After you have read the article ( Supervolcanoes within an ancient volcanic province in Arabia Terra, Mars (Michalski and Bleacher, 2013) , then you should answer the questions, see attachments.
Supervolcanoes within an ancient volcanic province in Arabia Terra, Mars (Michalski and
Bleacher, 2013)
The purpose of this assignment is to give you an idea of how understanding volcanic landforms
on Earth can be applied to increase our understanding of the geomorphic features, geologic
history, and climatic history of other planets.
After you have read the article, then you should answer the questions below.
1. List any terms that you are unfamiliar with or don’t understand as you read the article
and attempt to define them.
2. What is the hypothesis of this article?
3. Describe the main objectives for this study.
4. What is the evidence for volcanic origin of the features discussed in the article?
5. What are the issues with the authors hypothesis of volcanic origin?
6. Briefly summarize the volcanic history of Mars.
7. Describe Figure 5. What does show? How does it support the author’s hypothesis?
8. Describe the main results of this study.
9. How did the author’s apply their understanding of volcanic geomorphology on Earth to
better understand the formation of the craters in Arabia, Terra, Mars? Describe the
main methods used in this study.
10. After reading this article are you convinced that the features discussed are of volcanic
origin? Why or why not?
11. Did you enjoy this article? Why or why not? Do you have any further questions
regarding this article/research?
ARTICLE
doi:10.1038/nature12482
Supervolcanoes within an ancient
volcanic province in Arabia Terra, Mars
Joseph R. Michalski1,2 & Jacob E. Bleacher3
Several irregularly shaped craters located within Arabia Terra, Mars, represent a new type of highland volcanic construct
and together constitute a previously unrecognized Martian igneous province. Similar to terrestrial supervolcanoes, these
low-relief paterae possess a range of geomorphic features related to structural collapse, effusive volcanism and explosive
eruptions. Extruded lavas contributed to the formation of enigmatic highland ridged plains in Arabia Terra. Outgassed
sulphur and erupted fine-grained pyroclastics from these calderas probably fed the formation of altered, layered
sedimentary rocks and fretted terrain found throughout the equatorial region. The discovery of a new type of volcanic
construct in the Arabia volcanic province fundamentally changes the picture of ancient volcanism and climate evolution
on Mars. Other eroded topographic basins in the ancient Martian highlands that have been dismissed as degraded impact
craters should be reconsidered as possible volcanic constructs formed in an early phase of widespread, disseminated
magmatism on Mars.
The source of fine-grained, layered deposits1,2 detected throughout
the equatorial region of Mars3 remains unresolved, though the deposits are linked to global sedimentary processes, climate change, and
habitability of the surface4. A volcanic origin has been suggested on
the basis of the stratigraphy, morphology and erosional characteristics
of the deposits5. The case for a volcanic source is further strengthened
by the spectroscopic detection of sulphates in many of these deposits6
and detailed analyses of such rocks at the Meridiani Planum landing
site, which revealed materials altered under water-limited, acidic conditions that were probably governed by volcanic outgassing7. Yet,
although very fine-grained ash can be dispersed globally from a large
explosive eruption on Mars5,8, the currently known volcanic centres
are unlikely to have been the sources for thick, low-latitude layered
deposits in Arabia Terra9.
The lack of identifiable volcanic sources that could have produced
possible volcanogenic sediments in Meridiani Planum or in Gale
crater is not a unique problem. In fact, 70% of the crust was resurfaced
by basaltic volcanism, with a significant fraction emplaced from as yet
unrecognized sources10. Thus, undetected volcanic source regions
must exist within the ancient crust of Mars. Therefore, the following
questions arise: first, is ancient volcanism poorly understood because
higher Noachian erosion rates11 obliterated evidence for source
regions? Second, are ancient volcanoes highland volcanoes of fundamentally different character from the well-recognized, massive,
Hesperian shield volcanoes12,13? We suggest that the answer to the
second question is yes; we propose a new category of ancient volcanic
construct that has escaped detection until now.
Volcanism is the thread binding nearly every aspect of Mars’s
geological evolution. The crust of the planet was built through magmatism and effusive volcanism14, although an early phase of explosive
volcanism might have emplaced a significant amount of fragmented
material across the ancient crust15. Volatiles outgassed16 from volcanoes
controlled atmospheric chemistry17 and strongly affected climate18–20
throughout Martian history. The geochemistry and habitability of
Martian soils and sedimentary rocks are ultimately controlled by the
global sulphur cycle, which is fundamentally linked to volcanism21. It is
therefore critical to understand all styles and phases of Martian volcanism and how they have affected the Martian climate through time.
Evidence for volcanism in Arabia Terra
We present evidence for a new category of ancient volcanic construct
on Mars, ancient supervolcanoes, which together could have produced
vast amounts of lava and pyroclastic materials throughout Arabia
Terra and beyond. The features, which we call ‘plains-style caldera
complexes’, are characterized by the presence of collapse features,
low topographic relief (lower than that of typical paterae), and association with plains-style lavas and friable, layered deposits. Taken
together, the features, each with explosive outputs probably in excess
of terrestrial supervolcanoes, constitute a previously unrecognized
ancient volcanic province in Arabia Terra (Fig. 1).
The best example of a plains-style caldera complex is Eden patera,
which is a large, irregularly shaped topographic depression (dimensions ,55 km northwest–southeast and 85 km southwest–northeast)
located at 348.9u E, 33.6u N within Noachian–Hesperian ridged plains
of probable volcanic origin. The complex, which reaches a maximum
depth ,1.8 km below surrounding plains, includes at least three
linked depressions (Fig. 2) bounded by arcuate scarps and associated
with numerous faults and fractures. Although this feature has never
been differentiated from impact craters in the region, it lacks any
geological indicator of an impact origin, such as the presence of ejecta,
an uplifted rim, nearly circular geometry or the presence of a central
peak22. Its high ratio of depth to diameter is inconsistent with that of
an ancient impact crater that has been modified by erosion23. We
therefore rule out an impact origin for Eden patera.
We interpret Eden patera as a caldera complex on the basis of its
similarity to terrestrial calderas24 and its association with features that
indicate formation by means of collapse and volcanism both within
and outside the depression. The surrounding terrain comprises ridged
plains typical of Hesperian basaltic volcanism on Mars10. Within the
complex are fault-bounded blocks that display surfaces similar to the
adjacent ridged plain lavas (Fig. 2a). These blocks are tilted towards
the crater centre and are unrelated to headwall scarps that would
1
Planetary Science Institute, Tucson, Arizona 85719, USA. 2Department of Earth Sciences, Natural History Museum, London SW7 5BD, UK. 3NASA Goddard Space Flight Center, Greenbelt, Maryland
20771, USA.
3 O C T O B E R 2 0 1 3 | VO L 5 0 2 | N AT U R E | 4 7
©2013 Macmillan Publishers Limited. All rights reserved
RESEARCH ARTICLE
Study area
Semeykin crater
0° N
Oxus patera
Ismenia patera
Euphrates patera
0° E
y
otom
Dich dary
boun
Oxus cavus
Siloe patera
25 km
Eden patera
a
ia
rab
r
Ter
A
Elevation (m)
100 km
–3,000
–2,000
Figure 1 | Geographic context of the northern Arabia Terra region. The
dusty nature of Arabia Terra is shown in false-colour TES-derived albedo data
draped over MOLA hillshade data; bright colours correspond to dusty surfaces.
Recently named geographic features discussed in the text are labelled.
suggest a process similar to landslides. Graben associated with the interior fault blocks may have originally been linked with circumferential
graben outside the complex related to older collapses or progressive
formation through ‘piecemeal’, multicyclic evolution24. We interpret
a mound ,700 m high (11 km north–south and 23 km east–west)
within the complex to be a graben-related vent (Fig. 2b). Two sets of
nearly continuous terraces are found ,100 and 150 m above the caldera
floor. These terraces are strikingly similar to the ‘black ledge’ described
during the Kı̄lauea Iki eruption in 1959 (ref. 25), indicating high stands
of a drained lava lake26. A small mound (1 km across) several hundred
metres high and located between the two terraces shows surface cracks
similar to a tumulus27. Although tumuli clefts form during inflation, we
suggest that these cracks formed as the lava lake drained and the sinking
lake crust was draped onto caldera wall rocks.
The presence of volcanic features and significant faulting consistent
with collapse leads us to conclude that these linked depressions represent a large caldera complex formed in the Late Noachian to Early
Hesperian. A lacustrine origin for the terraces is unlikely due to the
a
60 km
c
d
Elevation (m)
–4,000
–3,000
b
d
e
e
c
Faults
Lava lake
Vent
Caldera 1
Caldera 2
20 km
Caldera 3
4 8 | N AT U R E | VO L 5 0 2 | 3 O C TO B E R 2 0 1 3
©2013 Macmillan Publishers Limited. All rights reserved
Figure 2 | The geology of Eden
patera. a, MOLA topographic data
are draped over THEMIS daytime
infrared data, showing the
morphology of Eden patera.
b, Geological mapping reveals the
presence of at least three calderas,
indicated by coloured shading.
c–e, Enlargements of the rectangles
in b. The caldera contains evidence
for fault blocks that preserve ridged
plain lavas on the upper surface (c),
a probable vent (d), and a series of
terraces that mark lava high stands of
a once active lava lake (white arrows)
and cracked crust (black arrows) due
to the draping of fragile crust onto
pre-existing surfaces during lava lake
drainage (e).
ARTICLE RESEARCH
paucity of channels found in or around the depression that could be
linked to aqueous surface processes. In addition, there is no apparent
evidence for lacustrine sediments within the basin, and the depression
is deeper than expected for a feature of this size that was partly filled by
outside sediment. The sequential development of this feature (calderas 1–3 in Fig. 2) seems to have undergone a transition from surface
sagging (caldera 1 in Fig. 2) to significant disruption of the crust and
subsequent down-dropping of large surface blocks (calderas 2 and 3 in
Fig. 2).
Several other features throughout the region have similar characteristics. Euphrates patera is an irregularly shaped depression that
reaches 700 m in depth below the surrounding lava plains and contains several benches in the interior that might be explained by
sequential episodes of collapse or lava-lake high stands (Fig. 3). The
irregular, rhombohedral form of the depression might relate to shortening in the southwest–northeast direction. Fractured surface textures in the centre of the depression are morphologically similar to
lava surfaces disrupted by collapse caused by the withdrawal of lava.
Other features in northern Arabia Terra contain evidence for collapse associated with volcanic activity. Siloe patera (6.6u E, 35.2u N) is
a set of nested deep depressions that reach ,1,750 m below the surrounding plains (Fig. 3). Similar to Eden patera, the nested craters are
characterized by steep-walled depressions linked by arcuate scarps
and faults. The primary structure is linked to a subtle northeast–
southwest-trending depression to the south that reaches ,700 m
depth, which we interpret as evidence for sagging due to the migration
of a magma body at depth. Although there is no evidence for impact
ejecta around the structure, there is a single set of lobate flows emanating from the southwest portion of the depression rim, which may
represent a single set of lava or pyroclastic flows reaching ,60 km
from the rim. Irregular mounds of friable materials inside the nested
craters are interpreted as pyroclastics from the volcano or as younger
friable deposits of another origin.
Some other depressions in the region contain less well preserved
evidence for volcanism, but in all cases they contain suites of features
Elevation (m)
–3,000
75 km
that are difficult to explain by other geological processes. Semeykin
crater is a large scalloped depression surrounded by lava plains and
friable deposits; it also contains mounds of friable materials in its
interior and ridged plains along the exterior. A suite of features,
Ismenia patera, Oxus patera and Oxus cavus, are located together near
0u E, 38.5u N. The two paterae have scalloped, breached rims composed
of layered materials. Oxus cavus is an elongated depression within a
broad mound 200–300 m high adjacent to a deep depression indicative
of sagging or collapse. Although none of these structures individually
contains as many pieces of evidence to clearly point to volcanism as are
seen at Eden patera, all of the features contain some evidence for
structural collapse, which is most likely to have occurred through
magmatic activity (although other hypotheses are considered below).
Eden patera and Euphrates patera represent the strongest evidence
for large calderas in Arabia Terra, on the basis of their association with
features that are diagnostic of surface disruption and collapse coupled
with evidence for effusive and explosive volcanism. Some of the other
features with fewer diagnostic features might not all represent caldera
formation, or they could have experienced a range of processes
responsible for the current morphology. Nonetheless, the region does
show strong evidence that several large depressions did not form as
impact craters and are most easily explained as volcanic calderas.
The roles of ice and impact
Some depressions throughout Arabia Terra have previously been
interpreted as thermokarst features28,29. There is no doubt that geological surfaces in and around the Arabia Terra region have been
modified by ice30, but we argue that it is unlikely that ice removal
could have created the collapse features themselves. Scalloped depressions in the Utopia Planitia region of Mars bear a striking resemblance
in size, shape and morphology to thermokarst features found on
Earth31,32; both terrestrial and Martian types form depressions on the
order of metres to tens of metres in depth33,34 (Fig. 4). Thus, those wellaccepted thermokarst features are orders of magnitude smaller than
the collapse features discussed here, whereas the proposed volcanic
10 km
b
–2,000
Siloe patera
Flow?
Collapse?
a
Elevation (m)
–3,000
25 km
d
5 km
–2,000
Euphrates patera
c
Figure 3 | The geology of Siloe and Euphrates paterae. MOLA data draped over CTX images show the morphologies of Siloe patera (a; rectangle enlarged in b)
and Euphrates patera (c; rectangle enlarged in d).
3 O C T O B E R 2 0 1 3 | VO L 5 0 2 | N AT U R E | 4 9
©2013 Macmillan Publishers Limited. All rights reserved
RESEARCH ARTICLE
A new category of Martian volcanic construct
Taken together, these features constitute a new category of Martian
volcano that can be described as plains-style caldera complexes, of
which Eden patera is the type example. Eden patera is not associated
with a major edifice. Each of the Martian low-slope paterae recognized previously12,13 shows a major edifice related to repeated volcanic
deposition of explosive and effusive products. Thus, Eden patera
seems to be a new class of Martian volcanic feature, formed through
a combination of magma withdraw, subsurface magma migration
(caldera 1) and/or major explosive episodes that would have distributed ash regionally or globally such that they did not accumulate near
the vent (calderas 2 and 3). These geomorphic features are most
analogous to those of a terrestrial supervolcano.
On Earth a supervolcano is defined as a volcano that can produce at
least 1,000 km3 of volcanic materials in an eruption. On Mars it is
generally not possible to link a single volcanic deposit to a particular
eruption event. However, erupted volumes can be constrained from
the volume of void space observed in the caldera itself, if that collapse
ss 2
ss 3
Class 1
Class 2
Class 3
Class 4
Cla
Clas
s4
a
Cla
2.5
2.0
Siloe patera
(inner basin)
Rim–floor depth (km)
structures in Arabia Terra are of the same scale and morphology as
terrestrial supervolcanoes35 (Fig. 4). If these proposed volcanic structures
are in fact the result of thermokarst, then they are a new type of thermokarst beyond any that has been definitively recognized previously.
In addition, the large volume of the collapse features is a strong
constraint on the possible origins. If they formed by collapse associated with the removal of subsurface ice, it necessarily implies that a
significant volume of ice was removed from each location, quickly
enough to cause the high strain rates required for faulting. However,
none of the features is associated with outflow channels, which are
typically cited as evidence for the rapid removal of surface or nearsurface ice. Furthermore, the amount of ice that could have existed
below such depressions can be constrained from quantitative models
of Martian subsurface porosity36. For example, if Eden patera had
been formed by the removal of subsurface ice, it would have been
necessary for all of the available void space to be entirely saturated
with ice to a depth of ,10 km (see Supplementary Information). We
therefore conclude that, although ice and thermokarst processes could
have been involved in the modification of the collapse features, it is
unlikely to explain the origin of the collapse or the presence of the
large depressions.
It is also possible that the depressions in Arabia Terra represent
degraded impact craters. However, none of the features described above
contain evidence for impact geology, such as the presence of ejecta,
raised crater rims, central peaks or inverted stratigraphy. It is possible
that erosion has removed such evidence, but the proposed calderas are
found adjacent to ancient impact craters of similar size (and possibly
similar age) that have preserved clear evidence for impact origins (Fig. 5).
Furthermore, impact craters that have been degraded by erosion37 have
much lower depth-to-diameter ratios than those measured in the proposed calderas (Fig. 5). The relations between depth and diameter
among the calderas are only consistent with depth-to-diameter ratios
of impact craters that are only partly modified; such craters have preserved at least some critical aspects of impact geology.
Siloe patera
(whole basin)
1.5
Euphrates patera
Eden patera
s
las
1.0
1
C
0.5
1,000
Martian scalloped terrains
Terrestrial thermokarst features
100
Terrestrial supervolcanoes
Length of minor axis (km)
Candidate Martian supervolcanoes
(this study: Eden patera, Euphrates
patera, Oxus patera, Siloe patera)
0
l
Vo
40
ni
ca
10
20
sm
b
60
80
100
120
Crater diameter (km)
Unnamed
crater
140
160
180
Ejecta
1
Siloe patera
(no rim, no ejecta)
st
r
ka
0.1
o
Rim
m
er
Th
0.01
Elevation (m)
–3,000
0.001
0.001
0.01
10
1
Length of major axis (km)
0.1
100
1,000
Figure 4 | Comparison of thermokarst features, terrestrial supervolcanoes
and the putative supervolcanoes on Mars. A plot of the dimensions of typical
terrestrial and Martian thermokarst features shows that they have roughly
similar sizes32,34. The proposed calderas in Arabia Terra have similar
dimensions to those of terrestrial supervolcanoes, which together are orders of
magnitude larger than known thermokarst features.
–2,000
Figure 5 | Comparison of the depth-to-diameter ratios of possible Martian
supervolcanoes with those of known impact craters. a, Plot of crater
measurements for all of the craters within the area of Fig. 1 with diameters of
1 km or more that have previously been categorized according to their level of
preservation37. Class 1 craters are the most degraded and class 4 are the least
degraded (essentially pristine). The proposed supervolcanoes plot along
trendlines associated with moderately modified craters that preserved impact
morphologies. b, However, the calderas clearly do not contain morphological
evidence for impact processes as seen in adjacent craters of similar size.
5 0 | N AT U R E | VO L 5 0 2 | 3 O C TO B E R 2 0 1 3
©2013 Macmillan Publishers Limited. All rights reserved
ARTICLE RESEARCH
is assumed to have occurred as a result of the removal of magma during
eruptive events. Focusing on a subset of these features including Eden
patera, Oxus cavus, Semeyken crater and Ismenia patera, the average
depression volume is more than 3,300 km3. This volume at each site
could have been produced by the removal of a comparable amount
of dense-rock-equivalent material. Assuming an average density of
2,800 kg m23 of the magma and an estimated density of 2,000 kg m23
for erupted lava or 1,300 kg m23 for tephra, it is possible to estimate the
amount of erupted material from each source. The average minimum
erupted volume could be more than 4,600–7,200 km3 for each of these
caldera complexes. Although this estimate cannot be linked to a single
eruption event, nor can we differentiate void space created by explosive
ejection from that created by magmatic subsidence, these features are
unlike other known Martian volcanoes and it is likely that they fall in
the category of terrestrial supervolcano, on the basis of both geomorphology and eruptive potential.
The question remains: Why would large calderas associated with
explosive volcanism occur in northern Arabia Terra? One possibility
is that volatile-rich crust was subducted beneath Arabia Terra during
an ancient episode of plate tectonics38. However, although the presence of northwest–southeast-trending scarps related to thrust faulting
in northern Arabia Terra related to southwest–northeast compression
might seem consistent with such an interpretation, the estimated
displacement on such faults is small and does not support the model
of an active plate margin28,39. It is more likely that the dichotomy
boundary evolved as a result of crustal thinning associated with endogenic processes39. The crust within Arabia Terra is relatively thin and
more similar to thicknesses modelled for the northern lowlands than
for the southern highlands40. Even so, we consider an origin due to
subduction to be an open question that merits further consideration.
We suggest that a combination of regional extension and thermal
erosion of the lower crust in the Late Noachian to Early Hesperian led
to a rapid ascent of magma in the northern Arabia Terra region. It is
not necessary that the magmas were of higher viscosity (more silicic)
or had higher volatile content than other Martian magmas. The lower
gravity and atmospheric pressure on Mars lead to bubble nucleation
at greater depths and greater gas expansion in comparison with
Earth41. As a result, pyroclastic eruptions would be more commonly
associated with basaltic volcanoes on Mars than on Earth, particularly
if the magma rapidly ascended and erupted and was not stored in
degassing magma chambers for long periods, as is thought to occur at
younger, large shield volcanoes42. In fact, it is possible that explosive
volcanism was more prevalent globally on early Mars because the
ancient crust was thinner, leading to less devolatilization of magmas
during ascent. The result may have been the deposition of vast quantities of tephra early in Mars’s history.
Links to global geology
Explosive eruptions of fine-grained materials might be linked to the
formation of fretted terrains that also occur in northern Arabia Terra,
the origin of which represents one of the major outstanding mysteries
in Mars science29 (Fig. 6). Youthful ice-related processes may have
modified the fretted terrains, but the sediment of which they are composed was probably deposited in the Noachian to Early Hesperian43.
These voluminous, fine-grained sediments may be tephra deposits
from explosive volcanic activity in northern Arabia Terra. In fact,
layered terrains throughout Arabia Terra might consist of tephra
deposits, but previous work has suggested that the source region was
the Tharsis province5. A much simpler explanation is that plains-style
calderas produced these sediments locally44 (Fig. 6).
Our understanding of volcanism45 on Mars continues to evolve as
numerous, small (tens of kilometres in diameter) and dispersed volcanic centres are recognized throughout the Tharsis region46–49, and
degraded, ancient volcanic centres are recognized in the southern
highlands12,13. Major volcanic constructs are now recognized in several distinct provinces throughout the Martian surface, although with
a paucity of features previously identified between Tharsis and Syrtis
Major (Fig. 4). The features identified here constitute a major volcanic
province in Arabia Terra, which fills a void in a large fraction of the
surface where volcanoes are expected to have occurred but have never
been recognized.
The origin of altered, fine-grained, layered, clay-bearing and sulphatebearing sediments throughout the equatorial region of Mars has yet to
be explained. A local volcanic source could explain the presence of
clastic materials composing these deposits and could serve as a significant source of volcanogenic sulphur that might have led to acidic
alteration at the surface and strongly perturbed the Martian climate,
sending it into periods of significant warming18 or substantial cooling19.
We suggest that fine-grained deposits at the Meridiani Planum and
Gale crater landing sites, as well as friable deposits in the equatorial
region of Mars, might ultimately have originated from volcanic sources
in Arabia Terra. Further mapping of plains-style caldera complexes
might reveal additional ancient volcanic source regions distributed
throughout the Martian highlands. Deciphering the nature of an early
This study
Layered sulphates
Volcanic provinces
Altered, layered terrains
Clays in stratigraphies
Sedimentary clays
Outflow channels
Fretted terrains, friable materials
Figure 6 | Links to global geology. The distribution of major volcanic provinces on Mars in relation to friable and fretted terrain, layered sulphates17 and layered
clay-bearing terrains50.
3 O C TO B E R 2 0 1 3 | VO L 5 0 2 | N AT U R E | 5 1
©2013 Macmillan Publishers Limited. All rights reserved
RESEARCH ARTICLE
phase of widespread, disseminated, explosive volcanism will be critical
to revealing the climate history and past habitability of Mars.
METHODS SUMMARY
The primary data sets used to evaluate the geomorphology of topographic depressions in the Arabia Terra region were gridded elevation data from the Mars Orbiter
Laser Altimeter (MOLA) and a global mosaic of daytime infrared images from the
Thermal Emission Imaging System (THEMIS). Additional data products included
high-resolution images and digital topographic data from the High Resolution
Stereo Camera (HRSC) aboard the Mars Express spacecraft, high-resolution
images from the High Resolution Imaging Science Experiment (HiRISE) and
Mars Context Imager (CTX) aboard the Mars Reconnaissance Orbiter, and the
Mars Orbiter Camera (MOC) aboard the Mars Global Surveyor spacecraft. These
data products are available within the publicly available Java Mission-planning and
Analysis for Remote Sensing (JMARS) software produced by Arizona State
University (available at http://jmars.mars.asu.edu). Image-based geological mapping was performed after geo-registering these data products within a geographic
information system (GIS). Data from the Thermal Emission Spectrometer (TES)
were used to evaluate dust cover and albedo.
Full Methods and any associated references are available in the online version of
the paper.
Received 8 May; accepted 15 July 2013.
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Malin, M. C. & Edgett, K. S. Sedimentary rocks of early Mars. Science 290,
1927–1937 (2000).
Edgett, K. S. & Malin, M. C. Martian sedimentary rock stratigraphy: outcrops and
interbedded craters of northwest Sinus Meridiani and southwest Arabia Terra.
Geophys. Res. Lett. 29(24), 2179, http://dx.doi.org/10.1029/2002gl016515
(2002).
Hynek, B. M. Implications for hydrologic processes on Mars from extensive
bedrock outcrops throughout Terra Meridiani. Nature 431, 156–159 (2004).
Bibring, J. P. et al. Global mineralogical and aqueous mars history derived from
OMEGA/Mars express data. Science 312, 400–404 (2006).
Hynek, B. M., Phillips, R. J. & Arvidson, R. E. Explosive volcanism in the Tharsis
region: global evidence in the Martian geologic record. J. Geophys. Res. Planets
108, 5111, http://dx.doi.org/10.1029/2003je002062 (2003).
Gendrin, A. et al. Sulfates in martian layered terrains: the OMEGA/Mars Express
view. Science 307, 1587–1591 (2005).
McCollom, T. M. & Hynek, B. M. A volcanic environment for bedrock diagenesis at
Meridiani Planum on Mars. Nature 438, 1129–1131 (2005).
Wilson, L. & Head, J. W. Explosive volcanic eruptions on Mars: tephra and
accretionary lapilli formation, dispersal and recognition in the geologic record.
J. Volcanol. Geotherm. Res. 163, 83–97 (2007).
Kerber, L., Head, J., Madeleine, J. B., Forget, F. & Wilson, L. The dispersal of
pyroclasts from ancient explosive volcanoes on Mars: implications for the friable
layered deposits. Icarus 219, 358–381 (2012).
Greeley, R. & Spudis, P. Volcanism on Mars. Rev. Geophys. 19(1), 13–41 (1981).
Golombek, M. P. et al. Erosion rates at the Mars Exploration Rover landing sites and
long-term climate change on Mars. J. Geophys. Res. Planets 111, E12S10,
http://dx.doi.org/10.1029/2006je002754 (2006).
Williams, D. A. et al. The Circum-Hellas Volcanic Province, Mars: overview. Planet.
Space Sci. 57, 895–916 (2009).
Xiao, L. et al. Ancient volcanism and its implications for thermal evolution of Mars.
Earth Planet. Sci. Lett. 323–324, 9–18 (2012).
Zuber, M. T. The crust and mantle of Mars. Nature 412, 220–227 (2001).
Bandfield, J. L., Edwards, C. S., Montgomery, D. R. & Brand, B. D. The dual nature of
the martian crust: young lavas and old clastic materials. Icarus 222, 188–199
(2013).
Lammer, H. et al. Outgassing history and escape of the Martian atmosphere and
water inventory. Space Sci. Rev. 174, 113–154 (2012).
Gaillard, F., Michalski, J., Berger, G., McLennan, S. M. & Scaillet, B. Geochemical
reservoirs and timing of sulfur cycling on Mars. Space Sci. Rev. 174, 251–300
(2013).
Halevy, I., Zuber, M. T. & Schrag, D. P. A sulfur dioxide climate feedback on early
Mars. Science 318, 1903–1907 (2007).
Tian, F. et al. Photochemical and climate consequences of sulfur outgassing on
early Mars. Earth Planet. Sci. Lett. 295, 412–418 (2010).
Johnson, S. S., Mischna, M. A., Grove, T. L. & Zuber, M. T. Sulfur-induced greenhouse
warming on early Mars. J. Geophys. Res. 113, E08005, http://dx.doi.org/10.1029/
2007JE002962 (2008).
King, P. L. & McLennan, S. M. Sulfur on Mars. Elements 6, 107–112 (2010).
French, B. M. Traces of Catastrophe: A Handbook of Shock-metamorphic Effects in
Terrestrial Meteorite Impact Structures (LPI Contribution no. 954, Lunar and
Planetary Institute, 1998).
Malin, M. & Dzurisin, D. Landform degradation on Mercury, the Moon, and Mars:
evidence from crater depth/diameter relationships. J. Geophys. Res. 82, 376–388
(1977).
24. Acocella, V. Caldera types: how end-members relate to evolutionary stages of
collapse. Geophys. Res. Lett. 33, L18314, http://dx.doi.org/10.1029/
2006gl027434 (2006).
25. Richter, D. H., Eaton, J. P., Murata, K. J., Ault, W. U. & Krivoy, K. L. Chronological
Narrative of the 1959–1960 Eruption of Kilauea Volcano, Hawaii (US Geological
Survey Professional Paper 539-E, 1970).
26. Stovall, W. K., Houghton, B. F., Harris, A. J. L. & Swanson, D. A. Features of lava lake
filling and draining and their implications for eruption dynamics. Bull. Volcanol. 71,
767–780 (2009).
27. Walker, G. P. L. Structure and origin by injection of lava under surface crust of
tumuli, ‘lava rises’, ‘lava-rise pits’, and ‘lava-inflation clefts’ in Hawaii. Bull. Volcanol.
53, 546–558 (1991).
28. McGill, G. E. Crustal history of north central Arabia Terra, Mars. J. Geophys. Res.
Planets 105, 6945–6959 (2000).
29. Sharp, R. P. Mars: Fretted and chaotic terrains. J. Geophys. Res. 78, 4073–4083
(1973).
30. Head, J. W., Mustard, J. F., Kreslavsky, M. A., Milliken, R. E. & Marchant, D. R. Recent
ice ages on Mars. Nature 426, 797–802 (2003).
31. Niu, F. J., Lin, Z. J., Liu, H. & Lu, J. H. Characteristics of thermokarst lakes and their
influence on permafrost in Qinghai-Tibet Plateau. Geomorphology 132, 222–233
(2011).
32. Bouchier, A. Response to Permafrost Failures on Hillslopes in the Brooks Range,
Alaska. MS thesis, Colorado School of Mines (2008).
33. Soare, R. J., Osinki, G. R. & Roehm, C. L. Thermokarst lakes and ponds on Mars in
the very recent (late Amazonian) past. Earth Planet. Sci. Lett. 272, 382–393 (2008).
34. Sejourne, A. et al. Scalloped depressions and small-sized polygons in western
Utopia Planitia, Mars: A new formation hypothesis. Planet. Space Sci. 59, 412–422
(2011).
35. Miller, C. F. & Wark, D. A. Supervolcanoes and their explosive supereruptions.
Elements 4, 11–15 (2008).
36. Clifford, S. M. et al. Depth of the Martian cryosphere: Revised estimates and
implications for the existence and detection of subpermafrost groundwater.
J. Geophys. Res. Planets 115, E07001, http://dx.doi.org/10.1029/2009je003462
(2010).
37. Robbins, S. J. & Hynek, B. M. A new global database of Mars impact craters $1 km:
1. Database creation, properties, and parameters. J. Geophys. Res. 117, E05004
http://dx.doi.org/10.1029/2011je003966 (2012).
38. Sleep, N. Martian plate tectonics. J. Geophys. Res. 99, 5639–5655 (1994).
39. Watters, T. R. Thrust faults along the dichotomy boundary in the eastern
hemisphere of Mars. J. Geophys. Res. 108 (E6) 5054 (2003).
40. Neumann, G. A. et al. Crustal structure of Mars from gravity and topography.
J. Geophys. Res. 109, E08002, http://dx.doi.org/10.1029/2004JE002262 (2004).
41. Wilson, L. & Head, J. W. Mars: review and analysis of volcanic eruption theory and
relationships to observed landforms. Rev. Geophys. 32, 221–263 (1994).
42. Wilson, L., Scott, E. D. & Head, J. W. Evidence for episodicity in the magma supply to
the large Tharsis volcanoes. J. Geophys. Res. 106(E1), 1423–1433 (2001).
43. Irwin, R. P., Watters, T. R., Howard, A. D. & Zimbelman, J. R. Sedimentary resurfacing
and fretted terrain development along the crustal dichotomy boundary, Aeolis
Mensae, Mars. J. Geophys. Res. Planets 109, E09011, http://dx.doi.org/10.1029/
2004je002248 (2004).
44. Kerber, L., Michalski, J., Bleacher, J. & Forget, F. in 44th Lunar and Planetary Science
Cconference, Lunar and Planetary Institute, Houston, Texas, USA, abstract 2290
(2013).
45. Hodges, C. A. & Moore, H. J. Atlas of Volcanic Landforms on Mars (US Geological
Survey Professional Paper 1534, 1994).
46. Bleacher, J. E., Greeley, R., Williams, D. A., Cave, S. R. & Neukum, G. Trends in
effusive style at the Tharsis Montes, Mars, and implications for the development of
the Tharsis province. J. Geophys. Res. Planets 112, E09005, http://dx.doi.org/
10.1029/2006je002873 (2007).
47. Hauber, E., Bleacher, J., Gwinner, K., Williams, D. & Greeley, R. The topography and
morphology of low shields and associated landforms of plains volcanism in the
Tharsis region of Mars. J. Volcanol. Geotherm. Res. 185, 69–95 (2009).
48. Bleacher, J. E. et al. Spatial and alignment analyses for a field of small volcanic
vents south of Pavonis Mons and implications for the Tharsis province, Mars.
J. Volcanol. Geotherm. Res. 185, 96–102 (2009).
49. Richardson, J. A., Bleacher, J. E. & Glaze, L. S. The volcanic history of Syria Planum,
Mars. J. Volcanol. Geotherm. Res. 252, 1–13 (2013).
50. Ehlmann, B. L. et al. Subsurface water and clay mineral formation during the early
history of Mars. Nature 479, 53–60 (2011).
Supplementary Information is available in the online version of the paper.
Acknowledgements We thank H. Frey, B. Hynek, S. Wright, J. Zimbelman and
L. Tornabene for discussions that improved the quality of the manuscript. Funding was
provided by the NASA Mars Data Analysis programme.
Author Contributions J.R.M. performed the initial observations, processed image and
topographic data and wrote most of the manuscript. J.E.B. wrote portions of the
manuscript, performed geological mapping and processed imaging and topographic
data. Both authors synthesized the results, developed the ideas and edited the paper.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial interests.
Readers are welcome to comment on the online version of the paper. Correspondence
and requests for materials should be addressed to J.R.M. (michalski@psi.edu).
5 2 | N AT U R E | VO L 5 0 2 | 3 O C TO B E R 2 0 1 3
©2013 Macmillan Publishers Limited. All rights reserved
ARTICLE RESEARCH
METHODS
Identification of volcanic features. Most well-recognized volcanic edifices on
Mars occur as central vent structures within topographically elevated terrain built
through sustained volcanism around the vent. Pavonis Mons (Supplementary Fig. 1)
is an example of typical shield-style volcanism on Mars. Note that Pavonis Mons
contains evidence for collapse and crustal sagging owing to removal or migration
of magma. The central caldera is a steep-sided, nearly circular crater that formed
during the latest stage of volcanic activity. However, that caldera is nested within a
larger set of ring-fractures that suggest more extensive collapse or additional
collapse events. Complex calderas are common on the Earth and Mars, and occur
as a result of collapse associated with magma withdrawal, owing to migration of
magma at depth, removal of magma during eruptions, or both. Tyrrhenus Mons
(Supplementary Fig. 1b) is an ancient volcano of different character on Mars. It
also is defined by a topographic rise with ring fractures. However, the flanks of
Tyrhennus Mons have a much lower profile than Pavonis Mons and are composed
of fingering, eroded, layered materials thought to indicate the presence of pyroclastic materials. Tyrrhena patera (the main caldera) might be the final location of
the central vent, but the caldera is breached and eroded, and there is evidence for
secondary calderas on the volcano. The plains-style caldera complexes that we
have identified in northern Arabia Terra bear characteristics similar to each of
these volcanoes, yet some other characteristics that are fundamentally different.
Most notably, the calderas in Arabia Terra do not occur on topographically elevated
volcanic constructs, which is probably one reason why they have never previously
been identified as volcanoes despite abundant evidence for volcanic processes.
The International Astronomical Union (IAU) formally named six features (five
paterae and one cavus) located in northern Arabia Terra in 2012. These features
have not been discussed by their proper names in previous literature. As discussed
in the main text, these features, as well as Semeykin crater (which was previously
named) have morpohological characteristics that are inconsistent with impact
origins. They are not the only depressions in Arabia Terra with enigmatic origins,
but they are the subset of features on which this paper is focused.
Supplementary Fig. 2 shows the morphology of all seven features discussed in
the text. Of these, Eden patera and Euphrates patera bear the strongest evidence
for ancient volcanism. The others probably formed through collapse, although
the link to volcanism is less clear in the other cases.
Calculation of volumes. One of the goals of this work is to constrain the amount of
collapse that occurred at each of the putative caldera complexes. Estimates of collapse
volume are important for placing minimum constraints on the amount of magma
involved in ancient igneous processes at each site, and for testing alternative hypotheses for the origins of these features (for example pseudokarst, described below).
To estimate the amount of collapse that has occurred, we mapped the features in
a GIS environment and used MOLA elevation data to calculate the volume of each
depression. The volume calculations are straightforward but depend on several
assumptions. First we describe the technical process, and then the assumptions.
For each site, gridded MOLA data were contoured and draped onto MOLA
hillshade and THEMIS daytime infrared data. The contoured data helped to delineate the maximum topographic level of the depression at each feature. We then
converted gridded MOLA elevation data to triangulated irregular networks (TINs)
at each site. The TINs provide a combined quantitative measure of surface elevation
and area. Then, for each site, we fitted a plane to the maximum allowable elevation
corresponding to the closest approximation to a closed depression. We then calculated the volume of void space beneath the plane, within the caldera at each site.
Examples of the volume calculations are shown in Supplementary Fig. 3. Note
that the fit of an elevation plane to each site is imperfect. One assumption we
make is that topography has not changed since the formation of the depressions.
This is clearly not so, but it is a limitation on our approach. There is clear evidence
that the entire region has been tilted towards the north since the formation of
these features. In addition, several of the calderas described in this work show
evidence that they were breached, which means that there is not an obvious closed
depression at most structures. Therefore, delineation of a single closed depression
grossly underestimates the actual volume of the structure because the calderas are
typically breached at some elevation along the rim. Last, younger impact craters
have been superimposed on each site, which further complicates the effort to define
a single elevation contour related exclusively to the caldera collapse itself. Given
these challenges, we have made every effort to perform the volume calculations with
the most conservative approach possible, to avoid overestimating the volume of
each depression. We have therefore chosen elevations that in each case are below the
rim of the depression, to provide the best estimate of a closed depression with the
knowledge that this decision results in an underestimate of the total caldera volume.
There are errors associated with these analyses, both in the direction of artificially increasing the volume estimates and in the direction of artificially decreasing them. One of the major errors resulting in underestimation of the volume
calculations is related to the fill deposits within the depressions themselves. Those
materials were probably sourced from the caldera in each case, but their topographic setting now is still considered part of the underlying terrain. In other
words, there is no way to identify the true caldera floor because friable fill deposits
bury the floor in most cases. We are calculating volumes of the void space that exists
above modern topographic depression in each case. Our calculations therefore
actually correspond to the volume of the caldera that has not been filled by friable
materials, lavas or colluvial deposits.
There are two sources of error that lead to overestimation of volumes. The first
is related to the erosional breaching of rims of the depression. In fitting a plane to
the best estimate of the closed depression, there is still some additional volume
added by calculating void space above the plains surrounding the breached
depression. However, we made every effort to avoid this bias as much as possible,
and the errors that did occur are likely to have been small. Another bias includes
the calculation of void space within superimposed impact craters that have interior depressions rivalling the depth of the caldera itself (see Supplementary Fig. 4).
However, these errors are again extremely small and do not change the calculated
volumes appreciably.
Could the depressions have formed by pseudokarst? Mars is in many ways a
periglacial planet. Permafrost is likely to be (and to have been) much more widespread and geologically important at the global scale on Mars than on Earth.
Catastrophically melted subsurface ice has been postulated as a likely source for
water that carved immense outflow channels on the surface. It has also implicated
in the formation of terrains bearing periglacial features such as fields of pitted
terrain, as seen in some parts of the Elysium basin. The possibility that the collapse
features described in this work could have formed from the removal of subsurface
ice there bears consideration.
To test this hypothesis we used the volume calculations described above to constrain how much ice must have been removed to produce the collapse by removal of
ice from the subsurface. Models describing the amount and distribution of subsurface ice on Mars have been produced36. These calculations include models of
subsurface porosity as a function of depth. Using those models of porosity, we
can then calculate the amount of pore space that could potentially have been filled
with ice beneath a given feature. In other words, is there enough pore space
available that, even if entirely filled with subsurface ice, would result in the collapse
volume of the depression if all of that ice were removed?
The best test case is Eden patera. Here, ,4,000 km3 of void space exists. If that space
was created by means of collapse that was related to removal of ice, it stands to reason
that the ice must have been present essentially beneath the feature itself. If the ice was
widely distributed in area, its removal would probably have produced multiple small
collapse pits (if any at all) or regional subsidence. We therefore focus on the area of the
depression itself. In the case of Eden patera, this area is roughly equal to 5,000 km2.
Supplementary Figure 4 shows the decay of porosity with depth on Mars and the
cumulative volume of void space beneath an area of 5,000 km2 beneath Eden
patera. Pore space decays to near zero by a depth of ,10 km. If all of the void
space to this depth were completely filled with ice, it would result in a total volume
of ,4,000 km3—roughly equal to the volume of collapse at Eden patera. Therefore,
the calculations, to first order, suggest that the volume of collapse at Eden patera
could potentially be explained theoretically by the removal of subsurface ice.
However, we suggest that the calculations present a compelling case that ice was
not solely responsible for the formation of the collapse at Eden patera because they
imply that all of the void space became filled with ice to a great depth and then all of
that ice was somehow removed from the subsurface without leaving any traces of
fluvial features (that is, outflow channels) that could be related to catastrophic melting.
These volume estimates provide some constraints on the amount of material
that was erupted from plains-style caldera complexes in the northern Arabia
Terra region. The volumes of the depressions represent, in the strictest sense,
the amount of void space produced by a combination of structural collapse and
eruption of lavas and/or pyroclastics. Structural collapse could occur as a result of
withdrawal of magma, or migration of a magma chamber at depth, and the voids
therefore do not necessarily relate directly to erupted volumes. However, explosive eruptions often continue to fragment magma within the volcano’s conduit,
and the final caldera volume can also be an underestimate of an eruption’s total
volume. These calculations therefore provide some guidance on the scale of the
eruptive potential of the Arabia volcanic province.
Assuming that the void space within calderas relates directly to the removal of
magma during eruptions, we can produce some simple scaling calculations to
estimate how much material may have been erupted. By assuming a dense rock
equivalent (DSE) of caldera volume equal to a typical mafic magma with density
of 2,800 kg m23, we can then scale the DSE for a lava density of 2,000 kg m23 or a
tephra of density 1,000-1,500 kg m23. Using these scaling factors and the volume
calculations described above, we calculated the estimated minimum erupted
volumes described in the text.
©2013 Macmillan Publishers Limited. All rights reserved