolper,
1996). Frey et al. (1990) have shown that the composition of the
post-shield volcanism is distinct from that during the primary
edifice construction. This compositional change points to the possibility of different source materials or melting conditions, which
would be consistent with minor continued volcanism after the
shield has been carried beyond the influence of the mantle plume
tail by the plate motion. On Mars, the lack of plate motion requires
that post-shield volcanism would only commence once the plume
itself had moved to a different location or ceased its activity
altogether. The transition from a high eruption rate during the
shield-building phase to a low eruption rate in the post-shield stage
is also supported by geomorphic analysis of the Tharsis Montes
(Bleacher et al., 2007). That study examined the exposed flows on
the surface of the edifice and rift apron, observing a transition from
tube- to channel-forming eruptive activity on the flanks that is
consistent with a declining magma supply rate.
The volcanic history of Olympus Mons and terrestrial shield
volcanoes may have parallels with the volcanic history of the Tharsis
R.J. Isherwood et al. / Earth and Planetary Science Letters 363 (2013) 88–96
province as a whole. A significant fraction of Tharsis formed during the
Noachian epoch (Anderson et al., 2001; Phillips et al., 2001), yet the
majority of the Tharsis plateau has a Hesperian age (Scott and Tanaka,
1986) and volcanism continued locally into the Amazonian (Neukum
et al., 2004). The formation of Tharsis may have involved a large
mantle plume that originated near the core–mantle boundary (Grott
and Breuer, 2010; Harder and Christensen, 1996; Hauber et al., 2011;
Kiefer and Li, 2009), analogous to, though on a much larger scale than,
the small plumes that have been implicated in hotspot volcanism. The
late-stage volcanism at Olympus Mons and Tharsis as a whole may be
a result of partial melt in the upper mantle generated by the
background mantle convection (Hauck and Phillips, 2002) or by the
thermal blanketing effect of the volcanically thickened crust
(Schumacher and Breuer, 2007). The late-stage magma may simply
take advantage of the existing magma conduits, or the magma ascent
may be enhanced by the flexural response to the load of the edifice
(Galgana et al., 2011).
Using a combination of paleotopography, flexural modeling, and
crater retention ages, this study has been able to more tightly
constrain the eruptive history of the Olympus Mons shield volcano
on Mars. This is the first study to place firm constraints on the
timing of early shield construction, finding that the bulk of Olympus
Mons post-dates 3.67 Ga. This analysis has revealed a similar
volcanic history to that observed in terrestrial hotspot volcanism,
with a high rate of early shield forming volcanism transitioning to a
low rate of post-shield volcanism. Thus, although the overall
patterns and styles of volcanic and tectonic activity on Earth and
Mars have differed greatly due to the dominance of plate tectonics
on the former, hotspot volcanism on these two bodies appears to
follow similar patterns of behavior.
Acknowledgments
We are grateful to Pat McGovern for the thorough and
thoughtful review of this paper. This work was supported in part
by a grant from the NASA Mars Fundamental Research Program.
RJI and LMJ were also supported by undergraduate research
fellowships provided by the Colorado Space Grant. Flexural
modeling was performed by RJI. Lava flow analyses and crater
retention ages were calculated by LMJ and JCJ. The manuscript
was written by RJI and JCAH.
References
Anderson, R.C., Dohm, J.M., Golombek, M.P., Haldemann, A.F.C., Franklin, B.J., Tanaka,
K.L., Lias, J., Peer, B., 2001. Primary centers and secondary concentrations of
tectonic activity through time in the western hemisphere of Mars. J. Geophys.
Res. 106, 20563–20585.
Bargar, K.E., Jackson, E.D., 1974. Calculated volumes of individual shield volcanoes
along the Hawaiian-Emperor chain. US Geol. Surv. J. Res. 2, 545–550.
Basilevsky, A.T., Werner, S.C., Neukum, G., Head, J.W., van Gesselt, S., Gwinner, K.,
Ivanov, B.A., 2006. Geologically recent tectonic, volcanic, and fluvial activity on
the eastern flank of the Olympus Mons volcano, Mars. Geophys. Res. Lett. 33,
L13201, http://dx.doi.org/10.1029/2006GL026396.
Bleacher, J.E., Greeley, R., Williams, D.A., Werner, S.C., Hauber, E., Neukum, G., 2007.
Olympus Mons, Mars: inferred changes in late Amazonian aged effusive activity
from lava flow mapping of Mars Express High Resolution Stereo Camera data.
J. Geophys. Res. 112, E04003, http://dx.doi.org/10.1029/2006JE002826.
Christensen, P.R., Bandfield, J.L., Hamilton, V.E., Ruff, S.W., Kieffer, H.H., Titus, T.N.,
Malin, M.C., Morris, R.V., Lane, M.D., Clark, R.L., Jakosky, B.M., Mellon, M.T.,
Pearl, J.C., Conrath, B.J., Smith, M.D., Clancy, R.T., Kuzmin, R.O., Roush, T.,
Mehall, G.L., Gorelick, N., Bender, K., Murray, K., Dason, S., Greene, E., Silverman, S., Greenfield, M., 2001. Mars Global Surveyor Thermal Emission Spectrometer experiment: investigation description and surface science results. J.
Geophys. Res. 106, 22823–22871.
Christensen, P.R., Jakosky, B.M., Kieffer, H.H., Malin, M.C., McSween, H.Y., Nealson,
K., Mehall, G.L., Silverman, S.H., Ferry, S., Caplinger, M., Ravine, M., 2004. The
Thermal Emission Imaging System (THEMIS) for the Mars 2001 Odyssey
Mission. Space Sci. Rev. 110, 85–130.
95
DePaolo, D.J., Stolper, E.M., 1996. Models of Hawaiian volcano growth and plume
structure: implications of results from the Hawaii Scientific Drilling Project.
J. Geophys. Res. 101, 11643–11654, http://dx.doi.org/10.1029/96JB00070.
Evans, A.J., Andrews-Hanna, J.C., Zuber, M.T., 2010. Geophysical limitations on the
erosion history within Arabia Terra. J. Geophys. Res. 115, E05007, http://dx.doi.org/
10.1029/2009JE003469.
Frey, F.A., Wise, W.S., Garcia, M.O., West, H., Kwon, S.-T., Kennedy, A., 1990.
Evolution of Mauna Kea volcano, Hawaii: petrologic and geochemical constraints on post-shield volcanism. J. Geophys. Res. 96, 1271–1300.
Fuller, E.R., Head, J.W., 2009. Olympus Mons, Mars: detection of extensive
preaureole volcanism and implications for initial mantle plume behavior.
Geology 31, 175–178.
Galgana, G.A., McGovern, P.J., Grosfils, E.B., 2011. Evolution of large Venusian
volcanoes: insights from coupled models of lithospheric flexure and magma
reservoir pressurization. J. Geophys. Res. 116, E03009, http://dx.doi.org/
10.1029/2010JE003654.
Grott, M., Breuer, D., 2010. On the spatial variability of the martian elastic
lithosphere thickness: evidence for mantle plumes? J. Geophys. Res. 115,
E03005, http://dx.doi.org/10.1029/2009JE003456.
Harder, H., Christensen, U., 1996. A one-plume model of Martian mantle convection. Nature 380, 507–509.
Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the evolution of
Mars. Space Sci. Rev. 96, 165–194.
Hauber, E., Broz, P., Jagert, F., Jodlowski, P., Platz, T., 2011. Very recent and widespread basaltic volcanism on Mars. Geophys. Res. Lett. 38, L10201, http://dx.d
oi.org/10.1029/2011GL047310.
Hauck, S.A., Phillips, R.J., 2002. Thermal and crustal evolution of Mars. J. Geophys.
Res., 107, http://dx.doi.org/10.1029/2001/JE001801.
Hiller, K.H., Janle, P., Neukum, G., Guest, J.E., Lopes, R.M.C., 1982. Mars: stratigraphy and gravimetry of Olympus Mons and its aureole. J. Geophys. Res. 87,
9905–9915.
Johnson, C.L., Solomon, S.C., Head, J.W., Phillips, R.J., Smith, D.E., Zuber, M.T., 2000.
Lithospheric loading by the northern polar cap on Mars. Icarus 144, 313–328.
Kiefer, W.S., Li, Q., 2009. Mantle convection controls the observed lateral variations
in lithospheric thickness on present-day Mars. Geophys. Res. Lett. 36, L18203,
http://dx.doi.org/10.1029/2009GL039827.
Lei, J., Zhao, D., 2006. A new insight into the Hawaiian plume. Earth Planet. Sci.
Lett. 241, 438–453.
Liu, L., Gurnis, M., 2010. Dynamic subsidence and uplift of the Colorado Plateau.
Geology 38, 663–666.
McGovern, P.J., Smith, J.R., Morgan, J.K., Bulmer, M.H., 2004a. Olympus Mons
aureole deposits: new evidence for a flank failure origin. J. Geophys. Res. 109,
E08008, http://dx.doi.org/10.1029/2004JE002258.
McGovern, P.J., Solomon, S.C., 1997. Filling of flexural moats around large
volcanoes on Venus: implications for volcano structure and global magmatic
flux. J. Geophys. Res. 102, 16303–16318.
McGovern, P.J., Solomon, S.C., Smith, D.E., Zuber, M.T., Simons, M., Wieczorek, M.A.,
Phillips, R.J., Neumann, G.A., Aharonson, O., Head, J.W., 2002. Localized gravity/
topography admittance and correlation spectra on Mars: implications for
regional and global evolution. J. Geophys. Res. 107, 5136, http://dx.doi.org/
10.1029/2002JE001854.
McGovern, P.J., Solomon, S.C., Smith, D.E., Zuber, M.T., Simons, M., Wieczorek, M.A.,
Phillips, R.J., Neumann, G.A., Aharonson, O., Head, J.W., 2004b. Correction to
‘‘Localized gravity/topography admittances and correlation spectra on Mars:
implications for regional and global evolution’’. J. Geophys. Res.109, 19-1–1925, http://dx.doi.org/10.1029/2004JE002286.
Michael, G.G., Neukum, G., 2010. Planetary surface dating from crater sizefrequency distribution measurements: partial resurfacing events and statistical age uncertainty. Earth Planet. Sci. Lett. 294, 223–229.
Neukum, G., Jaumann, R., Hoffmann, H., Hauber, E., Head, J.W., Basilevsky, A.T.,
Ivanov, B.A., Werner, S.C., van Gasselt, S., Murray, J.B., McCord, T., Team, H.C.-I.,
2004. Recent and episodic volcanic and glacial activity on Mars revealed by the
High Resolution Stereo Camera. Nature 432, 971–979.
Phillips, R.J., Zuber, M.T., Solomon, S.C., Golombek, M.P., Jakosky, B.M., Banerdt,
W.B., Smith, D.E., Williams, R., Hynek, B., Aharonson, O., Hauck, S.A.I., 2001.
Ancient geodynamics and global-scale hydrology on Mars. Science 291,
2587–2591.
Ribe, N., Christensen, U., 1999. A dynamical origin of the Hawaiian volcanism.
Earth Planet. Sci. Lett. 171, 517–531.
Robbins, S.J., Di Achille, G., Hynek, B.M., 2011. The volcanic history of Mars: highresolution crater-based studies of the calderas of 20 volcanoes. Icarus 211,
1179–1203.
Robinson, J.E., Eakins, B.W., 2006. Calculated volumes of individual shield
volcanoes at the young end of the Hawaiian Ridge. J. Volcanol. Geotherm.
Res. 151, 309–317.
Schumacher, S., Breuer, D., 2007. An alternative mechanism for recent volcanism
on Mars. Geophys. Res. Lett. 34, L14202, http://dx.doi.org/10.1029/
2007GL030083.
Scott, D.H., Tanaka, K.L., 1986. Geologic map of the western equatorial region
of Mars. US Geological Survey Miscellaneous Investigations Series, Map.
I-1802a.
Smith, D.E., Zuber, M.T., Frey, H.V., Garvin, J.B., Head, J.W., Muhleman, D.O.,
Pettengill, G.H., Phillips, R.J., Solomon, S.C., Zwally, H.J., Banerdt, W.B., Duxbury, T.C., Golombek, M.P., Lemoine, F.G., Neumann, G.A., Rowlands, D.D.,
Aharonson, O., Ford, P.G., Ivanov, A.B., McGovern, P.J., Abshire, J.B., Afzal, R.S.,
Sun, X., 2001. Mars Orbiter Laser Altimeter (MOLA): experiment summary
96
R.J. Isherwood et al. / Earth and Planetary Science Letters 363 (2013) 88–96
after the first year of global mapping of Mars. J. Geophys. Res. 106,
23,689–23,722.
Watson, S., McKenzie, D., 1990. Melt generation by plumes: a study of Hawaiian
volcanism. J. Petrol. 32, 501–537.
Werner, S.C., 2009. The global martian volcanic evolutionary history. Icarus 201,
44–68.
Willemann, R.J., Turcotte, D.L., 1981. Support of topographic and other loads on the
Moon and on the terrestrial planets. Lunar Planet. Sci. 12, 837–851.
Wilson, J.T., 1963. A possible origin of the Hawaiian islands. Can. J. Phys. 41,
863–870.
Wilson, L., Scott, E.D., Head, J.W., 2001. Evidence for episodicity in the magma
supply to the large Tharsis volcanoes. J. Geophys. Res. 106, 1423–1433.