i
Evidence for a Possible Supervolcano on Mars at Siloe Patera
by
Cheryl Wilkes Coker
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
May 4, 2014
Copyright 2014 by Cheryl Wilkes Coker
Approved by
Shawn P. Wright, Co-Chair, Postdoctoral Fellow in Geology
David T. King Jr., Co-Chair, Professor of Geology
Luke J. Marzen, Professor of Geography
Clinton I. Barineau, Professor of Geology (Columbus State University)
ii
Abstract
Highlands on Mars interpreted as impact craters may instead be supervolcanoes
within an ancient volcanic province in Arabia Terra, Mars. These volcanoes are
characterized by lower than normal topographic relief, collapse features, layered
deposits, as well as effusive volcanism and explosive eruptions. Seven features are to be
considered part of the new volcanic field, which includes Siloe Patera. Although an
alternative hypothesis for Siloe Patera?s origin is nested impact craters, evidence points
more towards multiple caldera collapses or a combination of impact and caldera
collapse events. Regions of interest in and around Siloe Patera include: (1) possible
volcanic-doming along the bench; (2) possible ring faulting on the eastern portion of the
bench; (3) a spire located on the floor of Siloe Patera; and (4) flow features around Siloe
Patera. Data from various Martian orbiters was used to analyze Siloe Patera, Ascraues
Mons, a nested crater and three random craters roughly the size of Siloe Patera. Data
collection included gathering raw Mars Orbital Laser Altimeter (MOLA) tracks over Siloe
Patera, Thermal Emission Imaging System (THEMIS) Night Infrared, Context Camera
(CTX), and Mars Orbiter Camera (MOC) images. Utilizing data currently available,
characteristics of Siloe Patera resemble that of a volcano more than that of an impact
crater based on topographic profiles, depth diameter ratios, slope angles, and
comparison of regions of interest.
iii
Acknowledgments
I would like to express a special appreciation and thanks to my advisors,
Dr. Shawn Wright and Dr. David King, you have been tremendous mentors for me. I
would like to thank you for encouraging my research and for allowing me to grow as a
research scientist. I would also like to thank my thesis committee members, Dr. Luke
Marzen, and Dr. Clinton Barineau for serving as my committee members even at
hardship. I also want to thank Dr. Jacob Bleacher and Dr. Joseph Michalski for allowing
me the opportunity to work with you on your exciting finding. It has been an
exhilarating challenge that would not have been possible without your generosity.
I want to acknowledge the extensive use of multiple publicly accessible imagery
databases which includes Arizona State University , U.S.G.S
Planetary GIS Web Server , MOLA Pedr Query <
http://ode.rsl.wustl.edu/mars/dataPointSearch.aspx >, HiWish: Public suggestion Page
, Google Earth 7.1, and JMARS.
I would like to dedicate this thesis to my wonderful family and husband for their
overwhelming support and advice to always go after your dream. I would also like to
give a special thanks to my husband Zachary for his unconditional support, love, and
especially patience over the past two years.
iv
Table of Contents
Abstract...............................................................................................................................ii
Acknowledgments .............................................................................................................iii
List of Tables.......................................................................................................................vi
List of Figures....................................................................................................................vii
Chapter 1?.........................................................................................................................1
Chapter 2 ???????????.............................................................................................7
Geologic History of Mars........................................................................................ 7
Martian Volcanism................................................................................................13
Calderas.....................................................................................................17
Collapse and Resurgent Calderas..............................................................19
Supervolcanos...........................................................................................23
Martian Impact Craters.........................................................................................25
Impact Crater Characteristics...................................................................24
Classification ........................................................................................................29
Chapter 3??....................................................................................................................33
Mars Orbital Laser Altimeter (MOLA)..................................................................33
High Resolution imaging Science Experiment (HiRISE)........................................35
Context Camera (CTX)..........................................................................................36
Mars Orbiter Camera (MOC)................................................................................36
v
Thermal Emission Imaging System (THEMIS).......................................................45
Chapter 4 ........................................................................................................................47
Geologic Map Classification.................................................................................49
ArcGIS Analysis.....................................................................................................53
3D Model..............................................................................................................58
Regions of Interest...............................................................................................61
Topographic Profiles............................................................................................63
Crater Database....................................................................................................75
Flow Features.......................................................................................................80
Chapter 5..........................................................................................................................82
Regions of Interest Analysis..................................................................................82
Nested Crater Comparison ..................................................................................92
Classification Map................................................................................................94
Suggested Volcanic History..................................................................................95
Chapter 6..........................................................................................................................96
References .......................................................................................................................98
vi
List of Tables
Table 1. HiRISE image suggestion list?????.........................................................37
Table 2. CTX, MOC, and HiRISE Product Identification and location downloaded from
Arizona State University?s Mars Image Explorer. Images and locations are displayed in
Figure 12. Letters are positioned in the top left corner of all image????..?..??41
Table 3. List of three classes and land cover types used to produce classification map in
Erdas Imagine (Figure 24)........................................................................................52
Table 4. List of slope angle from Figure 41????????........................................73
Table 5. Robbins and Barlow Crater Databases information collected on Siloe
Patera?????????????????????????????????????????..?..75
Table 6. Distance from Crater A and Crater B center (km) to edge of flow
features???????????????????????????????????????.?......82
vii
List of Figures
Figure 1: Image displaying the study area of the new proposed volcanoes/associated
features on Mars (Michalski and Bleacher,2013).?????????????.??.....?..2
Figure 2: Eden Patera shown here displaying multiple collapse events, as well as volcanic
vents and lava lakes (Michalski and Bleacher, 2013)????????????????.?3
Figure 3A: Siloe Patera divided into two craters. Crater A circled in blue and Crater B
circled in yellow??????????.??..???????????????...........................5
Figure 3B: Regions of interest: (1) volcanic-doming shown in red hexagons; (2) faulting
circled in a blue ellipse; (3) a spire encompassed by a green rectangle. The Bench (Crater
A?s floor) is highlighted in green????????????????..???????..????.6
Figure 4: Depiction of the physical and chemical history of Mars.?????...?.??.8
Figure 5: Global Mars map displaying various regions and features??????...?..9
Figure 6: Charitum Montes surrounding Argyre Planitia???????.?????..??10
Figure 7: Alba Patera and the Tharsis Bulge????????????????.??.???.11
Figure 8: Volcanism on Mars Multch et al., 1976. Circle indicates general area of Siloe
Patera and other supervolcanoes proposed????????????????.?..????15
Figure 9: Topographic Map of Mars ?M 25M RKN? (USGS, 2003) overlain with image of
volcanic units on Mars categorized by morphologic type (Greeley and Spudis, 1981).
According to this map, Siloe Patera falls into the region of highland patera and simple
flows. Circle indicates general region of Siloe Patera?s location????.???.???16
Figure 10: Simple flows near Hesperia Planum??????????????.??????18
Figure 11: Complex flows on the plains near Olympus Mons?????????..?..?18
Figure 12A: Olympus-type Martian caldera??????????????????..?.??.20
Figure 12B: Arsia-type Martian caldera??????????????????????....?.20
viii
Figure 13: Depiction of caldera collapse stages as described by Acecolla 2006)?..21
Figure 14: Depiction of the first six stages of resurgent calderas (Smith and Bailey, 1968).
Stage I, Regional tumescence and generation of ring fractures. Stage II, Caldera-forming
eruptions. Stage III, Caldera collapse. Stave IV, Preresurgence volcanism and
sedimentation. Stage V, Resurgent doming. Stage VI, Major ring-fracture volcanism.
Stage VII, not shown as volcanic activity is waning (Smith and Bailey, 1986). ?..?24
Figure 15: Comparison of Valles Caldera, New Mexico (top) and Siloe Patera (bottom).
Arrows point to volcanic domes as a result of ring faulting and circles encompass
resurgent calderas. For Siloe Patera, the circle could highlight a possible cinder cone or
dike instead of a resurgent caldera?????????????????????.?????.24
Figure 16A: IKONOS image of Kilauea caldera from NASA?s Earth Observatory Image of
the Day, December 10, 2005???????????????????????????..?..?26
Figure 16B: Kilauea Caldera viewed from the west. Image from USGS Hawaiian Volcano
Observatory website?????????????????????????????????..26
Figure 17: Depiction of simple and complex impact craters, their processes and final
results. Simple impact craters on the left and complex on the right (Osinski and Pierazzo,
2012)????????????????????????????????????????...?....28
Figure 18: Examples of Single Layer Ejecta (A), Double Layer Ejecta (B), and Multiple
Layer Ejecta (C) from Barlow 1990????.??????????????????????..30
Figure 19: Variation of crater degradation (Arvidson, 1974)??????????.??.30
Figure 20: Internal morphology of craters (Barlow, 1990)????????.??..??...31
Figure21: Spectral signatures for remote sensing are generally shown as a combination
of percent reflected back to the sensor and the wavelength in which it appears.
Different materials will present the same spectral signature independent of region,
time, and season (Meaden and Kapetsky, 1991)????????????????..??.33
Figure 22: Good coverage of MOLA tracks along Siloe Patera provides sufficient data to
analyze topography of the area. Siloe shown in box. Image rendered using ArcGIS,
topographic map of Mars, and MOLA tracks downloaded from the USGS Planetary GIS
Web Server (PIGWAD)???????????????????????????????..?.?34
Figure 23: Image showing HiRISE requests. Yellow boxes indicate areas suggested and
waiting to be taken where red indicates the areas that have been photographed.
ix
Photographed images are designated as 1 (being the more northern of the two), and 2
(being the southern). The center of Siloe is located at 35? N and 6? E???.???...37
Figure 24: HiRISE image 1 (ESP_033258_2160_MRGB) showing fluvial erosion and
possible thermokarst structures emanating from the wall. The brown strip in the middle
is shown in true color????????????????????????????????..??38
Figure 25: HiRISE image 2 (ESP_030344_2155_MRGB) is more interesting than the first
HiRISE image as it shows isolated mounds that could be the result of eroding ash,
definable layered deposits that continue for long distances, steep cliffs and a chaotic
and eroded terrane. As in HiRISE image 1, the brown strip shows true color?.?...39
Figure 26: Position and location of CTX, MOC, and HiRISE images covering Siloe Patera.
Letters are positioned in the top left corner of the image and some arrows point to the
top left corner, followed by the letter as labeled in Table 1??????.????..??40
Figure 27: Individual CTX images that were used. Image number and location are found
in Table 2. Continued on next three pages??????????????????..??...?42
Figure 28: Map of MOC images used. Image number and location can be found in Table
2????????????????????????????????????????????..?..45
Figure 29: MOC images showing several transects through Siloe Patera?..??..?.46
Figure 30: Siloe Patera shown in THEMIS Day IR (top) and THEMIS Night IR
(bottom)??????????????????????????????????????.?.?.?48
Figure 31: Color stretch of Siloe Patera with THEMIS Night IR, MOLA, and CTX mosaic
images combined into one image with each representing an individual band. Red
represents less reflective surfaces, suggesting high dust coverage and green shows areas
with hard rock ad are more reflective at night due to insolation absorption during the
day????????????????????????????????????????.???..?50
Figure 32: Geologic map of Siloe Patera showing areas of dust (tan), rock (brown), and
sediment fill-in (turquoise). Figure was rendered using THEMIS Night IR, MOLA, and CTX
mosaic images to produce an unsupervised classification of 20 classes???.???52
Figure 33: Geologic map of Siloe Patera rendered in ArcGIS and Adobe Photoshop using
a color stretch map, classification map, THEMIS Night IR, and CTX images??....?.54
Figure 34: Images are rendered using ArcGIS?s interpolation tool. Siloe (top left),
Ascraeus Mons (top right), and Crater A (right) were all generated with 30 classes and
set to a gray scale. All were generated using IDW from ArcGIS???????.?..??55
x
Figure 35: Siloe Patera (top left), Ascraeus Mons (top right), and Crater A (right) are
displayed here in contour intervals of 100 meters. This was accomplished by ArcGIS
through the use of the spatial analyst tool and imputing the IDW rendered layers and
assigning a contour interval??????????..???????????????????.?56
Figure 36: Images of Siloe Patera (top left), Ascraeus Mons (top right), and Crater A
(right) are shown as TIN layers. These layers are generated from contour maps and
display a digital surface. These maps can be used to create 3D models?????....57
Figure 37: Siloe Patera shown in a basic 3D color map (above) and in CTX (below).
Volcanic domes are outlined hexagons, the spire is outlined by a rectangle, and the
faulting region is circled by an ellipse?????????..???????????????...62
Figure 38: Siloe Patera shown in THEMIS Day IR (above) and in THEMIS Night IR (below).
Volcanic domes are outlined by hexagons, the spire is outlined by a rectangle, and the
faulting region is circled by an ellipse???????????????????????..??63
Figure 39: Topographic profile comparison of Siloe Patera against Ascraeus Mons
(caldera), Crater A, B, and C, and Crater F (nested crater). Profiles were downloaded
from ArcGIS after TIN maps were created of each feature. Data was imported into Excel
and manipulated to show their profiles????????????.??????????..?.64
Figure 40: Lines indicate areas the topographic profiles in figure 30 were rendered from.
A. Siloe Patera, B. Ascraeus Mons, C. Crater A, D. Crater B, E. Crater C, F. Crater F.
Topographic lines were drawn in an effort to best represent each feature??...?.65
Figure 41: Three topographic profiles of Siloe Patera showing the bench, a volcanic
dome, and the spire?????????????????????????????????..??67
Figure 42: Three topographic profiles of Siloe Patera showing the whole basin, lowest
bench elevation, and middle bench elevation????????????????..??.?..69
Figure 43: Topographic profiles from Figures 41 and 42???????????.??.?.70
Figure 44: Contour map of Siloe Patera with elevation labels??????????..?.72
Figure 45: Slope angles of Siloe Patera, Ascraeus Mons, Craters A, B, C, and F. Siloe
Patera contains the highest slope angle with Ascraeus Mons following????.??73
xi
Figure 46: Slope map of Siloe Patera. Red shows steeper slopes and white very shallow
slopes????????????????????????????????????????..........74
Figure 47: Crater databases used are displayed and each circle shows the area
determined to be the center of the crater according to Robbins and Barlow. Robbins
shown in yellow and Barlow in blue followed by the object number for each point in
their database????????????????????????????????????.?.?79
Figure 48: Flow features highlighted in tan and described in Table 2. Each flow has its
own characterizes ranging in length, width and general appearance. Box to the top left
of the image shows a zoomed in CTX image of an overlap in the southwestern
flow??????????????????????????????????????????.?.?81
Figure 49: Close up view of the spire located on the floor of Crater B in Siloe
Patera?????????????????????????????????????????..?..84
Figure 50: Possible movement history of a magma chamber below Siloe Pater?..88
Figure 51: Siloe Patera Crater A falls within the more degraded impact craters and Crater
B is considered to be new (Michalski and Bleacher, 2013)??????????..?.?..90
Figure 52: A comparison of Siloe Patera to a similar sized crater to the north
east?????????????????????????????????????????..?.?..91
Figure 53: Comparison of Siloe Patera (above) to Crater F (below). Many differences can
be seen here as described in text?????????????????????????.?..?93
1
CHAPTER 1
Recent work by Michalski and Bleacher (2013) suggest a possible volcanic field
located in northern Arabia Terra, Mars, along the Martian global dichotomy boundary
(Figure 1). The field contains seven possible volcanic calderas and associated features
that have been named Eden Patera, Euphrates Patera, Semeykin Crater, Ismenia Patera,
Oxus Patera, Oxus cavus, and Siloe Patera (Michalski and Bleacher, 2013; Figure 1). A
volcanic field associated with Siloe Patera is discussed here. These features are
proposed as volcanic in origin due to the presence of complex collapse features,
faulting, ridged plains likely to be volcanic in nature, and friable layered deposits, as well
as a significant lack of central peaks, uplifted rim, and ejecta typical of impact craters.
These supervolcanoes are defined as ?plains style caldera complexes? and are
characterized by lower than normal topographic relief, collapse features, and friable and
layered deposits (Michalski and Bleacher, 2013). Friable layered deposits and fretted
terrain are characterized by broad flat-floored, steep-walled valleys that are located
along the dichotomy boundary (Carr, 2001). Many fretted terrain regions are thought to
form from windblown volcanic ash or eolian sedimentary deposits of sand due to their
morphology and erosional characteristics (Irwin and Watters, 2004; Hynek et al., 2003;
Michalski and Bleacher, 2013). Sulfates have been detected in a nearby region known as
fretted terrain (Gendrin et al., 2005) and ash deposition is interpreted to be one
possible reason for the texture of the fretted terrain(Kerber et al., 2012). The sulfates
2
Fig
ure
1: Im
ag
e d
isp
layin
g t
he
stud
y a
rea
of
the ne
w
pro
po
sed v
olcan
oes
/ass
ocia
ted
fe
atures
on
M
ars (
M
iclaski and
Blea
cher,
20
13
).
3
likely formed in a volcanic outgassing environment where materials were altered under
water-limited, acidic conditions (Michalski and Bleacher, 2013).
The type volcano for this new category is Eden Patera (Figure 2), which contains
evidence for complex and irregularly shaped collapse events with arcuate scarps lining
the depressions (Michalski and Bleacher, 2013). Eden Patera has a depth of ~ 700 m and
displays many features interpreted as former lava lakes, vents, and cracks related to
lava lake drainage. Euphrates Patera also displays irregularly shaped depressions with a
similar depth of 700 m (Michalski and Bleacher, 2013). Eden Patera is interpreted to be
a caldera complex due to its similarity to terrestrial calderas and a lack of identifiable
Figure 2: Eden Patera shown here displaying multiple collapse events, as well as
volcanic vents and lava lakes (Michalski and Bleacher, 2013).
4
impact crater characteristics. This includes a surrounding terrain with ridged plains of
Hesperian basaltic volcanism and fault-bounded blocks with surfaces similar to adjacent
ridged plain lavas. Other caldera/volcanic characteristics include blocks tilted toward the
center of the depression, a mound interpreted as a graben-related vent, and continuous
terraces 100 and 150 meters above the floor, interpreted to be high stands of lava lake
drainage. The presence of apparent volcanic geomorphic features and faulting
consistent with caldera collapse are consistent with interpretations of Eden Patera as
volcanic in origin
Although multiple lines of evidence suggest a volcanic origin for Siloe Patera, it is
also possible that it originated as the result of multiple impact craters or a combination
of an overlapping impact crater and volcano. Comparison of Martian volcanoes and
impact craters are explored herein and will aid in the interpretation of Siloe Patera?s
origin.
Similar to Eden Patera, Siloe Patera contains quasi-circular nested depressions
and arcuate scarps, although it is almost 1 km deeper than Eden Patera (Figure 2),with a
total depth of ~1700 m. Herein, the overlapping depressions of Siloe Patera are referred
to as Crater A the largest and older depression and, Crater B the second and younger in
the southern half of Crater A (Figure 3A).Crater A is roughly 37 x 33km in diameter (NS-
EW) with an average depth of 1000m, ?100 m, while Crater B is ~28 x 24km with a depth
of 700m, ?100 m. Siloe Patera contains a number of notable features (Figure 3B),
including parallel mounds as a result of possible ring faulting, on the floor of Crater A. A
5
large mound on the floor of Crater B connects with the southeast rim of Siloe Patera and
is referred to as a spire for the purposes of this paper. The spire rises up in elevation
from the center of Crater B to the rim of Siloe Patera and is discussed in more detail
herein. Apparent flows lacking a definable source are found to the north, southeast, and
southwest of Siloe Patera?s rim with varying degrees of erosion, length and appearance
(Grant and Schultz, 1993). Five mounds found lining the fault scarp of Crater A and
Figure 3A: Siloe Patera divided into two craters. Crater A circled in blue and Crater B circled in
yellow.
6
Crater B are uncharacteristic of impact craters and will be discussed in more detail
below. Immediately south of Siloe Patera is a large, irregularly shaped depression that
links to Crater B at its southern flank. The significance of these regions is discussed
below and is critical for understanding the nature of Siloe Patera?s origin.
CHAPTER 2
Figure 3B: Regions of interest: (1) volcanic-doming shown in red hexagons; (2) faulting circled
in a blue ellipse; (3) a spire encompassed by a green rectangle. The Bench (Crater A?s floor) is
highlighted in green.
7
Mars is a dynamic planet with a complex history of geologic processes that have
shaped its surface through time. Volcanic features make up more than 35 percent of the
Martian surface and cover much of the planet with volcanic fields and plains, including
large shield volcanoes such as Olympus Mons (Greeley, 1985). Although volcanism is
widespread on Mars, impact craters are the more dominate feature and cover the entire
planet. Impact craters and volcanoes can be differentiated from one another other in
most cases; however, impact and volcanic processes can result in non-unique
geomorphic features that make unequivocal interpretations difficult. In these cases,
remote sensing is a very helpful tool as it provides methods that can help to identify
rock types by spectrographic signatures and provides insight into the process that may
have formed a particular feature.
Three main intervals (Figure 4) in Martian geologic history have been defined
based on impact crater size and frequency: Noachian (4.6 to 3.5 Ga); Hesperian (3.5 to
1.8 Ga); and Amazonian (1.8 to present; Scott and Carr, 1978; Figure 4). Tanaka (1986)
modified the original time-stratigraphic classification of Scott and Carr (1978) and
subdivided the three periods into series referred to as ?Upper,? ?Middle,? and ?Lower?
based on research in the western hemisphere of Mars, where detailed imagery
suggested stratigraphy in remotely observed outcrops. A second time scale was derived
from data gathered by the OMEGA Visible and Infrared Mineralogical Mapping
Spectrometer on board the Mars Express orbiter (Bibring et al., 2006). In this new time
scale there are three periods: Phyllocian (4.5 to 4.0 Ga); Theiikian (4.0 to 3.5); and
8
Siderikan (3.5 to present). These chemical periods are determined by the abundance of
specific minerals associated with water availability (Bibring, et al., 2006).
Physical History:
Noachian Period (4.5 to 3.5 Ga) - Noachian rocks are the oldest, generally the most
densely cratered by impacts, and are mainly found in the southern highlands (Scott and
Tanaka, 1986; Figure 5).Rocks which formed during this age developed during the final
stages of planetary accretion and subsidence associated with the heavy bombardment
period. Extensive flood lavas and surface modifications formed from eolian, fluvial, and
other surface processes after the heavy bombardment. It is thought that the Tharsis
bulge (Figure 5) formed during this time, as well as major flooding by liquid water
(Bibring, et al., 2006).
Lower Noachian Series - Recent mapping has distinguished stratigraphically lower
?basement material? from cratered terrain material that is scattered over many areas
on Mars (Scott and Tanaka, 1986). ?Basin rim material? and ?mountain material? units
Figure 4: Depiction of the physical and chemical history of Mars.
9
have been reclassified as basement material within the Lower Noachian Series and are
best observed in the uplifted rim of Charitum Montes (Figure 6) surrounding Argyre
Planitia. These rims are rugged, heavily cratered from impacts, and faulted with hills and
ridges that appear to be concentrically aligned with Argyre Planitia. Middle Noachian
cratered terrain material embays much of the Lower Noachian basement and can be
found around Hellas Basin (Tanaka, 1986; Figure 5). The Lower Noachian Series
represents the most primitive crust of Mars and formed by solidification of the molten
surface when the planet was forming (Tanaka, 1986).
Middle Noachian Series - The Middle Noachian Series contains rugged and scattered
cratered terrain with remnants in the northern plains. This unit has been mapped as
?cratered unit of the plateau sequence? (Tanaka, 1986) and is the most widespread of
all of the Noachian Systems. The Middle Noachian Series type region is west of Hellas
Figure 5: Global Mars map displaying various regions and features.
10
impact basin (Figure 5) and characterized by rugged cratered terrain with moderate high
relief, secondary impact craters, wrinkle ridges, scarps, and channels (Tanaka, 1986).
Upper Noachian Series - Defined as ?cratered plateau material,? the Upper
Noachian Series was originally classified as densely cratered with smooth and flat
intercrater areas, but recent mapping shows that both units are separate. The smooth,
flat portions over the cratered terrains embay the Middle Noachian Series (Tanaka,
1986).
Hesperian Period (3.5 to 1.8 Ga) - Hesperian formations and features display
extensive volcanism, tectonism, and canyon and channel formation as well as significant
Figure 6: Charitum Montes surrounding Argyre Planitia.
11
impact cratering (Scott and Tanaka, 1986). Hesperian material appears as thick sheets
that cover the highland terrain and partially cover the northern plains. The Tharsis
Montes, Alba Patera, and Syria Planum (Figure 7) formed from extensive lava flows
during this period which are estimated to have resurfaced more than half of Mars (Scott
and Tanaka, 1986).
Amazonian Period (1.8 to present) ?Materials from the Amazonian Period are
usually surrounded by Hesperian-aged materials (Scott and Tanaka, 1986). Amazonian
material is interpreted as youngest due to featureless characteristics and presence of
eolian material (Scott and Tanaka, 1986).
Figure 7: Alba Patera and the Tharsis Bulge
12
Chemical History:
Phyllocian (4.5 to 4.0 Ga) - The Phyllocian Period is characterized by phyllosilicates
that formed in an alkaline water environment which would match the Noachian Period
with its extensive flooding. This time period is thought to be the most likely for hosting
life (Bibring, et al., 2006).
Theiikian (4.0 to 3.5 Ga) - The Theiikian Period was dominated by acidic conditions as
sulfates from widespread volcanic outgassing and processes mixed with liquid water
present at that time (Bibring, et al., 2006).
Siderikan (3.5 to present)?During the Siderikan Period volcanism ceased and liquid
water appears to have been largely absent from the Martian surface. The physical
processes that dominate during this interval are largely limited to oxidation of pre-
existing rock. Iron-rich rocks reacting with atmospheric peroxides would have oxidized
the Martian surface, giving Mars its distinct red color (Bibring et al., 2006).
Arabia Terra (Figure 5) has a complex history of erosion, deposition and processes
resulting in varying topographic features, impact structures (McGill, 2000) and layered
deposits of fine-grained material (Fergason and Christianson, 2008). Supervolcanoes
mentioned above (e.g. Siloe Patera, Eden Patera) are found in Arabia Terra and along
the global dichotomy boundary separating the Northern Lowlands from the Southern
Highlands along a topographic break of 1-3 km (Irwin and Watters, 2004; Watters and
McGovern, 2006). Along this boundary are fretted terrain and fretted channels
characterized by smooth, flat, lowland areas separated by scarps and complex channels
13
(Carr, 2001). Fretted terrain is only found in northern Arabia Terra 40? N and between
280? W and 350? W and has been modified by water and ground ice (Sharp, 1973).Fine-
grained materialscould be linked to the formation of the fretted terrains and were likely
deposited in the early Noachian-Early Hesperian (Irwin and Watters, 2004). These fine-
grained materials may source from the proposed volcanic region of Arabia Terra,
suggesting the material may be volcanic ash.
The presence of supervolcanoes ultimately changes interpretation of Martian
history as magmatism plays a critical role in the formation of Martian crust and
habitability and atmospheric chemistry of the planet (Tian et al., 2010). Sulphur cycles,
driven by volcanism, control the habitability of soils and rocks on the Martian surface
(Halevy et al., 2007) and strongly affect the climate of the planet in a manner similar to
volcanic eruptions, here on Earth. Global cooling and warming are possible due to
volcanic eruptions as greenhouse gasses emitted can slow the escape of terrestrial
radiation, while volcanic aerosols erupted high into the atmosphere can either reflect or
absorb solar radiation, preventing it from influencing surface temperatures (Robrock
and Mao, 1992). Elucidating the volcanic history of Mars in general, and Arabia Terra
specifically, may provide profound insight into many unanswered question about the
planet.
Martian volcanism (Figure 8) is primarily analyzed through spectroscopy and
photographic data, and is classified based on general morphology. Unfortunately, Siloe
Patera falls within the dusty region of the Arabia Terra, rendering spectroscopy useless,
although thermal inertia can still be used for insight into the nature of the upper ~1 m
14
(Fergason and Christensen, 2008). Determination of volcanic constructs on Mars is
difficult to establish in detail, and are therefore categorized into two classes: central
volcanoes and volcanic plains (Figure 9). Central volcanoes are features that contain
?point source? vents that have accumulated volcanic material. This includes shield
volcanoes, domes and other similar features (Greeley and Spudis, 1981). Volcanic plains
are identified as broad, generally flat-lying units potentially exhibiting flow fronts and
other volcanic features associated with effusive eruptions (Greeley and Spudis, 1981).
Central volcanoes are subdivided into five categories that display point vent
features: Alba Patera, highland patera, shield volcanoes, dome volcanoes, and
miscellaneous central vent volcanoes (Greeley and Spudis, 1981). Alba Patera is a
unique feature that is similar to shield volcanoes with low relief, but also contains sheet
and tube-fed flows. The type volcano for Highland Patera is Tyrrhena Patera which is
characterized by low relief, complex calderas and radially textured channels. Shield
volcanoes on Mars are best typified by Olympus Mons, which contain tube-fed flows,
have moderate relief and broad slopes. Finally, domes are characterized as volcanoes
with steeper sides and are best represented by Tharsis Tholus. Although Siloe Patera
does not completely fit into any of the above categories, it does have some
characteristics typical of Highland Patera. Highland Patera is characterized by low
profiles, complex calderas and channels that radiate out from the caldera (Greeley and
Spudis, 1981). Siloe Patera does not include radial channels, but instead has a low
profile, a complex caldera and is situated in the northern hemisphere where Highland
Patera is generally found (Greeley and Spudis, 1981).
15
Volcanic plains are categorized into four types: simple flows, complex flows,
undifferentiated flows and questionable flows (Greeley and Spudis, 1981). Simple flows,
similar to those observed in Hesperia Planum (Figure 10), are regional and contain
wrinkle ridges without flow lobes. Complex flows have an abundance of flow lobes, are
complex and unlike simple flows, rarely have wrinkle ridges. These flows are best seen
in the Tharsis Plains near Olympus Mons (Figure 11). Undifferentiated flows, common in
the northern plains (Figure 5), rarely have flow lobes and provide very little insight into
their origins, but are likely volcanic. Lastly, questionable flows are heavily modified by
erosion, fracturing, and mass wasting. Like the central volcanoes, Siloe Patera does not
fully fit into any of the volcanic plain categories, but instead has a combination of
features typical of Martian volcanic regions. To the southwest of Siloe Patera (Figure
43), a possible lobe flow emanates out of Siloe Patera?s rim and extends for ~ 45 km.
Figure 8: Volcanism on Mars Multch et al., 1976. Circle indicates general area of Siloe
Patera and other supervolcanoes proposed.
16
Fig
ure
9:
To
po
grap
hic
M
ap
of
Mars
?M
25
M
RKN
? (
US
GS
, 2
00
3)
ov
erlai
n wit
h
im
ag
e o
f v
olcan
ic
un
its
on
M
ars
cate
go
riz
ed
by
m
orp
ho
log
ic
typ
e (Gr
ee
ley
an
d
Sp
ud
is,
19
81
). Acc
ord
ing
to
thi
s m
ap
, Si
loe
Pate
ra
fal
ls i
nto
th
e r
egi
on
of
hig
hlan
d
pate
ra
an
d si
mp
le
flo
ws.
Cir
cle
ind
icate
s g
eneral
regi
on
of
Silo
e P
ate
ra?s
locati
on
.
17
One area of overlap indicates multiple lobe flow events (Figure 43), while regions to the
northwest and southeast look very similar to mare-type wrinkle ridges. These flows are
thought to be the result of fissure eruptions that rapidly expel large amounts of lava and
are common as basin fill and resurfacing of highland terrain (Greeley et al., 2005).
Although calderas are generally similar in origin and geomorphic character,
independent of the planet they form on, the specific processes by which they develop
vary slightly and they range from complex features that form by collapse, explosive
eruptions and erosion (Greeley, 1994), to depressions formed in response to depletion
of a shallow magma chamber (Cattermole, 1989; Frankel, 1996).
Martian calderas are categorized into two main types: Olympus-type and Arsia-
type (Crumpler et al., 1994). Olympus-type caldera (Figure 12A) are characterized by
faulting that defines depression boundaries with arcuate scarps or nested depressions,
inward facing faults, and a scalloped appearance that adds to the chaotic nature of this
type of caldera. The Arsia-type caldera (Figure 12B) is defined by having a caldera that is
larger diameter in relation to its host volcano, a sag-like profile, gently sloping walls, and
multiple boundary faults. Of these two Siloe Patera is more typical of Olympus-type
caldera as they both display sharp arcuate faults, nested depressions, and a chaotic and
scalloped appearance (Crumpler et al., 1994).
All of the features described above can be identified due to their geologically
young age and the lack of time for erosion to modify the feature. Very little is known
about ancient Martian volcanism as older volcanic features on Mars are highly modified
18
or have been completely removed by erosion, or have been potentially misidentified as
impact features.
Figure 10: Simple flows near Hesperia Planum.
Figure 11: Complex flows on the plains near Olympus Mons.
19
Previous studies detailing the caldera collapse process can be used to compare
structures in and surrounding the area of Siloe Patera to those of caldera. Acocella
(2006) sorted caldera collapses into four stages (Figure 13) based on twelve
experimental sets that use a variety of different magma and crustal chemistries. Stage 1
of caldera collapse begins with elastic deformation resulting in gently inward-dipping
structures that creates broad regions of subsidence on the surface. Stage 2 is attained
when subsidence creates a faulted ring that propagates from the magma chamber out
to the surface and is characterized by outward-dipping reverse faults. Stage 3 is
characterized by flexures that form outside of the reverse faults. These flexures
eventually turn into a ring of normal faults that are separate from the smaller ring of
reverse faults form in stage 2. Stage 4 is the final stage and is reached when the inward-
dipping faults developed in stage 3 reach the surface and eventually begin slipping,
resulting in a caldera collapse (Acocella,2006). Siloe Patera has features which are
consistent with this model of caldera evolution, with the exception of the southern rim.
This region contains only one area of collapse, which could be the result of an
incomplete stage four in this area, or spatial overlap of ring faults from stage three and
four (Figure 13).
Resurgent calderas form in seven systematic stages associated with volcanic,
structural, and sedimentary processes (Smith and Bailey, 1968; Figure 14). These stages
20
Figure 12B: Arsia-type Martian caldera.
Figure 12A: Olympus-type Martian caldera.
21
are based primarily on work in Valles Caldera, as well as evidence from other calderas.
Stage I begins with regional tumescence and ring fault propagation due to magma
chamber pressure increasing and magma intrusions. This can result in eruptions along
ring and radial fractures. Stage II begins when large volume ash-flows erupt and
terminate the regional tumescence of Stage I. Ash-flow cycles generally generate
pyroclastic eruptions ranging from 50 to 500 cubic miles in volume. Debate continues as
Figure 13: Depiction of caldera collapse stages as described by Acecolla 2006).
22
to the duration of eruptions due to complex ash cycles and makes generalization
difficult. Caldera collapse starts off Stage III. As collapse and subsidence of the caldera is
a natural consequence of the removal of large volumes of material from the magma
chamber during Stage II. Stage IV is short and accompanied by large amounts of erosion
due to caving and avalanching of the newly formed steep caldera walls. During this time
of erosion, in-fill, and lake formation on top of the caldera, partial magma pressure is
restored. Resurgent doming starts off Stage V. Inital horizontal beds are uplifted to
steep dips that can exceed 65? to form the beginnings of a dome. In-fill continues and
lakes may overflow and breach the caldera. Ring fractures are reopened and provide
Figure 14: Depiction of the first six stages of resurgent calderas (Smith and Bailey, 1968).
Stage I, Regional tumescence and generation of ring fractures. Stage II, Caldera-forming
eruptions. Stage III, Caldera collapse. Stave IV, Preresurgence volcanism and
sedimentation. Stage V, Resurgent doming. Stage VI, Major ring-fracture volcanism.
Stage VII, not shown as volcanic activity is waning (Smith and Bailey, 1986).
23
settings for regional tumescence and potential resumption of Stage II eruption cycles.
Despite the active volcanism, caldera fill continues at a rate greater than erosion.
Following cessation of volcanic activity, fumaroles and hot-springs form, indicating the
beginning of the final stage, Stage VII. At this point the caldera reaches a terminal stage
due to waning volcanic activity with long-lived hydrothermal activity creating the
potential for major ore deposits (Smith and Bailey, 1968).
The region of northern Arabia Terra has been suggested as a new volcanic field
consisting of Martian supervolcanoes based on geomorphology; however, this may be
the result of explosive basaltic volcanism associated with crustal thinning and high gas
pressures. Supervolcanoes have been historically recognized on Earth only and are
characterized by subcircular, negative topographic features and a caldera size that
correlates with the eruption size (Miller and Wark, 2008). On Earth, these
supervolcanoes require high contents of volatiles such as H2O and CO2 as well as melts
of high enough viscosity to amplify pressure on the magma chamber. The main
difference between normal volcanoes and supervolcanoes is amount of the available
material that can be erupted from shallow magma chambers. Due to their enormous
eruptive size, the frequency of these types of eruptions correlates inversely to the
eruption size (Miller and Wark, 2008).
Valles Caldera in New Mexico (Figure15) provides a terrestrial analog for
comparison of extraterrestrial supervolcanoes, including Siloe Patera. Valles Caldera is
24
Fig
ure
15
: C
om
pari
so
n o
f V
all
es
Cald
era,
New
M
ex
ico
(to
p)
an
d Sil
oe
Pa
tera
(bo
tto
m). Arr
ows
po
int
to
vo
lca
nic do
mes
as a
resul
t o
f rin
g faul
tin
g and
circles e
nco
mp
ass r
esurg
ent
cald
eras. Fo
r Sil
oe
Pate
ra, t
he
circle
co
uld
hig
hli
gh
t a p
ossib
le
cin
der
co
ne
or di
ke
inst
ead
of a
resur
gent c
ald
era.
25
roughly 20-25 km in diameter and displays scalloped walls that rise hundreds to more
than 600 meters above the present floor, very similar to Siloe Patera. Redondo Dome, a
resurgent dome, lies in the center of the caldera with diameter of 10-15 km and rises
nearly 900 meters. Surrounding Redondo Dome to the north is a discontinuous ring of
10 isolated rhyolitic volcanic domes, with dimensions of 1-3 km at the base and a relief
of 150-600 meters (Smith and Bailey, 1968). Two main collapse events form Valles
Caldera: the first event created Toledo caldera and the second created Valles caldera as
we see it today. Only a small portion of Toledo caldera is preserved due to later
eruptions that crated Valles caldera. The formation of Valles caldera began with a large
eruption from a ring-fracture system, leading to the subsidence a 10-15 km block at a
depth of 600-900 meters (Smith and Bailey, 1968). Both Siloe Patera and Valles caldera
include deep depressions, although Siloe Patera has a depth almost double that of
Valles caldera. They share a number of commonalities, including evidence for multiple
collapse events, a discontinuous ring of volcanic vents/domes and a resurgent caldera or
cinder cone/dike similar to Kilauea caldera in Hawai?i (Figure 16a and 16B). The Kilauea
caldera is circular with deep, arcuate scarps and two main depressions, indicating
multiple collapse events like those proposed for Siloe Patera. These features are difficult
to explain with nested impact craters or multiple impact events that overlie each other.
Because of similarities in the surface expressions of caldera and impact sites,
impact crater morphology will be compared to Siloe Patera in order to establish a less
equivocal interpretation of its origin. Impact craters can be generally divided into simple
26
and complex craters (Pike, 1980; Figure 17). Fresh simple impact craters
characteristically show overturned strata with overlying ejecta and a bowl-shaped
Figure 16B: Kilauea Caldera viewed from the west. Image from USGS Hawaiian
Volcano Observatory website.
Figure 16A: IKONOS image of Kilauea caldera from NASA?s Earth
Observatory Image of the Day, December 10, 2005.
27
depression that is partially filled. Complex craters are similar to simple craters except
they may also contain interior morphology such as central peaks and walls which
develop from listric faulting, producing a chaotic terrane (Osinski and Pierazzo, 2012).
Impact crater ejecta is identified and classified into three categories (Figure 18):
(1) ?single layer ejecta? (SLE) that contains one ejecta layer; (2) ?double layer ejecta?
(DLE) which contains two complete layers; and (3) ?multiple layer ejecta? (MLE) that has
three or more partial or complete ejecta layers (Barlow et al., 2000). SLE craters typically
range in size of 3 to 50 km in diameter, DLE craters range between 4 to 26 km and MLE
range in between 8 to 59 km (Dohm et al., 2007). Siloe Patera, with a diameter of ~ 30-
40 km, falls into the same size category as SLE and MLE craters. SLE and MLE craters
also have depths similar to Siloe Patera, ranging from 1-2 km (Dohm et al., 2007).
Because of the similarity of dimensions between Siloe Patera and impact craters
containing SLE and MLE, we should expect to see either SLE or MLE ejecta if Siloe Patera
formed from impact craters. Various flow features that propagate out from Siloe
Patera?s rim could be interpreted as portions of SLE or MLE ejecta that have been
severely eroded (Figure 43).
Relative ages of impact craters can be determined by the modified nature of the
impact crater (Figure 19). Fresh impact craters display deep bowl-shaped depressions
with smooth walls and floors and contain very little erosional modification. Slightly
28
Figure 17: Depiction of simple and complex impact craters, their
processes and final results. Simple impact craters on the left and
complex on the right (Osinski and Pierazzo, 2012).
29
modified impact craters appear similar to fresh impact craters, being deep and fresh in
appearance, but with a flat floor and modification to walls. Modified impact craters
become shallower and are subjected to significant modification due to erosion. When
impact craters appear to be erosional remnants with flat floors and significantly shallow,
they are considered to be ghost impact craters (Grand and Shultz, 1993; Arvidson,
1974).Due to Siloe Patera?s age of Noachian/Hesperian it should be considered a ghost
crater having significant time to erode and modify. However, Siloe Patera?s depth to
diameter ratio plots it in between fresh and modified craters that are discussed later.
Central peaks are common in many impact craters as isolated mounds on the
crater floor (Smith, 1976). If Siloe Patera were the result of an impact crater, a single
central peak is expected on the floor of Crater B. There are nine classes of interior
structures that are associated with complex craters (Barlow, 1990): a) symmetric pit
crater; b) asymmetric pit crater; c) summit pit crater; d) central peak; e) peak ring; f) flat
floor ?pristine; g) flat floor ? deposits; h) complex; and i) unclassified (Figure 20). The
mound present in the center of Crater B of Siloe Patera (Figure 3B), is not isolated or
symmetric and extends out to the rim wall. Due to asymmetric nature of the internal
mound, its morphology would be classified as complex. The origin of this feature is more
likely a result of mass wasting from the caldera rim or a resurgent dome associated with
post-eruptive Stage V caldera processes.
Remote sensing is a very useful tool on Earth as every object reflects or emits
characteristic electromagnetic radiation that can be measured from a distance by
30
Figure 19: Variation of crater degradation (Arvidson, 1974).
Figure 18: Examples of Single Layer Ejecta (A), Double Layer Ejecta (B), and Multiple Layer
Ejecta (C) from Barlow 1990.
31
Figure 20: Internal morphology of craters (Barlow, 1990).
32
sensors designed for the purpose (Vincent, 1997). With the help of remote sensing,
many areas can be monitored remotely and used to distinguish different types of land
cover. This is done by measuring the wavelengths of certain objects with multi-spectral
sensors where each object has a specific spectral signature. The wavelengths typically
utilized for this purpose include visible light, (wavelengths of 0.4 ? 24.0 ?m: blue = 0.4
?m, green = 0.5 ?m, and red = 0.6 ?m) short wavelength infrared (SWIR: 1.0-3.0 ?m),
mid wavelength infrared (MWIR: 3.0-5.0 ?m), long wavelength infrared (LWIR: 8.0-14.0
?m), and very long wavelength infrared (VLWIR: 14.0 to 24.0 ?m). The region from 5.0-
8.0 is unnamed because the Earth absorbs almost all of the electromagnetic radiation
for this range (Vincent, 1997). Figure 21 shows different spectral signatures for typical
land cover found here on Earth.
Although remote sensing is highly effective in differentiating between different
types of surface cover on Earth, it presents difficult when attempting to analyze Martian
surfaces. The widespread presence of oxidized rock and sediment on Mars presents a
uniform color reddish-brown color, so visible light classifications are not useful for
differentiate geologic classes on the Martian surface. However, infrared radiation is very
useful when dealing with rocky regions as rocks will absorb energy during the day and
release that energy at night, appearing brighter than the surrounding area. Infrared
maps are generally more successful for producing geologic maps of Mars and other
planets (Vincent, 1997).
33
CHAPTER 3
Multiple data sets were acquired from a variety of sources and used to compare and
analyze Siloe Patera. Data sets include Mars Orbital Laser Altimeter (MOLA) , Mars
Orbital Camera (MOC), High Resolution Imaging Science Experiment (HiRISE), Context
Imager (CTX), and Thermal Emission Imaging System (THEMIS).
The Mars Orbital Laser Altimeter (MOLA) was launched on board the Mars
Global Surveyor (MGS) in 1996 from NASA with primary objectives to globally map the
topography, to measure atmospheric reflectance (1.064 micrometers near infrared) to
better understand the three-dimensional structure of the atmosphere, and to measure
surface roughness at 100 meter scale. Elevation points were converted into both
Figure21: Spectral signatures for remote sensing are generally shown as a
combination of percent reflected back to the sensor and the wavelength in which it
appears. Different materials will present the same spectral signature independent
of region, time, and season (Meaden and Kapetsky, 1991).
34
gridded and spherical models to compensate for the topography of Mars, resulting in
vertical and horizontal accuracies of ?1 m. MOLA measures the overall time it takes for a
laser pulse to reach its intended target, whether it be the Martian surface or
atmospheric layers, and return to the spacecraft. MOLA points are collected at an
interval of 300 m along their tracks and separated from other tracks by 4 km. The
resolution of MOLA reaches its finest detail at 37.5 cm on smooth surfaces but can
drastically increase to as much as 10 m on rough slopes (Smith et al., 2001).
A data set was created using MOLA tracks that transect Siloe Patera and its
surrounding area (Figure 22), generating ~ 150,000 MOLA points containing elevation,
latitude and longitude coordinates as well as orbital information. Points were collected
from the MOLA PEDR Query website between33.16 to 36.49 N and 6.23 to 8.9 E.
Coordinates were selected on the basis of covering the region of Siloe Patera to gather
Figure 22: Good coverage of MOLA tracks along Siloe Patera provides sufficient data to
analyze topography of the area. Siloe shown in box. Image rendered using ArcGIS,
topographic map of Mars, and MOLA tracks downloaded from the USGS Planetary GIS Web
Server (PIGWAD).
35
as much data as possible. MOLA tracks were also collected of Ascraeus Mon?s caldera,
generating 30,500 MOLA points in the region of 11.9 to 10.26 N and 255.8 to 256.0 W,
and a random impact crater, dubbed Crater A, resulting in 78,600 MOLA points covering
24.25 to 24.25 N and 342.2 to 343.13 W. Like that of Siloe Patera, tracks were chosen
based on their relevance to the region of interest and providing a sufficient amount of
points to analyze. Using MOLA tracks, depth/diameter (d/D) ratios can be derived from
typical impact craters and compared to Siloe Patera. Ratios d/D are calculated by the
rim crest height and compared to the depth of the impact crater as measured by MOLA.
The High Resolution Imaging Science Experiment (HiRISE) camera has a 0.5 meter
diameter primary mirror that can take images as large as 28 Gb (gigabytes) in as little as
6 seconds. Images display a detailed resolution at 0.25 to 1.3 m/pixel and can measure
topographic changes to better than 25 cm of vertical displacement. HiRISE orbits at an
altitude ranging from 255 - 320 km and has a minimum resolution of less than 30
cm/pixel at a distance of 200 km and a swath width of greater than 3 km from 200 km
(McEwen et al., 2007). HiRISE has a ground tracking velocity of 3.2 km/sec which can
cause difficulty when trying to take detailed images. In order to adjust for this issue,
multiple images were taken, up to 128 times, and averaged to increase the signal-to-
noise ratio, resulting in a detailed image. The top priority is to collect ~ 1000 stereo pairs
which provide two overlapping images that can be manipulated to generate an
elevation map as well as a 3D image (McEwen et al., 2007).The objective of this camera
is to photograph ~ 1% of the Martian surface in the first 2 years of its Primary Science
Phase (McEwen et al., 2007).
36
Through the HiRISE website, twelve suggestions were made for future imaging
sites in and around Siloe Patera (Figure 23). Currently, only one request has been
fulfilled (Figure 25) with an additional image of Siloe Patera not per my request.
Requests from another research observing fluvial features, covers the remaining areas
of Siloe Patera and was chosen for imaging (Figure 24). Image suggestions are rated by
importance which can render many, if not all, suggestions on a low priority which may
never be fulfilled. Table 1 shows a list of requested HiRISE image suggestions with image
82618 being the only one returned. Figures 24 and 25 show the HiRISE images taken for
this project
The Context Camera (CTX) on board the Mars Reconnaissance Orbiter launched
August 2005 and entered Mars?s orbit March 2006. CTX has a great ability to obtain
image at 30 km wide and 40 km long at a resolution of 5.0-6.5 m/pixel. Although this is a
lower resolution than those taken by HiRISE, it has the ability to image the majority of
the Martian surface with the most detail to date. This is a drastic improvement to Viking
images with a resolution of 8-30 m/pixel (Malin et al., 2007).
As CTX image provided the most detail over the largest area, these images are
the best choice for close observation. Nine CTX images were downloaded from Arizona
State University?s (ASU) Mars Image Explorer (Figures26, 27 and Table 2).
The Mars Orbiter Camera (MOC) was launched November of 1996 on the Mars
Global Surveyor (MGS) and reached Martian orbit by September, 1997. One of MOC?s
greatest discoveries was that Mars has areas that are kilometers deep in layered
material, suggesting massive erosion periods. MOC has four types of imaging: 1) global
1
2
37
Figure 23: Image showing HiRISE requests. Yellow boxes indicate areas where red indicates
the areas that have been photographed. Photographed images are designated as 1 (being
the more northern of the two), and 2 (being the southern). The center of Siloe is located at
35? N and 6? E.
ID Latitude Longitude ID Latitude Longitude
92040 35.398 6.746 84536 35.785 5.911
92018 34.538 7.674 84534 34.909 5.999
92014 34.785 5.89 82622 35.459 6.418
92012 34.942 6.729 82620 35.404 6.566
84792 34.493 7.092 82618 35.245 6.337
84538 35.109 6.609 82616 35.149 6.817
Table 1: HiRISE image suggestion list
38
Figure 24: HiRISE image 1
(ESP_033258_2160_MRGB)
showing fluvial erosion and
possible thermokarst
structures emanating from
the wall. The brown strip in
the middle is shown in true
color.
1
39
Figure 25: HiRISE image 2
(ESP_030344_2155_MRGB) is
more interesting than the
first HiRISE image as it shows
isolated mounds that could
be the result of eroding ash,
definable layered deposits
that continue for long
distances, steep cliffs and a
chaotic and eroded terrane.
As in HiRISE image 1, the
brown strip shows true color.
2
40
maps with 7.5 km/pixel resolution and swaths using red and blue wide-angle cameras
for weather monitoring; 2) narrow-angle images with 1.5-12 m/pixel resolution to
analyze geomorphology and geology; 3) context frames taken at wide-angle with 240
m/pixel, generally in red or blue; 4) wide-angle images ranging from 240-960 m/pixel
Fig
ure
26
: P
osi
tio
n an
d lo
catio
n o
f C
TX
, M
OC,
and
HiR
ISE
im
ag
es c
ov
erin
g Sil
oe
Pate
ra.
Let
ters
are
po
siti
on
ed in
th
e t
op
left
co
rner
of t
he
im
ag
e an
d s
om
e arr
ows
po
int
to
the t
op
left
co
rner
, fo
llo
we
d by
the le
tte
r as lab
eled in
Tab
le
1.
41
Table 2: CTX, MOC, and HiRISE Product Identification and location
downloaded from Arizona State University?s Mars Image Explorer.
Images and locations are displayed in Figure 12. Letters are
positioned in the top left corner of all images
centered around monitoring regional areas (Malin and Edgett, 2001).Five MOC images
were taken of Siloe Patera (table 1) labeled J ? N (Figure 28 and 29).
CTX Image Number Location
A B18_016828_2143_XI_34N354W 34.06?N 5.88?E
B P21_009325_2160_XN_36N354W 35.72?N 6.04?E
C G21_026388_2133_XN_33N353W 33.08?N 6.48?E
D B19_017184_2187_XI_38N353W 38.41?N 6.11?E
E B01_009892_2148_XI_34N353W 34.51?N 6.82?E
F B19_016973_2136_XN_33N352W 33.39?N 7.91?E
G P14_006464_2163_XI_36N352W 36.00?N 7.31?E
H B07_012516_2154_XN_35N352W 35.17?N 7.47?E
I D05_029355_2163_XN_36N352W 36.06?N 7.63?E
MOC Image Number
J E18/00445 35.06?N 6.51?E
K M23/01285 35.31?N 6.53?E
L S14/02434 35.19?N 6.64?E
M E11/00223 35.31?N 6.73?E
N S15/02431 35.58?N 6.70?E
HiRISE Image Number
O ESP_033258_2160 35.47?N 6.29?E
P ESP_030344_2155 35.23?N 6.34?E
42
A
B
Figure 27: Individual CTX images that were used. Image number and location are found in
Table 2. Continued on next three pages.
43
C D
44
E F
45
THEMIS is on board the 2001 Mars Odyssey mission with intent to investigate
the surface mineralogy using multi-spectral thermal-infrared images. These images
contain nine wavelengths with a concentration of 6.8 to 14.0 ?m, and five bands in the
visible/near infrared centered from 0.42 to 0.86 ?m. The goal of THEMIS is to map Mars
entirely in both day and night multi-spectral infrared images with a resolution of 100-m
Figure 28: Map of MOC images used. Image number and location can be
found in Table 2.
46
J K M L
Figure 29: MOC images showing several transects
through Siloe Patera.
47
per pixel. The two primary objectives of THEMIS are to resolve surface features at
scales less than 100 m, and to identify minerals and compositional units at 100 m spatial
scales (Christensen et al., 2004). THEMIS Day IR casts shadows and can increase visibility
to raised areas such as the flow feature to the southwest of Siloe Patera. However, due
to shadows, it can create discrepancies with thermal inertia of various rocks. THEMIS
Night IR does not maintain the effects of shadows and will display solid rock as white
and dust/sand as black (Figure 30).
In order to generate a rough geologic map of Siloe Patera, Erdas Imagine must be
used to classify different land covers. There are two primary methods to classify an
image; supervised and unsupervised. In a supervised classification, classes are generated
from training samples designated by the user. Unsupervised classifications are
generated by the computer and can later be adjusted by the user to determine what
land cover each class is. For this paper, unsupervised classification was used with
THEMIS Night IR, MOLA, and CTX images. The unsupervised classification used was
generated by the ISODATA clustering algorithm. This algorithm assigns an arbitrary
initial cluster center where each pixel is classified to a center that is closest. In the next
step, different cluster centers are chosen and each pixel is again classified to a cluster
center. This process is done until there is little change between the iterations (Omran et
al., 2002).
CHAPTER 4
48
Figure 30: Siloe Patera shown in THEMIS Day IR (top) and THEMIS Night IR
(bottom).
49
A clip of THEMIS Night IR, MOLA map set to 20 classes, and multiple CTX images
mosaicked together, were used to generate a geologic map of Siloe Patera in Erdas
Imagine. The THEMIS Night IR clip was cut from the global THEMIS Night IR map
provided by the U.S.G.S Planetary GIS Web Server (PIGWAD) website. The MOLA map
was rendered using MOLA track points that were previously collected and were placed
into 20 classes at equal intervals based on elevation. Multiple MOLA maps were created
for this project and classes ranged from 20 to 250 but fewer classes helped pull out
more features, so 20 classes were used in the final render. Unfortunately, there is no
global CTX mosaic image, so each individual CTX image had to be mosaicked by hand.
Images were easily georeferenced in ArcGIS, but were not easily mosaicked. Each CTX
image is surrounded by a black border that compensates for the image?s projection that
necessitated removal in Adobe Photoshop before all images could be mosaicked
together. Once the CTX images were mosaicked and imported into ArcGIS, they were
georeferenced to the THEMIS Night IR and MOLA images.
To minimize file size, the THEMIS Night IR clip, MOLA, and CTX mosaic images
were all set to 300 dpi with a width of 3300 pixels and a height of 2500 pixels and
exported from ArcGIS. Erdas Imagine 2011 was used to combine these three images into
one using the Layer Stack tool. THEMIS Night IR was imported onto band 1, MOLA was
imported onto band 2, and CTX mosaic was set to band 3. Once rendered and opened,
the image is displayed as a color stretch with green representing rockier regions and red
indicating very dusty areas (Figure 31).
50
An unsupervised classification of the color stretch image was generated to
produce a geologic map. Twenty classes were chosen and each class was analyzed by
color to determine into which land class it fell(Figure 32, Table 3). Three land types were
used in this classification Table 3): (1). Dust (tan); (2) Rock (brown); (3) Sediment fill-in
Fig
ure
31
: C
olo
r st
ret
ch
of
Silo
e P
atera w
ith T
HE
MIS
Nig
ht
IR, M
OL
A,
an
d CT
X m
osaic im
ag
es
co
mb
ined
into
on
e i
mag
e with
each
rep
resen
tin
g an
in
div
idu
al ba
nd
. Red
rep
res
en
ts l
ess r
efl
ectiv
e sur
fac
es,
sug
gest
ing
hig
h du
st
co
ver
ag
e an
d green
sho
ws
are
as w
ith
hard
ro
ck ad are
m
ore
refle
cti
ve
at
nig
ht
du
e
to ins
olati
on
abs
orp
tio
n du
rin
g t
he
day.
51
(turquoise). These land types were chosen because they are the most important land
cover types needed to create a geologic map for this research. Siloe Patera shows hard
rock in the scarps and a mainly hard rock floor with areas of sediment fill-in on the floor
and bench. It is not characteristic for a flat-floored impact crater to have rock exposed in
areas other than the central peak. The brown line across Figure 32 is the result of a gap
in between CTX images and should be ignored.
Combining the color stretch image, classified image, CTX, and THEMIS Night IR
images, a rough geologic map was rendered (Figure 33), similar to that of Mailam, et
al.(2003) and Martinez-Alonso et al, (2005) (Figure 31). This map was categorized into
nine different units/features and each was chosen to best represent a region based on
all images. Solid rock is exposed in much of Siloe Patera and contains small amounts of
dust located on the east side of the bench. Unit 1 fill-in and Unit 2 Fill-in are represented
as one unit in the classification map but under further review of CTX images, were
divided into two units. Unit 1 fill-in represents areas that appear to be older and have
had time to slightly alter or erode. Unit 2 fill-in is found in low lying flat areas that are
likely the result of younger runoff deposition according to its stratigraphic position.
Post-volcanic doming regions show up as solid rock in in low lying flat areas that are
likely the result of runoff deposition and must be younger according to its stratigraphic
position. Post-volcanic doming regions show up as solid rock in the classification map as
well as in the stretched color map, while the spire shows up as one of the brightest
areas in THEMIS Night IR and in the stretched color map.
52
Color Red Green Blue Class Land Type
Tan 0.823529 0.705882 0.54902 Class 1 Dust
Turquoise 0.25098 0.878431 0.815686 Class 2 Fill-in
Brown 0.647059 0.164706 0.164706 Class 6 Rock
Fig
ure
32
: G
eo
log
ic m
ap
of
Silo
e P
atera sh
owing
are
as
of du
st
(tan
), r
oc
k (br
own),
and
se
dim
ent
fill
-in
(turq
uo
ise). Fig
ure
w
as rend
ered
usin
g T
HEM
IS
Nig
ht
IR, M
OL
A, an
d C
TX
m
osaic i
mag
es
to
pro
du
ce
an
un
sup
erv
ised
classifi
cati
on
of
20
clas
ses.
Table 3: List of three classes and land cover types used to produce
classification map in Erdas Imagine (Figure 24).
53
The geographic coordinate system used was the GCS_Mars_2000_Sphere and
the projection was Equidistant Cylindrical with map units in kilometers. MOLA data was
collected from MOLA PEDR Query website and points ranged from 36.5 ? 33.0 N and 4.0
? 9.0 E for Siloe Patera, 10.0 ? 11.0 N and 103.0 ? 105.0 W for Ascraeus Mons, and 24.0
? 26.0 N and 17.45 ? 16.0 W for Crater A. These points were downloaded and imported
to Excel where columns were formatted for importing into ArcGIS. Once imported into
ArcGIS, XY data was displayed and exported into a shape file in order to be manipulated.
Once MOLA points were in shapefile format, they were used to produce interpolated
maps that fill in areas with no data. The Inverse Distance Weighted (IDW) function was
used to produce elevation map. The IDW method allows points in proximity of the
processing cell to have a weighted influence determined by its distance from the
processing cell. Therefore, closer points are likely to have similar elevations than those
points that are further away. This same process was done for Ascraeus Mons and Crater
A, B, C, and F providing detailed elevation maps.
Once IDW maps were produced (Figure 34), contour maps were rendered at
intervals of 5 meters. The 5 meter interval and a 100 meter interval were chosen to
provide detail of the topography as well as a less detailed map that covers a greater
area (Figure 35). Interpolated maps of MOLA points are necessary in order to generate
contour values as they cannot be generated from MOLA points alone.
Once Siloe Patera, Ascraeus Mons and Crater A, B, C, and F were completed
using IDW and resultant contour maps, Triangular Irregular Network (TIN) maps were
54
generated using the contour maps (Figure 36). TIN maps are essential for ArcGIS to
represent the surface morphology of features, unlike the flat surfaces seen in the IDW
and contour maps. Vertices are created and connected by a series of edges to produce a
network of triangles. Once these TIN maps were produced they were imported into
ArcScene to produce a 3D model. This 3D model was set to have a vertical exaggeration
of 24x which helps to accentuate features of Siloe Patera that were not previously
Figure 33: Geologic map of Siloe Patera rendered in ArcGIS and Adobe Photoshop using a
color stretch map, classification map, THEMIS Night IR, and CTX images.
55
Fig
ure
34
: Im
ag
es
are r
end
ered usin
g ArcGI
S?s
inte
rpo
lati
on
to
ol.
Sil
oe
(to
p left
), As
craeus
Mo
ns
(to
p rig
ht)
, and
Crat
er
A (ri
gh
t) w
ere
all
generat
ed
with 3
0 c
lasses
and
se
t to
a gra
y sc
ale. All
we
re
generated usin
g I
DW
fro
m
Ar
cGIS
.
56
Fig
ure
35
: Sil
oe
Pat
era (
top
left
), Ascraeus
M
on
s
(to
p rig
ht)
, and
Crat
er
A (ri
gh
t) are
disp
lay
ed her
e
in c
on
tou
r in
ter
vals
of
10
0 m
eters. T
his
w
as
acco
mp
lished
by
ArcGI
S thro
ug
h t
he
us
e o
f th
e
spa
tial an
alys
t to
ol and
im
pu
tin
g t
he
IDW
rend
ered la
yers
an
d assig
nin
g a
co
nto
ur in
terv
al.
57
noticed in 2D images. Higher and lower numbers exaggerate the image to the point that
it becomes unrecognizable. Once the 3D model was produced, CTX and THEMIS images
Fig
ure 3
6: I
mag
es
of S
ilo
e P
atera
(to
p l
eft),
Ascr
aeu
s
Mo
ns
(to
p rig
ht)
, and
Crate
r A (rig
ht)
ar
e sh
own as
TIN
lay
ers.
Th
ese
lay
ers
are
generat
ed fro
m
co
nto
ur
map
s and
disp
lay
a
dig
ital s
urface. Thes
e m
ap
s c
an
be
used t
o c
reat
e 3
D m
od
els.
58
were draped on top. This required assigning each image a layer to be draped on which
was achieved by viewing the properties, designate the CTX and THEMIS images as a float
layer, and then assign the layer it would be draped over.
All CTX, HiRISE, MOC, and THEMIS images were georeferenced to THEMIS day
global map downloaded from PIGWAD. For each image, at least 8 reference points were
placed with an acceptable root mean square error of less than one. Images were
downloaded with a standard black background that can be hidden.
Once TIN maps of Siloe Patera were created, they were imported to ArcScene
where a 3D model was generated. This was done by importing a TIN map generated at
five meter contour intervals in ArcGIS. Five meter contour intervals produced a detailed
view of Siloe Patera without crashing the software due to excessive file size. Once the
TIN map is imported, adjustments were made to that Siloe Patera could be accurately
depicted. Due to Siloe Patera?s vast diameter compared to its depth, vertical
exaggeration provides a better view of its features that are not noticeable without
vertical exaggeration.
Once the 3D model was generated, THEMIS and CTX images were overlain to
help identify features and points of interest. Draping images was completed by selecting
its properties, base height, and the layer over which it will be draped, which in our case
was the TIN layer. This produces an interactive 3D model that displays north in the
direction of the green arrow located in the top left corner (Figures 37 and 38). Scales for
59
Figure 37: Siloe Patera shown in a basic 3D color map (above) and in CTX (below).
Volcanic domes are outlined hexagons, the spire is outlined by a rectangle, and
the faulting region is circled by an ellipse.
60
Figure 38: Siloe Patera shown in THEMIS Day IR (above) and in THEMIS Night IR
(below). Volcanic domes are outlined by hexagons, the spire is outlined by a
rectangle, and the faulting region is circled by an ellipse.
61
the 3D model images are not displayed because scale is not continuous when viewing
Siloe Patera at different angles.
When viewing Siloe Patera in 3D, new features stand out where they did not
when viewed as a 2D image. These features include multiple peaks along the scarp
inside Siloe Patera, possible ring faulting along the north eastern portion of the bench,
and a feature ?spire? which emanates from the eastern wall of Siloe Patera and runs to
the center of Crater B?s floor (Figure 3B). These features are evident in, CTX, and THEMIS
day and night 3D images. Features are outlined in plain and CTX 3D models.
There are multiple regions in Siloe Patera which may provide insights into its
origin, including multiple raised mounds along the scarp line of the bench, faulting along
the north eastern portion of Crater A?s bench, flow features, a large depression to the
south and an elongated mound on the floor of Crater B referred to here as a spire.
The raised mounds along the fault scarp dividing Crater A from Crater B show
clear isolated peaks in MOLA data and are divided by channels. These five mounds
appear to be solid crystalline rock that was formed by volcanic eruptions and not the
result of sediment accumulation. Each one ranges in height and shape and appear bright
in the classification map and THEMIS Night IR indicating exposed rock (Figures 31 and
32). These mounds, however, are not as bright as the spire on the floor of Crater B and
could be covered by layers of dust or ash.
Along the right side of the bench of Crater A is a series of concentric exposures
that appear to be ring faults (Figure 3B). These faults run from the edge of the fault
scarp, toward the northern rim of Siloe Patera, slightly curving to the northwest. These
62
faults are degraded and likely covered by layers of dust or ash, leaving little room for
rock exposure. The classification map and THEMIS Night IR, however, suggest minor
exposures of rock along some of the faults (Figures 31 and 32).
Three flow features (Figure 43) propagate form the north, southeast, and
southwest regions of Siloe Patera. They vary in length with the southwest flow being the
longest, southeast flow much shorter and north flow slightly shorter than the southeast
(Table 6). Classification maps and THEMIS Night IR show areas of rock and dust
throughout the flows but the head of the flows appear brighter than the surrounding
area (Figures 23 and 24). The southwestern flow also shows an overlap of lobes
indicating at least two flow events created this feature (Figure 43).
To the south of Crater B?s southern rim is a large area of subsidence, likely once a
lake due to a high number of stream channels that flowed into it. The eastern-most
region is the shallowest and shows up with high thermal inertia in THEMIS Night IR and
in the classification map (Figures 31 and 32). The high thermal inertia is likely exposed
evaporate or mud deposits from the probable lake.
The spire located on the floor of Crater B is broken into two parts with one
isolated mound and an elongate mound. They extend and grade up in elevation until
they connect to the southeastern wall of Siloe Patera. MOLA data shows that it is higher
than the surrounding floor with elevations that rise toward the rim. It appears to be a
primary igneous feature, possibly an eroded cone or dike, rather than a clastic
sedimentary deposit. The classification map, as well as the THEMIS Night IR map, shows
the spire and surrounding area to be one of the brightest regions on the map (Figures 31
63
and 32). This means the area has high thermal inertia generated from crystalline rock
and is not likely to be a loosely consolidated sedimentary rock.
Topographic profiles of Siloe Patera, Ascraeus Mons, and Crater A, B, C, and F
(Figure 39) are essential for determining the true elevation of each feature and their
internal morphology. These profiles are produced in ArcGIS after TIN maps are created
of the areas of interest. Using the 3D analyst interpolate line tool, lines are drawn to
intersect Siloe Patera, Ascraeus Mons, and Craters A to provide a topographic profile of
each. Lines were drawn at an approximate distance of 60,000 meters to maintain similar
profile length for comparison.
To compare topographic profiles, all data are was and imported into Excel with
X, Y, and Z coordinates. Elevation and distance are reported in meters, but to simplify
distance variables were converted into kilometers. Elevation is displayed in relation to
the feature?s true elevation on Mars and not normalized to zero. This created difficulty
when comparing profiles as they would not plot over one another but instead above or
below. To correct this problem, elevation data was normalized such that all features
begin at zero distance and zero elevation. All graphs were overlain in order to facilitate
comparison (Figure 31 and 32).
Topographic profile comparison of Siloe Patera, Ascraeus Mons, Crater A, B, C
and F(Figure 40) shows that Siloe Patera resembles the profile of Ascraeus Mons more
than Craters A, B, C and F. This is shown by its arcuate scarped walls and steep slope
angle, unlike the more shallowly sloped wall angles and non-arcuate scarps of the
impact craters. Siloe Patera also lacks a raised rim and no central morphology, like
64
Fig
ure
39
: T
op
og
rap
hic pr
ofil
e c
om
pari
so
n o
f Sil
oe
Pat
era agai
nst
As
craeus
Mo
ns
(cald
era),
Crater
A, B,
and
C, and
Crate
r F
(neste
d crater
). P
rofil
es w
ere
do
wnlo
ad
ed
fro
m Ar
cGI
S after
TIN
m
ap
s w
ere
crea
ted
of
each
fe
atur
e. D
ata
was im
po
rted
into
Exc
el and
m
an
ipu
lated
to
sh
ow
their p
rofil
es.
65
central peaks or central pit craters, which are common in many impact craters. Crater A
is located at 25? N and 16? W and is ~ 35 km in diameter. Crater B is at 28? S and 54? E
and is roughly 30 km in diameter. Crater C has a diameter of 60 km and is found at 26? S
and 172? W. Crater F is found at 27?S and 173?W with a diameter of ~56 km.
Six topographic profiles (Figures 41, 42, and 43) and a contour map of Siloe
Patera were created to show its various regions including the spire, volcanic doming and
Fig
ure
40
: L
ines
ind
icat
e ar
eas
the
to
po
grap
hic pr
ofil
es in fig
ure
30
we
re re
nd
ere
d fro
m. A.
Silo
e P
atera, B. As
craeus
Mo
ns, C. Cr
ate
r A, D
. Crat
er
B, E
. Crat
er
C, F. C
rate
r F.
To
po
grap
hic
lin
es
we
re
dra
wn in
an eff
ort
to best
re
pre
sent
each f
eatur
e.
66
arcuate slopes. Figure 41 begins with a transect through the bench and across the spire
shown trending from the northeast to the southwest. This profile begins with a gentle
slope, and not a raised rim associated with impact craters, and moves down to the
bench. This portion of the bench is characterized by ring faulting that grade down to an
arcuate scarp. The profile transects the most central portion of the spire. Using only this
profile, the spire could be mistaken to be a central peak. However, incorporation of all
the transects across the spire make interpretation of this structure as an impact-related
central peak less likely. The second profile crosses the largest volcanic dome in Siloe
Patera, trending form the northwest to the south east. This dome is ~500 m higher than
the surrounding elevation of the bench and is approximately 2.5 times higher than a
projected impact crater rim from Crater B (Garvin, 2000). Typical impact craters have
raised rims that circle the entire crater; however, this is not the case with Siloe Patera as
it has isolated raised areas that exceed typical impact crater rim heights and are not
continuous as those associated with impact craters. The topographic profile then
transects the furthest point of the spire, giving the appearance of a central peak. Similar
to profile 1, this transect does not show the full extent of the spire that would disprove
a central peak. The third topographic profile shows the transect of the spire trending
from the southeast to the northwest. The spire grades down in elevation from the rim
toward the center of the floor of Crater B and has multiple raised areas. Viewing the
spire from this profile, it is apparent that the spire is not an isolated central peak but
67
Figure 41: Three topographic profiles of Siloe Patera showing the bench, a volcanic
dome, and the spire.
68
instead is a peninsula shaped formation that extends from the walls of Siloe Patera. As
the profile moves toward the northwest, there is an abrupt transition from the floor to
the bench and again from the bench to the walls of Crater A. Impact craters are known
for their characteristic bowl-shaped depressions; however, Siloe Patera does not display
the characteristic bowl-shape with its abrupt transitions seen on the northwestern side
of profile 3.
Figure 42 begins with a topographic profile of the whole basin trending from east
to west showing the highest elevation to the lowest elevation. There are no raised rims
associated with this profile; instead the area outside of Siloe Patera is relatively flat. Due
to the location this profile was drawn, it shows a profile similar to profile 3 of the spire
with arcuate scarps and abrupt changes from the floor and bench to the walls. The next
two profiles transect the lowest and roughly middle elevation of the bench. Profile 5
runs roughly north to south along a channel from the bench to the floor of Crater B that
likely had flowing water at one point. This channel is located in between the volcanic
dome (profile 2) and the ring faulting (profile 1). It is difficult to interpret the raised
areas of the bench to be the raised rim of an impact crater (Crater B) as the channel in
profile 5 would need to cut completely through and erode the rim down by ~500 m. The
lowest elevation of the bench reaches only ~500 m from the floor of Crater B. Profile 5
steep slopes but appears to have a slight U-shape due to the slope of the channel. The
last topographic profile, profile 6, has a broad wedge shape and transects the
approximate middle elevation of the bench in between two volcanic domes. This profile
69
Figure 42: Three topographic profiles of Siloe Patera showing the whole basin, lowest
bench elevation, and middle bench elevation.
70
trends from the northwest to the southeast with another transect of the spire and
displays slopes that are not as steep as the other profiles.
If Siloe Patera were the result of two nested impact craters, we would expect to
have a central peak that is isolated from the surrounding walls; instead the spire is not
isolated and forms a peninsula shape emanating from the southeastern wall of Siloe
Patera. A raised rim from Crater B would be expected on the bench and although there
are raised areas, they are isolated and are more than double the height of a raised rim
typical of impact craters. The majority of slopes shown in Figures X and Y are more steep
than would be expect for an impact crater and present a general wedge shape overall.
Figure 43: Topographic profiles from Figures 41 and 42.
71
Siloe Patera does show raised rims around Crater A; however, they are gentle slopes
and a characteristic part of the caldera evolution as described by Smith and Bailey
(1968). A contour map of Siloe Patera helps to emphasize the steep slopes, spire,
volcanic doming and lack of a raised rim (Figure 44).
Slope angles of impact craters, volcanoes, and Siloe Patera can also provide clues
as to Siloe Patera?s origin. Generally, impact craters are bowl shaped and have smooth
and gently sloping walls with low angles. Volcanoes, especially Ascraeus Mons, have
much more steeply angled slopes from the arcuate scarps that mark the extent of a past
collapse. Siloe Patera is much more similar to Ascraeus Mons than impact craters in this
aspect.
In an effort to measure the angles of the slopes, trendlines were inserted on the
topographic profiles of each feature (Figure 45). To do this, another graph was
generated with only the coordinates of the right slopes of each feature. All graphs were
generated and copied and pasted into one graph where a best fit trendline was placed.
This graph was then imported into Adobe Photoshop where the angles of the trendlines
were measured using the ruler tool. Trendlines of slope angles are as follows: Siloe
Patera - 69?, Ascraeus Mons - 67?, Crater A - 42?, Crater B - 49?, Crater C - 60?, Crater F -
64? (Table 4). Siloe Patera, with an angle of 69 degrees, is more like Ascraeus Mons (67
degrees) than the four impact craters of various preservation levels, which range from
42 to 60 degrees. Crater slope angles vary greatly depending on their degradation and
the slope measured which produces a wide range of angles; however, Siloe Patera
contains a slope angle even higher than Ascraeus Mons.
72
Although these slope angles give an idea as to the comparison of Siloe Patera,
Ascraeus Mons, and multiple impact craters, they do not speak for the overall slope
angle of each. Using ArcGIS, a slope degree map was produced that color codes slope
Fig
ure
44
: C
on
tou
r m
ap
of
Si
loe
Pat
era w
ith ele
vati
on
lab
els.
73
Siloe Patera 69?
Ascraeus Mons 67?
Crater A 42?
Crater B 49?
Crater C 60?
Crater F 64?
Figure 45: Slope angles of Siloe Patera, Ascraeus Mons, Craters A, B, C, and F.
Siloe Patera contains the highest slope angle with Ascraeus Mons following.
Table 4: List of slope angles from
Figure 41.
74
degree ranges. Figure 46 shows red as having the highest slope degree from 64? to 80?
and white having a relatively flat slope from 0.004? to 5?. Siloe Patera clearly has very
steep slopes on all sides with the volcanic doming showing steep slopes as well.
75
Other than the spire, Siloe Patera has no apparent internal morphology, raised
rim, or definable and uniform ejecta, all of which are indicators and characteristics of
impact craters. According to Robbins Crater Database 20120821 (Table 5), Siloe Patera
shows impact crater morphology of gullies, ejecta deposits, floor deposits, and slumps.
Despite the apparent lack of ejecta, and ejecta and morphology fields labeled null, Siloe
Patera has been labeled as an impact crater with the confidence of 4 out of 4. Barlow?s
Crater Database Version 1(Table 5) states that there is no definable internal morphology
or ejecta for Siloe Patera. Barlow only defines one crater for Siloe Patera whereas
Robbins defines two different impact craters (Figure 47); however there is little
Figure 46: Slope map of Siloe Patera. Red shows steeper slopes and white very shallow
slopes.
76
Robbins Crater Database
OBJECTID 28414
CRATER_ID 05-000145
LATITUDE_CIRCLE_IMAGE 35.344
LONGITUDE_CIRCLE_IMAGE 6.529
LATITUDE_ELLIPSE_IMAGE 35.503
LONGITUDE_ELLIPSE_IMAGE 6.532
DIAM_CIRCLE_IMAGE 33.59
DIAM_CIRCLE_SD_IMAGE 0.1
DIAM_ELLIPSE_MAJOR_IMAGE 35.16
DIAM_ELLIPSE_MINOR_IMAGE 33.41
DIAM_ELLIPSE_ECCEN_IMAGE 0.31
DIAM_ELLIPSE_ELLIP_IMAGE 1.05
DIAM_ELLIPSE_ANGLE_IMAGE 20
LATITUDE_CIRCLE_TOPOG 35.377
LONGITUDE_CIRCLE_TOPOG 6.518
DIAM_CIRCLE_TOPOG 36.03
DIAM_CIRCLE_SD_TOPOG
DEPTH_RIM_TOPOG -2.09
DEPTH_RIM_SD_TOPOG 0.11
DEPTH_SURFACE_TOPOG -2.19
DEPTH_SURFACE_SD_TOPOG 0.19
DEPTH_FLOOR_TOPOG -3.35
DEPTH_FLOOR_SD_TOPOG 0.05
DEPTH_RIMFLOOR_TOPOG 1.26
DEPTH_RIMHEIGHT_TOPOG 0.11
DEPTH_SURFFLOOR_TOPOG 1.16
PTS_USED_RIM_IMAGE 145
PTS_USED_RIM_TOPOG 71
PTS_USED_SURFACE 131
PTS_USED_FLOOR 29
PTS_USED_LAYER_1
PTS_USED_LAYER_2
PTS_USED_LAYER_3
PTS_USED_LAYER_4
PTS_USED_LAYER_5
NUMBER_LAYERS 3
MORPHOLOGY_CRATER_1 CpxUnc
MORPHOLOGY_CRATER_2 Gullies
77
MORPHOLOGY_CRATER_3 Ejecta Deposits / Floor
Deposits / Slump
Deposits
MORPHOLOGY_EJECTA_1 MLERS
MORPHOLOGY_EJECTA_2 HuBL
MORPHOLOGY_EJECTA_3
MORPHOLOGY_EJECTA_COMMENTS
DEGRADATION_STATE
CONFIDENCE_IMPACT_CRATER 4
LAYER_1_PERIMETER
LAYER_1_AREA
LAYER_1_LOBATENESS
LAYER_1_EJECTARAD_EQUIV
LAYER_1_EJECTARAD_REL
LAYER_2_PERIMETER
LAYER_2_AREA
LAYER_2_LOBATENESS
LAYER_2_EJECTARAD_EQUIV
LAYER_2_EJECTARAD_REL
LAYER_3_PERIMETER
LAYER_3_AREA
LAYER_3_LOBATENESS
LAYER_3_EJECTARAD_EQUIV
LAYER_3_EJECTARAD_REL
LAYER_4_PERIMETER
LAYER_4_AREA
LAYER_4_LOBATENESS
LAYER_4_EJECTARAD_EQUIV
LAYER_4_EJECTARAD_REL
LAYER_5_PERIMETER
LAYER_5_AREA
LAYER_5_LOBATENESS
LAYER_5_EJECTARAD_EQUIV
LAYER_5_EJECTARAD_REL
CRATER_NAME
Shape Point
OBJECTID 28500
CRATER_ID 05-000231
LATITUDE_CIRCLE_IMAGE 35.194
78
LONGITUDE_CIRCLE_IMAGE 6.598
LATITUDE_ELLIPSE_IMAGE 35.197
LONGITUDE_ELLIPSE_IMAGE 6.597
DIAM_CIRCLE_IMAGE 26.27
DIAM_CIRCLE_SD_IMAGE
DIAM_ELLIPSE_MAJOR_IMAGE 27.96
DIAM_ELLIPSE_MINOR_IMAGE 24.76
DIAM_ELLIPSE_ECCEN_IMAGE 0.47
DIAM_ELLIPSE_ELLIP_IMAGE 1.13
DIAM_ELLIPSE_ANGLE_IMAGE 86
LATITUDE_CIRCLE_TOPOG 35.203
LONGITUDE_CIRCLE_TOPOG 6.602
DIAM_CIRCLE_TOPOG 26.62
DIAM_CIRCLE_SD_TOPOG 10.04
DEPTH_RIM_TOPOG -2.71
DEPTH_RIM_SD_TOPOG 0.41
DEPTH_SURFACE_TOPOG -3.04
DEPTH_SURFACE_SD_TOPOG 0.32
DEPTH_FLOOR_TOPOG -3.89
DEPTH_FLOOR_SD_TOPOG 0.03
DEPTH_RIMFLOOR_TOPOG 1.18
DEPTH_RIMHEIGHT_TOPOG 0.33
DEPTH_SURFFLOOR_TOPOG 0.86
PTS_USED_RIM_IMAGE 186
PTS_USED_RIM_TOPOG 78
PTS_USED_SURFACE 34
PTS_USED_FLOOR 51
PTS_USED_LAYER_1
PTS_USED_LAYER_2
PTS_USED_LAYER_3
PTS_USED_LAYER_4
PTS_USED_LAYER_5
NUMBER_LAYERS
MORPHOLOGY_CRATER_1 CpxCPk
MORPHOLOGY_CRATER_2 Gullies
MORPHOLOGY_CRATER_3 Slump
Deposits
MORPHOLOGY_EJECTA_1 Rd
MORPHOLOGY_EJECTA_2
MORPHOLOGY_EJECTA_3
MORPHOLOGY_EJECTA_COMMENTS
79
DEGRADATION_STATE
CONFIDENCE_IMPACT_CRATER 4
LAYER_1_PERIMETER
LAYER_1_AREA
LAYER_1_LOBATENESS
LAYER_1_EJECTARAD_EQUIV
LAYER_1_EJECTARAD_REL
LAYER_2_PERIMETER
LAYER_2_AREA
LAYER_2_LOBATENESS
LAYER_2_EJECTARAD_EQUIV
LAYER_2_EJECTARAD_REL
LAYER_3_PERIMETER
LAYER_3_AREA
LAYER_3_LOBATENESS
LAYER_3_EJECTARAD_EQUIV
LAYER_3_EJECTARAD_REL
LAYER_4_PERIMETER
LAYER_4_AREA
LAYER_4_LOBATENESS
LAYER_4_EJECTARAD_EQUIV
LAYER_4_EJECTARAD_REL
LAYER_5_PERIMETER
LAYER_5_AREA
LAYER_5_LOBATENESS
LAYER_5_EJECTARAD_EQUIV
LAYER_5_EJECTARAD_REL
CRATER_NAME
Shape Point
Barlow Crater Database
FID 4283
Shape Point
SUBQUAD 05SW
ID 400
LATITUDE 35.52
LONGITUDE 6.59
DIAMETER 41.2
TERRAIN HCr
80
TYPE #c
EJECTA_MOR No
INTERIOR_M No
PIT_DIAMET 0
MIN_DIAMET 0
ANGLE 999
COMMENTS none
Table 5: Robbins and Barlow Crater
Databases information collected on
Siloe Patera.
Figure 47: Crater databases used are displayed and each circle shows the area determined to
be the center of the crater according to Robbins and Barlow. Robbins shown in yellow and
Barlow in blue followed by the object number for each point in their database.
81
evidence of typical impact crater characteristics present. Information filled out and
provided by Robbins and Barlow were downloaded from USGS PIGWAD website and
imported into ArcGIS to analyze and display data along a Mars_2000 geographic
coordinate system.
Robbins Crater Database was generated using THEMIS Daytime IR on board Mars
Odyssey (Robbins and Hynek, 2012). All visible impact craters larger than 1 km in
diameter were identified, analyzed and applied to this database. Robbins has many
fields that are left null but fields on impact crater and ejecta morphology have been
labeled as having impact crater characteristics. For object 28414, equivalent to Crater A
of Siloe Patera, Robbins indicates that it is a complex unclassified chaotic (CpxUnc)
impact crater where object 28500, Crater B of Siloe Patera, is classified as complex
central peaks (CpxCPk), referring to the feature we have dubbed as a spire. Object
28414 (Crater A)is classified as having two ejecta layers: 1. Multiple Layer Ejecta
Rampart Sinuous (MLERS), 2. Hummocky broad lobes (HuBL). Object 28500 (Crater B)
has one ejecta morphology labeled as radial ejecta. On the other hand, Barlow does not
indicate any ejecta or internal morphology from her Version 1 Crater Database although
she does indicate the terrain as being heavily cratered uplands (HCr) with no apparent
embayment by intercrater plains. Barlow has produced a second, more detailed, version
of the Crater Database but this version has not been released yet.
Flow features propagate from Siloe Patera?s rim to the north, southeast, and
southwest and could be interpreted as ejecta or lava flows (Figure 48). Since there are
two collapse/impact events, it is not possible to determine with which event the flow
82
features are associated. Much can be interpreted by the distance of each flow from the
rim as seen in Table 6. The north flow and southeastern flow could easily be interpreted
to represent ejecta layers as their distance falls within acceptable ranges from the
Fig
ure 4
8: Fl
ow f
eatu
res
hi
gh
lig
hted in
tan
an
d descr
ibed
in
Tab
le
2. Ea
ch
flo
w h
as its
own
ch
aract
eri
zes
ran
gin
g in len
gth, w
idth an
d general
app
earan
ce. B
ox
to
the t
op
lef
t o
f th
e im
ag
e sho
ws a
zo
om
ed
in CTX
im
ag
e o
f an
ov
erlap
in
the
so
uthw
est
ern
flo
w.
83
impact site. However, the southwestern flow extends almost triple the radius of Crater
A and Crater B of Siloe Patera
N Flow SE Flow SW Flow
Crater
A
28 38 61
Crater
B
36 30 56
Figure 48 flows are highlighted in tan shown on a THEMIS Day Infrared image.
Each flow feature appears to have its own characteristics and does not resemble each
other. The southwestern flow is very broad and expands outwards from the rim and
spans a great distance from Siloe Patera. This flow does not appear to have high
amounts of degradation or weathering and even displays an overlap (Figure 48)
indicating multiple flow events. The northern flow is highly degraded and flat and does
not extend more than one radius of Siloe Patera (17 km). The southeastern flow extends
a little over one radius of Siloe Patera but displays a more elongated and thin
appearance unlike the others that are flat and broad.
CHAPTER 5
Because features observed at Siloe Patera could be interpreted as both impact
and volcanic in origin, regions of interest are discussed below in the context of both.
Siloe Patera contains features interpreted as ring faults that have been partially covered
by dust and other debris. These faults exist in a region ~ 18 km long by 8 km wide and
are located on the northeast portion of Siloe Patera?s bench (Figure 3B). From the south,
faults along the bench begin to curve slightly until they tract north ? northwest at the
Table 6: Distance from Crater A and Crater B center
(km) to edge of flow features.
84
most northern portion of Siloe Patera. This is most easily seenseen in Figure 3B. If
interpreted as ring fractures, they would be associated with the beginning stages of
subsidence for Crater B before it collapsed. These features are shown to be bright in
THMIS Night IR and in the color stretch map which could be interpreted as exposed fault
scarps.
Siloe Patera contains features interpreted as ring faults that have been partially
covered by dust and other debris. These faults exist in a region ~ 18 km long by 8 km
wide and are located on the northeast portion of Siloe Patera?s bench (Figure 3B). From
the south, faults along the bench begin to curve slightly until they tract north ?
northwest at the most northern portion of Siloe Patera. This is most easily seen in Figure
3B. If interpreted as ring fractures, they would be associated with the beginning stages
of subsidence for Crater B before it collapsed. These features are shown to be bright in
THMIS Night IR and in the color stretch map which could be interpreted as exposed fault
scarps.
The spire located on the floor of Crater B is an odd structure that could be
interpreted in a number of ways. It can be divided into two parts (Figure 49) and
appears to have been covered by sediments. The region of the spire located directly in
the middle of Crater B is ~ 3 km long by 2 km wide and raises to a height of 80 meters
from the floor of Crater B. The remaining portion of the sprire extends from the middle
portion of the spire and connects with the wall of Siloe Patera on its southeastern side.
It is ~ 11 km long by 5 km wide and thickens as it approaches the wall, with its highest
point being 257 m from the Crater B?s floor. In the context of an impact origin, the spire
85
would be interpreted as a central peak due to its isolated location in the middle of the
crater floor. However, because the spire extends out from the southeastern wall (Figure
3 and Figure 13), it is more likely that it is the result of a mass wasting event or an
extension of the underlying rock defining the spire morphology. This suggests that it is
less likely to be the result of a mass wasting event and is more likely part of the spire
itself.
When observing the spire in THEMIS Night IR (Figure 22 and 38) and in the
stretch color map (Figure 31), we can see that the spire and surrounding area is among
the brightest regions on the map. This suggests that the spire and much of the floor of
Crater B is solid rock, uncharacteristic of impact craters, which have flat floors created
by fill-in of dust and sand. When looking at the THEMIS Night IR image (Figure 30), it is
apparent that impact craters floors around Siloe Patera are filled with dust and sand
Figure 49: Close up view of the spire located on the floor of Crater B in Siloe Patera.
Figure 49: Close up view of the spire located on the floor of Crater B in
Siloe Patera.
86
because they appear dark, suggesting low thermal inertia. In impact craters that have
central peaks, those peaks are generally bright in THEMIS Night IR, while the rest of the
surrounding crater floor is dark. This can be observed in two sets of conjoined impact
craters to the left and right of Siloe Patera (Figure 30).
Therefore, it is more likely that the spire has a volcanic or mass wasting origin
than an impact crater origin. In the context of a volcanic origin, the presence of the spire
is best explained as resurgent dome. If the spire is a resurgent dome we would expect to
see uplift to angles of more than 65? and distension faults. Due to dusty conditions, it is
possible that visitation, by human or rover, may be required to fully determine its origin.
Post-volcanic doming is another region of interest that was revealed during 3D
modeling of Siloe Patera as part of this research. Five to six potential post-volcanic
doming features are labeled in Figures 3B and 33. Elevations of peaks were determined
by manipulating classes and colors of an interpolated IDW map that was previously
constructed. Forty classes were applied and a range of elevations were given for each
class. The highest elevation in each given range was chosen to represent each feature?s
height, measured from the lowest elevation of Siloe Patera?s Crater B floor, at ?~3931m
below the mart. Dome heights from the floor of Crater B are as follows: (1) 957 m;(2)
1,116 m;(3) 841 m;(4) 1,240 m; and (5) 1,010 m. Dimensions of each dome range in size,
but are approximated: (1) 3.3 km x 1.7 km; (2) 4.3 km x 4.1 km ;(3) 2.1 km x 1.2 km;(4)
5.6 km x 4.3 km; and (5) 2.8 km x 1.4 km.
These domes are bright in THEMIS Night IR (Figure 30) and appear green in the
color stretch map (Figure 31). They could be explained as resurgent domes built in
87
Crater A on lake sediments and caldera fill from the first collapse event prior to the
younger caldera forming events associated with Crater B. In this case, these domes may
have developed along ring fractures associated with the development of Crater B,
where magma rising through the ring faults create a semi-circle of volcanic domes
(Figure 3B).
Siloe Patera has a number of sporadic flow features around the rim which vary in
shape, length, and other characteristics. In the THEMIS Night IR image (Figure 30) and
the color stretch map (Figure 31), the head of the flows appear as solid rock with the
rest covered in dust. The fact that these flows are not found all around Siloe Patera, but
only in specific areas is not supportive of the impact crater origin hypothesis because
ejecta from impact craters generally covers the significant regions around impact crater.
Definable ejecta accompanies the majority of impact craters and considering that an
impact origin for Siloe Patera requires two impact events, ejecta should be abundant in
the region surrounding Siloe Patera. These flows of limited spatial extent also vary
significantly in distance and appearance, which suggests they were not created by the
same process. The southwest flow (Table 6) extends more than double the distance of
the southeast and north flows and also maintains a broad, flat appearance. The
southeast and north flows extend away from Siloe Patera for distances typical of impact
craters and have a more fluid and finger-like appearance. It is possible that the flows are
ejecta deposits, with the southeast and north flows formed during a single impact
(Crater A) event, with the southwest flow formed during a subsequent impact event
(Crater B). This however, does not explain the notable lack of an extensive ejecta
88
blanket,. The lack of a clear and definable, extensive ejecta blanket make an impact
crater origin for Siloe Patera a less than satisfactory model of formation.
Evidence of subsidence to the south of Crater B is a unique feature not typical of
impact craters, but may be suggestive of volcanism. This region of subsidence (Figure
50) has a shallow, irregular V-shape with an area of ~2,015 km2and dimensions of ~36 x
18 km. Its lowest point is ~ -2015 m below Martian sea level which is ~ 1030 meters
above Siloe Patera?s lowest point and 700 meters below the average surrounding
surface. Most of this sag feature is suggestive of a material with high thermal inertia on
THEMIS Night IR and on the color stretch map, typical of solid rock. This subsidence
feature could represent the very beginning stages of caldera collapse, prior to the
development of ring fractures which have propagated to the surface, associated with a
magma chamber at depth which has migrated away from the area, or which cooled.
Figure 45 models one possible scenario for the succession and history of Siloe Patera?s
subsidence feature.
The depth to diameter ratio (d/D) of an impact crater is fairly standard and most
impact craters fall within the 0.03 to 0.17 interval. Due to degradation, modification,
and size, impact crater d/D ratios can vary greatly enough to encompass known volcanic
calderas like Ascraeus Mons. Although Siloe Patera and Ascraeus Mons fall within the
crater d/D ratio range, Siloe Patera and Ascraeus Mons fall very close to the edge of the
89
range at 0.03 (Gravin and Frawley, 1998). Siloe Patera has a d/D of 0.049, which is
almost identical to Ascraeus Mons at a d/D of 0.048. According to a graph of rim-floor
depth vs. impact crater diameter in a recent paper (Michalski and Bleacher, 2013), Siloe
Patera?s Crater A and Crater B contain a higher depth to diameter ratio than should be
expected due to its Noachian/Hesperian age (Figure 51). The degradation of impact
Fig
ure
50
: P
os
sib
le
mo
ve
ment hist
ory
of
a m
ag
ma
ch
am
ber
belo
w
Silo
e P
ate
ra.
90
craters is divided into 4 classes with class 1 being highly degraded and class 4 pristine.
Crater A falls along the class 2 trendline, suggesting it is older and has had time to
undergo erosion in an impact crater model of formation. Crater B is suggestive of
pristine and fresh impact craters. However, if Siloe Patera is considered a Hesperian
impact crater, it should plot as a Class 1. However, both Crater A and Crater B have d/D
ratios similar to young impact craters and should, therefore, have degradation features
typical of class, 2, 3, and 3 impacts. This conflict between the apparent age of the crater
based on d/D ratios and that based on degradation features suggests an impact origin
for Siloe Patera is problematic.
Garvin (2000) developed a relation between impact crater depth values and
crater diameter values for non-polar complex impact craters. The impact craters that
were examined for non-polar impact craters were best represented by a power law of
the form d = 0.19D0.55, where d and D are measured in kilometers. Impact craters that
have a diameter of 20 km should generate a depth of ~1km, a 40 km diameter crater
should produce a depth of 1.6 km and an 80 km diameter crater should create a depth
up to 2.5 km. If Siloe Patera is an impact crater, we would expect to see a depth of ~1.3
km as a fresh impact crater. Siloe Patera exceeds this calculation by ~0.5, but has had
significant time to fill in resulting in a flat floor. An impact origin for Siloe Patera would
suggest it is a relatively pristine crater.
If Craters A and B were relatively recent impacts, they would look similar to an
unnamed impact crater, roughly the size of Crater B, located to the northeast (Figure 52)
of Siloe Patera. This impact crater has not had significant amounts of degradation and
91
has a uniform ejecta blanket that completely surrounds it. We would expect to see the
same impact characteristics in Siloe Patera as this crater if it were the result of a
relatively young impact event as suggested by the d/D ratio. There is, however no
evidence for an extensive ejecta blanket, a central peak isolated from the crater rim and
the depth of Siloe Patera far surpasses the depth of the unnamed crater.
Figure 51: Siloe Patera Crater A falls within the more degraded impact craters and Crater B
is considered to be new (Michalski and Bleacher, 2013).
92
Streams and channels can also aid in distinguishing between an impact crater
and volcanic origin for Siloe Patera. Due to the raised rims around an impact crater;
water will flow down the rim, into the ejecta blanket and create a path. This path begins
Fig
ure
52
: A
co
mp
ari
so
n o
f Si
loe
Pat
era t
o a
sim
ilar
siz
ed crat
er
to t
he
no
rth eas
t.
93
at the top of the rim and will flow downhill toward the ejecta blanket (Figure 40). Siloe
Patera?s stream channels to the south, flow into the large area of subsidence, while
others flow into Crater B through a large channel on the southern flank of Siloe Patera
(Figure 50).Channels also flow from the bench of Crater A to the floor of Crater B,
reinforcing the lack of a raised rim around Crater B. Figure 3A highlights the channels
around Siloe Patera and shows their flow direction. For comparison, the small impact
crater along the western rim of Siloe Patera has a few channels that flow downhill from
the rim, typical of an impact crater?s raised rim.
Comparison of Crater F to Siloe Patera was chosen due to the fact that it
contains a secondary crater completely encompassed in a larger impact crater with one
wall being shared by both (Figure 53). Like Crater A and B of Siloe Patera, the shared
wall has arcuate scarps. However, the arcuate scarp was only present in a small area
before the hummocky and chaotic terrane of an impact crater took over. The fill of the
secondary impact crater is rough, modified and filled with slump blocks mass wasting
from the rim. These mass wasting events stay along the walls of the impact crater and
do not reach the center of the crater floor. A central peak is easily defined in the center
of the secondary impact crater. Moving from the secondary impact crater to the primary
impact crater, a uniform and prominent raised rim rests along the boundary between
both impact craters. Siloe Patera?s Crater B does not have a raised rim that rest on top
of Crater A, unlike the unnamed Crater F. The floor of the primary crater is relatively flat
with no features other than possible minor ejecta and a smaller impact crater. The rim
of the primary impact crater is similar to the rim of the secondary impact crater with
94
Figure 53: Comparison of Siloe Patera (above) to Crater F (below). Many
differences can be seen here as described in text.
95
chaotic terrain, blocks separating from the rim, and very uneven walls. It appears that
the only characteristics Siloe Patera and Crater F share are a secondary crater fully
contained within a primary crater.
The classification map and stretched color map were used to generate a geologic
map from their results. The classification map was divided into three types of materials:
(1) dust; (2) rock; (3) fill-in. The walls of Crater A and Crater B and the spire
unsurprisingly showed up as rock in the classification map, stretched color map, and
THEMIS Night IR. The majority of the floor of Crater B indicated an abundance of rock
uncharacteristic of impact craters, which typically have floors covered in sediment fill.
The majority of the surrounding area of Siloe Patera was classified as dust and only a
small portion inside Crater A and Carter B was identified as fill-in. Using the classification
and stretched color maps, a geologic map was rendered in ArcGIS based on nine
different features. The walls of Crater A, Crater B, parts of the bench and much of the
floor of Crater B were defined as solid rock with the spire in the middle of Crater B
identified as a possible volcanic feature. Five domes, labeled as post-volcanic domes,
rest on the bench of Crater A as well as a region of dust to the far right of the bench.
Two units of sediment fill-in were identified and divided by what appeared to be two
major episodes of deposition. The yellow unit likely formed during an early period of
deposition that allowed it to consolidate where the green unit appears to have been the
last depositional event of Siloe Patera?s history as it fills in low areas. Flows are
displayed as light green and appear to have formed during different periods due to
degradation, appearance and the length of each flow.
96
If Siloe Patera is a volcano, its features should be explained by volcanic processes
and stages of caldera development. Smith and Bailey (1968), and Acecolla (2006),
provide a framework for to describing the different features of Siloe Patera according to
their stages of a Resurgent-Caldron Cycle and Caldera Collapse (Figures 13 and 14) with
stages of the former model reported using Roman numerals (e.g. ?Stage III?) while
stages of the latter are reported using Arabic numerals (e.g. ?Stage 3?).
First, doming and sagging in the area would form due to rising magma that
would also induce the production of ring fractures (Stage I; Stage 1). Ash flows and lava
flows would then cover the area and the potential to erase signs of ring fractures (Stage
II). Figures 10A and 10B show the varying length and appearance of lava flows in the
Kilauea caldera. The flow features seen to the north, southeast, and southwest of Siloe
Patera could be the result of similar lava flows. Collapse of Crater A would then be
followed by another stage of ring faulting from resurgence of the magma chamber
(Stage III, IV, and V; Stage 2, 3, and 4). The possible ring faults highlighted in Figures 29
and 30 could be the result of resurgence on the floor of Crater A, prior to formation of
Crater B. Resurgence would continue and fractures would become more prominent,
magma could rise from the cracks of the ring faults and may have resulted in
development of the proposed volcanic domes, along the bench in Siloe Patera (Stage V
and VI). Shortly after these domes formed, ring faults would form and became the edge
of Crater B in its future collapsed (Stage III; Stage 2). The cycle would then started over
with Crater B and the spire on the floor of Crater B could be resurgence from the
magma chamber (Stage IV). It appears that this was the last stage for Crater B before
97
the magma chamber moved to the south where the sag feature is located. The history
of Siloe Patera proposed here is based on our current understanding of the geology of
the region and will be adjusted as more data becomes available.
CHAPTER 6
By utilizing 1.) topographic profile comparisons to known calderas and impact
craters; 2.) 3D analyses of regions of interest circum Siloe Patera and the surrounding
terrain; 3.) d/D comparisons to fresh and eroded Martian impact sites; and 4.)
classification of thermal, visual, and topographic data that aid in the generation of a
photogeologic map, suggest that a volcanic origin for Siloe Patera is more plausible than
impact origin interpretations
Striking similarities between Siloe Patera and the caldera of Ascraeus Mons
points toward a volcanic-based origin. Topographic profiles of Siloe Patera and Ascraeus
Mons are almost identical and are much more complex than their impact crater
counterparts. Topographic profiles of Siloe Patera show a very inconsistent terrain that
is difficult to explain using the double impact crater hypothesis; but is very consistent
with a volcanic origin hypothesis. Siloe Patera exceeds the d/D slope angle of Ascraeus
Mons by 2 degrees with Crater A, B, C, and F d/D slope angles much lower. Ring faulting
on the bench of Crater A is identified as rock with dust cover and not a sedimentary
feature, while 3D and high resolution analyses revealed a possible spire of volcanic
material with high thermal inertia, as well as steep, arcuate scarps with an angle of 69?.
These walls resemble Ascraeus Mons (67?) and not impact craters, which have shallower
slope angles (64? - 42?). Volcanic-doming along the boundary of Crater A and Crater B
98
show up as rock with probable dust or ash deposited around the area. A broad region of
subsidence to the south of Siloe Patera could have been the last volcanic event to
develop along a southeastward migrating magma chamber. Flow features around Siloe
Patera are more likely to be lava flows, than ejecta, due to the isolation and appearance
variation of each flow. These flows appear to be dust covered in many areas except at
the head of the flow, which shows high thermal inertia. Channels around Siloe Patera do
not flow in the typical direction observed in impact crater ejecta blankets, instead
flowing into the area of subsidence to the south. Ring faulting, the spire, volcanic-
doming, subsidence, and flow features show areas of high thermal inertia in THEMIS
Night IR, with the stretch color map and the classification map suggesting hard rock and
not sedimentary deposits. Also, the lack of obvious impact crater characteristics; such as
ejecta, a central peak, and a raised rim, implies a non-impact origin. These features can
easily be explained by volcanic processes; but are more difficult to explain through
impact processes.
It is difficult to make an unequivocal conclusion based on the currently available
data set. Dust cover presents a challenge with this research and makes mineralogic
analysis via spectroscopy virtually impossible. Evidence presented here is based
primarily on geomorphic features associated with Siloe Patera as compared to those of
Martian impact craters and volcanic calderas. Although the origin of Siloe Pater could be
explained by multiple impact craters, multiple caldera collapse events, or a combination
of the two, current evidence strongly suggests a volcanic origin and should be
considered a Martian supervolcano.
99
References
Acocella, V., Caldera types: How end?members relate to evolutionary stages of collapse,
Geophysical research letters 33, no. 18 (2006).
Aline, G., N. Mangol , J. Bibring, Y. Langevin, B. Gondet, F. Poulet, and G. Bonello,
Sulfates in Martian layered terrains: the OMEGA/Mars Express view,Science 307,
no. 5715: 1587-1591, 2005.
Arvidson, R. E. Morphologic classification of Martian craters and some
implications,Icarus 22, no. 3: 264-271,1974.
Barlow, N. G., J. M.Boyce,F. M. Costard,R. A. Craddock, J. B. Garvin,S. E.Sakimoto,R. O.
Kuzmin,D. J.Roddy, and L. A. Soderblom, Standardizing the nomenclature of
Martian impact crater ejecta morphologies,Journal of Geophysical Research:
Planets (1991?2012) 105, no. E11: 26733-26738, 2000.
Bibring, Jean-Pierre, et al. "Global mineralogical and aqueous Mars history derived from
OMEGA/Mars Express data." Science 312.5772 (2006): 400-404.
Blaschke, T., Object based image analysis for remote sensing,ISPRS Journal of
Photogrammetry and Remote Sensing 65, no. 1: 2-16, 2010.
Carr, M. H., Mars Global Surveyor observations of Martian fretted terrain,Journal of
Geophysical Research: Planets (1991?2012) 106, no. E10: 23571-23593, 2001.
Cattermole, P., Planetary Volcanism: A study of volcanic activity in the Solar
System,West Sussex, England. Ellis Horwood Limited, 1989.
100
Christensen, P. R., B. M. Jakosky,H. H. Kieffer,M. C. Malin, H. Y. McSween,K. Jr.Nealson, ,
and G. L. Mehall, The thermal emission imaging system (THEMIS) for the Mars
2001 Odyssey mission,Space Science Reviews 110, no. 1-2: 85-130, 2004.
Cole, J. W., D. M. Milner, and K. D. Spinks, Calderas and caldera structures: a
review,Earth-Science Reviews 69, no. 1: 1-26, 2005.
Crumpler, L. S., J. W.Head, and J. C. Aubele, Calderas on Mars: Characteristics, structure,
and associated flank deformation,Geological Society, London, Special
Publications 110, no. 1: 307-348, 1996.
Dohm, J. M., N. G. Barlow,R. C.Anderson, J.Williams, H. Miyamoto,J. C.Ferris, and R.
G.Strom, Possible ancient giant basin and related water enrichment in the Arabia
Terra province, Mars,Icarus 190, no. 1: 74-92, 2007.
Dr?gu?, L., E.Clemens, and S. Thomas, Local variance for multi-scale analysis in
geomorphometry, Geomorphology 130, no. 3: 162-172, 2011.
Fergason, R. L., and P. R. Christensen, Formation and erosion of layered materials:
Geologic and dust cycle history of eastern Arabia Terra, Mars,Journal of
Geophysical Research: Planets (1991?2012) 113, no. E12, 2008.
Frankel, C., Volcanoes of the Solar System, New York, NY. Cambridge University Press,
Gangale T. Geologic Time Scales of the Earth, Moon, and
Mars.
(Accessed 30 September 3013), 2007.
101
Garvin, J. B., and J. J.Frawley.,Geometric properties of Martian impact craters:
Preliminary results from the Mars Orbiter Laser Altimeter,Geophysical Research
Letters 25, no. 24: 4405-4408, 1998.
Grant, J. A., and P. H. Schultz, Degradation of selected terrestrial and Martian impact
craters, Journal of Geophysical Research: Planets (1991?2012) 98, no. E6: 11025-
11042, 1993.
Greeley, R., Planetary Landscapes, Second edition. New York, NY. Chapman & Hall, 1994
Greeley, R., and P. D. Spudis, Volcanism on Mar,Reviews of Geophysics 19, no. 1: 13-41,
1981.
Halevy, I., M.T. Zuber, and D. P. Schrag. A sulfur dioxide climate feedback on early
Mars,Science 318, no. 5858: 1903-1907, 2007.
Hynek, B. M., R. J. Phillips, and R. E. Arvidson, Explosive volcanism in the Tharsis region:
Global evidence in the Martian geologic record,Journal of Geophysical Research:
Planets (1991?2012) 108.E9, 2003.
Irwin, R. P., T. R. Watters,A. D. Howard, and J. R.Zimbelman, Sedimentary resurfacing
and fretted terrain development along the crustal dichotomy boundary, Aeolis
Mensae, Mars,Journal of Geophysical Research: Planets (1991?2012) 109.E9,
2004.
Kerber, L., J. W. Head,J.Madeleine,F.Forget, and L.Wilson, The dispersal of pyroclasts
from ancient explosive volcanoes on Mars: Implications for the friable layered
deposits,Icarus 219, no. 1: 358-381, 2012.
102
Malin, M. C., and K. S. Edgett, Mars global surveyor Mars orbiter camera: Interplanetary
cruise through primary mission,Journal of Geophysical Research: Planets (1991?
2012) 106, no. E10: 23429-23570, 2001.
Mart?nez?Alonso, S., B. M. Jakosky,M. T.Mellon, and N. E. Putzig, A volcanic interpretation
of Gusev Crater surface materials from thermophysical, spectral, and
morphological evidence,Journal of Geophysical Research: Planets (1991?2012)
110, no. E1, 2005.
Meaden, G. J., andJ. M. Kapetsky, Geographical information systems and remote sensing
in inland fisheries and aquaculture,FAO Fisheries Technical Paper. No. 318.
Rome, FAO. 262p, 1991.
McEwen, A. S., E. M. Eliason, J.W.Bergstrom, N. T. Bridges,C. J. Hansen,A. W.Delamere,
and J. A. Grant, Mars reconnaissance orbiter's high resolution imaging science
experiment (HiRISE),Journal of Geophysical Research: Planets (1991?2012) 112,
no. E5, 2007.
McGill, G. E., Crustal history of north central Arabia Terra, Mars,Journal of Geophysical
Research: Planets (1991?2012) 105, no. E3: 6945-6959, 2000.
Milam, K. A., K. R. Stockstill,J. E. Moersch,H. Y. McSween,L. L.Tornabene,A. Ghosh,M. B.
Wyatt, and P. R. Christensen, THEMIS characterization of the MER Gusev crater
landing site,Journal of Geophysical Research: Planets (1991?2012) 108, no.
E12,2003.
Miller, C. F., and D. A. Wark, Supervolcanoes and their explosive
supereruptions,Elements 4, no. 1: 11-15, 2008.
103
Omran, M., A. Salman, and A. P. Engelbrecht, Image classification using particle swarm
optimization,Proceedings of the 4th Asia-Pacific conference on simulated
evolution and learning. Vol. 2002, 2002.
Osinski, G. R. and E.Pierazzo, Impact Cratering: Processes and Products, in Impact
Cratering: Processes and Products, (eds G. R. Osinski and E. Pierazzo), John Wiley
& Sons, Ltd, Chichester, UK. doi: 10.1002/9781118447307.ch1, 2012.
Pike, Richard J. Formation of complex impact craters: Evidence from Mars and other
planets,Icarus 43.1: 1-19, 1980.
Robbins, S. J., and B. M.Hynek, A new global database of Mars impact craters? 1 km: 1.
Database creation, properties, and parameters,Journal of Geophysical Research:
Planets (1991?2012) 117, no. E5, 2012.
Robock, A., and J.Mao, Winter warming from large volcanic eruptions,Geophysical
Research Letters 19, no. 24: 2405-2408, 1992.
Scott, D. H., and M. H. Carr, Geologic map of Mars, Map 1-1083, US Geological Survey,
Reston, Va., 1978.
Scott, D. H., and K. L. Tanaka, Geologic map of the western equatorial region of
Mars,Geological Survey (US), 1986.
Smith, D. E., M. T.Zuber, H. V.Frey, J. B. Garvin,J. W. Head, D. O. Muhleman, and G.
H.Pettengill, Mars Orbiter Laser Altimeter: Experiment summary after the first
year of global mapping of Mars,Journal of Geophysical Research: Planets (1991?
2012) 106, no. E10: 23689-23722, 2001.
104
Smith, E. I., Comparison of the crater morphology-size relationship for Mars, Moon, and
Mercury,Icarus 28, no. 4: 543-550, 1976.
Smith, Robert L., and R. A. Bailey.Resurgentcauldrons, Geological Society of America
Memoirs 116: 613-662, 1968.
Tanaka, K. L. The stratigraphy of Mars,Journal of Geophysical Research: Solid Earth
(1978?2012) 91, no. B13: E139-E158, 1986.
Tian, F., M. W. Claire, J. D. Haqq-Misra,M.Smith, D. C. Crisp,D.Catling, K. Zahnle, and J.
F.Kasting, Photochemical and climate consequences of sulfur outgassing on early
Mars,Earth and Planetary Science Letters 295, no. 3: 412-418, 2010.
Vincent, R. K., Fundamentals of Geological and Environmental Remote Sensing, Upper
Saddle River, NJ. Prentice Hall, 1997.