SURFACE WAVE INVERSION OF THE UPPER MANTLE VELOCITY

STRUCTURE IN THE ROSS SEA REGION,

WESTERN ANTARCTICA

__________________________________

A Thesis

Presented to

The Graduate Faculty

Central Washington University

__________________________________

In Partial Fulfillment

of the Requirements for the Degree

Master of Science

Geology

__________________________________

by

James D. Rinke

June 2011

ii

CENTRAL WASHINGTON UNIVERSITY

Graduate Studies

We hereby approve the thesis of

James D. Rinke

Candidate for the degree of Master of Science

APPROVED FOR THE GRADUATE FACULTY

______________ _________________________________________ Dr. Audrey Huerta, Committee Chair

______________ _________________________________________ Dr. J. Paul Winberry

______________ _________________________________________ Dr. Timothy Melbourne

______________ _________________________________________ Dean of Graduate Studies

iii

ABSTRACT

SURFACE WAVE INVERSION OF THE UPPER MANTLE VELOCITY

STRUCTURE IN THE ROSS SEA REGION,

WESTERN ANTARCTICA

by

James D. Rinke

June 2011

The Ross Sea in Western Antarctica is the locale of several extensional basins

formed during Cretaceous to Paleogene rifting. Several seismic studies along the

Transantarctic Mountains and Victoria Land Basin’s Terror Rift have shown a general

pattern of fast seismic velocities in East Antarctica and slow seismic velocities in West

Antarctica. This study focuses on the mantle seismic velocity structure of the West

Antarctic Rift System in the Ross Embayment and adjacent craton and Transantarctic

Mountains to further refine details of the velocity structure.

Teleseismic events were selected to satisfy the two-station great-circle-path method

between 5 Polar Earth Observing Network and 2 Global Seismic Network stations

circumscribing the Ross Sea. Multiple filter analysis and a phase match filter were used

to determine the fundamental mode, and linearized least-square algorithm was used to

invert the fundamental mode phase velocity to shear velocity as a function of depth.

Observed velocities were then compared to the AK135-β reference earth model.

iv

Surface wave inversion results reveal three regions with distinct seismic velocity

structures: the East Antarctic craton, the Transantarctic Mountains orogenic zone, and

the extensionally rifted Ross Sea. The extensional zone of the Ross Sea displays slower

seismic velocities than the global average. A seismic velocity structure faster than the

global average is documented in the East Antarctic craton, while the Transantarctic

Mountains display seismic velocities more closely resembling the rift zone than the

craton.

The low velocity zone in the upper mantle of West Antarctica extends from the

Transantarctic Mountains through the Ross Sea to the Marie Byrd Land region. These

slow seismic velocities are suggestive of a warm upper mantle. A warm upper mantel is

difficult to reconcile with the lack of tectonic activity since approximately 30 Ma.

v

TABLE OF CONTENTS

Chapter Page

I INTRODUCTION ........................................................................................ 1  

Research Objectives and Significance.................................................... 1  Ross Sea Geologic Setting...................................................................... 2  Previous Seismic Research in the Ross Sea Region............................... 6  Surface Wave Theory ............................................................................. 8

II METHODS................................................................................................. 11  

Data Processing .................................................................................... 11  Data Selection....................................................................................... 16  Computer Programs in Seismology...................................................... 18

III INVERSION RESULTS ............................................................................ 22  

Surface Wave Inversion........................................................................ 22 IV DISCUSSION............................................................................................. 28  

Extension, Craton, and Orogeny Seismic Velocity Structures ............. 28 V SUMMARY................................................................................................ 32  

REFERENCES ........................................................................................... 33  

APPENDIXES............................................................................................ 39  

Appendix A - Starting Model ............................................................... 39  Appendix B - Inversion Parameters...................................................... 41  Appendix C - Inversion Profiles ........................................................... 42  

vi

LIST OF TABLES

Table Page

1 17 Station Pairs and Events Used for Group Velocity Analysis. ............... 17

2 Suggested Filter Parameter Values with Regards to Event-Station Distance ............................................................................... 19

3 Surface, Minimum and Maximum Velocities with Associated Depths of Each Interstation Path ................................................................ 25

A1 AK135-β Average Global Seismic Velocity Model................................... 39

B1 Layer Weighting for Inversion Process ...................................................... 41

vii

LIST OF FIGURES

Figure Page

1 Bedrock elevation map of Antarctica oriented grid north ............................ 3

2 Bedrock elevation site map of the Ross Sea................................................. 4

3 Ross Sea lithospheric thickness profile ........................................................ 6 4 Seismic velocity maps at several depth intervals from Ritzwoller et al., (2001)................................................................................ 7 5 2-D example of how surface waves travel along the earth’s surface ........... 9

6 Amplitude and wavelength of P-waves and surface waves........................ 10

7 Group velocity dispersion curve................................................................. 13

8 Phase velocity dispersion curve.................................................................. 14

9 Waveform time shift between two stations ................................................ 15

10 Accepted interstation paths......................................................................... 16

11 Unused group velocity dispersion curve with long wavelength interference ................................................................................................ 18 12 Shear wave velocity profile for station pair MPAT-LONW ...................... 20 13 Sensitivity kernel from the MILR-QSPA interstation path ........................ 21 14 Comparison of inversion profiles ............................................................... 23

15 Map of Ross Embayment showing interstation paths for associated velocity-depth profiles .............................................................. 24

16 Inversion profile comparison for cratons and rifts ..................................... 29 C1 Inversion profile for LONW-MPAT event 10/15/2009 ............................. 42 C2 Inversion profile for MPAT-LONW .......................................................... 43

viii

LIST OF FIGURES (continued) Figure Page

C3 Inversion profile for MPAT-MILR1 .......................................................... 44

C4 Inversion profile for MPAT-MILR2 .......................................................... 45

C5 Inversion profile for LONW-SIPL ............................................................. 46

C6 Inversion profile for SURP-LONW ........................................................... 47

C7 Inversion profile for QSPA-MPAT ............................................................ 48

C8 Inversion profile for MILR-QSPA ............................................................. 49

1

CHAPTER I

INTRODUCTION

Research Objectives and Significance

The extraction of geologic data needed to determine Antarctica’s obscure

evolution is impeded by immense ice sheets covering the continent. In most areas, the

only way to gather geologic information is through the use of geophysical techniques. In

this study the seismic velocity structure of the upper mantle of the Ross Sea in Western

Antarctica is examined using surface wave inversion techniques.

The Ross Sea is located within the West Antarctic rift system (WARS), a region

that underwent regional extension from the Cretaceous through Paleogene time.

However, the current volcanic activity in the Ross Sea is in conflict with the ~30 my of

tectonic inactivity of the WARS (Cande et al., 2000). A mantle plume(s) has been

suggested as the cause of the volcanic centers and regions of uplifted topography in the

region, but there has been no conclusive evidence confirming the existence of a plume.

Previous seismic studies in Antarctica document a seismically slow upper mantle

structure throughout West Antarctica. These continental scale studies lack the resolution

to determine regional details in the seismic structure.

This study uses surface waves recorded by several seismic stations that are part of

the Polar Earth Observing Network (POLENET) to delineate the upper mantle structure

in the Ross Sea and surrounding areas. Characterization of the seismic velocity of the

region will help to unravel the details of the tectonic evolution of the region.

2

Ross Sea Geologic Setting

The continent of Antarctica is bisected into West Antarctica (WA) and East

Antarctica (EA); (Fig. 1). WA is composed of several micro-blocks, while EA is Archean

craton. The geologic discontinuity is marked by the Transantarctic Mountains (TAM), a

3500 km long mountain chain that reaches 4500 m in elevation. The basement of the

TAM is comprised of rocks deformed during multiple orogenies: Beardmore, Nimrod

and Ross (Gunn and Warren, 1962; Stump et al., 1986; Goodge et al., 1991; Storey et al.,

1992). Sediments were then deposited unconformably over the deformed basement

(Storey et al., 1996) before uplift of the TAM occurred in the Cretaceous (Fitzgerald et

al., 1986), coinciding with rifting in WA.

WA is much younger than EA. The Marie Byrd Land (MBL), Antarctic

Peninsula (AP), Ellsworth-Whitmore Mountains (EWM) and Thurston Island (TI) blocks

were accreted onto the EA craton (Fig. 1) as the Phoenix Plate was being subducted

(Weaver et al., 1994) during the break up of Gondwana in the Mesozoic (Dalziel, 1991;

Fitzsimons, 2000; Boger et al., 2001).

Subduction of the Phoenix Plate continued until approximately 105 Ma

(Bradshaw, 1989; ten Brink et al., 1993; Luyendyk, 1995). Subsequent to convergence,

extension between EA and WA commenced sometime between 105 and 95 Ma (Weaver

et al., 1994). Extension produced thinning of the lithosphere in the area from the TAM

over to, and maybe including, the MBL dome and extending across WA. This stretched

and thinned area is referred to as the WARS (Figs. 1 and 2). Diffuse extension during the

Cretaceous produced a series of extensional basins in the Ross Sea region of the WARS

3

Figure 1. Bedrock elevation map of Antarctica oriented grid north. This map shows the East Antarctic craton and micro-blocks of WA: East Antarctica (EA), Antarctic Peninsula (AP), Ellsworth-Whitmore Mountains (EWM), Thurston Island (TI), and Marie Byrd Land (MBL). Diagonal lines bounded by dashed red lines represent the approximate location of the WARS, whose boundaries are still largely undetermined. The light gray area symbolizes the Ross Ice Shelf. The axes are distance from the South Pole in meters. The vertical color bar represents elevation in meters (bedrock elevation data from Lythe et al., 2000).

before a transition to focused extension in the Victoria Land Basin (VLB)/Terror Rift

(Fig. 2) during the Eocene to Oligocene (Cande et al., 2000). Sea floor magnetic

anomalies suggest that the Adare Trough was the continuation of the VLB and Northern

basin during the Eocene – Oligocene (Cande et al., 2000). The Adare Trough is located

4

Figure 2. Bedrock elevation site map of the Ross Sea. WA micro-blocks and EA craton are outlined in black: Ellsworth-Whitmore Mountains (EWM), Marie Byrd Land (MBL) and East Antarctic craton. Red circles are seismic stations used in this research. Black dots are seismic stations deployed by TAMSEIS and used by Watson et al., (2006) and Lawrence et al., (2006a,b,c). Bannister et al., (2000) performed surface wave inversion studies in the Terror Rift located in the VLB, represented by cross-hatch pattern. The line of cross section across the rifts is associated with the profile in figure 3 (bedrock elevation data from Lythe et al., 2000).

5

at the southern end of the Northern Basin (Fig. 2). Separation originating in the Adare

Trough provides a tectonic setting for uplift of the TAM, development of the VLB and

Northern Basin, and abundant plutonic and intrusive activity that were occurring

simultaneously (Tonarini et al., 1997; Cande et al., 2000).

An example of the deformation incurred during extension is observed within the

VLB. The Terror Rift is located in the VLB along the TAM scarp, in line with two active

volcanoes, Mt. Erebus and Mt. Melbourne (Fig. 2). The Terror Rift is 12 km deep, 70 km

wide, and the basement crust has been thinned to approximately 6 km (Behrendt, 1999;

Busetti et al., 1999; Trey et al., 1999; Fig. 3). There has been no record of tectonic

activity in the WARS region since the Oligocene extension (Donnellan and Luyendyk,

2004), and very little observed seismic activity. Despite no tectonic activity, there are

active volcanics located within the WARS (Fig. 2). WARS volcanics have an alkaline

basalt geochemical signature that is indistinguishable from ocean island basalts

(LeMasurier, 1990, 2008; Behrendt, 1999; Finn et al., 2005). Ocean island basalts are

understood to have a deep mantle source, which can be provided by a mantle plume. A

mantle plume may exist, but Finn et al. (2005) argues alkaline basalts can exist in WA

without the existence of a mantle plume.

6

Figure 3. Ross Sea lithospheric thickness profile. Generalized profile across the Ross Sea depicting lateral location, depth, and width of extensional basins in relation to the Mohorovičić discontinuity (MOHO), determined from geophysical surveys. Figure adapted from Busetti et al., (1999).

Previous Seismic Research in the Ross Sea Region

Surface wave inversion in the southern hemisphere (latitudes south of -55°) shows

that the craton of EA exhibits seismic velocities faster than average seismic velocities in

global earth models, while seismic velocities in WA are slower than average (Danesi and

Morelli, 2000, 2001, 2004; Ritzwoller et al., 2001; Fig. 4). The abrupt transition from

seismically fast EA to seismically slow WA is located along the TAM.

Regional studies in EA agree with findings of the continental seismic studies.

Seismic analysis in the TAM produced results of slower than average shear velocity (Vs)

extending from VLB, eastward under the front edge of the TAM (Bannister et al., 2000;

7

Figure 4. Seismic velocity maps at several depth intervals from Ritzwoller et al., (2001). The top left map shows seismic velocities at a depth of approximately 40-50 km depth (just below the crust), while the other three map depths are labeled above the corresponding map. Seismic velocities are shown as a percentage deviation from the global earth model AK135, which is the basis for the AK135-β used in this study. Lawrence et al., 2006a, c; Watson et al., 2006). Seismic velocities behind the TAM front

are in accordance with the rest of the craton (~4.6 km/s from Lawrence et al., 2006a, c),

while at the TAM scarp, seismic velocities drop from ~4.6 km/s to 4.2 km/s (Lawrence et

8

al., 2006c). The slow anomaly beneath the TAM scarp has been suggested as warm,

buoyant upper mantle from the Terror Rift, assisting in the uplift of the TAM (Bannister

et al., 2000; Lawrence et al., 2006c). The Terror Rift and VLB exhibit slower than

average seismic velocities, which may be associated with nearby active volcanics. To

date, there have been no seismic studies that go beyond the VLB and further into the

Ross Sea region of the WARS.

The WARS boundary opposite the TAM is presumed to be MBL. The MBL

dome is 1000 km by 550 km, and with over 2700 m of elevation. With substantial

topography, thicker than average crust is expected, but 25 km thick crust leaves MBL

dome out of airy isostatic equilibrium (Winberry and Anandakrishnan, 2004). This

presents a question of what supports the high elevation at MLB dome? Surface wave

inversions in MBL shows evidence of slow seismic velocities to a depth of 625 km,

which is suggestive of warm mantle plume (Sieminski et al., 2003).

This project will determine whether the anomalously slow seismic velocities

observed in the VLB extend across the Ross Sea region, as well as analyze its boundaries.

Details of the seismic velocity structure will provide insight to the tectonic history of the

rift system.

Surface Wave Theory

In this study, I use surface waves to determine the velocity structure of the

shallow earth (depths <~250 km) in the Ross Sea. Surface waves, which are energy

created by earthquakes, become concentrated near the earth’s surface as a result of their

geometric spreading (Fig. 5; Stein and Wysession, 2003). They spread and decay two-

dimensionally with distance from the source, thus their arrival is more prominent on

9

seismograms than body waves (Fig. 6), which attenuate with distance-squared from the

source (Stein and Wysession, 2003). The high peaks of surface wave arrivals are a

measure of their amplitude. The amplitude of surface waves is proportional to the

wavelength and decreases rapidly with depth (Mussett and Khan, 2000). Longer

wavelengths sample greater depths where material is denser due to greater pressure,

producing faster seismic velocities. These relationships allow the examination of discrete

depths within the earth by focusing on specific wavelengths within a seismic record.

Figure 5. 2-D example of how surface waves travel along the earth’s surface. Longer wavelength (λ2) surface waves sample deeper into the earth while both long and short wavelength surface waves lose amplitude with depth (Park Seismics LLC, 2007).

Direct measurements of seismic velocities in the earth are limited to the very near

surface. Below shallow depths, direct measurements of how the seismic velocity changes

at depths cannot be made and must be inferred. Inverse modeling, as used in this study,

is deducing the seismic velocity within the earth from observations. For example, an

earth model is formed based on estimations. The outcome of an analysis with the earth

10

Figure 6. Amplitude and wavelength of P-waves and surface waves. Seismograms of a February 25th, 2008 event recorded by POLENET stations MILR and SURP. Magnitude of body wave arrival is small compared to surface wave arrival. Thickness of blue bars show that first surface wave arrivals have longer wavelengths than later arrivals. This is due to longer wavelengths traveling deeper in the Earth where densities are higher, which allows the seismic waves to travel faster. Horizontal axis is time in 1000’s of seconds, and vertical axis is the unit-less amplitude. model is then compared to the original observations. The model is then modified until its

results match what is observed (Mussett and Khan, 2000).

In this study, the velocity of the upper mantle is determined by combining the

group velocity dispersion curves measured between two stations to create a single phase

velocity dispersion curve. The phase velocity dispersion curve is then the input for the

inversion. The inversion produces an optimized profile of seismic velocities as a function

of depth.

11

CHAPTER II

METHODS

Data Processing

Analysis of the seismic velocity structure of the Ross Sea region will provide

insight to the enigmatic evolution of the WARS. This analysis consisted of: 1) acquiring

data recorded by POLENET and Global Seismic Network broadband seismic stations in

WA; 2) selecting appropriate station pairs for two-station great-circle-path method; 3)

inspecting the quality of the signals; 4) isolating the fundamental mode within the group

velocity and phase velocity; and 5) implementing an inversion to estimate the VS of the

earth’s upper mantle as a function of depth.

All of the acquired seismic traces were evaluated using Seismic Analysis Code

(SAC; Goldstein et al., 2003), obtained from the Incorporated Research Institutions for

Seismology (IRIS). Only Rayleigh waves were examined since only the vertical

components of the seismic traces were analyzed.

Teleseismic events were selected using the JWEEDv3.2 program produced by

IRIS. JWEED allows events to be selected from the IRIS database according to date,

magnitude, depth, and hypocenter determination source. The traces of the described

events are then requested from the desired seismic stations.

Events were also selected based on their geographic location and orientation to

the field area to allow the use of the two-station great-circle-path method. This method

requires that the back azimuth (BAZ) from the far station to the event be within 6° of the

BAZ from the far station to the near station. The BAZ requirement, and high seismic

12

activity made Alaska, Japan and Sumatra regions of focused event selection. Spreading

ridges in the Southern Hemisphere were also targeted, as the seismic activity in spreading

ridges is relatively shallow, which produces strong surface wave arrivals.

Three computer programs developed by Herrmann and Ammon (2004) were used

to analyze and invert the raw surface wave traces: MFT96, POM96 and SURF96.

MFT96 is used to separate the individual modes within the waveform. The program uses

a multiple filter analysis and a phase match filter to identify and isolate the group velocity

fundamental mode (Fig. 7). The phase match filter technique is applied to each seismic

trace to isolate a single mode before the phase velocity stack is performed by POM96.

This is done to avoid yielding a fundamental mode that is still dependent of the amplitude

spectrum of the other modes. MFT96 allows the user to define the spectral amplitude

units, filter parameter, interactive identification of modes, and choice of phase match

filtering. The filter parameter is set based on the distance between the event and the

recording station, and reduces the spectral amplitude biasing on the group velocity

dispersion. Once the user indentifies the mode that is to be analyzed, a phase match filter

is chosen (fundamental mode, first overtone, second overtone, etc.). MFT96 then

isolates the modal spectral amplitude of waveforms as well as the group velocity in km/s

as a function of the period in seconds (Fig. 7).

POM96 performs a time shift and stack to produce a phase velocity dispersion

curve (Fig. 8). The time shift aligns the peaks of the waveforms (Fig. 9) and the

13

Figure 7. Group velocity dispersion curve. Dispersion curve from an event recorded at POLENET station LONW. The left graph shows modal spectral amplitude (count-sec); (vertical axis) vs. period (s); (horizontal axis) for each waveform measured. The right graph shows group velocity (km/s) vs. period (s). The largest amounts of energy are shown in red and are associated with the fundamental mode. The fundamental mode has been manually selected and is highlighted by white squares. The black trends are secondary waveform arrivals. stack integrates the signals from both stations. The user selects the period-velocity

window by selecting the minimum and maximum values. Values are then set for number

of ray parameters, type of surface wave (Love or Rayleigh), x-axis units, x-axis scale, and

length. The dispersion curve is created displaying the stack values indicated by colors,

red being the highest value (Fig. 8).

14

Figure 8. Phase velocity dispersion curve. Dispersion plot displaying phase velocity (km/s) vs. period (s). Waveforms are time shifted to align the peaks and troughs. After alignment, they are stacked resulting in an overall magnitude. The fundamental mode will have the largest magnitude (largest red area), while waveforms that are slightly misaligned produce slightly smaller magnitudes (smaller red areas).

Once the dispersion curve is determined, SURF96 uses the linearized least-square

inversion (LLSI) algorithm to determine Vs as a function of depth from the phase

velocity dispersion curve. LLSI compares the seismic velocity from each layer of the

earth model, and determines how well it fits the observed data. The earth model defines

Vs and compression velocities (Vp), as well as density for select depth intervals (layers)

within the earth. The starting earth model used here (AK135-β) is adapted from the

AK135-f model presented by Herrmann and Ammon (2004). The AK135-f is based

15

Figure 9. Waveform time shift between two stations. The same waveform recorded at two stations, 1 and 2, at some distance apart. The horizontal axis is time in 1000’s of seconds and the vertical axis is the unit-less amplitude. Station 2 is closer to the event, therefore, the arrival reaches it before station 1. The time shift moves the waveform along the time axis until the peaks are aligned. Once the time shift has been applied, the two seismic traces are integrated, and each mode within the two waveforms are summed.

on the AK135 velocity model from Kennett et al. (1995), but augmented with density and

attenuation (Q) based on travel times and free oscillations from Montagner and Kennett

(1995). AK135-β (Appendix A) takes the AK135-f model and adjusts the Vp/Vs and

density for the upper 30 km (depth of the MOHO) based on the recent findings of

receiver function analysis at Mt. Paterson (Svaldi, 2010).

16

Data Selection

Seismic records were requested from IRIS for events that occurred between

January 1st, 2008 and January 1st, 2010, a moment magnitude (Mw) of 4.0 or greater, and

depths of 0 km to 30 km. Requesting events at spreading ridges, Alaska, Japan and

Sumatra from the POLENET seismic array (LONW, MILR, MPAT, SIPL, and SURP)

and Global Seismic Network stations (QSPA and SBA; Fig. 10), produced approximately

2500 seismic traces from nearly 200 events. BAZ analysis yielded 216 event-station

pairs within 6°. After examining the traces for the signal-to-noise ratio,

Figure 10. Accepted interstation paths. Ross Sea location map showing station locations and accepted interstation paths after signal-to-noise ratio examination.

17

only 17 station pairs and events, shown in table 1, were used in the remaining analyses.

The other stations pairs were not used due to low signal-to-noise ratio from long

wavelength signals and multipathing (Fig. 11).

TABLE 1. 17 STATION PAIRS AND EVENTS USED FOR GROUP VELOCITY ANALYSIS. DATE, LOCATION, DEPTH OF EVENT, MOMENT MAGNITUDE

AND BAZ DIFFERENCE IN DEGREES.

Stations Event Date Lat Long Depth

(km) Mw ΔBAZ

LONW-MPAT 10/15/09 3.3° -103.8° 10 5.9 5.29°

LONW-SIPL 11/22/08 -37.2° -94.8° 10 5.7 2.66°

MILR-LONW 2/27/08 26.8° 142.4° 47 6.4 2.08°

MPAT-LONW 7/7/09 -26.7° 67.4° 4 5.6 4.52°

MPAT-MILR 8/22/08 -17.9° 65.4° 10 5.7 4.80°

MPAT-MILR 7/7/09 -26.7° 67.4° 4 5.6 5.43°

MPAT-MILR 10/12/09 -17.2° 66.0° 10 6.0 6.05°

MPAT-SBA 12/8/08 -53.0° 106.9° 10 6.4 3.08°

QSPA-MPAT 6/2/09 -62.8° -158.4° 10 5.6 3.59°

SBA-MPAT 11/22/08 -37.2° -94.8° 10 5.7 2.94°

SBA-SURP 12/19/08 47.1° -27.3° 10 5.9 3.28°

SIPL-LONW 12/8/08 -53.0° 106.8° 10 6.4 2.92°

SIPL_MPAT 6/2/09 -53.0° 106.9° 10 6.4 6.14°

SIPL-SURP 3/26/09 -27.5° 73.3° 10 5.8 9.84°

SIPL-SURP 7/7/09 -26.7° 67.4° 4 5.6 4.37°

SURP-LONW 12/13/08 -48.9° 123.3° 10 5.9 1.07°

SURP-LONW 6/3/09 -49.9° 120.6° 10 5.6 3.09°

18

Figure 11. Unused group velocity dispersion curve with long wavelength interference. Group velocity dispersion curve from LONW station from July 21, 2008 event where ray path passes through MILR and LONW at less then 6° difference. High amplitude, long wavelength interference can be seen on the right of the plot, in red, as well as a spectral hole in the fundamental mode near period of 30 s.

Computer Programs in Seismology

MTF96 is used to isolate the fundamental mode (Fig. 7). In the program, trace

units are set to counts (trace units must match the output from the seismometer), and the

filter parameter is adjusted according to the distance from the event. The greater the

distance between the event and the station, the larger the filter parameter is (3, 6.25, 12.5,

25, 50, 100, or 200). Table 2 gives distance ranges from the source with filter parameter

values recommended by Herrmann. The fundamental mode is selected by hand between

19

periods of ~15 s and ~120 s. Less than 15 s and the arrival has too little energy, and

beyond periods of 120 s the data becomes too attenuated. The selected mode is then

labeled as the fundamental mode for the phase match filter.

TABLE 2. SUGGESTED FILTER PARAMETER VALUES WITH REGARDS TO EVENT-STATION DISTANCE.

Distance Range (km) Filter Parameter (α) 1000 25 2000 50 4000 100 8000 200

Within the POM96 program, the dispersion window is defined by selecting a

minimum and maximum phase velocity of 2 and 5 km/s and periods of 10 and 170 s. The

surface wave is identified as a Rayleigh wave, as opposed to a Love wave or an unknown

waveform. The x-axis is set to a logarithmic scale, representing the period in seconds.

The phase velocity fundamental mode is hand selected from the dispersion curve (Fig. 8),

creating a dispersion file that is used for the inversion in SURF96.

Commands in SURF96 instruct the program to leave the upper 22 km of the data

fixed to the AK135-β starting model. The 10 km surrounding the 30 km deep MOHO

(24-34 km) are gradually given flexibility from the starting model (Appendix B). Since

this study is concerned with the upper mantle, the data have full flexibility to adjust from

the starting model from 34 km to 200 km. Beyond 200 km depth, the data gradually

become more constrained to the starting earth model, AK135-β, until it is fully

constrained at a maximum depth of 400 km, as surface wave signals become very weak at

such depths. The result is a shear wave velocity depth profile for the interstation path

(Fig. 12 and Appendix C).

20

Figure 12. Shear wave velocity profile for station pair MPAT-LONW. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model. The range of periods used in each inversion is determined with the use of

sensitivity kernels (Fig. 13). The kernel represents sensitivity of the phase velocity to a

perturbation of the shear wave speed at a particular depth (Simons et al., 1999). Longer

periods have longer wavelengths and sample deeper in the mantle, thus they are more

sensitive at greater depths. The shortest periods traveling through the crust have very

deep into the mantle and become too attenuated. The periods used for inversions in this

study range from ~15 s to ~149 s.

21

Figure 13. Sensitivity kernel from the MILR-QSPA interstation path. Sensitivity (unit-less) of several periods (s) are plotted as a function of depth (km). The sensitivity plot displays how greater periods are more sensitive to seismic velocity variations at greater depths. At depths of greater than ~250 km, surface waves begin to lose sensitivity as they lose energy due to dissipation and attenuation. Periods of ~15 to ~120 s are used in this study as they are most sensitive the upper mantle depths (~30-250 km).

22

CHAPTER III

INVERSION RESULTS

Surface Wave Inversion

Interstation paths crossing the WARS and the EA craton into the WARS, as well

as the path along the TAM, exhibit seismic velocities slower than the AK135-β earth

model in the upper mantle (Fig. 14). Each WA interstation inversion displays a similar

minimum Vs of 4.1-4.2 km/s at depths of ~80–160 km, compared to the approximate 4.5

km/s Vs of starting model. The one interstation path located entirely on the craton

displays a seismic velocity structure faster than the average earth model. This EA path

has a maximum seismic velocity of approximately 4.7 to 4.9 km/s at depths of ~70–120

km and ~150–250 km. Surface (0-2 km), minimum and maximum seismic velocities at

each interstation path are shown in table 3.

In the following inversion profiles, the starting earth model begins with the Vs of

3.6 km/s from 0 to 8 km depth. At 8 km, the Vs increases by 0.2 km/s every 4 km until it

reaches a seismic velocity of 4.5 km/s from 24 to 55 km. Below 55 km, the starting

model increases seismic velocity from 4.5 km/s to 4.6 km/s at 250 km. Determinations of

anomalously fast or slow are made in comparison to this starting Earth model AK135-β.

The MPAT-LONW inversion profile (Fig. 14) has a surface Vs of 3.4 km/s from

0-2 km depth (from here on out center of the depth interval will be used). The seismic

velocity then increases to 4.3 km/s below 30 km (the estimated MOHO depth in the Ross

Sea). The velocity structure between stations MPAT and LONW then remains slower

than the AK135-β model throughout the upper mantle. This slow seismic velocity

23

Figure 14. Comparison of inversion profiles. The colors outlining each inversion correspond to interstation paths in figure 15. Vs in km/s along the horizontal axis is plotted as a function of depth in km along the vertical axis. The red line represents the observed data and the blue dashed line represents the AK135-β model. Profiles traveling through the Ross Sea (MPAT-LONW, LONW-MPAT, MPAT-MILR1 & 2, and LONW-SIPL) exhibit Vs slower than AK135-β starting model. Profile QSPA-MPAT travels through both EA & WA and exhibits both slower and faster seismic velocities. Profile MILR-QSPA only travels through the EA craton and exhibits Vs faster than AK135-β. The SURP-LONW profile remains in the TAM and displays high seismic velocities at the estimated crust-mantle boundary.

24

Figure 15 . Map of Ross Embayment showing interstation paths for associated velocity-depth profiles. The colors of interstation paths correspond to inversion plots in figure 14. The gray area represents the extensional domain that may include the micro-blocks Marie Byrd Land (MBL) and Ellsworth-Whitmore Mountains (EWM). The double-hatched pattern represents the EA craton, while the V pattern displays the extent of the TAM orogenic domain. Several interstation paths sample across the extensional domain (LONW-MPAT, MPAT-LONW, LONW-SIPL AND MPAT-MILR), SURP-LONW samples only the TAM, MILR-QSPA samples the craton, and QSPA-MPAT samples portions of each domain.

25

TABLE 3. SURFACE, MINIMUM AND MAXIMUM VELOCITIES WITH ASSOCIATED DEPTHS OF EACH INTERSTATION PATH

Interstation Path

Surface Vs

(km/s)

Vs at 30 km

(km/s) Minimum Vs (km/s)

Depth of Minimum

(km)

Fastest Vs

(km/s)

Depth of Fastest (km)

West Antarctic Rift System MPAT-LONW 3.3 4.2 4.2 95-100 - - LONW-MPAT 3.4 4.3 4.2 90-95 - - MPAT-MILR 1 3.4 4.3 4.2 110-120 - - MPAT-MILR 2 3.3 4.2 4.1 130-140 - -

LONW-SIPL 3.3 4.2 4.2 120-130 - - Oregenic Belt (TAM) SURP-LONW 3.2 4.1 4.1 100-110 - -

East Antarctic Craton & West Antarctic Rift System QSPA-MPAT 3.2 4.1 4.2 110-120 4.7+ 250+

East Antarctic Craton MILR-QSPA 2.7 4.2 - - 4.9 230-240

Note: Profiles and station locations can be found on map in figure 14 and all profiles are shown again in figure 15. 30 km is the estimated depth of the MOHO. Fastest velocity and depth at QSPA-MPAT continue to increase beyond the maximum depth of the inversion (250 km).

anomaly reduces to a minimum of 4.2 km/s at a depth of ~90 km. This minimum Vs is

shallow compared to minimum seismic velocities in the other inversion paths (98–135

km).

An event along the Pacific-Cocos plate boundary produced the same interstation

path but from LONW to MPAT (Fig. 14). Compared to MPAT-LONW, LONW-MPAT

is nearly identical. There is a surface Vs of 3.3 km/s, 4.2 km/s at 30 km and minimum Vs

of 4.2 km/s at a depth of ~93 km. The overall trend of each inversion profile is the same.

There are two separate events recorded at the MPAT-MILR path that produce

nearly identical inversion profiles (Fig. 14). The inversion from the July 7, 2009 event

(MPAT-MILR1) produced a surface Vs of 3.4 km/s while the October 12, 2009 event

26

(MPAT-MILR2) produced a surface Vs of 3.3 km/s. The seismic velocity structure

increases to 4.3 km/s and 4.2 km/s, respectively, at 30 km. The only noticeable

difference is the depth of the minimum seismic velocities, 4.2 km/s at 115 km in MPAT-

MILR1 and 4.1 km/s at 135 km in MPAT-MILR2. The minimum velocity of later event

is 20 km deeper but the difference in Vs is negligible. The overall profiles are the same.

LONW-SIPL has an inversion profile (Fig. 14) with a surface Vs of 3.3 km/s,

which increases to 4.2 km/s at 30 km. Below that depth, the velocity structure decreases

to a minimum of 4.2 km/s at 125 km. LONW-SIPL inversion is similar to MPAT-MILR

inversions in that all three have minimum Vs at depths of 115-135 km and the slow

seismic velocity anomaly affects a broad range of depths.

The interstation path SURP-LONW travels across the TAM. The inversion

profile from the SURP-LONW (Fig. 14) interstation path has a surface Vs of 3.2 km/s,

increasing to 4.1 km/s at 30 km. Unlike the other inversions crossing through WA, the

Vs slightly exceeds the starting earth model at depths of 45-55 km. Below this depth, the

seismic velocity decreases to 4.1 km/s at 105 km.

The QSPA-MPAT inversion crosses both the EA craton and the WARS (Figs. 14

and 15) producing a unique profile. QSPA-MPAT, at 1 km depth, has a surface velocity

of 3.2 km/s that increases to 4.1 km/s at 30 km. At a depth of 115 km it reaches its

slowest seismic velocity of 4.2 km/s. Below 115 km the seismic velocity increases,

converges with the starting model at 200 km and 4.6 km/s, and continues to increase with

depth.

The inversion profile between MILR and QSPA lies within the EA craton (Fig.

15) and displays a much faster velocity structure compared to those of the Ross Sea (Fig.

27

14). The MILR-QSPA path along the TAM has a surface Vs of 2.7 km/s, which is the

slowest initial Vs analyzed. The velocity profile remains slower than the AK135-β model

to a depth of 50 km where it increases to 4.5 km/s. The velocity structure continues to

increases to 4.8 km/s at ~83 km before the Vs begins to decrease. At a depth of 125 km

the seismic velocity reaches a minimum of 4.7 km/s before increasing once again. The

MILR-QSPA profile reaches a maximum Vs of 4.9 km/s at a depth of 235 km.

28

CHAPTER VI

DISCUSSION

The area covered by this research encompasses three distinct geologic domains:

cratonic, orogenic, and extensional. Results from this study reveal that each domain

displays a distinct mantle seismic velocity profile. While the velocity structure of the

craton area is consistent with the seismic velocity structure of other cratons (Fig. 16), the

interpretation of the velocity structure beneath the Ross Sea (extensional domain) and the

TAM (orogenic domain) require additional explanation.

Extension, Craton, and Orogeny Seismic Velocity Structures

There are five seismic velocity profiles that stretch across the Ross Sea portion of

the WARS and they all display very similar slow seismic velocities. These slow seismic

velocities are consistent with documented seismic velocities in the VLB (Bannister et al.,

2000; Lawrence et al., 2006a,b; Watson et al., 2006). The similar seismic velocities are

interesting, as the Terror Rift should be expected to have much lower seismic velocities

due to the presence of active volcanics at either end of the rift. Laboratory and analytical

results have shown that melt produces very strong, slow seismic anomalies (Hammond

and Humphreys, 2000; Takei, 2002; Faul et al., 2004). The similarity between the

modestly slow anomalies in this study and the VLB are perplexing.

In comparison to several other rifts (Fig. 16), the mantle seismic velocity structure

of the Ross Sea region of the WARS is much faster than the East African rift

29

Figure 16. Inversion profile comparison for cratons and rifts. The top right graph plots seismic velocities from interstation paths that cross the Ross Sea in red (MPAT-LONW, LONW-MPAT, MPAT-MILR1, MPAT-MILR2, LONW-SIPL) plotted against seismic velocities from the East African rift (Weerartne et al., 2003), Rio Grand rift (West et al., 2004), and the Lau Back Arch Basin (Xu and Wiens, 1997). The top left graph plots the interstation path located on the craton against the East Antarctica (EA) craton results from Lawrence et al., (2006a), Tanzania craton (Weerartne et al., 2003), Siberia craton (Weerarnte et al., 2003), and Kaapvaal craton (Freybourger et al., 2001). The bottom right plots the QSPA-MPAT profile, which samples both the craton and rift, against several rifts and cratons. The bottom left plots the SURP-LONW interstation path in the orogenic domain with the other craton profiles.

30

(Weerartne et al., 2003) and the Lau Basin (Xu and Wiens, 1997). Both the East African

rift and Lau Basin are proposed locations of mantle plumes. Such slow seismic velocities

are often inferred to be warm material convecting upward from deep within the mantle.

The seismic profile of the Ross Sea most closely resembles the Rio Grande rift (West et

al., 2004). The Rio Grande rift is currently active, has high heat flow and is interpreted to

have a warm upper mantle. The WARS also has high heat flow and a warm upper mantle

has been proposed, but the WARS has been tectonically inactive for approximately 30

Ma (Cande et al., 2000). The factors that contribute to slow seismic velocities, high heat

flow and active volcanics in the Ross Sea region are still debated.

The interstation path MILR-QSPA is located on the EA craton and was included

in the study of the Ross Sea to test the validity of the methods used to determine seismic

velocity profiles in WA. The craton results display faster than average seismic velocities,

which are similar to the Tanzania (Weeraratne et al., 2003), Siberia (Weeraratne et al.,

2003), and Kaapvaal cratons (Freybourger et al., 2001), as well as EA craton results from

Lawrence et al. (2006a; Fig. 16). The elevated seismic velocity at depths greater than

150 km suggest the existence of a cold geochemically depleted continental keel

commonly associated with ancient cratons (Jordan, 1979; Forte and Perry, 2000;

Deschamps et al., 2002; Godey et al., 2004).

The interstation path QSPA-MPAT travels from the craton through the orogenic

TAM and across the rift system sampling each domain. The corresponding inversion

profile displays both slower and faster than average seismic velocities (Fig. 16). The

slower seismic velocities that resemble the extensional domain seismic mantle structure

are seen in the upper 200 km, while the faster than average seismic velocities are below

31

200 km. The faster seismic velocities may be attributed to the continental keel below the

EA craton.

The SURP-LONW interstation path samples the orogony domain in the TAM,

traveling from the rift edge of the TAM to the craton edge of the TAM. The seismic

velocity is much slower than the cratons from ~50–170 km, after which it accelerates to

seismic velocities similar to the other craton signatures (Fig. 16). The low velocity zone

may be due to an upper mantle that is heated and insulated by a thick crust with heat-

producing elements.

32

CHAPTER VI

SUMMARY

An analysis of the upper mantle velocity structure in the Ross Sea region of the

WARS, from the arrival of Rayleigh waves produced by teleseismic events, show

consistently slow seismic velocities across the Ross Embayment from depths of

approximately 60 km to 200 km (Figs. 14 and 16). While the velocity structure of the

craton area is consistent with the seismic velocity structure of other cratons (Fig. 16), the

interpretation of the velocity structure beneath the Ross Sea (extensional domain) and the

TAM (orogenic domain) require additional explanation.

The seismic velocity determined in this study is similar to results of other surface

wave inversion studies in Antarctica, including the VLB/Terror Rift area. Similar results

in the VLB suggest the upper mantle of the VLB is not unique from the rest of the WARS

in the Ross Embayment despite extremely thinned crust and the presence of active

volcanics.

The low elevation in WA suggests a dense lithosphere, which is associated with

low temperatures that would transmit faster seismic velocities than what is documented.

The slow seismic structure in the Ross Embayment is a result of higher than average

temperatures and/or variations in chemical compositions. The slow, rift like signature of

the WARS is consistent with a warm upper mantle and active volcanics, but contradicts

the tectonic inactivity of the rift system since the Paleogene.

33

REFERENCES

Bannister, S., Snieder R. K., and Passier, M. L., 2000, Shear-wave velocity under the Transantarctic Mountains and Terror Rift from surface wave inversion: Geophysical Research Letters, v. 27, n. 2, p. 281-284.

Behrendt, J. C., 1999, Crustal and lithospheric struture of the West Antarctic Rift System

from geophysical investigations – a review: Global and Planetary Change, v. 23, p. 25-44.

Behrendt, J. C., LaMasurier, W. E., and Cooper, A. K., 1992, The West Antarctic rift

system-a propagating rift “captured” by a mantle plume, in Kaminuma, K., and Yoshida, Y., eds., Recent Progress in Antarctic Earth Science: Terra Science, Tokyo, p. 315-322.

Blackman, D. K., Von Herzen, R. P., and Lawver, L. A., 1987, Heat flow and tectonics in

the western Ross Sea, Antarctica in Davey, F. J., ed., The Antarctic Continental Margin: Geology and Geophysics of the Western Ross Sea: Circum-Pacific Council for Energy and Natural Resources, Houston, TX, p. 179-189.

Boger, S. D., Wilson, C. J. L., Fanning, C. M., 2001, Early Paleozoic tectonism within

the East Antarctic craton: The final suture between east and west Gondwana, Geology, v. 29, n. 5, p. 463-466.

Bradshaw, J. D., 1989, Cretaceous Geotectonic Patterns in the New Zealand Region:

Tectonics, v. 8, n. 4, p. 803-820. Busetti, M., Spadini, G., van der Waternen, F. M., Cloetingh, S.A.P.L., and Zanolla, C.,

1999, Kinematic modeling of the West Antarctic rift system in the Ross Sea, Antarctica: Tectonophysics, v. 301, n. 1-2, p. 61-74.

Cande, S. C., Stocke, J. M., Muller, R. D., and Ishlhara, T., 2000, Cenezoic motion

between East and West Antarctica: Nature, v. 404, p. 145-150. Dalziel, I. W. D., 1991, Pacific margin of Laurentia and East Antarctica – Australia as a

conjugate rift pair: Evidence and implications for an Eocambrian supercontinent: Geology, v. 19, p. 598-601.

Danesi, S., and Morelli, A., 2000, Group velocity of Rayleigh waves in the Antarctic

region: Physics of the Earth and Planetary Interiors, v. 122, p. 55-66. Danesi, S., and Morelli, A., 2001, Structure of the upper mantle under the Antarctic Plate

from surface wave tomography: Geophysical Research Letters, v. 28, n. 23, p. 4395-4398.

34

Deschamps, F., Trampert, J., and Snieder, R., 2002, Anomalies of temperature and iron in

the uppermost mantle inferred from gravity data and tomographic models: Physics of the Earth and Planetary Interiors, v. 129, p. 245-264.

Donnellan, A., and Luyendyk, B. P., 2004, GPS evidence for a coherent Antarctic plate

and for post glacial rebound in Marie Byrd Land: Global and Planetary Change, v. 42, p. 305-311.

Faul, U. H., Fitz Gerald, J. D., and Jackson, I., 2004, Shear wave attenuation and

dispersion in melt-bearing olivine polycrystals: 2. Microstructural interpretation and seismological implications: Journal of Geophysical Research, v. 109.

Finn, C. A., Muller, R. D., and Panter, K. S., 2005, A Cenezoic diffuse alkaline magmatic

province (DAMP) in the southwest Pacific without rift or plume orogen: Geochemistry Geophysics Geosystems, v. 6, n. 1.

Fitzgerald, P. G., Sandiford, M., Barret, P. J., and Gleadow, J. W., 1986, Asymmetric

extension associated with uplift and subsidence in the Transantarctic Mountains and Ross Embayment: Earth and Planetary Science Letters, v. 81, p. 67-78.

Fitzsimons, I. C. W., 2000, Grenville-age basement provinces in East Antarctica:

Evidence for three separate collisional orogenies: Geology, v. 28, n. 10, p. 879-882. Forte, A. M., and Perry, A. C., 2000, Seismic-geodynamic evidence for a chemically

depleted continental tectosphere: Science, v. 290, p. 1940-1944. Freybourger, M., Gaherty, J. B., Jordan, T. H., and the Kaapvaal Seismic Group, 2001,

Structure of the Kaapvaal craton from surface waves: Geophysical Research Letters, v. 28, p. 2489-2492.

Godey, S., Deschamps, F., Trampert, J., and Sneider, R., 2004, Thermal and

compositional anomalies beneath the North American continent: Journal of Geophysical Research, v. 109.

Goldstein, P., Dodge, D., Firpo, M., and Minner, L., 2003, SAC2000: Signal processing

and analysis tools for seismologists and engineers in Lee, W. H. K., Kanamori, H., Jennings, P. C., and Kisslinger, C., eds., The International Association of Seismology and Physics of the Earth’s Interior International Handbook of Earthquake and Engineering Seismology, Academic Press, London.

35

Goodge, J. W., Fanning, C. M., Bennett, V. C., 1991, U-Pb evidence of ~1.7 Ga crustal tectonism during the Nimrod Orogeny in the Transantarctic Mountains, Antarctica: implications for Proterozoic plate reconstructions: Precambrian Research, v. 112, p. 261-288.

Gunn, G. M., and Warren, G., 1962, Geology of Victoria Land between the Mawson and

Mullock Glaciers, Antarctica: New Zealand Geological Survey Bulletin, 71, 157. Hammond, W. C., and Humphreys, E. D., 2000, Upper mantle seismic wave velocity:

Effects of realistic partial melt geometries: Journal of Geophysical Research, v. 105, p. 10,975-10,986.

Herrmann, R. B., and Ammon, C. J., 2004, surf96 from Computer Programs in

Seismology, surface waves, receiver functions and crustal structure, version 3.30, Department of Earth and Atmospheric Sciences, Saint Louis University.

Jordan, T. H., 1979, Mineralogies, densities and seismic velocities of garnet, Lherzolites

and their geophysical implications in Boyd, F. R., and Meyer, H. O. A., eds., The Mantle Sample: Inclusions in Kimberlites and Other Volcanics – Proceedings of the Second International Kimberlite Conference: American Geophysical Union, Washington D. C., p. 1-14.

Kennett, B. L. N., Engdahl, E. R., and Buland, R., 1995, Constraints on seismic velocities

in the Earth from travel times: Geophysical Journal International, v. 122, n. 1, p. 108-124.

Lawrence, J. F., Wiens, D. A., Nyblade, A. A., Anandakrishnan, S., Shore, P. J., and

Voigt, D., 2006a, Rayleigh wave phase velocity analysis of the Ross Sea, Transantarctic Mountains, and East Antarctica from a temporary seismograph array: Journal of Geophysical Research, v. 111.

Lawrence, J. F., Wiens, D. A., Nyblade, A. A., Anandakrishnan, S., Shore, P. J, and

Voigt, D., 2006b, Upper mantle thermal variations beneath Transantarctic Mountains inferred from teleseismic S-wave attenuation: Geophysical Research Letters, v. 33.

Lawrence, J. F., Wiens, D. A., Nyblade, A. A., Anandakrishnan, S., Shore, P. J., and

Voigt, D., 2006c, Crust and upper mantle structure of Transantarctic Mountains and surrounding regions from receiver functions, surface waves, and gravity: Implications for uplift models: Geochemistry Geophysics Geosystems, v. 7, n. 10.

LeMasurier, W. E., 1990, Late Cenozoic volcanism on the Antarctic plate: An overview

in LeMasurier, W. E., and Thomson, J. W., eds., Volcanoes of the Antarctic Plate

36

and Southern Oceans, American Geophysical Union Antarctic Research Series, v. 48, p. 1-17.

Luyendyk, B. P., 1995, Hypothesis for Cretaceous rifting in East Gondwana caused by

subducted slab capture: Geology, v. 23, p. 373-376. Lythe, M.B., Vaughan, D.G. and the BEDMAP Consortium. 2000. BEDMAP - bed

topography of the Antarctic. 1:10,000,000 scale map. BAS (Misc) 9. Cambridge, British Antarctic Survey.

Montagner, J. P., and Kennett, B. L. N., 1995, How to reconcile body-wave and normal-

mode reference Earth models: Geophysical Journal International, v. 125, n. 1, p. 229-248.

Morelli, A., and Danesi, S., 2004, Seismological imaging of the Antarctic continental

lithosphere: a review: Global and Planetary Change, v. 42, 155-165. Mussett, A.E. and Khan, M. A., 2000, An Introduction to Geological Geophysics,

Cambridge, Cambridge University Press, p. 144-162. Park Seismics LLC., 2007, Surface wave illustration: A website for surface wave

surveying: [http://www.parkseismic.com/SurfaceWaveSurvey.html]. Ritzwoller, M. H., Shapiro, N. M., Levshin, A. L., and Leahy, G. M., 2001, Crustal and

upper mantle structure beneath Antarctica and surrounding oceans: Journal of Geophysical Research, v. 106, n. 12, p. 30645-30670.

Sieminski, A., Debayle, E., and Leveque, J., 2003, Seismic evidence for deep low-

velocity anomalies in the tranzition zone beneath West Antarctica: Earth and Planetary Science Letters, v. 216, p. 645-661.

Stein, S. and Wysession, M., 2003, An introduction to seismology, earthquakes, and

Earth structure: Malden, Massachusetts, Blackwell, p. 29-412. Storey, B. C., Alabaster, T., Macdonald, D. I. M., Miller, I. L., Pankhurst, R. J., and

Dalziel, I. W. D., 1992, Upper Proterozoic rift-related rocks in the Pensacola Mountains: Precursors to supercontinent break up?: Tectonics, v. 11, p. 1392–1405.

Storey, B. C., Macdonald, D. I. M., Dalziel, I. W. D., Isbell, J. L., and Miller, I. L., 1996,

Early Paleozoic sedimentation, magmatism, and deformation in the Pensacola Mountains, Antarctica: The significance of the Ross orogeny: Geological Society of America Bulletin, v. 108, p. 685-707.

37

Stump, E., Smit, J. H., and Self, S., 1986, Timing of events during the late Proterozoic Beardmore Orogeny, Antarctica: Geological evidence from the La Gorce Mountains: Geological Society of America Bulletin, v. 97, n. 8, p. 953-965.

Svaldi, J., 2010, Crustal thickness of the Marie Byrd Land dome and its implications

[B.S. Thesis]: Ellensburg, Central Washington University, 29 p. Takei, Y., 2002, Effect of pore geometry on Vp/Vs: from equilibrium geometry to crack:

Journal of Geophysical Research, v. 107, n. 2043. ten Brink, U. S., Bannister, S., Beaudoin, B. C., and Stern, T. A., 1993, Geophysic

Investigation on the Tectonic Boundary Between East and West Antarctica: Science, v. 261, n. 5117, p. 45-50.

Tonarini, S., Rocchi, S., Armienti, P., Innocenti, F., 1997, Constraints on timing of Ross

Sea rifting inferred from Cainozoic intrusions from northern Victoria Land, Antarctica, in C.A. Ricci, C. A., ed., International Symposium on Antarctic Earth Sciences, vol. 7: Terra Antarctica Publication, Siena, p. 511–521.

Trey, H., Cooper, A. K., Pellis, G., Vedova, B. Della, Cochrane, G. R., Brancolini, G.,

and Makris, J., 1999, Transect across the West Antarctic rift system in the Ross Sea, Antarctica: Tectonophysics, v. 301, n. 1-2, p. 61-74.

Watson, T., Nyblade, A., Wiens, D. A., Anandakrishnan, S., Benoit, M., Shore, P. J.,

Voigt, D., and VanDecar, J., 2006, P and S velocity structure of the upper mantle beneath the Transantarctic Mountains, East Antarctic craton, and Ross Sea from travel time tomography: Geochemistry Geophysics Geosystems, v. 7, n. 7.

Weaver, S. D., Storey, B. C., Pankhurst, R. J., Mukasa, S. B., DiVenere, V. J., and

Bradshaw, J. D., 1994, Antarctic-New Zealand rifting and Marie Byrd Land lithospheric magmatism linked to ridge subduction and mantle plume activity: Geology, v. 22, n. 19, p. 811-814.

Weeraratne, D. S., Forsyth, D. W., Fischer, K. M., and Nyblade, A. A., 2003, Evidence

for an upper mantle plume beneath the Tanzanian craton from Rayleigh wave tomography: Journal of Geophysical Research, v. 108(B9), p. 2427-2441.

West, M., Ni, J., Baldridge, W. S., Wilson, D., Aster, R., Gao, W., and Grand, S., 2004,

Crust and upper mantle shear wave structure of the southwest United States: implications for rifting and support for high elevation: Journal of Geophysical Research, v. 109, B03309.

Winberry, J. P., and Anandakrishnan, S., 2004, Crustal structure of the West Antarctic

rift system and Marie Byrd Land hotspot: Geology, v. 32, n. 11, p. 977-980.

38

Xu, Y., and Wiens, D. A., 1997, Upper mantle structure of the southwest Pacific from

regional waveform inversion: Journal of Geophysical Research, v. 102, p. 27,439-27,451.

39

APPENDIXES

Appendix A

Starting Model

TABLE A1. AK135-β AVERAGE GLOBAL SEISMIC VELOCITY MODEL

Depth (km)

Thickness (km)

Vp (km/s)

Vs (km/s)

Density (g/cm3)

Qp Qs

2 2 6.2000 3.6000 2.7500 849 600 4 2 6.2000 3.6000 2.7500 849 600 6 2 6.2000 3.6000 2.7500 849 600 8 2 6.2000 3.6000 2.7500 849 600 10 2 6.6880 3.8000 2.9101 849 600 12 2 6.6880 3.8000 2.9101 849 600 14 2 7.0400 4.0000 3.0228 849 600 16 2 7.0400 4.0000 3.0228 849 600 18 2 7.3920 4.2000 3.1354 849 600 20 2 7.3920 4.2000 3.1354 849 600 22 2 7.7440 4.4000 3.2481 849 600 24 2 7.7440 4.4000 3.2481 849 600 26 2 8.0398 4.4844 3.3248 849 600 28 2 8.0398 4.4844 3.3248 849 600 30 2 8.0398 4.4844 3.3248 849 600 32 2 8.0398 4.4844 3.3248 849 600 34 2 8.0398 4.4844 3.3248 849 600 36 2 8.0398 4.4844 3.3248 849 600 38 2 8.0398 4.4844 3.3248 849 600 40 2 8.0398 4.4844 3.3248 849 600 42 2 8.0398 4.4844 3.3248 849 600 44 2 8.0398 4.4844 3.3248 849 600 46 2 8.0398 4.4844 3.3248 849 600 48 2 8.0398 4.4844 3.3248 849 600 50 2 8.0398 4.4844 3.3248 849 600 55 5 8.0398 4.4844 3.3248 849 600 60 5 8.0404 4.4856 3.3251 848 600 65 5 8.0409 4.4868 3.3253 848 600 70 5 8.0415 4.4880 3.3255 848 600 75 5 8.0421 4.4891 3.3258 848 600 80 5 8.0558 4.4886 3.3260 847 600 85 5 8.0433 4.4917 3.3262 848 600 90 5 8.0439 4.4929 3.3265 848 600 95 5 8.0445 4.4922 3.3267 848 600

40

TABLE B1 (continued)

Depth (km)

Thickness (km)

Vp (km/s)

Vs (km/s)

Density (g/cm3)

Qp Qs

100 5 8.0450 4.4952 3.3269 848 600 110 10 8.0525 4.4970 3.3273 847 600 120 10 8.0471 4.4984 3.3278 847 600 130 10 8.0642 4.5012 3.4206 116 76.1 170 10 8.1752 4.5096 3.3714 118 76.6 180 10 8.2030 4.5110 3.3607 120 77.2 190 10 8.2306 4.5131 3.3503 121 77.8 200 10 8.2589 4.5152 3.3399 123 78.4 210 10 8.2868 4.5172 3.3295 124 79.1 220 10 8.3189 4.5275 3.3285 211 134 230 10 8.3552 4.5457 3.3369 212 135 240 10 8.3913 4.5639 3.3453 213 135 250 10 8.4278 4.5821 3.3537 214 136

Note: Depth is of the bottom of the layer; Thickness is the layer thickness; Vp is the compressional velocity; Vs is the shear velocity; Qp is the compressional attenuation; and Qs is the shear attenuation.

41

Appendix B

Inversion Parameters

TABLE B1. LAYER WEIGHTING FOR INVERSION PROCESS

Depth (km) Layer Freedom

0-22 1-11 0.0 24 12 0.1 26 13 0.3 28 14 0.5 30 15 0.7 32 16 0.9

34-200 17-45 1.0 210 46 0.9 220 47 0.9 230 48 0.8 240 49 0.8 250 50 0.7

Note: Layer weighting for inversions. Freedom is the amount the data can vary from the model from 0.0 - 1.0; 0.0 fixes the data to the model and 1.0 is complete freedom.

42

Appendix C

Inversion Profiles

Figure C1. Inversion profile for LONW-MPAT event 10/15/2009. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model.

43

Figure C2. Inversion profile for MPAT-LONW event 7/7/2009. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model.

44

Figure C3. Inversion profile for MPAT-MILR1 event 7/7/2009. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model.

45

Figure C4. Inversion profile for MPAT-MILR2 event 10/12/2009. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model.

46

Figure C5. Inversion profile for LONW-SIPL event 11/22/2008. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model.

47

Figure C6. Inversion profile for SURP-LONW event 12/13/2008. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model.

48

Figure C7. Inversion profile QSPA-MPAT event 6/2/2009. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model.

49

Figure C8. Inversion profile for MILR-QSPA event 8/30/2008. The plot on the left has Vs (km/s) along horizontal axis and depth (km) along vertical axis. The red in the profile depicts the data while the blue dashed line shows the AK135-β model. On the right, phase velocity (km/s) along the vertical axis as a function of period (s) along the horizontal axis, show seismic velocity between the two stations. Black circles represent the data and red line is the modeled dispersion showing fit between the data and AK135-β model.