Copyright by Benjamin Thomas Cardenas 2014

Home , Green Valley (Mars)

The Thesis Committee for Benjamin Thomas Cardenas Certifies that this is the approved version of the following thesis:

Evidence for changes in coastline-controlled base level from

fluvial stratigraphy at Aeolis Dorsa, Mars

APPROVED BY SUPERVISING COMMITTEE:

Supervisor: David Mohrig Co-supervisor: Gary Kocurek

John W. Holt

Evidence for changes in coastline-controlled base level from

fluvial stratigraphy at Aeolis Dorsa, Mars

Benjamin Thomas Cardenas, B.S.

Thesis Presented to the Faculty of the Graduate School of The University of Texas at Austin in Partial Fulfillment of the Requirements for the Degree of

Master of Science in Geological Sciences

The University of Texas at Austin December 2014

Dedication

This thesis is dedicated to my parents, who have encouraged me in all my endeavors.

Acknowledgements

I would like to thank David Mohrig for his detailed feedback, and Gary Kocurek and Jack Holt for the guidance and discussions they provided. Thanks to David Brown for his help with the field aspect of this project. Thanks to David Mohrig, The University of Texas at Austin Graduate School, and the Jackson School of Geosciences for funding. I would also like to thank Nicole for her patience and support.

Evidence for changes in coastline-controlled base level from

fluvial stratigraphy at Aeolis Dorsa, Mars

Benjamin Thomas Cardenas, M.S. Geo. Sci.

The University of Texas at Austin, 2014

Supervisors: David Mohrig and Gary Kocurek

There is evidence that a subset of fluvial deposits at Aeolis Dorsa, a basin on Mars, preserve incised valleys carved and filled during changes in base level, which was likely controlled by water surface elevation of a large lake or sea.

Three low-albedo, channelized corridors, each several tens of kilometers long, contain relict point bars and scooped boundaries at their bases, indicating that the base and lateral extent of each corridor was defined by a migrating, net- erosional river. Above the basal deposits are stacks several tens of meters thick of “inverted sinuous ridges”, which are channel-filling deposits that have been exhumed and topographically inverted. Indicators of avulsions, channel reoccupations, an overall flattening of basal topography, and confinement of inverted sinuous ridges to the dark corridors are evidence of the gradual filling of a valley cut by the basal migrating river. Valley incision and fill are common responses to sea level change on Earth. Aeolis Dorsa is currently open to the

northern lowlands of Mars, where an ocean has been hypothesized to have once

existed, although a large lake could have also controlled base level. Cross-

cutting valleys require at least two episodes of base level fall and rise. The

magnitudes of the base level changes are estimated at about 80 meters, based

on the thickness of the valley-filling stratigraphy. Meander asymmetry is consistent with a southeastern flow direction, and is supported by a set of branching fluvial deposits 40 km to the southeast which, qualitatively, appear to be deltaic in origin.

vii

Table of Contents

List of Figures ...... ix

Chapter 1: Introduction ...... 1

Chapter 2: Background ...... 5

Chapter 3: Data...... 10

Chapter 4: Observations ...... 13

Chapter 5: Analysis ...... 16

Chapter 6: Interpretations ...... 24

Chapter 7: Conclusions ...... 34

Chapter 8: Figures ...... 36

References ...... 53

Vita ...... 59

viii

List of Figures

Figure 1: Regional context for Aeolis Dorsa ...... 36

Figure 2: Study area ...... 37

Figure 3: Corridor geometry ...... 38

Figure 4: Corridor contacts ...... 39

Figure 5: Fluvial stratigraphy ...... 40

Figure 6: Downstream increase in channel-filling deposits ...... 41 Figure 7: Aeolis Serpens ...... 42 Figure 8: Model explaining corridor formation ...... 43 Figure 9: Thermal inertia of corridors ...... 44 Figure 10: Characteristics of Brazos River valley, TX ...... 45 Figure 11: Preserved avulsions, channel re-occupations ...... 46 Figure 12: Meander asymmetry and paleo-flow direction ...... 47 Figure 13: Goosenecks and paleo-flow direction ...... 48 Figure 14: Valley incision depths ...... 49 Figure 15: Sinuosity over time ...... 50 Figure 16: Lateral channel migration ...... 51 Figure 17: Downstream deltaic deposit ...... 52

CHAPTER 1: INTRODUCTION

The existence of an ancient ocean spanning the northern hemisphere of

Mars has long been a point of contention. Observations point both ways.

Evidence for the existence of an ancient ocean includes large post-Noachian

outflow channels indicating large volumes of water once flowed from the

southern highlands into the northern lowlands from a variety of sources (Baker

2009), coastline morphology (Parker et al., 1989; Perron et al., 2007), comparisons of surface roughness between the northern and southern hemisphere at all scales (Head et al., 1999), sedimentary deposits along proposed shorelines (DiAchille and Hynek, 2010), and potential deep water deposits in the northern lowlands (Moscardelli et al., 2012; Moscardelli 2014).

There has been difficulty in reconciling the geochemical evidence with the geomorphic evidence. Had a long-lived ocean existed in the northern lowlands, carbonate formation would have been an expected result of the interaction between that ocean and a CO2-rich atmosphere. No significant exposures of

carbonates have been found at the surface. Instead, the majority of geochemical

evidence indicates subsurface interactions between water and rock (Ehlmann et al., 2011), although some scenarios allow for the existence of an ocean without carbonate formation (Fairén et al., 2004; Baker 2009).

Although three billion years’ worth of erosion and alteration is a challenge to overcome in most studies of the oldest Martian surfaces, erosion has also led 1

to the exhumation of stratigraphy, which has been a boon for some methods.

For example, DiBiase et al., (2013) use exhumed stratigraphy to map stratigraphic contacts and create a strong argument that a sedimentary deposit in the basin Aeolis Dorsa is the remnants of an ancient delta. Deltas have been identified and studied in closed crater basins (e.g., Wood 2006), but filling and draining these basins requires a relatively small volume of water. A delta in a basin open to the northern lowlands may be evidence for an ancient Martian ocean which spanned the northern lowlands (Parker et al., 1989). Landforms spanning hypothesized shorelines of a northern ocean have been interpreted as deltas due to surface morphology and relatively consistent elevations of topset- foreset slope breaks (DiAchille and Hynek, 2010), but on deposits that have

endured more than 3 billion years’ worth of erosion, stratigraphic studies are

valuable in strengthening interpretations made only using geomorphology

(DiBiase et al., 2013).

I offer another piece of stratigraphic evidence for the existence of a large

body of water on Mars during the late Hesperian to early Amazonian. I focus on

a subset of fluvial deposits exposed in a region of southeast Aeolis Dorsa

approximately 1,400 km2 in area and centered at 5.75º S, 153.54º E (Fig. 1).

These deposits make up three low-albedo, channelized corridors that are tens of

kilometers long (Fig. 2) and appear to be connected to the 600 kilometer

paleochannel Aeolis Serpens. Contained within each dark corridor are several

sets of continuous channel-filling deposits, which a high resolution digital

elevation model (DEM) shows are vertically stacked on top of each other.

An examination of the geometry of the dark channelized corridors, the patterns and thickness of fluvial stratigraphy contained within the corridors, and the meander asymmetry of the paleochannels suggests that the corridors are the exhumed remnants of sedimentary deposits filling incised valleys. This

interpretation has a significant implication. On Earth, valley incision and filling

are common fluvial responses to falls and rises in sea or lake level, if the river is

close enough to the shoreline to be affected. At Aeolis Dorsa, base level was

likely defined by the surface elevation of a large body of water towards the east-

southeast, either a large lake or sea, during the late Hesperian to early

Amazonian, and there are signs of at least two such changes in base level.

Observations of the dark channelized corridors at Aeolis Dorsa are used

to answer five specific questions:

1. Do the dark channelized corridors meet the geometric criteria for

incised valleys?

2. Is the fluvial stratigraphy contained within the dark channelized

corridors similar to valley-filling deposits on Earth?

3. If questions 1 and 2 can be answered yes, where was the shoreline?

4. How many times did base level change?

5. By how much did base level change?

Taken together, the answers to these questions lead to the interpretation

that the dark channelized corridors are the remnants of exhumed incised valleys,

which were formed in response to at least two episodes of sea or lake level fall and rise of over 50 meters.

CHAPTER 2: BACKGROUND

Aeolis Dorsa

Aeolis Dorsa is an open basin bounded to the east and west by two lobes of the Medusae Fossae Formation (MFF), and to the south by a cliff separating the high-elevation southern hemisphere and low-elevation northern hemisphere

(the well-known crustal dichotomy) of the planet (Fig. 1). The origin of the MFF has been debated by several researchers, but the morphology of the yardangs ubiquitous throughout the MFF seems to imply that the MFF was deposited by ignimbrites (Mandt et al., 2008) during the Hesperian, with the current surfaces exposed via erosion during the Hesperian through the early Amazonian

(Zimbelman and Scheidt, 2012).

Some of the best preserved fluvial deposits on Mars are located at Aeolis

Dorsa, in MFF deposits. The majority of the deposits have been preserved as inverted sinuous ridges (Burr et al., 2009), and although their origin has also been hypothesized as eskers or lava flows, the morphology of the ridges, the presence of sinuous ridge networks, and the evidence for lateral migration preserved within the deposits supports the conclusion that the vast majority of the inverted sinuous ridges at Aeolis Dorsa are fluvial in origin and have not been capped by lava (Burr et al., 2009; Burr et al., 2010; Lefort et al., 2012; Williams et al., 2013).

Using stratigraphic relationships between the fluvial deposits, different

MFF units, and the Cerberus lavas just north of Aeolis Dorsa, Jacobsen and Burr

(2013) dated the formation times of several fluvial deposits at Aeolis Dorsa to the

Hesperian-Amazonian transition. The youngest feature, a fluvial deposit over

600 kilometers in length named Aeolis Serpens, is dated to the early Amazonian.

Fluvial sedimentation likely lasted anywhere from 1 to 20 million years (Kite et al.,

2013).

Valley incision on Mars

Valleys and valley networks are widespread in the southern highlands of

Mars (Howard et al., 2005), but are rare north of the boundary marking the

crustal dichotomy. In the southern highlands, incised valleys often retain

morphology to varying degrees and only occasionally contain within them any

depositional remnants of channels. When they do contain fluvial deposits, it is a

single basal channel as opposed to an observable stratigraphic succession of

valley-filling deposits.

Ancient ocean

Aeolis Dorsa is open to the northern lowlands, where an ancient Martian ocean has been hypothesized to have existed during the late-Hesperian-

Amazonian on the basis of preserved coastline morphology (Parker et al., 1989;

Head et al., 1999; Perron et al., 2007), surface roughness at all scales (Head et

al., 1999), sedimentary deposits along proposed shorelines (DiAchille and

Hynek, 2010), and potential deep water deposits in the northern lowlands

(Moscardelli et al., 2012; Moscardelli 2014).

Parker et al. (1989) proposed a shoreline, later referred to as the “Arabia shoreline”, on the basis of morphologic features which appeared similar to terrestrial analogs, such as wave-cut cliffs. The Arabia shoreline falls in the

region of Aeolis Dorsa. Higher resolution data studied by Malin and Edgett

(1999) refuted the existence of any coastal landforms and argued against the

shoreline proposed by Parker et al., but the authors worked with a relatively small

dataset (14 images) and without a high resolution topographic dataset. Malin

and Edgett discuss the importance of using topographic data in tandem with

remotely-sensed images in identifying outcrops of ancient coastal deposits, even

on Earth. Head et al., (1999) used MOLA data to measure elevation variation

along the Arabia shoreline, and found elevations to vary by as much as 11

kilometers, with a standard deviation of 1.7 kilometers, which was too much for a

coastline. Perron et al., (2007) argued that deformation resulting from true polar

wander caused by crustal loading at Tharsis can explain the range of elevations

found along the Arabia shoreline. The amount true polar wander would also

bring the northern lowlands into tropical latitudes.

An ancient equipotential surface of water, very similar in elevation to the

Arabia shoreline, was interpreted by DiAchille and Hynek (2010). DiAchille and 7

Hynek made a global inventory of sedimentary deposits they interpreted as

deltas on the basis of slope breaks thought to occur where the delta surface

changes from subaerial to subaqueous. Their inventory included deltas enclosed within crater basins, as well as deltas in basins which are currently open to the northern lowlands. When every delta was considered, the elevations for the

noted slope breaks varied widely. When only the deltas in open basins were

considered, the slope-break elevations had a standard deviation of only 177

meters. This result is presented as evidence for a coastline centered on an

elevation of about -2,540 meters. This coastline would have reached Aeolis

Dorsa. Standing water at Aeolis Dorsa is also suggested by DiBiase et al.,

(2013), who interpreted a set of fluvial deposits 90 kilometers northeast of this

report’s study area as the remnants of a delta. They found that avulsion nodes

within those deposits were consistently at similar elevations, suggesting a base level control on avulsion.

In addition to coastal features, evidence for deep-water deposits in the northern plains has been presented. While the small-scale polygonal terrain in the northern plains has been explained by thermal contraction (e.g., Levy et al.,

2008), kilometer-scale polygons defy the same explanation. Moscardelli et al.,

(2012) compare these Martian features to similar-scale submarine polygons found on the seafloor of Earth’s oceans. Collections of clustered boulders in the

northern lowlands have been explained as boulders carried during submarine mass-transport events (Moscardelli 2014).

CHAPTER 3: DATA

Imagery

Visible imagery from the Context Camera (CTX) and High Resolution

Imaging Science Experiment (HiRISE) are abundant in the studied region of

Aeolis Dorsa. CTX takes a single band image (500 – 700 nanometers) of an

area 30 kilometers wide and up to 160 kilometers long with a spatial resolution of

about 6 meters (Malin et al., 2007). CTX is on the Mars Reconnaissance Orbiter

(MRO) satellite, which was launched in August, 2005 and entered a sun-

synchronous polar orbit around Mars in March, 2006. All CTX imagery was

acquired from the Mars Orbital Data Explorer PDS Geosciences Node at

Washington University in St. Louis (ode.rsl.wustl.edu/mars/).

The High Resolution Imaging Science Experiment (HiRISE) is able to

image in three channels. The red channel (570 – 830 nanometers) is able to

take images 6 kilometers wide. A blue-green channel (<580 nanometers) and a

NIR channel (>790 nanometers) are designed to sense the middle 1.2 kilometers

of the red channel’s 6 kilometer swath. HiRISE trades coverage for spatial

resolution of 25 centimeters at finest and 1.3 meters at coarsest (McEwen et al.,

2007). HiRISE is also a part of MRO’s instrument package. All HiRISE imagery

was downloaded at The University of Arizona’s PDS node (hirise-

pds.lpl.arizona.edu/PDS/).

CTX coverage of the study area is complete. This allowed for the creation of a single continuous mosaic using the USGS’s Integrated Software for Imagers and Spectrometers (ISIS). This mosaic was used as a basemap and imported into ArcGIS. Twelve HiRISE images of the study area are available, and were used wherever available.

In addition to visible imagery and DEMs, nighttime thermal infrared imagery from the Thermal Emission Imaging System (THEMIS) was also used in mapping. THEMIS is a multispectral sensor that measures emitted radiance in nine different wavelengths ranging from 6.8 – 14.9 micrometers at a spatial resolution of 100 meters per pixel, and five visible bands with a pixel size of 18 meters (Christensen et al., 2004). THEMIS had imaged Aeolis Dorsa in the thermal infrared during both day and night, allowing for comparisons between the two. THEMIS is onboard Mars Odyssey, which entered orbit around Mars in

October 2001. The THEMIS mosaic used in this study was assembled from data on Washington University’s PDS node.

Digital elevation models

One of the primary mission goals of HiRISE was to gather stereo pairs

(McEwen et al., 2007), and the majority of HiRISE images in the study area are paired. Stereo pairs were used to create DEMs with a spatial resolution of approximately 1 square meter and sub-meter vertical resolution (Kim and Muller,

2009). I completed this processing using the Ames Stereo Pipeline (ASP), an add-on to ISIS which can create a DEM from two stereo images of the same area. CTX stereo pairs were processed in a similar fashion and produced coarser but more extensive DEMs.

The Mars Orbiting Laser Altimeter (MOLA) provides a coarser, regional look at topography in the study area. MOLA measured the topography of the

Martian surface with a 1,064 nanometer laser by measuring the time it takes for the laser to be sent from the orbiter, reflect off the surface, and return to the orbiter. Resulting datasets have a spatial resolution as fine as 100 square meters, depending on the number of incident measurements, and a vertical accuracy of approximately 1 meter (Smith et al., 2001). The MOLA basemap for the study area has a spatial resolution slightly better than 500 square meters.

The DEMs created with stereo pairs are calibrated to specific MOLA points as part of the processing in ASP. All of the images and DEMs were imported into a

Geographic Information System (GIS) to facilitate mapping and measurements.

CHAPTER 4: OBSERVATIONS

Geometry of the dark corridors

The dark corridors are low-albedo, elongate surface features which are continuous for several tens of kilometers (Fig. 2). The corridors branch into three primary paths, each containing its own set of inverted sinuous ridges. The boundaries of the corridors consist of a series of scooped-shape features (Fig.

3). Deposits within the red corridor are truncated at the contact between the red and green corridors. No truncation is observed at the contact between the green and purple corridors (Fig. 4)

Fluvial deposits within the dark corridors

There are two distinct types of preserved fluvial deposits contained within each channelized corridor. The first type are thin inverted sinuous ridges (Burr et al., 2009). All of these inverted sinuous ridges are confined inside of the dark channelized corridors (Fig. 3), and single ridges can be traced continuously for tens of kilometers (Fig. 2). The sinuous ridges have varying heights, ranging from a few meters to more than ten meters in thickness. Although they are sinuous, their exposed internal stratification shows no evidence of significant lateral migration of channel bends (i.e., lateral movement exceeding 50 % of the preserved channel width).

The second type of fluvial deposit at the study area are sets curved strata

that appear linked to the lateral migration of channel bends (Fig. 5). These

curved strata, exposed on sub-horizontal erosional surfaces, are interpreted as

recording point-bar growth and motion, due to their similarity with such deposits on Earth (e.g., Fig. 1 from Nicoll and Hicken, 2010). Point bars form on the inner

banks of river bends, and grow into the channel towards the outer banks.

Subsequent erosion of the outer bank acts to maintain an approximately constant

channel width over time. Repetition of point-bar accretion and outer-bank erosion

process leads to channel bend migration in the direction of point bar growth.

Stacking patterns

HiRISE DEMs show that a similar vertical succession of fluvial deposits is preserved in each valley (Fig. 3, Fig. 5). The basal deposit exposed at the floor of each valley is composed of laterally amalgamated point-bar deposits. Stacked on top of this basal unit are 1 – 3 channel-filling deposits. These channel-filling deposits are more resistant to weathering than surrounding overbank deposits and are now exposed as the inverted sinuous ridges (Fig. 2). The number of sinuous channels defined by persistent channel-filling deposits increases towards the southeast in two of the three corridors. The red valley contains 2 inverted channels along the majority of its length, but increases to 3 stacked channels near the southeastern-most extent of the valley (Fig. 6). The green valley contains 2 inverted channels along the majority of its length (Fig. 2). The purple 14

valley contains 1 inverted channel deposit above the basal deposit for the

majority of its length, but an additional 2 channel deposits are exposed at the

southeastern end of the valley fill (Fig. 6).

Just northwest of the study area is a singular sinuous ridge that using CTX

imagery has been mapped as part of Aeolis Serpens (Fig. 1). Aeolis Serpens is

interpreted as a sinuous ridge of river-channel filling deposits (Williams et al.,

2013), which is continuous for over 600 kilometers. It has not previously been

mapped this far south, but CTX imagery clearly shows that Aeolis Serpens extends to the study area. Deposits composing Aeolis Serpens appear to make up the red channel contained within the green valley (Fig. 7).

CHAPTER 5: ANALYSIS

Recognizing incised valleys in the rock record

An incised-valley system is composed of the eroded valley surface and any sediment filling the valley. Boyd et al. (2006) present a set of criteria for

identifying an incised-valley system in the rock record that highlights four

particular elements: (1) the valley must truncate existing stratigraphy; (2) the

valley must be an erosional topographic feature; (3) valley fill deposits must

onlap the valley walls; and (4) the valley and its filling deposits must be regional

in extent.

If the dark corridors at Aeolis Dorsa are the erosional remnants of incised-

valley systems, we should be able to show that the corridors meet these

necessary conditions. A model placing the exhumed stratigraphy within the

incised-valley framework is illustrated in Fig. 8. Valley incision occurs during a

base level fall and lowstand and valley filling occurs during a base level rise and

highstand. Later erosion of the incised-valley succession generated an inverted

topography where the coarse channel-filling deposits are preferentially preserved

while the related floodplain deposits and the substrate into which the valley was

cut are preferentially eroded. The unusual presentation of these valley-filling

deposits requires that the criteria for identifying incised valleys in the rock record

be slightly revised. The reworked criteria are as follows:

1. Incised valley systems must be continuous over a regional extent.

2. Incised valleys must be erosional features, with elements of vertical

incision and lateral erosion greater than the maximum depth and width of

a single channel.

3. Valley walls must constrain the transport and deposition of sediment

associated with the valley-filling stage.

The dark corridors are of regional extent

Establishing the regional extent of ancient incised valleys using outcrops on

Earth is often difficult because tectonic deformation of strata following valley

formation and filling make regional correlations challenging (Aitken and Flint,

1995; Boyd et al., 2006). The lack of Martian tectonics and significant structural

deformation of outcrops at Aeolis Dorsa make the confirmation of this criterion

relatively simple and unambiguous. Demonstrating the regional extent of the

deposits at Aeolis Dorsa requires that we can define these features as unique

from the surrounding terrain, and then show that these features are continuous

over long distances.

In CTX and HiRISE visible imagery, the corridors can be identified by their

albedo, which is lower than that of the surrounding area (Fig. 2, Fig. 3). The

albedo of the dark corridors was the initial characteristic that the corridors were

mapped by. HiRISE images show textural differences between the corridors and

the surrounding landscape. Wind scour has created a very rough surface on the

terrain surrounding the corridors. The rock within the corridors is smooth by

comparison (Fig. 3). Mapping the corridors via surface roughness gives the

same boundaries as mapping by albedo. The corridors are also relatively bright in nighttime thermal inertia imagery taken by THEMIS (Fig. 9), suggesting a

greater degree of induration (Christensen et al., 2004), most likely cementation

(Pain et al., 2007; Burr et al., 2010), associated with the deposits making up the corridors. Using all of these characteristics, corridors have been mapped for over

40 kilometers in the transport direction (Fig. 2). Several of the channel-filling

deposits defining individual paleochannels within the corridors can be traced over

distances just as long.

The dark corridors are erosional features

The cross-sectional shape of an incised valley has not been preserved here,

but the vertical incision and lateral erosion that defines valley formation can be

inferred from the preserved stratigraphy (Fig. 8). Having only orbital imagery, the

most convincing evidence for the past occurrence of valley topography is the

observation that all of the channel-filling deposits are confined within the

boundaries of the corridors (Fig. 2). This observation holds true for at least 40

kilometers of stacked fluvial stratigraphy spread over multiple channel deposits,

indicating that original valley incision exceeded multiple channel depths.

Lateral erosion of the corridors can be inferred by the shape of the corridor boundaries. These boundaries are composed of multiple scoop-shaped bites

(Fig. 3) that preserve the planforms of the erosional outer banks of laterally migrating channel bends. On Earth, these characteristic scoop-shaped valley walls are observed on modern valleys (Fig. 10), as well as ancient valley successions preserved in the subsurface (Armstrong, 2012). Laterally amalgamated point-bar deposits indicating significant migration of a basal, net-

erosional river channel (Fig. 5) were likely lain down during the lateral cutting of

the valley that produced the scoop-shaped valley walls.

Several generations of channels were bounded by valley walls

Any topography associated with valley walls at Aeolis Dorsa has been

removed via subsequent erosion of the Martian surface. Without this topography

the former presence of valleys must be inferred using other observations. There

are several locations at Aeolis Dorsa where channel deposits run parallel to the

corridor boundaries for significant distances (Fig. 11), suggesting that a

paleochannel came into contact with, and was redirected along, a valley wall. As

an example, channels often abut valley walls of the modern Brazos River valley

(Fig. 10). Presently, the two active channels in the Brazos River valley are

situated at about 2 meters and 4 meters lower than the valley rim (Fig. 10).

Cores and seismic imagery show that the valley was carved to a depth of about

18 meters near this location (Sylvia and Galloway, 2006). Deposition following

incision has nearly filled in nearly 80% this valley’s original relief.

Determining paleoflow direction

Determining the flow direction of the paleochannels at Aeolis Dorsa has

proven challenging. There is topographic evidence that Aeolis Dorsa has been

moderately deformed sometime after water flowed through the region (Lefort et

al., 2012), so flow direction cannot be determined by measuring slope directions

for the channel-filling deposits. It has generally been assumed that flow at Aeolis

Dorsa was towards the north, but DiBiase et al. (2013) measured the dips of

contacts between stacked fluvial deposits and suggested that paleoflow was

towards the east. Here, an independent technique for determining the paleoflow

direction of these ancient fluvial deposits is introduced. The results of this

technique agree with the regional paleoflow direction suggested by DiBiase et al.

(2013).

Carson and Lapointe (1983) inspected several active rivers on Earth and discovered an inherent asymmetry in the shape of meandering river bends. They

define a river bend traverse as the river path between two points of maximum

curvature (Fig. 12). Each traverse contains one inflection point, which is most

often closer to the downstream point of maximum curvature. This is termed a

delayed inflection point. The greater the delay, the higher the degree of

downstream meander asymmetry for that particular traverse. Inflections may

also be premature, closer to the upstream point of maximum curvature, but

premature inflections were shown to be much less abundant than delayed

inflections (Fig. 12). If the delay is minor or non-existent, the traverse is called

symmetric. The asymmetry index, z, is a quantified measurement of this asymmetry, and is defined by Carson and Lapointe as

= 100 . (1) + 𝑢𝑢 𝑧𝑧 ∗ 𝑢𝑢 𝑑𝑑

The values u and d refer to the upstream concave and downstream concave

portions of a traverse (Fig. 12). Since the paleoflow direction is not known, z is

redefined as

= 100 , (2) + 𝑠𝑠 𝑧𝑧 ∗ 𝑛𝑛 𝑠𝑠

where s and n are the lengths of the convexities in the two possible flow

directions (Fig. 12). In equation (1), z is greater than 55 for traverses with

delayed inflection points, 45-55 for symmetric traverses, and less than 45 for

premature inflection points.

Over the entire study area, 59 traverses were measured by hand for n and

s values. The resultant z values were then calculated with equation (2). Values

of z greater than 55 indicate a delayed inflection towards the southeast, and z

values less than 45 imply a delayed inflection towards the northwest.

Carson and Lapointe (1983) also discuss goosenecks, which are

essentially extremely developed examples of downstream meander asymmetry.

Goosenecks tend to point upstream (Fig. 13).

Determining amount of base level change

On Earth’s surface and in laboratory experiments, a major control on

valley incision depth is the overall amount of sea level change. Although incision

depth cannot be measured at the study area directly, the total thickness of

exhumed valley-filling deposits should serve as an accurate proxy (Fig. 8).

HiRISE and CTX DEMs were used to construct topographic profiles perpendicular to the channelized-corridor axes at half-kilometer intervals. The amount of relief along each profile was measured, and organized into Fig. 14.

Profiles where the inverted sinuous ridges were not the primary source of relief, a result of post-depositional deformation (Lefort et al., 2012) or DEM error, were discarded. The relief measured in this fashion only puts a minimum boundary on

the amount of accumulated valley fill. Erosion may have removed upper portions

of the valley fill, and the basal meandering deposits, which could constitute tens

of meters of stratigraphy (e.g., Miall 2002), may not be completely exhumed.

A DEM of the incised valleys in the northwest Gulf of Mexico cut during the last lowstand (22 to 17 ka; Simms et al., 2007) is used here to assess the amount of base level change that is represented by valley incision. These valleys were cut in response to a 60 meter drop in eustatic sea level. Profiles across every Gulf of Mexico valley were taken at 10 kilometer intervals and incision depths were organized into Fig. 14.

CHAPTER 6: INTERPRETATIONS

Do the dark channelized corridors meet the geometric criteria for incised

valleys?

The dark corridors satisfy the criteria used to identify incised-valley

systems. The corridors have a regional extent and the mapped boundaries of

the corridors are consistent across multiple data sets. The corridors have been inferred to be net erosional in origin on the basis of the scoop-shaped geometry that characterizes their boundaries. The fluvial deposits contained within the corridors show that channels often interacted with, but were always contained within the boundaries that are interpreted here as the remnants of valley walls.

Each of the three corridors meets the criteria for an incised valley system, and are interpreted as such.

Does the fluvial stratigraphy of the dark corridors have a similar pattern to incised valley fill on Earth?

The fluvial stratigraphy observed within the corridors is consistent with stratigraphic successions found in incised valleys on Earth. Meander belt deposits composed of gravel and sand are ubiquitous along the base of the

Brazos River valley (Sylvia and Galloway, 2006), and at the bases of several other incised valleys studied along the northern coast of the Gulf of Mexico (Blum

and Aslan, 2006). Miall (2006) also identified the meander-belt deposits of a

laterally migrating river at the base of an incised valley in the Malay Basin, Gulf of

Thailand. Meander belts are common basal deposits in incised valleys, and the link between valley incision and channel migration of a net-erosional river has been demonstrated experimentally (Koss et al., 1994) and through numerical

models (Martin et al., 2011). Valley cutting, both vertical and lateral, during sea

level fall and lowstand increases the available coarse sediment within the river

system, providing the necessary sediment source to maintain the point-bar

deposition that is a necessary component of additional river-bend migration

(Strong and Paola, 2008; Martin et al, 2011). As sea level begins to rise, valley

cutting ends and the rate and total amount of channel migration diminishes. In

incised valleys of the northern coast of the Gulf of Mexico, valley fill above the

basal migrating deposits are dominantly sandy channel fills, avulsions, and fine- grained floodplains (Blum and Aslan, 2006; Sylvia and Galloway, 2006) that do not record evidence for substantial lateral migration of bends, similar to the deposits at Aeolis Dorsa.

The sinuosity of the inverted ridges stacked directly above the basal amalgamated point bars at the study area is not the product of channel-bend migration at the same rate expressed by the basal-channel deposits. The decrease in sinuosity observed in stratigraphically higher channels is likely caused by two factors. First, sediment supply from upstream is decreased as sea level rises and sedimentation within the valley becomes net-depositional.

The selective deposition of coarse sediment leads to less material being

available to drive additional point bar growth, which in turn decreases the ability

of the channel bends to grow in amplitude. Even so, some of the Aeolis Dorsa

paleochannels exhibit considerable sinuosity while lacking any evidence for

substantial migration, which brings us to the second cause of decreasing channel

sinuosity up-section. Younger channels are reoccupying the pathways of older channels. In the case that a younger river reoccupies an under-filled older channel, the sinuosity of the younger river is inherited from the older channel, rather than being formed by migration (Mohrig et al., 2000). This stacking pattern can be seen at the study area. Channels occupying the valley floor that is now directly above the basal meandering channel are more likely to reoccupy its course and inherit a well-developed sinuosity. Youngest channels located at stratigraphically higher positions are much less likely to be guided by the older, more sinuous channels since they have already been buried, and as a result have a smaller degree of sinuosity (Fig. 2, Fig. 3). The filling of the valley as demonstrated by the burial of older channels is further evidence that the valley has changed from a period of net-erosion to net-deposition, which is the expected response to rising sea level. Fig. 15 compares the paths of channels exhibiting maximum and minimum sinuosity in each valley. On the coast of the

Gulf of Mexico, avulsions and reoccupations similar to those at Aeolis Dorsa formed in valleys during sea level rise as well (Blum and Aslan, 2006).

Where was the shoreline?

Figure 12 is a histogram summarizing results of the meander asymmetry

analysis for this study. The abundance of traverses with z > 55, relative to

traverses with z < 45, indicates that the paleoflow direction, and therefore the

paleo-shoreline, was towards the southeast. There are several examples in the study area of goosenecks pointing to the northwest (Fig. 13), further suggesting paleoflow was towards the southeast.

Had several of the amalgamated point bars clearly indicated a migration direction as they have potential to (e.g., Nicoll and Hicken, 2010, Fig. 1), the meander asymmetry of the paleochannels would have been a secondary line of evidence for the transport direction, rather than the primary argument. However, the majority of the preserved record for channel migration was purely lateral, rather than containing an easily observed downstream component (Fig. 16, Fig.

12: note the high percentage of symmetric meanders). It’s possible that the high degree of lateral meanders is a result of the relatively low gravity of Mars. Sun et al., (1996) demonstrated that a flatter floodplain will encourage lateral meander migration over downstream migration. Since gravity’s effect on sediment transport is modified by the slope of the bed, a lower gravitational acceleration might be expected to have the same effect on meander migration as a flatter floodplain would. An abundance of primarily symmetric meanders should not be surprising. Application of the Carson and Lapointe (1983) method is necessary to 27

detect the subtle signature for the transport direction preserved in deposits of the

paleochannels (Fig. 12).

Two lines of valley-scale evidence also point toward a transport direction and shoreline to the southeast. First, the number of observed channels increases toward the southeast (Fig. 6). This increasing number of channels does not

appear to be a preservational artifact associated with differential erosion, but

rather can be interpreted to represent the consequence of more frequent

avulsions tied to greater rates of sediment deposition (Bryant et al., 1995; Mohrig et al., 2000) toward the paleo-shoreline.

Second, the overall valley system widens to the southeast when Aeolis

Serpens is considered its upstream extension (Fig. 7). Downstream-increasing amounts of coarse sediment load, partially mined from the valley walls promotes point bar growth along the inner banks of channel bends which, along with the erosion of the outer bank, drives lateral migration (Ikeda et al., 1981; Parker et al., 2011) and increases migration rates of channels (Smith 2012) which ultimately leads to increased valley widening (Martin et al., 2011; Meakin et al.,

1996).. Finally, the spatial arrangement of the red, green, and purple valleys defines a distributary pattern if the shoreline is located to southeast. Such a distributary pattern is commonly observed in coastal-zone valley systems on

Earth (Greene et al., 2007).

About 30 kilometers to the south and southeast of the study area is a large region of fluvial deposits (Fig. 17) which appear qualitatively similar to the delta identified by DiBiase et al., (2013). The zone separating the study area and these deposits at least partially covered by an aeolian dune field, so definitive correlation between the two sites using exhumed stratigraphy is not possible.

However, the location and arrangement of these deposits suggest that they may be the remnants of a deltaic shoreline positioned at the downstream end of the incised valleys discussed here.

How many times did base level change?

The truncation of the red valley strata by the green valley (Fig. 4) indicates that the entire red valley fill is older, and points to at least two separate episodes of base level rise and fall. Filling of the red valley is interpreted to have contributed to a redirection of the river system when incision began during a subsequent base-level fall (e.g., Brazos River valley, Anderson et al., 1996). In this scenario, the green system would have removed the strata associated the upstream portion of the previously cut and then filled red valley. The only record of the red valley is now found downstream of the point where a channel avulsion shifted cutting of the green valley eastward of the older red-valley deposits. A different relationship is observed between the green and purple valleys (Fig. 4) where no truncated channel fills can be found. This lack of truncation is

interpreted as evidence for the coeval incision and filling of the green and purple valleys.

How much did base level change?

The valley relief estimates at Aeolis Dorsa and the northwest Gulf of

Mexico during the previous lowstand in sea level are presented in Fig. 14. The greatest thickness of exhumed Martian valley deposits is 50 meters in the green valley. The most frequently measured thicknesses within each of the three corridors are in the 11-20 meter range. The valleys at the northwestern coast of the Gulf of Mexico were most frequently incised to depths of 26-30 meters, although incision depths reached as high as 50 meters.

The results from the Gulf of Mexico emphasize the fact that incision depth and fill thickness can only serve as less-than-minimum estimations of the amount of sea or lake level change, as valleys are never observed to incise as deeply as the total sea level fall. The valley incision depths measured with this dataset were formed during a eustatic 60 meter sea level fall, yet the most frequent incision depths are about half of the total sea level fall or less (26-30 meters).

Peak values are less than the total sea level fall.

Other authors have observed this as well. Mattheus et al., (2007) measured the dimensions of 9 incised valleys along the northern coast of the

Gulf of Mexico near their modern bayhead deltas. Incision depth varied from

about 5 to 30 meters, and valley widths varied from about 500 meters to over 16 kilometers. The variability in valley dimensions, which were all incised during the

same eustatic sea level drop, is partly explained by variations in drainage basin

area and sediment supply. Sea level drop triggers incision if an underwater

slope break is subaerially exposed, and increased rates of sea level drop can

deepen incision depths (Martin et al., 2011), but it is not the only control on valley

dimensions.

Although valley widths generally increase downstream (Martin et al.,

2011), depths tend to vary by, at most, a few tens of meters. The Brazos River

Valley is incised about 22 meters into the surrounding terrain at the modern coastal plain. Cores and well logs over 100 kilometers upstream show average incision depths to be about 18 meters (Sylvia and Galloway, 2006). Incised valleys of the Alabama and Tombigbee rivers have a relatively constant incision depth of 30 meters over a distance of 200 kilometers from the modern shoreline

(Greene et al., 2007). Incision depths are deepest at the inner continental shelf, where depths of about 40 meters were measured for the Brazos, Colorado, and

Trinity River valleys (Anderson et al., 1996).

Experiments have yielded similar results. An experiment run by Paola et al., (2001) showed incision depths of about 33% of the total water surface level change. In this experiment, incision depths were nearly equal for a rapid and a

slow change in water surface level, although incision rates differed and valley formation was much more pronounced during the rapid fall.

Shoreline progradation during a sea level fall inhibits deeper incision.

Progradation creates more distance over which the fluvial system will reach the ocean, leading to gentler fluvial slopes and buffering downcutting.

Considering that the most frequent incision depths in the northwestern

Gulf of Mexico are about half of the eustatic sea level drop (Fig. 14), and that the maximum incision depths are less than the total sea level drop, the frequent depths of 11-20 meters and upper limits of 50 meters at Aeolis Dorsa could reasonably be attributed to a sea or lake level change of over 50 meters, especially when the limitations of measuring relief are considered, such as erosion of exposed inverted sinuous ridges and incomplete exhumation of basal, migratory deposits. The magnitude of this change in water-surface elevation is falls within a wider range of values estimated by DiAchille and Hynek (2010) using interpreted shorelines. The advantage of the water-level change estimate generated by this study is that it is tied to a single location and constrained by coherent successions of strata.

Sequence of events

The sequence of events that led to the creation of the deposits at the study area is as follows:

1. Incision of the red valley, which extended upstream into the position of

the current green valley, during an episode of rapid sea or lake level fall of

at least 50 meters.

2. Complete filling of a downstream reach of the red valley during an episode

of sea or lake level rise of at least 50 meters.

3. Incision of the green and purple valleys, including the reworking of unfilled,

upstream portions of the red valley, during an episode of rapid sea or lake

level fall of at least 50 meters.

4. Filling of the green and purple valleys during an episode of sea or lake

level rise of at least 50 meters, with the last period of filling in the green

valley by the red channel.

If the red channel is the downstream portion of Aeolis Serpens as it appears to be, then the final episode of valley filling can be dated to the early Amazonian.

Aeolis Serpens has been dated to this period, due to its incision into the

Cerberus lava plains (Jacobsen and Burr, 2013).

CHAPTER 7: CONCLUSIONS

There are multiple lines of evidence demonstrating that the dark channelized corridors at Aeolis Dorsa are the remnants of valleys that were incised by rivers during periods of sea level fall, and filled by rivers during periods of sea level rise. The lateral extents of the corridors were created by the erosion of the outer banks of migrating channels. Incision depths of several tens of meters are implied by the fact the deposits left by younger channels never leave the area bounded by the basal migrating channels and are likely only a fraction of the total sea or lake level fall, which we have estimated at 50 meters or greater.

The distance covered by these features is over 40 km and identifiable in multiple data sets, ruling out a localized deep scour. The fluvial stratigraphy exhibited within the corridors, which contains the deposits of a rapidly migrating channel at the base and deposits which show little evidence of migration stratigraphically above that, is common in incised valleys on Earth and an expected response to changes in coarse sediment supply caused by changes in sea level. Fluvial deposits further downstream appear to be deltaic. The implication of a nearby large body of water, such as a sea or large lake, is corroborated by the identification of deltaic deposits within the region (DiAchille and Hynek, 2010;

DiBiase et al., 2013) and by hypothesized shorelines of an ancient Martian ocean

(Parker et al., 1989, Perron et al., 2007). The age of the MFF and Cerberus lavas which these deposits have reworked vary from Hesperian to early

Amazonian, indicating that the suggested episodes of sea or lake level change occurred around the same time.

CHAPTER 8: FIGURES

Fig. 1 – This MOLA DEM shows the location of the Aeolis Dorsa basin and the surrounding lobes of the Medusae Fossae Formation relative to Gale Crater. The sharp increase in elevation (black dashed line) marks the boundary between the southern highlands and northern lowlands (the crustal dichotomy of Mars) (Tanaka et al., 2014). Elevation on Mars is measured relative to a geoid. The northern lowlands, including Aeolis Dorsa, exist up to several kilometers below the geoid and have a negative elevation. A black arrow points to the dark corridors in southeastern Aeolis Dorsa. The inset more precisely defines the location of the study area.

Fig. 2 – Top: CTX mosaic of the study area. Arrows point to the low-albedo corridors. The corridors are branching, channelized features several tens of kilometers long. Each corridors contains several channel-filling deposits preserved as inverted sinuous ridges. Bottom: This interpreted version of the top image maps three unique corridors (red, green, and purple) and the inverted sinuous ridges contained each corridor. Paleoflow direction is interpreted to be towards the right of the image (southeast).

Fig. 3 – The clusters of sinuous ridges, channel bends, and scooped corridor boundaries associated with cutting by the erosive outer banks of these bends. The arrows point to the stratigraphically highest sinuous ridge in the green corridor (Fig. 2, red channel), which is rarely drawn to the location of lower channels and has a relatively low sinuosity. The relative smoothness of the dark corridors compared to the rough surrounding terrain can be used to distinguish the dark corridors from the surrounding terrain.

Fig. 4 – Left, top and bottom: The green valley truncates deposits from the red corridor, indicating that the red valley formed first, and upstream portions were later reworked by flow through the green corridor. Right, top and bottom: Deposits at the contact between the green and purple corridors have not been truncated.

Fig. 5 – Top: An example of fluvial stratigraphy observed within one of the dark corridors. A few basal, curved linea are mapped with black lines, and are interpreted as point bar deposits. These deposits are extensive throughout the corridors. Above the basal deposits are two generations of channel-filling deposits, preserved as inverted sinuous ridges. Bottom: A HiRISE-derived DEM of the same location, which resolves the different elevations associated with the basal point bar deposits and the following two generations of sinuous ridges. There is approximately 40 meters of relief between the basal and upper deposits. Post-depositional deformation described by Lefort et al., (2012) likely caused the rapid increase in elevation towards the southeast. 40

Fig. 6 – The number of unique sinuous ridges in the red valley (top, left and right) and purple valley (bottom, left and right) increases to three towards the southeast, the interpreted downstream direction.

Fig. 7 – The fluvial deposit mapped as part of Aeolis Serpens (subset, tan sinuous ridge) just upstream of the study area. Aeolis Serpens shows no signs of significant migration, such as the amalgamated point bar deposits seen throughout the rest of the study area (e.g., Fig. 4). However, the narrow, dark region surrounding Aeolis Serpens begins to widen towards the study area, and features a scooped boundary in the bottom-right corner of the image. The length of the inset scale bar is 10 kilometers.

Fig. 8 – The proposed origin of the dark corridors presented in three steps: 1. Valley incision and lateral valley growth via an actively migrating, eroding river during base level fall. 2. Valley filling by single-thread, aggradational rivers. 3. Erosion of the surrounding MFF and floodplain deposits, with preferential preservation of the coarse-grained, cemented channel-filling deposits. The maximum relief is an approximation of the thickness of the valley fill and, assuming complete valley-filling, incision depth. 43

Fig. 9 – The valleys are visible in THEMIS nighttime thermal infrared imagery. Brighter regions have a higher thermal inertia. The valley outlines were mapped over CTX and HiRISE imagery, and coincide with brighter regions in the thermal image. Cementation, a common reason fluvial deposits on Earth are topographically inverted, would result in a denser rock and explain the higher thermal inertia.

Fig. 10 – Top: An SRTM DEM of the Brazos River valley, about 85 kilometers upstream from the Gulf of Mexico shoreline, Texas, USA. This valley is wider than the valleys on Mars, but many features used to identify the Martian valleys are visible here, such as the scooped shape of the valley walls, the rivers abutting valley walls, and the river confinement. Bottom: The elevation profile from A to A’. The two channels visible in this image occupy floodplains separated by two meters of elevation. The different floodplain elevations are marked by the black arrows. 45

Fig. 11 – This figure is a mapped version of Fig. 4. The yellow channel is younger than the brown channel and stacked on top of it. Near the center of the image, the yellow channel had avulsed away from the lower brown channel (a). The yellow channel maintains contact with the inner boundary of the valley, indicating the channel may have been redirected along the valley wall (b). At some locations, the yellow channel has reoccupied the under-filled brown channel, and been guided by its topography (c).

Fig. 12 – Top: Components of the meander asymmetry analysis. Inflection points are dark circles, traverse endpoints (curvature maximums) are hollow circles labeled 1, 2, and 3. The terms s and n describe the portion of a traverse which is convex towards the south or to the north, respectively. The inflection point in traverse 1-2 is delayed from the center of the traverse, making traverse 1-2 asymmetric. The larger s1,2 relative to the n1,2 suggests flow was towards the south, but many meanders must be sampled. Traverse 2-3 is symmetric, meaning s2,3 equals n2,3, and no paleoflow direction is implied. Bottom: A histogram of meander asymmetry index (z) values from measurements made from channel-filling deposits at Aeolis Dorsa and active rivers on Earth from Carson and Lapointe (1983). In both examples, the majority of bends have z values greater than 55. The higher occurrence of z>55 indicates paleoflow at Aeolis Dorsa was towards the south. 47

Fig. 13 – Top: A gooseneck in the Brazos River points upstream as predicted (Carson and Lapointe, 1983). Bottom: Goosenecks at Aeolis Dorsa point north- northwest, indicating paleoflow was towards the south-southeast.

Fig. 14 – Top: The distribution of relief measurements across each Martian valley is similar. The most frequent range is from 11-20 meters. Bottom: Valley relief in the northwestern Gulf of Mexico during the last lowstand formed during a 60 meter drop in eustatic sea level. The most frequent incision depths measured are half of that or less, and no incision depth is measured to be greater than that. Assuming a similar distribution, and with a discussion of the limits of the Martian data, we estimate base level change at Aeolis Dorsa to have been at least 50 meters.

Fig. 15 – The lighter lines represent the paths of maximum sinuosity in each valley, and the bold lines represent the path of least sinuosity. Sinuosity is defined as the path length divided by the length of the valley axis, and has decreased over time. The purple valley contains only a single sinuous ridge for most of its path.

Fig. 16 – These deposits show point bar migration was generally lateral, away from the valley axis, rather than downstream. The examples here are representative of the study area, and are the primary reason why determining paleoflow direction has been challenging. The high degree of laterally-migrating meanders is represented by the abundance of symmetric meanders (Fig. 12).

Fig. 17 – Top: The set of fluvial deposits circled in orange appears to be deltaic. These deposits are downstream from the incised valleys of the study area. Bottom: Qualitatively, the deposits at the downstream end of the study area (left) are similar to the delta identified by DiBiase et al., (2013) (right). Both regions have tightly clustered sets of branching, distributary fluvial deposits, and the scale is similar.

REFERENCES

Aitken, J.F., and Flint, S.S., 1995, The application of high-resolution sequence stratigraphy to fluvial systems: a case study from the Upper Carboniferous Breathitt Group, eastern Kentucky, USA: Sedimentology, v. 42, p. 3-30. Anderson, J.B., Abdulah, K., Sarzalejo, S., Siringan, F., and Thomas, M.A., 1996, Late Quatenary sedimentation and high-resolution sequence stratigraphy of the east Texas shelf: Geological Society Special Publication, no. 117, p. 95-124. Armstrong, C.P., 2012, 3D seismic geomorphology and stratigraphy of the late Miocene to Pliocene Mississippi River delta: Fluvial systems and dynamics: Thesis, The University of Texas at Austin. Baker, V.R., 2009, Megafloods and global paleoenvironmental change on Mars and Earth: Geological Society of America Special Papers, no. 453, p. 25- 36. Blum, M.D., and Aslan, A., 2006, Signatures of climate vs. sea-level change within incised valley-fill successions: Quaternary examples from the Texas Gulf Coast: Sedimentary Geology, v. 190, p. 177-211, doi:10.1016/j.sedgeo.2006.05.024. Boyd, R., Dalrymple, R.W., and Zaitlin, B.A., 2006, Estuarine and incised-valley facies models: SEPM Special Publication No. 84, p. 171-235, ISBN 1- 56576-121-9. Bryant, M., Falk, P., and Paola, C., 1995, Experimental study of avulsion frequency and rate of deposition: Geology, v. 23, no. 4, p. 365-368. Burr, D.M., Enga, M.-T., Williams, R.M.E., Zimbelman, J.R., Howard, A.D., and Brennand, T.A., 2009, Pervasive aqueous paleoflow features in the Aeolis/Zephyria Plana region, Mars: Icarus, v. 200, p. 52-76, doi:10.1016/j.icarus.2008.10.014. Burr, D.M., Williams, R.M.E., Wendell, K.D., Chojnacki, M., and Emery, J.P., 2010, Inverted fluvial features in the Aeolis/Zephyria Plana region, Mars: Formation mechanism and initial paleodischarge estimates: Journal of Geophysical Research, v 115, E07011, doi:10.1029/2009JE003496. Carson, M.A., and Lapointe, M.F., 1983, The inherent asymmetry of river meander planform: The Journal of Geology, v. 91, p. 41-55.

Christensen, P.R., Jakosky, B.M., Kieffer, H.H., Malin, M.C., McSween, Jr., H.Y., Nealson, K., Mehall, G.L., Silverman, S.H., Ferry, S., Caplinger, M., and Ravine, M., 2004, The Thermal Emission Imaging System (THEMIS) for the Mars 2001 Odyssey mission: Space Science Reviews, v. 110, p. 85- 130. Di Achille, G., and Hynek, B.M., 2010, Ancient ocean of Mars supported by global distribution of deltas and valleys: Nature Geoscience, v. 3, p. 459- 463, doi:10.1038/ngeo891. DiBiase, R.A., Limaye, A.B., Scheingross, J.S., Fischer, W.W., and Lamb, M.P., 2013, Deltaic deposits at Aeolis Dorsa: Sedimentary evidence for a standing body of water in the northern plains of Mars: Journal of Geophysical Research: Planets, v. 118, p. 1285-1302, doi:10.1002/jgre.20100. Ehlmann, B.L., Mustard, J.F., Murchie, S.L., Bibring, J.-P., Meunier, A., Fraeman, A.A., and Langevin, Y., 2011, Subsurface water and clay mineral formation during the early history of Mars: Nature, v. 479, p. 53-60, doi:10.1038/nature10582. Fairén, G.A., Fernández-Remolar, D., Dohm, J.M., Baker, V.R., and Amils, R., 2004, Inhibition of carbonate synthesis in acidic oceans on early Mars: Nature, v. 431, p. 423-426, doi:10.1038/nature02911. Greene, D.L., Rodriguez, A.B., and Anderson, J.B., 2007, Seaward-branching coastal-plain and piedmont incised-valley systems through multiple sea- level cycles: Late Quaternary examples from Mobile Bay and Mississippi Sound, U.S.A.: Journal of Sedimentary Research, v. 77, p. 139-158, doi:10.2110/jsr.2007.016. Head III, J.W., Hiesinger, H., Ivanov, M.A., Kreslavsky, M.A., Pratt, S., and Thomson, B.J., 1999, Possible ancient oceans on Mars: Evidence from Mars Orbiter Laser Altimeter Data: Science, v. 286, p. 2,134-2,137, doi:10.1126/science.286.5447.2134. Howard, A.D., Moore, J.M., and Irwin III, R.P., 2005, An intense terminal epoch of widespread fluvial activity on early Mars: 1. Valley network incision and associated deposits: Journal of Geophysical Research, v. 110, E12S14, doi:10.1029/2005JE002459. Ikeda, S., Parker, G., and Sawai, K., 1981, Bend theory of river meanders. Part 1. Linear development: J. Fluid Mech., v. 112, p. 363-377.

Jacobsen, R.E., and Burr, D.M., 2013, Local-scale stratigraphy of inverted fluvial features in Aeolis Dorsa, western Medusae Fossae Formation, Mars: 44th Lunar and Planetary Science Conference, poster #2165. Kim, J.R., and Muller, J.-P., 2009, Multi-resolution topographic data extraction from Martian stereo imagery: Planetary and Space Science, v. 57, p. 2095-2112, doi:10.1016/j.pss.2009.09.04 Kite, E.S., Lucas, A., and Fassett, C.I., 2013, Pacing early Mars river activity: Embedded craters in the Aeolis Dorsa region imply river activity spanned > (1-20) Myr: Icarus, v. 225, p. 850-855, doi:10.1016/j.icarus.2013.03.029. Koss, J.E., Ethridge, F.G., and Schumm, S.A., 1994, An experimental study of the effects of base-level change on fluvial, coastal plain and shelf systems: Journal of Sedimentary Research, v. B64, no. 2, p. 90-98. Lefort, A., Burr, D.M., Beyer, R.A., and Howard, A.D., 2012, Inverted fluvial features in the Aeolis-Zephyria Plana, western Medusae Fossae Formation, Mars: Evidence for post-formation modification: Journal of Geophysical Research, v. 117, E03007, doi:10.1029/2011JE004008. Levy, J.S., Head, J.W., Marchant, D.R., and Kowalewski, D.E., 2008, Identification of sublimation-type thermal contraction crack polygons at the proposed NASA Phoenix landing site: Implications for substrate properties and climate-driven morphological evolution: Geophysical Research Letters, v. 35, L04202, doi:10.1029/2007GL032813. Malin, M.C., Bell III, J.F., Cantor, B.A., Caplinger, M.A., Calvin, W.M., Clancy, R.T., Edgett, K.S., Edwards, L., Haberle, R.M., James, P.B., Lee, S.W., Ravine, M.A., Thomas, P.C., and Wolff, M.J., 2007, Context Camera investigation on board the Mars Reconnaissance Orbiter: Journal of Geophysical Research, v. 112, E05S04, doi:10.1029/2006JE002808. Malin, M.C., and Edgett, K.S., 1999, Oceans or seas in the Martian northern lowlands: High resolution imaging tests of proposed coastlines: Geophysical Research Letters, v. 26, no. 19, p. 3049-3052. Mandt, K.E., de Silva, S.L., Zimbelman, J.R., and Crown, D.A., 2008, Origin of the Medusae Fossae Formation, Mars: Insights from a synoptic approach: Journal of Geophysical Research, v. 113, E12011, doi:10.1029/2008JE003706. Martin, J., Cantelli, A., Paola, C., Blum, M., and Wolinsky, M., 2011, Quantitative modeling of the evolution and geometry of incised valleys: Journal of Sedimentary Research, v. 81, p. 64-79, doi:10.2110/jsr.2011.5. 55

Mattheus, C.R., Rodriguez, A.B., Greene Jr., D.L., Simms, A.R., and Anderson, J.B., 2007, Control of upstream variables on incised-valley dimension: Journal of Sedimentary Research, v. 77, p. 213-224, doi:10.2110/jsr.2007.022. McEwen, A.S., Eliason, E.M., Bergstrom, J.W., Bridges, N.T., Hansen, C.J., Delamere, W.A., Grant, J.A., Gulick, V.C., Herkenhoff, K.E., Keszthelyi, L., Kirk, R.L., Mellon, M.T., Squyres, S.W., Thomas, N., and Weitz, C.M., 2007, Mars Reconnaissance Orbiter’s High Resolution Imaging Science Experiment (HiRISE): Journal of Geophysical Research, v. 112, E05S02, doi:10.1029/2005JE002605. Meakin, P., Sun, T., Jøssang, and Schwarz, K., 1996, A simulation model for meandering rivers and their associated sedimentary environments: Physica A, v. 233, p. 606-618. Miall, A.D., 2002, Architecture and sequence stratigraphy of Pleistocene fluvial systems in the Malay Basin, based on seismic time-slice analysis: AAPG Bulletin, v. 86, no. 7, p. 1201-1216. Mohrig, D., Heller, P.L., Paola, C., and Lyons, W.J., 2000, Interpreting avulsion process from ancient alluvial sequences: Guadalope-Matarranya system (northern Spain) and Wasatch Formation (Colorado): Geological Society of America Bulletin 2000, v. 112, no. 12, p. 1787-1803, doi:10.1130/0016- 7607. Moore, J.M., Howard, A.D., Dietrich, W.E., and Schenk, P.M., 2003, Martian layered fluvial deposits: Implications for Noachian climate scenarios: Geophysical Research Letters, v. 30, no. 24, 2292, doi:10.1029/2003GL019002. Moscardelli, L., 2014, Boulders of the Vastitas Borealis Formation: Potential origin and implications for an ancient martian ocean: GSA Today, v. 24, no. 2, p. 4-10, doi:10.1130/GSATG197.A.1. Moscardelli, L., Dooley, T., Dunlap, D., Jackson, M., and Wood, L., 2012, Deep- water polygonal fault systems as terrestrial analogs for large-scale Martian polygonal terrains: GSA Today, v. 22, no. 8, p. 4-9, doi:10.1130/GSATG147A.1. Nicoll, T.J., and Hickin, E.J., 2010, Planform geometry and channel migration of confined meandering rivers on the Canadian prairies: Geomorphology, v. 116, p. 37-47, doi:10.1016/j.geomorph.2009.10.005.

Pain, C.F., Clarke, J.D.A., and Thomas, M., 2007, Inversion of relief on Mars: Icarus, v. 190, p. 478-491, doi:10.1016/j.icarus.2007.03.017. Paola, C., Mullin, J., Ellis, C., Mohrig, D., Swenson, J.B., Parker, G., Hickson, T., Heller, P.L., Pratson, L., Syvitski, J., Sheets, B., and Strong, N., 2001, Experimental Stratigraphy, GSA Today, v. 11, no. 7, p. 4-9. Parker, G., Shimizu, Y., Wilkerson, G.V., Eke, E.C., Abad, J.D., Lauer, J.W., Paola, C., Dietrich, W.E., and Voller, V.R., 2011, A new framework for modeling the migration of meandering rivers: Earth Surf. Process. Landforms, v. 36, p. 70-86, doi:10.1002/esp.2113. Parker, T.J., Saunders, R.S., and Schneeberger, D.M., 1989, Transitional morphology in west Deuteronilus Mensae, Mars: Implications for modification of the lowland/upland boundary: Icarus, v. 82, p. 111-145. Perron, J.T., Mitrovica, J.X., Manga, M., Matsuyama, I., and Richards, M.A., 2007, Evidence for an ancient martian ocean in the topography of deformed shorelines: Nature, v. 447, p. 840-843, doi:10.1038/nature05873. Simms, A.R., Anderson, J.B., Milliken, K.T., Taha, Z.P., and Wellner, J.S., 2007, Geomorphology and age of the Oxygen isotope stage 2 (last lowstand) sequence boundary on the northwestern Gulf of Mexico continental shelf: Seismic Geomorphology: Applications to Hydrocarbon Exploration and Production. Geological Society, London, Special Publications, v. 277, p. 29-46. Smith, D.E., Zuber, M.T., Frey, H.V., Garvin, J.B., Head, J.W., Muhleman, D.O., Pettengill, G.H., Phillips, R.J., Solomon, S.C., Zwally, H.J., Banerdt, W.B., Duxbury, T.C., Golombek, M.P., Lemoine, F.G., Neumann, G.A., Rowlands, D.D., Aharonson, O., Ford, P.G., Ivanov, A.B., Johnson, C.L., McGovern, P.J., Abshire, J.B., Afzal, R.S., and Sun, X., 2001, Mars Orbiter Laser Altimeter: Experiment summary after the first year of global mapping on Mars: Journal of Geophysical Research, v. 106, E10, p. 23,689-23,722. Strong, N., and Paola, C., 2008, Valleys that never were: time surfaces versus stratigraphic surfaces: Journal of Sedimentary Research, v. 78, p. 579- 593, doi:10.2110/jsr.2008.059. Sun, T., Meakin, P., Jøssang, T., and Schwarz, K., 1996, A simulation model for meandering rivers: Water Resources Research, v. 32, no. 9, p. 2937- 2954.

Sylvia, D.A., and Galloway, W.E., 2006, Morphology and stratigraphy of the late Quaternary lower Brazos valley: Implications for paleo-climate, discharge and sediment delivery: Sedimentary Geology, v. 190, p. 159-175, doi:10.1016/j.sedgeo.2006.05.023. Tanaka, K.L., Skinner, J.A., Dohm, J.M., Irwin, R.P., Kolb, E.J., Fortezzo, C.M., Platz, T., Michael, G.G., and Hare, T.M., 2014, Geologic Map of Mars: United States Geological Survey Scientific Investigations Map 3292. Smith, V.B., 2012, Geomorphology of a coastal sand-bed river: lower Trinity River, Texas. Dissertation, The University of Texas at Austin. Williams, R.M.E., Irwin III, R.P., Burr, D.M., Harrison, T., and McClelland, P., 2013, Variability in martian sinuous ridge form: Case study of Aeolis Serpens in the Aeolis Dorsa, Mars, and insight from the Mirackina paleoriver, South Australia: Icarus, v. 225, p. 308-324, doi:10.1016/j.icarus.2013.03.016. Wood, L. J., 2006, Quantitative geomorphology of the Mars Eberswalde delta: GSA Bulletin, v. 118, no. 5/6, p. 557-566, doi:10.1130/B25822.1. Zimbelman, J.R., and Scheidt, S.P., 2012, Hesperian age for western Medusae Fossae Formation, Mars: Science, v. 336, p. 1,683, doi:10.1126/science.1221094.

Vita

Benjamin Thomas Cardenas was born in Fort Stewart, Georgia, USA.

Following his graduation in 2005 from the Northeast School of the Arts High

School in San Antonio, Texas, he eventually attended The University of Texas at

San Antonio, where he earned his B.S. in geology and mathematics minor in

2012. He pursued his M.S. in geology at The University of Texas at Austin and graduated in 2014. He currently lives in Austin, TX, with his girlfriend Nicole and guinea pigs, Snickers and Goggles. In 2015, he will enter the PhD program at

The University of Texas at Austin.

Email address: [email protected]

Website: sites.google.com/a/utexas.edu/benjamincardenas

This thesis was typed by Benjamin Thomas Cardenas.