The first 5 pages, shown as text so search engines and readers can find this document.
Page 1
Articles
https://doi.org/10.1038/s41561-019-0512-61GEOPS – Géosciences Paris Sud, Univ. Paris-Sud, CNRS, Université Paris-Saclay, Orsay, France. 2IMCCE – Observatoire de Paris, CNRS-UMR 8028, Paris,
France. 3California Institute of Technology, Pasadena, CA, USA. 4Geosciences Environnement Toulouse, UMR 5563 CNRS, IRD & Université de Toulouse,
Toulouse, France. 5Laboratoire de Planétologie et Géodynamique, CNRS UMR 6112, Université de Nantes, Université d’Angers, Nantes, France.
6Lunar and Planetary Laboratory, University of Arizona, Tucson, AZ, USA. 7Institut de Minéralogie, de Physique des Matériaux, et de Cosmochimie
(IMPMC) – Sorbonne Université- Muséum National d’Histoire Naturelle, UPMC Université Paris 06, UMR CNRS 7590, IRD UMR 206, Paris, France.
8Department of Earth, Environmental and Planetary Sciences, Rice University, Houston, TX, USA. *e-mail: sylvain.bouley@u-psud.fr
The crust of Mars has been shaped by 4.5 Gyr of meteorite impacts,
volcanism, tectonism and surface processes (Fig. 1a,b). Its most
prominent crustal features are the hemispheric dichotomy, the
Tharsis volcanic province and several large impact basins. The hemi-
spheric dichotomy describes Mars’s north–south asymmetry, where
the northern lowlands have roughly half the crustal thickness of the
southern highlands. The Tharsis volcanic province represents the
thickest region of the crust and is associated with a prominent topo-
graphic rise responsible for deformation of the lithosphere at a plan-
etary scale1. Whereas the crustal thickness of Tharsis is considered
to be the result of magmatic intrusions and volcanic eruptions, the
origin of the hemispheric dichotomy is more enigmatic. Past studies
have suggested that the hemispheric crustal dichotomy is the result
of either giant impact(s)2,3 or hemispheric-scale mantle upwelling4,5.
While previous studies have quantified the contribution of
Tharsis to Mars’s gravity field1, crustal thickness6 and topogra-
phy7, the contribution of impact basins to the crustal structure has
never been quantified. Mars has four giant (diameters >1,000 km)
unequivocal impact basins with unambiguous gravity field anom-
alies and expressions in crustal thickness maps (Hellas, Argyre,
Utopia and Isidis), but only three of them have a significant topo-
graphic expression (Hellas, Argyre and Isidis). Utopia has a muted
topographic signature, due to its early formation and possible crustal
relaxation8. The four impacts (Hellas, Argyre, Isidis and Utopia)
predate Tharsis, and extensively modified Mars’s crustal structure.
For example, Hellas is surrounded by an annulus of high-standing
topography and thickened crust, which has been interpreted to be
the consequence of ejecta deposits and crustal thickening during
crater excavation and collapse9–11 (Extended Data Fig. 1). Several
outcrops in Terra Cimmeria and Terra Sirenum are interpreted as
a veneer of Hellas ejecta over older crust—roughly four crater radii
away from the centre of the basin12. Investigations of lunar gravity
and topography, coupled with hydrocode impact simulations, have
suggested that the Moon’s hemispheric dichotomy (where the far-
side crust is about twice as thick as the nearside crust) may be par-
tially explained by the deposition of a thick ejecta blanket around
the giant South Pole–Aitken impact basin13,14. Extrapolating these
results to Mars, it appears necessary to determine the contribution
of large impact basins to its present crustal structure.
Reconstructing Mars without impact basins and volcanoes
There are many methods for isolating, characterizing and removing
the contribution of impact basins and volcanoes from the gravity
field and topography of planets—each with varying degrees of com-
plexity1,6,15,16. In this work we develop a method for removing these
features using crustal thickness maps17,18 (Fig. 1c,d). These maps—
which are derived from topography and gravity data (Methods)—
can be used for reconstruction of the early crustal structure with
conservation of mass arguments. The present crustal thickness
model (model B, ref. 18) accounts for higher crustal densities in vol-
canic complexes (2,900 kg m−3), lower crustal densities elsewhere
(2,582 kg m−3) and a mantle density of 3,500 kg m−3. To simplify
calculations, the crustal thickness was rescaled at every location to
the same crustal density (that is, we increase (decrease) the crustal
A thick crustal block revealed by reconstructions
of early Mars highlands
Sylvain Bouley 1,2*, James Tuttle Keane 3, David Baratoux4, Benoit Langlais5, Isamu Matsuyama 6,
Francois Costard1, Roger Hewins7, Valerie Payré8, Violaine Sautter7, Antoine Séjourné1,
Olivier Vanderhaeghe 4 and Brigitte Zanda2,7
The global-scale crustal structure of Mars is shaped by impact basins, volcanic provinces, and a hemispheric dichotomy with a
thin crust beneath the northern lowlands and a thick crust beneath the southern highlands. The southern highlands are com-
monly treated as a coherent terrain of ancient crust with a common origin and shared geologic history, plausibly originating
from a giant impact(s) or a hemispheric-scale mantle upwelling. Previous studies have quantified the contribution of volcanism
to this crustal structure; however, the influence of large impacts remains unclear. Here we present reconstructions of the past
crustal thickness of Mars (about 4.2 Gyr ago) where the four largest impact basins (Hellas, Argyre, Isidis and Utopia) are
removed, assuming mass conservation, as well as the main volcanic provinces of Tharsis and Elysium. Our reconstruction shows
more subdued crustal thickness variations than at present, although the crustal dichotomy persists. However, our reconstruc-
tion reveals a region of discontinuous patches of thick crust in the southern highlands associated with magnetic and geochemi-
cal anomalies. This region, corresponding to Terra Cimmeria–Sirenum, is interpreted as a discrete crustal block. Our findings
suggest that the southern highlands are composed of several crustal blocks with different geological histories. Such a complex
architecture of the southern highlands is not explained by existing scenarios for crustal formation and evolution.
NATuRe GeOSCIeNCe | www.nature.com/naturegeoscience
Page 2
Articles Nature GeoscieNcethickness in regions of high (low) density until the crustal density is
globally uniform). This lets us use crustal mass and crustal volume
conservation interchangeably.
We assume that impacts and volcanoes modify the crustal struc-
ture symmetrically about their centres, and that impact basins
reshape the crust in a manner that approximately conserves crustal
12 km
10 km
8 km
6 km
4 km
2 km
0 km
−2 km
−4 km
−6 km
−8 km
Topography
a b
c d
e f
55 km
50 km
45 km
40 km
35 km
30 km
25 km
20 km
15 km
Crustal thickness
55 km
50 km
45 km
40 km
35 km
30 km
25 km
20 km
15 km
Crustal thickness,
with impact basins
and volcanoes
removed
? ?
Tharsis
Rise
Alba
Patera
Elysium
Mons
Hellas
Planitia
Utopia
Planitia
Borealis
Isidis
Planitia
Argyre
Planitia
Olympus
Mons
Hellas
Planitia
Tharsis
Rise
Isidis
Planitia
Utopia
Planitia
Argyre
Planitia
Alba
Patera
Ascraeus
Mons
Pavonis
Mons
Arsia
Mons
Terra
Cimmeria
Terra
Sirenum
Terra Cimmeria–
Sirenum crustal
block
Borealis
Gale
Crater
Thaumasia Thaumasia
Fig. 1 | A global view of the crustal structure of Mars. a,b, The Mars Orbiter Laser Altimeter topography of Mars with the features of interest labelled9.
c,d, The crustal thickness of Mars based on crustal model B (ref. 18, Methods). e,f, The crustal thickness of Mars after removing all of the large impact basins and
volcanic features. The Cimmeria–Sirenum crustal block is enclosed by a dash–dot line. The question mark to the east of the block indicates the unclear boundary
between the Thaumasia region and the Cimmeria–Sirenum block due to the possible specific origin of Thaumasia19. The maps are in Lambert azimuthal equal-
area projection, centred on the equator at longitudes 0° (left column) and 180° (right column). Each map covers all of Mars except for a small region antipodal
to the map centre. The maps are overlaid on the present-day topography for reference. Grid lines are in increments of 30° of latitude and longitude.
NATuRe GeOSCIeNCe | www.nature.com/naturegeoscience
Page 3
ArticlesNature GeoscieNcemass. Figure 2 shows an example of our fitting process for the Hellas
basin. To remove the feature of interest, we first calculate the azi-
muthally averaged crustal thickness as a function of distance from
the centre of the feature (Fig. 2a). Next, we define an annulus out-
side the feature (the ‘transition zone’, Fig. 2a,b), which we take to
represent the background crustal thickness. We calculate the mean
crustal thickness in the transition zone, and then calculate the mass
of crustal material above and below that datum within the ‘fitting
zone’. If the mass above the datum (the mass excess) equals the mass
below the datum (the mass deficit), then we deem the background
crustal thickness to be a mass-conserving solution. We include a
cosine taper to prevent discontinuities in the corrected crustal
thickness map. For each feature of interest, we performed a param-
eter-space search for mass-conserving crustal structure correc-
tions—testing a range of possible transition zone radii and widths
(Extended Data Fig. 3). In cases where multiple mass-conserving
solutions are found, we select the solution that makes the smallest
change to the crustal structure. In cases where no mass-conserving
solution is found, we select the solution that comes closest to con-
serving mass. While the formation of impact basins should conserve
crustal mass, the same cannot be assumed for large volcanic edifices
(as the erupted mass may be derived from the mantle and not the
crust). Thus, for removing volcanoes, we prescribe a nominal angu-
lar radius of the feature based on the slope break at its edge that
is identified on the azimuthally averaged crustal thickness profiles.
We then remove the crustal mass above the background, defined
from an average value in an annulus surrounding the volcanic prov-
ince (Extended Data Fig. 4).
Using this technique, we sequentially isolated and removed the
axisymmetric crustal thickness structures associated with the four
largest impact basins (Hellas (Fig. 2), Argyre, Isidis and Utopia),
Elysium Mons and the Tharsis Rise, plus the five largest individual
volcanoes: Olympus Mons, Arsia Mons, Pavonis Mons, Ascraeus
Mons and Alba Patera. This method is iterative and nonlinear. That
is, starting with the present crustal thickness of Mars, we remove
one feature, and then use that iterated crustal thickness as the
starting point for removing the next feature, and so on (Extended
Data Fig. 5). This means that the resulting corrected crustal struc-
ture depends a priori on the order in which we remove features,
although the effect of the order is minimal if the features are distant
enough from each other, which is the case here. To evaluate this
dependence, we repeated this analysis using randomized removal
sequences. We find that the solution is only weakly dependent on
the order of removal (Methods). With this precaution taken, our
nominal solution uses a removal sequence based on our current
knowledge of the relative chronology of basins and volcanoes8, from
youngest to oldest: Elysium, Olympus and Tharsis montes, Tharsis
Rise, Argyre, Isidis, Hellas and Utopia.
Subdued thickness variations of the early Martian crust
Removing impact basins and volcanic constructs reveals more sub-
dued crustal thickness variations (Figs. 1e,f and 3). The hemispheric
dichotomy persists in the reconstruction. Hellas contributes to the
southern highlands, but excavation of the basin and emplacement
0 15 30 45 60 75 90 105 120 135 150 165 180
Distance (°)
0
20
40
60
80
100
Crustal thickness (km)
0
a
b
1 2 3 4 5 6 7 8 9 10
Distance (feature radii)
0 1 2 3 4 5 6 7 8 9 10
Distance (feature radii)
0 15 30 45 60 75 90 105 120 135 150 165 180
0
0.5
1
Filter
Distance (°)
Mass
excessMass
deficit
Feature
radius
Fitting zone Transition
zone
Fitting zone
filter, h
Transition zone
filter, g
Fig. 2 | The radial crustal structure of Hellas Planitia. a, The azimuthally
average crustal thickness about Hellas (solid black line with hatching
beneath). We calculate the mean crustal thickness in the transition zone,
and then calculate the mass of crustal material above and below that
datum within the ‘fitting zone’. The mass above the mass-conserving
crustal thickness (the mass excess, red) integrated in an annulus around
the feature is equal to the mass below it (the mass deficit, blue). The
peak at 180° distance is Alba Patera, which is approximately antipodal
to Hellas. b, The fitting and transition zone filters used in this
reconstruction (Methods).
42.8 ± 12.8
40.0 ± 7.8
30.3 ± 3.7 45.0 ± 7.7
43.1 ± 4.8
50.8 ± 2.7
54.2 ± 20.3
10 15 20 25 30 35 40 45 50 55 60 65 70 >75
0
1
2
3
4
5
6
a
b Fractional surface area (%)
0
1
2
3
4
5
6
7
Fractional surface area (%)
Mars’s present crustal thickness
Mars’s crustal thickness without volcanoes and impact basins
Crustal thickness (km)
10 15 20 25 30 35 40 45 50 55 60 65 70 >75
Crustal thickness (km)
Southern
highlands
(excluding
Cimmeria–
Sirenum)
Northern
lowlands
Total
Cimmeria–
Sirenum
Southern
highlands
Northern
lowlands
Total
Tharsis and
Elysium
Tharsis and
Elysium
(excluding
Cimmeria–
Sirenum)
8
31.4 ± 2.9
38.9 ± 6.7
Fig. 3 | Histograms of crustal thickness of three domains (northern
lowlands, southern highlands and Cimmeria–Sirenum) of the Martian
crust. a, Mars's present crustal thickness. b, Mars’s crustal thickness
without volcanoes and impact basins. Note that the contribution of
Tharsis and Elysium is isolated to the northern lowlands to visualize the
residual contribution after removal. The black histograms show the global
distribution of crustal thickness; when summed together, those coloured
histograms equal the global histogram. The horizontal error bars indicate
the mean and standard deviation for each respective distribution.
NATuRe GeOSCIeNCe | www.nature.com/naturegeoscience
Page 4
Articles Nature GeoscieNceof ejecta are unable to explain the entire dichotomy; the mass excess
in the southern highlands is about four times larger than the mass
deficit within the Hellas basin.
However, after removing all of the large impact basins and vol-
canic features, the crustal thickness map highlights a crustal block
characterized by discontinuous, thick (>50 km) regions corre-
sponding to Terra Cimmeria–Sirenum (Fig. 1e,f). We delineated
the boundary of this block such that it includes all of the neighbour-
ing >50-km-thick crustal segments (Extended Fig. 6). We excluded
the Thaumasia region, even though it has comparable thickness,
because it may have a different tectonic (orogenic) history19. The
relation between the most eastern part of the Terra Cimmeria–
Sirenum crustal block and Thaumasia is unclear (Fig. 1e,f).
The distribution of crustal thicknesses suggests that the early
Martian crust can be divided into three main crustal domains hav-
ing characteristic thickness ranges that partially overlap (Fig. 3b):
the northern lowlands with a thickness of between 20 and 38 km
(centred at 33 km); the southern highlands—including Arabia Terra,
Hellas and Argyre—with an intermediate thickness of between 30
and 55 km (centred at 46 km); and the Terra Cimmeria–Sirenum
block with a thickness of between 42 and 60 km (centred at 50 km).
A thick crustal block in the southern highlands
The Cimmeria–Sirenum crustal block is also noticeable because it
shares many of the characteristics of terrestrial continental crust.
On Earth, continental crust is derived via differentiation and re-
melting of pre-existing basaltic crust. It possesses an intermediate
composition between felsic and mafic, is enriched in incompatible
elements, and is associated with thicker crust and elevated topogra-
phy. The Terra Cimmeria–Sirenum region is associated with anom-
alously elevated pre-Tharsis topography20 (Extended Data Fig. 7)
and includes the strongest crustal magnetic anomalies21–24 (Fig. 4a),
which, by comparison with the Earth, could correspond to the accre-
tion of terranes25. Terra Cimmeria–Sirenum is also associated with
some of the highest surface (down to the first metre) abundances of
potassium (K) and thorium (Th) (Fig. 4b,c)26. This region is the only
place in the southern terrains where K and Th are correlated. Within
this region, lower concentrations of K and Th are noted in the region
near the south pole, where ground ice may mask the true (ice-free)
K and Th (ref. 26). Only the most eastern part of the crustal block
(~10% of its surface area) is not associated with these geochemi-
cal and magnetic anomalies. Possible fragments of continental crust
have been found with in situ observations by the Curiosity rover at
Gale crater27 on the northern edge of the Cimmeria–Sirenum block
(Fig. 1a). While this interpretation appears inconsistent with the
vast amount of spectroscopic observations suggesting that the sur-
face of this region is dominated by basaltic rock28, it is plausible that
these basalts are surficial deposits overlying a geochemically distinct
crust. This would explain why exposed continental components are
rare in the southern hemisphere29.
The existence of (at least) one crustal block on Mars makes
the pre-Noachian geologic history of Mars—and the origin of the
hemispheric dichotomy—much more complex than previously rec-
ognized. We identify a regional-scale crustal block covering 10%
of Mars. Geophysical and geochemical signatures in this region
are suggestive of a geochemically evolved crustal component.
Combined with the identification of continental crust-like material
on Mars in specific outcrops30, and inferred from meteoritic samples
containing zircons27,31,32, our findings suggest that the Cimmeria–
Sirenum crustal block had a complex geologic history, whether or
10
Magnetic
field
intensity (nT)
100
1,000
a
b
c
10,000
0.2
Potassium
abundance
(wt%)
0.3
0.4
0.5
Geochemically altered
0.3
Thorium
concentration
(ppm)
0.4
0.5
0.6
0.7
0.8
0.9
Geochemically altered
Fig. 4 | The geophysical and geochemical signature of the Cimmeria–
Sirenum block. a, The magnetic field intensity, evaluated at the surface
of Mars24. b, The potassium concentration (wt%)26. c, The thorium
concentration (ppm)26. In b and c, the northern hemisphere is hatched out
to focus attention on the signatures south of the dichotomy. The maps are
in Lambert azimuthal equal-area projection, centred on 0°, and cover all
of Mars except for a small region on the opposite hemisphere. The maps
are overlaid on the present-day topography for reference. The Cimmeria–
Sirenum crustal block is enclosed by a dash-dot line.
NATuRe GeOSCIeNCe | www.nature.com/naturegeoscience
Page 5
ArticlesNature GeoscieNcenot it is analogous to terrestrial continental crust. Since this ter-
rain is overprinted by the Hellas and Argyre basins, it is probably
the oldest part of the crust. Subsequent large impacts significantly
altered Mars’s crustal structure and contributed to the observed
hemispheric dichotomy. Therefore, the southern highlands should
not be considered as a single, homogeneous unit (requiring a single
formation mechanism), but rather a collection of crustal blocks with
plausibly different origins, variably affected by magmatic, tectonic
and impact processes. The formation mechanism of several crustal
blocks during the pre-Noachian (before 4.1 Gyr ago) remains to
be deciphered and has important implications for the planet’s
evolution, including its climate, the surface environment and the
mantle dynamics.
Online content
Any methods, additional references, Nature Research reporting
summaries, source data, extended data, supplementary informa-
tion, acknowledgements, peer review information; details of author
contributions and competing interests; and statements of data and
code availability are available at https://doi.org/10.1038/s41561-
019-0512-6.
Received: 26 January 2019; Accepted: 19 November 2019;
Published: xx xx xxxx
References
1. Zuber, M. T. & Smith, D. E. Mars without Tharsis. J. Geophys. Res. 102,
28673–28686 (1997).
2. Wilhelms, D. E. & Squyres, S. W. The martian hemispheric dichotomy may be
due to a giant impact. Nature 309, 138–140 (1984).
3. Frey, H. & Shultz, R. A. Large impact basins and the mega-impact origin for
the crustal dichotomy on Mars. Geophys. Res. Lett. 15, 229–232 (1988).
4. Zhong, S. & Zuber, M. T. Degree-1 mantle convection and the crustal
dichotomy on Mars. Earth Planet. Sci. Lett. 189, 75–84 (2001).
5. Roberts, J. H. & Zhong, S. Degree-1 convection in the Martian mantle and
the origin of the hemispheric dichotomy. J. Geophys. Res. 111, E06013 (2006).
6. Andrews-Hanna, J. C., Zuber, M. T. & Banerdt, W. B. The Borealis basin and
the origin of the martian crustal dichotomy. Nature 453, 1212–1215 (2008).
7. Matsuyama, I. & Manga, M. Mars without the equilibrium rotational figure,
Tharsis, and the remnant rotational figure. J. Geophys. Res. Planets 115,
12020 (2010).
8. Fassett, C. I. & Head, J. W. Sequence and timing of conditions on early Mars.
Icarus 211, 1204–1214 (2011).
9. Smith, D. E. et al. The global topography of Mars and implications for surface
evolution. Science 284, 1495–1503 (1999).
10. Zuber, M. T. et al. Internal structure and early thermal evolution of Mars
from Mars Global Surveyor topography and gravity. Science 287,
1788–1793 (2000).
11. Zuber, M. The crust and mantle of mars. Nature 412, 220–227 (2001).
12. Irwin, R. P., Tanaka, K. L. & Robbins, S. J. Distribution of Early, Middle,
and Late Noachian cratered surfaces in the Martian highlands: implications
for resurfacing events and processes. J. Geophys. Res. Planets 118,
278–291 (2013).
13. Melosh, H. J. et al. South Pole–Aitken basin ejecta reveal the Moon’s upper
mantle. Geology 45, 1063–1066 (2017).
14. Zuber, M. T., Smith, D. E., Lemoine, F. G. & Neumann, G. A. The shape and
internal structure of the Moon from the clementine mission. Science 266,
1839–1843 (1994).
15. Keane, J. T. & Matsuyama, I. Evidence for lunar true polar wander and
a past low-eccentricity, synchronous lunar orbit. Geophys. Res. Lett. 41,
6610–6619 (2014).
16. Garrick-Bethell, I., Perera, V., Nimmo, F. & Zuber, M. T. The tidal-rotational
shape of the Moon and evidence for polar wander. Nature 512,
181–184 (2014).
17. Genova, A. et al. Seasonal and static gravity field of Mars from MGS, Mars
Odyssey and MRO radio science. Icarus 272, 228–245 (2016).
18. Goossens, S. et al. Evidence for a low bulk crustal density for Mars from
gravity and topography. Geophys. Res. Lett. 44, 7686–7694 (2017).
19. Nahm, A. L. & Schultz, R. A. Evaluation of the orogenic belt hypothesis for
the formation of the Thaumasia highlands, Mars. J. Geophys. Res. 115,
E04008 (2010).
20. Bouley, S. et al. Late Tharsis formation and implications for early Mars.
Nature 531, 344–347 (2016).
21. Connerney, J. E. P. et al. The global magnetic field of Mars and implications
for crustal evolution. Geophys. Res. Lett. 28, 4015–4018 (2001).
22. Arkami-Hamed, J. A coherent model of the crustal magnetic field of Mars.
J. Geophys. Res. 109, E09005 (2004).
23. Langlais, B., Purucher, M. E. & Mandea, M. Crustal magnetic field of Mars.
J. Geophys. Res. 109, E02008 (2004).
24. Langlais, B. et al. A new model of the crustal magnetic field of Mars using
MGS and MAVEN. J. Geophys. Res. Planets 124, 1542–1569 (2019).
25. Fairén, A. G., Ruiz, J. & Anguita, F. An origin for the linear magnetic
anomalies on Mars through accretion of terranes: implications for dynamo
timing. Icarus 160, 220–223 (2002).
26. Boynton, W. V. et al. Concentration of H, Si, Cl, K, Fe, and Th in the
low- and mid-latitude regions of Mars. J. Geophys. Res. 112, E12S99 (2007).
27. Sautter, V. et al. In situ evidence for continental crust on early Mars. Nat.
Geosci. 8, 605–609 (2015).
28. Rogers, A. D. et al. Global spectral classification of Martian low-albedo
regions with Mars Global Surveyor Thermal Emission Spectrometer
(MGS-TES) data. J. Geophys. Res. 112, E02004 (2007).
29. Baratoux, D. et al. Petrological constraints on the density of the Martian
crust. J. Geophys. Res. Planets 119, 1707–1727 (2014).
30. Christensen, P. R. et al. Evidence for magmatic evolution and diversity on
Mars from infrared observations. Nature 436, 504–509 (2005).
31. Humayun, M. et al. Origin and age of the earliest Martian crust from
meteorite NWA 7533. Nature 503, 513–516 (2013).
32. Bouvier, L. C. et al. Evidence for extremely rapid magma ocean
crystallization. Nature 586, 586–589 (2018).
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.
© The Author(s), under exclusive licence to Springer Nature Limited 2020
NATuRe GeOSCIeNCe | www.nature.com/naturegeoscience