ARTICLE IN PRESS

Planetary and Space Science 57 (2009) 895–916

Contents lists available at ScienceDirect

Planetary and Space Science

0032-06

doi:10.1

� Corr

E-m

mccord

lmanfre

(J.J. Plau1 N2 Te

journal homepage: www.elsevier.com/locate/pss

The Circum-Hellas Volcanic Province, Mars: Overview

David A. Williams a,�, Ronald Greeley a, Robin L. Fergason a,1, Ruslan Kuzmin b, Thomas B. McCord c,Jean-Phillipe Combe c, James W. Head IIId, Long Xiao a,e, Leon Manfredi a, Franc-ois Poulet f, Patrick Pinet g,David Baratoux g, Jeffrey J. Plaut h, Jouko Raitala i, Gerhard Neukum j,2, the HRSC Co-Investigator Teama School of Earth and Space Exploration, Arizona State University, Box 871404, Tempe, AZ 85287-1404, USAb Vernadsky Institute, Russian Academy of Sciences, Kosygin Street 19, Moscow 117975, GSP-1, Russiac The Bear Fight Center, 22 Fiddler’s Road, Box 667, Winthrop, WA 98862-667, USAd Department of Geological Sciences, Brown University, Providence, RI 02912, USAe Faculty of Earth Sciences, China University of Geosciences, 388 Rumo Rd., Hongshan Dist., Wuhan 430074, Chinaf Institut d’Astrophysique Spatiale, Batiment 121, Universite Paris-Sud, 91405 Orsay Cedex, Franceg Observatoire Midi-Pyrenees, Laboratoire Dynamique Terrestre et Planetaire, UMR 5562, CNRS, Universite Paul-Sabatier, 14, Avenue Edouard Belin, 31 400 Toulouse, Franceh NASA Jet Propulsion Laboratory, 4800 Oak Grove Drive, Mail Stop 183-501, Pasadena, CA 91109, USAi Astronomy Division, Department of Physical Sciences, University of Oulu, Oulu, Finlandj Freie Universitaet Berlin, Department of Earth Sciences, Institute of Geosciences, Planetary Sciences and Remote Sensing, Malteserstr. 74-100, Building D, D-12249 Berlin, Germany

a r t i c l e i n f o

Article history:

Received 14 April 2008

Received in revised form

30 July 2008

Accepted 5 August 2008Available online 23 August 2008

Keywords:

Mars volcanism

Mars express

Remote sensing

Crater statistics

33/$ - see front matter & 2008 Elsevier Ltd. A

016/j.pss.2008.08.010

esponding author. Tel.: +1480 965 7029; fax:

ail addresses: David.Williams@asu.edu (D.A

tb@aol.com (T.B. McCord), jean-philippe_co

d@asu.edu (L. Manfredi), francois.poulet@ias.

t), jraitala@sun3.oulu.fi (J. Raitala), gneukum

ow at: United States Geological Survey, Astro

l.: +49 30 83870 579.

a b s t r a c t

Building on previous studies of volcanoes around the Hellas basin with new studies of imaging

(High-Resolution Stereo Camera (HRSC), Thermal Emission Imaging System (THEMIS), Mars Orbiter

Camera (MOC), High-Resolution Imaging Science Experiment (HiRISE), Context Imager (CTX)),

multispectral (HRSC, Observatoire pour la Mineralogie, l’Eau, les Glaces et l’Activite (OMEGA)),

topographic (Mars Orbiter Laser Altimeter (MOLA)) and gravity data, we define a new Martian volcanic

province as the Circum-Hellas Volcanic Province (CHVP). With an area of 42.1 million km2, it contains the

six oldest central vent volcanoes on Mars, which formed after the Hellas impact basin, between 4.0 and

3.6 Ga. These volcanoes mark a transition from the flood volcanism that formed Malea Planum �3.8 Ga,

to localized edifice-building eruptions. The CHVP volcanoes have two general morphologies: (1) shield-

like edifices (Tyrrhena, Hadriaca, and Amphitrites Paterae), and (2) caldera-like depressions surrounded

by ridged plains (Peneus, Malea, and Pityusa Paterae). Positive gravity anomalies are found at Tyrrhena,

Hadriaca, and Amphitrites, perhaps indicative of dense magma bodies below the surface. The lack of

positive-relief edifices and weak gravity anomalies at Peneus, Malea, and Pityusa suggest a fundamental

difference in their formation, styles of eruption, and/or compositions. The northernmost volcanoes, the

�3.7–3.9 Ga Tyrrhena and Hadriaca Paterae, have low slopes, well-channeled flanks, and smooth

caldera floors (at tens of meters/pixel scale), indicative of volcanoes formed from poorly consolidated

pyroclastic deposits that have been modified by fluvial and aeolian erosion and deposition. The �3.6 Ga

Amphitrites Patera also has a well-channeled flank, but it and the �3.8 Ga Peneus Patera are dominated

by scalloped and pitted terrain, pedestal and ejecta flow craters, and a general ‘softened’ appearance.

This morphology is indicative not only of surface materials subjected to periglacial processes involving

water ice, but also of a surface composed of easily eroded materials such as ash and dust.

The southernmost volcanoes, the �3.8 Ga Malea and Pityusa Paterae, have no channeled flanks, no

scalloped and pitted terrain, and lack the ‘softened’ appearance of their surfaces, but they do contain

pedestal and ejecta flow craters and large, smooth, bright plateaus in their central depressions.

This morphology is indicative of a surface with not only a high water ice content, but also a more

consolidated material that is less susceptible to degradation (relative to the other four volcanoes).

We suggest that Malea and Pityusa (and possibly Peneus) Paterae are Martian equivalents to Earth’s

giant calderas (e.g., Yellowstone, Long Valley) that erupted large volumes of volcanic materials, and that

Malea and Pityusa are probably composed of either lava flows or ignimbrites. HRSC and OMEGA spectral

ll rights reserved.

+1480 965 8102.

. Williams), greeley@asu.edu (R. Greeley), rfergason@usgs.gov (R.L. Fergason), rok@geokhi.ru (R. Kuzmin),

mbe@bearfightcenter.com (J.-P. Combe), James_Head@brown.edu (J.W. Head III), longxiao@cug.edu.cn (L. Xiao),

u-psud.fr (F. Poulet), barataoux@dtp.obs-mip.fr (P. Pinet), pinet@dtp.obs-mip.fr (D. Baratoux), plaut@mail.jpl.nasa.gov

@zedat.fu-berlin.de (G. Neukum).

geology Research Program, 2255 N. Gemini Drive, Flagstaff, AZ 86001, USA.

ARTICLE IN PRESS

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916896

data indicate that dark gray to slightly red materials (often represented as blue or black pixels in HRSC

color images), found in the patera floors and topographic lows throughout the CHVP, have a basaltic

composition. A key issue is whether this dark material represents concentrations of underlying basaltic

material eroded by various processes and exposed by aeolian winnowing, or if the material was

transported from elsewhere on Mars by regional winds. Understanding the provenance of these dark

materials may be the key to understanding the volcanic diversity of the Circum-Hellas Volcanic

Province.

& 2008 Elsevier Ltd. All rights reserved.

1. Introduction

More than half the surface of Mars is covered by volcanicdeposits, including lava flows and inferred pyroclastic materials,reflecting volcanic activity that has spanned much of the knownhistory of the planet (see reviews by Greeley and Spudis, 1981;Mouginis-Mark et al., 1992; Neukum et al., 2004a). In the early1970s, the first successful Mars orbiter, Mariner 9, returned low-resolution images (1–3 km/pixel) that revealed the geologicaldiversity of the surface, including the presence of features nearthe Hellas impact basin that appeared to be volcanoes (McCauleyet al., 1972; Carr, 1973; Carr et al., 1973; Potter, 1976; Peterson,1977). These features had radial patterns of ridges and channels,hints of slightly circular craters, and relatively low topographicrelief. The term patera was applied to these and similar featuresseen on Mars. In planetary nomenclature (http://planetaryna-mes.wr.usgs.gov/jsp/append5.jsp), patera refers to a shallow craterwith a scalloped or complex edge (Strobell and Masursky, 1990).However, in planetary literature for Mars, the term also commonlyrefers to volcanic constructs (e.g., Carr, 1981). The term highland

paterae was applied to a category of low-profile volcanoes foundmostly in the cratered highlands (Plescia and Saunders, 1979) andis also commonly used, especially for the features in the Hellasregion.

As reviewed below, in the years following Mariner 9, geologicalmapping and many topical investigations were conducted for thepaterae of the Hellas region, which can now be enhancedsubstantially by new and better data. Building on this body ofwork and the insight provided by the current Mars missions, wepropose the formal designation of the Circum-Hellas Volcanic

Province (CHVP), shown in Figs. 1a and b.

2. Background

In their geological analysis of the Mariner 9 data, McCauleyet al. (1972), Carr (1973), and Carr et al. (1973) recognized thevolcanic origins of the features around the Hellas basin andsummarized the volcanic history of Mars, with ages for thesethree paterae (later named Tyrrhena, Hadriaca, and Amphitrites)of 3.5–4.0 Ga, based on crater counts (Carr, 1976). As part of thesystemic Mars geological mapping program (see also Scott andCarr, 1978), Potter (1976) analyzed the eastern part of the Hellasbasin and Amphitrites Patera, and noted flow-like patterns thathe suggested were produced by fluid lava flows. At about the sametime, Peterson (1977) mapped the western part of the Hellas basinand identified many of the volcanic centers, later codified in hisprescient paper (Peterson, 1978), in which he suggested that thepaterae are centered on ring fractures associated with the Hellasimpact structure (Fig. 1). Later studies by Frey et al. (1991) inferredfour rings from a previously unrecognized basin (centered onwhat is now recognized as Pityusa Patera) with diameters of 600,1200, 1700, and 2200–2400 km, but they admitted that some ofthe features noted as part of these rings could also apply to ringsof the Hellas basin, as suggested by Peterson (1978). Peterson

(1978) also noted the numerous mare-like wrinkle ridges inthe areas around the paterae, and suggested that these could besurface manifestations of feeder dikes. Because the paterae havelow topographic relief, Peterson (1978) suggested that theyare composed either of ultramafic lavas, which were erupted asvery low-viscosity, fluid lavas, or as ash-flows, analogous to thoseof Earth’s Yellowstone volcanic region.

In the mid-1970s, the Viking orbiters began returning imagessubstantially improved over those of Mariner 9, enabling refine-ments of the ideas proposed earlier (Carr et al., 1977a).For example, Greeley and Spudis (1981) suggested that Tyrrhenaand Hadriaca Paterae are multistage volcanoes that began as low-profile ash-shields built from explosive eruptions resulting frommagma rising through water-rich mega-regolith. As near-surfacegroundwater became less abundant due to depletion or freezing,explosivity diminished, and eruptions evolved into an effusivephase to form the lava flows inferred near the summits.Subsequently, the ash deposits were eroded to form the channelson the flanks of the paterae. Later image analyses, geologicmapping, and eruption modeling studies (Greeley and Crown,1990; Crown et al., 1992; Crown and Greeley, 1993, 2007; Wilsonand Head, 1994; Gregg et al., 1998; Gregg and Farley, 2006) furtherelucidated these processes.

The Hellas paterae were mapped as part of the post-Viking

global geologic mapping program at a scale of 1:15 M, in whichGreeley and Guest (1987) mapped Tyrrhena and Hadriaca Pateraeand Tanaka and Scott (1987) mapped Amphitrites Patera and thenow named Peneus Patera. Malea and Pityusa Paterae (namedlater) were mapped as volcanic materials (Pityusa Patera as a‘volcanic patera’), and much of the entire area west–southwest ofHellas, now named Malea Planum, was mapped and interpreted tobe composed of various volcanic units, including fluid lavaflows and/or pyroclastic flows (Tanaka and Scott, 1987). Later,1:5 M-scale geologic mapping was done by Tanaka and Leonard(1995; the published US G.S. map is by Leonard and Tanaka, 2001),in which they studied the general geology of the Hellas area anddescribed both the basin-filling material and Malea Planum,including the paterae. Tanaka and Leonard (1995) suggested thatthe south side of the Hellas basin was filled with material mostlikely to be lava flows 41 km thick, erupted from Malea Planum(MOLA data do show a topography suggestive of flows from MaleaPlanum flowing into the basin), and later Tanaka et al. (2002)suggested catastrophic erosion of the southern Hellas basin rimoccurred because of magmatic intrusion into volatile-rich rocks.Tanaka and Leonard (1995) mapped northern Malea Planum ascontaining upper-Hesperian to lower-Amazonian pyroclastic flowdeposits, heavily dissected by the fluvial channels of Axius Valles,derived from Amphitrites and Peneus Paterae, and suggested thatthe distinctive caldera of Peneus indicated eruptions of massivevolumes of magma, whereas the less pronounced morphology ofAmphitrites was more indicative of a dissected shield that eruptedlower volumes of magma. They noted a 20-km-diameter depres-sion on the west flank of Amphitrites Patera that they thoughtmight be a collapsed vent structure. Tanaka and Leonard (1995)analyzed the Hellas basin rim slopes dissected by Axius Valles,

ARTICLE IN PRESS

Fig. 1. (a) MOLA shaded relief map of the Circum-Hellas Volcanic Province (CHVP), outlined in black (dashed where inferred), with an area of 42.1 million km2. The white

circles delineate the calderas of the CHVP volcanoes, and white lines represent their respective flanks (dashed where inferred), based on our image analyses and

comparisons to the geologic maps of Gregg et al. (1998), Leonard and Tanaka (2001), Crown and Greeley (2007), and Kolb and Tanaka (in review). The black ‘‘rings’’ of the

Hellas impact structure are from Peterson (1978), in which he inferred that the formation of Hellas influenced the formation of the named volcanoes. (b) (Color online)

MRO95a gravity anomaly map of Mars, with a blowout of the CHVP (area approximate to (a)). Strong positive gravity anomalies are present at Tyrrhena (TP), Hadriaca (HP),

and Amphitrites Paterae (AP), whereas weaker ones appear to correlate with Peneus (PP), Malea (MP), and Pityusa Paterae (PityP). We suggest that the strong gravity

anomalies could indicate the presence of solidified, relatively dense mafic magma bodies underneath these volcanoes. Gravity data courtesy of G.A. Neumann, NASA

Goddard SFC. Gravity model MRO95a by Alexander Konopliv, JPL (available at: http://pds-geosciences.wustl.edu/missions/mro/gravity.htm).

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 897

and suggested that the presence of a hilly surface reflecteddegraded lava flows, lahars, or volcaniclastic materials. Incontrast, they mapped southern Malea Planum (i.e., south ofAmphitrites and Peneus Paterae) as dominated by upper-Noa-chian to lower-Hesperian ridged plains, which they interpreted aslow-viscosity lava flows from Amphitrites and Peneus.

Crumpler et al. (1996) analyzed calderas on Mars anddescribed Amphitrites and Peneus Paterae based on Viking data.They described the radial and concentric pattern of mare-typeridges around Amphitrites and noted their resemblance to the so-called ‘‘arachnoids’’ on Venus, the category of low-profile volcanofirst seen in low-resolution Venera 15/16 images (Barsukov et al.,1986) and later in better detail in Magellan images (Head et al.,1992).

Some of the first Mars Global Surveyor (MGS) data analysisof this region was by Head and Pratt (2001a), who characterizedthe topography of the ‘Malea Planum volcanic province’ based onMOLA data. They found embayed craters, suggesting flooding byflows, and quantified the elevations of terrain, showing that Malea

Planum stands some 1000–1500 km above Mars datum, and thatAmphitrites is a distinctive shield 1.5 km high and contains acaldera 300–600 m deep, among other measurements.

With time, data from the current generation of Mars orbitersbegan to accumulate over the southern latitudes in the areas thatstill are difficult to observe because of atmospheric conditions,enabling testing of ideas that had been proposed previously.For example, Plescia (2003, 2004) used MOLA data and agreedthat Amphitrites and Peneus are well-defined calderas based ontheir topography, but suggested that the features now calledMalea and Pityusa Paterae are heavily eroded or buried cratersand not necessarily volcanic vents. He also noted that the region isheavily mantled, with the terrain showing ‘‘scalloped’’ morphol-ogies in MOC images, and that the paucity of small craterssuggests extensive resurfacing.

Most recently, Kolb and Tanaka (in review) completedgeological mapping of the south polar region and identified Maleaand Pityusa Paterae as volcanoes, although heavily altered byother processes. Larson (2007), in an unpublished Master’s thesis,

ARTICLE IN PRESS

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916898

conducted a thorough study of the paterae of Malea Planum usingMOLA, MOC, and THEMIS data, and suggested that Malea andPityusa Paterae might have been intrusive, and could not identifyany lava flow features associated with them.

In summary, highland paterae appear to be the oldestrecognized volcanoes on Mars that formed in association withedifice-building eruptions. Although earlier volcanism is inferredto have occurred on Mars (see e.g., Tanaka et al., 1992) there are novisible morphological indicators such as flows, and the inter-pretation is based on remote-sensing data that suggest iron andmagnesium-rich compositions, such as basalt (McCord et al.,1982; Pinet and Chevrel, 1990; Bandfield et al., 2000). By analogy

Fig. 2. THEMIS daytime-IR mosaics of the volcanoes of the CHVP, with spatial resoluti

301S, 931E) and Tyrrhena Patera (right, caldera center coordinates 211S, 106.51E); (b) M

center coordinates 57.71S, 52.71E), and Malea Paterae (caldera center coordinates 63.81S,

mosaics, the black-lined regions delineate crater count areas for the ‘‘Edifice’’ as given in

all six volcanoes. The arrow in (b) points to a dome that could be a small volcanic constru

suggesting post-eruption deformation. Black zones are gaps in coverage. Image proces

with lunar lava flows and flood basalts on Earth, the putative earlyMartian basalts are thought to have erupted from fissures, tracesof which have been buried by their own products (i.e., ridgedplains). Recent data have revealed evidence for extensive ridgesinterpreted to be exposed dikes related to flood basalt-likeemplacement of the plains (e.g., Head et al., 2006). Thus, in thisinterpretation, highland paterae may represent a change in thestyle of volcanism on Mars from fissure eruptions (that typicallyinvolve rapid outpourings of large volumes of magma to producevast sheets of flood lavas, e.g., Hesperia Planum) to local centralvents that involved smaller ‘‘batches’’ of magma erupted at lowerrates. We test these ideas in the work described below.

ons of 100 m/pixel. (a) Mosaics of Hadriaca Patera (left, caldera center coordinates

osaic of Amphitrites (caldera center coordinates 58.71S, 60.71E), Peneus (caldera

521E); (c) Mosaic of Pityusa Patera (caldera center coordinates 671S, 38.51E). For all

Table 1; the white-lined regions delineate crater count areas for the ‘‘Caldera’’ for

ct. Mare-like ridges cut across the floors of Amphitrites, Peneus, and Malea Paterae,

sing by Chris Edwards and Robin Ferguson, Arizona State University.

ARTICLE IN PRESS

Fig. 3. Cumulative impact crater size–frequency distribution (SFD) plots for the six volcanoes of the CHVP: (a) Hadriaca Patera; (b) Tyrrhena Patera; (c) Amphitrites Patera;

(d) Peneus Patera; (e) Malea Patera; (f) Pityusa Patera. Refer to Fig. 2 for count areas. For detailed crater statistics, see Table 1.

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 899

ARTICLE IN PRESS

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916900

3. Data and methods

Image analyses constitute the primary data used in this study,coupled with crater counts, topographic measurements, andmultispectral analyses, on which details are provided in appendix.The most complete areal coverage is from Mariner 9, Viking

Orbiters 1-2, the Mars Express (MEX) High-Resolution StereoCamera (HRSC: Neukum et al., 2004b), and the Mars Odyssey

(MO) Thermal Emission Imaging System (THEMIS: Christensenet al., 2004). Very high-resolution images of limited coverage areavailable from the MGS Mars Orbiter Camera (MOC: Malin andEdgett, 2001) and the Mars Reconnaissance Orbiter (MRO) High-Resolution Imaging Science Experiment (HiRISE: McEwen et al.,2007) and Context Imager (CTX: Malin et al., 2007). In addition,visible and near-infrared multispectral data at high-spectralresolution are available for much of the area from the Mars

Express Observatoire pour la Mineralogie, l’Eau, les Glaceset l’Activite (OMEGA) spectrometer (Bibring et al., 2004). HRSCdata are also used as 5-channel multispectral data. It should benoted, however, that images and multispectral data for geo-morphic analyses are often compromised in the more southerlylatitudes of Mars by clouds and dust in the atmosphere.Topographic data from the Mars Global Surveyor Mars OrbiterLaser Altimeter (MOLA: Zuber et al., 1992; Smith et al. 1999, 2001),MGS and MRO gravity data (Lemoine et al., 2001; G.A. Neumann,pers. comm., 2008), and subsurface imaging from the MEX MarsAdvanced Radar for Subsurface and Ionosphere Sounding(MARSIS: Picardi et al., 2004) were also used in this study.

4. Results

In the following sections, we describe sets of paterae thatconstitute the CHVP, beginning with an overview of the well-known pair, Tyrrhena and Hadriaca Paterae northeast of the Hellas

Fig. 4. (Color online) HRSC false color mosaics of Hadriaca Patera (caldera center coord

Note the blue patches in crater and channel floors and other topographic lows, and on so

consistent with mafic (typically basaltic) materials (McCord et al., 2007).

basin (Fig. 2a), discussion of the new data for Amphitritesand Peneus Paterae southwest of the Hellas basin (Fig. 2b), andconsideration of the potential volcanic nature of Malea andPityusa Paterae (Figs. 2b and c).

4.1. Tyrrhena and Hadriaca Paterae

Tyrrhena and Hadriaca Paterae are the northernmost volcanoesof the CHVP (Figs. 1 and 2a), and the least modified. Their geologyhas been relatively well described in Viking and post-Viking dataanalyses (e.g., Greeley and Spudis, 1981; Greeley and Crown, 1990;Crown et al., 1992; Crown and Greeley, 1993, 2007; Gregg et al.,1998; Gregg and Farley, 2006; Williams et al., 2007, 2008).The first volcanic stage to form the constructs of these pateraeappears to have involved eruptions of ash, perhaps driven byexplosions resulting from water–magma interactions, to producelow-profile, shield-like edifices and large summit calderas.The volcanoes were then eroded, as evidenced by radial channelsseparated by mesas. This phase was followed by eruptions of fluidlavas that flooded the calderas and formed local flows on theflanks of the volcanoes (Greeley and Crown, 1990; Crown et al.,1992; Crown and Greeley, 1993). Since the cessation of volcanismand either concurrent or subsequent fluvial activity, the dominantprocess acting on Tyrrhena and Hadriaca Paterae has beendifferential aeolian erosion and deposition, as evidenced by brightmantling and dune deposits visible in MOC images (Williamset al., 2007).

Further studies based on HRSC, MOC, and THEMIS imagesinclude crater counts to assign cratering model ages (Williamset al., 2007, 2008) to units defined by the Viking image-basedgeological mapping (Gregg et al., 1998; Crown and Greeley, 2007).Both Hadriaca and Tyrrhena Paterae formed �3.7–3.9 Ga, after theHellas impact (�4 Ga: Werner and Neukum, 2003; Werner, 2005).Major volcanic event(s) occurred on their flanks and calderas

inates 301S, 931E) and Tyrrhena Patera (caldera center coordinates 211S, 106.51E).

me mesas. Previous HRSC color-spectral studies indicated such blue color units are

ARTICLE IN PRESS

Fig. 5. MOLA profiles across the four southern paterae of the CHVP. Profile AA1 crosses Peneus (PP) and Amphitrites Paterae (AP, left to right). Profile BB1 crosses Pityusa

(PityP), Malea (MP), and Amphitrites Paterae (left to right). Profile CC1 crosses Malea and Peneus Paterae (bottom to top). The analysis of these profiles is consistent with

Head and Pratt (2001a), such that a definite shield-like edifice exists for Amphitrites Patera, but not for Peneus Patera. If Peneus, Malea, and Pityusa Paterae are volcanoes,

then their eruption styles are different from Amphitrites, Hadriaca, and Tyrrhena Paterae.

Fig. 6. (Color online) HRSC false color mosaic of Amphitrites, Peneus, and Malea

Paterae. Arrows mark locations of concentrated (black arrows), buried or partially

buried (white arrows) dark deposits. Previous HRSC color-spectral studies

indicated such blue to black color units are consistent with mafic (basaltic)

materials (McCord et al., 2007). Image processing by David Williams, ASU.

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 901

through �3.3–3.7 Ga, and later resurfacing activity by fluvial,volcanic and/or aeolian processes occurred from �3.3 to as recentas 1.1 Ga at Hadriaca Patera (Williams et al., 2007) and as recent as0.8 Ga at Tyrrhena Patera (Williams et al., 2008).

For this project, we used THEMIS daytime-IR mosaics at100 m/pixel spatial resolution as a consistent image base toperform crater counts to assess the ages of all six paterae in theCHVP (Figs. 2 and 3). The crater counting followed standardpractices (Crater Statistics Analysis Group, 1979), and wedetermined N(1) cratering model ages following the method ofHartmann and Neukum (2001); see also methods discussion inthe appendix). Our results give a cratering model age of 3.7 Ga forHadriaca Patera (whole volcano, excluding the caldera: Fig. 3a).This result is consistent with Williams et al. (2007), who countedsmaller and different areas of the edifice on higher-resolutionHRSC nadir images, and found N(1) model ages ranging3.7–3.9 Ga. For Tyrrhena Patera (whole volcano, excluding thecaldera: Fig. 3b), our results give a cratering model age of 3.8 Ga.This result is in good agreement with Williams et al. (2008), whocounted different areas of the edifice on HRSC nadir images andfound N(1) cratering model ages of formation ranging 3.7–4.0 Ga.Given the similarities in cratering model ages resulting from theHartmann and Neukum (2001) method applied to both higher-resolution HRSC images and lower-resolution THEMIS day-IRmosaics, we have greater confidence in our crater counts using theTHEMIS mosaics for the other four volcanoes in the CHVP thathave not been studied previously with this technique.

We also used HRSC color images (Fig. 4) to assess variations inmaterials composing Hadriaca and Tyrrhena Paterae. The HRSCimages show darker areas in topographic lows (crater and channelfloors) or downslope from the summits, and in some cases on thetops of mesas, which we interpret as concentrations of basalticmaterial from the interiors of the volcanoes exposed by aeolianwinnowing, or perhaps produced by glacial or fluvial erosion ofunderlying basalt and concentrated by local or regional winds(e.g., Baratoux et al., 2007b). This is consistent with previousinterpretations of dark material observed in HRSC color data asbeing basaltic in composition (McCord et al., 2007). Alternatively,these darker areas could be basaltic sand or dust depositstransported from elsewhere on Mars and deposited by aeolianprocesses, although it is unclear where the long-distancesource(s) are located. The relative brightness of these volcanoflanks is indicative of considerable dust mantling, likely deposited

ARTICLE IN PRESS

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916902

by regional winds around the Hellas basin and attesting to thedominant role of current aeolian processes on Mars.

Finally, we compared MGS (Lemoine et al., 2001) and MROgravity data (G.A. Neumann, pers. comm., 2008; gravity modelMRO95a by Alexander Konopliv, JPL, available at: http://pds-geos-ciences.wustl.edu/missions/mro/gravity.htm) for the CHVP region(Fig. 1b). Results show strong positive gravity anomalies corre-lated with Tyrrhena, Hadriaca, and particularly AmphitritesPaterae (Lemoine et al., 2001), and much weaker anomalies forPeneus, Malea, and Pityusa Paterae. Although a comprehensiveanalysis of these martian gravity data has not been performed, wesuggest that these data showing strong gravity anomalies couldindicate the presence of relatively dense magma bodies under-neath these edifices.

4.2. Amphitrites and Peneus Paterae

Most of this study focused on post-Viking images and otherdata to understand the four features in Malea Planum interpretedto be volcanoes: Amphitrites, Peneus, Malea, and Pityusa Paterae.In this section we describe Amphitrites and Peneus Paterae.MOC and HiRISE images and MOLA profiles (Fig. 5) providetantalizing clues to the origin of these features. However, MOChigh-resolution frames (�4 m/pixel) are isolated and cover toosmall an area to enable an integrated analysis of the features.HRSC nadir (�12 m/pixel) and THEMIS visible (18 or 36 m/pixel)images, along with HRSC color (Fig. 6) and THEMIS (Fig. 2b)daytime-IR images, both at 100 m/pixel, provide coverage overboth features under relatively clear atmospheric conditions.

Fig. 7. (a) MOLA context for Amphitrites Patera (16 pixel/deg). (b and c) THEMIS VIS

scalloped and pitted terrain. The subdued and ‘‘softened’’ appearance suggests geo

V14889005 (34 m/pixel).

MOLA profiles (Fig. 5) show that Amphitrites Patera is about280 km across and stands 1000–1300 m above the surroundingplains, forming a broad shield-shaped edifice superposed on olderridged plains of Malea Planum to the south. The summit ofAmphitrites Patera contains a caldera 112 km in diameter witha floor 300–600 m below the rim. Much of the surface ofAmphitrites is modified by Barnard crater, a 130 km diametercentral-ring impact structure superposed on the southern flank ofthe volcano. Although ejecta flow lobes from the impact mantlethe patera floor, some process(es) (e.g., periglacial modification,aeolian mantling, or post-impact eruptions) have erased smallerfeatures, including secondary impact craters.

Where not obviously covered by ejecta from large impacts, theflanks of Amphitrites consist of ridges as large as 8 km wide by100 km long that are approximately radial to the caldera (Fig. 2b).Segments of channel-like depressions 5 km wide and 35 km longoccur between some of the ridges. Lobate flows are seen in someof these depressions and in many parts of the flank, especially inthe western zone (Fig. 7). Cross-sections of flank materialsexposed in the channel walls do not appear to be layered, atleast at the limit of MOC resolution (�4 m/pixel). Smaller channelsegments about 300 m wide and chains of small (200 m), elongatecraters suggest the presence of lava flow channels and partlycollapsed lava tubes (Fig. 8). Although the crater chains areapproximately radial to Barnard crater and could be secondarycraters, they appear to follow local topography, which would notbe expected for ballistically emplaced impacts. A small (�18 km indiameter) dome on the southwest flank of Amphitrites Pateracould be a volcanic construct (Fig. 2b).

images of the patera’s floor, showing lobate-like patterns marking the scarps of

morphic degradation. (b) THEMIS VIS V06577004 (17 m/pixel). (c) THEMIS VIS

ARTICLE IN PRESS

Fig. 8. Extract from THEMIS day-IR image mosaic showing an irregular channel

segment 3–5 km wide, within which are found small elongate craters (arrows)

interpreted as possible collapsed sections of lava tubes, seen on the northwest

flank of Amphitrites Patera.

Fig. 9. Cumulative impact crater size–frequency distribution plot for a section of

dissected plains in Malea Planum. Refer to Fig. 2b for count area. For detailed crater

statistics, see Table 1. The cratering model age for the dissected terrain is �3.8 Ga,

much older than the late-Hesperian to early-Amazonian age predicted by Tanaka

and Leonard (1995).

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 903

Leonard and Tanaka (2001) suggested that late-Hesperian toearly-Amazonian pyroclastic deposits from both Amphitrites andPeneus Paterae flowed north over the rim of the Hellas basin.These relative ages correspond to cratering model ages o3.5 Ga toas young as 2.1 Ga (the Hesperian–Amazonian boundary occursat 3.3 Ga (Neukum system) to 2.9 Ga (Hartmann system) as givenin Hartmann and Neukum, 2001; Hartmann, 2005 lists theboundary as 3.3–2.0 Ga). We tested this hypothesis by selectinga section of dissected terrain north of Peneus and AmphitritesPaterae on the THEMIS 100 m/pixel mosaic for crater counting(Fig. 2b). We found a cratering model age of �3.8 Ga for thisdissected unit (Fig. 9), indicating that its formation occurred muchearlier in martian history. Leonard and Tanaka (2001) alsosuggested that late-Noachian to early-Hesperian lava flowsmake up the ridged plains south of Amphitrites and PeneusPaterae (�3.8–43.5 Ga: Hartmann and Neukum, 2001). We usedthe terrain surrounding the Pityusa Patera caldera (the Pityusa‘‘edifice’’: Table 1) as an example of ridged plains, and we found acratering model age of �3.8 Ga for this unit. This result isconsistent with the mapping of Leonard and Tanaka (2001),indicating the ridged plains of Malea Planum are late-Noachian inage. For Amphitrites Patera as a whole (whole volcano, excludingthe caldera: Figs. 2b and 3c), we determined a cratering model ageof 3.6 Ga. This is slightly younger within uncertainties with earlierresults for the formation of Hadriaca and Tyrrhena Paterae(Williams et al., 2007, 2008), and younger than the ridged plainsof Malea Planum. Amphitrites’ caldera also has a cratering modelage of �3.6 Ga.

The Amphitrites Patera caldera (Figs. 2b, 6, and 7) is ringedwith discontinuous, shallow graben, and subtle ridges. Much ofthe caldera floor is smooth in comparison to the flanks of the

volcano. Ejecta flow lobes from Barnard crater cover parts of thesouthern floor, and secondary craters and ejecta from smaller,relatively fresh impact craters are traced over parts of thenorthern and eastern floor. However, the central and westernparts of the caldera floor are smooth and lack obvious chains ofsecondary craters from Barnard crater and are interpreted to post-date the impact. The caldera floor is disrupted by mare-like ridges,some of which extend over the caldera rim onto the northernflank, suggesting post-eruption deformation of the summit area.Because of atmospheric effects in the HRSC color data, the extentof dark material in the SW floor of the caldera and southeasternflank of the volcano is unclear (Fig. 6).

Peneus Patera (Figs. 2b, 6, and 10) is about 290 km across and,unlike Amphitrites Patera, shows little topographic relief from itsouter flanks to the central caldera (Fig. 5). Radial ridges andchannel segments are present, but are less prominent than onAmphitrites and are seen mostly on the northern and easternflanks of the volcano. Similar features might also be on the otherflanks, but ejecta from impact craters are superposed on thesurface, masking the underlying features. The craters include the100 km diameter impact crater Henry Moore, the crater Chamanon the southern flank and the crater Eilat on the northwesternflank (Fig. 2b). Small sinuous channels �200 m wide and localflows are also seen in the topographically lower parts of the flanksof Peneus Patera. Our crater counts on the THEMIS mosaic indicatea cratering model age of �3.8 Ga (total shield, excluding thecaldera: Fig. 3d). This age is equivalent to that of the dissected andridged plains in Malea Planum, and suggests that volcanic activityat Peneus began before that at Amphitrites. The caldera of Peneus

ARTICLE IN PRESS

Table 1Crater count statistics for the CHVP using THEMIS 100 m daytime-IR mosaics and technique described in Neukum (1983), Hartmann and Neukum (2001), and Michael and

Neukum (2008)

Name Area

(km2)

Number of

craters

Crater size range for

age estimate (km)

Cumulative crater

density N(1)

N(1) error

(7 or +, �)

Cumulative crater

density N(10)

N(10) error

(7 or +, �)

Cratering model age t

(Ga), +error, �error

Notes

Tyrrhena

Patera

660 This Study

Caldera

floora

346 45 0.5–0.9 1.63E�3 6.6E�4 2.33E�5 9.4E�6 3.2, +0.3, �1.2 From Williams

et al. (2008)

Edifice 30,080 10 7–10 1.19E�2 4.9E�3,

5.8E�3

1.7E�4 7E�5, 8E�5 3.8, +0.1, �0.1 This study

Hadriaca

Patera

758 This study

Caldera

floora

1908 559 0.7–1.3 2.52E�3 1.01E�4 3.60E�5 1.52E�6 3.5, +0.1, �0.3 From Williams

et al. (2007)

Edifice 68,032 9 7–40 5.14E�3 2E�3 7.34E�5 3E�5, 4E�5 3.7, +0.1, �0.2 This study

Amphitrites

Patera

288 This study

Caldera

floor

5,990 10 1.3–5 3.83E�3 2E�3 5.47E�5 2E�5 3.6, +0.1, �0.2 This study

Edifice 153,448 12 9–30 4.41E�3 2E�3 6.30E�5 2E�5 3.6, +0.1, �0.1 This study

Peneus

Patera

233 This study

Caldera

floor

4,537 9 1.1–2 2.87E�3 1.2E�3 4.10E�5 1.7E�5 3.5, +0.1, �0.3 This study

Edifice 91,827 5 12–60 8.09E�3 4E�3, 5E�3 1.16E�4 6E�5, 7E�5 3.8, +0.1, �0.2 This study

Malea

Patera

122 This study

Caldera

floor

13,360 9 3–10 1.21E�2 5E�3, 6E�3 1.73E�4 7E�5, 8E�5 3.8, +0.1, �0.1 This study

Edifice 41,045 8 9–30 1.11E�2 5E�3, 6E�3 1.58E�4 7E�5, 8E�5 3.8, +0.1, �0.1 This study

Pityusa

Patera

297 This study

Caldera

floor

27,000 23 1.5–11 3.31E�3 9E�4, 1E�3 4.73E�5 1E�5 3.6, +0.1, �0.2 This study

Edifice 161,283 20 14–40 1.05E�2 3E�3, 4E�3 1.50E�4 5E�5 3.8, +0.1, �0.1 This study

Dissected

Plains

57,853 38 3–20 8.28E�3 2E�3 1.18E�4 3E�5 3.8, +0.1, �0.1 This study

Refer to Figs. 2a–c for count areas. Crater densities normalized per square kilometer.

Notes: N(1) ¼ cumulative number of craters with diameters X1 km/km2. N(10) ¼ cumulative number of craters with diameters X10 km/km2. N error bars ¼7N(x)/sqrt(n),

sqrt ¼ square root, n ¼ number of craters counted (Crater Analysis Techniques Working Group, 1979). Cratering model age for formation events calculated from martian

chronology model of Neukum (1983), Hartmann and Neukum (2001), and Ivanov (2001): N(1) ¼ 2.68�10�14[e6.93t�1]+4.13�10�4t.

a Counts made on HRSC images of Tyrrhena and Hadriaca Paterae from earlier studies.

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916904

has a cratering model age of �3.5 Ga, similar to Amphitrites’caldera.

The caldera of Peneus Patera is about 100 km across, as markedby a series of four concentric, discontinuous graben. Sinuousdepressions and irregular-shaped crater-forms within graben onthe northwest and southeast parts of the caldera suggest fluidflow and collapse. Some of the depressions are superficiallysimilar to those seen in the south polar region of Mars, leading tothe formation of so-called ‘‘Swiss cheese’’ terrain (Malin andEdgett, 2001), which in the south polar region is related toseasonal accumulation and sublimation of CO2 ice. Because MGSThermal Emission Spectrometer (TES: Christensen et al., 2001)and THEMIS data analysis does not suggest the presence of CO2 iceat this latitude, this scalloped and pitted terrain is thought toinvolve surface degradation resulting from sublimation of near-surface water ice (Lefort et al., 2005; Russell et al., 2005;Morgenstern et al., 2007). Our preliminary mapping of the regionusing THEMIS data shows that this scalloped and pitted terraincovers about 2% of the area and is found primarily on the flanks ofAmphitrites and the outer margins of the Peneus caldera. Thisoccurrence could reflect the presence of ash or other fine-grainedmaterial capable of containing water or ice at the time offormation of the terrain. Most of the caldera floor is generallyflat and smooth except for two wrinkle ridges and a zone ofirregular depressions that might be collapse features. HRSC colordata (Fig. 6) show a low albedo zone in the northern half of the

caldera floor. We interpret this zone to be residual concentrationsof underlying basaltic material exposed by aeolian winnowing(perhaps buried lava flows or pyroclastic material) covered bydust. (This region is also noted for many dust devil tracks, andmultiple active dust devils observed in HRSC and SRC images. seeOberst et al., 2008). Alternatively, aeolian processes could causebasaltic sand to accumulate in topographic lows. However,the HRSC preliminary DTM for orbit 2133 (spatial resolution200 m/pixel) shows that the west-central interior floor has asimilar elevation as the dark zone; yet there are no dark depositsthere, possibly supporting the volcanic interpretation for the darkzone.

The surfaces of both paterae have a ‘‘softened’’ appearance,typical of terrains in the higher latitudes of Mars and suggestive ofmodification by periglacial processes such as downslope move-ment of fragmental material aided by incorporated water andice (Squyres and Carr, 1986). The presence of volatiles in surfacematerials at the time of modification is also suggested byabundant ‘‘ejecta flow’’ or pedestal craters (Fig. 11), consideredto represent impacts into volatile-rich targets (Carr et al., 1977b;Barlow et al., 2000). The occurrence of mass-wasting flow lobes onthe inner walls of Barnard and the other large impact craters isconsistent with the presence of fragmental material lubricated bywater or ice (e.g., Marchant and Head, 2007). HiRISE and CTXimages of this region show polygonal fractures and mounds in andaround the scalloped and pitted terrain (Figs. 12 and 13), thought

ARTICLE IN PRESS

Fig. 10. (a) MOLA context for Peneus Patera (16 pixel/deg). (b) THEMIS VIS image V07551005 (17 m/pixel), showing a degraded impact crater and the stepped rim of Peneus.

This topography may suggest an embedded caldera, resulting from multiple episodes of explosion and collapse. (c) MOC-NA image E1202776 (2.9 m/pixel) of the SW part of

the patera’s rim showing degraded crater rim and interior, containing smaller-scale scallops and dust devil tracks. These images clearly show that small positive relief

circular structures that might appear as domes in lower-resolution data are actually buried and exhumed impact craters. (d) THEMIS VIS image V16961008 (17 m/pixel)

showing irregular scalloped and pitted terrain that covers parts of Amphitrites and Peneus Paterae. This terrain, although superficially similar to that noted near the south

pole in MOC images (Malin and Edgett, 2001), is suggested to result from water ice sublimation (Lefort et al., 2005; Russell et al., 2005).

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 905

to be indicative periglacial processes involving water icesublimation (Lefort et al., 2005; Levy and Head, 2005;Morgenstern et al., 2007; Zanetti et al., 2008). Furthermore, thewidespread presence of pedestal craters (also called ‘‘perched’’craters: McCauley, 1973; Kadish et al., 2008a, b) suggest thatdifferences exist in the blockiness of the ejecta compared tothe surrounding terrain, leading to differences in the thermalregime and differential devolatilization of the materials, in whichthe ejecta is more resistant to erosion (Meresse et al., 2006). MOCimages show that the ejecta for these craters is surfaced withboulders, supporting the ideas regarding differential erosion.

In summary, the evidence supporting the volcanic hypothesisfor Amphitrites Patera include: (a) the shield-like topography anddissected flanks, similar to Tyrrhena and Hadriaca; (b) thecratering model age similarity (�3.6 Ga) to a time of inferredvolcanic activity at Tyrrhena and Hadriaca; (c) a strong positivegravity anomaly, perhaps indicative of a dense subsurface magmabody, similar to Tyrrhena and Hadriaca; (d) dark material in colordata, perhaps evidence of residual concentrations of maficvolcanics (alternatively, aeolian-transported basaltic sand anddust), similar to that observed at Tyrrhena and Hadriaca, and(e) rare evidence of radial channel segments (perhaps collapsedlava tubes?).

The evidence supporting a volcanic hypothesis for PeneusPatera is somewhat more ambiguous: (a) a caldera-like centraldepression, but no positive-relief edifice; (b) a cratering model age(�3.8 Ga) correlative with the formation ages of Tyrrhena,Hadriaca and the plains of Malea Planum; (c) a positive (although

weak) gravity anomaly; and (d) dark material in color data in thenorth-central Peneus floor, perhaps evidence of residual concen-trations of mafic volcanics (alternatively, aeolian-transportedbasaltic sand and dust), similar to that observed at Tyrrhena andHadriaca. No clear evidence of lava flows, flow fronts, smallshields or cones, or fine-scale layering has thus far been found inhigh-resolution SRC, MOC, CTX, or HiRISE images, even thoughcoverage is scarce; rather, these images show pervasive scallopedand pitted terrain (i.e., large-scale layering) and buried andexhumed impact craters, which may indicate emplacement ofpyroclastic, aeolian or circum-polar mantles (e.g., Head and Pratt,2001b). These results suggest to us that Peneus Patera may be bestinterpreted as an ‘‘embedded caldera’’, in which multiple episodesof explosion and collapse produced the multiple graben in itsinterior. Nevertheless, it is clear that, at the latitude of Amphitritesand Peneus (�581S), repeated surface modification by aeolianand periglacial processes and/or emplacement of circum-polarmantles may have destroyed or hidden much of the evidence oftheir ancient volcanism.

4.3. Malea and Pityusa Paterae

MOLA data (Figs. 1 and 5) show that Malea and Pityusa Pateraeconsist of two irregular depressions in the surface of the ridgedplains of Malea Planum. Relatively weak gravity anomalies(Fig. 1b) occur in association with these features (Lemoine et al.,2001; G.A. Neumann, pers. comm., 2008). The floor of Malea

ARTICLE IN PRESS

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916906

Patera lies at an average of �10 m in elevation, below a south andsouthwest topographic high with a maximum elevation�1600–1800 m. The north end of the Malea Patera floor isbordered by a �1000 m topographic high, sloping up �0.31 beforedescending into Peneus Patera. The ‘‘rim’’ of Malea Patera isill-defined in MOLA and THEMIS data; wrinkle ridges in partdefine a possible caldera (Fig. 2b). Similar to Peneus andAmphitrites, the floor of Malea Patera contains wrinkle ridges,

Fig. 11. Portion of THEMIS mosaic of a relatively fresh ‘‘ejecta-flow’’ crater, 5 km in

diameter southwest of the Peneus Patera caldera rim. Such crater ejecta patterns

represent impacts into volatile-rich target materials (Carr et al., 1977b; Barlow et

al., 2000).

Fig. 12. (a) Part of HRSC nadir image showing a relatively young, buried and exhumed

irregular pits a few hundred meters across within the ejecta zone of the crater

PSP_004867_1220_RED, showing greater detail on the scalloped and pitted terrain. (b)

the ejecta zone of the crater.

pedestal craters, dark zones in HRSC color data (Fig. 6), and anunusual heart-shaped layered mesa �35 km wide (W–E) by�37 km long (N–S). MOC and THEMIS images (Fig. 14) show thatthis mesa is layered on its margin, and has an upper surface thatcontains dust devil tracks and many fine pits at the tens tohundreds of meters scale, which is expected at higher southernlatitudes (Head et al., 2003). However, Malea lacks the scallopedand pitted terrain found on Peneus and Amphitrites, suggestingthat Malea’s floor materials are in some way different from thosein Peneus and Amphitrites. Subsurface data from several MARSIStracks show an apparent layer under the center of Malea Patera(Fig. 15) at least 100 km long and at a depth of about 300–500 m;however, it is currently unclear what this layer is, as its location isnot the same as the heart-shaped layered mesa. Possibilitiesinclude lava flow, pyroclastic deposits, or ice, each of which hasimplications for the types of geologic activity that occurred withinMalea Patera. Because this is a relatively shallow feature andMARSIS is better at imaging deeper subsurface features (41 km:Picardi et al., 2004, 2005; Plaut et al., 2007), it may require higherspatial resolution SHARAD data to better understand thisinterface. Crater counts on the THEMIS 100 m daytime-IR mosaicfound a cratering model age of �3.8 Ga for Malea Patera,suggesting a very old structure.

Comparison of the HRSC nadir and color images yieldedidentification of several dark features within the inner rim ofMalea Patera (Fig. 16). These appear to be point sources of darkmaterial that, although being modified by local winds, are notaccumulations in topographic lows, like many of the other darkdeposits in the CHVP. We suggest these may be point sources ofvolcanic materials, perhaps vents being exhumed. We comparedthe HRSC color data to the OMEGA data for this orbit using severaldifferent techniques (principal component analysis: Pinet et al.,2007; nonlinear spectral unmixing: Poulet et al., 2007; linearspectral unmixing: Combe et al., 2008a, b) (Fig. 17a); OMEGA dataunambigously indicate the composition of this dark material isconsistent with mafic volcanics, rich in clinopyroxene,orthopyroxene, and localized deposits of olivine. Additionally,the spectroscopic information inferred from OMEGA datasupports the idea of a basaltic material with closecharacteristics to those observed on the Syrtis Major shield(Baratoux et al., 2007a, b; Pinet et al., 2007; Poulet et al., 2007),that can be exposed or seen intermittently through the dust

impact crater on the northern part of Amphitrites caldera floor, and the unusual

(HRSC image H2525_0000.nd4; 12.5 m/pixel). Inset is part of HiRISE image

Portion of THEMIS mosaic showing the low-temperature signature correlated with

ARTICLE IN PRESS

Fig. 13. (Color online) (a) Part of THEMIS day-IR mosaic merged with HRSC color data of Peneus Patera (100 m/pixel). White arrows mark the region of dark material in

HRSC color data, which contains extensive dust devil tracks. (b) CTX image P03_002296_1217_XI_58S306W of SE Peneus floor (6 m/pixel), showing scalloped inner rim

(center) and scalloped and pitted terrain (lower right). (c) HiRISE image PSP_002296_1215 (50 cm/pixel) showing detail on the scalloped and pitted terrain that is pervasive

across much of Amphitrites and Peneus Paterae. (d1 and d2) Full-resolution enlargements of sections of the HiRISE image, showing polygonal fracturing and mounds

indicative of periglacial processes. There are hints of layering in the shadowed cliff face in d2).

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 907

coverage exhibiting clear ferric oxide phase characteristics typicalof bright terrains across Mars.

HRSC images are also used as 5-band multispectral data(440 nm—blue, 530 nm—green, 650 nm—nadir, 750 nm—red,970 nm—infrared), after calibration into radiance factor I/F(I: observed surface radiance, F: observed solar flux or irradiance)and corrected for incidence angle variations (division by thecosine of the incidence angle). Those spectra are mostly sensitiveto two surface materials (dark mafics and bright red iron oxides)represented by two image spectral endmembers, and shade.We have found that we can use the spectral unmixing method ofCombe et al. (2008b) on multispectral HRSC data to map theproportions of dark material (Combe and McCord, 2008) in theCHVP (Fig. 17b), to better identify the extents and locations of

residual concentrations of possible volcanic materials (e.g., SWMalea Patera floor). In future studies we will be able to targetthese locations with OMEGA data to assess if any compositionalvariations are present in these dark materials (e.g., Tirsch et al.,2008).

In MOLA data, Pityusa Patera occurs as an irregular centraldepression surrounded by a pronounced rim on the south, west,and north sides (Fig. 2c), in which the central depression contains:(a) a southeastern rugged zone of layered hills, (b) a northern darksand dune field, and (c) a west-central, bright, smooth, irregular,cratered plateau. The lowest levels of the Pityusa centraldepression are around 0 m elevation, and the bright plateau risesto �300 m. The central depression is �170 km wide (W–E) by�200 km long (N–S), and the rim is crosscut by radial to sub-

ARTICLE IN PRESS

Fig. 14. The interior of Malea Patera: (a) MOLA shaded relief index map (16 pixel/deg); (b) MOC wide-angle index image M0302176 (250 m/pixel); (c) The heart-shaped

mesa has the same texture at that resolution, which in HiRISE images is suggestive of periglacial surfaces: MOC narrow-angle image M0302175 (2.76 m/pixel); (d) Clear

layering is exposed at the margins of the heart-shaped mesa: THEMIS VIS V07089006 (17 m/pixel); (e) Widespread aeolian degradation is occuring—note the dust devil

tracks at upper center: THEMIS VIS V16593019 (17 m/pixel).

Fig. 15. (Color online) MARSIS radargram from Mars Express orbit 2648, showing a

potential linear reflector at a depth of �300–500 m in central Malea Patera. x-Axis

is in kilometers, y-axis in microseconds. As the MOLA context shows, the location

of this feature is not the same as the heart-shaped layered mesa observed in

imaging data, and it is currently unclear what this inferred subsurface layer

represents: pyroclastic deposits, lava flows, ice, or some other material.

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916908

parallel wrinkle ridges. The surrounding plains contain many largecraters and near-concentric wrinkle ridges. Crater counts on theTHEMIS 100 m day-IR mosaic (Fig. 2c) give a cratering model ageof �3.8 Ga for the feature, and �3.6 for the central depression(dominated by the bright irregular plateau). Similar to MaleaPatera, Pityusa lacks the scalloped and pitted terrain found onPeneus and Amphitrites, but does contain many pedestal craters.MOC and THEMIS data reveal details about the materials in thePityusa central depression (Figs. 18 and 19). The �100 km wideby �120 km long bright, irregular plateau is quite diverse, witha striated and heavily cratered surface on its north side, and a

heavily cratered surface that lacks striations on the south side.The plateau has pits both lacking and containing dark sand, inwhich wind streaks from the filled pits point to and feed the darkdune field to the north. The hills in the southeastern part of thedepression are layered on the scale of hundreds of meters, andheavily eroded. Between the hills the surface is cut by sub-parallelsets of ridges and grooves, indicative of extensive degradation ofthese layered materials.

We used OMEGA data (Fig. 20) to assess the composition of thedark sand dune field in Pityusa Patera, in the form of modalmineralogy derived using the nonlinear unmixing model ofShkuratov et al. (1999), Poulet et al. (2002), and Poulet andErard (2004). The dark sands are composed mostly of plagioclaseand high-Ca pyroxene, with smaller amounts of low-Ca pyroxene,olivine, and dust (Table 2). We interpret this composition asbasaltic sand, derived from mafic volcanics.

The nature of the bright, irregular plateau is particularlypuzzling (Figs. 18 and 19). Based on MOLA data it is severalhundred meters thick. It has a well-defined boundary around itsperimeter, and appears well-consolidated. It is unclear whetherthe layered hills to the southeast are remnants of part of theplateau that has been eroded. It is also unclear whether the darksands are being exposed from underneath the plateau, or whetherthey have simply been deposited north of the plateau by windsand the dark sand originates elsewhere on Mars. Although adefinitive explanation for the bright plateau cannot be made withavailable data, we suggest that it, and perhaps the heart-shapedmesa in Malea Patera consists of ignimbrite deposits. Ignimbrites(Marshall, 1935; Gilbert, 1938) are pyroclastic flow deposits(also called ash-flow tuffs), and may be welded or unweldedand composed of mostly pumice and ash (Schminke, 2004).Welded ignimbrites can be somewhat more resistant to erosion,

ARTICLE IN PRESS

Fig. 16. (Color online) (a) HRSC false color image of the north ‘‘rim’’ and heart-shaped layered mesa in Malea Patera (orbit 2133, 100 m/pixel, coordinates 631S, 52.71E). Black

arrows point to dark features apparently being eroded out of the rim, and not occurring in topographic lows. We suggest that these may be buried volcanic sources that are

now being exhumed. (b) Corresponding HRSC nadir image (25 m/pixel).

Fig. 17. (Color online) (a) Spectral analyses of Mars Express data from orbit 2133 (Center coordinates 591S, 531E). Left: color composite of HRSC images corrected for

incidence angle variations (R: 970 nm, G: 650 nm, B: 530 nm). Center and bottom right: OMEGA mapping of the mafic minerals from image ORB 2133_4 superimposed on

HRSC color composite, following the spectral unmixing method by Combe et al. (2008) (R: Olivine, G: Orthopyroxene (OPX), B: Clinopyroxene (CPX)). OPX is detected

almost everywhere in bright and moderately dark areas, and is correlated with iron oxides (not shown here). Moderately dark areas are enriched in CPX. The darkest spots

are likely enriched in both CPX and olivine, but OPX is probably also present. Upper right: OMEGA spectrum modeled by linear spectral unmixing. Each color curve is a

spectrum of a pure mineral weighted by its mixing coefficient. The sum of all the modeled contributions plus the residual is superimposed on the actual OMEGA spectrum

(top). (b) Map of proportion of dark endmember in HRSC color data from orbit 2133 covering Peneus and Malea Paterae (refer to (a)). Spectral unmixing analysis from the

method described in Combe et al. (2008).

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 909

resulting from the welding of the pyroclastic particles that occursby the heat retained in the flow after motion ceases. A recentreview of studies of the Medusae Fossae Formation suggests thatit most likely contains the remnants of martian ignimbritedeposits (Scott and Tanaka, 1982; Zimbelman et al., 1998;see discussion in Mandt et al., 2007, 2008), and we note thatthere are some morphological similarities between the Pityusa

irregular plateau and at least one terrestrial ignimbrite deposit(cf., Mandt et al., 2008, Fig. 10).

In summary, our overview of the available data for Malea andPityusa Paterae strengthens the case for the role of volcanism inthe formation of these features, but does not categoricallydemonstrate that these are volcanoes. In fact, their morphologiesare quite different from Tyrrhena, Hadriaca, and Amphitrites

ARTICLE IN PRESS

Fig. 18. Features in the interior of the Pityusa Patera central depression: (a) MOLA shaded relief index map (16 pixel/deg); (b) portion of the THEMIS VIS image V16993020

(17 m/pixel), showing a striated upper surface to the large, bright, irregular, plateau. Holes in the plateau are filled with dark sand, and wind streaks mark a generally

northly wind direction, in which dark sands are deposits forming a dune field; (c) portion of the THEMIS VIS image V18172008 (17 m/pixel); (d) THEMIS VIS image

V06303008 (17 m/pixel), showing eroded, layered hills; (e and f) fragments of the THEMIS VIS image V06303008.

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916910

Paterae. Both Malea and Pityusa formed �3.8 Ga, at the same timeas Tyrrhena, Hadriaca, and Peneus Paterae, but before AmphitritesPatera. Like Peneus, but unlike Tyrrhena, Hadriaca, or Amphitrites,these features have no well-defined volcanic edifice; rather theyhave central depressions formed in the ridged plains of MaleaPlanum. Like Peneus, but unlike Tyrrhena, Hadriaca, or Amphi-trites, these features have no strong positive gravity anomalies,perhaps indicative of a different style of volcanism (moreexplosive), or simply of collapse after early eruption(s) emptiedthe magma chambers, forming the central depressions. Darkmaterials detected in HRSC color data are identified in OMEGAdata unambiguously as basaltic in composition; however, theirprovenance as localized mafic volcanics, particularly at Pityusa,cannot be definitively determined. No lava flows, flow fronts, orunequivocal vents have been identified at Malea and Pityusa;however, we propose that the heart-shaped mesa at Malea andthe bright, irregular plateau at Pityusa could be ignimbritedeposits. If so, these would indicate a different style of volcanismat Malea and Pityusa compared to the other volcanoes in theProvince.

5. Discussion

5.1. Ages

Crater size–frequency distribution (CFD) results (Table 1,Fig. 21) show that the six highland paterae are in the same age-range, but that Tyrrhena, Peneus, Malea, and Pityusa are slightly olderat 3.8 Ga, while Hadriaca and Amphitrites are younger at 3.7 and3.6 Ga, respectively. These differences may not be that significant,considering the 100 Ma error bar on the Hartmann and Neukum

technique. However, the caldera ages all range from 3.8 to 3.2 Ga,and are substantially older than those observed on the floors ofthe Tharsis volcanoes, most of which are younger than 0.5 Ga(Neukum et al., 2004b). Based on the observations from the newdata presented here and from previous studies, we suggest thatthe main constructs of the highland patera around the Hellas

basin formed over the period of 4–3.6 Ga, making them the oldestpreserved volcanoes on Mars and (in agreement with previousstudies) reflecting a change in eruptive style from the putativeearlier fissure-fed flood eruptions to local edifice-buildingeruptions. The ages of Peneus, Malea, and Pityusa Paterae arealso consistent with the ages of valley networks (�3.7–3.5 Ga:Fassett and Head, 2008) and the time when extensive amounts ofwater existed (at least temporarily) on the surface (e.g., Carr,1996). If Malea and Pityusa Paterae are the Martian equivalentof ancient giant calderas or ‘‘supervolcanoes’’ (e.g., Self, 2006) thatproduced ignimbrite deposits, then these eruptions likelyinvolved substantial amounts of volatiles (such as water) in themagma. If early Mars was characterized by generally wetterconditions (i.e., abundant groundwater in the shallow crust), thenthese conditions might have enabled these types of largeexplosive eruptions.

5.2. Structure and topography

MOLA profiles and the MGS (Lemoine et al., 2001) and MROgravity data show dissimilarities among the six volcanoes of theCHVP. The three volcanoes with positive-relief edifices in MOLAdata (Tyrrhena, Hadriaca, and Amphtrites) also have strongpositive gravity anomalies, perhaps indicative of dense magmabodies below the surface. The other three volcanoes (Peneus,Malea, and Pityusa) have central depressions within the surround-ing plains, and lack both the shield-like edifices and the strongpositive gravity anomalies observed at the other three volcanoes.If these paterae are the remnants of giant Martian calderas thatproduced explosive eruptions and that, in the case of Malea andPityusa Paterae, produced ignimbrite deposits, then there was alarger diversity of volcanic eruption styles in the CHVP than haspreviously been considered.

5.3. Morphology and surface degradation

If this is a volcanic province, then where are all the volcanicvents and lava flow boundaries? The detailed MOC, THEMIS, and

ARTICLE IN PRESS

Fig. 19. The multilayered structure of the large, bright, irregular plateau within the central part of the Pityusa Patera, which we suggest could be some type of volcanic flow,

perhaps an ignimbrite deposit?: (a) MOLA shaded relief index map (16 pixel/deg); (b) MOC wide-angle index image E1300618 (250 m/pixel); (c) MOC-NA image E1300617

(4.34 m/pixel); (d) portion of the MOC image E1300617. The upper portion of the plateau shows degraded craters at a range of sizes, indicative of an old surface.

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 911

HRSC images show that the original volcanic surface of the CHVPis covered by a surface mantle material with a thickness of tens toseveral hundreds of meters. The low-density distribution of fresh,small (o500 m diameter) impact craters on much of the surface ofthe CHVP, and the generally smoothed appearance of the surfaceat the scale of hundreds of meters, are indicative of relativelyrecent formation of much of the surface material composingthe mantle cover (likely of Amazonian age). We suggest that thenature of the material is closely related to deposition of aeoliandust and sand in combination with enrichment by water ice fromcondensation of atmospheric water vapor, and from arealvariations in seasonal polar cap deposits due to periodic variationsof Mars’ orbital parameters. For example, throughout theHesperian, extensive south circum-polar deposits (the Dorsa

Argentea Formation, DAF) were emplaced. One interpretation isthat these represent south polar and related ice-rich deposits(Head and Pratt, 2001b) that could have been characterized bymantling deposits that extended well beyond the border of theDAF to lower latitudes.

As we have shown, the volcanic morphology of the pateraesouth of Hellas at regional scales (tens to hundreds of kilometers)has some similarities and some differences to Tyrrhena andHadriaca Paterae. However, based on MOC, CTX, and HiRISEimages, the surface morphology of the southern part of the CHVPat the scale of tens to one hundred meters looks mostly identical,due to widespread manifestations of the periglacial features(e.g., mounds, polygonal fractures, scalloped and pitted terrain).These features formed because of freezing and ice enrichment of

ARTICLE IN PRESS

Fig. 20. (Color online) OMEGA and HRSC color data for Pityusa Patera. (A) Map of the pyroxene spectral parameter over the eastern part of the dark deposits located in

Pityusa Patera. The parameter is calculated as described in Poulet et al. (2007). (B) OMEGA spectra extracted from two locations of the deposits. One (indicated by red

dashed arrow) comes from thicker dune deposits, one (indicated by blue dashed arrow) from dark sand deposits that are thinner over brighter substrate material. (C) The

spectral ratio (red over blue) indicates deeper mafic bands for the thick dune deposit. (D) HRSC false color mosaic of Pityusa Patera and plains to the north (50 m/pixel). The

black box shows the location of the OMEGA image in (A). The large region of dark material on the north side of the Pityusa central depression is composed of dark sand

dunes, clearly an aeolian deposit in a topographic low. The OMEGA-derived mineralogy is consistent with a basaltic sand. A key question is whether these basaltic sands

were transported from underneath the bright plateau that fills much of the central depression, or from another region of Mars. Many of the dark zones around Pityusa bear

closer examination. HRSC image processing by David Williams, ASU.

Table 2OMEGA-derived modal mineralogies on a dark unit found in the CHVP

Location of unit Mineral component Abundance (vol%)

Pityusa Patera dark dune field Plagioclase 4876

High-Ca pyroxene 2673

Low-Ca pyroxene 872

Olivine (100mm) 574

Dust 1377

Methodology of mineralogy derivation discussed in Poulet and Erard (2004) and

Poulet et al. (2007).

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916912

the surface mantle and former volcanic deposits. The dominanttype of the permafrost features in the Province, as observed inHiRISE images, is a polygonal relief formed by frost cracking.

The features form a polygonal system of the shallow troughs,which surround central mounds a few meters in size (see alsoMarchant and Head, 2007). These permafrost features are mostlyfound in the area of Amphitrities and Peneus Paterae (Fig. 13) andoccur in and around scalloped depressions (Plescia, 2003; Lefortet al., 2005). A combination of processes such as sublimation ofinterstitial ice and wind deflation is the main mechanism for theformation of the depressions (Costard et al., 2008; Zanetti et al.,2008). These processes are widespread within the area of theAmphitrities and Peneus Paterae. In contrast, the ice-rich depositsto the south are less disturbed by these processes. Nevertheless, itis now clear that the presence of the ice in the surface materialswithin the CHVP enabled the smoothing of the surface relief,because of sublimation, cryogenic weathering, and ground icecreep processes, all of which masks most of the evidence of theancient volcanism.

ARTICLE IN PRESS

Fig. 21. Age correlation plot of the volcanoes and plains in the CHVP. Cratering

model ages (with error bars) plotted relative to the martian geologic timescale of

Hartmann and Neukum (2001). TP, Tyrrhena Patera. HP, Hadriaca Patera. AP,

Amphitrites Patera. PP, Peneus Patera. MP, Malea Patera. PityP, Pityusa Patera. Dis

Pln, Dissected Plains. N, Noachian Period. H, Hesperian Period. A, Amazonian

Period. Lines mark the period boundaries (a gray box marks the uncertainty in the

Hesperian-Amazonian boundary) and the formation of the Hellas basin. From this

plot it is clear that the volcanoes and plains of the CHVP formed in a relatively

short interval (3.9–3.6 Ga) after the formation of the Hellas basin at �4.0 Ga

(Werner and Neukum, 2003; Werner, 2005).

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 913

5.4. Compositional information

Previous studies of the Martian surface based on thermal-infrared and visible/near-infrared spectroscopy (e.g., Bandfieldet al., 2000; Bibring et al., 2005; McCord et al., 2007) indicatedthat materials that appear dark blue to black in ‘‘true color’’images have compositions consistent with basaltic volcanicdeposits. HRSC color data at 50–100 m/pixel cover wide areas ofthe surface, such that through the application of spectraldeconvolution method (Combe et al., 2008) we can map theseimages for percentage of dark material in a given region (Fig. 17).Clearly, there are many areas in the CHVP that contain this darkmaterial; the key question then becomes whether this darkmaterial is concentrated from underlying basaltic volcanics andexposed by aeolian winnowing or produced by glacial or fluvialerosion and concentrated by local winds (e.g., Baratoux et al.,2007a, b), or is it basaltic sand and dust transported fromelsewhere on Mars by regional winds? As there is no directindication of source regions outside the CHVP, and there are somerare outcrops clearly exposed on steep slopes such as crater orcaldera rims (Fig. 16), we favor the interpretation that the darkmaterial is concentrated from underlying basaltic volcanics.However, additional high-resolution imaging (HiRISE) and spec-tral analysis (CRISM) is required to better address this question.

6. Future work

Now that we have completed our initial reconnaissance of theCHVP, the next step is to focus future studies on specific aspects orregions to better assess the volcanic nature of the exposed surface.The following types of analyses are needed:

(1)

Use available MOC, SRC, HRSC, and THEMIS images to study indetail all locations of dark material exposed in the CHVP, and,if possible, acquire new HiRISE and CTX images of thesefeatures to assess their geologic nature.

(2)

Conduct a complete spectral deconvolution study of all goodquality HRSC color data covering the CHVP, to assess theproportions of partially buried dark material in the surfaceand their relationship to exposed dark material.

(3)

Analyze OMEGA and CRISM data (as it become available) ofexposed dark material to assess compositional variability.

(4)

Expand the study of MARSIS data and include SHARAD data toinvestigate the nature of layering at key locations throughoutthe CHVP.

Additionally, there is a region of western Promethei Terra on theeast side of the Hellas basin (36–501S, 90–1061E) that containssmooth plains interpreted to be volcanic materials (Raitala et al.,2007) that needs to be more thoroughly studied. As we havedemonstrated in this paper, only with integrated data analyses ofall available spacecraft data for this region will it be possible tounravel the nature of the ancient volcanism and the complexgeologic history of the Circum–Hellas Volcanic Province.

Acknowledgments

We thank Laszlo Keszthelyi and Renee Weber of the US G.S.,and two anonymous reviewers, for offering helpful comments andsuggestions. We were gratified when one reviewer commented,‘‘this is one of the best HRSC papers I’ve ever seen!’’. We thankChris Edwards and the THEMIS team for producing several of theTHEMIS mosaics and for useful discussions on interpretingTHEMIS data. The authors acknowledge the use of Mars OrbiterCamera images processed by Malin Space Science Systems thatare available at http://www.msss.com/moc_gallery/. We thank themembers of the following instrument teams for acquiring newdata that are changing our understanding of Mars: MOLA, TES,GRS, HiRISE, CTX, and MARSIS. We particularly thank Devin Wallerfor assistance in interpreting the MARSIS data. We thank the HRSCExperiment Teams at DLR Berlin and the Freie Universitat Berlin aswell as the Mars Express Project Teams at ESTEC and ESOC for theirsuccessful planning and acquisition of data, as well as for makingthe processed data available to the HRSC Team. We acknowledgethe effort of the HRSC Co-Investigator Team members and theirassociates who have contributed to this investigation in thepreparatory phase and in scientific discussions within the Team.This research was supported by NASA through the Jet PropulsionLaboratory for US participation in the ESA Mars Express mission,and through the NASA Planetary Geology and GeophysicsProgram.

Appendix A

A.1. Crater count methods

Crater size–frequency distributions (CFDs) for the six volca-noes of the CHVP were obtained to derive a general sequence ofevents and their relation to general Martian history. We producedmosaics covering Tyrrhena, Hadriaca, Amphitrites, Peneus,Malea, and Pityusa Paterae using the THEMIS daytime-IR data at100 m/pixel as a standardized base. Because there is controversyregarding CFDs for dating surfaces on Mars especially with regardto estimating ‘‘absolute’’ dates and the use of statistics based onsmall craters (McEwen et al., 2005), our approach in this studywas to: (1) use craters 4800 m diameter to avoid the additionof secondary craters which would yield erroneously old ages, and(2) use the results primarily for age comparisons within individualvolcanoes, and for comparisons among the patera and othervolcanoes on Mars (i.e., even though the ‘‘absolute’’ dates might

ARTICLE IN PRESS

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916914

not be accurate, the relative ages can be used for suchcomparisons). Recent work based on impact craters discoveredin MOC data to have formed over the last 7 years on Mars(Malin et al., 2006) suggests that the cratering rate for Marsextrapolated from the Moon is essentially correct (Hartmann,2005). Nevertheless, restricting our CFDs to large craters andusing ages for relative comparisons should produce moreconservative results.

Crater measurements were tabulated, partitioned into bins ofincreasing crater diameter based on standard practices, and thesebinned data were used to produce cumulative CFD plots withcorresponding statistical errors (see e.g., Hartmann, 1966;Neukum and Wise, 1976; Crater Analysis Techniques WorkingGroup, 1979; Neukum and Hiller, 1981; Neukum, 1983) using theCRATERSTATS package developed by the Free University, Berlin(Michael and Neukum, 2008: http://hrscview.fu-berlin.de/crater-stats.html). In this technique, the cumulative CFDs were analyzedto determine crater densities at specific reference diameters, andcratering model ages following Hartmann and Neukum (2001).Cumulative crater densities for 1-, 2-, 5-, 10- and 16-km diametercraters were used by Hartmann and Neukum (2001) to assessrelative ages for martian geologic units and to place units intothe martian chronologic system using key units as referents(e.g., Tanaka et al., 1992). In general, the lower the crater density(the fewer the craters that formed or have been retained(preserved) on the surface), the younger the age. A cratering

model age (in Ga) is calculated from the cumulative crater densityat a reference crater diameter of 1 km using an establishedcratering chronology model for Mars (Neukum, 1983; Hartmannand Neukum, 2001; Ivanov, 2001), which is typically extrapolatedfrom the lunar model (in which crater frequencies are correlatedwith radiometric ages from Apollo samples) that has beenadjusted for the different orbital mechanics, crater scaling, andimpact flux for Mars relative to the Moon. The transfer of the lunarcratering chronology model to Mars may introduce a systematicerror of up to a factor of 2. This means that the typical uncertaintyin cratering model ages could vary by a factor of 2 for ageso3.5 Ga (in the constant flux range), whereas the uncertainty isabout 7100 Ma for ages 43.5 Ga (Hartmann and Neukum, 2001).Extensive testing and application of these techniques, however,have shown that the applied martian cratering chronology modelresults in ages for basin formation and volcanic surfaces that arein good agreement with Martian meteorite crystallization ageswith respect to ‘‘peak’’ activity periods (Neukum et al., 2007),which suggests that the chronology model is correct within anuncertainty of less than 20% (Werner, 2005). Specific error bars foreach cratering model age are included in Table 1. The resultsobtained for Hadriaca Patera and Tyrrhena Patera were comparedwith previous CFDs made from counts on higher-resolution HRSCimages (Williams et al. 2007, 2008). Ages for the ‘‘Edifice’’were obtained from homogeneous surfaces that are consideredto represent the main structure of the volcano. ‘‘Caldera’’ ageswere obtained for the smoothest surfaces that are consideredto represent eruptions of flood lavas or ash, or deposits ofnon-volcanic origin. All of these surfaces have been highlymodified by degradational processes. In most cases, CFDs show‘‘kinks’’ in the cumulative distributions that are considered torepresent resurfacing events, in which the smaller craters areobliterated, while the larger craters are still visible and can becounted. Resurfacing events can include mantling by windblown,volcanic, or other deposits, or surface degradation processes, suchas mass wasting or periglacial activity, that might differentiallyremove smaller craters. Because of the uncertainty in theprocess(es) leading to these ‘‘kinks,’’ and the focus of this studyon the major volcanic events, this aspect of the CFDs is notconsidered.

A.2. Spectroscopy methods: Fig. 17

OMEGA spectra are calibrated into radiance factor (I/F) andcorrected for atmospheric absorptions using an empirical method,which consists of using a synthetic atmospheric transmittancecorresponding to the ratio of two spectra acquired at the top andbottom of Olympus Mons and scaling it to the depth of the 2mmCO2 band. We restricted the analysis to the OMEGA ShortWavelength Infrared (SWIR) detector (0.96–2.57mm) to eliminateinter-detector calibration and registration discrepancies. Further-more, the SWIR detector covers a wavelength range that containsthe most diagnostic bands.

The method of spectral unmixing applied here (Combe et al.,2008) is a pixel-by-pixel spectral linear deconvolution that relieson a reference spectral library of 20 pure minerals. Mineralsrepresented include mafics, iron oxides, sulfates, clays, carbonates,and water ice. Two modeled spectra account for atmosphere andsurface photometric effects. Each OMEGA spectrum is modeled bya maximum of four mineral components. This limitation preventsinsignificant endmembers from being included in the model.The maps of the minerals vary with abundances, but they are notproportions, since mixing coefficients are also sensitive to grainsize variations.

A.3. Spectroscopy methods: Fig. 18 and Table 2

Modal mineralogy is derived using the nonlinear unmixingmodeling based on the Shkuratov radiative transfer model(Shkuratov et al., 1999; Poulet et al., 2002; Poulet and Erard,2004). Poulet et al. (2002) showed the degree of realism andefficiency of this model relative to other scattering models, and inparticular to the Hapke model and its derivatives. This model hasalso been tested to determine the type of mixture (sand, areal,or bedrock), the relative proportions, and the grain sizes ofcomponents of laboratory mineral samples (Poulet and Erard,2004), as well as applied to spectra of several planetary surfaces.For each spectrum the model had to satisfy two major constraints:the depth and the shape of the absorption(s) band(s) and theaverage value of the reflectance. One additional free parameterwas used to modestly adjust the continuum spectral sloperesulting from aerosols and/or photometric effects. The data werefit using a simplex minimization algorithm.

Although nonlinear mixing of reflectance spectra is a powerfulway to explore data sets, the method has limitations. First, the setof optical constants used as endmembers must be representativeof materials that are under study. Because we apply the nonlinearunmixing method to low albedo regions that are dominated bybasaltic materials, the minerals used in this paper are pyroxene(both low- and high-calcium), olivine (both Fe- and Mg-rich),feldspar (labradorite) and a dark oxide (magnetite). The opticalconstants were derived using the scheme described in Poulet andErard (2004) from reflectance spectra.

References

Bandfield, J.L., Hamilton, V.E., Christensen, P.R., 2000. A global view of Martiansurface compositions from MGS-TES. Science 287, 1626–1630.

Baratoux, D., Pinet, P.C., Gendrin, A., Kanner, L., Mustard, J., Daydou, Y., Vaucher, J.,Bibring, J.P., the OMEGA Science Team, 2007a. The mineralogical structure ofthe subsurface of Syrtis Major from OMEGA observations of lobate ejectablankets. J. Geophys. Res. 112, E08S05.

Baratoux, D., Mangold, N., Arnalds, O., Gregoire, M., Plavoet, B., Bardinzeff, J.-M.,Pinet, P., 2007b. Formation, transport and mineralogical evolution of basalticsands on Earth and Mars. In: Seventh International Conference on Mars,Abstract #3048. Lunar and Planetary Institute, Houston (CD-ROM).

Barlow, N.G., Boyce, J.M., Costard, F.M., Craddock, R.A., Garvin, J.B., Sakimoto, S.E.H.,Kuzmin, R.O., Roddy, D.J., Soderblom, L.A., 2000. Standardizing the nomen-

ARTICLE IN PRESS

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916 915

clature of Martian impact crater ejecta morphologies. J. Geophys. Res. 105(E11), 26,733–26,738.

Barsukov, V.L., et al., 1986. The geology and geomorphology of the Venus surface asrevealed by radar images obtained by Veneras 15 and 16. J. Geophys. Res. 91,378–398.

Bibring, J.-P., Soufflot, A., Berthe, M., Langevin, Y., Gondet, B., Drossart, P., Bouye, M.,Combes, M., Puget, P., Semery, A., Bellucci, G., Formisano, V., Moroz, V., Kottsov,V., the OMEGA Co-Investigator Team, 2004. OMEGA: Observatoire pour laMineralogie, l’Eau, les Glaces et l’Activite. ESA Special Publication, vol. 1240,pp. 37–50.

Bibring, J.-P., Langevin, Y., Gendrin, A., Gondet, B., Poulet, F., Berthe, M., Soufflot, A.,Arvidson, R., Mangold, N., Mustard, J., Drossart, P., the OMEGA Team, 2005.Mars surface diversity as revealed by the OMEGA/Mars Express observations.Science 307, 1576–1581.

Carr, M.H., 1973. Volcanism on Mars. J. Geophys. Res. 78, 4049–4062.Carr, M.H., 1976. The volcanoes of Mars. Sci. Am. 234, 33–43.Carr, M.H., 1981. The Surface of Mars. Yale University Press, New Haven, CT,

232pp.Carr, M.H., 1996. Water on Mars. Oxford University Press, New York, NY, 248pp.Carr, M.H., Masursky, H., Saunders, R.S., 1973. Generalized geologic map of Mars.

J. Geophys. Res. 78, 4031–4036.Carr, M.H., Blasius, K.R., Greeley, R., Guest, J.E., Murray, J.E., 1977a. Observations on

some Martian volcanic features as viewed from the Viking orbiters. J. Geophys.Res. 82, 3985–4015.

Carr, M.H., Crumpler, L.S., Cutts, J.A., Greeley, R., Guest, J.E., Masursky, H., 1977b.Martian impact craters and emplacement of ejecta by surface flow. J. Geophys.Res. 82, 4055–4065.

Christensen, P.R., et al., 2001. Mars Global Surveyor Thermal Emission Spectrometerexperiment: investigation description and surface science results. J. Geophys.Res. 106, 23,823–23,871.

Christensen, P.R., et al., 2004. The Thermal Emission Imaging System (THEMIS) forthe Mars 2001 Odyssey mission. Space Sci. Rev. 110, 85–130.

Combe, J.-Ph., McCord, T.B., 2008. Mars Express/HRSC spectral data of MER landingsites analyzed by a Multiple-Endmember Linear Spectral Linear UnmixingModel (MELSUM). In: Lunar and Planetary Science XXXIX, Abstract #2381.Lunar and Planetary Institute, Houston (CD-ROM).

Combe, J.-Ph., Le Mouelic, S., Sotin, C., Gendrin, A., Mustard, J.F., Le Deit, L., Launeau,P., Bibring, J.-P., Gondet, B., Langevin, Y., Pinet, P., the OMEGA Science Team,2008. Analysis of OMEGA/Mars Express data hyperspectral data using aMultiple-Endmember Linear Spectral Unmixing Model (MELSUM): Methodol-ogy and first results, Planet. Space Sci. 56, 951–975.

Costard, F., Forget, F., Madeleine, J.B., Soare, R.J., Kargel, J.S., 2008. The origin andformation of scalloped terrain in Utopia Planitia: insight from a GeneralCirculation Model. In: Lunar and Planetary Science XXXIX, Abstract #1274.Lunar and Planetary Institute, Houston (CD-ROM).

Crater Analysis Techniques Working Group, 1979. Standard techniquesfor presentation and analysis of crater size–frequency data. Icarus 37,467–474.

Crown, D.A., Greeley, R., 1993. Volcanic geology of Hadriaca Patera and the easternHellas region of Mars. J. Geophys. Res. 98, 3431–3451.

Crown, D.A., Greeley, R., 2007. Geologic map of MTM quadrangles �30262 and�30267, Hadriaca Patera region of Mars. US Geol. Surv. Sci. Invest. Ser. Map2936.

Crown, D.A., Price, K.H., Greeley, R., 1992. Geologic evolution of the east rim of theHellas basin, Mars. Icarus 100, 1–25.

Crumpler, L.S., Head, J.W., Aubele, J.C., 1996. Calderas on Mars: characteristics,structure, and associated flank deformation. In: McGuire, W.J., Jones, A.P.,Neuberg, J. (Eds.), Volcano Instability on the Earth and Other Planets.Geological Society Special Publication, vol. 110, pp. 307–348.

Fassett, C.I., Head, J.W., 2008. The timing of martian valley network activity:Constraints from buffered crater counting. Icarus 195, 61–89.

Frey, H., Schultz, R.A., Reidy, A.-M., Wolfe, H., 1991. A large pre-Hellas impact basinin the southern hemisphere of Mars (abstract). Lunar Planet. Sci. Conf. XXII,421–422.

Gilbert, C.M., 1938. Welded tuff in eastern California. Geol. Soc. Am. Bull. 49,1829–1862.

Greeley, R., Crown, D.A., 1990. Volcanic geology of Tyrrhena Patera, Mars.J. Geophys. Res. 95, 7133–7149.

Greeley, R., Guest, J.E., 1987. Geologic map of the eastern equatorial region of Mars.US Geological Survey Miscellaneous Investigations Series Map I-1802B, scale1:15,000,000.

Greeley, R., Spudis, P.D., 1981. Volcanism on Mars. Rev. Geophys. Space Phys. 19,13–41.

Gregg, T.K.P., Farley, M.A., 2006. Mafic pyroclastic flows at Tyrrhena Patera, Mars:constraints from observations and models. J. Volcan. Geotherm. Res. 155,81–89.

Gregg, T.K.P., Crown, D.A., Greeley, R., 1998. Geologic map of part of Tyrrhena Pateraregion of Mars (MTM Quadrangle �20252). US Geol. Surv. Map I-2556.

Hartmann, W.K., 1966. Martian cratering. Icarus 5, 565–576.Hartmann, W.K., 2005. Martian cratering 8: isochron refinement and the

chronology of Mars. Icarus 174, 294–320.Hartmann, W.K., Neukum, G., 2001. Cratering chronology and the evolution of

Mars. Space Sci. Rev. 96, 165–194.Head, J.W., Pratt, S., 2001a. Malea Planum Hesperian volcanic province:

characterization using MOLA data. In: Lunar and Planetary Science XXXII,Abstract #1627. Lunar and Planetary Institute, Houston (CD-ROM).

Head, J.W., Pratt, S.F., 2001b. Extensive Hesperian-aged south polar ice sheet onMars; Evidence for massive melting and retreat, and lateral flow and pondingof meltwater. J. Geophys. Res. 106 (E6), 12,275–12,299.

Head, J.W., Crumpler, L.S., Aubele, J.C., Guest, J.E., Saunders, R.S., 1992.Venus volcanism: classification of volcanic features and structures, associa-tions, and global distribution from Magellan data. J. Geophys. Res. 97 (E8),13,153–13,197.

Head, J.W., Mustard, J.F., Kreslavsky, M.A., Milliken, R.E., Marchant, D.R., 2003.Recent ice ages on Mars. Nature 426, 797–802.

Head, J.W., Wilson, L., Dickson, J.L., Neukum, G., the HRSC Co-Investigator Team,2006. The Huygens-Hellas giant dike system on Mars: implications for LateNoachian–Early Hesperian volcanic resurfacing and climatic evolution.Geology 34, 285–288.

Ivanov, B., 2001. Mars/Moon cratering rate ratio estimates. Space Sci. Rev. 96,87–104.

Kadish, S.J., Head, J.W., Barlow, N.G., 2008a. Pedestal craters on Mars: distribution,characteristics, and implications for Amazonian climate change. In: Lunar andPlanetary Science XXXIX, Abstract #1766. Lunar and Planetary Institute,Houston (CD-ROM).

Kadish, S.J., Head, J.W., Barlow, N.G., Marchant, D.R., 2008b. Martian pedestalcraters: marginal sublimation pits implicate a climate-related forcingmechanism. Geophys. Res. Lett. 35, L16104, doi:10.1029/2008GL034990.

Kolb, E.J., Tanaka, K.L., 2008. Geologic map of the Planum Australe region of Mars.US Geological Survey Miscellaneous Investigations Series, in review.

Larson, S.K., 2007. The origins of four paterae of Malea Planum, Mars. UnpublishedM.S. Thesis, Brigham Young University, Provo, UT, 120pp.

Lefort, A., Russell, P., Thomas, N., 2005. Ice sublimation landforms in Peneus andAmphitrites Patera. In: Lunar and Planetary Science XXXVI, Abstract #1626.Lunar and Planetary Institute, Houston (CD-ROM).

Lemoine, F.G., Smith, D.E., Rowlands, D.D., Zuber, M.T., Neumann, G.A., Chinn, D.S.,Pavlis, D.E., 2001. An improved solution of the gravity field of Mars (GMM-2B)from Mars Global Surveyor. J. Geophys. Res. 106 (#E10), 23,359–23,376.

Leonard, G.J., Tanaka, K.L., 2001. Geologic map of the Hellas region of Mars. USGeol. Surv. Misc. Invest. Ser. Map I-2694.

Levy, J.S., Head, J.W., 2005. Evidence for remnants of ancient ice-rich deposits:Mangala Valles outflow channel, Mars. Terra Nova 17 (#6), 503–509.

Malin, M.C., Edgett, K.S., 2001. Mars Global Surveyor Mars orbiter camera:interplanetary cruise through primary mission. J. Geophys. Res. 106, 23,429–23,570.

Malin, M.C., Edgett, K.S., Posiolova, L.V., McColley, S.M., Noe Dobrea, E.Z., 2006.Present-day impact cratering rate and contemporary gully activity on Mars.Science 314, 1573–1577.

Malin, M.C., et al., 2007. Context camera investigation on board theMars Reconnaissance Orbiter. J. Geophys. Res. 112, E05S04, doi:10.1029/2006JE002808.

Mandt, K.E., de Silva, S.L., Zimbelman, J.R., Crown, D.A., 2007. A synoptic approachto evaluating the origin of the Medusae Fossae Formation, Mars. In: Lunar andPlanetary Science XXXVIII, Abstract #1823. Lunar and Planetary Institute,Houston (CD-ROM).

Mandt, K.E., de Silva, S.L., Zimbelman, J.R., Crown, D.A., 2008. The origin of theMedusae Fossae Formation, Mars: insights from a synoptic approach.J. Geophys. Res., in press.

Marchant, D.R., Head, J.W., 2007. Antarctic dry valleys: microclimate zonation,variable geomorphic processes, and implications for assessing climate changeon Mars. Icarus 192, 187–222.

Marshall, P., 1935. Acid rocks of the Taupo-Rotorua volcanic district. Trans. R. Soc.N. Z. 64, 323–366.

McCauley, J.F., 1973. Mariner 9 evidence for wind erosion in the equatorial and mid-latitude regions on Mars. J. Geophys. Res. 78, 4123–4137.

McCauley, J.F., Carr, M.H., Cutts, J.A., Hartmann, W.K., Masursky, H., Milton, D.J.,Sharp, R.P., Wilhelms, D.E., 1972. Preliminary Mariner 9 Report on the geologyof Mars. Icarus 17, 289–327.

McCord, T.B., Adams, J.B., Hawke, B.R., Head, J.W., Mouginis-Mark, P.J., Huguenin,R.L., Pieters, C.M., Singer, R.B., Soderblom, L.A., Zisk, S.H., 1982. Mars: definitionand characterization of global surface units. J. Geophys. Res. 87, 10,129–10,148.

McCord, T.B., et al., 2007. Mars Express High Resolution Stereo Camera spectro-photometric data: characteristics and science analysis. J. Geophys. Res. 112,E06004, doi:10.1029/2006JE002769.

McEwen, A.S., Preblich, B.S., Turtle, E.P., Artemieva, N.A., Golombek, M.P., Hurst, M.,Kirk, R.L., Burr, D.M., Christensen, P.R., 2005. The rayed crater Zunil andinterpretations of small impact craters on Mars. Icarus 176, 351–381.

McEwen, A.S., et al., 2007. Mars Reconnaissance Orbiter’s High Resolution ImagingScience Experiment (HiRISE). J. Geophys. Res. 112, E05S02, doi:10.1029/2005JE002605.

Meresse, S., Costard, F., Mangold, N., Baratoux, D., Boyce, J.M., 2006. Martianperched craters and large ejecta volume: evidence for episodes of deflation inthe northern lowlands. Meteorit. Planet. Sci. 41, 1423–1690.

Michael, G., Neukum, G., 2008. Surface dating: software tool for analyzing cratersize–frequency distributions including those showing partial resurfacingevents. In: Lunar and Planetary Science XXXIX, Abstract #1780. Lunar andPlanetary Institute, Houston (CD-ROM).

Morgenstern, A., Hauber, E., Reiss, D., van Gasselt, S., Grosse, G., Schirrmeister, L.,2007. Deposition and degradation of a volatile-rich layer in Utopia Planitia andimplications for climate history on Mars. J. Geophys. Res. 112, E06010.

Mouginis-Mark, P.J., Wilson, L., Zuber, M.T., 1992. Physical volcanology. In: Kieffer,H.H., et al. (Eds.), Mars. University of Arizona Press, Tucson, pp. 424–452.

ARTICLE IN PRESS

D.A. Williams et al. / Planetary and Space Science 57 (2009) 895–916916

Neukum, G., 1983. Meteoritenbombardement und Datierung planetarer Ober-flachen. Habilitation Dissertation for Faculty Membership, Ludwig-Maximi-lians University, Munchen, Munich, Germany, 186pp.

Neukum, G., Hiller, K., 1981. Martian ages. J. Geophys. Res. 86, 3097–3121.Neukum, G., Wise, D.U., 1976. Mars—a standard crater curve and possible new time

scale. Science 194, 1381–1387.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., the HRSCCo-Investigator Team, 2004a. Recent and episodic volcanic and glacialactivity on Mars revealed by the High Resolution Stereo Camera. Nature 432,971–979.

Neukum, G., Jaumann, R., the HRSC Co-Investigator Team, 2004b. HRSC: the HighResolution Stereo Camera of Mars Express. In: Mars Express: The ScientificPayload, European Space Agency Special Publications, vol. 1240, pp. 17–36.

Neukum, G., and 11 coauthors, 2007. The geologic evolution of Mars: episodicity ofresurfacing events and ages from cratering analysis of image data andcorrelation with radiometric ages of Martian meteorites. In: Lunar andPlanetary Science XXXVIII, Abstract #2271. Lunar and Planetary Institute,Houston (CD-ROM).

Oberst, J., Schwarz, G., Behnke, T., Hoffmann, H., Matz, K.-D., Flohrer, J., Hirsch, H.,Roatsch, T., Scholten, F., Hauber, E., Brinkmann, B., Jaumann, R., Williams, D.,Kirk, R., Duxbury, T., Leu, C., Neukum, G., 2008. The imaging performance of theSRC on Mars Express. Planet. Space Sci. 56, 473–491.

Peterson, J.E., 1977. Geologic map of the Noachis quadrangle of Mars. US Geol. Surv.Misc. Invest. Ser. Map I-910.

Peterson, J.E., 1978. Volcanism in the Noachis-Hellas region of Mars, 2. In:Proceedings of the Ninth LPSC, pp. 3411–3432.

Picardi, G., et al., 2004. MARSIS: Mars advanced radar for subsurface andionosphere sounding. In: Mars Express: The Scientific Payload. EuropeanSpace Agency Special Publication, vol. 1240, pp. 51–69.

Picardi, G., et al., 2005. Radar soundings of the subsurface of Mars. Science 310,1925–1933.

Pinet, P., Chevrel, S., 1990. Spectral identification of geological units on the surfaceof Mars related to the presence of silicates from earthbased near-infraredtelescopic CCD imaging. J. Geophys. Res. 95 (#B9), 14435–14446.

Pinet, P.C., Heuripeau, F., Clenet, H., Chevrel, S., Daydou, Y., Baratoux, D.,Rosemberg, C., Bibring, J.-P., Poulet, F., Gondet, B., Mustard, J., LeMouelic, S.,Bellucci, G., the OMEGA Team, 2007. Mafic mineralogy variations across SyrtisMajor shield and surroundings as inferred from visible-near-infrared spectro-scopy by OMEGA/Mars Express. In: Seventh International Mars Conference,Pasadena, CA, Abstract #3146. Lunar and Planetary Institute, Houston(CD-ROM).

Plaut, J.J., et al., 2007. Subsurface radar sounding of the South Polar LayeredDeposits of Mars. Science 316, 92–95.

Plescia, J.B., 2003. Amphitrites-Peneus Paterae/Malea Planum geology. In: Lunarand Planetary Science XXXIV, Abstract #1478. Lunar and Planetary Institute,Houston (CD-ROM).

Plescia, J.B., 2004. Morphometric properties of Martian volcanoes. J. Geophys. Res.109, E03003.

Plescia, J.B., Saunders, R.S., 1979. The chronology of Martian volcanoes. In:Proceedings of the 10th Lunar Planetary Science Conference, pp. 2841–2859.

Potter, D.B., 1976. Geologic map of the Hellas quadrangle of Mars. US Geol. Surv.Misc. Invest. Ser. Map I-941.

Poulet, F., Erard, S., 2004. Nonlinear spectral mixing: quantitative analysis oflaboratory mineral mixtures. J. Geophys. Res. 109, E02009.

Poulet, F., Cuzzi, J.N., Cruikshank, D.P., Roush, T., Dalle Ore, C.M., 2002. Comparisonbetween the Shkuratov and Hapke scattering theories for solid surfaceplanetary surfaces: application to the surface composition of two Centaurs.Icarus 160, 313–324.

Poulet, F., Gomez, C., Bibring, J.-P., Langevin, Y., Gondet, B., Pinet, P., Belluci, G.,Mustard, J., 2007. Martian surface mineralogy from Observatoire la Miner-alogie, l’Eau, les Glaces et l’Activite on board the Mars Express spacecraft(OMEGA/Mex): global mineral maps. J. Geophys. Res. 112, E08S02.

Raitala, J., Kostama, V.-P., Korteniemi, J., Ivanov, M.A., Tormanen, T., Neukum, G.,2007. Western Promethei Terra smooth plains region of Mars: implications ofvolcanic origin and analysis of surface features. In: European Mars Science and

Exploration Conference: Mars Express and ExoMars, Abstract #1118199, ESTEC,Noordwijk, The Netherlands, 12–16 November 2007.

Russell, P.S., Lefort, A., Thomas, N., 2005. Modeling ice stability with topography ona local scale, Mars. In: Lunar and Planetary Science XXXVI, Abstract #1554.Lunar and Planetary Institute, Houston (CD-ROM).

Schminke, H.-U., 2004. Volcanism. Springer, Berlin, 324pp.Scott, D.H., Carr, M.H., 1978. Geologic map of Mars. USGS Misc. Invest. Ser.

Map I-1083, 1:25M.Scott, D.H., Tanaka, K.L., 1982. Ignimbrites of Amazonis Planitia region of Mars.

J. Geophys. Res. 87, 1179–1190.Self, S., 2006. Defining super-eruptions: purpose, prejudices, and limitations. EOS

Trans. AGU 87(52), Fall Meet. Suppl., Abstract V23G-02.Shkuratov, Y., Starukhina, L., Hoffmann, H., Arnold, G., 1999. A model of spectral

albedo of particulate surfaces: implications for optical properties of the Moon.Icarus 137, 235–246.

Smith, D., Zuber, M., Solomon, S., Phillips, R., Head, J., Garvin, J., Banerdt, W.,Muhlemann, D., Pettengill, G., Neumann, G., Lemoine, F., Abshire, J., Aharonson,O., Brown, C., Hauck, S., Ivanov, A., McGovern, P., Zwally, H., Duxbury, T., 1999.The global topography of Mars and implications for surface evolution. Science284, 1495–1503.

Smith, D.E., et al., 2001. Mars Orbiter Laser Altimeter: experiment summary afterthe first year of global mapping of Mars. J. Geophys. Res. 106, 23,689–23,722.

Squyres, S.W., Carr, M.H., 1986. Geomorphic evidence for the distribution of groundice on Mars. Science 231, 249–252.

Strobell, M.E., Masursky, H., 1990. Planetary nomenclature. In: Greeley, R., Batson,R.M. (Eds.), Planetary Mapping. Cambridge University Press, New York, 296pp.

Tanaka, K.L., Leonard, G.J., 1995. Geology and landscape evolution of the Hellasregion of Mars. J. Geophys. Res. 100, 5407–5432.

Tanaka, K.L., Scott, D.H., 1987. Geologic map of the polar regions of Mars. US Geol.Surv. Misc. Invest. Map I-1802-C.

Tanaka, K.L., Scott, D.H., Greeley, R., 1992. Global stratigraphy. In: Kieffer, H.H., et al.(Eds.), Mars. University of Arizona Press, Tucson, pp. 345–382.

Tanaka, K.L., Kargel, J.S., MacKinnon, D.J., Hare, T.M., Hoffman, N., 2002.Catastrophic erosion of Hellas basin rim on Mars induced by magmaticintrusion into volatile-rich rocks. Geophys. Res. Lett. 29, 1195.

Tirsch, D., Jaumann, R., Poulet, F., Matz, K.-D., Bibring, J.-P., Neukum, G., 2008. Aglobal view on the mineralogical composition of dark dunes on Mars. In: Lunarand Planetary Science XXXIX, Abstract #1693. Lunar and Planetary Institute(CD-ROM).

Werner, S., 2005. The Martian crater size–frequency distribution and aspects of theevolutionary history of Mars. Ph.D. Dissertation, Free University, Berlin,Germany, and online in English at: /http://www.diss.fu-berlin.de/2006/33/indexe.htmlS.

Werner, S.C., Neukum, G., 2003. The end of the heavy bombardment as reflected inthe ages of Martian impact basins. In: Lunar and Planetary Science XXXIV,Abstract #1986. Lunar and Planetary Institute, Houston (CD-ROM).

Williams, D.A., Greeley, R., Zuschneid, W., Werner, S.C., Neukum, G., Crown, D.A.,Gregg, T.K.P., Gwinner, K., Raitala, J., 2007. Hadriaca Patera: insights into itsvolcanic history from Mars Express High Resolution Stereo Camera. J. Geophys.Res. 112, E10004.

Williams, D.A., Greeley, R., Werner, S.C., Michael, G., Crown, D.A., Neukum, G.,Raitala, J., 2008. Tyrrhena Patera: geologic history derived from Mars ExpressHRSC. J. Geophys. Res., in press.

Wilson, L., Head, J.W., 1994. Mars: review and analysis of volcanic eruption theoryand relationships to observed landforms. Rev. Geophys. 32, 221–264.

Zanetti, M., Hiesinger, H., Reiss, D., Hauber, E., Neukum, G., 2008. Scallopeddepressions in Malea Planum, southern Hellas basin, Mars. In: Lunar andPlanetary Science XXXIX, Abstract #1682. Lunar and Planetary Institute,Houston (CD-ROM).

Zimbelman, J.R., Johnston, A.K., Patel, S.D., 1998. Photogeologic constraints on theemplacement of the Medusae Fossae Formation, Mars. In: Lunar and PlanetaryScience XXIX, Abstract #1085. Lunar and Planetary Institute, Houston(CD-ROM).

Zuber, M.T., Smith, D.E., Solomon, S.C., Muhleman, D.O., Head, J.W., Garvin, J.B.,Abshire, J.B., Bufton, J.L., 1992. The Mars Observer Laser Altimeter investigation.J. Geophys. Res. 97, 7781–7798.