JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 110, E12S18, doi:10.1029/2005JE002427, 2005
Complex geology of two large impact craters in Tyrrhena Terra, Mars:
Detailed analysis using MEX HRSC camera data
J. Korteniemi,1 V.-P. Kostama,1 T. Törmänen,1 M. Aittola,1 T. Öhman,1,2
H. Lahtela,1 J. Raitala,1 and G. Neukum3
Received 7 March 2005; revised 15 May 2005; accepted 31 August 2005; published 30 November 2005.
[1] Tyrrhena Terra, located just north of the roughly 200 km Hellas impact basin, is a
typical region of the ancient cratered southern highlands on Mars. Its base material is a
mixture of Hellas ejecta, prebasin remnants, and highland terrain, all later saturated with
smaller (<200 km) impact craters. Fluvial and lacustrine deposits, erosion, tectonic
movements, volcanic materials and aeolian processes have subsequently modified the
region further. In this study we take a closer look at two adjacent unnamed craters in
Tyrrhena Terra, located at 24.5°S, 80.8°E (crater ‘‘A’’) and 23.9°S, 79.3°E (crater ‘‘B’’).
The craters are covered by the 20–80 m/pixel Mars Express High Resolution Stereo
Color imager (HRSC) multispectral data, which together with MOC, THEMIS, and
MOLA data sets allows us to make very detailed analysis of the area. We describe several
identified geological and geomorphological units with their individual characteristics (e.g.,
morphology, color, cratering records, and elevation differences) and interpret their
individual evolution. The crater floors show several unique material types of both
depositional and erosional origin. Crater A has a large 200 m high central massif and a
central peak ring as well as a unique low-lying terrain type, ‘‘honeycomb terrain,’’ with
narrow 50–200 m high ridges and intervening pits. Additionally, both craters exhibit
erosional features similar to each other and nearby craters, indicating episodes of
deposition of various material types in both craters. We find that the crater pair reveals
many characteristics which are typical of the N Hellas rim but are not found elsewhere.
Citation: Korteniemi, J., V.-P. Kostama, T. Törmänen, M. Aittola, T. Öhman, H. Lahtela, J. Raitala, and G. Neukum (2005),
Complex geology of two large impact craters in Tyrrhena Terra, Mars: Detailed analysis using MEX HRSC camera data,
J. Geophys. Res., 110, E12S18, doi:10.1029/2005JE002427.
1. Introduction
[2] The surface of Mars exhibits areas of both smooth
northern lowlands and intensely cratered highland terrain in
the south. The cratering record can be used to obtain the age
of the surface [e.g., Kreiter, 1960; Hartmann and Neukum,
2001; Neukum et al., 2004b], while individual craters’
original morphology hints to the target material properties
at the time of impact [Melosh, 1989; Öhman et al., 2005a].
Additionally, the craters provide natural topographical sinks
for later deposits, thus showing glimpses of the evolution of
the region they reside in.
[3] The 2000 km diameter Hellas impact basin (Figure 1)
dominates about half of the southern hemisphere of Mars.
The region has volcanic, tectonic, glacial, fluvial and
aeolian features [e.g., Carr and Schaber, 1977; Crown et
al., 1992; Crown and Greeley, 1993; Leonard and Tanaka,
1
Astronomy Division, Department of Physical Sciences, University of
Oulu, Oulu, Finland.
2
Department of Geosciences, University of Oulu, Oulu, Finland.
3
Institut für Geologische Wissenschaften, Department of Earth
Sciences, Freie Universität Berlin, Berlin, Germany.
Copyright 2005 by the American Geophysical Union.
0148-0227/05/2005JE002427$09.00
2001; Raitala et al., 2004; Öhman et al., 2005a, and
references therein], as well as a multitude of large and
small impact craters. Tyrrhena Terra, on the northeastern
Hellas basin rim [Tanaka and Leonard, 1995] (Figure 2a)
has been interpreted to be a mixture of Hellas ejecta,
uplifted crustal material and cratered highland [Schaber,
1977; Greeley and Guest, 1987], in places cut by dendritic
fluvial channels [Leonard and Tanaka, 2001].
[4] Many of the craters on Mars, especially around the
Hellas region, exhibit anomalous depressions on their floors
(Figure 1) [Korteniemi, 2003; Korteniemi et al., 2003, 2005;
Ansan and Mangold, 2004; Mest and Crown, 2004, 2005;
Moore and Howard, 2005; Ansan et al., 2005; Wilson and
Howard, 2005]. These features occur only within craters in
a specific region, and thus are indicative of a distinct
evolution process (see more in section 3.3). In this study,
we take a close look at a pair of large impact craters with
such depressions, located in Tyrrhena Terra. The eastern
crater (24.5°S, 80.8°E, d = 95 km), referred to as crater A,
and the western crater referred to as crater B (23.9°S,
79.3°E, d = 75 km), form an adjoined pair of same-size
craters (Figure 3a), which both harbor a complex morphology composed of several identifiable material units with
distinct topographies. The features include, e.g., deep pits
with intervening filament-like ridges, a peak ring, a large
E12S18
1 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
central massif of layered material and rugged, highly eroded
terrain. These features are measured, analyzed and interpreted mostly using a multitude of data sets, in order to
explain the geologic modification history of the two craters.
E12S18
This, in part, also helps us understand the changes in the
environment the pitted or ‘‘depressed’’ craters were subjected to. The results also reveal some of the evolutional
phases of the region, possibly related to the Hellas basin
Figure 1
2 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
interaction with its surroundings [e.g., Raitala et al., 2004;
Ivanov et al., 2005].
2. Data and Methods
2.1. Data Sets
[5] The studied area has been covered both extensively
and in detail by the multispectral MEX HRSC (Mars
Express High Resolution Stereo Color imager) [Neukum
et al., 2004a] orbit 389 image set. In the studied region, the
spatial resolutions of the nadir, stereo and color (+photometric) channels were constantly at 20, 40 and 80 m/
pixel, respectively. In conjunction with HRSC, we also used
THEMIS (Thermal Emission Imaging System on board the
Mars Odyssey probe) IR data (resolution 100 m/pixel) to
find possible differences in surface materials [Christensen et
al., 2003]. An assortment of better resolution Mars Global
Surveyor Mars Orbiter Camera (MGS MOC) NA (narrow
angle, 1.4– 5 m/pixel) and THEMIS visual (17 m/pixel)
images were used in order to determine the morphologic
characteristics of the surface units. However, these two data
sets were usable only in places, due to their limited and
scattered coverage. The topography of the region was
deducted from MGS MOLA DTMs (Mars Orbiter Laser
Altimeter digital terrain models, 16-bit data, 128 pixels/
degree), in conjunction with 3D-anaglyph images made
using the HRSC stereo and photometric channels. Imaging
devices before the HRSC have covered the area with either
higher spatial resolution but less areal coverage, or with
large coverage but poorer spatial resolution. The HRSC
images provide an additional, multispectral data set, with
good constant resolution and large areal coverage, i.e., a
very good and unprecedented basis for detailed studies such
as this.
2.2. Crater Counting
[6] The crater counting done in this study (Figure 4) was
used mainly to identify and separate the different surface
units from each other, and generally not for dating the
surfaces. Crater production functions [e.g., Neukum and
Ivanov, 2001, and references therein] work well only for the
ages of unit surfaces, not for the ages of the actual geologic
materials. This comes to play when the unit surfaces are,
e.g., protected from impacts by overlying deposits, and only
later become exhumed. Additional erosion processes may
also alter the surface, removing impact features. We interpret that these factors have played a major part in the
evolution of many units in the studied region. Thus the
E12S18
actual ages of the mapped geologic units cannot be determined from the cratering record of their exposed surfaces,
as the surface ages may easily be much younger than the
geologic units themselves. Additionally, the surface areas of
the mapped geological units and the amount of sufficiently
large craters (1+ km) are so small that no statistically
accurate dating would be possible [e.g., McEwen et al.,
2005; Hartmann, 2005; Plescia, 2005]. However, crater
counting does have one very important application in this
study, namely the distinction between the surfaces of the
various geological units. For more on the topic of crater
calculations, see section 5.5.
3. Regional Setting
3.1. General Geology
[7] The studied region has a 0.5-degree southward slope
toward the Hellas basin. It consists of highly cratered
highlands (unit Npl1, Figure 3b) mixed with Hellas basin
rim and ejecta (unit Nh) as well as scattered mountains of
prebasin crustal material (unit Nm) [Schaber, 1977; Greeley
and Guest, 1987; Leonard and Tanaka, 2001]. The area has
a high concentration of large and small craters (the largest of
which are mapped as C1, C2 and C3) and it is generally of
Noachian age. Additionally, the floors of the craters A and
B exhibit several depositional and erosional units with
distinct ridges and intervening pits (Nmh) or with cratered,
smooth, rugged and etched appearances (Nmc, HNms, Nmr
and HNme, respectively). All these floor units are described
in section 4 and analyzed further in section 5.
[8] Fluvial valley networks are a feature found often in
the Martian cratered highlands [e.g., Carr, 1996], and the
studied region is no exception. Dendritic channel systems
have previously been mapped north and east of the studied
craters [e.g., Leonard and Tanaka, 2001; Mest and Crown,
2004]. Several additional fluvial channel networks are seen
in the HRSC images, mostly on the immediate southern side
of the craters (Figures 3b and 5). The channels originate
from topographic highs such as the elevated crater rims or
local mountains. When continuing downhill, they often
merge with each other to form larger dendritic channel
networks. In places, they disappear due to later modification
events and then appear again further down the slope. A
concentration of locally uniform pits (Figure 6) is often
observed where the channels become indiscernible. Some of
the pits exhibit rims, indicating that they are caused by
either multiple impacts or secondary ejecta from larger
impact craters. However, several characteristics speak
Figure 1. (a) This shaded relief map from the greater Hellas region (0 – 70°S, 20– 110°E) shows the distribution of found
depressions on crater floors near the Hellas basin. The Viking study made from 30+ km craters earlier (white circles)
[Korteniemi, 2003] shows a large concentration of the features on the W side of Hellas. The study by Moore and Howard
[2005] shows a similar distribution, with a few variations (X marks). The HRSC images do not yet cover that area (white
outlines show locations of HRSC images released before May 2005). The study made from the HRSC data using 2+ km
craters (black circles) [Korteniemi et al., 2005] showed there to be many times more depressed craters on the N and E sides
of the basin. The black box shows the location of Figure 2. Mest and Crown [2005] found several craters with floor deposits
and/or pits in that region (+ marks). (b) Example of ‘‘type 1’’ depression, i.e., not crater-related faults going through the
craters (21.6°S, 200.7°E). (c) Two ‘‘type 2’’ examples of floor-fractured craters, one with rim-circling depression (lower
left) and another with a fully developed fracture system (center, 6°S, 108°E). (d) Example of isolated pit on crater floor
(31.7°S, 40.95°E), ‘‘type 3.’’ Note the triangular shape; walls are often straight, indicating that they may be tectonically
influenced. Example images are from Viking MDIM2.
3 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
E12S18
craters, where almost no fluvial channels are seen. Instead,
the pits tend to occur on the craters’ southern side, and even
there they only occupy the local lows connected with fluvial
channels. These pits mostly form clusters rather than, e.g.,
chains characteristic of tectonically controlled collapse pits
[Tanaka and Golombek, 1989]. Instead, they are interpreted
to have been associated with the release of volatiles from
the sediments, originally deposited by the channels. The pits
are thus most probably caused by ice sublimation and
subsequent ground collapses [Kostama et al., 2003].
[9] The images from the area show two sparse sets of
wrinkle ridges trending N-S and E-W (Figure 6a) to both the
north and south of the studied craters on the cratered
highlands [Chicarro et al., 1985]. These structures indicate
large-scale compressional stresses within highly competent
crustal materials. They may be, but not necessarily are of
volcanic origin [Greeley and Spudis, 1978]. Chicarro et al.
[1985] showed that the wrinkle ridge patterns in the highlands are aligned with the large impact basins. Öhman et al.
[2005b] made a survey of polygonal impact craters in the
Hellas region, and found that the regional zones of weakness are also dominated by directions radial and concentric
to Hellas. Thus the regional tectonism in the study area is
clearly dominated by this ancient basin.
Figure 2. (a) The southwestern Tyrrhena Terra region
(600 by 800 km; 18– 30°S, 72– 88°E) shown in MOLA
gray scale DTM. The contours are set 2 km apart. The
region exhibits a general 0.5° slope toward the southwest.
Craters marked as A and B are studied in detail in this
study; Millochau and Terby have been analyzed previously
[Mest and Crown, 2004, 2005; Ansan and Mangold, 2004;
Ansan et al., 2005; Wilson and Howard, 2005]. The craters
numbered 1 –5 are examples of craters of roughly the same
size discussed further in Table 1. (b) The crater types of the
region, according to the Barlow [2000, 2003] crater
catalogue: Unclassified craters (gray circle), almost destroyed craters (white circle), and two crater types with
lobate (black circle) and ballistic (black dashed circle) ejecta
fields may either have flat crater floors (white infilling) or
include central peaks (vertical lines), peak rings (horizontal
lines), summit pits (upward diagonal lines), or central pits
(downward diagonal lines). Note that crater A is the only
peak ring crater within the region.
against all or even most of the pits being impact created.
(1) None of the pit clusters exhibit any ‘‘herringbone’’
patterns, which are characteristic to secondary impacts.
(2) The pit clusters are not aligned with each other, or with
any large ‘‘parent’’ craters in their neighborhood. (3) The
distribution of the pits within the region is not uniform:
They do not occupy as much the northern side of the studied
3.2. Regional Crater Morphology
[10] The Barlow crater catalogue [Barlow, 2000, 2003],
available at ftp://ftpflag.wr.usgs.gov/dist/pigpen/mars/
crater_consortium and http://webgis.wr.usgs.gov/mars.htm,
was used to find the general characteristics of the crater
floors in the region surrounding the studied craters
(Figure 2; 600 by 800 km; 18– 30°S, 72– 88°E). The
catalogue includes all craters on the planet larger than 5 km in
diameter. Special morphology types are given for those
craters whose ejecta blankets are identified. Though somehow incomplete, the catalogue shows (Figure 2b) that most of
the classified craters in this region have fluidized ejecta [Carr
et al., 1977] and roughly half of the craters have central peaks,
central pits or summit pits. According to the catalogue and our
own findings (see section 5.2), crater A is the only peak-ring
crater in the region. This infers that the formation of this
particular crater has been unique within the region and
deserves more detailed investigation.
[11] Fresh complex craters on Mars with diameters of
60– 100 km have typically raised rims, a central peak and/or
a peak ring and a generally flat floor with slumping on the
inner walls [Melosh, 1989; Strom et al., 1992]. These
features are subsequently modified, altered and generally
smoothened by additional geological processes characteristic for the area, e.g., erosion, sedimentation and later
cratering.
[12] A parameter, which gives some insight to the crater
degradation state, is the ratio of depth (d) to diameter (D).
For topographically fresh complex craters between 7 and
100 km in diameter, the weighted power law function
according to Garvin et al. [2003] is
d ¼ 0:36 D0:49 ;
ð1Þ
where all parameters are in km. They calculated also a
formula for the rim height H for the same population, i.e.,
4 of 21
H ¼ 0:02 D0:84 :
ð2Þ
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
Figure 3
5 of 21
E12S18
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
The average diameter and height parameters measured from
MOLA, Viking and HRSC data for nine 60– 170 km craters
in the region (see Figure 2) are shown in Table 1, and
compared to derived ‘‘fresh’’ crater values obtained using
formulas (1) and (2). It is apparent that all the craters in the
region have floors mantled by deposits or have highly
eroded rims, most probably both [e.g., Craddock et al.,
1997; Boyce et al., 2003, 2004b; Forsberg-Taylor et al.,
2004]. The craters A and B are among the most modified
ones, having 1.4 and 2.1 km of floor infilling,
respectively (see Figure 2).
3.3. Pits and Depressions on Crater Floors on Mars
[13] One quite rare type of postimpact crater modification
is the creation of depressions on the crater floor. We have
previously done a planet-wide study on these anomalous
crater floors [Korteniemi et al., 2003], using the Viking
MDIM2 data set with 231 m/pixel resolution. While the
Barlow catalogue includes 45000 craters (d > 5 km)
planet-wide, we found there to be only roughly 400 craters
(d > 30 km) with depressions. These are further categorized
into three generalized types with distinct characteristics and
distribution patterns:
[14] 1. Crater-independent collapses most often do not
follow the structural patterns (e.g., circular shape) of the
crater and appear not to be related to the crater itself. They
are rather related to large-scale regional deformations, and
include mostly grabens and associated features crosscutting
and deforming both the crater rim and floor (Figure 1b).
This type of depression is found mainly in the regions
modified heavily by tectonism, such as Claritas and Memnonia Fossae or Tempe Terra [Korteniemi et al., 2003].
[15] 2. Floor-fractured craters exhibit mainly narrow
fissures that are restricted to the crater floor, causing it to
have both concentric and radial depressions and a general
web-like appearance (Figure 1c). The evolution of these
features is studied in more detail by Newsom [2001].
Similar crater floors have also been identified on the Moon,
Venus and possibly Earth and have been interpreted to be a
result of local volcanic activity made possible by fracturing
driven by the impact event [Schultz, 1976; Schultz and
Orphal, 1978; Wichman and Schultz, 1995]. On Mars, this
crater type is usually seen close to the dichotomy boundary,
on its southern highland side, with examples found also in
the fluvially deformed regions near Valles Marineris
[Korteniemi et al., 2003].
E12S18
[16] 3. Isolated pits on crater floors are generally small
(width<10% of crater diameter) and irregular collapse and
depression pits are found only in parts of the crater floor
(Figure 1d) [Korteniemi et al., 2003; Moore and Howard,
2005]. Usually no other crater deforming features such as
grabens or fluvial channels are associated with these pits
or their ‘‘parent’’ craters. Furthermore, there are generally
no depressions or other resembling features found outside
the impact craters. Many depression walls exhibit layering, suggesting sedimentary deposits around them. The
walls are additionally often straight and parallel with each
other, either within one crater or in a group of nearby
craters. The latter indicates that the walls have been either
created or altered by a tectonic trend, of regional origin
[Korteniemi et al., 2003], rather than being controlled by
the shape of the crater itself. This depression type is most
common in craters in the Hellas rim region, with a clear
concentration (70%) on the west and northwestern sides
of the basin. In the study made using Viking data we
identified 78 craters in a cluster near and around Hellas
(Figure 1a). The trends described above show some
correlations with zones of weakness indicated by the
shapes of polygonal craters in the Hellas region [Öhman
et al., 2005a]. This suggests that the Hellas-created
tectonism may be a controller of the depression shapes.
[17] Using the available HRSC data, we have now done a
new search for depressions on crater floors (crater d > 2 km)
within the Hellas region [Korteniemi et al., 2005]. The new
data set covers only parts of the northern and eastern rim
regions of the Hellas basin, with a total covered area of over
4 106 km2 (Figure 1a). Thus we do not see most of the
features found in the earlier study, but some interesting
preliminary findings were already made:
[18] 1. 113 craters were identified to exhibit depressions
on their floors, in clear contrast to the 11 examples found
from Viking data in the same area. This is probably due
to the better resolution of HRSC images and lighting
conditions.
[19] 2. The type 2 (described above) web-like fissure
depressions were identified in some of the craters on the
Hellas basin floor (Figure 1a). This may indicate that the
craters in question were subjected to conditions similar to
that of the fluvially deformed regions east of Tharsis.
[20] 3. A new depression type (type 4) was categorized,
i.e., rugged depression, with a characteristically knobby,
pitted or etched surface. The pits in question are signifi-
Figure 3. The area of study: 23.1 – 25.6°S, 78.4– 81.9°E. (a) The mosaic of enhanced color HRSC orbit 389 RGB image
with complementing THEMIS day-IR data shows the studied craters. The white star and arrow on the upper right show the
direction of sunlight. THEMIS images in mosaic: I07026002, I07313002, I08062005, I08524002, I08836002, and
I10296002. (b) Geological map from the same area showing the identified material and terrain units. The dotted circle
represents the location of the peak ring in crater A. Regional units (first three adapted from Leonard and Tanaka [2001]):
Npl1 (cratered plains unit; modified highland material), Nh (hilly plains unit; modified Hellas ejecta), Nm (massif unit; preHellas resistant crustal material), blue line (sinuous channel; mostly of fluvial origin), black dotted line (wrinkle ridge;
compressional feature), and white circle (collapse pit). Crater units: C1A (Crater A), C1B (Crater B), C2 (craters with no
ejecta, some highly eroded), C3 (craters with ejecta fields), and black line w. double stroke (crater rim crest). Crater floor
units: Amd (dune unit; aeolian deposits), HNms (smooth unit; relatively fresh sediments), Nmr2 (upper rugged unit;
medium-eroded sediments), Nmr1 (lower rugged unit; highly eroded sediments), HNme (etched unit; highly eroded
sediments), Nmc2 (upper cratered unit; partly eroded cratered sediments), Nmc1 (lower cratered unit; partly exhumed
cratered sediments), Nmh (honeycomb unit), and white line (honeycomb ridge crest). The area surrounded by plus marks in
crater B represents an area which has undergone an episode of mass wasting.
6 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
cantly smaller than the type 3 depression (described above).
Additionally, rather than forming only one or a few depressions on the crater floor, they create entire floor units with
numerous depressions and intervening highs. The knobs
and other high elevation areas are usually of different
albedo than the pit floors and other lower levels. The rugged
terrain is often characterized by typical erosion marks such
as yardangs. We have interpreted [Korteniemi et al., 2005]
this terrain type to be a result of a layered structure on top of
the original basement, created by, e.g., fluvial or other
mantling processes. Afterward the more friable deposit
materials are eroded away, leaving behind the rugged
surface we see today. This depression type is observed in
clusters, especially on the eastern side of the Hellas rim
plains. It is found also outside impact craters, supporting the
idea of a simple sediment-related cause.
[21] In the region under study here we have identified
four craters with floor depressions; craters A, B, Terby and
Millochau. Terby (30°S, 75°E, d = 170 km) has a 2 km
deep, W-shaped closed depression with rim-high protruding
lobes superposed on it [Ansan and Mangold, 2004; Ansan et
E12S18
al., 2005; Wilson and Howard, 2005]. Millochau (21.4°S,
85°E, d = 115 km) has a 400 m high tilted central plateau
with deeply etched terrain partly surrounding it [Mest and
Crown, 2005]. Both Terby and Millochau exhibit layering in
their surface units, many of which are interpreted as the
result of volatile-rich sedimentation [Mest and Crown,
2004, 2005; Ansan and Mangold, 2004; Ansan et al.,
2005; Wilson and Howard, 2005]. Both craters in our study
(especially the eastern crater A) display characteristics
similar to those of Terby and particularly Millochau. The
accuracy of the HRSC data set, however, enables us to study
these two chosen example craters in more detail. The
existence of these several large craters with similarly
Figure 4. Size-frequency distribution curves of small
craters (diameters 90– 9000 m) calculated from the studied
crater floor materials and terrain units (from Figure 3b). The
crater counts indicate the relative ages of the units in
question, with old terrains having more and larger craters
than young ones. The existence of only a few >1 km craters
makes absolute age determination [Tanaka, 1986] risky at
best, as is seen from the provided overlapping error bars.
The crater counts were done from HRSC orbit 389 nadir
images. The unit names are followed by the number of
craters used for that curve. Two curves are also noted with
percentages; they indicate the amount of coverage of HRSC
images within that unit. 100% would be full coverage; 50%
would be half. (a) The two separate curves (gray symbols)
given for the cratered unit show the division into two
different subunits (Nmc2 shown as diamond and Nmc1
shown as square). The honeycomb terrain unit (plus signs)
and etched material unit (x marks) show curves with
shallower gradients in the small crater end. This is probably
due to their high-relief topography, which was caused by
deep-cutting erosion of overlying material and subsequent
exhuming of the unit material. Additionally, due to this
topography, most of the later small impacts onto the steep
slopes have no doubt been buried or destroyed by mass
wasting caused either by the impact event itself or by later
processes. Thus no visible craters will be left on those
occasions. (b) A comparison of the only two material units,
which are found in both craters: the rugged (triangles) and
smooth materials (circles). The parts of units which are
located in crater A are marked with gray symbols (HNmsA
and NmrA), whereas the ones in crater B are marked with
white symbols (HNmsB and NmrB). The rugged parts
coincide with each other quite well throughout the curve,
supporting the interpretation that they are indeed different
parts of one geological unit. The smooth unit in crater B
appears to be somewhat older than its counterpart in crater
A. This is probably an artifact, caused by the existence
several secondary crater clusters inside crater B. The
smooth material inside crater B is more modified (e.g., by
fluvial channels) than in crater A, which would indicate that
the smooth material filling ended earlier inside crater B. It is
also notable in the case of crater B that the areas from which
the crater counts were made represent only parts of the
whole mapped unit in that crater. This is due to the HRSC
nadir image not extending to the westernmost third of the
crater and leads to a larger uncertainty in the accuracy of the
particular size-distribution curves.
7 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
E12S18
multiple events and processes have to be applied in order to
find the right interpretation(s). In order to discover the
evolution sequence, we have first defined the geomorphologic units and their individual characteristics. Each unit
Figure 5. Examples of features found in the study region,
part 1. (a) The black boxes show the locations of sample
Figures 5b– 5c, 6a – 6c, 9b– 9j, and 10b – 10d. (b) A large
fluvial channel entering crater B. It may be a sapping
channel, judging from its width to length ratio. Cutout from
THEMIS IR image I08836002. (c) Fluvial activity has
created dense networks of dendritic channels on the craters’
outer rim. Cutout from HRSC nadir image.
deformed floors within the region suggests that they were
subjected to similar geologic processes and events. This, in
turn, may be indicative to the evolution of the whole Hellas
region [e.g., Leonard and Tanaka, 2001; Raitala et al.,
2004]. The geologic map (Figure 3b) shows the distribution
of various crater units within the study area.
4. Studied Craters
[22] The floors of the studied craters (Figure 3) show
topographic and morphologic anomalous features compared
to ‘‘regular’’ impact craters [e.g., Melosh, 1989; Strom et
al., 1992]. The two craters are of slightly different ages,
inferred by the degradation of their rims. At the same time,
the craters show similarities in their floor features due to the
geologic postimpact processes. The complex and versatile
crater floor morphologies and units require that a set of
Figure 6. Examples of features found in the study region,
part 2. (a) The fluvial channels are often interrupted by pit
clusters. Some may be secondary impacts (impacts marked
black in the inset), while others are not; we interpret the
clusters in local lows to be sublimation pits. (b) A set of
highland wrinkle ridges (arrows; traced by lines in inset)
continue onto the crater B floor. This indicates that the
regional compression continued after the crater formation.
(c) A possible mass wasting episode has modified the crater
B rim near crater A. The slope is much shallower there, and
the area exhibits clear deposits at the foot of the wall
extending to the HNme unit. White lines in inset are
channels; black lines are ridges or other local highs. All
figures are cutouts from the HRSC 389 nadir image.
8 of 21
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
E12S18
E12S18
Table 1. Geometric Properties of Nine Large Impact Craters (Figure 2a) in the Region 18 – 30°S, 72 – 88°E Compared With Theoretical
Estimatesa
Crater
Diam. D
Depth d
d/D
Rim
Heightb H
Fresh
Depthc,d df
Fresh
Rimc,d Hf
Deposit Depth
df d
Rim Erosion
H Hf
Crater Ae
Crater Be
Terbyd,e
Millochaud,e
Crater 1e
Crater 2e
Crater 3e
Crater 4e
Crater 5e
95
78
170
115
90
80
65
84
67
1.91
0.91
2.10
1.60
2.40
2.11
1.58
2.18
1.15
1/50
1/86
1/81
1/72
1/37
1/38
1/41
1/39
1/58
0.71
0.41
0.25
0.64
0.59
0.59
0.35
0.07
0.19
3.35
3.04
(4.46)
(3.68)
3.26
3.08
2.78
3.16
2.83
0.92
0.78
(1.49)
(1.08)
0.88
0.79
0.67
0.83
0.68
1.44
2.13
(2.36)
(2.08)
0.86
0.97
1.20
0.98
1.68
0.21
0.37
(1.24)
(0.44)
0.29
0.20
0.32
0.76
0.50
a
Comparing the measured values (depth, diameter, and rim height) to the calculated ones (‘‘fresh’’ crater parameters derived from Garvin et al. [2003])
gives an estimation of crater filling deposit and rim erosion. Note: All units are in km.
b
Rim height is measured from the surrounding intercrater plains to the rim crest.
c
Formulas df = 0.36 D0.49 and Hf = 0.02 D0.84 from Garvin et al. [2003].
d
Craters Terby and Millochau are larger than the diameter range for which the above formulas are given (7 < D < 110 km). Thus the calculated values do
not apply for them.
e
Example craters are designated in Figure 2a.
was found to have a distinct topography level of its own as
well as a characteristic extent where each of them does
occur (Figure 7). This gives a certain clue to the layered
nature of the materials and allows further detailed studies of
the deposition-erosion history of the units.
4.1. Crater A: The Peak Ringed and Pitted Youngster
[23] Crater A (the eastern crater, Figure 3a) is 95 km in
diameter with its rim intact all around the crater. The
elevation difference between crater floor and surrounding
plains is 1200 m, and the almost intact rim towers 750 m
above the plains in average. The terrain outside the crater
rim, especially in the east and northwest, is rugged and may
represent the remains of a highly eroded ejecta blanket
(mapped in Figure 3b). The inner walls are cut by mainly
mass wasting furrows, but some more fluvial-like channels
are also recognized (Figure 3b), though they are here much
less prominent than on the walls of the western crater (see
section 4.2). A large 200 m high central massif connected to
a degraded inner ring structure and a multitude of honeycomb-like pits dominate the crater floor.
[24] After the main impact, approximately 680 smaller
craters (diameters 100 – 9000 m) visible in the HRSC image
have formed on the floor of crater A. They allow unique
size-frequency distribution curve calculations for each floor
unit described in section 4.3 (Figure 4a). Only 12 craters
larger than one km in diameter have accumulated onto the
crater floor. This and the small unit sizes (500– 1500 km2)
increase the crater calculation error margins and make age
determination [Tanaka, 1986] quite impossible.
4.2. Crater B: Fluvially Scarred and Ridged for Life
[25] The western crater B (Figure 3a) has a diameter of
75 km. It appears to be the more modified and older of the two
craters studied. Crater B has shallow topography and a highly
modified and broken rim, which is observable only in places.
The slope of the crater’s eastern inner wall has a much
shallower slope (4 – 7 degrees) than elsewhere around its
circumference (10 –16 degrees). For reference, the slope of
crater A is 6– 10 degrees, nearly uniform overall. On the crater
floor just next to this shallower slope region are some deposit
lines (mapped in Figure 3b).
[26] The elevation difference between the floor and
surrounding plains is 500 m, and from rim to plains
400 m. Numerous small channels cut the inner walls of the
crater. Some of them appear to be created by fluvial processes
and extend well onto the crater floor (Figure 3b). The largest
channel (roughly 1 km wide) enters the crater from the
northwest. Some ridges on the crater floor resemble the
wrinkle ridges outside the crater, and have similar topography
and north-south orientation. They also breach the southern
rim, indicating direct continuation from the surrounding plain
(Nh) wrinkle ridges (Figure 6d). However, the erosion, which
has modified the ridge-like formations on the crater floor,
does not allow absolute distinctions of their origin to be
made.
[27] Unfortunately, the HRSC images do not yet cover the
westernmost third of crater B. The only extensive enough
data set to continue over the rest of the crater was a
THEMIS IR image mosaic, which has 1/5 of the HRSC
spatial resolution. This results in several problems: (1) Unit
identification becomes more difficult, since some of the
morphological differences between the units were recognized only at HRSC’s 20 m scale. (2) Crater counting can be
misleading in the THEMIS-mapped region because of
differing illumination circumstances between different
orbits. We did a test, trying to correlate the amount of
craters in overlapping HRSC and THEMIS images. This
resulted in some both very good correlations (60% of
HRSC-mapped craters were found in THEMIS), as well
as some quite bad ones (20% of HRSC craters recognized
from THEMIS). This was solely due to lighting conditions
in the area in question. Thus we decided that we use only
HRSC-derived crater calculations to keep the numbers
comparable to those of crater A, even though the images
cover some floor units only partly. 443 smaller craters were
found in the HRSC-covered areas of crater B, with 12 of
them being larger than one km.
4.3. Floor Units: Evidence of the Past Environments
4.3.1. Amazonian Dune Materials
[28] The youngest unit in both crater floors consists of
Amazonian dune materials (Amd, Figures 3b and 9a), which
are superposed by only a few very small (d < 50 m) very
9 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
fresh craters. The unit Amd is composed of small and sparse
dark dunes or megaripples with wavelengths of 10– 20 m,
which are visible in almost all of the crater-floor depressions, superposing uniformly all other units. This unit is
also characterized by a very dark albedo in the THEMIS
nighttime IR images (arrows in Figure 8). However, as this
unit can only be accurately distinguished from other units
from scattered and small MOC NA images, its accuracy in
the geological map is not as good as for the other units.
4.3.2. Smooth Materials
[29] Being Early Hesperian or Late Noachian in age, the
smooth material of HNms (Figure 3b) forms the second
youngest unit in both craters. It creates 45– 70 km wide
donut-shaped rings just inside both craters’ inner walls, with
a topography, which slopes very steeply close to the walls
but levels quickly down to a gentle slope toward the crater
center. The unit material morphology is generally smooth
and featureless, with some scattered craters (Figure 9b).
However, it does reveal some minor larger-scale textures
such as morphology indications of underlying ridges (see
E12S18
section 4.3.6) and buried channels. In THEMIS night-IR
images (Figure 8) the HNms unit appears varied, with bright
regions usually associated with the aforementioned, partly
or almost exhumed underlying features and darker color
relating to smoother areas. While the smooth unit is topographically relatively uniform in crater A, with elevations of
1200– 1250 m below the surrounding plains, it is more
diverse in crater B. There, the donut ring elevation is 450–
550 m below the plains, except for the northeastern parts.
That section is occupied by a tongue-shaped rise, 50– 200 m
above the average unit elevation, protruding from the crater
wall and penetrating into the etched unit (section 4.3.3). In
addition, deposits of darker albedo materials, as well as
several fluvial channels occupy the eastern and northwestern sections of the HNms unit in crater B.
4.3.3. Etched Material
[30] The etched material unit (HNme, Figures 3b and 9c)
is located on the northeastern floor of crater B, inside the
HNms unit area and at the end and sides of the aforementioned protruding rise. It is characterized by 50 –100 m deep
etched pits with clear layering visible on their walls. The
fluvial channels and darker material deposits on the smooth
unit extend well onto the HNme unit floor.
4.3.4. Cratered Materials
[31] The cratered surface material (Nmc, Figures 3b, 9d
and 9e) occupies the 35 35 km central plateau massif of
crater A. The massif raises high at the crater center with
steep, up to 60-degree slopes. Mesas similar in topography
to the massif are found at the edges of the honeycomb unit
(section 4.3.6) in a semi-circular pattern (Figures 3 and 7).
In average, the massif unit is 150 – 200 m above the HNms
elevation. Numerous 100 to 9000 m diameter craters and
pits cover the entire plateau surface. The massif walls show
Figure 7. The topographic MOLA profiles across the
studied craters show that the individual material units occur
at distinct elevation levels. (a) Black lines show the
locations of the profiles in Figures 7b and 7c. (b) Profiles
from crater A; black line west-east from point A to point A0,
gray line north-south from B to B0. On both profiles, the
smooth material unit HNms occupies the regions closest to
the crater rim (seen best on the left side). Going toward the
crater center, it is replaced by the central massif and the
remnant peak ring (PR). Note the two-level topography of
the central massif, seen clearly in the E-W profile, and
represented by the two cratered subunits, Nmc2 (higher) and
Nmc1 (lower). The unit with the lowest elevation, the
honeycomb terrain Nmh, surrounds the massif in the north,
south, and east. Farther eastward, the black profile continues
into the rugged terrain, which is halved into the shallow slope
subunit Nmr2 and the lower lying subunit Nmr1. (c) Profiles
from crater B; gray line northwest-southeast from point C to
point C0 and black line northeast-southwest from D to D0. The
gray profile shows the typical slope created by the smooth
material unit HNms on the left. The black profile traverses
more shallowly along the anomalous high-rising tongue-like
formation before being cut by the etched unit HNme
depressions. Both profiles are afterward taken over by the
rugged shallow slope unit Nmr2 and, further, at the deepest
areas by the Nmr1 subunit. Note the small remnant peaks of
Nmr2 on the gray profile.
10 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
E12S18
Figure 8. THEMIS night-IR mosaic indicates the material differences [Christensen et al., 2003] on the
crater floors. In crater B the rugged units are easily distinguished from each other and the smooth unit, as
are the pits with dark dunes on their floors (arrows) in the honeycomb terrain of crater A. The mosaic was
compiled from THEMIS images I05459009, I05846005, I05921005, I06208011, I06595007, I06620008,
I06645011, I07756015, I08455009, I09441011, I10040010, and I10639007.
layers of very bright material underneath a darker surface
material visible in HRSC and MOC narrow angle images
covering the edges of the massif (Figure 9e). In places, this
bright material appears to extend into the honeycomb terrain
unit forming the basement material there (section 4.3.6).
The southwestern edge of the plateau slopes gently into a
depression 50 meters below the HNms unit elevation. Crater
statistics put the Nmc unit clearly into the Noachian period.
[32] The Nmc unit is divided into two subunits according
to topography, crater density and albedo. The brighter Nmc2
has more craters than the darker Nmc1, as can be seen from
their different size-frequency distribution curves (Figure 4a).
It is also 100 m higher in topography, occupying the SW, E
and N parts of the massif, with a remnant in the middle,
surrounded by the entirety of Nmc1. The contact between the
two subunits is obvious, but diffuse. It is noteworthy that the
central Nmc2 area appears a bit brighter and more degraded
than the rest of Nmc2. This is, however, at least partly due to
the abrupt steep change into Nmc2 coloring at that locale, and
on the other hand the diffuse change experienced on other
contacts from Nmc1 to Nmc2. The central remnant is most
probably the degrading bottom part of the Nmc1 deposit.
Nmc1 has been covered by Nmc2 and is now being exhumed,
hence its lower crater density.
4.3.5. Rugged Materials
[33] The region next to the east rim of crater A and the
area at the center of crater B are identified as rugged
materials (Nmr, Figure 3b). They are characteristically
50– 100 m lower in elevation than the lowest areas of the
smooth unit HNms. The Nmr unit is generally rugged in
appearance (Figure 9f), with a typical feature diameter of
50– 200 meters. In both craters, the unit resides in a local
basin with two distinct subunits: Nmr2 makes up the slope
downward from the HNms unit onto the basin floor, which
is occupied by Nmr1. The elevations of the subunits are 30–
100 m and 80– 130 m below the smooth unit, respectively.
The basin floor is rough and coarse, and clearly brighter
than the knobby, almost hummocky material of the slope.
The slope has also more recognizable impact craters than
the basin floor. The albedo difference is apparent also in
11 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
THEMIS night-IR images (Figure 8), where the bright Nmr2
appears to continue deep into the smooth and etched units,
although there is no morphologic evidence of that from
HRSC data. In crater B, small pieces of darker Nmr2
materials remain inside Nmr1, making the appearance of
the central rugged regions appear patchy and complex.
[34] The crater size-frequency curve of the Nmr unit
inside crater A converges with that of HNms around craters
of 300 – 800 m in size, suggesting that they would be
approximately of the same age. Nevertheless, as there are
E12S18
several observable highly eroded circular features, which
may indeed be, but cannot be unambiguously identified as
craters, the curve is probably slightly shifted to the younger
end of the diagram and the material is actually older than
HNms.
4.3.6. Honeycomb Terrain
[35] The most striking unit in crater A is the honeycomb
terrain Nmh (Figures 3b, 9g and 9h). This geomorphologic
unit, with its distinctive honeycomb- and filament-like
ridges and intervening pits, forms a crescent-shaped area
between the central massif and the Nmr unit. The Nmh unit
differs clearly from other material units in the crater and it
also appears to be quite unique compared to units in other
craters in the whole northeastern Hellas rim. This is the
reason why its morphology and material deserve a more
detailed description, given in the following paragraphs.
[36] The ridges are approximately at the same level or
sometimes slightly lower than the HNms unit elevation,
while the intervening pits range 400– 1200 m in diameter
and 50 –200 m in depth. The ridges separating the pits are
generally covered by a dark-looking material, which, at
least on some wider ridges, appears to be remnant of some
overlying material, perhaps similar to the cratered material
of the central plateau (unit Nmc). The ridges interact with
the polygonal impact crater structure (see section 5.5)
located on the northern side of the massif (Figures 3a and
10) by deforming its walls and even protruding through it.
[37] Many pit floors are covered by the Amd dunes, and
thus appear very dark in the THEMIS nighttime IR images
(Figure 8). The remaining floors reveal a material with
much higher albedo in both THEMIS and HRSC images.
This material appears to form the floors of the pits. The
Figure 9. Samples of the material and terrain units with
locations designated in Figure 3a. All images except
Figures 9a, 9e, and 9h are HRSC orbit 389 nadir images.
(a) Dune material Amd covers most of the pits and
depressions in both craters as seen in MOC image
M0202276. The location of this close-up is seen in
Figure 9c. (b) Smooth material unit HNms is generally flat
with only minor underlying topographical features protruding
through; note the ridge caused by a buried channel in the
middle of the image. (c) The etched material HNme resembles
the honeycomb terrain in crater A but has no narrow ridges or
different deposits in the pits. The white rectangle shows the
location of Figure 9a. (d) Cratered material Nmc covers the
entire central massif (lower left). Note the brighter layers on
the walls. (e) The bright layers continue from the massif walls
down to the surface of the lower material unit Nmh (upper
right). MOC image M0905009. (f) Rugged material is divided
into two subunits: Nmr2 (darker and hummocky, ‘‘1’’)
occupies the higher ground, while Nmr1 (brighter and rough,
‘‘2’’) covers the deepest areas of the crater. (g) The
honeycomb terrain Nmh exhibits ridges and intervening pits.
Some of the ridges (arrows) are filament-like. (h) A closer
view shows these dark more filament-like ridges and bright
material on the pit floors. MOC image M0703863. Arrows
point to layered deposits on the ridge walls. Figures 9i and 9j
are examples of the mesas representative as the remnant peak
ring structure: (i) from the northern large mesa and (j) from
southern edge of the honeycomb terrain.
12 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
Figure 10. Crater A harbors three 9 km craters with different characteristics. One has a polygonal shape
(crater p), the second has a fluidized ejecta field (crater f), and the third has a ballistic dry ejecta field
(crater b). The provided profiles show their topographies and image their characteristics. (a) The locations
of the profiles: white arrows correspond to gray profiles; black ones correspond to black profiles. (b) The
polygonal shape of this 9 km crater indicates tectonic deformation. Note the thick layers of bright
materials (black arrows) on the crater and the central massif walls. HRSC orbit 389 nadir image. The
close-up on the right (MOC NA image M1102707) shows ridges that may be dikes protruding through
the inner northeast walls of the crater (white arrows). (c) This is the crater with ballistic, but still traceable,
ejecta, located on top of the central massif. (d) The rampart ejecta was caused most likely by the target
properties; floor unit reveals the higher volatile content of the local subsurface.
13 of 21
E12S18
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
same material is also visible in a few places on the walls of
the ridges where overlying darker material has eroded away.
This brighter material is layered as can be seen from the
high-resolution MOC narrow-angle images covering parts
of the terrain unit (arrows in Figure 9h). As observed above,
this same material appears to extend in places also beneath
the cratered material (Nmc) unit of the central plateau.
Additionally, the bright material can be seen to create
100 – 200 m thick layers on the walls of the 500 m deep
polygonal crater just north of the central massif (arrows in
Figure 10b). An assumption, that these bright layers are
connected to each other beneath the covering darker material deposits, allows us to estimate the thickness of the
bright layer. Height measurement, from the lowest occurrence at the polygonal crater wall to the highest layer on the
central plateau walls, reveals that the bright deposit thickness would be about 300 m. However, the extent of the
actual vertical continuity of the bright layer is unknown.
[38] The Nmh unit morphology is most distinct in the
southeastern part of the unit area. The northern parts are
more degraded; there the unit is superposed by the HNms
material from the north and by the Nmc2 type surface from
the west. Ridges of Nmh continue under both of these units,
and must thus be stratigraphically older than both of them.
However, its crater density curve shows that its surface is
one of the youngest on the whole crater floor. This indicates
that in the past the unit was covered by overlying materials
and was then exhumed.
5. Evolution History of the Siblings
5.1. Setting: Crater Formation
[39] The craters are of slightly different ages, inferred by
their degradation stages. The rim of crater A is considerably
higher and better-preserved, covering the entire circumference, while the rim of crater B is highly eroded and in
places not preserved at all. Additionally, there is a large area
of the crater B inner wall, just next to crater A, the slope of
which is shallower than the usual wall of the crater. This,
and the remnants of the deposits on the crater floor next to
this shallow slope indicate that the crater’s eastern wall has
undergone a significant mass wasting episode, with some
remnants of the apron created by the landslide event still
visible. The mass wasting may have been caused by A
impact, as it is situated just next to the area.
[40] The age differences we have determined are relative,
not absolute. We do know that the two impact events took
place in Noachian [Schaber, 1977; Greeley and Guest,
1987; Leonard and Tanaka, 2001], most probably during
or shortly after the heavy bombardment. Though temporally
not very significantly separate (geologically speaking; maybe
thousands to millions of years, separation is not known), the
target material properties may well have changed between the
two impacts. For example, Laskar et al. [2002] and
Kreslavsky and Head [2004] have shown there to be
large changes in the obliquity of Mars’ axis and subsequent insolation within relatively short time periods (i.e.,
millions of yrs). Together with a thicker atmosphere early
in the planet’s history [e.g., Carr, 1999] this would easily
enable climate conditions where water would be transported from one place to another [see, e.g., Craddock and
Howard, 2002].
E12S18
[41] The tongue-shaped high terrain in the northeastern
edge of crater B next to HNme may be the result of ejecta
from crater A and/or mass wasting material deposits originating in the crater wall. The origin of the tongue is
unclear, as the surface has been covered by later deposits
(HNms). It may also be just a thick deposit of HNms
material. However, it is impossible to determine its origin
with current data.
[42] The basement of the main central massif in crater A
may be the result of a massive outburst of impact melt
during the crater formation, or an unusually large central
peak structure. This is, however, speculation.
5.2. A Peak Ring?
[43] The outline shape of the central plateau inside crater
A and the degraded mesas surrounding the honeycomb
terrain suggest the presence of a large and highly degraded
ring structure at the center of the crater. On Mars, the
average transition from a central peak crater to a proto
basin with a peak ring (i.e., a central peak with an additional
encircling ring) [Head, 1978; Wood, 1980; Boyce, 1982]
takes place at crater diameters of 90 km [Garvin et al.,
1999]. The peak ring is an uplift of fractured target material,
similar to the central peak [Melosh, 1989]. The crater (D)
and peak ring (Dpr) diameters have been shown to follow
the rule
Dpr ¼ 0:54 D 8:19 km;
ð3Þ
where all parameters are in km [Wood, 1980]. When
calculating the estimated peak ring diameter from formula
(3) for crater A (DA = 95 km), we get Dpr = 43.1 km, which
coincides almost perfectly with the observed ring structure
(DApr = 45 km). Thus we are able to concur with Barlow’s
[2000, 2003] catalogue and interpret this feature to indeed
be a peak ring.
[44] The observed ring structure is highly deformed,
following the general erosion pattern of the crater, with
more of its structure intact in the west and north, and less in
the south and east. The remnant ring consists of several
mesas and not a continuous ring. This is either a feature of
the original structure, or an acquired characteristic. The
former would require that the peak ring was only partial to
begin with, or that the mesas we see now are only the
highest peaks of the ring structure. This may very well be
the case; As is calculated in Table 1, crater A was probably
filled with thick deposits, and sediments may accordingly
mostly or at least partly overlay the ring and prevent it from
being observed. An alternative explanation is that the peak
ring was modified, i.e., partly destroyed during progressive
crater evolution. This most probably happened by erosion
processes. The original material of the ring comes deep
within the crust, with most probably igneous (or metamorphic?) composition, and is thus expected to endure erosion
quite well. However, the material has also been highly
fractured by the impact event [Melosh, 1989], which makes
it much more friable and susceptible to erosion than, e.g.,
surrounding plains, if they would consist of the same
materials. Unfortunately, the floor of the crater is covered
by sedimentary materials, so none of the given possibilities
on the peak ring evolution can be confirmed.
[45] The existence of the peak ring implies that the
evolution of this particular crater has been unusual in the
14 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
region, since the other roughly same size craters do not
exhibit this feature. One possibility for the difference is that
the crater formation was different in the case of crater A,
and thus it would be the only one in the region having that
feature. In this case, the reason for this uniqueness would
most probably be the difference in target material properties
at the time of the impact event. The impact may have
occurred at a time when the soil was infested with large
amounts of volatiles. Alternatively, other craters in the
region also have the peak ring, but they are not visible.
This would be due to either (1) larger amounts of deposits in
other craters, (2) a smaller rate of exhumation in the other
craters, or (3) larger amounts of erosion in the other craters,
resulting in the destruction of their peak rings. We favor
alternatives 1 and 2 because they seem plausible.
5.3. Burial: Deposits, Volatiles, and Intrusions
[46] After their formations, the two craters were to some
extent filled by sedimentary deposits, creating several layers
of materials on the original floors. The layers seen, e.g., in
the central plateau walls and in most of the floor units as well
as in the multitude of the floor unit types, display clearly that
the deposition was the result of several consequent episodes
of sedimentation rather than of one single event. The
deposition sequence was neither continuous nor homogenous, and it worked side by side with erosional processes.
The deposits were at least once partly eroded away (see
section 5.4) and subjected to additional episodes of deposition/erosion, ultimately resulting in the units we see today.
[47] In addition to the maximum sediment load indicated
in Table 1, we can also estimate the minimum thicknesses
and volumes of the sedimentary layers we see today, i.e., the
amount of previously deposited materials being eroded
away. To do this, we have to assume that the deposits were
planar and covered the entire crater and that now we do not
observe the original crater floor but only the deposits
superposed on it. Thus the minimum sediment load for
crater A, measured from the Nmh pit floors to the top of the
Nmc2 unit, was 400 m thick and 6500 km3 in volume. For
crater B, the same value, measured from bottom of Nmr1 to
the top of the HNms unit, were 300 m and 4500 km3,
respectively.
[48] The deposition processes are not known, but most
probable candidates are layers of lavas, pyroclastics and
aeolian driven materials. The source regions for all of these
are indistinct. The pyroclasts may originate from the two
highland volcanoes (Hadriaca and Tyrrhena Paterae) 500 –
1000 km to the east, both of which are attributed to have
experienced mostly explosive volcanism [e.g., Greeley and
Spudis, 1978; Crown and Greeley, 1993; Leonard and
Tanaka, 2001, and references therein]. There are no indications of any volcanic vents in the immediate surroundings
of the studied area. However, the material of the cratered
highlands has been suggested to be of basaltic/andesitic
origin [e.g., McSween et al., 2003; Mustard et al., 2005, and
references therein]. This hints that there may have been lava
sources available in the region early in the planet’s evolution, perhaps in a similar manner as for the old volcanic
provinces of Hesperia and Syria Planae.
[49] Some sedimentary deposits may be of fluvial origin.
The rim of crater B is cut by several fluvial channels, most
prominently in the north and east (Figure 3), revealing
E12S18
fluvial processes as the prime sedimentation suspect. Crater
A rim is relatively intact, with only a couple small channels
leading to the crater floor. Their volume is not nearly
enough to explain the thick layers of sediments. Thus there
is no undisputable evidence of fluvial, lacustrine or glacial
activity in that crater. Certain floor unit morphologies (Nmh,
HNms; see below) imply, however, that at least some of the
sediments were rich in volatiles, in particular inside crater
A. This may be explainable by the erosion experienced by
the sediments later on: the fluvial channels could easily
have been eroded away, as the underlying deposits were
being exhumed.
[50] Several studies support the idea of volatile-rich
sedimentary deposits filling crater floors on Mars [e.g.,
Boyce et al., 2003, 2004b]. Like the craters studied here,
the craters Terby and Millochau have been recognized as
sinks of large amounts of local deposits [Mest and Crown,
2004, 2005; Ansan and Mangold, 2004; Ansan et al., 2005]
dating from Noachian to the present-day [Wilson and
Howard, 2005].
[51] The idea of local volatile-rich sediments is further
supported by the morphologies of two relatively pristine
9 km C3 craters, located in the southern area of crater A floor
(Figures 3 and 10). The topographies of these craters indicate
that they have excavated the surface approximately to the
same depth. The one created on the Nmc unit has a dry
ballistic ejecta field, while the other on the unit HNms is
clearly a rampart crater, revealing the volatile-rich nature of
the target floor subsurface. The craters are roughly of the same
age, since their rims are approximately of the same height, and
they both exhibit traceable ejecta fields. The ejecta fields
show that either the smooth floor unit (HNms) or the deposits
under it have been richer in volatiles than the massif (the Nmc
and underlying units). This may further support the idea of the
massif basement being either a peak ring or injection of
impact melt.
[52] It is seen from unit embayment that the HNms unit is
a deposit which has been put in place after the erosion of
Nmr and Nmc. In the NW part of the Nmc unit the contact
indicates that the smooth unit overlies the cratered unit. The
surface of Nmc is older than that of HNms (lots of fresh and
degraded craters versus almost no craters; this applies even
whether the craters are secondaries or regular impact craters). The only place where Nmc appears to overlie HNms is
the depression on the west side of crater A floor. However,
this depression is rather anomalous, and it is not even
absolutely clear if its floor material is HNms. This 6 8 km depression has straight walls and a flat floor. We
interpret it to be a small plate on the crater floor, which has
undergone a 50-meter movement downward (for more on
tectonism, see section 5.5). This movement has caused the
massif wall to the east to collapse down and create the slope
from the depression up to the Nmc unit. The contact
between the HNms and Nmc units in the location is diffuse,
and thus no stratigraphic relation can be determined there.
[53] The honeycomb terrain Nmh that is observed in
crater A is a complex unit. In order to discuss plausible
hypotheses concerning to its formation, we have to take into
account the following alterations, at least.
[54] 1. The ridges may represent large ancient lithified
dunes, which have been buried under later sediments.
However, this scenario fails to explain the straight and more
15 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
filament-like ridges (Figures 9g and 9h), which have additional hints of tectonic origin (see section 5.5), and may in
fact be erosion-resistant dikes.
[55] 2. Tectonic origin of the ridges is likely as an
additional process creating some of the features, but not
all. Because some ridges are linear or slightly arcuate (see,
e.g., Figure 9h), they could be of tectonic origin, perhaps
steeply sloped synclines. However, we do not favor this
interpretation to be the main cause for the ridges, because
large-scale tectonic deformation required for a closely
packed set of features is not seen anywhere else in the crater
or in its surroundings. Additionally, some ridges are layered.
[56] 3. The ridges may result from polygonal cracking of
a surface layer, which is seen on Mars in various places,
sizes and extents [e.g., Lucchitta, 1984; Seibert and Kargel,
2001, and the references therein]. Being abundant in the
Noachian environment [cf. Segura et al., 2002], volatiles
probably existed also in the early deposits in the crater and
were gathered into the places where the Nmh unit now
resides. This may have been facilitated by the existence of
the peak ring acting as a circular dam and keeping the
volatile-rich materials inside. The area inside the peak ring
may have been slightly deeper, due to, e.g., now-buried
impact craters in that region. After the deposition, due to
changes in the local temperature (e.g., insolation or climate
change), these sediments froze or dried, making the surface
crack and creating polygonal sets of fractures. To transform
fractures into ridges, we must have a secondary process
emplacing more erosion-resistant material into the fissures.
This may have been created in three ways: with the help of
igneous or clastic dikes, or by secondary mineralization.
[57] 4. Igneous dike intrusions may have been the process
to fill the fractures. Since volcanic dikes are able to reach
great distances from their sources [e.g., Ernst et al., 1995]
and impacts are able to cause such large fissure systems for
dikes to propagate into [Schultz, 1976; Schultz and Orphal,
1978; Wichman and Schultz, 1995], it is possible, that the
ridges were formed by dike injections from the subsurface,
such as from a pluton. The nearest volcanic centers (e.g.,
Hadriaca Patera some 600 km to the E-SE) could also
have acted as sources for this material. The igneous origin
of the ridges seems, however, unlikely since we do not
observe any similar features in other nearby craters or
elsewhere in the region, and the cause of the visible layering
in some of the ridges would remain unsolved.
[58] 5. The formation of clastic dikes is the easiest way to
explain the features seen on the Nmh terrain. As the
fractures opened, further sedimentation caused them to be
intruded and filled by harder material. Alternatively, the
sediments underlying the fractured layer were forced upward, for example due to the weight of the overlaying mass.
Whatever was the method in the fracture-filling event, this
injected material was consequently cemented and formed
dikes of harder material.
[59] 6. Secondary mineralization is a third viable option
for filling the cracks with more erosion-resistant materials
(J. Plescia, personal communication, 2005). This mineralization may well have happened due to the presence of water.
5.4. Exhuming: Erosion and Revelation of the Puzzle
[60] The sediments emplaced into the craters as well as
the fundamental characteristics of the craters (e.g., rims or
E12S18
the peak ring) were later at least partly eroded [e.g.,
Craddock et al., 1997; Boyce et al., 2004a], removing the
top sediments, exhuming the underlying older units and
exposing them to further sedimentation and erosion. This
also explains the upturned crater size-distribution curves
(Figure 4, discussed further in section 5.6).
[61] In crater A, the western side occupies a much higher
terrain in general than the (south)eastern side. The central
massif extends much further into the west than to the east.
The remnant peak ring also degrades in that region, though
there are still several large remnant mesas in place (see
Figure 9) as testament of the previously circular form
(Figure 3b). The sedimentary layers, e.g., Nmr and Nmh,
are exhumed deepest in the SE. The reason for all of this is
either unequal material distribution to begin with, of which
there is no evidence of, or more probably a prevailing
erosion scheme. The eastern region of the crater is much
more erosion-prone.
[62] The crater has apparently lost a lot of deposited
materials, judging from the multitude of the eroded surfaces. The erosion method may have been partly fluvial,
although no actual fluvial features are seen anymore.
However, having only fluvial erosion in crater A would be
counter-productive, since there are no outlets; material
would just be deposited at the lowest elevation inside the
crater, i.e., at the end or bottom of a river or a lake, and no
material (except water) could actually escape. Thus we
suggest the main method of material removal from the
crater to be likely to have been aeolian, acting over a very
long period of time. Additionally, if one dominating wind
direction is invoked, the east-west asymmetry in eroded
materials can easily be created. The same erosion scheme
applies for crater B, with the exception that in this case it is
the southern area which is much more eroded. Earlier
studies by Forsberg-Taylor et al. [2004] on Sinus Sabaeus
craters have shown that the general erosion method for the
crater rims in that region has been mainly fluvial.
[63] The differences in material durability governs the
resistance to erosion. The infilling sedimentary materials are
generally easily erodable. However, our proposed evolution
scheme calls that at least some of the sediments have been
volatile-rich, thus perhaps at times cementing in place and
being rather congealed and erosion-resistant during the
water-rich epoch. Only after the binding water is removed
from the sediments, they become more friable. Additionally,
more durable materials, such as the peak ring, should endure
more erosion.
[64] The rugged unit Nmr (Figure 3b) represents well the
surficial style of erosion, where an active process, most
probably wind, has removed the most brittle material from
the surface, revealing the underlying material. The Nmr
appearance in craters A and B resembles very much the
pitted unit in crater Millochau (designated ‘‘mp’’) [Mest and
Crown, 2004, 2005]. The similarity suggests a similar, if not
the same deposition and erosion processes for all the three
craters, at least for this particular unit. Although we mapped
the unit Nmr to be of Noachian age, a point should be made:
it is unknown if it is of Noachian or of Hesperian age. Its
surface has been eroded so much that no dating of its
surface features (e.g., craters) can accurately be done and
be called representative of the geological unit. Stratigraphically and topographically Nmr superposes Nmh, and is
16 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
further superposed by HNms. Nmr has no contact with
Nmc, and there is no direct indication that Nmr would
continue under the Nmc unit, or that Nmc has ever occupied
the area of Nmr. Parts of Nmc may be features of the crater
structure, i.e., a central peak and/or impact melt. This would
allow Nmr to be deposited at a later date to the area
surrounding Nmc, possibly even during the Hesperian
period. However, we suggest that Nmr does indeed underlie
Nmc. The Nmr deposit has a vast lateral extent, as it is
observed in many places inside crater A and additionally in
craters B and Millochau. This makes it very plausible that it
may underlie Nmc, if it can. At this point, the nature of Nmc
is the key issue: if it is mainly a construct of sedimentary
deposits, as we suppose, then it most probably also superposes Nmr. This would put Nmr clearly in the Noachian
period.
[65] The evolution of the honeycomb terrain Nmh unit,
after the sedimentation, cracking and dike intrusions discussed above, was dominated by erosion, though some
tectonics was also involved (section 5.5). As the remaining
volatile-rich material sublimated, or the material was otherwise eroded, karst-like depressions were created between
the hardened ridge materials. This happened due to either
aeolian removal, collapse or sinking of the surface material.
In the case of the similar unit in crater B, HNme, there are
clear fluvial channels running into the formation, but it is
unclear whether this happened only after or also during the
pit formation. The etched units in both crater B and crater
Millochau [Mest and Crown, 2004, 2005] were probably
formed in similar manner to the honeycomb unit, i.e., by the
collapse of the topsoil due to release of underground
volatiles, although in a more localized way and without as
vast a ridge-forming process.
[66] The surfaces of the Nmh and HNme pits appear
pristine, i.e., harbor only few impact craters. This is most
probably due to their highly varying topography, steep
slopes and consequential effective destruction of later induced small impact craters by mass wasting processes. The
latest episode of activity in the craters has been the minor
(re)deposition of materials by mass wasting and aeolian
processes. The floors of many of the pits are, like many
other small depressions in the craters, covered with small
dark dunes (unit Amd). Furthermore, judging from color
and morphology variations in HRSC images from the
eastern crater A, part of the dark and bluish Nmc1 material
(Figure 3a) from the top of the central massif has been
transported northeastward by wind. This material has covered the floor of the polygonal crater, as well as somewhat
softened the Nmh ridges and brought bluish color onto the
surfaces Nmh of HNms in crater A. However, this deposit
cannot be very thick, as it is not observed in THEMIS nightIR images.
5.5. Guide to Erosion: Tectonism
[67] Local and regional stresses and zones of weakness
are the driving forces of tectonism. An impact causes large
scale fracturing and movements in the target rock beneath
the forming crater [Melosh, 1989]. The fracturing may
induce an uprising of local volcanism, which may cause
further instabilities on the crater floor, e.g., add weight and
cause slabs of crater floor material to move, tilt and turn
[Schultz, 1976; Schultz and Orphal, 1978; Wichman and
E12S18
Figure 11. Many straight lines of walls or features inside
the crater indicate tectonic influence on the erosion and
other processes. Here we show some examples. White lines:
linear ridges, which often traverse in parallel with each
other. White dot lines: Ambiguous straight lines, often
caused by, e.g., ridge continuations beneath the topmost
layers. Black lines with a cross-line: linear or slightly
arcuate scarps. The circles represent the outlines of the
studied craters, and the dashed line circle represents the
remnant peak ring. Most of the lines on both crater floors
have directions generally concentric or radial to the Hellas
impact basin (to the direction of the arrow pointing down).
The old wrinkle ridge formations (lines with dots) also
follow this pattern.
Schultz, 1995]. In the highlands of Mars, there are additional regional compressional stresses [Chicarro et al., 1985]
and zones of weakness [Öhman et al., 2005a, 2005b],
indicating that the large impact basins (Isidis, Argyre, Hellas,
etc) basins have controlled the regional tectonism in the
region for very long periods.
[68] These tectonic stresses have supervised much of the
evolution of the craters under study. The first markings of
such forces are the ancient compressional wrinkle ridges
extending onto the southern floor of crater B (Figure 6d).
On the northern floor of crater A is another tectonically
modified feature; a 9-km polygonal depression, which is an
old, highly modified impact crater (Figure 10). Polygonal
craters occur in great numbers on the highlands of Mars,
and have been shown to indicate fracture fields within the
Hellas and Argyre regions [Öhman et al., 2005a, 2005b].
The weakness zone direction patterns in the Hellas region
are generally radial and/or concentric to the basin. This
applies also to this particular polygonal crater: its longest
sides point toward Hellas. Similarly, many of the linear
features in crater A, which can be interpreted to be of
tectonic origin or whose formation is at least influenced by
tectonism (such as linear or slightly arcuate ridges and linear
scarps) also follow this pattern (Figure 11). This is in
accordance to our earlier works on crater floor collapse
feature directions on the western side of Hellas [Korteniemi
et al., 2003], and implies that the Hellas-dominated weakness zones operate also inside a given younger crater,
17 of 21
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
E12S18
Figure 12
E12S18
18 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
dominating over the craters’ own fracture system. Interestingly, in this case the Hellas-concentric direction coincides
(within 5 degrees) with the direction toward Hadriaca
Patera, perhaps giving some support to the igneous dike
hypothesis of the Nmh ridge formation.
[69] The 9 km polygonal crater in Figure 10 is additionally a good stratigraphic marker. It has clearly been created
in the late stages or after the deposition on the central
massif; otherwise it would have been obscured by the
deposition. Its shape may or may not have been polygonal
to begin with, but additional later tectonic movements in the
Nmh terrain have radically deformed it. Clear compressional
and folding movements deform the crater walls and rim,
some features can be seen even protruding through the
crater wall (arrows in Figure 10b). This indicates that
tectonism played a part at least in the later stages of Nmh
evolution, perhaps creating fractures and/or routes for
igneous injections.
5.6. Crater Counting Versus Stratigraphy Revisited
[70] As stated in section 2.2, the usage of crater counting
for dating of units can at least in this case be somewhat
misleading. We try to illustrate this problem here, giving an
example from crater A. The stratigraphic construct of its
floor units is quite straightforward, if one assumes that all
units have been deposited on the original surface at successive occasions, uniformly to all around the crater. It is seen
from the topography (Figure 7) that the honeycomb terrain
(Nmh) occupies the lowest level and should thus be the
oldest unit, overlain with the rugged unit (Nmr), which is
further superposed by the cratered unit (Nmc), the latest
being the youngest of the aforementioned. Now, the smooth
unit (HNms) is topographically clearly in between Nmr and
Nmc units, and would thus appear to be between them also
temporally. This may very well be the case for the material
unit underlying HNms, but it has actually neither been
observed nor mapped. Instead, the HNms material on the
surface has been emplaced after the erosion of both Nmc
and Nmr, as it embays both those other units, best seen in
the northern part of crater A. Some Nmr unit surface spots
are visible here and there under the northernmost parts of
HNms, indicating the larger extent of this unit underneath
the superposing layer. Generally there is only a small
amount of erosion features on the smooth HNms unit
surface, making it the youngest major unit in the crater.
[71] Looking at the cratering records of the four units
(Nmh, Nmc, Nmr and HNms), one would come to a totally
different conclusion, as the given size-distribution curves
(Figure 4) show: the Nmr and Nmh curves appear to be
lower (i.e., younger) than that of the HNms. However, if
one takes the error bars into account, from 500+ meter
craters onward the curves rather coincide with than differ
from each other. And, if the values for small craters (<1 km)
are totally disregarded, as should be done when dating
surfaces with craters accurately [e.g., McEwen et al.,
2005; Hartmann, 2005; Plescia, 2005], the only curve
E12S18
standing out is that of the topmost cratered unit Nmc2,
within the error bars.
[72] The surfaces of only the HNms and Nmc units are
construed to be characteristic for those specific geologic
units. If this interpretation is correct, they can be used to
give some insight to the unit ages. However, in our opinion
they do not provide enough material for absolute age
determination, for the reasons given earlier (e.g., small
surface areas, amount of 1+ km craters, etc). All the units
other than Nmc and HNms (i.e., Nmh, Nmr, HNme) have
been so much eroded, superposed by other deposits and/or
exhumed that their crater populations are not representative
of the age of the actual geological unit. In this study the
units are put into their stratigraphic places mainly by
determining their relations to surrounding units.
6. Conclusions
[73] 1. The chronologic evolutional history of the studied
crater pair has had several phases of varying processes (see
Figure 12). They were controlled by the initial form of the
impact structures, and afterward by the sediments they were
filled with. Additionally, regional tectonism played a major
part in their development. Several material units, each with
their own elevation level of occurrence have been identified.
The craters were formed on the highly modified, Hellasdominated Noachian highlands, patterned with numerous
wrinkle ridges and fluvial channels. The total sediment
thickness is 1400 m for the eastern crater A and 2100 m
for the western crater B. The minimum thickness of the
sediments, which were estimated to have been eroded, is
400 and 300 m, respectively.
[74] 2. Although no doubt geologically complex, the
older crater B is the simpler of the two. After the impact
event, regional tectonic forces continued to compress the
region, creating the wrinkle ridges on the crater floor.
Stemming from the somewhat later crater A impact event,
crater B was partly filled with ejecta and experienced severe
mass wasting on its eastern wall.
[75] Crater B was thereafter dominated by overlapping
periods of sedimentation and erosion. First the floor was
covered with the now-underlying rugged unit Nmr1, and
later by another, Nmr2. Fluvial channels started to cut into
the crater walls and floor, continuing (though with diminishing capacity) along with the deposition of smooth material HNms unit. Later on, erosion took the upper hand and
dug deep (150 m) into the central areas of the crater,
revealing a cross-section of all three layers. A second center
of erosion, the location of etched unit HNme, was created
into the smooth material at the northeast. This area formed
near a high protruding tongue of smooth material, with
100 m deep eroded on its sides. The etched terrain also
acted as a sink for late fluvial and aeolian (Amd) deposits.
[76] 3. Probably due to the target material, the slightly
larger crater A has had a more complex geological history.
The impact event formed a peak-ring structure and perhaps
Figure 12. Stratigraphy of the units designated in Figure 2b for the studied region. The two craters were formed after the
formation of the Hellas basin, and after and during the formation and modification of the surrounding plains. The region has
later been subjected to several tectonic episodes, cratering fluvial deformation, and other forms of erosion. The crater floors
have several recognizable material units superposing each other and becoming exhumed under other layers due to later
erosion.
19 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
a large central peak. Afterward the crater was filled with
three different deposit types: volatile-rich, erodable and
more erosion-resistant materials.
[77] In crater A, the stratigraphically oldest observable
sediments are the bright materials visible on the polygonal
crater walls and on the floors of the Nmh pits. The same or
similar bright material also occupies the base of the central
massif. Later sedimentation included or retained more volatiles, which first gathered inside the peak ring, and then froze
or dried, creating polygonal fractures. This resulted in voids
for intrusions of more durable and hardwearing materials
from the following sedimentation process, which occupied
the fractures and cemented into them. The crater floor was
further superposed by several layers of materials, creating the
basis for the now-rugged subunits labeled as Nmr1 and Nmr2.
Continuing deposition grew the central massif with bright
sediments and the cratered material units (Nmc1 and Nmc2).
[78] At this time, the 9 km impact crater was emplaced
onto the floor on the northern side of the massif. It, like the
whole crater A floor, was modified and transformed by
erosion and influenced by tectonism. This deformation also
probably resulted in some of the endogenic dike injections
into the honeycomb terrain and the walls of the nowpolygonal crater.
[79] The crater floor was later slightly modified by fluvial
channels, and thereafter filled from the sides with the
smooth HNms unit sediments. Though erosion had been
an active force all along, it overpowered deposition and dug
deep into the sediments. This resulted in the exposure of the
rugged terrains to the east, the degradation of the central
massif, and, while digging deeper in the central areas,
exhuming of the erosion-resistant material ridges as testament of the ancient polygonal fractures in the Nmh unit.
[80] 4. The evolution of the craters was been dominated
by sedimentation and erosion. The existence of a peak ring
in the eastern crater resulted in the capture of volatiles, and
the creation of a very unusual landform, the honeycomb
terrain. The erosion process was dominated by zones of
weakness aligned with the Hellas basin.
[81] As similar collapse and erosion features inside craters are found mostly only in the vicinity of the Hellas basin
[Korteniemi, 2003; Korteniemi et al., 2003, 2005], this
raises a question on the nature of the special circumstances
occurring in the region. The type 3 collapses described in
section 3.3 may be due to material deposits found only in
the Hellas region. Alternatively, they are perhaps the result
of subsurface activity, which has not reached the surface.
Hellas region has volcanism on the southern and eastern rim
regions. If the volcanic activity was somewhat uniform
around the basin, one would expect some forms of endogenic heat to be released also in the west and northern rim
regions, i.e., in the areas where the pits are found. The pits
may thus be representations of, e.g., volatiles being released
form the ground by subsurface heating.
[82] The usage of the unprecedented MEX HRSC data
with other existing and future data sets improves the
possibilities in analyzing different landforms on Mars, and
identifying their origins.
[83] Acknowledgments. We thank A. Basilevsky, M. Ivanov, and
M. Kreslavsky for innovative discussions, new ideas, and constructive
criticism. We also gratefully acknowledge J. M. Boyce, the anonymous
E12S18
reviewer, and associate editor J. Plescia for their useful reviews and
improvement suggestions. The Mars Express flight team and the HRSC
DLR and FU-Berlin groups are recognized for obtaining the studied
images. We sincerely thank the HRSC Science Co-Investigator Team
(J. Albertz, A. T. Basilevsky, G. Bellucci, J.-P. Bibring, M. Buchroithner,
M. H. Carr, E. Dorrer, T. C. Duxbury, H. Ebner, B. H. Foing, R. Greeley,
E. Hauber, J. W. Head III, C. Heipke, H. Hoffmann, A. Inada, W.-H.
Ip, B. A. Ivanov, R. Jaumann, H. U. Keller, R. Kirk, K. Kraus, P. Kronberg,
R. Kuzmin, Y. Langevin, K. Lumme, W. Markiewicz, P. Masson,
H. Mayer, T. B. McCord, J.-P. Muller, J. B. Murray, F. M.
Neubauer, G. Neukum (PI), J. Oberst, G. G. Ori, M. Pätzold, P. Pinet,
R. Pischel, F. Poulet, J. Raitala, G. Schwarz, T. Spohn, and S. W. Squyres)
for planning the imaging sequences, and for continuing scientific input.
Additionally, the efforts made by the HRSC Photogrammetry Team in
processing the digital image data are recognized and highly appreciated.
The NRPIF we acknowledge for the use of working facilities.
References
Ansan, V., and N. Mangold (2004), Impact crater paleolakes in Hellas
and Thaumasia areas, Mars, in Second Conference on Early Mars
[CD-ROM], abstract 8006, Lunar and Planet. Inst., Houston, Tex.
Ansan, V., N. Mangold, A. Lucas, A. Gendrin, S. Le Mouélic, F. Poulet, J.-P.
Bibring and the OMEGA Co-Investigator Team (2005), Analysis of the
layered deposits in Terby crater (Hellas region, Mars) using multiple
datasets MOC, THEMIS and OMEGA/MEX data, Lunar Planet. Sci.
[CD-ROM], XXXVI, abstract 1324.
Barlow, N. G. (2000), Updates to the Catalog of Large Martian Impact
Craters, Lunar Planet. Sci. [CD-ROM], XXXI, abstract 1475.
Barlow, N. G. (2003), Revision of the Catalog of Large Martian Impact
Craters, in Sixth International Conference on Mars [CD-ROM], abstract
3073, Lunar and Planet. Inst., Houston, Tex.
Boyce, J. M. (1982), Peaks, peak-rings, and multiple ring spacing in impact
basins: Discontinuities in the function relationship, Lunar Planet. Sci.,
XIII, 61 – 62.
Boyce, J. M., P. J. Mouginis-Mark, J. B. Garvin, and H. Garbeil (2003),
Evidence for a thick mantle of volatile-rich materials in the Utopia basin,
Mars, based on crater depth/diameter measurements, Lunar Planet. Sci.
[CD-ROM], XXXIV, abstract 1967.
Boyce, J. M., P. Mouginis-Mark, and H. Garbeil (2004a), Predicted effects
of surface processes on Martian impact crater depth/diameter relationships, Lunar Planet. Sci. [CD-ROM], XXXV, abstract 1816.
Boyce, J. M., P. Mouginis-Mark, and H. Garbeil (2004b), Depth to diameter
relationships of craters in the high latitudes (70 – 80 deg) of Mars:
Implications for geologic history of those areas, Lunar Planet. Sci.
[CD-ROM], XXXV, abstract 1129.
Carr, M. H. (1996), Water on Mars, 229 pp., Oxford Univ. Press, New York.
Carr, M. H. (1999), Retention of an atmosphere on early Mars, J. Geophys.
Res., 104(E9), 21,897 – 21,909.
Carr, M. H., and Schaber G. G. (1977), Martian permafrost features,
J. Geophys. Res., 82, 4039 – 4054.
Carr, M. H., L. S. Crumpler, J. A. Cutts, R. Greeley, J. E. Guest, and
H. Masursky (1977), Martian impact craters and emplacement of ejecta by
surface flow, J. Geophys. Res., 82, 4055 – 4065.
Chicarro, A. F., P. H. Schultz, and P. Masson (1985), Global and regional
ridge patterns on Mars, Icarus, 63, 153 – 174.
Christensen, P. R., et al. (2003), Morphology and composition of the surface of Mars: Mars Odyssey THEMIS results, Science, 300(5628),
2056 – 2061.
Craddock, R. A., and A. D. Howard (2002), The case for rainfall on
a warm, wet early Mars, J. Geophys. Res., 107(E11), 5111, doi:10.1029/
2001JE001505.
Craddock, R. A., T. A. Maxwell, and A. D. Howard (1997), Crater morphometry and modification in the Sinus Sabaeus and Margaritifer Sinus
regions of Mars, J. Geophys. Res., 102(E6), 13,321 – 13,340.
Crown, D. A., and R. Greeley (1993), Volcanic geology of Hadriaca Patera
and the eastern Hellas region of Mars, J. Geophys. Res., 98(E2), 3431 –
3451.
Crown, D. A., K. H. Price, and R. Greeley (1992), Geologic evolution of
the east rim of the Hellas basin, Mars, Icarus, 100, 1 – 25.
Ernst, R. E., J. W. Head, E. Parfitt, E. Grosfils, and L. Wilson (1995), Giant
radiating dike swarms on Earth and Venus, Earth Sci. Rev., 39, 1 – 59.
Forsberg-Taylor, N. K., A. D. Howard, and R. A. Craddock (2004), Crater
degradation in the Martian highlands: Morphometric analysis of the
Sinus Sabaeus region and simulation modeling suggest fluvial processes,
J. Geophys. Res., 109, E05002, doi:10.1029/2004JE002242.
Garvin, J. B., S. E. H. Sakimoto, C. Schnetzler and J. J. Frawley (1999),
Global geometric properties of Martian impact craters: A preliminary
assessment using MOLA, in Fifth International Conference on Mars
[CD-ROM], abstract 6163, Lunar and Planet. Inst., Houston, Tex.
20 of 21
E12S18
KORTENIEMI ET AL.: TYRRHENA TERRA IMPACT CRATERS, MARS
Garvin, J. B., S. E. H. Sakimoto and J. J. Frawley (2003), Craters on Mars:
Global geometric properties from gridded MOLA topography, in Sixth
International Conference on Mars [CD-ROM], abstract 3277, Lunar and
Planet. Inst., Houston, Tex.
Greeley, R., and J. E. Guest (1987), Geological map of the Eastern
equatorial region of Mars, U.S. Geol. Surv. Misc. Invest. Ser., Map
I-1802.
Greeley, R., and P. D. Spudis (1978), Volcanism in the cratered terrain
hemisphere of Mars, Geophys. Res. Lett., 5, 453 – 455.
Hartmann, W. K. (2005), Adventures (arrrggghh!) in crater counting: Small
crater controversies, Lunar Planet. Sci. [CD-ROM], XXXVI, abstract
1427.
Hartmann, W. K., and G. Neukum (2001), Cratering chronology and the
evolution of Mars, Space Sci. Rev., 96(1 – 4), 165 – 194.
Head, J. W. (1978), Origin of central peaks and peak rings: Evidence from
peak-ring basins on Moon, Mars and Mercury, Lunar Planet. Sci., IX,
485 – 487.
Ivanov, M. A., J. Korteniemi, V.-P. Kostama, M. Aittola, J. Raitala,
M. Glamoclija, L. Marinangeli, and G. Neukum (2005), Major episodes of
the hydrologic history in the region of Hesperia Planum, Mars, J. Geophys.
Res., doi:10.1029/2005JE002420, in press.
Korteniemi, J. (2003), Collapses and depressions post-dating crater formation in Martian impact structures—Distribution and consequences, in
Third International Conference on Large Meteorite Impacts [CDROM], abstract 4091, Lunar and Planet. Inst., Houston, Tex.
Korteniemi, J., V.-P. Kostama, and J. Raitala (2003), Post-impact depressions on Martian impact crater floors: Preliminary results and a case study
of the greater Hellas region, in 38th Vernadsky-Brown Microsymposium
on Comparative Planetology [CD-ROM], abstract MS048, Vernadsky
Inst., Moscow.
Korteniemi, J., H. Lahtela, J. Raitala, G. Neukum, and the HRSC CoInvestigator Team (2005), Anomalous depressions on the circum-Hellas
crater floors as seen in the first year MEX HRSC images, Lunar Planet.
Sci. [CD-ROM], XXXVI, abstract 1669.
Kostama, V.-P., M. A. Kreslavsky, and J. W. Head (2003), Morphology of
the high-latitude mantle in northern plains on Mars, in Sixth International
Conference on Mars [CD-ROM], abstract 3011, Lunar and Planet. Inst.,
Houston, Tex.
Kreiter, T. J. (1960), Dating lunar surface features by using crater frequencies, Publ. Astron. Soc. Pac., 72, 393 – 398.
Kreslavsky, M., and J. W. Head (2004), Periods of active permafrost layer
formation in the recent geological history of Mars, Lunar Planet. Sci.
[CD-ROM], XXXV, abstract 1201.
Laskar, J., B. Levrard, and J. F. Mustard (2002), Orbital forcing of the
Martian polar layered deposits, Nature, 419, 375 – 377, doi:10.1038/
nature01066.
Leonard, G. J., and K. L. Tanaka (2001), Geologic map of the Hellas region
of Mars, U.S. Geol. Surv. Misc. Invest. Ser., Map I-2694.
Lucchitta, B. K. (1984), Small-scale polygons on Mars, in Reports in
Planetary Geology, NASA Tech. Memo., NASA-TM 86246, 205 – 208.
McEwen, A., B. Preblich, E. Turtle, D. Struder, N. Artemieva, M. Golombek,
M. Hurst, R. Kirk, and D. Burr (2005), Distant secondary craters and age
constraints on young Martian terrains, Lunar Planet. Sci. [CD-ROM],
XXXVI, abstract 2111.
McSween, H. Y., Jr., T. L. Grove, and M. B. Wyatt (2003), Constraints on
the composition and petrogenesis of the Martian crust, J. Geophys. Res.,
108(E12), 5135, doi:10.1029/2003JE002175.
Melosh, H. J. (1989), Impact Cratering, 245 pp., Oxford Univ. Press, New
York.
Mest, S. C., and D. A. Crown (2004), Geologic map of MTM-20272
and -25272 quadrangles, Tyrrhena Terra region of Mars, in Annual
Meeting of Planetary Geologic Mappers, U.S. Geol. Surv. Open File
Rep., 2004-1100.
Mest, S. C., and D. A. Crown (2005), Millochau crater, Mars: Infilling and
erosion of an ancient highland impact crater, Icarus, 175(2), 335 – 359.
Moore, J. M., and A. D. Howard (2005), Layered deposits and pitted terrain
in the circum Hellas region, Lunar Planet. Sci. [CD-ROM], XXXVI,
abstract 1512.
E12S18
Mustard, J. F., F. Poulet, A. Gendrin, N. Mangold, J.-P. Bibring, Y. Langevin,
B. Gondet, G. Belluci, F. Altieri, and the OMEGA Science Team
(2005), Compositional diversity of the Martian crust from OMEGA
observations, Lunar Planet. Sci. [CD-ROM], XXXVI, abstract 1316.
Neukum, G., and B. A. Ivanov (2001), Crater production function for Mars,
Lunar Planet. Sci. [CD-ROM], XXXII, abstract 1757.
Neukum, G., R. Jaumann, and the HRSC Co-Investigator and Experiment
Team (2004a), HRSC: The High Resolution Stereo Camera of Mars
Express, in Mars Express—The Scientific Payload, edited by A. Wilson,
Eur. Space Agency Spec. Publ., ESA-SP 1240, 17 – 35.
Neukum, G., et al. (2004b), Recent and episodic volcanic and glacial activity on Mars revealed by the High Resolution Stereo Camera, Nature,
432, 971 – 979.
Newsom, H. E. (2001), Central remnant craters on Mars—Localization of
hydrothermal alteration at the edge of crater floors?, Lunar Planet. Sci.
[CD-ROM], XXXII, abstract 1402.
Öhman, T., M. Aittola, V.-P. Kostama, and J. Raitala (2005a), The preliminary analysis of polygonal impact craters within greater Hellas region,
Mars, in Impact Tectonics, edited by C. Koeberl and H. Henkel, pp. 131 –
160, Springer, New York.
Öhman, T., M. Aittola, V.-P. Kostama, M. Hyvärinen, and J. Raitala
(2005b), Preliminary study of polygonal impact craters in Argyre region,
Mars, Lunar Planet. Sci. [CD-ROM], XXXVI, abstract 1731.
Plescia, J. (2005), Small-diameter Martian craters: Applicability for
chronology—or not, Lunar Planet. Sci. [CD-ROM], XXXVI, abstract 2171.
Raitala, J., M. Ivanov, M. Aittola, V.-P. Kostama, and J. Korteniemi (2004),
The history and nature of material in Hellas basin, Mars, Lunar Planet.
Sci. [CD-ROM], XXXV, abstract 1134.
Schaber, G. G. (1977), Geologic map of the Iapygia quadrangle of Mars,
U.S. Geol. Surv. Misc. Invest. Ser., Map I-1020.
Schultz, P. H. (1976), Floor-fractured lunar craters, Moon, 15, 241 – 273.
Schultz, P. H., and D. L. Orphal (1978), Floor-fractured craters on the Moon
and Mars, Meteoritics, 13, 622 – 625.
Segura, T. L., O. B. Toon, A. Colaprete, and K. Zahnle (2002), Environmental effects of large impacts on Mars, Science, 298, 1977 – 1980.
Seibert, N. M., and J. S. Kargel (2001), Small-scale Martian polygonal
terrain: Implications for liquid surface water, Geophys. Res. Lett.,
28(5), 899 – 902.
Strom, R. G., S. K. Croft, and N. G. Barlow (1992), The Martian impact
cratering record, in Mars, edited by H. H. Kieffer et al., pp. 383 – 423,
Univ. of Ariz. Press, Tucson.
Tanaka, K. L. (1986), The stratigraphy of Mars, Proc. Lunar Planet. Sci.
Conf. 17th, Part 1, J. Geophys. Res., 91, suppl., E139 – E158.
Tanaka, K. L., and M. P. Golombek (1989), Martian tension fractures and
the formation of grabens and collapse features at Valles Marineris, Proc.
Lunar Planet. Sci. Conf. 19th, 383 – 396.
Tanaka, K. L., and G. J. Leonard (1995), Geology and landscape evolution
of the Hellas region of Mars, J. Geophys. Res., 100(E3), 5407 – 5432.
Wichman, R. W., and P. H. Schultz (1995), Floor-fractured impact craters
on Venus: Implications for igneous crater modification and local magmatism, J. Geophys. Res., 100(E2), 3233 – 3244.
Wilson, S. A., and A. D. Howard (2005), Geomorphic and stratigraphic
analysis of layered deposits in Terby crater, Mars, Lunar Planet. Sci.
[CD-ROM], XXXVI, abstract 2060.
Wood, C. A. (1980), Martian double ring basins: New observations, Proc.
Lunar Planet. Sci. Conf. 11th, 2221 – 2241.
M. Aittola, J. Korteniemi, V.-P. Kostama, H. Lahtela, T. Öhman, J. Raitala,
and T. Törmänen, Astronomy Division, Department of Physical Sciences,
University of Oulu, P.O. Box 3000, FIN-90014 Oulu, Finland. (jarmo.
[email protected])
G. Neukum, Institut für Geologische Wissenschaften, Department of
Earth Sciences, Freie Universität Berlin, Malteserstrasse 74-100, Bldg. D,
D-12249 Berlin, Germany.
21 of 21