GEOPHYSICAL RESEARCH LETTERS, VOL. 33, L11201, doi:10.1029/2006GL025946, 2006
Recent high-latitude icy mantle in the northern plains of Mars:
Characteristics and ages of emplacement
V.-P. Kostama,1 M. A. Kreslavsky,2,3 and J. W. Head2
Received 3 February 2006; revised 23 April 2006; accepted 2 May 2006; published 3 June 2006.
[1] Martian high-latitude zones are covered with a smooth,
layered ice-rich mantle containing characteristic polygonal
patterns. The mantle is especially uniform and
homogeneous in the northern plains. We performed a
survey of decameter-scale mantle textures and circular
features, including impact craters, in the northern lowlands
down to 50°N latitude. Strongly altered and mantled craters
of impact and non-impact origin form a population of subtle
circular features below the mantle. Pits of non-impact origin
are numerous in some regions at lower latitudes. Impact
craters superposed on the mantle are small and very sparse.
The inferred mean crater retention age of the mantle is
0.1 Ma. The spatial distribution of young craters suggests
an age difference between the highest latitudes (younger)
and some lower-latitude regions (older). Latitudinal trends
in polygon textures agree with impact crater evidence for a
latitudinal age progression of mantle properties.
Citation: Kostama, V.-P., M. A. Kreslavsky, and J. W. Head
(2006), Recent high-latitude icy mantle in the northern plains of
Mars: Characteristics and ages of emplacement, Geophys. Res.
Lett., 33, L11201, doi:10.1029/2006GL025946.
1. Introduction
[2] A global zonality of surface textures on Mars was
first seen in Mariner 9 images by Soderblom et al. [1973]
who hypothesized the presence of eolian mantles at high
latitudes. Different types of latitude-dependent terrain softening were observed in Viking images [e.g., Squyres et al.,
1992]. Data collected by Mars Orbiter Laser Altimeter
onboard Mars Global Surveyor (MGS) showed remarkable
smoothing of subkilometer-scale topography at high latitudes, which was attributed to mantling deposits
[Kreslavsky and Head, 2000]; this mantle has a specific
decameter-scale surface texture clearly seen in high resolution images taken by Mars Orbiter Camera (MOC) onboard
MGS [Malin and Edgett, 2001] (Figure 1). The mantle is at
least tens of meters thick and is comprised of several layers
[Kreslavsky and Head, 2002]. In mid-latitude zones (30 –
50° latitude) the mantle is dissected and eroded [Mustard et
al., 2001], which has been attributed to desiccation of icerich material forming more continuous mantles at higher
latitudes. The results from the gamma ray and neutron
spectrometers (GRS suit) onboard Mars Odyssey [e.g.,
1
Department of Physical Sciences, University of Oulu, Oulu, Finland.
Department of Geological Sciences, Brown University, Providence,
Rhode Island, USA.
3
Also at Astronomical Institute, Kharkov National University, Kharkov,
Ukraine.
2
Copyright 2006 by the American Geophysical Union.
0094-8276/06/2006GL025946$05.00
Boynton et al., 2002] showed that the mantle indeed
contains much ice. The boundaries of the high hydrogen
concentration according to the GRS data are located at
about 60° latitude [e.g., Tokar et al., 2002]; at least the
upper meter of the mantle at lower latitudes is desiccated.
The mantle is geologically young, because it has few
superposed impact craters [Mustard et al., 2001]. Formation
of the mantle is probably related to redistribution of ice on
Mars due to change of latitudinal insolation pattern controlled by obliquity variations. The uppermost mantle layers
were proposed to form during the most recent obliquity
peaks (0.4 – 2 Ma ago) [Head et al., 2003] by migration of
H2O from polar caps, while Levrard et al. [2004] concluded
that the mantle was formed due to the general decrease of
oscillating obliquity 3 – 5 Ma ago by migration of H2O from
equatorial ice deposits.
[3] To learn more about the nature, age, formation and
modification of the mantle, we performed a systematic
survey of small impact craters and mantle morphology in
the mantled high-latitude regions. Here we report on our
results for the northern plains of Mars, where the geological
units, and hence the mantle substrate, are exceptionally
homogeneous.
2. Survey
[4] For our survey we used narrow-angle MOC images of
the northern lowlands taken during cycles E01 to E05
(February – June 2001). This corresponds to the summer
season in the northern hemisphere (solar longitude 110–
180°), and the observational conditions were mostly good.
To make the survey as homogeneous as possible, we used
only images of 4.8 m/pixel resolution. We limited the
survey to terrains northward of 50°N. In total, we examined
about 700 images; 270 of them were of good quality and
contained large areas of typical mantle. This small percentage of images with typical mantle does not reflect the
percentage of area occupied by the mantle; it reflects the
preferential targeting of MOC to certain features.
[5] We systematically reviewed and classified the mantle
texture, studied the morphology and measured the diameters
of all circular features of 50 m to 2 km, as craters larger than
this size are usually not completely covered by individual
MOC images. Features that are smaller than 50 m in
diameter are often identifiable, but size estimates for them
are too inaccurate, and their morphological details are often
not apparent. In total, we measured the diameters of more
then 1100 circular features. Although we made every
attempt to undertake the survey as evenly and as homogeneously as possible, the ability to identify features and
textures sometimes differed from image to image due to
differences in atmospheric conditions, illumination geome-
L11201
1 of 5
L11201
KOSTAMA ET AL.: RECENT HIGH-LATITUDE ICY MANTLE ON MARS
L11201
[7] Local variations of the patterns are often correlated
with and modulated by kilometer-scale topography. The
wrinkle texture on the slopes of km-scale knobs has a radial
orientation. We observed a number of examples of barchan
dunes that have apparently traveled over the basketball
texture without modifying it, suggesting that the texture is
made of rather competent material.
[8] There are latitudinal variations in the occurrence of
the mantle surface texture patterns (Figure 2). The common
basketball texture occurs at all latitudes between 50– 80°N.
The linear texture tends to occur more commonly at higher
latitudes (70 – 80°N). The polygonal texture occurs only at
high latitudes above 70°N. A tendency for the S1 pattern to
occur the highest latitudes was also noted by Mangold
[2005], although he also reported a few sites with such a
pattern below 70°N. The wrinkle texture tends to occur at
lower latitudes (50 – 70°N).
Figure 1. Typical textures of the mantle in the northern
plains: (a) Basketball texture, (b) linear texture, (c) wrinkle
texture, and (d) polygonal texture. Portions of E01/01975,
E04/00026, E01/01868, and E04/00028, respectively.
Samples are 0.9 0.9 km, illumination is from lower left.
try, presence of contrasting albedo features, electronic settings of the camera, etc. Consequently, in analyzing the
survey results, we remain aware of possible latitudinal
(illumination) and regional (albedo pattern) biases. Nevertheless, we are very confident that in the surveyed images
we identified all fresh features larger than 50 m that are
superposed on (disrupt) the mantle.
3. Results
3.1. Mantle Textures
[6] The mantle has a very distinctive decameter-scale
surface pattern that is well seen in the images (Figure 1).
The ‘‘basketball’’ texture, formed by very evenly spaced
knobs, (Figure 1a; see also Malin and Edgett [2001],
Figure 123d) is the most common and typical texture. The
highest quality, highest-resolution (1.6 m/pix) low-sun
images show that the knobs are dome-shaped, with the
steepest slopes <6°. In some places, the knobs are organized
into highly coherent linear structures (Figure 1b, see also
Malin and Edgett [2001], Figure 123e) forming a linear
texture. The wrinkle texture (Figure 1c) is also very common.
The difference between knobs and wrinkles is gradational.
Wrinkle orientations are locally consistent but vary regionally; they do not have an obvious orientation preference.
Basketball and wrinkle textures correspond to patterned
ground types S2 and S3 in Mangold’s [2005] global survey
of patterned ground on Mars. The polygonal texture
(Figure 1d; Mangold’s [2005] type S1) is formed by small,
rather irregular crack networks and occurs rarely and in
relatively small patches (several km). The spatial scale of
the textures differs within 15–50 m range. The polygonal
texture of the mantle differs from polygons that are observed
on the floors of some large impact craters (type LT of
Mangold [2005]); these polygons have an order of magnitude
larger spacing and a different morphology.
3.2. Mantle Morphology
[9] The mantle is almost continuous in the high latitudes
(>65°N); only layered material of the polar deposits, icy
polar cap outliers, dunes and a few very steep slopes are not
covered by the mantle. There are several examples where
the mantle is eroded on steep slopes, exposing its layered
structure. In a number of images the uppermost meter-scalethick layer of mantle is not continuous, and forms patches
outlined by a gentle scarp.
[10] Below 60– 65°N erosion is ubiquitous and occurs
not only on slopes, but also on flat surfaces. Often, erosion
of the mantle produces specific dissected patterns described
and surveyed by Mustard et al. [2001]. We analyzed many
clear examples showing that the typical dissection described
by Mustard et al. [2001] is dissection of the mantle
originally characterized by basketball texture. In many sites,
however, dissection does not take place, and the eroded
edges of the mantle form lobate scarps.
3.3. Pits
[11] In the 50– 70°N latitude zone, there are a number of
circular or quasi-circular pits (Figure 3). They are different
Figure 2. Latitudinal distribution of the (left) mantle
texture occurrence and (right) pits and fresh craters. Plotted
is the percentage of images of good quality having the
specific features within the latitudinal zones.
2 of 5
L11201
KOSTAMA ET AL.: RECENT HIGH-LATITUDE ICY MANTLE ON MARS
Figure 3. (a) Pits covered with the mantle. Small sharp
feature shown by arrow is interpreted to be a fresh impact
crater (40 m). Portion of MOC NA image E02/01984,
1.4 km wide, illumination is from lower left. (b) This area is
located approximately 50 km from that shown in Figure 3a
and apparently belongs to the same huge cluster of pits.
Impact crater (arrow) differs from the pits in its morphology
(raised rim, circular shape). Portion of THEMIS VIS image
V05897016, 5.5 km wide, illumination from left.
from typical raised-rim impact craters in their morphology.
Some of them clearly have a collapse origin. Some of the
pits are fresh; they cut the uppermost mantle layer, and
disrupt the mantle texture, but typically they are softer and
are covered with the textured mantle (Figure 3a). Some pits
appear to be distributed randomly, but typically they form
dense clusters. THEMIS images (Figure 3b) show that the
clusters are huge and contain a large number of pits. Unlike
the randomly distributed pits, these clusters of pits are
observed in a single region of the northern plains to the N
and NE of Alba Patera (240 – 280°E). These pits are
apparently related to the local geology of this region,
characterized by lava flows from Alba Patera.
3.4. Mantled Crater-Like Features
[12] There are numerous circular features at high latitudes
interpreted by Malin and Edgett [2001] to be mantled
craters (see their Figures 118 and 119). These features span
a wide range of morphologies from well-expressed impact
craters covered with a patterned surface (Figure 4a) to more
subtle circular chains of lineaments and/or arcuate albedo
markings (Figure 4b). We identified a continuous series of
transitional morphologies, which could be considered as
evidence of an impact origin of the subtle features. On the
other hand, the rims of the circular features, when seen,
sometimes are irregular, deviating from the more common
circles of impact craters (Figure 5a). These features are
clustered to a much greater degree than would be expected
for a production population of impact craters. Cratered
cones similar to those observed in Isidis Planitia [Bridges
et al., 2003] (Figure 5b) outside the mantle extent, could be
candidates for some of the mantled circular features.
[13] The size-frequency distribution of the largest (1 –
2 km) mantled crater-like features in the 50– 80°N latitude
zone is close to the production function for an Early
Amazonian age (Figure 6). The deflection from the Amazonian/Hesperian boundary (the age of the Vastitas Borealis
Formation (VBF), the substrate below the mantle in the
Northern Lowlands) is less than a factor of two in crater
density. This subpopulation of features may actually be the
same as the production population of the VBF. For smaller
L11201
Figure 4. (a) Impact crater, covered with mantle. (b) Subtle
circular feature in the upper center, fresh small impact crater
(60 m) in the lower right (arrow). Two portions of E02/
01380, 1.4 km wide. Illumination is from lower left.
Figure 4b has 30% higher contrast than Figure 4a.
diameters (below 1 km), the crater retention age is much
younger. Furthermore, a break in slope in the size-frequency
distribution (arrow in Figure 6) may be related to the
characteristic onset size of a non-impact subpopulation of
circular features, such as those shown in Figure 5a.
3.5. Fresh Craters
[14] We identified 26 small circular features >50 m in
diameter, almost certainly impact craters superposed on the
mantle (see examples in Figures 3a and 4b); the largest of
these is 200 m. All these craters are somewhat degraded;
they do not have apparent ejecta. In addition to that, 20–
25 small features that we conservatively classified as pits,
could conceivably be somewhat softened impact craters
superposed on the mantle. The mean density of superposed
impact craters in the survey area, 0.0016 km 2, corresponds
to a mean crater retention age of 0.1 Ma, according to the
Neukum-Hartmann production function [Ivanov, 2001].
This estimate is uncertain (within an order of magnitude
and perhaps more), because (1) the scaling from the Moon
to Mars is accurate within a factor of 2 [Ivanov, 2001];
(2) the extrapolation of the projectile flux established for the
hundreds of Ma time scale down to the 1 Ma time scale is
Figure 5. (a) Lager crater-like features covered by the
mantle. Portions of E05/00602, 1.4 km wide. Illumination is
from lower left. (b) Impact (upper part) and non-impact
(lower part) features in Isidis Planitia [Bridges et al., 2003].
Portion of M03/02859; 1.4 km wide.
3 of 5
L11201
KOSTAMA ET AL.: RECENT HIGH-LATITUDE ICY MANTLE ON MARS
Figure 6. Cumulative size-frequency distributions of
mantled craterlike features (upper distribution with arrow)
and of fresh superposed craters (in the lower left corner).
Vertical lines correspond to 85% formal confidence interval.
The arrow points to a faint but statistically significant break
in slope. Thin lines show impact crater production isochrons
dividing the periods of geological history of Mars (the
shaded band is the Hesperian), as well as 0.1 Ma and 4.6 Ma
isochrons. The isochrons are computed using the NeukumHartmann production function recalculated to Mars [Ivanov,
2001]. The dropdown of the isochrons at the right edge is
due to removal of craters larger than 2 km from the
production function.
not firmly grounded; and (3) the atmospheric attenuation of
the small projectile flux is uncertain.
[15] Several images at latitudes below 70°N (e.g., E04/
01263, E05/01597) contain a set of small superposed
craters, apparently randomly scattered over the image; one
or two of them are larger than 50 m, while a few others are
smaller. This situation is very similar to that which would be
expected for an accumulating population of impact craters.
The crater density calculated over several such images gives
an age estimate on the order of 2 Ma. On the other hand, a
large number of images certainly do not have impact craters
superposed on the mantle. This observation suggests that
the crater retention age of the mantle in different locations
may differ by at least an order of magnitude. Despite
statistics of only a few craters, the retention age difference
between individual images is statistically significant at a
high confidence level.
[16] Figure 2 clearly shows a concentration of superposed
craters at lower latitudes. Despite poor statistics, the fivefold difference in crater density is statistically significant.
Thus, in addition to the wide local age variations, there is a
global latitudinal age trend.
4. Discussion
[17] The inferred mean crater retention age of the mantle
(0.1 Ma) formally corresponds better to the mantle formation during the last obliquity peaks (0.3+ Ma ago) [Head et
al., 2003] than to the general decrease of obliquity 3 – 5 Ma
ago [Levrard et al., 2004], although the latter cannot be
L11201
confidently excluded. On the other hand, the observed wide
age differences suggest that the formation of the mantle was
occurring at least in part during the period from 5 to 3 Ma
ago.
[18] We propose two end-member scenarios to explain
the relatively wide differences in the crater retention age of
the mantle: (1) In the first scenario, older layer(s) of the
mantle were emplaced in a wide area prior to 3 Ma ago.
Then, during the last ‘‘ice age’’ from 2 to 0.3 Ma ago
additional layer(s) of the mantle were emplaced in a smaller
area, poleward from 60°N. The contacts between mantles
of different age may not be recognized as such because the
MOC high-resolution coverage is too patchy. (2) In the
second scenario, the icy mantle material was deposited more
or less simultaneously, but the accumulation onset of the
observable crater population is delayed until 10s of meters
of ice sublimate, and the surface texture is formed. At high
latitudes this delay was essentially longer. Some of the
patches of bare ice material outside the polar layered
deposits (polar cap outliers) could be areas still undergoing
delayed ice removal. An intermediate scenario may include
formation of mantle layers of different age, and repeated
deposition and removal of protective decameter-thick layers
of pure ice.
[19] The mantle texture has been interpreted as sublimation polygons [Marchant and Head, 2005; Mangold, 2005]
that are formed due to thermal cracking of permafrost and
further localized sublimation enhanced by cracks. Within
this scenario the polygonal texture (Figure 1d) represents
young polygons, while the basketball texture is a mature
pattern. The concentration of the immature patterns at high
latitudes is consistent both with the age trend inferred from
craters and with the slower desiccation rates at higher
latitudes. In the context of the second scenario, the local
patches of polygonal texture are in the places that were last
to lose the protective ice cover.
References
Boynton, W. V., et al. (2002), Distribution of hydrogen in the near
surface of Mars: Evidence for subsurface ice deposits, Science, 297,
81 – 85.
Bridges, J. C., et al. (2003), Selection of the landing site in Isidis Planitia of
Mars probe Beagle 2, J. Geophys. Res., 108(E1), 5001, doi:10.1029/
2001JE001820.
Head, J. W., J. F. Mustard, M. A. Kreslavsky, R. E. Milliken, and D. R.
Marchant (2003), Recent ice ages on Mars, Nature, 426, 797 – 802.
Ivanov, B. A. (2001), Mars/Moon cratering rate ratio estimates, Space Sci.
Rev., 96, 87 – 104.
Kreslavsky, M. A., and J. W. Head III (2000), Kilometer-scale roughness of
Mars’ surface: Results from MOLA data analysis, J. Geophys. Res., 105,
26,695 – 26,712.
Kreslavsky, M. A., and J. W. Head III (2002), Mars: Nature and evolution
of young latitude-dependent water-ice-rich mantle, Geophys. Res. Lett.,
29(15), 1719, doi:10.1029/2002GL015392.
Levrard, B., F. Forget, F. Montmessin, and J. Laskar (2004), Recent ice-rich
deposits formed at high latitudes on Mars by sublimation of unstable
equatorial ice during low obliquity, Nature, 431, 1072 – 1075.
Malin, M. C., and K. S. Edgett (2001), Mars Global Surveyor Mars Orbiter
Camera: Interplanetary cruise through primary mission, J. Geophys. Res.,
106, 23,429 – 23,570.
Mangold, N. (2005), High latitude patterned grounds on Mars: Classification, distribution and climatic control, Icarus, 174, 336 – 359.
Marchant, D. R., and J. W. Head (2005), Equilibrium landforms in the Dry
Valleys of Antarctica: Implications for landscape evolution and climate
change on Mars, Lunar Planet. Sci., XXXVI, abstract 1421.
Mustard, J. F., C. D. Cooper, and M. K. Rifkin (2001), Evidence for recent
climate change on Mars from the identification of youthful near-surface
ground ice, Nature, 412, 411 – 414.
4 of 5
L11201
KOSTAMA ET AL.: RECENT HIGH-LATITUDE ICY MANTLE ON MARS
Soderblom, L. A., M. C. Malin, J. A. Cutts, and B. C. Murry (1973),
Mariner 9 observations of the surface of Mars in the north polar region,
J. Geophys. Res., 78, 4197 – 4210.
Squyres, S. W., S. M. Clifford, R. O. Kuz’min, J. R. Zimbelman, and F. M.
Costard (1992), Ice in the Martian regolith, in Mars, pp. 523 – 554, Univ.
of Ariz. Press, Tucson.
Tokar, R. L., W. C. Feldman, T. H. Prettyman, K. R. Moore, D. J.
Lawrence, R. C. Elphic, M. A. Kreslavsky, J. W. Head III, J. F. Mustard,
and W. V. Boynton (2002), Ice concentration and distribution near the
L11201
south pole of Mars: Synthesis of odyssey and global surveyor analyses,
Geophys. Res. Lett., 29(19), 1904, doi:10.1029/2002GL015691.
J. W. Head and M. A. Kreslavsky, Department of Geological Sciences,
Brown University, Providence, RI 02912, USA. ([email protected].
edu)
V.-P. Kostama, Department of Physical Sciences, University of Oulu, P.O.
Box 3000, FIN-Oulu, Finland.
5 of 5