REMOTE SENSING OF SHALLOW-MARINE IMPACT CRATERS ON MARS
Except where reference is made to the work of others, the work described in this thesis is
my own or was done in collaboration with my advisory committee. This thesis does not
include proprietary or classified information.
______________________________________
Germari de Villiers
Certificate of Approval:
_______________________ _______________________
Luke J. Marzen, Co-Chair David T. King, Jr., Co-Chair
Associate Professor Professor
Geography Geology
_______________________ _______________________
Willis E. Hames George T. Flowers
Profesor Interim Dean
Geology Graduate School
REMOTE SENSING OF SHALLOW-MARINE IMPACT CRATERS ON MARS
Germari de Villiers
A Thesis
Submitted to
the Graduate Faculty of
Auburn University
in Partial Fulfillment of the
Requirements for the
Degree of
Master of Science
Auburn, Alabama
December 17, 2007
iii
REMOTE SENSING OF SHALLOW-MARINE IMPACT CRATERS ON MARS
Germari de Villiers
Permission is granted to Auburn University to make copies of this thesis at its discretion,
upon request of individuals or institutions at their expense. The author reserves all
publication rights.
______________________________
Signature of Author
______________________________
Date of Graduation
iv
VITAE
Germari de Villiers, daughter of Gerard and Gerda de Villiers, was born on 28
February, 1983, in Pretoria, South Africa, as the first of three daughters. She grew up in
Randburg, South Africa, and attended both Laerskool Unika and Ho?rskool Randburg.
She graduated from Reeltown High School, Alabama, in 2001 and matriculated from
Greenside High School, Johannesburg, in 2002. She graduated from the University of
Pretoria in 2005 with a Bachelor of Science degree in Geology and entered Auburn
University Graduate School in January 2006.
v
THESIS ABSTRACT
REMOTE SENSING OF SHALLOW-MARINE IMPACT CRATERS ON MARS
Germari de Villiers
Master of Science, December 17, 2007
(B.Sc., University of Pretoria, 2005)
161 Typed Pages
Directed by David T. King, Jr. and Luke J. Marzen
Impact craters are common on solid planetary bodies in our solar system and are
one of the most important physical features from which the surface history of these
planetary bodies can be deduced. Remote sensing is a crucial tool in planetary science
and is essential in the detailed study of impact structures on planetary surfaces.
Oceans have been proposed to have existed on Mars during its history, the
shorelines of which would coincide roughly with the crustal dichotomy that divides the
smooth, northern lowlands with the cratered, southern highlands. Arabia Terra is a
region on Mars that straddles the dichotomy and three proposed shorelines are located in
the area. If Mars had a large ocean during its early history, Arabia Terra would be a
vi
continental shelf area and hence an ideal location for the preservation of shallow-
marine impact craters.
Shallow-marine impact craters on Earth exhibit characteristic morphological
features. Due to the sub-marine formation and the influence of the water column, the
morphologies of these craters are distinctly different from that of craters formed on land.
Common attributes of marine impact craters include features of wet mass movement such
as gravity slumps and debris flows; radial gullies flowing into the crater depression;
resurge deposits and blocks of dislocated materials; a central peak terrace or peak ring
terrace; crater rim collapse or breaching of the crater wall; and subdued topography.
These features are visible from orbital imagery, and can thus be used to evaluate craters
on Mars for possible marine origin. This study designed a simple quantification system
that can be used to crudely judge and rank shallow-marine impact crater candidates based
on the features observed in previously proposed shallow-marine impact crater candidates
as well as features observed in terrestrial analogs. With the use of Mars Orbiter Laser
Altimeter topographic data and Mars Orbiter Camera and Thermal Emission Imaging
System imagery, the area bounded by 20? and 40? north as well as 20? west and 20? east is
explored for evidence of shallow-marine impact craters. Based on the quantification
system, 77 potential shallow-marine impact craters are found within Arabia Terra of
which nine exemplary candidates were ranked with total scores of 70% or more.
vii
ACKNOWLEDGEMENTS
The author would like to thank Dr. David King, Dr. Luke Marzen, and Dr. Willis
Hames, members of her thesis committee, for endless support and guidance in this
endeavor. The author is exceedingly grateful to Dr. David King, who took particular
effort to ensure that the time spent at Auburn University was a positive experience. The
author is grateful also to Dr. Luke Marzen for taking the initiative to apply for funding to
NASA?s Experimental Program to Stimulate Competitive Research (EPSCoR). Due to
his resourcefulness, this study was funded by the Alabama Space Grant Consortium
through NASA EPSCoR.
As far as inspiration goes, the author would like to acknowledge Jens Orm? who
made valuable suggestions and contributions, and who essentially planted the seed for
this study. The author would like to thank Trent Hare and ArunKumar Jayakeerthy for
assistance with image processing as well as Filippo Bianchi and Gerard de Villiers for
valuable comments and suggestions concerning software applications. The author
acknowledges the extensive use of Mars Orbiter Camera (MOC) images that are available
at http://www.msss.com/moc_gallery/ as well as THEMIS images that are available at
http://themis.asu.edu/mars-bin/webmap.pl.
The author would like to dedicate this thesis to her family who stood closely by
her through these challenging two years:
viii
?Pappa and Mamma, thank you for inspiring me to work and teaching me to
dream. Lorida and Marina, thank you for always cheering me up and making me laugh.
Filippo, thank you for loving me with all that you have and letting me be your crazy,
South African scientist. And finally, to my Heavenly Father, for showing me that all
things are possible through Him.?
ix
Journal style used: Meteoritics and Planetary Science
Computer software used: Adobe? Reader? 7.0
ESRI? ArcGIS? 9.2
ISI 2
Microsoft? Office Excel 2003
Microsoft? Office Power Point? 2003
Microsoft? Office Word 2003
GRIDVIEW
WinZip?
x
TABLE OF CONTENTS
LIST OF FIGURES .........................................................................................................xiii
LIST OF TABLES.........................................................................................................xviii
INTRODUCTION .............................................................................................................. 1
Impact Craters.............................................................................................................. 3
Distribution of craters........................................................................................... 3
Types of craters .................................................................................................... 4
Objectives .................................................................................................................. 13
Significance ............................................................................................................... 18
LITERATURE REVIEW ................................................................................................. 19
Water on Mars ........................................................................................................... 20
Oceans on Mars ......................................................................................................... 24
Large Noachian ocean and smaller Hesperian ocean......................................... 24
Smaller local lakes.............................................................................................. 31
Evidence for oceans on Mars ............................................................................. 31
Physical properties and characteristics of shallow-marine impact craters ................ 34
Depth-diameter ratios ......................................................................................... 37
Slopes ................................................................................................................. 39
Ejecta deposits.................................................................................................... 40
Wet Mass Movement (WMM) ........................................................................... 42
xi
Radial Gullies (RG)............................................................................................ 43
Resurge Deposits (RD)....................................................................................... 44
Central Terrace (CT) .......................................................................................... 44
Rim Collapse (RC) ............................................................................................. 45
Subdued Topography (ST) ................................................................................. 45
Examples of shallow-marine craters.......................................................................... 46
Chesapeake Bay.................................................................................................. 47
Chicxulub ........................................................................................................... 48
Lockne ................................................................................................................ 48
Mj?lnir................................................................................................................ 49
Wetumpka........................................................................................................... 49
Remote Sensing ......................................................................................................... 50
Summary.................................................................................................................... 54
METHODS ....................................................................................................................... 55
Background................................................................................................................ 56
Mars Global Surveyor ........................................................................................ 57
Mars Odyssey ..................................................................................................... 59
Data Acquisition ........................................................................................................ 60
MOLA data......................................................................................................... 60
MOC and THEMIS data..................................................................................... 64
Data Analysis............................................................................................................. 67
Data Interpretation..................................................................................................... 69
xii
RESULTS ......................................................................................................................... 76
Orm? et al.?s candidate crater results ........................................................................ 78
Results from Set A..................................................................................................... 80
Results from Set B..................................................................................................... 83
Exemplary candidates................................................................................................ 87
INTERPRETATIONS ...................................................................................................... 91
Type I candidates....................................................................................................... 93
Crater D .............................................................................................................. 93
Crater 24 ............................................................................................................. 97
Crater 66 ........................................................................................................... 101
Type II candidates.................................................................................................... 105
Crater 6 ............................................................................................................. 105
Crater 17 ........................................................................................................... 108
Crater 45 ........................................................................................................... 112
Crater 54 ........................................................................................................... 116
Crater 55 ........................................................................................................... 120
Type III candidates .................................................................................................. 124
Crater 58 ........................................................................................................... 124
Summary.................................................................................................................. 128
CONCLUSIONS............................................................................................................. 130
REFERENCES ............................................................................................................... 133
xiii
LIST OF FIGURES
Fig. 1. Morphology of a complex crater (b) differs substantially from that of a simple
crater (a). Modified from French (1998)............................................................................ 6
Fig. 2. Different layers present in the target (A-F) as well as the ejecta blanket (G) of a
complex crater. Also shown are a terrace (t) and a block of rim material (b) both
products of slumping. From Pike (1980). .......................................................................... 6
Fig. 3. Cross-section of the inverted sombrero morphology observed at Chesapeake Bay
impact crater. The section runs from the western rim to the central peak due east. The
wide annular trough and the slumped blocks where the rim collapsed are shown. From
Horton et al. (2006)............................................................................................................. 9
Fig. 4. Mars topographic map as viewed from the equator and central meridian (a).
Elevation increases from blue to red. The study area lies mainly within Arabia Terra,
(shown in b). Modified from MOLA Topographic Map, GIS I-2782, USGS................. 15
Fig. 5. Regional geological map of the northern plains of Mars, as viewed from the
North Pole. Modified from Tanaka et al. (2005). ............................................................ 17
Fig. 6. Locations of shorelines proposed by Parker et al. (1989) as viewed from the
North Pole. A thick black line indicates Contact 1, whereas Contact 2 is the thinner grey
line. The Ismenius Lacus quadrangle, which forms the upper left hand corner of the
study area, is shaded. From Parker et al. (1989).............................................................. 25
Fig. 7. Locations of two mapped shorelines, here labeled Contact 1 and Contact 2, as
viewed from the equator in an equidistant cylindrical projection of the surface. Meridiani
shoreline (as proposed by Clifford and Parker, 2001) is also shown. From Orm? et al.
(2004)................................................................................................................................ 27
Fig. 8. Locations of shorelines proposed by Fair?n et al. (2003) as viewed from the North
Pole. Contact 1 (now referred to as Shoreline 1) is indicated by a solid black line,
whereas Contact 2 (now referred to as Shoreline 2) is the thinner grey line. The dashed
line north of Arabia Terra is the orginal location of Contact 1, but has since been
modified to run further south (see discussion). From Fair?n et al. (2003). ..................... 28
Fig. 9. Flooding of the northern lowlands (a) up to a maximum depth of 1490 m (filled to
level of Contact 2); and (b) up to a maximum depth of 3570 m (filled to level of Contact
1) (Head et al. 2003). ........................................................................................................ 30
Fig. 10. Stages (A-H) of formation, excavation, and modification for marine impact
xiv
craters. Note the inverted sombrero morphology of the crater. From Johnson (2007)... 35
Fig. 11. Parameters in depth-diameter calculations include rim-rim diameter and rim
height. Also shown here are d
s
(depth of crater compared to surrounding topography) and
d
r
(depth of crater from rim). From Boyce et al. (2005).................................................. 37
Fig. 12. Illustrated difference between a slump (left) and a flow (right). Modified from
Geology Web Pages at http://www.nicholas.duke.edu/eos/geo41/geo41.htm (accessed
2007). ................................................................................................................................ 42
Fig. 13. Resurge gullies as observed in the Lockne and Kamensk craters in Eurasia.
From Orm? and Muinonen (2000).................................................................................... 43
Fig. 14. Locations of selected terrestrial analogues for Martian shallow-marine craters ?
three out of five are located on the American plate, and the remaining two on the
Scandinavian plate. ........................................................................................................... 46
Fig. 15. Spacecraft that have gone to Mars include Mariner 4, launched 28 November
1964 (a); Viking Orbiter Lander 1 and 2, launched 20 August 1975 and 9 September
1975 respectively (b); Mars Global Surveyor, launched 7 November 1996 (c); and Mars
Odyssey (still operating), launched 7 April 2001 (d). All images from NASA (available
from http://mars.jpl.nasa.gov/gallery/spacecraft/index.html)........................................... 53
Fig. 16. Timeline depicting the stages in Mars exploration by Viking Orbiters 1 and 2
(V1 and V2), Mars Global Surveyor (MGS), and Mars Odyssey (MO). White areas
indicate cruise time, and darker areas indicate sensor-operating time.............................. 56
Fig. 17. Different data products that can be derived from raw MOLA data. MEGDRs are
used in a GIS, and are therefore used in this study........................................................... 61
Fig. 18. Tiling scheme for the 128 pixel/degree MEGDRs from the Planetary Data
System (PDS), indicating the two tiles that are used as base maps in this study.............. 63
Fig. 19. The study area falls within four of the central quadrangles in the Northern
Hemisphere ? Mare Acidalium, Oxia Palus, Ismenius Lacus, and Arabia. From Malin
Space Science Systems (MSSS). ...................................................................................... 65
Fig. 20. The locations of Craters A, B, C, and D (as indicated by Orm? et al., 2004) in
Viking images F529A09 and F072A32, respectively. Images from the Planetary Data
System (PDS).................................................................................................................... 69
Fig. 21. THEMIS raster image of a Crater B on top of MOLA DEM (a) and
georeferenced MOC image draped over THEMIS image (b)........................................... 70
Fig. 22. Subdued topography and no elevated rim (a) (E0600102). Central terrace (b)
(V12594005). Large radial gullies (c) (R0500631). Large-scale wet mass movement (d)
xv
(R1402326). Images from Malin Space Science Systems (MSSS) and PIGWAD. ......... 72
Fig. 23. Weighted distribution of the six classes of shallow-marine crater characteristics.
Note that wet mass movement (WMM) is split into two sub-categories.......................... 74
Fig. 24. Locations of potential shallow-marine impact craters in relation to the three
shorelines proposed by Parker et al. (1989) as well as Edgett and Parker (1997). North
polar projection hill-shade base map is from the USGS and is based on MOLA
topographic data. Study area is shown by the gray rectangle.......................................... 77
Fig. 25. Stacked columns with individual contributions to the overall rank for the craters
proposed by Orm? et al. (2004). ....................................................................................... 79
Fig. 26. Stacked columns with individual contributions to the overall rank for the craters
in Set A. ............................................................................................................................ 82
Fig. 27. Stacked columns with individual contributions to the overall rank for the craters
in Set B.............................................................................................................................. 86
Fig. 28. Breakdown of characteristics present in Crater D and legend with color-coded
list of features.................................................................................................................... 87
Fig. 29. Breakdown of characteristics present in the exemplary candidates (see legend in
Fig. 28).............................................................................................................................. 88
Fig. 30. Locations of exemplary candidates on MOLA topographic raster image.......... 92
Fig. 31. Composite image for Crater D with MOLA base map and MOC wide and
narrow-angle images as high-resolution overlays............................................................. 93
Fig. 32. A close-up of Fig. 30 showing MOC narrow angle image overlay in more detail.
Radial gullies are indicated by white dashed lines and features labeled A-B are discussed
in text. ............................................................................................................................... 94
Fig. 33. Profile view of Crater D showing varying depth (in km) with variation in
latitude (in degrees) from north (left) to south (right). Vertical exaggeration is 21:1..... 96
Fig. 34. Composite image for Crater 24 with MOLA base map, THEMIS image and
MOC narrow-angle images as high-resolution overlays. ................................................. 97
Fig. 35. A close-up of Fig. 34 showing MOC narrow angle image overlay in more detail.
Radial gullies are indicated by white dashed lines and features labeled A-E are discussed
in text. ............................................................................................................................... 99
Fig. 36. Profile view of Crater 24 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 9:1. ........................ 100
xvi
Fig. 37. Composite image for Crater 66 with MOLA base map and THEMIS images as
high-resolution overlays.................................................................................................. 101
Fig. 38. A close-up of Fig. 36 showing THEMIS image overlay in detail. Features
labeled A-E are discussed in the text. ............................................................................. 102
Fig. 39. Profile view of Crater 66 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 8:1. ........................ 103
Fig. 40. Composite image for Crater 6 with MOLA base map, THEMIS image, and
MOC narrow-angle images as high-resolution overlays. ............................................... 105
Fig. 41. Close-up of Fig. 40 showing MOC and THEMIS image overlays in detail.
Features labeled A-B are discussed in the text. .............................................................. 106
Fig. 42. Profile view of Crater 6 showing varying depth (in km) with variation in latitude
(in degrees) from north to south. Vertical exaggeration is 10:1. ................................... 107
Fig. 43. Composite image for Crater 17 with MOLA base map, THEMIS image, and
MOC narrow-angle images as high-resolution overlays. ............................................... 108
Fig. 44. Close-up of Fig. 42 showing MOC image overlay over THEMIS. Features
labeled A and B are discussed in the text. ...................................................................... 110
Fig. 45. Profile view of Crater 17 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 29:1. ...................... 111
Fig. 46. Composite image for Crater 45 with MOLA base map and THEMIS images as
high-resolution overlays.................................................................................................. 112
Fig. 47. Close-up of Fig. 46 showing THEMIS image overlay. Features labeled A-C are
discussed above............................................................................................................... 114
Fig. 48. Profile view of Crater 45 from north to south. Vertical exaggeration is 13:1. 115
Fig. 49. Composite image for Crater 54 with MOLA base map THEMIS image as a
high-resolution overlay. .................................................................................................. 116
Fig. 50. Close-up of Fig. 49 showing THEMIS image overlay. Features labeled A-D are
discussed in the text above.............................................................................................. 118
Fig. 51. Profile view of Crater 54 from north to south. Vertical exaggeration is 6:1... 119
Fig. 52. Composite image for Crater 55 with MOLA base map and MOC narrow-angle
images as high-resolution overlays................................................................................. 120
Fig. 53. Close-up of Fig. 52 showing MOC image overlay. Features labeled A and B are
xvii
mentioned in the text above. ........................................................................................... 122
Fig. 54. Profile view of Crater 55 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 17:1. ...................... 123
Fig. 55. Composite image for Crater 58 with MOLA base map and MOC wide and
narrow-angle images as high-resolution overlays........................................................... 124
Fig. 56. Close-up of Fig. 54 showing MOC and THEMIS image overlay on top of
MOLA topography. Features labeled A and B are discussed in the text. Radial gullies
are indicated with white dashed lines. ............................................................................ 126
Fig. 57. Profile view of Crater 58 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 15:1. ...................... 127
Fig. 58. Spatial distribution of candidate craters shown by type. Type I craters are
shown in blue, type II craters shown in green, and type III craters shown in red. ......... 128
xviii
LIST OF TABLES
Table 1. Marine impact craters on Earth. Modified from Orm? and Lindstr?m (2000)
and Dypvik and Jansa (2003)............................................................................................ 12
Table 2. Geological time-scale for Mars based on cratering chronology. Modified from
Hartmann and Neukum (2001). ........................................................................................ 16
Table 3. Main events associated with impacts into aqueous targets (adapted from
Artemieva and Shuvalov 2002). ....................................................................................... 36
Table 4. Classification of ejecta morphology according to Barlow?s Catalog of Large
Martian Impact Craters and examples from THEMIS images (Barlow 2006a). .............. 41
Table 5. Physical properties of selected terrestrial shallow-marine impact craters........ 47
Table 6. Common remote sensing fields in the infrared section of the electromagnetic
radiation spectrum and their associated wavelengths. ...................................................... 51
Table 7. Instrument specifics in terms of resolution for MOLA, MOC, and THEMIS
(Malin et al. 1992; Abshire et al. 2000; Christensen et al 2002; Kirk 2005).................... 57
Table 8. Evaluation of Orm? et al.?s (2004) potential shallow-marine crater candidates.
........................................................................................................................................... 78
Table 9. Evaluation of characteristics of small shallow-marine crater candidates.
Exemplary ranking candidates are shown in italics.......................................................... 81
Table 10. Evaluation of characteristics of medium shallow-marine crater candidates... 81
Table 11. Evaluation of characteristics of large shallow-marine crater candidates........ 82
Table 12. Evaluation of characteristics of small shallow-marine crater candidates....... 84
Table 13. Evaluation of characteristics of large shallow-marine crater candidates........ 84
Table 14. Evaluation of characteristics of medium shallow-marine crater candidates.
Exemplary ranking candidates are shown in italics.......................................................... 85
Table 15. Physical parameters as measured from MOLA data for exemplary candidates.
All depths are negative, in other words, below mean surface level (see text for
explanation of abbreviations)............................................................................................ 89
xix
Table 16. Depths as measured from MOLA data for exemplary candidates and estimated
depths as calculated from depth-diameter relationships in literature (Garvin 2000;
Howenstein 2006). Factors of difference (see text) are given for each estimation.......... 90
1
INTRODUCTION
The study of impact craters began in 1609 when Galileo Galilei turned his
telescope to the moon for the first time. His discoveries, documented in Sidereus
Nuncius in 1610, changed the world?s perspective of the universe and became the
foundation upon which the science of impact craters was built (Koeberl 2001).
The formation of cosmic impact structures is a major geologic process not only on
Earth, but also on the solid surfaces of other planetary bodies throughout the solar
system. Impact crater formation is one of the most fundamental processes in the solar
system and is thought to be responsible for many important characteristics of the
terrestrial planets. Impact events have been responsible for the formation and
preservation of numerous ore deposits on Earth, and impacts have been responsible for at
least one major extinction in the Earth?s history (French 1998). Even the formation of
our Moon is commonly considered to have occurred when a large impactor collided with
Earth (Hartmann and Davis 1975; Canup and Asphaug 2001). Furthermore, physical
evidence of the impact processes, in the shape of nearly circular rimmed depressions
(Melosh 1980), can be seen on all solid planetary bodies in our solar system, and craters
are still formed throughout the solar system today (Melosh 1989).
2
Impact craters are the dominant physical features on most, if not all, solid
planetary surfaces and are one of the most important physical features from which the
surface history and composition of these bodies can be deduced. The study of craters can
provide important information about the evolutionary history of planetary bodies in our
solar system, and since the impact process has been described and studied in detail, the
initial shape of these features can be predicted (Malin et al. 1992; Boyce et al. 2005).
The study of impact craters plays a large role in understanding the properties of the target
surfaces, their ages, and the physical conditions of these surfaces at the time of impact.
The physical conditions of planetary surfaces can be deduced from impact craters
by looking at the morphology of these craters. One type of physical condition that is
particularly interesting is surface water cover. A layer of water on the surface influences
the shape of the final crater (Orm? et al. 2002), and thus the morphology of craters
formed in terrestrial environments differs from that of craters formed in marine
environments. This project investigates the morphology of impact craters in an area on
Mars suggested to be a shallow continental shelf environment (Parker 1989; Edgett and
Parker 1997; Fair?n et al. 2003).
A brief discussion on the background of impact cratering introduces this study,
followed by a discussion on the objectives and significance of the project.
3
Impact Craters
Distribution of craters
Scientists assume that comets and asteroids strike all regions of a planetary body
at approximately the same rate over a given span of time and thus the crater density of a
planetary surface indirectly indicates the relative age of that surface (Hartmann 1977;
Hartmann and Neukum 2001). This flux can change over time, but simply stated, regions
with higher crater densities tend to be older than regions with lower crater densities. For
example, the highly cratered surface of Mercury is expected to be much older than that of
the Earth. Furthermore, different planets have different rates of crater formation
depending on factors such as distance from the sun and atmospheric density. The
presence of an atmosphere on a planetary body shields the surface from some impacts by
eliminating smaller impactors before they can strike the surface. Thus, one can assume
that Mercury?s surface is not only older than that of the Earth, but also Mercury probably
had less protection in contrast to Earth. In addition to the factors mentioned above, the
size of a planetary body is also a factor in the rate of crater formation due to the influence
of the gravity with which it will affect projectiles.
If a planet (or moon) is geologically active or has an atmosphere or hydrosphere,
then processes such as volcanism, tectonism, weathering, and erosion can partially or
completely erase or degrade craters. The time it takes for a crater to be completely erased
from the surface is referred to as the crater retention age and it naturally depends on the
original diameter of the crater (Hartmann 1966). Crater degradation may take place on
4
the surfaces of inactive planetary bodies through the influence of younger impacts and
their ejecta, but these processes take place at a much slower rate than geological
processes such as regional volcanism or tectonics. Clearly, numerous processes have
modified and degraded impact craters on the surface of Mars, hence we can conclude that
Mars was either geologically active, or had an atmosphere or hydrosphere, or a
combination of these, at some point in the planet?s history.
Preserved craters are relatively rare on Earth, because weathering and other
geological processes have removed many of them. Currently there are 174 confirmed
impact crater structures (Earth Impact Database 2007) and approximately 564 probable
and/or possible suspected impact craters (Suspected Earth Impact Sites Database 2007)
on Earth. At present, there are more than 42,000 known large crater structures on Mars
(Barlow et al. 2003). Due to significantly smaller amounts of atmosphere-surface
interaction, the craters on Mars have not eroded as fast as those on Earth. Mars is also
much less geologically active at present, and therefore it is intuitive that the crater
population on Mars is much larger than that on Earth.
Types of craters
The formation of different types of impact craters depends on numerous factors,
such as the size, velocity, and composition of the impactor as well as the composition of
the target material (Melosh 1980; Melosh 1989; Holsapple 1993; Orm? et al. 2002).
The shape of the final crater is directly linked to size and velocity factors. If one
increases the size or the speed of a projected object, it is logical to assume that the extent
of morphological change (or alteration) is directly proportional to these two factors. The
kinetic energy that is released upon impact is responsible for the creation of the crater
shape. A simple relationship exists between this energy and the size and velocity of the
impactor, and is formulated as follows: KE = mv
2
. The exponential relationship
indicates that the velocity of the projectile is much more influential on crater shape than
the mere mass of the object.
There are three types of crater morphologies: a) simple craters, b) complex
craters, and c) multi-ring basins (Melosh 1989; French 1998). Simple craters are small
and bowl-shaped, whereas complex craters are large and flat-floored (see Fig. 1) (French
1998). Complex craters also usually have central peaks that formed during the rebound
of the transient crater, while simple craters exhibit no further structural features. Multi-
ring basins are large crater structures where the basins can extend for hundreds of
kilometers and usually consists of series of concentric rims much like circular mountain
ranges (French 1998).
On Mars, simple-to-complex crater transition occurs at diameters of 3-8 km (Pike
1980); where on Earth it is about 4 km for crystalline targets and roughly 2 km for
sedimentary targets (French 1998). Craters in this study are mainly complex craters. In
this study, most of the complex craters exhibit only a central peak; however, some of the
5
craters do exhibit peak rings. Most of the complex craters in this study also show signs
of structural rim failure in the form of slumped terraces and dislocated blocks of rim
material (see Fig. 2).
(a)
(b)
Fig. 1. Morphology of a complex crater (b) differs substantially from that of a simple
crater (a). Modified from French (1998).
Fig. 2. Different layers present in the target (A-F) as well as the ejecta blanket (G) of a
complex crater. Also shown are a terrace (t) and a block of rim material (b) both
products of slumping. From Pike (1980).
The composition of the impactor also plays a role in the morphology of the crater,
particularly when the composition can be linked to density. Most projectiles are either
6
7
comets or asteroids. Both comets and asteroids are small, rocky objects; however,
comets also contain significant amounts of volatile ices. Comets are mostly less dense
than asteroids, but they are often originally much larger and travel at higher velocities
(Melosh 1989).
The influence of the target material composition is a wide and popular field in
impact studies and it may even be more important in shape determination than the
properties of the projectile (Melosh 1989). For impacts into solid targets such as
crystalline bedrock, the crater shape will be uncomplicated and straightforward. For
impacts into soft targets, such as unconsolidated sands and water, the crater shape will be
much more complicated, and often larger. Impacts on land are often into crystalline
bedrock, unless the environment is sedimentary. Impacts in water are more complicated
due to the various layers that are present in the target material (Orm? et al. 2002). If the
water overlays a region of unconsolidated sediment, which in turn overlays crystalline
bedrock, then this tri-layer composition of the target makes the morphology of the
resulting crater much more intricate, often resulting in an ?inverted sombrero?
morphology (Kenkmann 2005; discussed in more detail below).
Beyond the influence of the target composition (further discussed below), other
factors that contribute to the morphology of craters include the extent of atmospheric
interference, the effects of modification, and the obliqueness of impact. Although these
factors are significant, they are of less importance in this specific study. It is assumed
8
that the craters in this study formed under roughly the same atmospheric conditions and
that they have been modified for equivalent amounts of time by similar factors. Oblique
impacts are much more common than vertical impacts, yet, most oblique impacts leave
approximately concentric crater forms much like vertical impacts do. In the case of
marine impacts, a projectile entering the water at an oblique angle would experience
more traveling time within the water column than a projectile at a vertical angle, thus
lowering the chances of seafloor crater formation (Artemieva and Shuvalov 2002). The
crater population in this study is limited to circular impact craters, but no calculations on
water column depth were done and thus no conclusions were drawn on the difference in
morphology between impacts at vertical and oblique angles.
As mentioned before, the influence of the target material composition is very
important. Sub-aerial craters (craters formed on land) usually exhibit an uncomplicated
structure; whereas sub-aqueous craters (craters formed in water) usually have more
complicated, and often larger, structures (Orm? and Lindstr?m 2000).
Observations of Earth-based marine impacts show that the water column greatly
influences the shape, size, and lithology of the resulting crater fill (Orm? et al. 2002;
2004). This gives impacts into marine environments a completely different nature from
impacts into dry, sub-aerial targets. Marine-target environments generally exhibit two-
fold rheology: a weak, volatile-rich upper layer, and a hard, crystalline lower layer.
Owing to the difference in the strength of the two layers, the crater shape that is often
created is that of an inverted sombrero (Kenkmann 2005). This inverted sombrero shape
is an outer ring with a large, flat annular trough surrounding a central peak or peak ring,
as shown in Fig. 3. Furthermore, the two-fold rheology is also responsible for creating
planes on which slump blocks can easily slide down, resulting in a larger crater with a
wider annular trough (Collins and W?nneman 2005).
Fig. 3. Cross-section of the inverted sombrero morphology observed at Chesapeake Bay
impact crater. The section runs from the western rim to the central peak due east. The
wide annular trough and the slumped blocks where the rim collapsed are shown. From
Horton et al. (2006).
More than 70% of the Earth?s surface is covered by water, and has been for a long
time, therefore it is expected that most cosmic impacts on Earth would have occurred at
sea. However, only 25 out of 564 suspected impact craters on Earth (roughly 5%) are
considered to be marine impact craters (Dypvik and Jansa 2003; Dypvik et al. 2004;
Suspected Earth Impact Sites Database 2007). Furthermore, only one of these marine
impacts, Eltanin, in the Southern Ocean, occurred in a deep marine setting (Gersonde et
al. 1997).
9
10
It is important to understand why there are so few marine impact craters on Earth
so that we can search in the right areas for marine impact craters on Mars. Some reasons
why there are only a few marine impact craters on Earth include a) reduced force of
impact; b) plate-tectonics; and c) limited exploration (Artemieva and Shuvalov 2002;
Gersonde et al. 2002). These reasons may also be applied to explain why no marine
impact craters have been identified on Mars.
Firstly, the strength of the protecting water column weakens the kinetic energy of
the impact, and therefore the chance of forming a crater on the ocean floor is smaller.
Seafloor craters form only when the impactor interacts with the sub-sea bedrock, so great
ocean depths preclude crater formation beyond the water column and it is thus more
likely to find distinctively marine crater shapes in shallow water than in deep water. If
the ratio (h/D) of water depth (h) to crater diameter (D) is larger than 0.4, there is no
formation of a seafloor crater and if the h/D ratio is larger than 4, there is no formation of
any impact-related features (Artemieva and Shuvalov 2002). Projectile size is also a
factor to be considered: if the diameter of the projectile is less than the target water-depth,
some or all of the crater shape is formed in the water (Orm? et al. 2006). Due to
numerous modification processes that take place immediately following the contact and
excavation stages of crater formation, the transient crater does not last very long (French
1998). In marine impact craters, this effect is magnified due to the violent return of
displaced water that rushes back and re-deposits the sediments (W?nneman and Lange
11
2002), modifying the crater rim and floor almost instantaneously. Seafloor-craters are
therefore rare on Earth, and possibly also on Mars, because of the protection of the
surface by the water layer.
Secondly, plate tectonics usually dictates that denser oceanic plates subduct
beneath lighter continental plates. Continental plates are therefore typically much older
than oceanic plates. This dynamic movement of the lithosphere periodically eradicates
crater forms on the ocean floor as the oceanic plate subducts beneath a continental plate.
Marine impact craters are therefore not that common on Earth because on a large part of
the planet?s surface, the evidence of these craters is continuously removed. This might
also be the case on Mars where large igneous events resurfaced the planet during its
earlier history, removing evidence of impact craters.
Lastly, the ocean floor is relatively unexplored when compared to the continental
crust. Not only is it easier for geologists to explore features on land, but it is also easier
to observe terrestrial features from satellites in orbit around Earth. Furthermore, craters
on the seafloor may not only be buried by water but also by sediments and are therefore
harder to recognize than craters on land. Therefore, it is not unexpected to have found
evidence of so few marine impacts. Even though Mars does not currently have an ocean
that obscures crater forms from satellite imagery, it is true that the vast amounts of
satellite imagery is yet to be fully analyzed and explored due to the large volume.
12
Even though few marine impact craters are identified on Earth, it should be noted
that exceptional preservation is required to confirm marine origin. Thus, most confirmed
marine impacts are well-preserved examples. When seafloor craters are formed, their
post-modification shapes are usually better preserved in aqueous environments than
similar craters in dry, land environments. This is true because of rapid sediment burial
(Orm? and Lindstr?m 2000). The sediment burial preserves the shape of the crater
(Dypvik et al. 2004) similar to a layer of sediments preserving the shape of a fossil
skeleton. To study marine impact craters on Mars, it is important to refer back to
terrestrial examples. Table 1 lists the locations, diameters, and ages of 17 confirmed
marine impact craters on Earth.
Table 1. Marine impact craters on Earth. Modified from Orm? and Lindstr?m (2000)
and Dypvik and Jansa (2003).
Crater Locality Diameter (km) Age (Ma)
Avak Alaska, USA 12 > 95
Chesapeake Bay Virginia, USA 85 35.5 +/- 0.3
Chixulub Yucutan, Mexico ~180 64.98 +/- 0.05
Eltanin South Pacific ? 2 to 15
Gusev Donets, Russia 3 49.0 +/- 0.2
Granby Link?ping, Sweden 3 470
Kaluga Kaluga, Russia 15 380 +/- 5
Kamensk Donets, Russia 25 49.0 +/- 0.2
Kara Kara Sea, Russia 65 70.3 +/- 0.3
K?rdla Hiiumaa, Estonia 4 455
Karikkoselk? L?si-Suomi, Finland 1.3 440 to 445
Lockne ?stersund, Sweden 13.5 > 455
Mj?lnir Barents Sea, Norway 40 142 +/- 2.6
Montagnais Nova Scotia, Canada 45 50.5 +/- 0.76
Neugrund Gulf of Finland, Estonia 20 535
Ust Kara Kara Sea, Russia 25 70.3 +/- 2.2
Wetumpka Alabama, USA 7.6 81.0 +/- 1.5
13
Objectives
This study aims to answer the question of evidence of marine impact craters on
the surface of Mars. Although it is believed that these structures should exist, and some
structures have been identified as potential shallow-marine impact craters (Orm? et al.
2004), as yet there is still no catalogue of potential candidates, and large areas of
continental shelf environment lies unexplored.
This project aims to investigate imagery and topographic data collected from the
Mars Global Surveyor and Mars Odyssey missions to study the general topography and
morphology of impact craters on selected parts of Mars?s surface. The main goal is to
map and describe impact structures on a part of Mars described by some as a shallow
continental shelf area (Parker 1989; Edgett and Parker 1997; Fair?n et al. 2003; Orm? et
al. 2004). The use of remote-sensing techniques is essential in the detailed study of
impact structures on planetary surfaces. Use of different datasets obtained from the Mars
Global Surveyor and Mars Odyssey spacecraft allows us to gain insight into the
topography and morphology of selected areas on Mars, at higher resolutions that possible
with previous datasets such as that of the Viking and Mariner spacecraft. These data are
used to identify and rank potential marine impact crater candidates in an area where there
is evidence of a historical shallow-marine environment on Mars.
All candidate craters in the study area are rated in an effort to quantify the
evidence of marine origin. The criteria for the quantification system are mainly collected
14
from literature on terrestrial marine impact craters and include characteristics such as
evidence of wet mass movement, radial gullies, resurge deposits, central terraces,
collapsed rims, and subdued topography. Exemplary candidates are discussed in detail
and data (remotely sensed and field-based) from shallow-marine impact structures on
Earth are used to compare characteristics of these craters with those of shallow-marine
impact craters on Mars (e.g., Von Dalwigk et al. 2001; Dypvik and Jansa 2003; Horton et
al. 2006; King et al. 2006; Orm? et al. 2006).
The selected area of study on Mars falls largely within northwestern Arabia Terra
(NWAT) with small sections in Acidalia Planitia and Chryse Planitia (Fig. 4). The study
area lies roughly right above the equator on the central meridian line and is bounded by
the 20? N and 40? N latitude lines as well as the 340? E and 20? E longitude lines. Arabia
Terra is a large, flat region straddling the distinct geologic boundary, commonly referred
to as the crustal dichotomy, which roughly separates the northern and southern
hemispheres. The average elevation of the study area is around 1-2 km below mean
surface level. Assuming that a large sea once covered the northern lowlands early in its
history (Parker 1989; Edgett and Parker 1997; Fair?n et al. 2003), water would have
covered all areas north of the dichotomy, thus creating a shallow-water, continental shelf
environment of varying width all along the dichotomy. On Earth, a continental shelf is
an ideal setting for the preservation of shallow-marine impact craters in the seafloor since
a seafloor crater is expected to form and rapid sedimentation should keep further
erosional processes from destroying it. Similarly, NWAT, which arguably was once part
of a continental shelf, makes a suitable study area for shallow-marine impact craters.
(a) (b)
Fig. 4. Mars topographic map as viewed from the equator and central meridian (a).
Elevation increases from blue to red. The study area lies mainly within Arabia Terra,
(shown in b). Modified from MOLA Topographic Map, GIS I-2782, USGS.
15
16
Based on cratering chronology, the geological evolution of Mars is divided into
three epochs: Noachian (N), Hesperian (H), and Amazonian (A) (see Table 2).
Table 2. Geological time-scale for Mars based on cratering chronology. Modified from
Hartmann and Neukum (2001).
Age Epoch Characteristics
2.9 Ga ? date Amazonian Little geological activity
3.7 Ga ? 2.9 Ga Hesperian Intense volcanic activity
4.5 Ga ? 3.7 Ga Noachian Heavy cratering
The study area is largely Noachian in age. Some small parts of the study area are
of Hesperian and Amazonian age, but these are less than 20% of the total surface
exposure. The study area includes two major geological units, Noachis Terra of
Noachian age (shown in orange) and Vastitas Borealis of Amazonian age (shown in
green) (see Fig. 5).
Fig. 5. Regional geological map of the northern plains of Mars, as viewed from the
North Pole. Modified from Tanaka et al. (2005).
This study focuses on craters within a size range of 10 to 100 kilometers in
diameter. The spatial resolution of the topographic data determined the lower limit,
whereas convenience is mainly responsible for the upper limit. However, it should be
noted that most of the Earth analogues for shallow-marine impact craters, fall within this
range (refer back to Table 1).
17
18
Significance
In unison with Goal 3, Sub-goal 3-C of NASA?s Strategic Plan (NASA 2006),
this project seeks to contribute to the scientific knowledge about the origin and history of
the solar system by investigating the environmental history of Mars. The existence and
extent of oceans on the surface of Mars is a highly debated question that is yet to be
answered. Evidence of shallow-marine impact craters on Mars will aid in understanding
the morphological evolution of the planet and its former oceans, and could potentially
support the theory of the occurrence of one or more large oceans on the surface of Mars
during its early history. Research on shallow-marine impact craters on Mars is still in an
early stage and a catalog of potential shallow-marine impact craters is a useful addition to
the Martian crater catalogs that are already available. The results of this type of study are
useful in helping develop a general classification and characterization of potential marine
craters.
19
LITERATURE REVIEW
To give a proper overview of the published literature in the field of shallow-
marine impact craters, this chapter is subdivided into several sections, the first of which is
a brief discussion on the evidence for water on Mars. A second section follows, more
focused on the arguments for the existence of an ocean(s) on Mars. The third section
explores the different physical properties of shallow-marine craters and focuses
specifically on the characteristics for shallow-marine impact craters, and the fourth
section discusses some examples of shallow-marine impact craters on Earth. Lastly, a
short background of remote sensing is given, and the methods employed in this study are
introduced.
20
Water on Mars
Water has played, and is still playing, a large role in Mars? geological evolution.
Until recently, proving the presence of water on Mars has been an elusive task, regardless
of the fact that numerous scientists have expected it for decades. Remote sensing
combined with ground-based rover observations and laboratory-based Martian meteorite
analysis make it possible to support the presence of water and to analyze the influence it
has had on the geology of this planet (Squyres 1989; Baker 2001).
At present, Mars has surficial and near-surface water (Solomon et al. 2005;
McSween 2006). Water in a liquid form is not stable on the surface due to low
temperatures and pressures, but occurs frozen at the poles and below the surface.
According to Solomon et al. (2005), there is evidence for interaction of liquid water with
the Martian surface particularly during the Noachian epoch. The Noachian is the period
during which heavy bombardment took place on the surface of the planet. It is during
this period that Mars is thought to have been warm and wet. Heavy resurfacing due to
volcanic activity characterized the Hesperian, and the Amazonian is generally believed to
have been geologically inactive.
McSween (2006) lists the following three observations as indicators of water on
the surface: a) the geomorphology (Baker 2006); b) the presence of altered as well as
evaporite minerals; c) and the composition of Martian meteorites. Geomorphology is
remotely studied by means of high-resolution visual imagery of the surface, and more
21
recently, also by shallow, ground penetrating radar. The surface mineral deposits are also
remotely studied, mainly through spectrometers on both rovers and orbital spacecraft.
Martian meteorites are the only true samples we have from this planet, and are thus the
only objects that can be analyzed in the laboratory.
Geomorphologic indicators of water include channels, valleys, alluvial fans, and
sediments (Solomon et al. 2005; Baker 2006). Even though scientists have observed such
structures on Mars?s surface for decades, it is only recently that imagery became
advanced enough to show the details of these water-related features. Some of the recent
images show small gullies running through crater rims, and these have been interpreted as
evidence for recent melting of near-surface ground ice, also referred to as seepage (Malin
and Edgett 2000; Solomon et al. 2005; Baker 2006). Further evidence of subsurface ice
can be found in the presence of a) grooved textures on channel and crater walls, b)
unusual crater ejecta shapes caused by fluidization of ice, and c) glacial landforms
(Christensen 2006). Indicators of flowing water are found throughout the surface of the
planet, but valleys and channels are more prevalent in the mid-latitude regions. This
distribution indicates that water might have once flowed from the higher southern
latitudes to the lower northern latitudes, forming an ocean in the northern plains.
In addition to an abundance of igneous minerals, the Martian surface shows
evidence of chemically altered minerals as well as evaporitic sedimentary deposits
(Wyatt and McSween 2006). The robotic Mars Exploration Rovers (MERs) have
22
discovered much about the nature of the surface of the planet. Spirit rover landed on 4
January 2004 in Gusev crater and found drenched, altered volcanic ash at nearby
Columbia Hills. Opportunity rover landed on 25 January 2004 in Eagle crater and found
rippled evaporite sediments in the area. The discovery of sulfate minerals indicates that
the basaltic crust may have been intensely weathered by water, and the presence of
hematite supports the theory that water once flowed across the surface at Meridiani
Planum (Squyres et al. 2004). Besides the mineralogy, the cross-stratification observed
in rock outcrops implies that the area was affected by transport of sediment, either under
aqueous or aerial conditions. According to Squyres et al. (2004; 2006a; 2006b), the size
(wavelengths of a few centimeters) and the geometry of these ripples, suggest the former.
The trough cross-bedding found in Eagle crater as well as in Endurance and Erebus
craters nearby, displays festoon or concave-upward geometry that on Earth is seen only in
aqueous environments (Herkenhoff et al. 2004; Squyres et al. 2006a; 2006b).
Further support for a warmer and wetter Mars in the past is provided by the
results of the spectrometers onboard the Mars Odyssey and Mars Express spacecraft.
THEMIS, a thermal emission spectrograph onboard Mars Odyssey, found evidence of
hematite at Meridiani Terra long before the Opportunity rover did. More recently,
OMEGA, a reflectance spectrometer onboard Mars Express, found evidence of multiple
ancient sedimentary deposits that formed in an aqueous environment (Paige 2005).
Reflectance spectrometers can gather data from fine-grained deposits that would
normally give a very weak emission signal, which is why this instrument has identified
23
more aqueous sedimentary deposits than THEMIS or other thermal spectrometers.
Of the more than 24,000 meteorites found to date, only 36 are known to be
Martian in origin (Leshin and Vicenzi 2006). Three classes of Martian meteorites exist
(shergottites, nakhlites, and chassignites, which are known collectively as SNCs) based
on mineralogical similarities to type meteorites that fell in India, Egypt, and France,
respectively (Lodders 1998). Martian meteorites have isotope ratios that are all very
similar to one another, yet vary greatly from isotope ratios of asteroids, comets, and
terrestrial rocks. Even though the SNC meteorites are igneous in origin, they contain
traces of water-precipitated minerals such as carbonates and sulfates (Gooding 1992;
Leshin and Vicenzi 2006). Furthermore, Gooding (1992) concluded that the SNC
secondary mineral precipitated from saline liquid water. These aqueous precipitates
indicate that water was present on Mars at least at the time that the meteorites formed,
and maybe even beyond that.
The evidence for water on Mars is unequivocal. Aqueous activity was, and
maybe still is, common on Mars. The question is no longer if water was present on the
surface, but how much and for how long (Baker 2001; Malin and Edgett 2003).
24
Oceans on Mars
Oceans on Mars have been proposed by several authors (Parker et al. 1989; Edgett
and Parker 1997; Fair?n et al. 2003) and two main oceans have been suggested; one large
ocean of Noachian age (at least ~4 Ga; Clifford and Parker 2001) and a smaller Hesperian
ocean (of ~2-3 Ga; Clifford and Parker 2001). Since then, a few smaller seas and lakes
have also been proposed, but these are much younger and smaller and therefore do not
play a large role in this study.
Large Noachian ocean and smaller Hesperian ocean
Using Viking data, Parker et al. (1989) interpreted the dichotomy as the remnant
of a shoreline of an ocean or lake. In fact, Parker et al. (1989) delineated two potential
shorelines, and called them Contact 1 and Contact 2 (see Fig. 6). Contact 1 lies nearly
along the boundary between the upland and lowland surfaces on Mars, which falls along
the crustal dichotomy. Contact 1 encloses the northern lowlands, runs through Arabia
Terra and exhibits two types of boundaries; a gradational, transitional boundary and a
sharper, fretted boundary (Parker 1989). Contact 2 lies between Contact 1 and the rest of
the northern plains, in other words, it encloses a slightly smaller part of the northern
lowlands. Both of these contacts are drawn in Fig. 6. Based on the relationship between
these two shorelines, it seems that the ocean contained by Contact 1 was not only much
larger and expansive than the ocean contained by Contact 2, but also much older and thus
its shoreline is less well preserved.
Fig. 6. Locations of shorelines proposed by Parker et al. (1989) as viewed from the
North Pole. A thick black line indicates Contact 1, whereas Contact 2 is the thinner grey
line. The Ismenius Lacus quadrangle, which forms the upper left hand corner of the
study area, is shaded. From Parker et al. (1989).
Edgett and Parker (1997) proposed the existence of a large ocean in the vicinity of
Arabia Terra during Mars? early history based on three factors: a) the existence of an
interpreted shore-like contact between the southern highlands and the northern lowlands;
b) the occurrence of polygonal evaporite structures; and c) the presence of sand deposits.
The difference in elevation and roughness around the crustal dichotomy can easily be
interpreted as evidence for an ancient shoreline. Thus, the potential of the dichotomy to
represent a shoreline is by now fairly accepted. Large (3-8 km across) polygonal
structures in Sinus Meridiani (just south of Arabia Terra) could indicate that water was
25
26
present on the surface here and that the water has since evaporated (Edgett and Parker
1997). Layered sand deposits are found in and around craters in Arabia Terra. The sand
deposits do not drape over the topography the way eolian features are expected to do, but
rather seem to lap up against ridges and crater rims as if emplaced by lacustrine
processes. These findings are in accord with the proposed Noachian shoreline that runs
through Arabia Terra, also referred to as Contact 1.
Furthermore, Edgett and Parker (1997) also mention that large valley networks in
Arabia Terra are almost entirely absent, and suggest that this is because Arabia Terra was
under water at the time the valley networks formed. The formation of the major valley
networks on Mars have been dated as Noachian (Carr 1995; Hynek and Phillips 2003),
which corresponds with the Noachian shoreline proposed by others (Parker 1989; Fair?n
et al. 2003).
More recently, the existence of a northern ocean or Oceanus Borealis, within the
shorelines suggested by Parker (1989) was suggested because the dichotomy not only
resembles a shoreline, but the outflow channels and valley networks terminate at the
boundary much like rivers terminate in deltas; the low density of craters in the northern
plains is due to the water cover of a large sea; and the spectrographic signatures of
carbonates and sulfates indicate evaporite deposits (Fair?n et al. 2003). Some of these
arguments have been stated before, for example, Helfer (1990), argued that oceans in the
northern plains are to be expected based on the low crater density observed in this region.
Smith et al. (2001) stated that MOLA data, for regions with large channels, support
sustained flow and multiple flooding events as well as potential existence of an early
ocean. Most importantly, Fair?n et al. (2003) highlighted the existence of another
shoreline suggested by Clifford and Parker (2001). Clifford and Parker suggested that the
Noachian shoreline, Contact 1, may have a slightly different location than first proposed.
The location of Contact 1 has been modified to accommodate for the large elevation
differences across the boundary, to include more of the valley networks that end abruptly
south of Arabia Terra, and to coincide with the crustal thickness dichotomy (Fair?n et al.
2003). Contact 1 is also referred to as the Arabia shoreline, and Contact 2 as the
Deuteronilus shoreline (Clifford and Parker 2001). Both these shorelines, as well as the
Meridiani shoreline, which later formed part of Fair?n et al.?s (2003) Shoreline 1, are
shown in Fig. 7.
Fig. 7. Locations of two mapped shorelines, here labeled Contact 1 and Contact 2, as
viewed from the equator in an equidistant cylindrical projection of the surface. Meridiani
shoreline (as proposed by Clifford and Parker, 2001) is also shown. From Orm? et al.
(2004).
27
The three shorelines discussed by Fair?n et al. (2003) are drawn from a north
polar perspective in Fig. 8. One can see from this figure that the original ocean,
encompassed by either the Meridiani or Arabia shorelines, was much more extensive than
the more recent ocean encompassed by the Deuteronilus shoreline.
Fig. 8. Locations of shorelines proposed by Fair?n et al. (2003) as viewed from the North
Pole. Contact 1 (now referred to as Shoreline 1) is indicated by a solid black line,
whereas Contact 2 (now referred to as Shoreline 2) is the thinner grey line. The dashed
line north of Arabia Terra is the orginal location of Contact 1, but has since been
modified to run further south (see discussion). From Fair?n et al. (2003).
28
29
Thick sedimentary layers cover Arabia Terra as well as most of the northern
lowlands. Sedimentary deposits could indicate that a large ocean once existed in the
northern hemisphere (Edgett and Parker 1997; Edgett and Malin 2002; Venechuk et al.
2005; Baker 2006). The main ~100 m thick sedimentary layer is commonly referred to as
the Vastitas Borealis Formation (VBF) and is found across the smooth northern plains
(Smith et al. 1998; Carr and Head 2003). Numerous origins for this sedimentary layer
have been suggested, including paleopolar, volcaniclastic, eolian, and sub-aqueous
sedimentary, however, Mars Orbiter Camera (MOC) images have not clearly identified
one of these theories as the true origin (Edgett and Malin 2002). Boyce et al. (2005)
made a study of the depth-diameter ratios of a large population of impact craters in the
northern lowlands and showed that the VBF is draped over most of the lowlands giving it
a smooth appearance. The VBF could only have formed as sub-aqueous sedimentation
beneath a large body of water, thus, it is highly likely that an ocean with an average depth
of ~430 m could have once existed in the northern lowlands (Boyce et al. 2005). Based
on their conclusions, an ocean existed in the northern lowlands during Late Hesperian
(approximately 3,5 to 1,8 Ga ago). This would correspond with the second, smaller, and
younger ocean, and shoreline Contact 2.
Mars Orbiter Laser Altimeter (MOLA) data further provide evidence for an ocean
on Mars (Head et al. 1999; Ivanov and Head 2001). According to Head et al. (1999), the
current high-resolution altimetry data of the northern lowlands of Mars affirm the
hypothesis that elevations around Contact 2 are close to an equipotential line and that it
indicates the level shoreline of a large standing body of water somewhere during the
Hesperian epoch. In addition, Head and Ivanov (2001) also found that the Hesperian-age
outflow channels all enter the northern plains at similar elevations and that the
morphology of these channels change rapidly from sub-aerial to sub-marine at the base
level.
If water filled the northern lowlands up to the Meridiani shoreline, the equivalent
global depth would be 1510 m (Carr and Head 2003). Similarly, if water filled the
lowlands up to the Arabia shoreline, the global depth would be 599 m and for the
Deuteronilus shoreline, it would be 130 m (Carr and Head 2003). The average height of
the Arabia shoreline is -1680 m, which would yield an ocean with maximum depth of
3570 m over the North Polar basin (Head et al. 2003). Similarly, the average height of
the Deuteronilus shoreline is -3760 m, which would yield and ocean with maximum
depth of 1490 m over the North Polar Basin (see Fig. 9) (Head et al. 2003).
(a) (b)
Fig. 9. Flooding of the northern lowlands (a) up to a maximum depth of 1490 m (filled to
level of Contact 2); and (b) up to a maximum depth of 3570 m (filled to level of Contact
1) (Head et al. 2003).
30
31
Smaller local lakes
As discussed in the previous section on water on Mars, the results from the Mars
Exploration Rover (MER) explorations indicate that Meridiani Planum has been
periodically flooded (Squyres et al. 2004). However, the timing of these flooding
episodes is hard to assess. According to Squyres et al. (2004), cratering rates suggest that
the rocks in this area could be of Noachian age, or at least a few billion years old.
However, it seems more likely that these small, localized seas were of Amazonian age.
Since Meridiani Planum was likely once a large, shallow playa lake, it is easy to deduce
that there may have been numerous small seas on Mars? surface during its history.
Evidence for oceans on Mars
Since the shorelines of three different large water masses have been proposed,
numerous critical evaluations of these shorelines have been done. Malin and Edgett
(1999), as well as Carr and Head (2003), evaluated relations along the crustal dichotomy
on Mars with the use of high-resolution imagery from Mars Orbiter Camera (MOC) and
Thermal Emission Imaging System (THEMIS). The morphologic evidence for a large
standing body of water is equivocal in some places, and no visual evidence was found for
a shoreline along the crustal dichotomy (Malin and Edgett 1999; Carr and Head 2003;
Chapman 2003; Ghatan and Zimbelman 2006). Most recently, Ghatan and Zimbleman
(2006) studied 735 MOC and 447 THEMIS images and found only four images with
potential candidates for coastal ridges such as spits and barrier islands. Ghatan and
Zimbleman (2006) reason that either a) there was no northern ocean, b) factors such as
32
low wave energy and low sediment input were responsible for no formation of ridges, or
c) large amounts of erosion caused destruction of ridges. Malin and Edgett (1999)
mention that it may be hard to find evidence of coastlines from remotely sensed data
alone and that even coastlines on Earth are hard to discern from orbital or airborne data.
In addition, even though orbital cameras photograph the surface intensely, there may still
not be enough data for producing conclusions. It may even be that coastal features that
are common on Earth, are not well developed on Mars (Chapman 2003), and therefore we
cannot assume that no evidence of coastal features mean there was no coast.
Additionally, there are significant variations in elevation along the Noachian
shoreline, Contact 1. Sea level, by definition, should be more or less level; therefore,
changes in elevation do not support the existence of a sea level. Contact 1 was first
defined by Parker (1989) with a difference in elevation in places was as much as 11 km.
The modified Contact 1, or Shoreline 1 (Fair?n et al. 2003), decreases this elevation
difference to approximately 3 km.
Until recently, the difference in elevation was large enough to cast doubt on an
ancient, Noachian ocean encompassed by the Arabia shoreline. However, Perron et al.
(2007) suggested a global mechanism - polar wander - for altering elevations along the
dichotomy while retaining its original shape. Polar wander is the wobbling of a planet?s
axis and it is responsible for many climatic variations. Schultz and Lutz (1988)
suggested that polar wander has played a role in Mars?s geological history, but not until
33
Perron et al. (2007) modeled true polar wander in their recent study, did it become
apparent how influential this wandering has been in shaping the surface of Mars.
Polar motion changes a planet?s geographic surfaces, including the sea level
(Perron et al. 2007). For each shoreline, there is a paleopole that, when altered, will
deform the shape of the shoreline accordingly. Perron et al. (2007) shows that the
deformation associated with true polar wander can account for the trends of variation in
the shoreline elevations. True polar wander on Mars may have been induced by the
formation of the massive volcanic province, Tharsis (Zuber, 2007).
Perron et al.?s (2003) analysis removes the main argument against oceans on
Mars. Because it seems that oceans existed on Mars for at least some time in the Martian
history, the surface should display numerous crater forms that fit the Earth-based
description of a shallow-marine impact. The Martian surface could have up to 1,400
marine impact craters, based on the length of time that oceans existed on the planet and
the size of the possible ocean(s) (Orm? et al. 2004). Their study was based on the
minimum and maximum ages of duration for two large bodies of water in the northern
lowlands, in combination with the cratering rates for the planet during those times. Thus,
some previously submerged regions on the surface of Mars are potentially excellent areas
of study for shallow-marine impact craters.
34
Physical properties and characteristics of shallow-marine impact craters
In order to discuss the physical properties and characteristics of shallow-marine
impact craters, a brief discussion on the stages of formation of marine target craters is
necessary. The three main stages (contact/compression, excavation, modification) of
crater formation are the same for both terrestrial and marine craters (Melosh 1989);
however, the details of these stages differ slightly. In the contact/compression stage, the
projectile enters the water column, subsequently the unconsolidated sediment layers, and
finally the crystalline basement. During the excavation stage, a transient crater is formed,
often largely within the water column. Lastly, during the modification stage, numerous
physical changes occur within the crater, many related to water resurge, and some of
these processes continue until the crater is completely erased from the surface. Fig. 10
illustrates these different stages and shows the expected inverted sombrero morphology.
According to Artemieva and Shuvalov (2002), when a projectile enters a water
target there are numerous events that occur in the water column and on the seafloor. The
main events are listed in Table 3 below. These events, couple with the stages of
formation, are used to formulate a list of characteristics that are expected to be present in
shallow-marine craters.
Fig. 10. Stages (A-H) of formation, excavation, and modification for marine impact
craters. Note the inverted sombrero morphology of the crater. From Johnson (2007).
35
36
Table 3. Main events associated with impacts into aqueous targets (adapted from
Artemieva and Shuvalov 2002).
Event Effects Time
Contact Stage Strong shock-wave <1 second
Excavation Flow Surge formation 1-10 seconds
Possible craterform
Crater Collapse Tsunamis Minutes to hours
Craterform modification
Impact debris fall-back
Several physical parameters are used to describe impact craters such as the depth
to diameter ratios, the slope of the crater rim, and the shape and extent of the ejecta
deposits. Furthermore, several indicative characteristics can be associated with shallow-
marine impact craters on Earth, and these include features indicative of wet mass
movement (Dypvik et al. 2004), radial gullies (Von Dalwigk and Orm? 2001), resurge
deposits, central terraces, collapsed rims, and subdued topography. These features, if
present, could be visible on images and could therefore potentially be used as key
characteristics in the identification of Martian shallow-marine impact craters. Orm? et al.
(2004) recognized some possible shallow-marine craters with the use of low-resolution
Viking imagery based on expected morphological characteristics. All these parameters
are discussed in more detail below.
Depth-diameter ratios
Rim diameter and crater depth are two of the most important morphologic
elements of an impact crater, both on Earth as well as Mars (Smith et al. 2001; Boyce et
al. 2005). The ratio between depth and diameter is often a reliable indicator of the type
of crater (i.e., the environment in which it was formed) and the extent of crater
modification (Garvin et al. 2000; Aharonson et al. 2001; Boyce et al. 2005). Diameter
may not change much during crater modification and can be used as a good estimate of
original crater size, but in contrast, surface processes significantly influence crater depth
over time (Boyce et al. 2005). Fig. 11 shows the two depths that can be measured; one
from the rim to the center of the crater (d
r
), and one from the average surrounding
topography to the center of the crater (d
s
). The difference between these two depths (d
r
and d
s
) is the rim height, which is commonly about 4% or less of the crater diameter
(Orm? and Lindstr?m 2000).
Fig. 11. Parameters in depth-diameter calculations include rim-rim diameter and rim
height. Also shown here are d
s
(depth of crater compared to surrounding topography) and
d
r
(depth of crater from rim). From Boyce et al. (2005).
37
38
Common depth-diameter ratios for simple, terrestrial craters range from one third
to one fifth, depending on the nature of the target material (Orm? and Lindstr?m 2000).
More particularly, d = 0.29D
0.93
for simple terrestrial craters, and d = 0.15D
0.43
for
complex terrestrial craters (Grieve 1987) where d is the depth of the crater from the rim,
and D is the rim-to-rim diameter, both in kilometers. Notice that complex craters are
shallower than simple craters.
Studies of depth-diameter ratios for lunar impact craters (diameters <250 km)
have been done by Pike (1977) who found the relationship to be d = 0.23D
0.94
for small
craters and d = 0.75D
0.30
for larger craters. Howenstein (2006), made a study of the
depth-diameter ratios for large Martian craters (diameters of 20 to 2000 km), and found
the overall depth-diameter ratio to be d = 0.61D
0.33
. Garvin et al. (2000) found depth-
diameter relationships for polar and non-polar craters to be d = 0.03D
1.04
and d =
0.19D
0.55
respectively. For fresh, complex craters on Mars, this relationship has been
defined as d = 0.33D
0.53
(Smith et al. 2001). More specifically, craters in Arabia Terra
have a d/D ratio ranging from 0.07 to 0.09 (Barlow 1993). Depth-diameter ratios for
Martian craters seem to decrease with an increase in latitude due to an increase in sub-
surface volatiles and ices (Barlow 1993).
Modeling experiments suggest that marine craters should be wider than craters
formed on land under similar conditions (Gault and Sonnett 1982). According to Dypvik
and Jansa (2003), marine craters are characterized by larger diameters because of radial
39
enlargement. Some processes have been suggested to be responsible for the widening of
the crater, of which the most common are rim failure due to resurge or erosion by water
currents or due to sediment instability. Boyce et al. (2005) concluded that the low depth-
diameter ratios suggest infilling by sedimentary deposits, thus supporting the theory of a
large ocean depositing what has been described as the Vastitas Borealis Formation
(VBF).
Slopes
The slopes of crater rims are indicative of the style and duration of the subsequent
modification processes (Garvin et al. 2000; Aharonson et al. 2001). Crater slopes can be
modified during impact by resurge activity (Orm? and Lindstr?m 2000) or after impact
by thermal creep, a process wherein the difference in temperatures cause particles to slide
down a slope (Sharp 1968).
Kreslavsky and Head (2006) used MOLA data to measure the steepness of crater
slopes in the northern plains and found that some craters are very shallow. Shallow
craters could indicate that sediment infilling has taken place. The steepest crater wall
slopes in the population studied by Kreslavsky and Head (2006) are at angles of 30?, but
the majority of the 130 craters that were studied have more gently sloping crater walls.
This is in line with observations of marine craters on Earth because modification
processes in unconsolidated sediments often yield larger, more subdued craters.
40
Ejecta deposits
Further distinctive attributes of shallow-marine impact craters include the extent,
shape and thickness of the ejecta deposits that are blown out of the crater upon impact, as
well as the volume of the crater depression. Studies have shown that crater and ejecta
morphology depend greatly on sediment strength, which is directly linked to water
content (Wohletz and Sheridan 1983). The presence of water in the target material leads
to the formation of water vapor, which in turn accelerates and increases the formation of
an ejecta layer, resulting in wider ejecta blankets for marine target craters than for land
target craters (Melosh 1989; Dypvik 2004; Schaefer et al. 2006). Schaefer et al. (2006)
modeled impacts into wet and dry sandstone and found that ejecta velocities from the wet
sandstone were up to 50% higher than for the dry sandstone.
Barlow (2006a) classified ejecta morphologies based on appearance in her
Catalog of Large Martian Impact Craters. Table 4 lists the different morphologies and
some examples of each. The ejecta morphologies important in this study are the rampart
craters. Rampart craters are fluidized craters with lobate ejecta morphologies (Wohletz
and Sheridan 1983) and form as a result of impacts into volatile rich environments or
fluid-rich substrates containing water or ice, usually the latter (Mouginis-Mark 1987).
Table 4. Classification of ejecta morphology according to Barlow?s Catalog of Large
Martian Impact Craters and examples from THEMIS images (Barlow 2006a).
Name Description Example Image
SL Single lobe rampart
DL Double lobe
rampart
ML Multiple lobe
rampart
Di Diverse morphology
Pn Pancake ejecta
41
Wet Mass Movement (WMM)
Mass movement can occur at various rates, but most important in this study is the
rapid movement of unconsolidated sediment. Two types of mass movement are
particularly important: slides, where the movement occurs in a well-defined plane; and
flows, where the movement is more fluid in behavior (see Fig. 12). A slump is a type of
slide where material moves downward as a parcel or unit, often with a backward rotation
on a curved displacement surface (Kennett 1982). Listric or normal faults form as the
result of this displacement. A debris flow is a type of flow where particles chaotically
move downslope in a saturated sediment slurry while supported by cohesion strength
(Kennett 1982).
Fig. 12. Illustrated difference between a slump (left) and a flow (right). Modified from
Geology Web Pages at http://www.nicholas.duke.edu/eos/geo41/geo41.htm (accessed
2007).
Slumps, or block collapse of rim material into the crater are indicative of weak or
unconsolidated target material (Dypvik and Jansa 2003). Slump blocks are present in
both the Mj?lnir and Chesapeake Bay crater structures, but not at Lockne (Orm? and
Lindstr?m 2000). Flows, or tongues of rounded deposits, have been observed in both
42
Chicxulub (Kring 2005) and Lockne (Von Dalwigk 2001) craters. Flows are different
from slumps in that they are continuous and often rounded in shape.
Radial Gullies (RG)
Radial channels carved by sediment-loaded waters induced by violent resurge are
common in marine target craters (Orm? and Muinonen 2000; Von Dalwigk and Orm?
2001). Resurge gullies have been observed in both the Lockne (Lindstr?m et al. 1996)
and Kamensk (Orm? and Lindstr?m 2000) craters (see Fig. 13). Even though resurge
gullies are distinctive of marine impact craters, they tend to occur only in deeper water
where the water depth exceeds the projectile diameter (Orm? et al. 2002). The water
depth at the time of impact for Lockne has been estimated at 1000 m (Orm? et al. 2002)
and for Kamensk at 100 to 200 m (Orm? and Lindstr?m 2000). Thus, one can assume
that shallow water depth limits the potential for the formation of radial gullies.
Fig. 13. Resurge gullies as observed in the Lockne and Kamensk craters in Eurasia.
From Orm? and Muinonen (2000).
43
44
Resurge Deposits (RD)
Resurge deposits may form inside or outside of the crater, depending on the depth
of the target water, the strength of the crater material, and the location of the crater.
Intra-crater terrain is the evidence of resurge in the form of avalanches, slides, and
slumps of mixed blocks inside the crater (Dypvik and Jansa 2003). Extra-crater terrain is
the evidence of resurge in the form of resurge sediments and mixed blocks outside the
crater (Dypvik and Jansa 2003). As discussed by King et al. (2006), the ?extrastructure
terrain? observed at Wetumpka is likely a product of collapse of the rim in respond to
resurge flow.
Resurge deposits on earth have been confirmed through drilling (von Dalwigk and
Orm? 2001) and from fieldwork (King et al. 2006) in combination with subsequent
sedimentological description.
Central Terrace (CT)
A central peak terrace or a peak ring terrace is a large structure in the center of the
crater, sometimes associated with equally flattened concentric rings (Dypvik and Jansa
2003). Development of flat-topped central uplifts has been predicted by some studies
(Gault and Sonett 1982) and could be an indication of marine origin and that the structure
has been buried under water for some time.
45
Rim Collapse (RC)
Crater rims are usually tough features formed from crystalline bedrock. When
craters form in sedimentary environments, the rims are not as pronounced or as strong.
Often no remnant of rim is visible due to large amounts of inward slumping (Dypvik and
Jansa 2003). This is sometimes also referred to as structural rim failure. Structural rim
failure may be due to resurge activity or due to instability of the rim (King et al. 2006).
Subdued Topography (ST)
Craters that exhibit subdued topography have little to no elevation of the rim
above the surrounding topography. Subdued topography is indicative of the amount of
erosion that has taken place since formation, and thus the amount of time that has passed.
As erosion is often induced by water, it could also indicate that large amounts of water
were present on the surface for an extended period of time (Orm? and Lindstr?m 2000).
Lack of an elevated rim is considered by Dypvik and Jansa (2003) to be one of the most
distinctive features of marine impacts.
Note that none of the characteristics listed above can be used with certainty to
imply marine origin. Other processes may be responsible for generating similar features
in different locations; for example, complex craters on Earth often show signs of
slumping without having formed in marine environments. However, if more than two of
these characteristics are observed, it is likely that marine origin can be implied.
Examples of shallow-marine craters
The study of terrestrial impact craters in shallow-marine environments helps to
understand the formation and properties of these craters on the surface of Mars. Good
terrestrial analogues formed in a continental shelf environment and now show a well-
preserved structure and features that are described in detail. Some analogues are
particularly useful in this study, and they include the larger Chesapeake Bay and
Chicxulub craters, the medium Lockne and Mj?lnir craters, as well as the smaller
Wetumpka crater (Fig. 14). Table 5 lists the locations and sizes of five terrestrial marine
impact craters, as well as the estimated water depth upon impact for each of the events.
Fig. 14. Locations of selected terrestrial analogues for Martian shallow-marine craters ?
three out of five are located on the American plate, and the remaining two on the
Scandinavian plate.
46
47
Table 5. Physical properties of selected terrestrial shallow-marine impact craters.
Crater
Location Diameter
(km)
Water Depth
(m)
Chesapeake Bay Virginia, USA 85 ?340 m
Chicxulub Yucutan, Mexico 180 < 50
Lockne ?stersund, Sweden 13.5 > 200
Mj?lnir Barents Sea, Norway 40 300-500
Wetumpka Alabama, USA 7.6 30-100
Chesapeake Bay
The Chesapeake Bay impact structure is a ~85 km diameter crater that occurred
about 35 million years ago into the shallow Atlantic Coastal Plain with an estimated
water depth of ?340 m (Poag 1997; Poag et al. 2004; Horton et al. 2006). The structure,
currently buried beneath 150-400 m of post-impact sediments (Horton et al. 2006), is
very well preserved but not visible from the surface. At the time of impact, the target
consisted of crystalline basement overlain by 600-1000 m of unconsolidated sedimentary
rocks and 200-500 m of saline water (Poag et al. 2004).
Collins and W?nneman (2005) modeled the Chesapeake Bay impact event and
concluded that the main factor responsible for the morphology of the crater is the
variation in strength of the layers present in the target. Without this variation in strength
amongst the layers, it is likely that the diameter of the Chesapeake structure would have
been around 40 km (Collins and W?nneman 2005). This is roughly half the size of the
actual 85 km that has been determined through seismic profiling.
48
Chesapeake Bay crater exhibits an inverted sombrero morphology with an outer
zone or annular trough and an inner zone or moat (Poag 1997; Horton et al. 2006). The
rim has been subject to large-scale collapse and slumping resulting in crater-wall failure
(Poag 1997; Horton et al. 2006). Numerous extensional collapse structures are observed
in the seismic profiles, and these structures result in radial enlargement of the crater
(Poag 1997; Horton et al. 2006).
Chicxulub
The Chicxulub crater, with a diameter of approximately180 km, is the largest
crater that formed in a marine environment on Earth. Chicxulub occurred about 65
million years ago into a shallow sea with water depths of no more than 100 m (Pierazzo
and Melosh 1999). The crater is well-preserved beneath a layer of sediments and not
visible at the surface. Surge deposits are particularly well-preserved and include
deposition of massive sands, debris flows, and collapse of the central peak into a peak
ring (Dypvik et al. 2004; Kring 2005).
Lockne
The Lockne crater is a ~13 km diameter impact structure that occurred about 455
million years ago (Orm? and Miyamoto 2002; Sturkell et al. 1998) in Sweden. Water
depth at the time of impact was at least 200 m (Von Dalwigk 2001). Lockne crater
contains very good examples of resurge gullies that have been associated with large-scale
resurge flows (Von Dalwigk 2001). Four gullies as well as a few debris flow units have
49
been described and the rim has been classified as breached in more than one location
(Von Dalwigk 2001).
Mj?lnir
The 40 km diameter Mj?lnir crater in the Barents Sea (off the coast of Norway)
formed about 140 million years ago (Tsikalas 1998). The estimated diameter of the
asteroid is 1-3 km and the depth of the water roughly 300-500 m (Shuvalov and
Trubestkaya 2002). The Mj?lnir crater is not exposed at the surface, but instead is
located on the seafloor and presently covered by ~350 m of water and 50-400 m of
sediments (Dypvik and Jansa 2003). The structure was discovered from geophysical
data, much like most of the shallow-marine craters. The Mj?lnir crater rim is
characterized by terraces that are bordered by faults and possible gullies (Dypvik and
Jansa 2003). This crater also exhibits the classic inverted sombrero morphology with an
inner zone of 8 km and an outer zone of 12 km (Dypvik and Jansa 2003). Rotated fault
blocks are found in the annular trough (Orm? and Lindstr?m 2000).
Wetumpka
Wetumpka impact crater formed roughly 80 million years ago in a shallow sea of
depths 30-100 m. The crater is 7.6 km in diameter and is the smallest of the five analogs
discussed in this study. The remnants of the Wetumpka impact crater are not completely
covered by sediments. The Wetumpka crater exhibits signs of numerous intra- and extra-
crater resurge deposits as well as a structurally disturbed rim (King et al. 2002; 2006).
50
Remote Sensing
Remote sensing is a crucial tool in studying the surfaces of objects, particularly in
planetary science. In broad terms, remote sensing is the studying of a surface with
electromagnetic radiation sensors that are at a distance from the studying site. Planetary
exploration occurs almost entirely remotely, even with the exception of human space
travel, remote sensing is still involved. (Ramsey and Christensen 1998; Carr and Garvin
2001)
Remote sensing can either occur passively by means of recording reflected light
from surfaces or actively through sending and receiving signals. Common examples of
passive remote sensing include photography such as the images taken by Mars Orbital
Camera (MOC) and Thermal Emission Imaging System (THEMIS) as well as
spectrography such as the spectra gathered by THEMIS. Examples of active remote
sensing include RADAR (radio detection and ranging) or LIDAR (light detection and
ranging), where time of flight of signals is measured. Data captured by Mars Orbiter
Laser Altimeter (MOLA) are included here (Neumann 2001).
The different branches of passive remote sensing can be classified based on
wavelength. Wavelength (?) and frequency (?) are related through the velocity of light in
a vacuum (c), which is constant, by the following equation: c = ? ?. Remote sensing
commonly makes use of photography, which falls in the visual part of the light spectrum.
Other common wavelengths are ultraviolet and infrared (Table 6).
51
Table 6. Common remote sensing fields in the infrared section of the electromagnetic
radiation spectrum and their associated wavelengths.
Remote Sensing Field Wavelength
Near-infrared (NIR) 0.7-1.4 ?m
Mid-infrared (MIR) 1.4-3.0 ?m
Far-infrared (FIR) 3-1000 ?m
Thermal infrared (TIR) is a popular branch of remote sensing that falls within the
far-infrared region (Ramsey and Christensen 1998). Martian surface minerals can
broadly be classified by comparing the high-resolution thermal infrared spectra of surface
minerals on Mars to thermal infrared spectra of common rock-forming minerals found on
Earth (Christensen et al. 2000). To aid this process, Christensen et al. (2000) published a
preliminary database (available at http://rruff.geo.arizona.edu/doclib/hom/) containing
thermal emission spectra of common terrestrial rock-forming minerals. Thermal
emission spectra of rock-forming minerals are obtained from Earth-orbiting satellites
such as the Thermal Infrared Multispectral Scanner (TIMS) and the Advanced
Spaceborne Thermal Emission and Reflection radiometer (ASTER), as well as the series
of Earth Observing System (EOS) satellites, which all have similar spectral resolutions to
those in orbit around Mars (Wright 2003; Christensen et al. 2000; 2004).
Active remote sensing involves sending a signal to a surface and measuring the
time it takes the signal to be reflected back to the sensor. This is how altimetry data for
the generation of topographic maps are gathered (Neumann 2001). Knowing the radius
of the planet, the exact height of the sensor at the time, and the angle of the laser pulse,
52
the topographic height of the surface is determined through geometric principles (Abshire
et al. 2000; Neumann 2001).
Planetary exploration of Mars by means of spacecraft started in 1964 with the
launch of Mariner 4 (Kirk 2005). This spacecraft carried a camera that took about 20
close-up images of the planet. Mariner 6, 7, and 9 followed shortly and carried onboard
narrow and wide angle cameras as well as ultraviolet and infrared spectrometers. The
Viking orbiters and landers followed in 1975 and completely mapped the surface of Mars
for the first time from orbit (Kirk 2005). Only in 1996 did another spacecraft
successfully go to Mars. This time it was Mars Global Surveyor, carrying onboard the
MOC and MOLA instruments (Smith et al. 2001). Mars Odyssey is currently in orbit
around Mars and it carries a variety of remote sensing instruments, including a gamma
ray spectrometer, a radiation spectrometer, and a multi-wavelength thermal spectrometer
that can measure in both visual and infrared wavelengths (THEMIS). Fig. 15 shows four
of the spacecraft deployed in the remote study of the surface of Mars, one of which is still
operational today. Other current operational Mars orbiters include the Mars
Reconnaissance Orbiter (NASA) and Mars Express (ESA) (Carr and Garvin 2001; Kirk
2005).
a)
b)
c)
d)
Fig. 15. Spacecraft that have gone to Mars include Mariner 4, launched 28 November
1964 (a); Viking Orbiter Lander 1 and 2, launched 20 August 1975 and 9 September
1975 respectively (b); Mars Global Surveyor, launched 7 November 1996 (c); and Mars
Odyssey (still operating), launched 7 April 2001 (d). All images from NASA (available
from http://mars.jpl.nasa.gov/gallery/spacecraft/index.html).
Based on data gathered through remote sensing, scientists have been able to draw
conclusions about the morphology, mineralogy, and topography of Mars (e.g.,
Christensen et al. 2000; Aharonson 2001; Christensen et al. 2004; Boyce et al. 2005; Kirk
2005).
53
54
Summary
This section discussed the many aspects of the study of shallow-marine impacts
on Mars. First of all it is important to consider the evidence for water on the surface of
Mars, and then to assume that the proposed shorelines of ancient oceans did indeed exist.
Shallow-marine impact craters have distinct physical properties and characteristic
features, some of which can be measured and/or seen from data captured by orbital
sensors. Terrestrial analogs are few, yet hold important clues to the expected
morphologies for shallow-marine impact craters. Remote sensing is a crucial tool in the
study of planetary surfaces, particularly that of Mars, where humans are yet to set foot.
The study of shallow-marine impact craters on planetary surfaces - a new branch of
planetary geology - is just beginning to contribute to geoscience, and many facets of this
topic are yet to be exposed.
55
METHODS
This study employs datasets gathered by three different instruments. They are the
a) Mars Orbiter Laser Altimeter (MOLA) and b) Mars Orbital Camera (MOC), which are
onboard the Mars Global Surveyor (MGS) spacecraft, and c) Thermal Emission Imaging
System (THEMIS) onboard the Mars Odyssey spacecraft. The spatial resolutions of
these datasets are substantially better than that of the Viking images, which were
previously used in areomorphological studies of this nature (Kirk 2005). After discussing
the background of the missions involved in this study, the methods employed are
discussed in three parts: data acquisition, data analysis, and data interpretation. The data
acquisition section describes where the data are obtained from as well as the manners in
which the data are acquired and processed. This section also explains how the shallow-
marine impact crater candidates were chosen. The data analysis section deals with the
quantitative investigation, particularly in terms of the physical parameters of the
candidate craters. The data interpretation section categorizes the crater candidates by
evaluating the marine characteristics through the use of a standardized ranking system.
Background
Three of the most productive missions that have been sent to Mars are the Viking
orbiters (V1 and V2), Mars Global Surveyor (MGS), and Mars Odyssey (MO). A
timeline of events associated with these missions is outlined in Fig. 16. A variety of
highly successful imaging sensors has been incorporated into these missions, including
Mars Orbiter Laser Altimeter (MOLA), Mars Orbiter Camera (MOC), and Thermal
Emission Imaging System (THEMIS); however, THEMIS is the only sensor of the three
that currently is still operating.
V1
V2
MGS
MO
1975 1980 ? 1995 2000 2005 2007
Fig. 16. Timeline depicting the stages in Mars exploration by Viking Orbiters 1 and 2
(V1 and V2), Mars Global Surveyor (MGS), and Mars Odyssey (MO). White areas
indicate cruise time, and darker areas indicate sensor-operating time.
The main improvement of the MOLA, MOC, and THEMIS sensors over Viking
imagery is in the category of spatial resolution (Table 7). Viking imaging sensors
mapped the entire surface of planet in 150-300 m/pixel resolutions and only a few
selected areas at lower resolutions (Kirk 2005). MOC images are generally at 2-10
m/pixel resolutions when taken with the narrow angle camera (Malin et al. 1992).
THEMIS images are not quite as high-resolution, but still generally have resolutions of
56
57
around 20 m/pixel. MOLA resolution is of an entirely different kind and cannot be
directly compared to that of MOC or THEMIS because MOLA data are topographic
while MOC and THEMIS data are visual images. Topographic data by definition does
not have a spectral resolution, which is why this information is not listed for MOLA in
Table 7. MOC has the best radiometric resolution and is thus provides the most sensitive
sensor, followed by MOLA and then by THEMIS. All three sensors have low temporal
(related to frequency of image capture) resolutions because none of them were designed
to capture surface change.
Table 7. Instrument specifics in terms of resolution for MOLA, MOC, and THEMIS
(Malin et al. 1992; Abshire et al. 2000; Christensen et al 2002; Kirk 2005).
MOLA MOC THEMIS
Full Name
Mars Orbiter
Laser Altimeter
Mars
Orbiter Camera
Thermal Emission Imaging
System
Maximum Visual
Spatial Resolution
130 m 1.4 m 18 m
Spectral
Resolution
-
3 visual
5 visual
9 infrared
Radiometric
Resolution
16 bit 32 bit 8 bit
Temporal
Resolution
Low Low Low
Mars Global Surveyor
NASA launched Mars Global Surveyor (MGS) on November 7, 1996, which went
into orbit around Mars in September the following year (Smith et al. 2001). MGS carries
the Mars Orbiter Laser Altimeter (MOLA) and the Mars Orbiter Camera (MOC),
as well as two other scientific instruments (Abshire et al. 2000; Smith et al. 2001).
58
The main objective of the MOLA investigation was to generate an accurate global
topographic map of Mars with sufficient resolution to produce elevation data that can be
useful in studies of planetary geology at a large scale (Abshire et al. 2000; Smith et al.
2001). Shortly after MGS went into orbit, MOLA started transmitting altimetric data.
The MOLA spatial resolution is approximately 130 m/pixel and horizontal resolution
approximately 930 m/pixel (Smith et al. 2001). The vertical accuracy is approximately 1-
2 m (Roark et al. 2004; Howenstein 2006). Signal strength, pulse width, detector noise,
and background level are all factors that affect the accuracy of the signal (Abshire et al.
2000). Prior to MOLA, topographic measurements were based on photogrammetric
analyses of Viking stereo image pairs, with horizontal errors as large as a few kilometers
(Smith et al. 1998; Abshire et al. 2000; Kirk 2005).
The main objective of the MOC investigation is to obtain high-resolution imagery
of the surface with which more in-depth studies of areomorphological features can be
done (Malin et al. 1992). MOC has two wide-angle cameras, one in the blue part of the
visual spectrum and one in the red, with 280 m/pixel spatial resolution and a narrow-
angle camera with a maximum spatial resolution of 1.4 m/pixel.
Mars Global Surveyor is not operating anymore. MOLA sent its last signal in
2000 and MOC in 2006, yet the data that these instruments collected (MOC alone
collected more than 212,000 images; Malin Space Science Systems MOC database at
http://www.msss.com/) have not been fully analyzed and interpreted
59
Mars Odyssey
Mars Odyssey, launched by NASA on April 7, 2001, arrived at Mars in October
of that same year and is still operational. Odyssey carries onboard with it the Thermal
Emission Imaging System (THEMIS) as well as two other scientific instruments.
THEMIS has five visual bands with a spatial resolution of 18 m/pixel and nine infrared
bands with a spatial resolution of about 100 m/pixel. The main objective of THEMIS is
to compile a mineralogical map of the surface of Mars based on thermal infrared
spectroscopy (Christensen et al. 2000), but a large amount of visual imagery data are also
being gathered to fill the gap between the high-resolution MOC data and the lower
resolution Viking data. More than 78,000 visual images have been taken to date with
THEMIS, with a further 70,000 images taken in the infrared band of the spectrum
(Arizona State University THEMIS database at http://themis-data.asu.edu/).
60
Data Acquisition
The datasets are available either through an interactive website called Planetary
Interactive GIS on-the-Web Analyzable Database (PIGWAD), which is maintained by
the astrogeology team of the USGS, or through the Planetary Data System (PDS), which
is maintained by NASA (Hare and Tanaka 2000; Hare and Tanaka 2001; Hare et al.
2003). The Malin Space Science Systems (MSSS) host MOC images in an online
database, and often the MSSS database is easier to access than the PDS database due to a
simpler, interactive map interface. Arizona State University (ASU) host THEMIS
images in a similar online database and the images are also accessible through an
interactive map interface.
MOLA data
MOLA data are collected in vector format as binary points. The datasets contain
thousands of vector points recorded for each individual track. Altimetric data from
MOLA can be manipulated to yield profiles, and can also be used to calculate slope and
elevation statistics (Aharonson et al. 2001).
MOLA data as captured in the original format cannot be used directly in a
Geographic Information System (GIS). The data have to be processed and this requires
several steps as shown in Fig. 17 (Smith et al., 2001). The raw data are stored as
Aggregated Experiment Data Records (AEDRs), a Level 0 binary file product in a vector
format, which are produced directly from downlink telemetry (Smith et al. 2001). Level
1 processing applies orbit calibration factors to the AEDRs to produce Precision
Experiment Data Records (PEDRs). The output format is again a binary point vector.
The PEDRs are subsequently converted to a raster format that is called Experiment
Gridded Data Records (EGDRs) (Smith et al. 2001). EGDRs are stored in two forms,
Initial EGDRs and Mission EGDRs. MEGDRs are topographic maps generated from
Level 2 ASCII text files, and can be used to generate GIS shapefiles (Hare and Tanaka
2000).
Fig. 17. Different data products that can be derived from raw MOLA data. MEGDRs are
used in a GIS, and are therefore used in this study.
To interpret MOLA data as altimetric profiles, precision orbit information is
merged with the data to locate each spot on the planet and determine its radius at that
spot. Unlike the Earth, Mars does not have a standard sea level, but rather a reference
datum representing a mean surface elevation. This reference datum is referred to as the
areoid (comparable to the Earth?s geoid). The radius minus the areoid is the elevation.
61
62
MEGDRs usually consist of at least three, sometimes four, objects: a) a planetary radius
as measured by the MOLA instrument, b) an areoid model, c) a topographic map
computed as the difference between the radius and the areoid, and d) a count of the
number of MOLA hits per map unit.
The elevation data used in this study are MEGDRs, released through the PDS on
May 7, 2003, and are at 128 m/pixel resolution. All the global maps are in simple
cylindrical projection using a planetocentric, east-positive longitude coordinate system
assigned by the International Astronomical Union in 2000 (Kirk 2005). The 128 m/pixel
MEGDRs are stored as 16 image tiles, based on coordinates (Fig. 18). The location of
the study area lies well within two of these tiles that cover the area between 0 and 44
degrees north, and between 270 and 90 degrees east. For this study, the topographic
northern hemisphere tiles for areas bounded by 0?-44? N and 0?-90? E
(megt44n000hb.img) and by 0?-44? N and 270?-0? E (megt44n270hb.img) are
downloaded as Imagine image files. Along with their respective label files (also
downloaded from the PDs), the gridded records are imported into ArcGIS as base maps.
The vast amount of MOLA data products make it necessary to specify exactly
which particular files are used and why. File names for the global image maps use the
format of MEGpxxnyyyrv.IMG. Table i in the Appendix lists the symbols used in the
file path and their respective definitions. Topography files, as opposed to count files, are
all that are needed in this study because they are sufficient for drawing profiles in the
software program GRIDVIEW ? a scientific visualization tool developed by NASA?s
Goddard Space Flight Center specifically for use with gridded MOLA data (Roark et al.
2004).
63
88?N
Fig. 18. Tiling scheme for the 128 pixel/degree MEGDRs from the Planetary Data
System (PDS), indicating the two tiles that are used as base maps in this study.
A raster MOLA product from the USGS is also used as a base map for mapping
the location of the study area and the candidate craters. This raster file is a hillshade
alteration of original topographic data for the north pole and not a gridded data product
like the topography files mentioned above.
The coordinate system of the MOLA base map is an equidistant cylindrical
projection with central meridian at 180? and a datum called GCS_Mars_2000_sphere,
downloaded from PIGWAD.
44?N
44?S
0?
88?S
90? 0? 180? 270?
360?
64
MOC and THEMIS data
MOC and THEMIS datasets are collected in raster format from PIGWAD, MSSS,
and ASU. There are a few limitations to the visual data, including limited coverage of
the northern plains due to cloud cover. In addition, some errors in the transmitted
telemetry stream cause gaps in images, and even though some images can be corrected, a
few cannot, leaving parts of these images useless (Caplinger et al. 1999).
The Malin Space Science Systems (MSSS) website holds most of the MOC
images at http://www.msss.com/moc_gallery/. On this website, each biannual group of
orbits has a global map of Mars that is divided into 30 quadrangle sections (Fig. 19), and
sections can be entered by clicking on it. Each section then has an enlarged picture on
which there are hyperlinks pinned to each image in the location in which it was taken.
Images within the study area are captured by opening these hyperlinks and downloading
the data. Composite images and images with technical problems are excluded from the
dataset, and all images have center coordinates that fall strictly within the study area.
Furthermore, a few of the last images taken by MOC are not included in this study as
these were never uploaded to the MSSS website. Considering that these images were
taken in the last three months of a nine-year study, not too much emphasis is placed on
trying to locate these images.
65
Fig. 19. The study area falls within four of the central quadrangles in the Northern
Hemisphere ? Mare Acidalium, Oxia Palus, Ismenius Lacus, and Arabia. From Malin
Space Science Systems (MSSS).
Mare Acidalium Ismenius Lacus
Oxia Palus Arabia
The downloaded images are all processed and map-projected. As the images are
downloaded, the image identification number and center coordinates are entered into a
database. The database is used to organize the data collection, to simplify the process of
going back for THEMIS data, and to get a visual idea of the quantity of images in the
various areas. From the Arabia Terra region, the 868 available images are screened for
shallow-marine crater characteristics. Firstly, 85 useful images are identified. These
images show interesting features that could potentially be marine in origin. Secondly,
potential shallow-marine crater sites are documented in the database from the population
of useful images, based on the presence of shallow-marine crater characteristics such as
evidence for wet mass movement and resurge deposits, as discussed in the previous
chapter. For all the useful images, the original raw data are also downloaded in
66
.imq (different from the .img Imagine image files) file format so that the images could be
processed as described below.
Raw MOC data are stored as compressed Standard Data Products, or .imq files.
Raw data need to be decompressed, corrected, and projected for use in a GIS. A common
software application, Integrated Software for Imagers and Spectrometers (ISIS), is used
to areo-correct the imported planetary datasets (Hare et al. 2003). The compressed .imq
files are downloaded and radiometrically calibrated to give Level 1 image files.
Subsequently, these files are converted to map-projected Level 2 image files, which are
used as layers in a GIS. ISIS runs in the command line from a UNIX based operating
system. Once ISIS is installed, a few basic commands will spatially register the file to a
specified coordinate system and projection (Table ii in Appendix).
THEMIS images are hosted by ASU and can be accessed directly through an
interactive map on their Mars Odyssey website, or through PIGWAD. For all the
potential candidates, THEMIS visual images are also downloaded. The resolution of
THEMIS images is a little lower than that for MOC images, but the THEMIS images
cover a larger area. Thus, the THEMIS images are used to place MOC images in context
of location (e.g., de Villiers et al. 2006). To complete the image database, a second
search is done in PIGWAD to ensure that all available MOC and THEMIS imagery for
each candidate site has been acquired.
67
Data Analysis
GIS is particularly useful for this project due to its ability to overlay multiple
datasets such as MOC, MOLA, or THEMIS imagery or even a geological map and other
graphical datasets tied to a geodetic framework.
The topographic maps, or digital elevation models (DEMs) produced with MOLA
are used as base maps upon which high-resolution images can be overlain. Vector
features such as shorelines and crater locations are digitized as new layers directly from
the base map or from the high-resolution images. The images are spatially aligned with
the basemap through a process called georeferencing. This process involves picking
ground control points on the base map and on the image and manually aligning the new
image with the base map.
Using ArcGIS, the crater features are digitized and the crater depths and
diameters are calculated, providing the necessary comparative data on some of the
physical properties of Martian craters. Layers are created in ArcCatalog where spatial
properties are also defined for each layer, after which the layers are imported to ArcMap.
Once the coordinate system has been correctly defined and the map projection correctly
assigned, then the diameters of craters are directly measured with the measure tool.
Depths of craters are calculated by querying the point values inside and outside the crater.
Crater depths were measured at four locations both inside (depth of crater, or dc) and
outside (height of surroundings, or ds) the crater, as well as at four location on the rim
68
(height of the rim, or dr). The average of the four locations, generally located north, east,
south, and west of the crater, are used. The depth of the crater is calculated by
subtracting the height of the crater floor (dc) from the height of the rim (dr).
With the use of GRIDVIEW, profiles of exceptional candidates are drawn and
included in the results. To draw a profile in GRIDVIEW, a gridded Imagine (.img)
image file such as a MEGDR is uploaded and the coordinates and resolution defined.
The image can be enlarged to view a particular area and the color stretch can be
maximized to enhance topographical differences. Once the profile is drawn, the point
data are exported as a text file that is subsequently imported into a spreadsheet where the
profile is re-drawn and simplified for clarity.
Data Interpretation
Once the database of MOC images and their physical properties is compiled, a list
of shallow marine crater characteristics was constructed by which the craters would be
evaluated. The characteristics were chosen based on those reported in literature for
marine craters on Earth and from four Martian candidate craters identified by Orm? et al.
(2004). Orm? et al. (2004) identified four shallow-marine impact craters within Arabia
Terra, from Viking images. Three of the four candidates are located together at site 1
near 39?N and 349?W and the remaining candidate is located at site 2 which is at
approximately 39?N and 001?W (Fig. 20).
D
69
Fig. 20. The locations of Craters A, B, C, and D (as indicated by Orm? et al., 2004) in
Viking images F529A09 and F072A32, respectively. Images from the Planetary Data
System (PDS).
B
A
C
The candidate sites were analyzed in more detail with the use of high-resolution
MOC and THEMIS imagery. For site 1 in the eastern section of the study area, a series
of MOC and THEMIS images is available for Crater B, but not much is available for
either Crater A or Crater C. For site 2 in the center of the study area, a total of 15 MOC
images are available, but no THEMIS data. Fig. 21 shows the overlay of imagery that
was done for Crater B.
Legend
MOLA MDIM_2.1.jpg
Va l u e
High : 255
Low : 0
V125 940051
Va l u e
High : 255
Low : 0
5km 15km
(a)
Fig. 21. THEMIS raster image of a Crater B on top of MOLA DEM (a) and
georeferenced MOC image draped over THEMIS im
The following characteristics are observed
radial gullies entering the crater from one or m
deposits; d) central terraces; e) collapsed rims
mass movement occurs when saturated sedim
manner (Dypvik and Jansa 2003). Radial gull
70
Legend
THEMIS VISUAL.jpg
Value
High : 255
Low : 0
be0201413d1
Value
High : 255
Low : 21
(b)
age (b).
: a) evidence of wet mass movement; b)
ore location on the crater rim; c) resurge
; and f) subdued topography (Fig. 22). Wet
ent is mobilized, either in a brittle or ductile
ies are common in deeper continental
71
environments where the water resurge breaches the crater rim (Orm? and Muinonen
2000; Von Dalwigk and Orm? 2001). Resurge deposits are sometimes difficult to discern
from satellite imagery, but include blocks of rim material inside or outside the crater
(Dypvik and Jansa 2003). Central terraces indicate long-standing water, potentially
spanning from right after formation until the end of the oceanic phase (Gault and Sonett
1982; Dypvik and Jansa 2003). Collapsed rims are usually associated with wet mass
movement and resurge (Dypvik and Jansa 2003), but are different in the sense that wet
mass movement can occur without a breach in the rim. Rim collapse is usually
associated with radial gullies, however, it is also observed without the presence of
identifiable gullies. Finally, subdued topography does not only indicate that water may
have persisted on the surface for an extended period of time, but it is also the result of
weathering and therefore relates to the age of the crater population under examination. If
an ocean once existed in the northern lowlands on Mars, it occurred during the late
Noachian, very early in Mars?s history (Parker 1989; Fair?n et al. 2003), leaving plenty
of time for the craters to become degraded and the surrounding topography to become
subdued.
a)
b)
c)
d)
Fig. 22. Subdued topography and no elevated rim (a) (E0600102). Central terrace (b)
(V12594005). Large radial gullies (c) (R0500631). Large-scale wet mass movement (d)
(R1402326). Images from Malin Space Science Systems (MSSS) and PIGWAD.
72
73
The features identified in the craters proposed by Orm? et al. (2004) form the
basis of the quantification system in this study. Features reported in literature for
terrestrial analogs supplement these characteristics. Therefore, following from the
observed features listed above, the six categories in the candidate database are wet mass
movement; radial gullies; resurge deposits; central terraces; rim collapse; and subdued
topography.
The six categories are weighted to ensure that more important characteristics
carry more value. Wet mass movement is likely the most important characteristic,
however, it should be noted that there are numerous ways of forming wet mass
movement features and that these features do not only occur in shallow-marine
environments. Resurge gullies are likely the most indicative characteristic, however,
these features are more likely to form in deeper water (von Dalwigk and Orm? 2001) and
are thus not expected to be common in Arabia Terra.
Within the quantification system, wet mass movement comprises 40% of the
total. This includes 20% for evidence of slumping (small or large scale) and 20% for
evidence of debris flows (small or large scale). Radial gullies contribute a further 20%,
where one gully is 10%, and two or more gullies is 20%. Resurge deposits, central
terraces, rim collapse, and subdued topography each contribute 10% to the overall rating.
Fig. 23 shows the distribution of weight for each of the characteristics. These values
were chosen to reflect relative importance of various characteristic features. Wet mass
movement is an important, relatively indicative factor, and as such comprises 40% of the
score. Radial gullies, though not as common in shallow water as in deeper water, is also
indicative and therefore comprises a further 20% of the score. Resurge deposits are more
difficult to discern from orbital imagery, resulting in the factor only comprising of 10%
of the score. Rim collapse, central terrace, and subdued topography are all factors that
are less indicative of marine origin, and thus make up 10% each of the total score.
WMM Slump
WMM Flow
Radial Gullies
Resurge Deposits
Central Terrace
Rim Collapse
Subdued Topography
Fig. 23. Weighted distribution of the six classes of shallow-marine crater characteristics.
Note that wet mass movement (WMM) is split into two sub-categories.
With this system it is possible to crudely quantify the features observed and to
rate each crater based on properties associated with marine craters on Earth. Each
candidate crater is assigned a total score and these values can be used to rank and classify
types of shallow-marine crater candidates.
74
75
The database of shallow-marine crater candidates is included in this thesis in
electronic format along with all the images that are discussed on a CD. In addition to the
database, all images used in this thesis along with all figures are included on the attached
CD. Interpretation of the results includes a discussion on the physical properties and
characteristics of some exemplary candidates.
76
RESULTS
The results of the analysis of crater characteristics are discussed in detail, because
the crater characteristics are the parameters used in the quantification system. The
quantification system was designed to rank potential crater candidate sites and to assign
total scores indicating degrees of certainty of marine origin to these craters. As
mentioned before, it should be noted that none of the crater characteristics can be used
with certainty to apply marine origin, and as such, the total score assigned to a crater is
only an estimation of the potential for that crater to have been formed in a shallow-
marine environment. The candidate sites were analyzed in more detail with the use of
high-resolution MOC and THEMIS imagery. Three sets of analyses were done: a) a pilot
analysis to act as reference for subsequent sets; b) an initial analysis based on available
MOC imagery for the study area; and c) a second analysis based on MOLA topography.
The pilot analysis was done on the potential craters proposed by Orm? et al. (2004). The
initial analysis is referred to as Set A and the second analysis as Set B. The three
analyses are discussed separately, but all candidate craters are shown in Fig. 24. The
analysis of physical parameters were not the primary aim of this study, but do contribute
some value and are therefore briefly discussed at the end of this section.
All of the crater candidate sites fall north of (within) the Meridiani shoreline
(shown in red on Fig. 24) proposed by Edgett and Parker (1997), and only a few fall
north of (within) the Arabia shoreline (original Contact 2 ? shown in yellow). The
Deuteronilus shoreline (Parker 1989; Edgett and Parker 1997; Fair?n et al. 2003) is likely
the result of a smaller, more recent ocean and is indicated in blue in the figure. The
location of this shoreline is largely outside the study area and is therefore not of further
relevance in this study. The study area is indicated by the gray rectangle and can be seen
to encompass a large part of the Arabia Terra continental shelf.
Fig. 24. Locations of potential shallow-marine impact craters in relation to the three
shorelines proposed by Parker et al. (1989) as well as Edgett and Parker (1997). North
polar projection hill-shade base map is from the USGS and is based on MOLA
topographic data. Study area is shown by the gray rectangle.
77
78
Orm? et al.?s candidate crater results
The four craters proposed by Orm? et al. (2004) as potential shallow marine
impact craters were rated in the quantification system and the results are shown are in
Table 8. The total value is a number ranging from 0.0 to 1.0, reflecting the percentage
confidence that a particular crater is of marine-origin. All four of the craters attained
total scores of 0.5 or greater, yet only Crater D has a total score higher than 0.7.
Table 8. Evaluation of Orm? et al.?s (2004) potential shallow-marine crater candidates.
Characteristics of Orm? et al. Candidate Craters
Candidate WMM RG RD CT RC ST TOTAL
20% 20% 20% 10% 10% 10% 10%
(Slump) (Flow)
A 0.5 0 1 0 0 1 1 0.50
B 1 0 0.5 0 1 1 1 0.60
C 0.5 0 1 0 0 1 1 0.50
D 1 0.5 1 0 1 1 1 0.80
WMM ? Wet Mass Movement; RG ? Radial Gullies; RG ? Resurge Deposits; CT ? Central Terrace; RC ? Rim
Collapse; ST ? Subdued Topography
To visually interpret these results, a stacked histogram is plotted with all the
variables shown as individual blocks (Fig. 25). Total scores range from 0.5-0.8 as even
these prime examples do not exhibit all possible characteristics on the list. From Fig. 25
it is clear that Crater D is the most likely of the marine impact crater candidates proposed
by Ormo et. al (2004) to be a shallow-marine impact crater. A total score of 0.8 is
assigned to Crater D. Five of the six categories have features present in Crater D, and
both slumps and flows are observed in the Wet Mass Movement category. Resurge
deposits are not evident from the imagery, and this is likely due to the difficulty in
observing such a feature from orbital imagery. Crater D is discussed in the next chapter.
Orm? et al. Craters
-
0.20
0.40
0.60
0.80
1.00
ABCD
WMM Slump WMM Flow Radial Gullies Resurge Deposit Central Terrace Rim Collapse Subdued Topography
Fig. 25. Stacked columns with individual contributions to the overall rank for the craters
proposed by Orm? et al. (2004).
79
80
Results from Set A
In the database of MOC images in Arabia Terra there are 868 images that were
taken during the period September 1997 to September 2005. From these images, 86 are
classified as useful, and 65 of these images are potential candidates for marine targets
(see Table iii in the Appendix). After carefully analyzing these images, a residual of 51
proved to have some significance to this study. Some of the 51 images are of the same
crater, and thus there are a final 40 candidate craters which are identified from the
imagery. THEMIS images were subsequently downloaded for the 40 candidate sites.
The view of the candidate site was expanded with the use of lower resolution context
images provided by Malin Space Science Systems as well as available THEMIS images.
This proved to be useful particularly in larger craters where only parts of the structure are
visible in any one image.
Fifteen of the candidate craters were found to be too small (diameters <10 km) to
fall within the population for this study and were thus discarded. The 25 remaining
craters were divided into three groups based on size. These divisions were created
because the size of the crater could affect the morphology of the crater (Dypvik and Jansa
2003), and as such an effective assessment of craters can only be done if similar craters
are compared with each other. Small craters have diameters of 10-30 km, medium craters
have diameters from 30-50 km, and large craters have diameters 50-100 km. The 25
shallow-marine crater candidates were rated for shallow-marine origin and the craters and
their ratings are listed in Tables 9, 10 and 11.
81
Table 9. Evaluation of characteristics of small shallow-marine crater candidates.
Exemplary ranking candidates are shown in italics.
Characteristics of Small Candidate Craters
Candidate WMM RG RD CT RC ST TOTAL
20% 20% 20% 10% 10% 10% 10%
(Slump) (Flow)
1
0.5 1 0.5 0 0.5 1 0.5 0.60
2
1 1 0 1 0 1 0.5 0.65
3
1 1 0 0 0.5 1 1 0.65
5
1 0.5 0 0 0 0 1 0.40
6
1 0 1 0.5 0.5 1 1 0.70
7
1 0 0 1 0 1 0.5 0.45
9
1 0 0 1 1 0 1 0.50
14
0 0.5 0 0 1 0 1 0.30
15
1 0 0.5 0 1 1 1 0.60
16
1 0 0 1 0 1 1 0.50
17
1 0 0.5 1 1 1 1 0.70
18
1 0.5 0 1 0.5 0.5 1 0.60
19
1 0.5 0 1 0.5 0 1 0.55
24
1 1 1 1 1 1 0.5 0.95
27
1 0.5 0 0 0 0 1 0.40
30
0.5 0 0 0 1 0 1 0.30
32
1 0 0 1 0 0 1 0.40
37
0.5 0 0.5 0 0 0 1 0.30
38
1 0 0 0.5 1 1 1 0.55
40
1 0.5 0 1 0 1 1 0.60
WMM ? Wet Mass Movement; RG ? Radial Gullies; RG ? Resurge Deposits; CT ? Central Terrace; RC ? Rim
Collapse; ST ? Subdued Topography
Table 10. Evaluation of characteristics of medium shallow-marine crater candidates.
Characteristics of Medium Candidate Craters
Candidate WMM RG RD CT RC ST TOTAL
20% 20% 20% 10% 10% 10% 10%
(Slump) (Flow)
Medium
8
1 0 0 1 1 0 1 0.50
12
1 0 0 0 1 0 1 0.40
25
1 0 0.5 1 1 0 0.5 0.55
35
0.5 0 0 0 1 0 1 0.30
WMM ? Wet Mass Movement; RG ? Radial Gullies; RG ? Resurge Deposits; CT ? Central Terrace; RC ? Rim
Collapse; ST ? Subdued Topography
Table 11. Evaluation of characteristics of large shallow-marine crater candidates.
Characteristics of Large Candidate Craters
Candidate WMM RG RD CT RC ST TOTAL
20% 20% 20% 10% 10% 10% 10%
(Slump) (Flow)
26
0.5 0.5 0 1 0 1 1 0.50
WMM ? Wet Mass Movement; RG ? Radial Gullies; RG ? Resurge Deposits; CT ? Central Terrace; RC ? Rim
Collapse; ST ? Subdued Topography
In Fig. 26, ratings of all the craters from Set A, the craters are organized based on
size. The first twenty craters are small (D = 10-30 km), the next four are medium (D =
30-50 km), and the last one is large (D = 50-100 km). It is evident from the data that the
population is skewed to the smaller sizes. This occurrence is addressed in the next
section. Scores range from 0.2-0.95 with an average of 0.52. The three highest ranked
craters are Crater 6, Crater 17, and Crater 24, and all are small. These three craters, as
well as the associated ArcGIS overlays and GRIDVIEW profiles, are discussed in the
next chapter.
All Craters A
-
0.20
0.40
0.60
0.80
1.00
123567914151617181924273032373840812253526
WMM Slump WMM Flow Radial Gullies Resurge Deposit Central Terrace Rim Collapse Subdued Topography
Fig. 26. Stacked columns with individual contributions to the overall rank for the craters
in Set A.
82
83
Results from Set B
Due to the high resolution of the MOC imagery and smaller field of view (FOV),
much of the larger picture and context of the images as they are situated within the craters
were lost. The candidates within the initial database are all relatively small, mostly
ranging from 10-30 km in diameter. To expand this population, a second search was
completed, this time only by looking at the MOLA base map, which has a larger FOV.
From the second search, 48 additional potential candidates were identified. Candidate
craters from Set B are larger in size and more similar to the potential candidates
identified by Orm? et al. (2004).
High-resolution data (both MOC and THEMIS) were downloaded for the new
potential sites (165 images in total). Originally, 54 new potential sites were identified,
but 4 of these exceed the population size and 2 are highly modified in the recent past, to
the point where shallow-marine features could not be clearly distinguished. Recent
modification is evident from minimal erosion resulting in sharper surface features. The
candidates in the second set were judged by the same system of quantification as those in
the first and were again classified based on size. The results are listed in Tables 12, 13
and 14. From Set B there are four small craters (D = 10-30 km), twenty-nine medium
craters (D = 30-50 km), and fifteen large craters (D = 50-100 km).
84
Table 12. Evaluation of characteristics of small shallow-marine crater candidates.
Characteristics of Small Candidate Craters
Candidate WMM RG RD CT RC ST TOTAL
20% 20% 20% 10% 10% 10% 10%
(Slump) (Flow)
42 0.5 1 0 0 0 1 1 0.5
43 0 0 1 0 0 0 1 0.3
44 0.5 0 0 1 0 1 1 0.4
77 1 0.5 0 0 0 1 1 0.5
WMM ? Wet Mass Movement; RG ? Radial Gullies; RG ? Resurge Deposits; CT ? Central Terrace; RC ? Rim
Collapse; ST ? Subdued Topography
Table 13. Evaluation of characteristics of large shallow-marine crater candidates.
Exemplary ranking candidates are shown in italics.
Characteristics of Large Candidate Craters
Candidate WMM RG RD CT RC ST TOTAL
20% 20% 20% 10% 10% 10% 10%
(Slump) (Flow)
47 1 0.5 0 0 1 1 1 0.6
48 0 0.5 0 1 1 1 1 0.5
49 1 0 0 0 0 1 1 0.4
58 0.5 1 0.5 1 1 1 0.5 0.75
60 1 0 0 0 0.5 0 0.5 0.3
64 1 0 1 0 0.5 0 1 0.55
65 1 1 0 0 0 0 1 0.5
69 1 1 0 0 1 0 1 0.6
70 0.5 0.5 0 0 0 0 1 0.3
73 0 0 0 0 0 0 1 0.1
75 1 0 0 0 0.5 0.5 1 0.4
79 0 0 0 0 0 0 1 0.1
81 0.5 0.5 0 0.5 0 0.5 1 0.4
85 0.5 0.5 0 0 0 0.5 1 0.35
92 0 0 0 0 0 1 1 0.2
WMM ? Wet Mass Movement; RG ? Radial Gullies; RG ? Resurge Deposits; CT ? Central Terrace; RC ? Rim
Collapse; ST ? Subdued Topography
The largest group of craters in set B is the medium craters. This is expected since
the Orm? et al. (2004) craters, that were also medium in size, were the examples after
which the search was modeled.
85
Table 14. Evaluation of characteristics of medium shallow-marine crater candidates.
Exemplary ranking candidates are shown in italics.
Characteristics of Medium Candidate Craters
Candidate WMM RG RD CT RC ST TOTAL
20% 20% 20% 10% 10% 10% 10%
(Slump) (Flow)
41 0.5 0.5 0 1 1 0.5 1 0.55
45 0.5 1 0 1 1 1 1 0.7
46 1 0 0 0.5 0.5 0 1 0.4
50 1 0 0 0 1 0 1 0.4
51 0.5 1 0 1 1 0 1 0.6
52 1 1 0 1 0.5 0 1 0.65
53 1 0 0 0 0.5 1 1 0.45
54 1 1 0 1 0 1 1 0.7
55 0.5 1 1 0 0 1 1 0.7
56 1 0 0.5 0.5 0.5 1 0.5 0.55
57 1 0 0 0 0 0 1 0.3
61 0 0.5 0.5 0.5 0 1 1 0.45
62 0 0 0 0 0 1 1 0.2
66 1 1 0 1 0 1 1 0.7
68 1 0 0 0 0 1 1 0.4
72 0 0 0 0 0 1 1 0.2
78 0 0 0 0 1 0 1 0.2
80 0 0 0 0 1 0 1 0.2
82 0 0 0 0.5 0 0 1 0.15
83 0 0 0 0.5 0.5 0 1 0.2
84 0 0.5 0 1 0.5 1 1 0.45
86 0 0 0 0.5 0.5 1 1 0.3
87 0 0.5 0.5 1 0.5 1 1 0.55
88 1 0.5 0.5 0 0.5 1 1 0.65
89 0 0 0 0.5 0 0 0.5 0.1
90 0.5 0.5 0.5 1 0 1 1 0.6
91 0 0 0 1 0 1 1 0.3
93 1 0.5 0 1 0 1 0.5
94 0 0 1 0 0 1 1 0.4
WMM ? Wet Mass Movement; RG ? Radial Gullies; RG ? Resurge Deposits; CT ? Central Terrace; RC ? Rim
Collapse; ST ? Subdued Topography
The global scale of the MOLA imagery and the large sizes of the craters make
judging the second set of potential candidates challenging. Therefore, here as before,
problems were encountered due to scale. The global scale of the imagery means that the
resolution is too low to judge individual features such as slumps and flows. In addition,
the larger the craters, the smaller parts of the crater is covered by a single image. This is
a problem because entire features cannot be seen and the context of features is lost. The
available imagery does not sufficiently cover the potential sites to assign total scores with
certainty, and therefore the average total scores are lower for the second set than for the
first. Fig. 27 shows the ratings of all the craters in Set B. Total scores range from 0.1-
0.75 with an average of 0.42. The highest ranked crater is Crater 58, with four craters
(Crater 45, 54, 55, and 66) following closely. Crater 58 is large, but the other four top-
rated craters are all medium in size. Crater 66 falls within both Arabia and Meridiani
shorelines, thus the probability that it formed in a marine environment is much higher
than for any of the other top-rated candidates in this set. Craters 45, 54, and 55 are all
within close proximity of the Arabia shoreline. These three craters are grouped together
as typical shallow-marine impact craters. These shallow-marine craters, as well as
Craters 58 and 66, are discussed in the next chapter.
All Craters B
-
0.20
0.40
0.60
0.80
1.00
41 43 45 47 49 51 53 55 57 60 62 65 68 70 73 77 79 81 83 85 87 89 91 93
WMM Slump WMM Flow Radial Gullies Resurge Deposit Central Terrace Rim Collapse Subdued Topography
Fig. 27. Stacked columns with individual contributions to the overall rank for the craters
in Set B.
86
Exemplary candidates
From the 77 craters in the database, nine are chosen for further discussion based
on ranking. Craters D, 6, 17, 24, 45, 54, 55, 58, and 66 (see electronic appendix for full
database) have total scores of 0.7 or higher, and are the highest-ranking examples in this
study. All of these craters exhibit signs of slumping, rim collapse and subdued
topography. Furthermore, 77% of these show signs of debris flows and resurge deposits,
and 66% have radial gullies and/or central terraces. These data are shown in Figures 28
and 29, where pie charts for each of the craters are drawn. Wet mass movement is
divided into two categories: slump (WMM-S) and flow (WMM-F). The sizes of the
exemplary candidates range from 20 to 60 km in diameter, with most of the craters falling
in the medium (30-50 km diameter) range. Profiles of these craters are included in the
next chapter.
Fig. 28. Breakdown of characteristics present in Crater D and legend with color-coded
list of features.
Crater D
WMM-S WMM-F
RG RD
CT RC
ST Remainder
87
Fig. 29. Breakdown of characteristics present in the exemplary candidates (see legend in
Fig. 28).
Crater 6
Crater 17
Crater 24
Crater 45
Crater 54
Crater 55
Crater 58
Crater 66
88
89
Physical parameters that were analyzed in this study include diameters, depths,
and aspect ratios or depth-diameter relationships. The accuracy with which these
parameters are measured depends directly on the resolution of the MOLA data. Depths
were measured by subtracting the average height of the crater floor (dc) from the average
height of the crater rim (dr). Heights of the surrounding topography (ds) were also
measured, and when compared to the rim heights, it seems that there may be some
limitations to the use of this parameter. Rim heights are expected to be elevated or at
least at the same height as the surrounding topography, yet, in four of the nine crater
sites, this is not the case. This could be the result of local variations in topography and an
error in the data and/or method is not necessarily implied. Further investigation is needed
to accurately assess the depths of the craters in this study. Table 15 lists the depths,
diameters, and depth-diameter ratios (d/D ratios) as measured and calculated for the
exemplary candidates. Note that the d/D ratios range from approximately 0.01 to 0.03,
which is much lower than the ratios of 0.07 to 0.09 measured for all craters in Arabia
Terra by Barlow (1993).
Table 15. Physical parameters as measured from MOLA data for exemplary candidates.
All depths are negative, in other words, below mean surface level (see text for
explanation of abbreviations).
Crater
ds
(m)
dr
(m)
dc
(m)
Depth
(m)
Diameter
(km)
Depth/Diameter
ratio
D 3493 3599 3746 147 37 0.0040
6 2217 2325 2874 549 26 0.0211
17 2039 2022 2394 372 23 0.0162
24 3769 3698 4375 677 21 0.0322
45 2242 2152 2783 631 32 0.0197
54 3003 3468 4135 667 36 0.0185
55 3142 3411 4155 744 36 0.0207
58 2098 1902 3051 1148 60 0.0191
66 3362 3642 4387 745 32 0.0233
90
Using the depth-diameter relationships defined by Howenstein (2006) and Garvin
et al. (2000), one can estimate the depth by substituting the diameter. For the relationship
defined by Howenstein (2006), d = 0.61D
0.33
, the estimated depths (A; see Table 16) are
much higher than the depths estimated (B; see Table 16) with d = 0.19D
0.55
, as defined by
Garvin et al. (2000) for non-polar craters. Estimated depths for the exemplary candidates
are listed in Table 16. Garvin et al.?s (2000) relationship (B) more closely predicts the
depths measured in this study. The factors with which the estimated depths differ from
the measured depths are also listed in Table 16. Crater D stands out as potentially
anomalous with nearly an order of magnitude difference between the measured value and
the estimated value. The rest of the depths vary consistently with a factor around two,
possibly explained by faster rates of infilling for craters in shallow-marine environments.
Table 16. Depths as measured from MOLA data for exemplary candidates and estimated
depths as calculated from depth-diameter relationships in literature (Garvin 2000;
Howenstein 2006). Factors of difference (see text) are given for each estimation.
Crater
Depth
(m)
Estimated depth A
(m)
Factor
Estimated depth B
(m)
Factor
D 147 2008 12 1384 9
6 549 1788 3 1140 2
17 372 1717 5 1066 3
24 677 1666 3 1014 2
45 631 1914 3 1278 2
54 667 1990 3 1364 2
55 744 1990 2 1364 2
58 1148 2356 2 1806 2
66 745 1914 2 1278 2
The results of this study are included in the format of spreadsheets on a CD-ROM
that accompanies this thesis. All images and figures are also included.
91
INTERPRETATIONS
My shallow-marine impact crater candidate database contains the four craters
identified by Orm? et al. (2004), 25 craters identified in Set A, and 48 craters identified
in Set B. From this database, a few exemplary candidates were chosen and their
locations relative to the proposed shorelines are shown in Fig. 30. In this figure, the blue
line represents the location of the Deuteronilus shoreline, and the yellow line represents
the location of the Arabia shoreline. The Meridiani shoreline runs further southeast and
all craters fall within its boundaries. The low depth-diameter (d/D) ratios of the
exemplary candidates could be an indication of age. Older craters have rims that have
been more eroded and crater shapes that are shallower due to sediment infilling (Reiss et
al. 2006). The d/D ratios (refer back to table 15 in Results) for the exemplary candidates
are lower than expected, yet this can be interpreted as a result of the old age (Noachian)
of the craters in the population of potential candidates. Furthermore, resurge deposits and
post-impact sedimentation from the water column would fill the crater faster than craters
on land, and could therefore be responsible for the low crater depths. Estimated depths
vary greatly from the measured depths (refer back to Table 16 in Results), possibly due to
the difference in crater populations. Garvin et al. (2000) considered all non-polar craters
and Howenstein (2006) considered all large craters. The exemplary candidates would
therefore be shallower than average craters on the surface.
The exemplary candidates are divided into three classes based on total scores and
locations. Type I includes well-developed, mostly medium-sized examples; type II
includes typical small and medium examples; and type III includes other large potential
candidates.
Fig. 30. Locations of exemplary candidates on MOLA topographic raster image.
92
Type I candidates
Crater D
Orm? et al. (2004) originally identified Crater D as a potential marine crater,
based on certain morphological features observed by looking at Viking imagery. Crater
D is located at roughly 39.00?N and 001.00?W and well within the Meridiani shoreline as
defined by Edgett and Parker (1997) and Fair?n et al. (2003).
This candidate is more closely evaluated with the use of several MOC images.
The MOC images are draped over the MOLA base map in ArcGIS to create a layered,
composite, context images with high-resolution details (Fig. 31).
Fig. 31. Composite image for Crater D with MOLA base map and MOC wide and
narrow-angle images as high-resolution overlays.
93
A
B
Fig. 32. A close-up of Fig. 30 showing MOC narrow angle image overlay in more detail.
Radial gullies are indicated by white dashed lines and features labeled A-B are discussed
in text.
94
95
Of the potential marine craters suggested by Orm? et al. (2004), Crater D has the
highest total score of 0.8. In Fig. 32, the radial gullies are the most remarkable features
of this candidate crater (indicated by white dashed lines). One large gully enters the
crater from the southwest and is likely responsible for the large resurge deposit (A)
located in the southwestern section of the crater. Features of wet mass movement are
also evident in both the form of slumping of the crater wall as well as the flow of some
debris flows, particularly in the southern part of the rim (B).
Crater D is roughly 37 km in diameter and hence falls in the medium class (30-50
km diameter). The depths between the surrounding topography and the bottom of the
crater floor vary greatly, but average around 200 m. A north-south profile of Crater D is
shown in Fig. 33. The d/D ratio was calculated to be 0.004. Even though low values are
expected for marine craters on Mars, this value could be anomalously low considering
that the average d/D ratio for the exemplary candidates is 0.02. As mentioned earlier in
this chapter, low d/D ratios are indicative of old age and large amounts of sedimentation.
Impacts into marine environments should exhibit more sediment infilling than impacts on
land. Resurge and tsunami deposits could contribute largely to this rapid infilling;
beyond regular fast rates of burial.
Fig. 33. Profile view of Crater D showing varying depth (in km) with variation in
latitude (in degrees) from north (left) to south (right). Vertical exaggeration is 21:1.
From the profile shown in Fig. 33 one can see the terraced central uplift and what
seems to be a central ring. The subdued topography of the crater walls is clear on the
northern rim. Also visible in the southern half of the profile is a terraced resurge unit,
likely to have been deposited by a large gully entering from the southwest.
96
Crater 24
Crater 24 is located at roughly 38.39?N and 006.02?W, and well within both the
Arabia and Meridiani shorelines. MOC images are draped over THEMIS images, which
in turn are overlaid on top of the MOLA base map to create a layered, composite, context
images with high-resolution details (Fig. 34).
Fig. 34. Composite image for Crater 24 with MOLA base map, THEMIS image and
MOC narrow-angle images as high-resolution overlays.
97
98
Crater 24 has a total score of 0.95 ? the highest ranking of all craters in the
database. This crater exhibits all the characteristics for a shallow-marine candidate,
except that the topography is not as subdued as expected. This casts some doubt on the
age of this crater. Thus, even though the confidence rating is high, it is arguable that this
crater did in fact form during the Noachian. However, in terms of morphology, this
candidate exhibits some good examples of the characteristics that are thought to be
present in shallow-marine impact craters (see Fig. 35). Deposits that have a possible
resurge origin are the most remarkable features of this candidate crater (A). Gullies seem
to enter the crater from the northwest as well as the north northeast (shown in white
dashed lines) and are likely responsible for the large resurge deposits located in the both
these areas. Large amounts of slumping can be seen around the crater rim (B), and in
some places the rim has started to collapse (D). Debris flows dominate the eastern half of
the crater (E). Some of the flows, particularly in the south southeast, are much younger
than the crater itself and are the result of subsequent draining into the crater. Large
valleys can be seen in the south southeast and these should not be confused with the
flows related to the impact such as those further north. A central terrace is present in this
crater (C), but large amounts of post-impact sediments seem to cover most of the intra-
crater terrain.
A
B
C
D
E
Fig. 35. A close-up of Fig. 34 showing MOC narrow angle image overlay in more detail.
Radial gullies are indicated by white dashed lines and features labeled A-E are discussed
in text.
99
Crater 24 is roughly 21 km in diameter (the smallest of the exemplary candidates)
and hence falls in the small class (10-30 km diameter). A north-south profile of Crater 24
is shown in Fig. 36. From the profile shown in Fig. 36 one can see that this crater is
fairly deep and that the rim is still somewhat elevated above the surrounding areas. A
thick sedimentary unit is visible in the southern half of the profile, and though it is
possibly a resurge unit resulting from the gully to the south southwest, it is impossible to
exclude post-impact sedimentation from younger drainage systems. This resurge unit
may be responsible for the low d/D ratio (0.03) calculated for this crater. Regardless of
the fact the d/D ratio is low compared to that of other craters on Mars, it is still the
highest d/D ratio measured in this study, indicating that this crater may be younger than
the other exemplary craters.
Fig. 36. Profile view of Crater 24 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 9:1.
100
Crater 66
Crater 66 is located at roughly 38.21?N and 002.48?W, and lies within both
Arabia and Meridiani shorelines. THEMIS images are draped over the MOLA base map
to create a layered image with high-resolution details (Fig. 37).
Fig. 37. Composite image for Crater 66 with MOLA base map and THEMIS images as
high-resolution overlays.
Crater 66 has a total score of 0.7. Particularly striking in Fig. 38, are the features
of wet mass movement. Both slumps and flows are present in this crater, particularly
near the southern and northern parts of the crater (A and B). This is in accord with the
general location of the crater in relation to the shoreline. According to the
paleogeography, the resurge waters would have entered the crater from the north and
potentially affected the northern walls more severely than the eastern or western walls.
The crater?s wall has collapsed in the southernmost parts of the crater (C).
101
102
A
E
C B
D
Fig. 38. A close-up of Fig. 36 showing THEMIS image overlay in detail. Features
labeled A-E are discussed in the text.
Topography in and around Crater 66 is subdued and the rim is hardly elevated
above average surface height (see D in Fig. 38). The central peak ring seem also to have
been affected by resurge waters by having the entire northern half obliterated (E).
Crater 66 is roughly 32 km in diameter and therefore falls in the medium class
(30-50 km diameter). Crater 66 has a low d/D ratio (0.02), which is average for this
population of craters. A north-south profile of Crater 66 is shown in Fig. 39.
Fig. 39. Profile view of Crater 66 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 8:1.
Slumping is evident from the profile shown in Fig. 39, particularly in the northern
parts of this crater. One can clearly see a topographic high located in the southern half,
but this uplift is only the southern part of the central peak ring terrace. The resurge
waters have likely completely eroded the northern part of the central peak ring. The
103
104
sedimentary deposit directly north of this collapsed inner ring probably formed as a result
of water resurge from the north, leaving blocks of rim and other material on the crater
floor. The location of the gully that may have dumped this sediment is uncertain, but the
water may have been too shallow to allow gully formation. The depths between the
surrounding topography and the bottom of the crater floor vary between 300 m in the
south to 600 m in the north. This supports the observation of structural rim failure and
subdued topography in the southern crater wall.
Type II candidates
Crater 6
Crater 6 is located at roughly 35.34?N and 350.75?W. MOC images are draped
over THEMIS images, which in turn are overlaid on top of the MOLA base map to create
a layered, composite, context images with high-resolution details (Fig. 40).
Fig. 40. Composite image for Crater 6 with MOLA base map, THEMIS image, and
MOC narrow-angle images as high-resolution overlays.
Crater 6 has a total score of 0.7. As seen in Fig. 41, slumping is evident all
around the rim (A), but no debris flows are present. The central terrace is not well
developed and may be covered in post impact sediments. A few small gullies are present
(shown with white dashed lines), but they are not distinct and possibly even post-impact.
105
It is mainly the rim collapse sedimentary deposit (B) in the northwest corner of the
structure that indicates some resurge.
B
A
A
B
B
A
Fig. 41. Close-up of Fig. 40 showing MOC and THEMIS image overlays in detail.
Features labeled A-B are discussed in the text.
106
Crater 6 has a diameter of approximately 26 km and falls in the small class (10-30
km diameter). The d/D ratio for Crater 6 is approximately 0.02; around average again for
the exemplary candidates. A north-south profile of Crater 6 is shown in Fig. 42.
Fig. 42. Profile view of Crater 6 showing varying depth (in km) with variation in latitude
(in degrees) from north to south. Vertical exaggeration is 10:1.
From the profile shown in Fig. 42, the terraced central uplift is the most
prominent feature. Furthermore, large slumps are present in the northern half of the
structure. The subdued topography of the crater walls is clear on the southern rim.
107
Crater 17
Crater 17 is located at approximately 37.43?N and 354.30?W. MOC images are
draped over THEMIS images, and both sets of images are draped over the MOLA base
map to create a layered, context images with high-resolution details (Fig. 43).
Fig. 43. Composite image for Crater 17 with MOLA base map, THEMIS image, and
MOC narrow-angle images as high-resolution overlays.
Crater 17 has a total score of 0.7. This crater has been filled with post-impact
sediments, but some of the characteristic features are still present in this structure. Fig.
44 shows the large, shallow, filled crater. Most of the central terrace is covered in
sediments, and resurge sediments are present just south of the northern rim (A).
Topography is subdued and the crater rim is not elevated above the surrounding
108
109
topography (B). The rim is breached in more than one place, and the breach in the north
is possibly responsible for the sedimentary resurge deposit just south of the rim (A). The
remains of what might have been a gully entering the crater from the north can also be
seen in this area (Fig. 44). Features of wet mass movement are also evident in the form
of slumping of the crater wall, evident particularly in the east.
110
A
B
Fig. 44. Close-up of Fig. 43 showing MOC image overlay over THEMIS. Features
labeled A and B are discussed in the text.
Crater 17 is roughly 23 km in diameter and falls in the small class (10-30 km
diameter) along with Crater 6. Furthermore, similar to Crater 6, the d/D ratio for Crater
17 is approximately 0.02. Fig. 45 shows a north-south profile of Crater 17.
Fig. 45. Profile view of Crater 17 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 29:1.
From the profile shown above the subdued topography of the crater walls is clear
on both the northern and southern rims. Some deposits are evident in the center of the
crater. Judging by the appearance of the crater floor in the images, these deposits are
more likely the result of post-impact sedimentary fill than syn-impact resurge.
111
Crater 45
Crater 45 is located at roughly 36.18?N and 351.96?W. THEMIS images are
draped over the MOLA base map to create a layered, high-resolution image (Fig. 46).
Fig. 46. Composite image for Crater 45 with MOLA base map and THEMIS images as
high-resolution overlays.
Crater 45 has a total score of 0.7. The slumped resurged deposits (A) visible in
the northern and southern parts of the structure (Fig. 47) could be post-impact. The
smooth appearance of the crater floor clearly indicates a second, more recent
sedimentation event. This could be from drainage into the crater basin. No drainage
channels are visible, which suggests that the sedimentation might be from water rising
from below the structure. If sediments are fluidized at a later stage from below, slumping
112
113
and flowing of the rim wall would be expected. It is therefore uncertain if the sediments
were deposited at the time of impact by a resurge event. However, it seems plausible that
this deposit could have been related to resurge, based on its relative location. The
proposed ocean would lie north of the crater, resulting in forceful waves coming from the
northerly direction. Deposits in the center of the crater (B) may or may not be impact-
related. Relative spatial location suggests some link to the possible resurge deposit in the
north. The crater is not elevated above the surrounding topography, indicating that either
no rim formed initially, or that the rim has been modified and eroded, possibly through
fluids related to resurge activity and/or long-standing water covering the crater. The
breach in southern part of the rim could be part of rim collapse due to sediment saturation
(C).
114
A
B
C
Fig. 47. Close-up of Fig. 46 showing THEMIS image overlay. Features labeled A-C are
discussed above.
Crater 45 is roughly 31.5 km in diameter and hence falls in the medium class (30-
50 km diameter). Fig. 48 shows a north-south profile of Crater 45. It is clear that a
resurge or slump deposit is located in the northern parts of the crater. Judging by the
visual images, the spike in the southern wall is not representative of the whole rim.
Fig. 48. Profile view of Crater 45 from north to south. Vertical exaggeration is 13:1.
115
Crater 54
Crater 54 is located at roughly 34.94?N and 005.24?W. One THEMIS image is
draped over the MOLA base map to create a layered image with selected parts in high
resolution (Fig. 49).
Fig. 49. Composite image for Crater 54 with MOLA base map THEMIS image as a
high-resolution overlay.
Crater 54 has a total score of 0.7. Only one high resolution THEMIS image could
be draped over the MOLA topography, but from the composite image in Fig. 50,
numerous features could be identified. Features of wet mass movement are evident in
both the form of slumping of the crater wall (A) as well as the flow of some debris flows.
116
117
A large well-developed debris flow tongue (B) is visible in the north northeast of the
crater, and numerous smaller flows are also visible in the THEMIS images. The crater
rim is not elevated above the topography, and in some places, the rim has collapsed (C).
No gullies are observed in this crater, and no central terrace is evident. Some structural
blocks are found on the crater floor (D), possibly deposited as rim material being washed
into the crater as the water returns.
118
B
A
C
D
Fig. 50. Close-up of Fig. 49 showing THEMIS image overlay. Features labeled A-D are
discussed in the text above.
Crater 54 is 36 km in diameter (very similar to Crater D) and hence falls in the
medium class (30-50 km diameter). It is also close to Crater D in relative location (Fig.
58). A north-south profile of this crater is drawn in Fig. 51. The depths between the
surrounding topography and the bottom of the crater floor vary greatly, with a maximum
of 1500m. The central uplift and the subdued topography are the two features that are
most evident from the profile. Some slumping is apparent in the southern rim area and
sediment fills the northern half of the crater, possibly induced by debris flow.
Fig. 51. Profile view of Crater 54 from north to south. Vertical exaggeration is 6:1.
119
Crater 55
Crater 55 is located at approximately 36.37?N and 001.29?W. MOC images are
overlaid on top of the MOLA base map to create a layered image with selected parts in
high resolution (Fig. 52).
Fig. 52. Composite image for Crater 55 with MOLA base map and MOC narrow-angle
images as high-resolution overlays.
120
121
Crater 55 has a total score of 0.7, as most of the craters in category type II. In
Fig. 52, the radial gullies are the most remarkable features of this candidate crater (shown
in dashed white lines). The gullies are not visible in any of the MOC images, but the
MOLA topography has a high enough resolution to show these features. One large gully
enters the crater from the southwest and is likely responsible for the large resurge deposit
located in the southwestern section of the crater. The central terrace is shown in Fig. 53,
indicated as feature A. Slumping is not common in this crater, but wet mass movement is
observed in the form of debris flows in the south (B). Rim collapse is evident wherever
gullies enter the crater depression, and also directly east of the crater?s center (refer back
to Fig. 52).
122
A
B
Fig. 53. Close-up of Fig. 52 showing MOC image overlay. Features labeled A and B are
mentioned in the text above.
Crater 55 is also roughly 37 km in diameter, similar to Crater D and 54, and thus
joins these craters in the medium class (30-50 km diameter). A profile of this crater is
shown in Fig. 54. Craters 45, 54, and 55 all have low d/D ratios of around 0.02.
Fig. 54. Profile view of Crater 55 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 17:1.
From the profile shown in Fig. 54 one can see a central uplift as well as the
evidence of slumped units on both the northern and southern parts of the crater rim. The
average change in elevation between the surrounding topography and the crater floor is
700m. The southern half of the profile also shows a terraced sedimentary unit, likely
deposited by the gully entering from the southwest.
123
Type III candidates
Crater 58
Crater 58 is located at roughly 27.38?N and 353.95?W, the farthest south of all of
the exemplary candidates. It is in a class of its own because of its large size and because
of its relative location to the shorelines and other candidates. It falls within the Meridiani
shoreline, but is not even close to the Arabia shoreline, thus decreasing its chances of
having formed in an oceanic setting. THEMIS images are overlain on top of the MOLA
base map to create a layered image with high-resolution features (Fig. 55).
Fig. 55. Composite image for Crater 58 with MOLA base map and MOC wide and
narrow-angle images as high-resolution overlays.
124
125
Crater 58 has a total score of 0.75. Despite its southern location, numerous
indicative features are found in this structure Fig. 56 shows the resurge unit that enters
the crater structure from the north (labeled A). This resurge unit has been subsequently
modified and eroded. This is in accord with assumed age of the craters in this crater
population. Radial gullies seem to enter the crater from a few different areas, mainly
from the north (shown with white dashed lines). The MOLA topography indicates a
large low-lying area directly to the north of this crater, and the rim in this vicinity is
totally collapsed. Rim collapse indicates resurge, and the relative location of the crater
with respect to the proposed shorelines make this a likely scenario. Features of wet mass
movement are also present in both the form of slumping of the crater wall as well as
debris flows, but these features are not well defined in this crater.
126
A
B
Fig. 56. Close-up of Fig. 54 showing MOC and THEMIS image overlay on top of
MOLA topography. Features labeled A and B are discussed in the text. Radial gullies
are indicated with white dashed lines.
Crater 58 is roughly 60 km in diameter and is the only exemplary candidate that
falls in the large class (50-100 km diameter). A north-south profile of Crater 58 is shown
in Fig. 57. From the profile shown in Fig. 57 one can see resurge deposit in the northern
half of the crater. The rim here is largely destroyed. This resurge unit is likely
responsible for the crater?s low d/D ratio of around 0.02. The topography is subdued,
particularly in the north.
Fig. 57. Profile view of Crater 58 showing varying depth (in km) with variation in
latitude (in degrees) from north to south. Vertical exaggeration is 15:1.
127
Summary
Spatial correlation with the different types of candidates was not investigated in
detail. However, based on only the nine exemplary candidates, there seems to be a direct
correlation between location of the crater and its rating. Type I candidates are located in
close proximity to the Arabia shoreline; and type II candidates are located a small
distance away from the Arabia shoreline, in the direction of the Meridiani shoreline (see
Fig. 58). In Fig. 58, Type I candidates are shown in blue, Type II candidates in green,
and the single Type III candidate in red. It is impossible to comment on the spatial
distribution of type III craters since there is only one. Based on the spatial distribution of
the exemplary candidates, it seems more likely that the Arabia shoreline, instead of the
Meridiani shoreline, was in fact the shoreline of the ancient Noachian ocean.
Fig. 58. Spatial distribution of candidate craters shown by type. Type I craters are
shown in blue, type II craters shown in green, and type III craters shown in red.
128
129
Low depth-diameter ratio values were calculated for all of the exemplary
candidates. The values may be anomalous, but low values are expected, especially if
large amount of resurge deposition filled the crater shortly after formation.
Based on the quantification system designed in this study, nine craters were rated
with total scores of 70% or higher and are subsequently classified as exemplary
candidates. All of these craters exhibit signs of slumping, rim collapse and subdued
topography. Furthermore, 77% of the exemplary candidates show signs of debris flows
and resurge deposits, and 66% have radial gullies and/or central terraces.
130
CONCLUSIONS
This study found 77 impact craters in the area of Arabia Terra on Mars that
potentially formed in a shallow-marine environment. These craters are identified based
on high-resolution imagery data from Mars Orbiter Camera (MOC) and Thermal
Emission Imaging Spectrometer (THEMIS), in combination with topographic data from
Mars Orbiter Laser Altimeter (MOLA). The classification and ranking of these craters is
based on the presence of certain morphological features identified from shallow-marine
craters on earth as well as potential candidates on Mars. A quantification system is
designed based on the following features:
? Wet mass movement in the form of slumping (a brittle movement) and
flowing (a ductile movement) as a result of sediment weakness induced by
saturation
? Radial gullies or channels carved by sediment-loaded waters induced by
violent resurge of water almost immediately after crater formation
? Resurge deposits in the form of mixed structural blocks and/or
sedimentary deposits related to the return of water, either inside or outside
of the crater
? Central terrace or flat-topped central uplifts indicative of marine origin
and/or that the structure has been buried under water for some time
131
? Rim collapse and large amounts of inward collapse due to resurge activity
or instability of the rim
? Subdued topography indicative of the amount of erosion that has taken
place since formation.
The factors listed above are weighted based on importance to create a
quantification system for rating and ranking potential candidates. From the quantification
system, nine craters are rated with total scores of 0.7 or higher and are subsequently
classified as exemplary candidates. All of these craters exhibit signs of slumping, rim
collapse and subdued topography; suggesting that these features may be most important
in shallow marine impact craters. Furthermore, 77% show signs of debris flows and
resurge deposits, and 66% have radial gullies and/or central terraces.
The exemplary candidates are classified into three groups or types. Type I
includes well-developed, mostly medium-sized examples; type II includes typical small
and medium examples; and type III includes other large potential candidates. Three of
the nine craters fall in the type I group; a further five falls in the type II group; and one
falls within the type III group.
The results of this study are useful in helping develop a general classification and
characterization of potential marine craters. However, a few limitations should be
considered:
132
? The images either show a large amount of detail with very little context, or
good context but little detail
? Many of the features that have been listed as characteristics of shallow-
marine impact craters can also be formed in other ways and are therefore
not entirely predictive
? Not much is known about the geomorphology of terrestrial shallow-
marine impact craters, particularly from a remote-sensing point of view,
and thus it is hard to compare terrestrial analogs with Martian examples.
This study only looked at MOLA, MOC, and THEMIS data for a small part of the
Martian surface. Further studies could include the following:
? More imagery is already available with missions such as Mars Express
and Mars Reconnaissance Orbiter, and will also be available in the near
future with the Mars Phoenix mission that is on its way
? Analysis of spectrographic data of the candidate sites in order to determine
composition and to draw conclusions about target material
? Other areas of the proposed shorelines could also be scrutinized for similar
craters, as well as areas within large basins that may have been filled with
water for extended periods of time.
This study concludes that evidence for potential shallow-marine impact craters
can be found on the surface of Mars as exemplified by Arabia Terra.
133
REFERENCES
Abshire J.B., Sun X., and Afzal R.S. 2000. Mars Orbiter Laser Altimeter: receiver model
and performance analysis. Applied Optics 39(15):2449-2460.
Aharonson O., Zuber M.T., and Rothman D.H. 2001. Statistics of Mars? Topography
from the Mars Orbiter Laser Altimeter: Slopes, Correlations, and Physical
Models. Journal of Geophysical Research 106(23):723-735.
Artemieva N.A. and Shuvalov V.V. 2002. Shock metamorphism on the ocean floor
(numerical simulations). Deep-Sea Research II 49:959-968.
Arizona State University Thermal Emission Imaging System Database (Accessed 25 October 2007)
Baker V.R. 2001. Water and the martian landscape. Nature 412:228-236.
Baker V.R. 2006. Geomorphological Evidence for Water on Mars. Elements 2(3):139-
143.
Barlow N.G. 1993. Depth-diameter ratios for Martian impact craters: Implications for
target properties and episodes of degradation. LPI Technical Report 93-06 Part 1.
Barlow N.G. 2006a. Impact craters in the northern hemisphere of Mars: Layered ejecta
and central pit characteristics. Meteoritics and Planetary Science 41(10):1425-
1436.
Barlow N.G. 2006b. Martian impact craters and their implications for target
characteristics. Proceedings, 40
th
ESLAB Symposium ? First International
Conference on Impact Cratering in the Solar System. pp. 17-25.
Barlow N.G., Barnes C.W., Barnouin-Jha O.S., Boyce J.M., Chapman C.R., Costard
J.M., Craddock R.A., Garvin J.B., Greeley R., Hare T.M., Kuzmin R.O.,
Mouginis-Mark P.J., Newsom H.E., Sakimoto S.E.H., Stewart S.T., and
Soderblom L.A. 2003. Utilizing GIS in Martian Impact Crater Studies (abstract).
Advances in Planetary Mapping Workshop - International Society for
Photogrammetry and Remote Sensing.
Boyce J.M., Mouginis-Mark P., and Garbeil H. 2005. Ancient oceans in the northern
134
lowlands of Mars: Evidence from impact crater depth/diameter relationships. Journal of
Geophysical Research 110(E03008):15p.
Canup R.M. and Asphaug E. 2001. Origin of the Moon in a giant impact near the end of
the Earth?s formation. Nature 412:708-712.
Caplinger M.A., Jensen E., Edgett K., Eliason E.M., and Garcia P.A. 1999. An archive
of MGS/MOC imaging data (abstract #1659). 30th Lunar and Planetary Science
Conference. CD ROM.
Carr M.H. 1995. The Martian drainage system and the origin of valley networks and
fretted channels. Journal of Geophysical Research 100(E4):7479-7507.
Carr M.H. and Garvin J.B. 2001. Mars Exploration. Nature 412:250-253.
Carr M.H. and Head J.W. III 2003. Oceans on Mars: An assessment of the observational
evidence and possible fate. Journal of Geophysical Research 108(E5):5042-
5050.
Chapman M.G., 2003. Sub-ice volcanoes and ancient oceans/lakes: A Martian challenge.
Global and Planetary Change 35:185-198.
Christensen P.R., Bandfield J.L., Hamilton V.E., Howard D.A., Lane M.D., Piatek J.L.,
Ruff S.W., and Stefanov W.L. 2000. A thermal emission spectral library of rock?
forming minerals. Journal of Geophysical Research 105(E4):9735-9739.
Christensen P.R., Jakosky B.M., Kieffer H.H., Malin M.C., McSween H.Y. Jr., Nealson
K., Mehall G.L., Silverman S.H., Ferry S., Caplinger M., and Ravine M. 2004.
The Thermal Emission Imaging System (THEMIS) for the Mars 2001 Odyssey
Mission. Space Science Reviews 110:85-130.
Christensen P.R. 2006. Water at the Poles and in Permafrost Regions of Mars. Elements
2(3):151-155.
Clifford S.M. and Parker T.J. 2001. The Evolution of the Martian Hydrosphere:
Implications for the Fate of a Primordial Ocean and the Current State of the
Northern Plains. Icarus 154:40-79.
Collins G.S. and W?nneman K. 2005. How big was the Chesapeake Bay impact?
Insight from numerical modeling. Geology 33(12):925-928.
De Villiers G., King D.T. Jr., and Marzen L.J. 2006. A study of Impact Craters on Mars
with the use of GIS and Remote Sensing Applications (abstract). 3
rd
Annual GIS
Symposium and Workshop.
Dypvik H. and Jansa L.F. 2003. Sedimentary signatures and processes during marine
135
bolide impacts: a review. Sedimentary Geology 161:309-337.
Dypvik H., Burchell M.J., and Claeys P. 2004. Impacts into Marine and Icy
Environments. In Cratering in marine environments and on ice, edited by Dypvik
H., Burchell M.J., and Claeys P. Berlin: Springer-Verlag. pp.1-20.
Earth Impact Database (Accessed: 18 July
2007)
Edgett K.S. and Malin M.C. 2002. Martian sedimentary rock stratigraphy: Outcrops and
interbedded craters of northwest Sinus Meridiani and southwest Arabia Terra.
Geophysical Research Letters 29(24):2179-2183.
Edgett K.S. and Parker T.J. 1997. Water on early Mars: Possible subaqueous
sedimentary deposits covering ancient cratered terrain in western Arabia Terra
and Sinus Meridiani. Geophysical Research Letters 24(24):2897-2900.
Fair?n A.G., Dohm J.M., Baker V.R. de Pablo M.A., Ruiz J., Ferris J.C., and Anderson
R.C. 2003. Episodic flood inundations of the northern plains of Mars. Icarus
165:53-67.
Garvin J.B., Sakimoto S.E.H., Frawley J.J., and Schnetzler C. 2000. North Polar Region
Craterforms on Mars: Geometric Characteristics from the Mars Orbiter Laser
Altimeter. Icarus 144:329-352.
Gault D.E. and Sonnett C.P. 1982. Laboratory simulation of pelagic asteroidal impact:
atmospheric injection, benthic topography and the surface wave radiation field. In
Geological Implications of Impacts of Large Asteroids and Comets on the Earth
edited by Silver L.T. and Schultz P.H. Geological Society of America Special
Paper 190:69-92.
Geology Web Pages (Accessed 7
September 2007)
Gersonde R., Kyte F.T., Bleil U., Diekmann B., Flores J.A., Gohl K., Grahl G., Hagen R.,
Kuhn G., Sierro F.J., Volker D., Abelmann A., and Bostwick J.A. 1997.
Geological record and reconstruction of the late Pliocene impact of the Eltanin
asteroid in the Southern Ocean. Macmillan Magazines 390(6658):357-363.
Gersonde R., Deutsch A., Ivanov B.A., and Kyte F.T. 2002. Oceanic impacts ? a
growing field of fundamental geoscience. Deep-Sea Research II 49:951-957.
Ghatan G.J. and Zimbelman J.R. 2006. Paucity of candidate coastal constructional
landforms along proposed shorelines on Mars: Implications for a northern
lowlands-filling ocean. Icarus 185:171-196.
Gooding J.L. 1992. Soil Mineralogy and Chemistry on Mars: Possible Clues from Salts
136
and Clays in SNC Meteorites. Icarus 99:28-41.
Grieve R.A.F. 1987. Terrestrial Impact Craters. Annual Reviews of Earth and Planetary
Science 15:245-270.
French B.M. 1998. Traces of Catastrophe: A Handbook of Shock-Metamorphic Effects
in Terrestrial Meteorite Impact Structures. LPI Contribution No. 954, Lunar and
Planetary Institute, Houston. 120p.
Hare T.M. and Tanaka K.L. 2000. Using MOLA and MOC in a GIS (abstract #1907).
31
st
Lunar and Planetary Science Conference. CD ROM.
Hare T.M. and Tanaka K.L 2001. Planetary Interactive GIS-on-the-Web Analyzable
Database (PIGWAD) (abstract). 20
th
International Cartographic Congress.
Hare T.M., Tanaka K.L., and Skinner J.A. 2003. GIS 101 for planetary research
(abstract). Advances in Planetary Mapping Workshop - International Society for
Photogrammetry and Remote Sensing.
Hartmann W.K. 1966. Martian Cratering. Icarus 5:565-576.
Hartmann W.K. and Davis D.R. 1975. Satellite-Sized Planetesimals and Lunar Origin.
Icarus 24:504-515.
Hartmann W.K. and Neukum G. 2001. Cratering Chronology and the Evolution of Mars.
Space Science Reviews 96:165-194.
Head J.W. III., Hiesinger H., Ivanov M., Kreslavsky M.A., Pratt S., and Thomson B.J.
1999. Possible Ancient Oceans on Mars: Evidence from Mars Orbiter Laser
Altimeter Data. Science 286:2134-2137.
Helfer H.L. 1990. Of Martian Atmospheres, Oceans, and Fossils. Icarus 87:228-235.
Herkenhoff K.E., Squyres S.W., Arvidson R., Bass D.S., Bell J.F. III, Bertelsen P.,
Ehlmann B.L., Farrand W., Gaddis L., Greeley R., Grotzinger J., Hayes A.G.,
Hviid S.F., Johnson J.R., Jolliff B., Kinch K.M., Knoll A.H., Madsen M.B., Maki
J.N., McLennan S.M., McSween H.Y., Ming D.W., Rice. J.R. Jr., Richter L.,
Sims M., Smith P.H., Soderblom L.A., Spanovich N., Sullivan R., Thompson S.,
Wdowiak T., Weitz C., and Whelley P. 2004. Evidence from Opportunity?s
Microscopic Imager for Water on Meridiani Planum. Science 306:1727-1730.
Holsapple K.A. 1993. The scaling of impact processes in planetary sciences. Annual
Reviews of Earth and Planetary Science 21:333-373.
Horton J.W. Jr., Orm? J., Powars D.S., and Gohn G.S. 2006. Chesapeake Bay impact
structure: Morphology, crater fill, and relevance for impact structures on Mars.
137
Meteoritics and Planetary Science 41(10):1613-1624.
Howenstein J. 2006. Analysis of the Depth-Diameter Relationship of Martian Craters. A
Capstone Experience Thesis, Mount Holyoke College, South Hadley,
Massachusetts, USA.
Ivanov M.A. and Head J.W. 2001. Chryse Planitia, Mars: Topographic configuration,
outflow channel continuity and sequence, and tests for hypothesized ancient
bodies of water using Mars Orbiter Laser Altimeter (MOLA) data. Journal of
Geophysical Research 106(E2):3275-3296.
Johnson R.C. 2007. Wetumpka Impact Structure modeled as the exposed remains of a
large, shallow-water, marine-target impact crater for analysis and interpretation of
two drill cores taken from near the structure?s geographic center. M.S. Thesis,
Auburn University, Auburn, Alabama, USA.
Kenkmann T. and Sch?nian F. 2005. Impact Craters on Mars and Earth: Implications by
Analogy (abstract #3017). 36
th
Lunar and Planetary Science Conference. CD
ROM.
Kennett J.P. 1982. Marine geology. Englewood Cliffs: Prentice-Hall. 813p.
King D.T. Jr., Neathery T.L., Petruny L.W., Koeberl C., and Hames W.E. 2002.
Shallow-marine impact origin of the Wetumpka structure (Alabama, USA). Earth
and Planetary Science Letters 202:541-549.
King D.T. Jr., Orm? J., Petruny L.W., and Neathery T.L. 2006. Role of water in the
formation of the Late Cretaceous Wetumpka impact structure, inner Gulf Coastal
Plain of Alabama, USA. Meteoritics and Planetary Science 41(10):1625-1631.
Kirk R.L. 2005. Mars. Photogrammetric Engineering and Remote Sensing 71(10):1111-
1114.
Kring D.A. 2005. Hypervelocity collisions into continental crust composed of sediments
and an underlying crystalline basement: comparing the Ries (~24 km) and
Chicxulub (~180 km) impact craters. Chemie der Erde 65:1-46.
Koeberl C. 2001. Craters on the Moon from Galileo to Wegener: A short history of the
impact hypothesis, and implications for the study of terrestrial impact craters.
Earth, Moon and Planets 85-86:209-224.
Kreslavsky M.A. and Head J.W. 2006. Modification of impact craters in the northern
plains of Mars: Implications for Amazonian climate history. Meteoritics and
Planetary Science 41(10):1633-1646.
Leshin L.A. and Vicenzi E. 2006. Aqueous Processes Recorded by Martian Meteorites:
138
Analysing Martian Water on Earth. Elements 2:159-164.
Lindstr?m M., Sturkell E.F.F., Orm? J., and T?rnberg R. 1996. The Lockne Marine
Impact in the Ordovician of Central Sweden. Meteoritics and Planetary Science
31:A80-A81.
Lodders K. 1998. A survey of shergottite, nakhlite and chassigny meteorites whole-rock
compositions. Meteoritics and Planetary Science 33:A183-190.
Malin M.C., Danielson G.E., Ingersoll A.P., Masursky H., Veverka J., Ravine M.A., and
Soulanille T.A. 1992. The Mars Observer Camera. Journal of Geophysical
Research 97(E5):7699-7718.
Malin M.C. and Edgett K.S. 1999. Oceans or seas in the Martian northern lowlands:
High resolution imaging tests of proposed coastlines. Geophysical Research
Letters 26(19):3049-3052.
Malin M.C. and Edgett K.S. 2000. Evidence for Recent Groundwater Seepage and
Surface Runoff on Mars. Science 208:2330-2335.
Malin M.C. and Edgett K.S. 2003. Evidence for Persistent Flow and Aqueous
Sedimentation on Early Mars. Science 302:1931-1934.
Malin Space Science Systems Mars Orbiter Camera Database
(Accessed 25 October 2007)
McSween H.Y. Jr. 2006. Water on Mars. Elements 2(3):135-137.
Melosh H.J. 1980. Cratering Mechanics: Observational, Experimental, and Theoretical.
Annual Reviews of Earth and Planetary Science 8:65-93.
Melosh H.J. 1989. Impact Cratering: A Geologic Process. New York: Oxford
University Press, Inc. 245 p.
Mouginis-Mark P.J. 1987. Water or Ice in the Martian Regolith? Clues from Rampart
Craters Seen at Very High Resolution. Icarus 71:268-286.
NASA 2006. Strategic Plan. National Aeronautics and Space Administration. (NP-
2006-02-423-HQ).
Neumann G.A. 2001. Some aspects of processing extra-terrestrial LIDAR data:
Clementine, NEAR, MOLA. International Archives of Photogrammetry and
Remote Sensing 34:73-80.
Orm? J. and Lindstr?m M. 2000. When a cosmic impact strikes the sea bed. Geological
Magazine 137(1):67-80.
139
Orm? J. and Muinonen P. 2000. Impact Craters as Indicators for Oceanic Phases on
Mars (abstract #1266). 31
st
Lunar and Planetary Science Conference. CD ROM.
Orm? J., Shuvalov V.V., and Lindstr?m M. 2002. Numerical modeling for target water
depth estimation of marine target impact craters. Journal of Geophysical
Research 107(E12):3-11.
Orm? J., Dohm J.M., Ferris J.C., Lepinette A., and Fair?n A.G. 2004. Marine?target
craters on Mars? An assessment study. Meteoritics and Planetary Science
39(2):333-346.
Orm? J., Lindstr?m M., Lepinette A., Martinez?Frias J., and Diaz?Martinez E. 2006.
Cratering and modification of wet?target craters: Projectile impact experiments
and field observations of the Lockne marine?target crater, Sweden. Meteoritics
and Planetary Science 41(10):1605-1612.
Paige D.A. 2005. Ancient Mars: Wet in Many Places. Science 307:1575-1576.
Parker T.J., Saunders R.S., and Schneeberger D.M. 1989. Transitional morphology in the
west Deuteronilus Mensae region of Mars: Implications for modification of the
lowland/upland boundary. Icarus 82:111-145.
Pike R.J. 1977. Apparent depth/apparent diameter relation for lunar craters.
Proceedings, 8
th
Lunar Science Conference. Pp. 3427-3436.
Pike R.J. 1980. Formation of Complex Impact Craters: Evidence from Mars and Other
Planets. Icarus 43:1-19.
Perron J.T., Mitrovica J.X., Manga M., Matsuyama I., and Richards M.A. 2007.
Evidence for an ancient martian ocean in the topography of deformed shorelines.
Nature 447:840-843.
Pierazzo E. and Melosh H.J. 1999. Hydrocode modeling of Chicxulub as an oblique
impact event. Earth and Planetary Science Letters 165(2):163-176.
Poag C.W. 1997. The Chesapeake Bay bolide impact: a convulsive event in Atlantic
Coastal Plain evolution. Sedimentary Geology 108:45-90.
Poag C.W., Koeberl C., and Reimold W.U. 2004. The Chesapeake Bay Crater: Geology
and Geophysics of a Late Eocene Submarine Impact Structure. New York:
Springer-Verlag. 522 p.
Ramsey M.S. and Christensen P.R. 1998. Mineral abundance determination:
Quantitative deconvolution of thermal emission spectra. Journal of Geophysical
Research 103(B1):577-596.
Reiss D., Hauber E., Ivanov B.A., Michael G., Jaumann R., Neukum G., and The HRSC
140
Co-investigator Team 2006. Rampart craters in Thaumasia Planum, Mars: Onset
diameters, ages, and implications for the Hesperian hydrology. Geophysical
Research Abstracts (volume 8 abstract #06506). 3
rd
European Geoscience Union
General Assembly.
Roark J.H., Masuoka C.M., and Frey H.V. 2004. GRIDVIEW: Recent improvements in
research and education software for exploring Mars topography (abstract #1833).
35
th
Lunar and Planetary Science Conference. CD ROM.
Schaefer F., Thoma K., Behner T., Nau S., Kenkman T., W?nneman K., Deutsch A. and
the MEMIN-Team 2006. Impact experiments on dry and wet sandstone.
Proceedings, 40
th
ESLAB Symposium ? First International Conference on Impact
Cratering in the Solar System. pp. 131-136.
Schultz P.H. and Lutz A.B. 1988. Polar wandering of Mars. Icarus 73:91-141.
Sharp R.P. 1968. Surface Processes Modifying Martian Craters. Icarus 8:472-480.
Shuvalov V.V. and Trubestkaya I.A. 2002. Numerical Modeling of Marine Target
Impacts. Solar System Research 36(5):417-430.
Smith D.E. , Zuber M.T., Frey H.V., Garvin J.B., Head J.W., Muhleman D.O., Pettengill
G.H., Phillips R.J., Solomon S.C., Zwally H.J., Banerdt W.B., and Duxbury T.C.
1998. Topography of the Northern Hemisphere of Mars from the Mars Orbiter
Laser Altimeter. Science 279:1686-1692.
Smith D.E., Zuber M.T., Frey H.V., Garvin J.B., Head J.W., Muhleman D.O., Pettengill
G.H., Phillips R.J., Solomon S.C., Zwally H.J., Banerdt W.B., Duxbury T.C.,
Golombek M.P., Lemoine F.G., Neumann G.A., Rowlands D.D., Aharonson O.,
Ford P.G., Ivanov A.B., Johnson C.L., McGovern P.J., Abshire J.B., Afzal R.S.,
and Sun X. 2001. Mars Orbiter Laser Altimeter: Experiment summary after the
first year of global mapping of Mars. Journal of Geophysical Research
106(23):689-722.
Smith D.E., Neumann G., Arvidson R.E., Guinness E.A., and Slavney S. 2003. Mars
Global Surveyor Laser Altimeter Mission Experiment Gridded Data Record.
NASA Planetary Data System, MGS-M-MOLA-5-MEGDR-L3-V1.0.
Solomon S.C., Aharonson O., Aurnou J.M., Banerdt W.B., Carr M.H., Dombard A.J.,
Frey H.V., Golombek M.P., Hauck S.A. II, Head J.W. III, Jakosky B.M., Johnson
C.L., McGovern P.J., Neumann G.A., Phillips R.J., Smith D.E., and Zuber M.T.
2005. New Perspectives on Ancient Mars. Science 307:1214-1220.
Squyres S.W. 1989. Water on Mars. Icarus 79:229-288.
Squyres S.W., Grotzinger J.P., Arvidson R.E., Bell J.F. III, Calvin W., Christensen P.R.,
Clark B.C., Crisp J.A., Farrand W.H., Herkenhoff K.E., Johnson J.R.,
141
Klingelh?fer G., Knoll A.H., McLennan S.M., McSween H.Y. Jr., Morris R.V.,
Rice J.W. Jr., Rieder R., and Soderblom L.A. 2004. In Situ Evidence for an
Ancient Aqueous Environment at Meridiani Planum, Mars. Science 306:1709-
1714.
Squyres S.W., Knoll A.H., Arvidson R.E., Clark B.C., Grotzinger J., Jolliff B.L.,
McLennan S.M., Tosca, N., Bell J.F. III, Calvin W.M., Farrand W.H., Glotch
T.D., Golombek M.P., Herkenhoff K.E., Johnson J.R., Klingelh?fer G., McSween
H.Y. Jr., and Yen A.S. 2006a. Two Years at Meridiani Planum: Results from the
Opportunity Rover. Science 313:1403-1407.
Squyres S.W., Aharonson, O., Arvidson R.E., Bell J.F. III, Christensen P.R., Clark B.C.,
Crisp J.A., Farrand W.H., Glotch T., Golombek M.P., Grant J., Grotzinger J.,
Herkenhoff K.E., Johnson J.R., Jolliff B.L., Knoll A.H., McLennan S.M.,
McSween H.Y. Jr., Moore J.M., Rice J.W. Jr., and Tosca N. 2006b. Planetary
Science: Bedrock formation at Meridiani Planum. Nature 443:E1-E2.
Sturkell E.F.F, Ekelund, A., and T?rnberg, R. 1998. Gravity modeling of Lockne, a
marine impact structure in J?mtland, central Sweden. Tectonophysics 296:421-
435.
Suspected Earth Impact Sites Database
(Accessed: 19
July 2007)
Tanaka K.L., Skinner J.A. Jr., and Hare, T.M. 2005. Geologic Map of the Northern
Plains of Mars. Prepared for the National Aeronautics and Space Administration
by the United States Geological Survey to accompany Scientific Investigations
Map 2888. 32p.
Tsikalas F., Gudlaughsson S.T., and Faleide J.I. 1998. The anatomy of a buried complex
impact structure: The Mj?lnir Structure, Barents Sea. Journal of Geophysical
Research 103:30469-30484.
Venechuk E.M., Allen C.C., and Oehler D.Z. 2005. Layers, bright rim craters, and
faulting in the Arabia Terra, Mars. Proceedings, 21
st
LPI Summer Intern
Conference. pp.34-36.
Von Dalwigk I. and Orm? J. 2001. Formation of resurge gullies at impacts at sea: The
Lockne crater, Sweden. Meteoritics and Planetary Science 36:359-369.
Wohletz K.H. and Sheridan M.F. 1983. Martian Rampart Crater Ejecta: Experiments and
Analysis of Melt-Water Interaction. Icarus 56:15-37.
Wright S.P. 2003. Thermal infrared data analyses of Meteor Crater, Arizona:
Implications for Mars spaceborne data from the Thermal Emission Imaging
142
System. M.S. Thesis, University of Pittsburgh, Pittsburgh, Pennsylvania, USA.
W?nnemann K. and Lange M.A. 2002. Numerical modeling of impact-induced
modifications of the deep-sea floor. Deep-Sea Research II 49:969-981.
Wyatt M.B. and McSween H.Y. Jr. 2006. The Orbital Search for Altered Materials on
Mars. Elements 2(3):145-150.
Zuber M.A. 2007. Mars at the tipping point. Nature 447:785-786.