ABSTRACT

Title of Thesis:

Thesis Directed By:

The lunar subsurface is a primary science target for future missions to the Moon

and serves as a potential host location for resources such as water ice, void spaces

for astronaut shelter, and important ore bodies for in-situ manufacturing and

building materials (Horz, 1985; Coombs and Hawke, 1992; Wendell, 2017). Here

we conducted laboratory experiments to investigate how the seismic signature

obtained at or near the lunar surface is related to subsurface material properties

and structures. A range of analog basaltic samples collected from the San

Francisco Volcanic Field (SFVF) in Arizona, Kilbourne Hole (KH) in Southern

New Mexico, and Lava Beds National Monument (LBNM) in Northern California

are collected and their geophysical properties are measured. The relationships

between the seismic wave velocity and porosity of basaltic rocks from different

locales are obtained and the measured velocities are compared to local seismic

surveys of the sample locale. Using laboratory techniques, the SFVF basalts



P-wave and S-wave velocities range from ~5 km/s to ~6 km/s +/- 0.1 km/s and

~2.1 km/s to ~3 km/s +/- 0.1 km/s, respectively. For the KH basalts P-wave and

S-wave velocities range from ~3.5 km/s to ~4.4 km/s +/- 0.1 km/s and ~1.8 km/s

to ~2.3 km/s +/- 0.1 km/s, respectively. For LBNM basalts P-wave and S-wave

velocities range from ~3.1 km/s to ~6 km/s +/- 0.1 km/s and ~1.2 km/s to ~2.4

km/s +/- 0.1 km/s, respectively. A relationship between the porosity and velocity

of basalts from these three locations was determined. Using the field seismic data

for each study site and the relationship derived from the laboratory seismic data

and porosity, the amount of possible large-scale fracturing is derived. Based on

these experimental measurements, we estimate the large-scale fractures and voids

are ~84 vol% at the SFVF, ~42 vol% at the KH, and ~39 vol% at the LBNM.

These results provide quantitative constraints for subsurface exploration in future

lunar surface missions.



EXPERIMENTAL CONSTRAINTS ON PHYSICAL PROPERTIES OF VOLCANIC
ROCKS WITH IMPLICATIONS FOR LUNAR EXPLORATION

by

Casey M Braccia

Thesis submitted to the Faculty of the Graduate School of the
University of Maryland, College Park, in partial fulfillment

of the requirements for the degree of
Master of Science

2023

Advisory Committee:
Professor Wenlu Zhu, co-advisor (University of Maryland)
Professor Nicholas Schmerr, co-advisor (University of Maryland)
Professor Laurent Montesi (University of Maryland)



© Copyright by
Casey M Braccia

2023



Table of Contents

Table of Contents………………………………………………………………….ii
List of Tables………………………………….……...………………………..…iii
List of Figures……………………………………………………………………..v
Chapter 1: Introduction

1.1: Formation of the Moon.…………..…………………………….…….1
1.2: Lunar Crustal Composition and Structures…………………………...3
1.3: What we know from the Apollo Missions…………………………....6
1.4: GEODES………………………………………………………..…...11
1.5: Research Goals……………………………………………….……..17

Chapter 2: Study Sites
2.1: History of Selected Study Sites………………….……………....….20
2.2: Sample Material and Collection Process.….…………………….….25
2.3: Major Variations Due to Sample Locale…….……………………....34

Chapter 3: Technical Methodology
3.1: Velocity of Basalt………….…………………………………..…….36
3.2: Experimental Setup………………………………………...………..37
3.3: Velocity-Porosity Relationship…………..……………………….…42
3.4: Velocity-Effective Moduli Relationship……..…………………...…45

Chapter 4: Results
4.1: Velocity Data………………………………………………….…….50
4.2: Mechanical Data………………………………………………….…65

Chapter 5: Discussion
5.1: Derived Velocity-Porosity Relationship………...………………..…76
5.2: Clay Content………...........................................................................86
5.3: Extrapolation of Apollo Velocity Data…………………………..….90
5.4: Field-scale Porosity Relationship………………………………...…93

Chapter 6: Concluding Remarks………………………………………………....97
Appendix………………………………………………………………………....99
Bibliography……………………………………………………………………119

II



List of Tables

Table 1. Elastic and seismic properties of samples collected during the Apollo
missions. Data are from Wang et al., 1971; Todd et al., 1972; Chung, 1973;
Mizutani and Newbigging, 1973.

Table 2. Sample description for all samples from the San Francisco Volcanic
Field.

Table 3. Sample description for all samples from Kilbourne Hole in New Mexico.

Table 4. Sample description for all samples from Lava Beds National Monument.

Table 5. Field data collected for the three study sites

Table 6. Sample pore volume, porosity from the saturation method and derived
from CT scans, dry and wet bulk density, and average solid density for all samples
from The San Francisco Volcanic Field.

Table 7. Sample pore volume, porosity from the saturation method and derived
from CT scans, dry and wet bulk density, and average solid density for all samples
from Kilbourne Hole in New Mexico.

Table 8. Sample pore volume, porosity from the saturation method and derived
from CT scans, dry and wet bulk density, and average solid density for all samples
from Lava Beds National Monument.

Table 9. Relationships between E, , K, , , and . The first column are inputν µ 𝑉
𝑝

𝑉
𝑠

values and the row below are output values.

Table 10. Velocity data for all basalts at The San Francisco Volcanic Field.

Table 11. Velocity data for all basalts at Kilbourne Hole.

Table 12. Velocity data for all basalts at Lava Beds National Monument.

Table 13. Linear regressions for P-wave velocities for all three study sites at 20
MPa, 40 MPa, 60 MPa, and an average velocity.

Table 14. Linear regressions for S-wave velocities for all three study sites at 20
MPa, 40 MPa, 60 MPa, and an average velocity.

Table 15. Measured elastic moduli values and calculated velocities for basalts
from LBN

III



Table 16. Calculated elastic moduli values and calculated corresponding
velocities for basalts from LBNM.

Table 17. Clay content for selected samples from the three study sites.

IV



List of Figures

Figure 1. Schematic representation of the lunar crustal composition at depth
described in Heiken et al., 1991. The four main sections can be described as a
fine-grained regolith layer, a large-scale ejecta layer, a layer of structurally
disturbed large crustal blocks, a layer of fractured crust, and finally in place
bedrock.

Figure 2. Samples from the Apollo missions. (A) Apollo 14 sample 14310 (B)
Apollo 15 sample 15058 (C) Apollo 15 sample 15065. Images from NASA
Apollo sample catalog.

Figure 3. Sample 15016, also known as the “Seatbelt Rock”. A vesicular olivine
basalt collected during the Apollo 15 mission .From NASA Apollo sample
catalog.

Figure 4. (A) Image taken by SELENE in the Marius Hills region of a suggested
65 m diameter lava tube. (B) Enlarged image of Marius Hills Hole. (C-F) Angled
imagery looking into Marius Hills Hole; a possible skylight into the lava
tube.Taken from Haruyama et al., 2009.

Figure 5. Examples of an analog field study conducted in June 2022 by the
NASA GEODES team. Shown are members of the team measuring fractures
inside of Skull Cave at Lava Beds National Monument. People for scale. Photo
taken by Molly Wasser from the NASA GEODES team. Credit: Molly Wasser,
2022.

Figure 6. Different techniques are used to measure elastic wave velocity at
different scales. From the left to the right: Aerial view of SP Crater and it’s
associated lava flow is shown on a kilometer scale, Google Earth 2023. Seismic
surveys can be conducted to investigate the subsurface on a meter scale; Credit
Molly Wasser, 2022. Experimental rock mechanics can be conducted to
investigate the subsurface on a scale of centimeters. Three-dimensional
tomographic imaging can be conducted to investigate the subsurface on a scale of
microns.

Figure 7.Map of Arizona. The San Francisco volcanic field is highlighted in
light gray (Moore et al., 1976).

Figure 8. Geologic map of Kilbourne Hole in New Mexico (Reiche, 1940). This
crater consists of multiple layers of volcanic material indicated by the legend.

Figure 9.Map of Lava Beds National Monument (National Park Services, 2023).

V

https://agupubs.onlinelibrary.wiley.com/authored-by/Haruyama/Junichi


Figure 10. Selected quarry walls located in the San Francisco Volcanic Field
where sampling was accessible.

Figure 11. The six selected samples, varying in porosity, used for the velocity
experiments.

Figure 12. Selected sampling area located in the southern rim of Kilbourne Hole.
The red arrows indicate various geologic layers including, from top to bottom,
unconsolidated volcanic ash and volcanic surge beds, a basaltic lens, and
sedimentary rock.

Figure 13. The four selected samples, varying in porosity, used for the velocity
experiments.

Figure 14. Aerial view of the entrance of Skull Cave and surrounding area used
during sampling. Sampling site locations for all LBNM samples are indicated by
marker.

Figure 15. Basaltic samples varying in porosity were selected from each sampling
site providing nine samples in total. Three from inside Skull Cave (top), and three
from the surface: lobe 1 (middle), and three from the surface: lobe 2 (bottom)
used in the velocity experiments.

Figure 16. The sample assembly for deformation experiments consists of two
ultrasonic transducers on either end of a basalt core and jacket to keep the sample
dry.

Figure 17. Data from Han and Morgan, 1986. Linear fits between velocity and
porosity and clay content. Top left: Vp relative to porosity, top right: Vp relative
to clay content, bottom left: Vs relative to porosity, and bottom right: Vs relative
to clay content.

Figure 18. Results from Borgomano, 2019. The uncracked glass is in dark blue
while the cracked saturated glass is in light blue and the cracked dry glass is in
green. Notice the lower velocities for the cracked dry glass at lower pressures.

Figure 19. P-wave velocity data (top) and S-wave velocity data (bottom) for the
six basalt samples from The San Francisco Volcanic Field.

Figure 20. P-wave velocity data (top) and S-wave velocity data (bottom) for the
four basalt samples from Kilbourne Hole in New Mexico.

Figure 21. P-wave velocity data (top) and S-wave velocity data (bottom) for the
eight basalt samples from Lava Beds National Monument.

VI



Figure 22. P-wave velocity data (top) and S-wave velocity data (bottom) for
basalts from all three study sites.

Figure 23. P-wave velocity plotted against porosity of all samples from the three
study sites. Velocities were collected at 20 MPa confining pressure (top), 60 MPa
confining pressure (middle), and the mean velocity for each sample (bottom).

Figure 24. S-wave velocity plotted against porosity of all samples from the three
study sites. Velocities were collected at 20 MPa confining pressure (top), 60 MPa
confining pressure (middle), and the mean velocity for each sample (bottom).

Figure 25. Stress-strain curves for the three basalt sample from Lava Beds
National Monument inside Skull Cave. The slope of the elastic portion is the
Young’s modulus of the sample.

Figure 26. Stress-strain curves for the three basalt sample from Lava Beds
National Monument at lobe 1. The slope of the elastic portion is the Young’s
modulus of the sample.

Figure 27. Stress-strain curves for the three basalt sample from Lava Beds
National Monument at lobe 2. The slope of the elastic portion is the Young’s
modulus of the sample.

Figure 28. Stress-strain curves for the three basalt sample from Lava Beds
National Monument inside Skull Cave. The slope of the elastic portion is the bulk
modulus of the sample.

Figure 29. Stress-strain curves for the three basalt sample from Lava Beds
National Monument at lobe 1. The slope of the elastic portion is the bulk modulus
of the sample.

Figure 30. Stress-strain curves for the three basalt sample from Lava Beds
National Monument at lobe 2. The slope of the elastic portion is the bulk modulus
of the sample.

Figure 31. P-wave velocities measured using the transducer assembly (black),
derived from the raw mechanical data (red), and derived from the corrected
mechanical data (blue square- Poisson’s ratio from Vp/Vs ratio and blue circle-
from average Poisson’s ratio).

Figure 32. S-wave velocities measured using the transducer assembly (black),
derived from the raw mechanical data (red), and derived from the corrected
mechanical data (blue square- Poisson’s ratio from Vp/Vs ratio and blue circle-
from average Poisson’s ratio).

VII



Figure 33. P-wave velocities plotted against porosity (left). S-wave velocities
plotted against porosity (right). Both figures use similar y-axis scales for
comparison.

Figure 34. P-wave velocities plotted against porosity for SFVF. The saturation
method is used to determine porosity for the plots on the top and the CT scans are
used for the plots on the bottom.

Figure 35. P-wave velocities plotted against porosity for KH. The saturation
method is used to determine porosity.

Figure 36. P-wave velocities plotted against porosity for LBNM. The saturation
method is used to determine porosity for the plots on the top and the CT scans are
used for the plots on the bottom.

Figure 37. Spherical pore (left) and long, narrow crack (right). The pressure
applied to each pore is the same, but the stress is higher on the crack because the
area is smaller.

Figure 38. Variation in pore shape and diameter among the four tomographic
scans for basalts from the SFVF. Yellow circles mark spherical pores and red
circles mark narrow elongated cracks in order to show the two end members of
pore shape. Pores in between spherical and narrow are unmarked.

Figure 39. Variation in pore shape and diameter among the eight tomographic
scans for basalts from LBNM. The two end members of pore shape, spherical (in
yellow) and narrow elongated cracks (in red) marked. From top left to bottom
right, samples are in order of decreasing pressure dependence on velocity
measurements.

Figure 40. Density and porosity relationship for basalts at the three study sites.

Figure 41. Density of the solid portion of each basalt sample from all three study
sites plotted against their respective p-wave velocity.

Figure 42. Density of the solid portion of each basalt sample from all three study
sites plotted against their respective s-wave velocity.

Figure 43. Linear regression for p-wave velocity (top) and s-wave velocity
(bottom) for basalts in all three study sites. Extrapolation of these relationships
can be used to estimate the total porosity of the shallow lunar surface.

Figure 44. Field scale velocity plotted along with the laboratory velocity data for
all three study sites, specifically showing the difference between the field velocity
and laboratory velocity for the San Francisco Volcanic Field.

VIII



Figure 45. Field scale velocity data plotted along with the laboratory velocity data
for all three study sites, specifically showing the difference between the field
velocity and laboratory velocity for Kilbourne Hole.

Figure 46. Field scale velocity data plotted along with the laboratory velocity data
for all three study sites, specifically showing the difference between the field
velocity and laboratory velocity for Lava Beds National Monument.

IX



Acknowledgements

I would like to thank my advisor, Dr. Wenlu Zhu, for her guidance and

patience. I thank Dr. Laurent Montesi and Nicholas Schmerr for serving on the

thesis committee and helping me improve the presentation of the thesis. I would

also like to thank the following institutions and individuals for their contributions.

The University of Texas High-Resolution X-ray CT Facility, specifically Matt

Colbert, for conducting scans on selected basalt samples used in this study. The

Planetary Environments Laboratory at NASA’s Goddard Space Flight Center,

specifically Christine Knudson and Amy McAdam, for conducting XRD analysis

on selected basalt samples used in this study. This work was supported by the

NASA SSERVI GEODES grant 80NSSC19M0216.

X



Chapter 1: Introduction

1.1: Formation of the Moon

The most widely accepted theory for how our Moon was created is the

giant-impact theory (Hartmann and Davis, 1975; Benz et al., 1989). This theory

involves the collision of Earth with a Mars-sized planetary body, forming ejecta

material that orbited the Earth and eventually coalesced into the present-day

Moon. Some studies suggest that the collision occurred near the end of Earth’s

formation ~4.5 billion years ago (Urey, 1969; Schmitt, 1975; Canup and Asphaug,

2001).

Taylor, (1975) and Heiken et al., (1991) discuss the evolution of the Moon

after the initial impact. Due to the energy released from the collision, as the Moon

formed it was made entirely of molten material. As it cooled and minerals

crystallized out of the magma, heavier minerals such as olivine, pyroxene, and

ilmenite sank and lighter minerals such as plagioclase cooled to form the

anorthositic crust of the Moon. During this time, about 4 billion years ago, the

Moon was also being heavily bombarded by meteorites that produced the

highlands and impact basin topography we see today. The highlands are

composed of anorthositic ancient lunar rock, primarily made up of plagioclase

feldspar, and possibly materials left behind from impacts. Deeper crustal and

mantle material is hypothesized to be more mafic in composition (Taylor and

Bence, 1975). At the surface, the highlands are also composed of breccias and

melted rock fragments due to the intense meteorite bombardment as well as loose

1

https://www.sciencedirect.com/science/article/pii/S0016703714004463?casa_token=3ZZtFbXPhg0AAAAA:9EOQlDIhd28w8sPCFS1qXzgUQIxXia-ZSTVhRd6yZxK9AvWK2agCHO1svmNqSLLn_2KLlrRKPA#b0025


unconsolidated soil-like material called regolith. After this period of intense

bombardment, lava flowed through fractures in the lunar crust and cooled over a

large area of the Moon’s surface. This lava cooled and formed the mare basalts, an

iron-rich volcanic rock commonly found on Earth. Further studies provide insight

on the emplacement history of lunar mare basalts and find that lunar lavas are

consistent with the textures found in terrestrial pāhoehoe flows, indicating a

low-viscosity, effusive volcanic history (Gawronska et al., 2022).

The giant-impact theory provides an explanation for the similar

composition and minerals found on both the Earth and the Moon (Canup, 2012).

Much like the Earth, the interior structure of the Moon is composed of a crust,

mantle, and core, although details of the lunar interior, especially at great depths,

are not well constrained. Data from the GRAIL mission suggests that the upper

30-40 km of lunar crust is similar in composition to areas of the Earth’s upper

crust (Wieczorek et al., 2013). Thermodynamic modeling and geophysical

observations have been used to estimate the bulk composition of the Moon at

depths. Similar to the lunar crust, the lunar mantle is hypothesized to consist of

similar elements as Earth. The lunar mantle is likely exposed to the surface in

regions with large craters. Lunar orbital satellites have identified regions of these

deep craters to be composed of olivine-rich rocks, hypothesized to make up the

lunar mantle (Yamamoto et al., 2023). Lunar Laser Ranging (LLR) experiments

have been used for many years and can provide insight into the distance between

the Earth and Moon, lunar orbital predictions, and the precision of Newton’s

gravitational constant (Bender et al., 1973; Müller and Biskupek, 2007). Based on

2



density data from Lunar Laser Ranging, the lunar core is likely composed of iron

and nickel (Williams and Boggs, 2009). There is debate about if the lunar core is

an entirely viscous layer or split into a viscous outer core and solid inner core, but

the latter may be used to explain the evolution of the Moon’s magnetic field

(Weber, 2011; Briaud et al., 2023). Although data to understand the complexities

of the Moon’s composition is very limited, the bulk composition appears to be

similar to that of Earth.

1.2: Lunar Crustal Composition and Structures

The Moon’s crust can be characterized using remote sensing techniques

such as aerial imaging and spectrometry (Vondrak, 2010), ground-based field

work such as seismic surveys (Latham et al., 1972), and sampling characterized

through petrographic techniques such as sequential instrumental neutron

activation analysis (Wanke et al., 1973; Laul and Schmitt, 1973). Nevertheless,

these techniques have many limitations and ground-based surveys and sampling

are not extensive enough to confidently determine the entire composition of the

lunar crust, especially below the immediate surface.

With these limitations in mind, velocity and attenuation data collected

from the Apollo Passive Seismic Network can be used to broadly extrapolate the

overall structure of the lunar crust. Latham et al., (1972) used data from three of

the seismometers of the Apollo Passive Seismic Network at the Descartes landing

site in order to work out the structure of the Moon’s interior up to about 100 km

depth. They determined that the Moon is composed of a layered crust that extends

to a depth of about 65 km; where a jump in velocity from 7 km/s to 9 km/s can be

3



observed, marking the boundary to the lunar mantle. Additional studies have

provided more context to Latham’s early work, exploring deeper into the

complexities of the lunar crustal structure using the Apollo seismic data. Due to

the long history of heavy bombardment on the Moon, the structure of the lunar

near-subsurface is composed of unconsolidated regolith, ejecta material, and

highly fractured bedrock (Wetherill, 1975; Besserer et al., 2014). Strong

heterogeneities in the seismic signature of the shallow lunar surface are also

indicative of this heavily fractured and unconsolidated material (Li et al., 2022).

Using seismic data and data from Heiken et al., (1991), the overall

structure of the lunar crust can be summarized from the following layers: (1) a

shallow section of surficial regolith composed of roughly 5-20 meters of loose,

fine-grained, porous ejecta material, (2) a 2 km section of coarse-grained ejecta

material and melt sheets, (3) an upper megaregolith section composed of roughly

10 km of deposited layers of brecciated material that has been transported from its

original depositional location due to subsurface movement, (3) a lower

megaregolith section composed of roughly 25 km of in place but still highly

fractured bedrock- decreasing in fracture density with depth, and finally (4)

unfractured bedrock beneath extending to the mantle (Figure 1).

4



5



1.3: What we know from the Apollo Missions

Our understanding of the lunar crust can come from a better understanding

of each of these four broad compositional layers; regolith, ejecta material,

megaregolith, and unfractured bedrock. Crewed missions to the Moon allowed for

the first time collection of intact samples of the lunar crust. Between the years of

1969 and 1972, the Apollo missions brought back to Earth ~840 pounds of rock

and soil totaling 2,196 samples. (Taylor, 1975). Every year, roughly 400 samples

are sent out to scientists and institutions in order to continue studying the lunar

surface (Stansbery, 2022). Currently, material properties of lunar crustal rock

come primarily from this limited set of samples collected during the Apollo

missions and meteorites discovered on the Earth’s surface (Warren, 2005).

Each Apollo mission focused on specific scientific questions that involved

sampling from various locations of the lunar surface (Taylor, 1975). Some of the

sampling conducted on each mission and images of samples is as follows (Figure

2):

Apollo 11’s main objective was a successful Moon landing. Samples from

Apollo 11 were mainly composed of basalts and breccias (Taylor, 1975). Apollo

12 sampled mostly basaltic rocks, and some highly fractionated rocks known as

KREEP which give insight into the lunar mantle composition (Taylor, 1975).

KREEP rocks are enriched in potassium (K), rare earth elements (REE), and

phosphorus (P) and are believed to have solidified during the early stages of the

Moon’s formation. Basalts from Apollo 11 and Apollo 12 formed ~500 million

years apart, indicating mare volcanism occurred over an extended period of time

6



(Levinson and Taylor, 1971). Apollo 14 focused on sampling from material

ejected by the meteorite impact that formed the Imbrium Basin. Samples included

many breccias and can be used to provide further insight into the types of

meteorite impacts necessary to form these composite materials (Taylor, 1975).

Many of the Apollo 14 samples are enriched in KREEP. Apollo 15 and 16

sampled a combination of breccias, basalts, and lunar highland rocks (Taylor,

1975). Apollo 17 sampled lunar soils composed of the three major lunar rock

types: basalt, anorthosite, and breccia as well as glass particles (Taylor, 1975).

Some of the Apollo 17 samples are believed to have formed in the Moon’s lower

crust. Since the Apollo missions, there have been many publications on the

physical properties of the samples collected (Wang et al., 1971; Todd et al., 1972;

Chung, 1973; Mizutani and Newbigging, 1973; Table 1).

7



8



9



Although these landings allowed for sampling that had never been

conducted before, given the vast size of the Moon and still unexplored areas, it is

far from representative of the lithologies or structures that exist on the Moon.

Even here on Earth with abundant sampling opportunities, there exists many

geologic questions. While some material property data has been successfully

collected and cataloged for the lunar crust, there is much more to discover,

especially at depths. The majority of the lunar crust, from ~25 km to 65 km depth,

consists of an intact, dense crustal material that is not easily sampled. Sample

15016, better known as the “Seatbelt Rock”, is a vesicular basalt collected during

the final stages of the Apollo 15. Sample 15016 is unique because it lacks the

abundant fracturing, consistent with near-surface bombardment, present in other

samples collected on the lunar surface. (Chappell and Green 1973; Jolliff and

Robinson, 2019; Figure 3). This sample provides some insight on the structures

possible in the lunar crust, but outside of this singular sample there has been no

sampling of vesicular basalts without fracturing. In addition, due to difficulties

sampling and transporting large rocks from the Moon back to Earth, lunar

samples on average are relatively small. Small-scale samples are less than ideal in

capturing large-scale structures and fractures. Because of this, there is a lack of

understanding of the overall structure of the lunar subsurface and how to interpret

geophysical data of the subsurface.

10



1.4: GEODES

This study is part of a NASA Solar System Exploration Research Virtual

Institute (SSERVI) project called NASA Geophysical Exploration of the

Dynamics and Evolution of the Solar System (GEODES). GEODES main science

targets include the surfaces of the Moon, near-Earth asteroids, and Mars, with this

study focusing primarily on lunar analogs. Research interests for GEODES are

composed of four primary sections: lava tubes and void spaces, ice deposits,

regolith, and magma-tectonic systems. To investigate these subsections, GEODES

conducts field expeditions here on Earth and uses geophysical techniques,

including but not limited, to seismic surveys, ground penetrating radar scans,

11



magnetometry, and LiDAR to characterize and further explore the subsurface

(Schmerr et al., 2020). Earth-based field expeditions will aid in refining the

techniques that will be used to study the lunar subsurface during future crewed

missions to the Moon.

Esmaeili et al., (2020), used ground penetrating radar to map the location

of lava tubes at LBNM. This remote sensing technique is capable of locating

subsurface features, such as lava tubes, because the interface between solid rock

and void space reflects radar pulses. This study was successful in locating the

ceilings and width of lava tubes, which is ground-truthed using light detection and

ranging (LiDAR) data of the tubes’ dimensions. Esmaeili et al., (2020) provides

evidence that ground penetrating radar may be a useful geophysical tool to use on

the lunar surface to locate and characterize subsurface structures. This study also

provides evidence showcasing how co-located geophysical surveys are useful in

ground-truthing and interpreting data, particularly on other planetary bodies.

One of the goals of future lunar missions is to sustain a human presence on

the surface of the Moon (Smith et al., 2020). Extended activities on the lunar

surface will require a protected shelter to house astronauts and equipment. Lunar

surface threats pose a real concern for future crewed missions to the Moon.

Because the lunar surface lacks a substantial atmosphere, the surface is exposed to

micrometeorite bombardment, cosmic rays, and solar radiation making for a harsh

environment (Hörz, 1985; Coombs and Hawke, 1992). The surface of the Moon

can also have extreme temperature fluctuations up to 280⁰C, making long-term

surface activities difficult.

12



One proposed solution is to use natural lunar caves in order to protect

astronauts from these surface threats (Hörz, 1985; Coombs and Hawke,

1992).These natural caves are likely formed from ancient lava flows that cool and

harden near the surface while allowing molten lava to drain below, forming a

hollow structure known as a lava tube (Greeley, 1971; Gornitz, 1973; Coombs and

Hawke, 1992). Although there is no direct evidence of the existence of lava tubes

on the lunar surface, there is plenty to suggest they are there. Sinuous rilles,

thought to be ancient lava flow channels, are a lunar feature discovered in the

early 1970’s that may be indicative of the presence of a lava tube beneath the

surface. These sinuous rilles are evident in aerial data and may be collapsed

sections of lunar lava tubes (Greeley 1971; Gornitz, 1973; Cruikshank and Wood

1972; Peterson et al., 1994). Sections of these sinuous rilles also contain

depression features which may be evidence of intact lava tubes (Haruyama et al.,

2009). There have also been recent discoveries by SELENE of vertical holes

around a sinuous rille at the Marius Hills region, suggested to be evidence of a

skylight- another collapsed feature of a lunar lava tube (Haruyama et al., 2009;

Figure 4).

13

https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2009GL040635#grl26500-bib-0020
https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2009GL040635#grl26500-bib-0020
https://agupubs.onlinelibrary.wiley.com/authored-by/Haruyama/Junichi
https://agupubs.onlinelibrary.wiley.com/authored-by/Haruyama/Junichi


14



These lunar lava tubes possess a unique opportunity to sustain long-term

lunar missions. Their presence can allow for construction of lunar bases where the

thick rock walls and roof can protect from micrometeorites and intense solar

radiation (Hörz, 1985; Angelis et al., 2002). While there are extreme temperature

fluctuations on the surface of the Moon, the temperature inside a lava tube is

expected to be relatively consistent and mild, at ~20⁰C (Hörz, 1985). Since the

inside of a lunar lava tube is protected from these surface conditions, they can

also provide a chance to better understand an unaltered lunar environment that

persisted during the time the tube was formed (Haruyama et al., 2012). These

lunar lava tubes may also be home to useful volatiles, such as water ice, that need

a consistent temperature below freezing to remain intact (Haruyama et al., 2012).

Lunar lava tubes are an ideal candidate for sustained lunar exploration.

Currently, there are limited abilities to validate their presence using remote

sensing techniques. Ground-based field work using geophysical techniques, such

as seismic surveys, can be used during future lunar missions to confidently

identify these void spaces. However, ground-based lunar missions are expensive

and time-consuming. Therefore, Earth-based studies of analogous terrestrial lava

tubes and volcanic regions are used to improve the efficiency and capabilities of

future lunar missions. Studying terrestrial lava tubes can provide valuable

information on their structure, dimensions, and volcanic history that can aid in the

investigation of lunar lava tubes. Lava tubes and volcanic environments similar to

those found on the Moon are found and studied near in many locations including

Lava Beds National Monument (LBNM) in California (Waters et al., 1990), the

15



Hawaiian Island volcanoes (Greeley, 1987), and Iceland (Sam et al., 2020). For

this study, three Earth-based analog field sites were chosen: The San Francisco

Volcanic Field in Arizona, Kilbourne Hole in New Mexico, and Lava Beds

National Monument in California. Earth-based studies are less expensive, easier

to access, and can provide the unique opportunity to ground-truth techniques

because the location of lava tubes are already known (Figure 5).

16



1.5: Research Goals

In order to refine interpretations of the subsurface using these geophysical

techniques, a solid understanding of the geotechnical properties of the material

composing the surface is necessary. One way to achieve this involves

experimental rock mechanics.

This study will investigate the seismic signature of small-scale basalt

samples from three volcanic field sites and refine the relationship between their

seismic signature and material properties. We will refine the relationship between

the velocity of lunar analog material with a range of porosities, a major control on

the seismic signature of a material. Clay content will be measured to determine if

it plays a significant role in velocity variations. Additionally, the mechanical

properties of the basalts will be measured and velocity will be derived from these

mechanical properties in order to ground-truth the original technique used to

collect velocity measurements. Three-dimensional tomographic imaging will be

used to quantify the pore space and pore structure of selected basalts. Finally,

material properties measured in the laboratory will be useful as input parameters

for future modeling of the lunar surface.

Han et al., (1986) measured the velocity of sandstones with varying

amounts of porosity and clay content in order to determine a relationship between

these material properties and velocity. They found that both porosity and the

amount of clay are linearly related to both the compressional and shear velocity of

the sandstone; both reduce the overall velocity. Han et al., (1986) also determined

that these porosity and clay content affect the elastic moduli of the sandstones.

17



Higher clay content and higher porosity will result in lower shear modulus and

bulk modulus, with the shear modulus being affected greater. Similarly, there are

variations in porosity and clay content in basalt. We hypothesize that similar

relationships may exist for basalt. Characterizing the effects these parameters

have on the velocity of basalt will aid in constraining the non-uniqueness of the

lunar surface’s velocity.

Geophysical methods, such as seismic surveys, are not new techniques for

investigating the subsurface. This study adds an experimental method to capture a

material’s seismic signature to these techniques, thereby investigating the

subsurface on a smaller scale. The GEODES team uses Earth-based analog field

studies to investigate the subsurface of volcanic regions using co-located

techniques. Seismic surveys, ground penetrating radar, and LiDAR scans are all

conducted in the same location where samples are collected for experimental

characterization. Data collected from these techniques can therefore cover a wide

spatial scale (Figure 6). Geophysical methods can investigate the surface on the

scale of kilometers to meters. With experimental techniques, the addition of

smaller spatial scales is achieved. Experimental rock mechanics investigates the

subsurface in the centimeter scale and three-dimensional tomographic imaging

investigates the subsurface on a micron scale. This study aims to characterize

lunar analog material and provide insights on material properties and seismic

signatures on a smaller scale than is possible using field measurements or remote

sensing techniques alone.

18



19



Chapter 2: Study Sites

To optimize the effectiveness of planetary studies, scientists may first

conduct analog studies here on Earth in order to test new instruments, work out

the logistics of data collection, train future astronauts, and prepare a workflow for

future missions (Groemer & Ozdemir, 2020). Analog studies also provide an

approach to refine these methods in a more cost effective and accessible manner.

Previous studies have used terrestrial volcanic fields as analog field sites to the

lunar surface (Hargraves and Buddington, 19070; Cruikshank and Wood, 1972;

Léveillé, 2010; Bell et al., 2021; Esmaeili et al., 2020). This study will use

samples collected from multiple analog field sites and focuses on laboratory

classification. These field sites include the San Francisco Volcanic Field in

Arizona, Kilbourne Hole in Southern New Mexico, and Lava Beds National

Monument in Northern California.

2.1: History of Selected Study Sites

1. The San Francisco Volcanic Field

The San Francisco Volcanic Field is composed of hundreds of cinder

cones and associated lava flows, composed primarily of alkali olivine basalt,

covering an area of approximately 5,000 square kilometers (Moore, 1976, Figure

7). While most eruptions consisted of basaltic material, small amounts of

intermediate to silicic magmas are located around San Francisco Mountain and

O’Leary Peak, as well as some ultramafic to mafic xenoliths. The sedimentary

rocks beneath the lava flows dip slightly northeast and therefore the lava flows are

20



layered in that direction. There is evidence of a fracture system that trends

northwest in the southern section and north-northeast in the northern section,

based on vent elongation and alinements (Moore, 1976). Classification for these

cones and lava flows are based on five parameters: stratigraphy and physiography,

weathering and erosion patterns, tree-ring age, and chemical and petrographic

information. Based on these five parameters, Moore, 1976 classify these rocks

into five groups from oldest to youngest: basalt of cedar ranch (6-5 Ma),

Woodhouse (3-0.8 Ma), Tappan (0.7-0.2 Ma), Merriam (<150,000 yrs), and

Sunset age (1064 A.D.).

21



These rocks vary in both composition and structure. The basalt of Cedar

Ranch is classified as ophitic or subophitic and has altered olivine phenocrysts.

Robinson (1913) named these rocks “augite basalt” likely for their high augite

content. The basaltic rocks of Woodhouse age form more smooth and thin sheets

with thicknesses from 2 to 15 meters (Moore, 1976). These rocks typically

contain tiny gabbroic nodules and some larger grains of plagioclase,

clinopyroxene and olivine. These rocks also have an ophitic texture and some are

characterized by a high percentage of titanium dioxide. The basaltic rocks of

Tappan age are classified by well-defined lava flows ranging in thickness from 1

to 60 meters and identifiable by volcanic flow features such as hornitos and

pressure ridges. Most of the basalts in this group are alkali olivine and alkali-rich

high-alumina basalts and vary widely from nonporphyritic to porphyritic with

little to many phenocrysts of plagioclase, clinopyroxene, and olivine (Moore,

1976). The basalts of Merriam age consist of younger lava flows that contain

well-defined levees and surface structures. These flows are also alkali olivine and

alkali-rich high-alumina basalts with variable amounts of phenocrysts. The

basaltic rocks of sunset age are the youngest of the five groups and the lava flows

associated with this age have very fresh, rough, and well-preserved surfaces and

contain structures that suggest both pahoehoe and aa flows. They consist of alkali

olivine basalt with some phenocrysts of plagioclase, clinopyroxene, and olivine

(Moore, 1976). Basalts for this study come primarily from the black point lava

flow and are Woodhouse age.

22



2. Kilbourne Hole

Kilbourne Hole is a maar crater that runs ~3 km long and a quarter mile

wide with a rim surrounding the depression that ranges in height from ten to 150

feet (Reiche, 1940, Figure 8). It sits at the eastern edge of the Potrillo volcanic

field; active from about 1.2 million to 20,000 years ago (Hawley, 1981).

Extension and thinning of the crust in this region were likely the trigger for

volcanic activity. The formation of this crater is a result of the mixing of

groundwater during a magmatic event, resulting in a radial explosive eruption and

formation of this mostly circular crater.

23



Samples in this study were collected in the southeastern region of the

crater. The composition of the rim walls in this region can be simplified into four

distinct layers: the sedimentary rock at the base, a basaltic lens, volcanic surge

beds, and unconsolidated volcanic ash. Pre-eruption sandstones sit at the base of

the crater and continue roughly 10-20 meters up the rim wall. The basaltic lens,

formed during a pre-eruption lava flow, thins out near the southern portion of the

crater but contains sections of visible differences in pore abundance, weathering,

and olivine phenocrysts. All of the samples for this study from this location are

from this basaltic lens. The volcanic surge beds form during hot pyroclastic

eruptions where gas, steam and ash are expelled. These eruptions are believed to

have occurred repeatedly, creating fragile cross-stratified and layered deposits.

The final layer consists of more ash deposits similar in composition to the surge

beds but unconsolidated.

3. Lava Beds National Monument

Lava Beds National Monument was formed from repeated lava flows and

eruptions from Medicine Lake shield volcano over millions of years up to about

1,000 years ago (Juillerat, 2015). There are more than 700 lava tubes in the entire

region. Specifically, this study samples from the lava tube known as Skull Cave

(Figure 9). Most of the volcanic rocks in the region, including the samples for this

study, are basalt. Studies have indicated that this region may be highly (but

variably) permeable given the somewhat random distribution of fractures, spaces,

and tubes on both a macroscopic and microscopic scale (Hotchkiss, 1968). The

characteristics of the volcanic rocks in this area vary widely as well, given the

24



long eruption period. Basalts and lava flows range from vesicular olivine basalts

with rough surfaces, fresh and mostly unweathered, to reddish brown weathered

scoria. Lava flows can also be composed of glass shards, pumice, brecciated

scoria and even more andesitic rocks.

2.2: Sample Material and Collection Process

These samples were collected during field expeditions during which

field-scale seismic data was also collected. The combination of field data and

25



laboratory measurements conducted during this study will provide further insight

into the geology and structure of each locale.

1. San Francisco Volcanic Field Samples

Samples at the San Francisco Volcanic Field were collected by Jacob

Richardson during the 2019 GEODES field expedition. Samples were selected

based on visible variations in physical properties such as color, density, and pore

structure and sampling followed a vertical transect of a quarry wall (Figure 10).

Above this quarry wall a seismic survey was conducted in order to measure the

field-scale seismic activity. One additional sample was also collected from SP

Crater. Six samples in total were collected and sent to the laboratory for analysis

(Table 2, Figure 11).

26



27



2. Kilbourne Hole Samples

Samples at Kilbourne Hole were collected during the 2021 RISE2 and

GEODES field expedition. These samples were also selected based on visible

variations in physical properties such as color, density, and pore structure.

Sampling was conducted in each of the four layers shown below (Figure 12). The

28



top most layer is composed of an unconsolidated ash layer. In order to sample this

layer a 5-gallon bucket was filled. The second layer is composed of volcanic

surge beds. These samples were collected but unfortunately were too fragile to

survive transit back to The University of Maryland. The next layer is composed of

a basaltic lens and four samples were collected. The bottom-most layer is

composed of sedimentary rock and three samples were collected. Field-scale

seismic lines were conducted at each of these layers. The basalt samples are used

in this study (Table 3, Figure 13).

29



30



3. Lava Beds National Monument Samples

Samples at Lava Beds National Monument were collected during the 2022

GEODES field expedition. These samples were also selected based on visible

variations in physical properties such as color, density, and pore structure.

Samples for this study were collected in and around Skull Cave. Skull Cave is a

static ice cave formed from a lava flow from Medicine Lake Volcano

approximately 10,500 to 65,00 years ago. The lava flow drained downslope

leaving behind a lava tube. In the winter, Skull Cave circulates cold air throughout

its cave system. In the summer, since hot air rises, Skull Cave traps the cold

31



winter air in lower caverns, allowing for water ice to remain frozen year-round.

Skull cave is also one of the largest caves in Lava Beds National Monument,

roughly 180 meters in length and has both a complex geometry and rocks that

vary visibly in porosity (Bell et al, 2022).

Sampling was conducted in three locations (Figure 14). Area one, labeled

SKU_IN was located inside Skull Cave, ~30 meters from the entrance. Area two,

labeled SKU_LB1 was located on the surface, directly above Skull Cave. Area

three, labeled SKU_LB2 was located at an area of exposed basalt hypothesized to

be a secondary lobe of the lava flow that formed Skull Cave. A seismic survey

was conducted on the surface across Skull Cave where SKU_LB1 samples were

collected. Three samples were collected from each of the three locations (Table 4,

Figure 15).

32



33



34



35



2.3: Major Variations due to Sample Locale

The lunar surface is composed largely of basaltic materials that can have

physical properties that vary widely. Some of these variations include, but are not

limited to; a range of porosity values, varying amounts of clay content, size and

shape of pores, and weathering. These variations will play a significant role in the

basalt’s seismic signature. This study mainly focuses on quantifying the seismic

effects porosity and clay content have on basalt. Interpretation of future

geophysical data collected of the lunar surface must therefore be accompanied

with a solid understanding of the material properties of basaltic material.

Due to the nature of sampling from three different study sites located in

three different regions of the United States, natural variations based on sample

locale will likely be present.

Active source seismic experiments were conducted at all three sample

locations: The San Francisco Volcanic Field, Kilbourne Hole, and Lava Beds

National Monument in the vicinity of sample collection. Because of this

co-location of field data and sampling, parallels can be drawn between field-scale

seismic signatures and laboratory seismic signatures. The range of seismic

velocities of the basaltic regions of each location are as followed: The San

Francisco Volcanic Field has a velocity ranging from 500 m/s - 3000 m/s for Vp

and 50 m/s to 200 m/s for Vs, the basaltic lense at Kilbourne Hole has a velocity

ranging from 500 m/s to 2000 m/s for Vp and 300m/s to 600 m/s for Vs, and the

area above the opening of Skull Cave at Lava Beds National Monument has a

36



velocity ranging from 500 m/s to 4000 m/s for Vp and an unresolved velocity

range for Vs (Table 5).

37



Chapter 3: Technical Background

3.1: Velocity of Basalt

Seismology plays an important role in the future exploration of the

Moon’s surface. Seismology on Earth has been used for decades to aid in the

discovery of many subsurface resources and structures. For example, some of the

earliest geophysical exploration focused on using geophysics to aid in the

discovery of petroleum (Sheriff and Geldart, 1995). In June of 1921, what is

considered the first seismograph exploration party, consisting of Karcher,

Haseman, Perrine, and Kite, recorded some of the first seismic reflections

between two rock bodies (Schriever 1952). Since the 1920’s, there have been

many advances in the study and application of seismology including studies of

deep earthquakes (H H Turner, 1922), the Earth’s inner core (Lehmann, 1936),

and studies on other planetary bodies (Latham, 1969). In future missions to the

Moon, seismology will be used in various prospecting efforts including searching

for resources such as water ice, hydrogen and other volatiles trapped in the

subsurface.

One of the most abundant volcanic materials on the Moon is basalt,

making up about a quarter of the lunar surface. Future lunar missions will likely

focus in highly basaltic regions and therefore understanding the seismic properties

of basalt will be important in interpreting the structures present on the Moon.

The seismic signature of basalt can vary widely based on the exact

composition and/or porosity (Carlson, 2014). Because of this, there is a range of

38



velocities that have been recorded for different basalt samples around the world.

In order to constrain the interpretation of the seismic velocity of basalt, the

relationship between velocity and porosity as well as velocity and clay content

must be studied.

3.2: Experimental Setup

Sample Preparation

Prior to conducting experiments, porosity was measured using the

saturation method. Samples were cored into cylinders with diameters of 18.41

mm and lengths of 38.10 mm. In order to measure porosity, samples were

saturated in deionized water and put into a vacuum chamber at 60 C for 24 hours.◦

The samples were weighed three times each and an average saturated weight was

derived. The samples are then placed back into the oven for an additional 24

hours in order to completely dry. The samples were weighed three additional

times after dried and an average dry weight was derived. Pore volume was

determined for each sample by calculating the difference between the average

saturated weight and the dried sample weight and dividing by the total sample

volume which was calculated from the sample’s dimensions. Porosity values

using the saturation method have some limitations when the pore structure is not

well connected as the water does not have a way to enter some microscopic pores.

Because of this there is an error of ~2% in the porosity measurements derived

using the saturation method.

39



Select samples were sent to Texas A&M where three-dimensional

tomographic images were collected by The University of Texas High-Resolution

CT Facility. Images were then segmented and analyzed using Avizo imaging

software in order to visualize the pore structures and obtain a precise porosity

measurement for samples from two of the three locations: The San Francisco

Volcanic Field and Lava Beds National Monument. These porosity values were

compared to the porosity values obtained through the saturation method. Because

of the limitations using the saturation method, when available in further

calculations a sample’s CT scan derived porosity is used. Porosity was measured

using the saturation method and using the 3-D tomographic images for samples at

The San Francisco Volcanic Field and Lava Beds National Monument. Porosity

was measured using only the saturation method for samples at Kilbourne Hole

(Tables 6-8).

40



Prior to experiments, cored samples are jacketed first with polyolefin

heat-shrink tubing and second with a thick rubber heat-shrink tubing. Two jackets

are fitted to ensure there is no leakage of confining fluid into the pore space of the

samples. Piezoelectric transducers are placed on either end of the sample in order

to capture the ultrasonic seismic signature of the sample during the hydrostatic

experiments. Between the sample and the piezoelectric transducers, a thin layer of

shear wave couplant is used to maximize the coupling between transducer and

41



sample end. 18-gauge steel tie wires are placed on either side of the assembly and

two-step epoxy is placed at the ends of the jackets and allowed to cure overnight.

This sample assembly is then placed in the Autolab 1500 pressure vessel (Figure

16).

42



Experiments in the Autolab 1500

Experiments were conducted using the NER Autolab 1500 deformation

apparatus. The Autolab 1500 is a servo-controlled deformation apparatus that can

use mineral oil as a confining medium. Although this machine was designed with

industrial engineering purposes in mind, it can be used on an academic level as

well.

In the experiments shown in this study, two experimental setups were used

for each sample. First, a hydrostatic experiment was conducted. A confining

pressure was applied to the sample assembly up to 60 MPa at a rate of 2

MPa/min. From the hydrostatic experiment velocity data was collected as well as

mechanical data in order to directly derive a bulk modulus. For select samples a

second experiment was conducted using an axial setup. An axial load was applied

to the sample at a steady rate up to ~60 MPa. Velocity data was collected at a 5

MPa increments and mechanical data was collected in order to directly derive a

Young’s modulus.

The method for collecting a seismic signature of each sample varied

slightly by location:

The first set of experiments were conducted on samples from the San

Francisco Volcanic Field. For the hydrostatic experiments, the confining pressure

increased to a maximum of 60 MPa and seismic signatures were collected at a

frequency of ~700 kHz using the piezoelectric transducer assembly at intervals of

10 MPa. Two samples from this locale were chosen in order to conduct the axial

experiment along with the hydrostatic experiment.

43



The second set of experiments were conducted on samples from Kilbourne

Hole. For the hydrostatic experiments, the confining pressure increased to a

maximum of 60 MPa and seismic signatures were collected at a frequency of

~700 kHz using the piezoelectric transducer assembly at intervals of 2 MPa. No

axial experiments were conducted with samples at this locale.

The third set of experiments were conducted on samples from Lava Beds

National Monument. For the hydrostatic experiments, the confining pressure

increased to a maximum of 60 MPa and seismic signatures were collected at a

frequency of ~700 kHz using the piezoelectric transducer assembly at intervals of

2 MPa. This hydrostatic experiment was repeated with the same conditions for a

second run. All samples from this locale were used in order to conduct the axial

experiment as well.

3.3: Velocity-Porosity Relationship

The velocity of a medium consists of a combination of the velocities and

physical properties of all the material that make up that medium. Variations in

mineralogy, pore geometry, pore fluid, and temperature are all known to affect

velocity of a material. One parameter that significantly affects the velocity of

materials is porosity (Yu et al., 2016).

Fractures in rocks are very compliant and therefore affect the bulk elastic

properties. A rock with high porosity will have lower elastic moduli values and in

turn will have a lower velocity value than a rock with low porosity. Studies such

as Han et al., 1986 determine the relationship between velocity and porosity and

44



velocity and clay content using sandstone and determine that higher porosity and

greater clay content reduce the compressional and shear velocity (Figure 17).

45



Using the derived linear fits of velocity and porosity and velocity and clay

content for the sandstones a relationship can be established.

𝑉
𝑝
 =  5. 59 −  6. 93ϕ −  2. 18𝐶

and

𝑉
𝑠
 =  3. 52 −  4. 91ϕ −  1. 89𝐶

where is the P-wave velocity, is porosity, C is volume clay content, and is𝑉
𝑝

ϕ 𝑉
𝑠

the S-wave velocity. This study will adapt the systematic derivation of these

relationships for basalts.

Other studies determine that there is likely a different role porosity plays

on the Earth than on the Moon. On Earth, fractured rocks are often saturated with

fluids while the Moon is composed of extremely dry cracked rocks. Borgomano,

2019 found that saturated cracked glass behaves similarly to uncracked glass

while dry cracked glass has a lower starting velocity at lower pressures.

Therefore, porosity will have a much greater effect on in-situ lunar rocks than

in-situ terrestrial rocks and velocities are expected to be much lower on the highly

fractured, dry, lunar surface (Borgomano, 2019; Figure 18).

46



Natural, in-situ velocity measurements of terrestrial basalts may therefore

not best represent the velocities expected on lunar basalts. Experimental studies,

like this one, in which saturation can be controlled and limited may be a better

representation of the velocities of lunar rocks.

47



3.4: Velocity-Effective Moduli Relationship

At certain conditions, the mechanical behavior of rocks can be described

using the theory of elasticity. Rocks can be structurally characterized by

measuring or calculating properties known as elastic moduli. The elastic modulus

of a material describes that material’s resistance to being deformed when a given

stress is applied. The speed of an elastic wave propagating through a material is

related to the elastic moduli of that material. Therefore, elastic moduli can be

derived from velocity measurements as well. If the density is known and the

P-wave ( ) and S-wave ( ) velocity are measured, various elastic moduli of𝑉
𝑝

𝑉
𝑠

that material can be derived and vice versa. In this study, elastic moduli are

calculated using velocity measurements in order to ground-truth the velocity

obtained through experimental methods. These derived elastic moduli values are

compared with measured elastic moduli values from deformation experiments.

Two primary elastic moduli are a material’s bulk modulus (K) and

Young’s modulus (E). The bulk modulus of a material describes a material’s

resistance to compressibility. It is defined as volumetric stress/volumetric strain.

The Young’s modulus of a material describes a material’s resistance to axial

stretching and is defined as the ratio of axial stress/axial strain. After two elastic

properties of a material are independently determined, various other properties

can be calculated to further characterize a material (Gueguen and Palciauskas,

1994).

48



For this study, the bulk modulus and Young’s modulus are independently

measured during deformation experiments and the shear modulus ( ) and Poissonµ

ratio ( ) are subsequently derived. The shear modulus of a material describes aν

material’s likelihood to shear when subjected to a force and is defined as shear

stress/shear strain. The Poisson ratio is a measure of the material’s deformation

that occurs perpendicular to the direction of loading. Most materials have values

that range from 0.0 to 0.5, where a perfectly incompressible material will have a

Poisson’s ratio of exactly 0.5 (Gueguen and Palciauskas, 1994).

The Young's modulus is the slope of the stress-strain curve from an axial

deformation experiment. The differential pressure is measured directly from the

Autolab’s load cell and the axial strain is derived using the following equation:

Eq. 1ε
𝑎
 =  ((𝑑 −  ((1/1950) *  σ

𝑑
)) / 38. 1) *  1000

where is the axial strain, is the displacement, and is the differential stress.ε
𝑎

𝑑 σ
𝑑

The bulk modulus is the slope of the stress-strain curve from a hydrostatic

deformation experiment. The confining pressure is measured directly from the

Autolab’s pressure sensor and the volumetric strain is derived using the following

equation:

Eq. 2ε
𝑣
 =  3 *  (((𝑑 −  ((1/1950) *  σ

𝑣
)) / 38. 1) *  1000)

where is the volumetric strain, is the displacement, and is theε
𝑣

𝑑 σ
𝑣

volumetric/mean stress.

The is multiplied by three in order to account for the strain in the x, y,ε
𝑣

and z directions. This can lead to complications if the sample is not completely

49



homogenous. In order to address this possible discrepancy in the bulk modulus

value, not only is the shear modulus ( ) and Poisson ratio ( ) derived directlyµ
𝑚

ν
𝑚

from the Young’s modulus ( ) and bulk modulus ( ), the shear modulus ( )𝐸
𝑚

𝐾
𝑚

µ
𝑐

and bulk modulus ( ) are derived from the measured Young’s modulus ( ) and𝐾
𝑐

𝐸
𝑚

a range of Poisson’s ratio ( ) values from 1 - 4.5. The Poisson’s ratio can also beν
𝑐

calculated from the frequency (denoted as a superscripted number on and )𝑉
𝑝

𝑉
𝑠

dependent velocities, and , in order to verify which Poisson’s ratio𝑉
𝑝
700 𝑉

𝑠
700

between 1 - 4.5 might be most accurate. The equation used is as follows:

Eq. 3ν
𝑚

 =  1
2

(𝑉
𝑝
700 /𝑉

𝑠
700)

2
 − 2

(𝑉
𝑝
700 /𝑉

𝑠
700)

2
 − 1

P-wave velocities ( ) and S-wave velocities ( can be calculated from𝑉
𝑝
𝐼 𝑉

𝑠
𝐼)

the bulk modulus ( ) and shear modulus ( ) values for each basalt sample𝐾
𝑚

µ
𝑚

(Table 9). These P-wave and S-wave velocities can be compared to the P-wave (

) and S-wave ( velocities calculated from the calculated bulk modulus ( )𝑉
𝑝
𝐼𝐼 𝑉

𝑠
𝐼𝐼) 𝐾

𝑐

and calculated shear modulus ( ). These calculated P-wave and S-waveµ
𝑐

velocities will therefore be frequency independent because they are derived

directly from mechanical data. The frequency independent velocities can then be

compared to the frequency dependent velocities, and measured with the𝑉
𝑝
700 𝑉

𝑠
700

transducer assembly in the Autolab 1500, in order to determine if frequency plays

50



a significant role in measuring velocities in a laboratory setting. A table

containing the nomenclature for all variables can be found in the appendix.

Table 9. Relationships between E, , K, , , and . The first column areν µ 𝑉
𝑝

𝑉
𝑠

input values and the row below are output values.

Young’s
Modulus
( )𝐸

𝑚

Poisson’s
Ratio
( )ν

𝑚

Bulk
Modulus
( )𝐾

𝑚

Shear
Modulus
( )µ

𝑚

Bulk
Modulus
( )𝐾

𝑐

Shear
Modulus

( )µ
𝑐

P-wave
velocity
( / )𝑉

𝑝
𝐼 𝑉

𝑝
𝐼𝐼

S-wave
Velocity
( / )𝑉

𝑠
𝐼 𝑉

𝑠
𝐼𝐼

,𝐸
𝑚

ν
𝑚

𝐸
𝑚

ν
𝑚

- - 𝐸
𝑚

3(1 − 2ν
𝑚

)

𝐸
𝑚

2(1 + ν
𝑚

)
- -

,𝐾
𝑚

µ
𝑚

- - 𝐾
𝑚

µ
𝑚

- - 𝐾
𝑚

+ 4
3 µ

𝑚

ρ

µ
𝑚

ρ

,𝐾
𝑐

µ
𝑐

- - - - 𝐾
𝑐

µ
𝑐 𝐾

𝑐
+ 4

3 µ
𝑐

ρ

µ
𝑐

ρ

In total four elastic moduli, bulk modulus, Young's modulus, shear

modulus, and Poisson ratio, are determined for all samples with axial data. This

includes all basalts from Lava Beds National Monument. Basalts from The San

Francisco Volcanic Field and Kilbourne Hole have only hydrostatic data and

therefore these calculations do not apply. This study plans to utilize these

equations used to describe the relationship between velocity, elastic moduli, and

porosity, specifically to basaltic material. The main goal is to create a velocity

model that can describe and predict porosities of basalts in hope of benefiting the

interpretation of the geophysical signature of the lunar surface.

51



Chapter 4: Results

4.1: Velocity Data

A series of hydrostatic experiments were conducted at three volcanic

regions of the US: (1) The San Francisco Volcanic Field, (2) Kilbourne Hole, and

(3) Lava Beds National Monument. During the hydrostatic experiment, a seismic

signature of each sample was collected using the transducer assembly at various

pressures. The relationship between rock properties and seismic data due to

sample variations and site variations were analyzed. The complete dataset can be

found in the appendix.

1. The San Francisco Volcanic Field

Six basalt samples, varying in porosity, from the San Francisco Volcanic

Field were chosen. The P-wave velocity for the six samples ranges from ~5 km/s

to ~6 km/s +/- 0.1 km/s. The S-wave velocity for the six samples ranges from

~2.1 km/s to ~3 km/s +/- 0.1 km/s (Table 10, Figure 19). There is little variation

in velocity due to loading. For both the P-wave and the S-wave, velocity

measurements taken when increasing the confining pressure from 10 MPa to 60

MPa are mostly consistent with the velocity measurements taken when decreasing

the confining pressure from 60 MPa to 10 MPa. Linear regressions for each

individual basalt have relatively small positive correlations with low R-squared

values. There is almost no increase in the P-wave velocity at increasing pressures,

indicative of no pressure dependence on the P-wave velocity for all six basalt

samples. There is an increase in the S-wave velocity at increasing pressures,

52



indicative of a pressure dependence on the S-wave velocity, evident in all six

samples. Linear regressions for the S-wave velocity of each individual basalt have

positive correlations and high R-squared values, with a median value of 0.74

(Table 10). The Vp/Vs ratio can be calculated from the P-wave and S-wave

velocities. This Vp/Vs ratio is used in order to determine a Poisson’s ratio. The

Poisson’s ratio for all San Francisco Volcanic Field samples ranges from 0.34 -

0.39.

53



54



55



2. Kilbourne Hole

Four basalt samples, varying in porosity, from Kilbourne Hole in New

Mexico were chosen. The P-wave velocity for the four samples ranges from ~3.5

km/s to ~4.4 km/s +/- 0.1 km/s. The S-wave velocity for the four samples ranges

from ~1.8 km/s to ~2.3 km/s +/- 0.1 km/s (Table 11, Figure 20). There is little

variation in velocity due to loading. For both the P-wave and the S-wave, velocity

measurements taken when increasing the confining pressure from 10 MPa to 60

MPa is mostly consistent with the velocity measurements taken when decreasing

the confining pressure from 60 MPa to 10 MPa. There is almost no increase in the

P-wave velocity at increasing pressures, indicative of no pressure dependence on

the P-wave velocity for all basalt samples except sample KH 1. Sample KH 1 has

a higher positive correlation by one to two orders of magnitude compared to the

other samples and a high R-squared value of 0.85. Sample KH 1 therefore has a

slight pressure dependence on its P-wave velocity. There is no increase in the

S-wave velocity at increasing pressures, and therefore no pressure dependence, on

all samples except KH 1. Again, KH 1 has increasing S-wave velocities at

increasing pressure, indicative of a pressure dependence. The Vp/Vs ratio can be

calculated from the P-wave and S-wave velocities. This Vp/Vs ratio is used in

order to determine a Poisson’s ratio. The Poisson’s ratio for all Kilbourne Hole

samples ranges from 0.13 - 0.33 (Table 11).

56



57



58



3. Lava Beds National Monument

Eight basalt samples, varying in porosity, from Lava Beds National

Monument were chosen. The P-wave velocity for the eight samples ranges from

~3.1 km/s to ~6 km/s +/- 0.1 km/s. The S-wave velocity for the eight samples

ranges from ~1.2 km/s to ~2.4 km/s +/- 0.1 km/s (Table 12, Figure 21). There is

little variation in velocity due to loading. For both the P-wave and the S-wave,

velocity measurements taken when increasing the confining pressure from 10

MPa to 60 MPa is mostly consistent with the velocity measurements taken when

decreasing the confining pressure from 60 MPa to 10 MPa. There is almost no

increase in the P-wave velocity at increasing pressures, indicative of no pressure

dependence on the P-wave velocity for all basalt samples except samples SKU

IN1 and SKU IN2. Samples SKU IN1 and SKU IN2 have a higher positive

correlation compared to the other samples and therefore a positive correlation

between P-wave velocity and pressure. All eight samples have increasing S-wave

velocities at increasing pressure, indicative of a pressure dependence. The Vp/Vs

ratio can be calculated from the P-wave and S-wave velocities. This Vp/Vs ratio

is used in order to determine a Poisson’s ratio. The Poisson’s ratio for all

Kilbourne Hole samples ranges from 0.25 - 0.40 (Table 12).

59



60



61



4. Study Site Variations

The P-wave velocities of samples from The San Francisco Volcanic Field

and Kilbourne Hole, New Mexico plot in distinct clusters, while samples from

Lava Beds National Monument have a wider range of P-wave velocities and

overlap with the two other sites. The S-wave velocities of samples from The San

Francisco Volcanic Field and Kilbourne Hole, New Mexico also plot in distinct

clusters. Samples from Lava Beds National Monument have a wider range of

S-wave velocities, overlapping with samples from Kilbourne Hole, New Mexico

(Figure 22).

62



63



For all three study sites a negative correlation is evident between velocity

and porosity. For the P-wave, the SFVF basalts and KH basalts have a strong

correlation, with the SFVF samples having an R-squared value ranging from

0.92-0.98 and the KH samples having an R-squared value ranging from 0.58-0.86.

The LBNM basalts have more scatter between samples and a lower correlation

with an R-squared value ranging from 0.04-0.08 (Table 13, Figure 23).

64



65



66



The S-wave velocity for all three sites has a slightly weaker correlation

with porosity. The SFVF samples have an R-squared value ranging from

0.60-0.83. The KH samples have an R-squared value ranging from 0.03-0.60 and

the LBNM basalts have an R-squared value ranging from 0.11-0.14 (Table 14,

Figure 24).

67



68



69



4.2: Mechanical Data

Experiments conducted at Lava Beds National Monument contain

hydrostatic deformation data in which a bulk modulus can be derived.

Experiments conducted with samples from Lava Beds National Monument also

include axial mechanical data in which the Young’s modulus can be derived. The

Young’s modulus combined with the velocity-derived Poisson’s ratio values are

used to derive a bulk modulus, and a shear modulus. These moduli values are then

used to calculate a P-wave velocity and S-wave velocity independent of

frequency.

The Young’s modulus for the eight samples at Lava Beds National

Monument ranges from 10 - 25 GPa (Figure 25-27, Table 15). The bulk moduli of

basalts from Lava Beds National Monument range from 4 - 6 GPa (Figure 28-30,

Table 15). The shear modulus of the eight basalts, calculated from the measured

Young’s modulus and measured bulk modulus is 4 - 27 GPa. The P-wave velocity

calculated from the bulk and shear modulus ranges from 3.2 - 4.0 km/s and the

S-wave velocity calculated from the bulk and shear modulus ranges from 1.5 - 3.4

km/s (Table 15).

70



71



72



73



74



75



76



77



A range of bulk modulus values are calculated using the samples Young’s

modulus value and a Poisson’s ratio derived from the frequency dependent

velocity. Error bars for the Poisson’s ratio range from possible values from 0.1 -

0.45. These bulk modulus values range from 12-42 GPa depending on the sample

and Poisson’s ratio value used. A range of shear modulus values are calculated

using the samples Young’s modulus value and a Poisson’s ratio as well. The shear

modulus values range from 4-10 GPa depending on the sample and Poisson’s ratio

value used (Table 16).

78



79



The P-wave velocities derived from the raw bulk modulus and shear

modulus do not correspond to the P-wave velocities measured using the

transducer assembly. The P-wave velocities measured using the transducer

assembly plots closest to the P-wave velocities calculated from the calculated

bulk modulus and calculated shear modulus- with the P-wave velocities calculated

from the average Poisson’s ratio value plotting the closest (Figure 31). The error

bars represent the range of P-wave velocities possible for each sample with a

given Young’s modulus and having a range of Poisson’s ratio values from 0.1 -

0.45 and the actual value plotted uses an average Poisson’s value. The agreement

between both the measured P-wave velocities and the calculated P-wave

velocities indicates little to no frequency dependence on the measured velocities.

80



The S-wave velocities derived from the raw bulk modulus and shear

modulus correspond to the S-wave velocities measured using the transducer

assembly better than the P-wave velocities. The S-wave velocities measured using

the transducer assembly plots closest to the S-wave velocities calculated from the

calculated bulk modulus and calculated shear modulus- with the S-wave velocities

calculated from the average Poisson’s ratio value plotting the closest (Figure 32).

The error bars represent the range of S-wave velocities possible for each sample

with a given Young’s modulus and having a range of Poisson’s ratio values from

0.1 - 0.45 and the actual value plotted uses an average Poisson’s value..Although

the measured values do not plot on top of the derived values, the values plot

closer to one another than the P-wave velocities. This is most evident when the

same y-axis scale is used between the P-wave velocities and S-wave velocities

(Figure 33). This may indicate little frequency dependence on the measured

velocities as well.

81



82



Chapter 5: Discussion

5.1: Derived Velocity-Porosity Relationship

Basalts from all three study sites demonstrate a negative correlation

between their velocity and porosity (Figure 34-36). Porosity is determined using

two different methods for the basalts at SFVF and LBNM: the saturation method

and the tomography method and only using the saturation method at KH. There is

a stronger correlation using the saturation method than the tomographic images

which may be due to the nature of the tomography method. A subvolume of each

basalt sample is used in calculating the porosity and this may lead to greater error

in the porosity of the full sample.

83



84



85



86



The generalized mixture rule (Ji, 2004), expressed as a two-phase system,

can help understand the relationship between velocity and porosity.

Eq. 4𝑀
𝑇
 =  𝑉𝑜𝑙

𝑠
𝑀

𝑠
 + 𝑉𝑜𝑙

𝑝
𝑀

𝑝
  

Where = the total velocity, and are the volume fraction of𝑀
𝑇

𝑉𝑜𝑙
𝑠

𝑉𝑜𝑙
𝑝

the solid and pore space respectively, and and are the velocities of the𝑀
𝑠

𝑀
𝑝

solid and the pore space respectively. The total velocity ( ) of a basalt with a𝑀
𝑇

greater volume fraction of solid rock will have a greater contribution of the solid

rock velocity and vice versa.

The geometry and density of pores and cracks can significantly influence

the velocity of a material (Walsh, 1965; Popp and Kern, 1994). At the same

confining pressure, the distribution of stress on a spherical pore will differ from

the stress on a crack tip. A higher stress concentration will occur at the tips of the

long, narrow crack because the area is smaller and exceeds the elastic regime for

deformation (Figure 37). The stress concentrations are lower for the spherical

pore and at low pressures will behave elastically and therefore not close.

87



This phenomenon, known as crack closure, may explain the pressure

dependence on the P-wave and S-wave velocities observed in some of the basalts.

As the pressure increases during the experiment, the spherical voids are behaving

elastically, but cracks throughout the sample may close therefore decreasing the

volume fraction of pore space to solid rock and increasing the overall velocity.

The larger the ratio of cracks to spherical voids the greater the pressure

dependence on velocity.

Three-dimensional tomographic imaging of four basalts from the SFVF

and all nine basalts from LBNM were collected prior to the deformation

experiments. The pore structure can therefore be utilized in order to describe the

relationship between velocity and porosity as well. The compressional waves

(P-waves) travel along the axis of the sample and longitudinal waves (S-waves)

88



travel perpendicular to the axis of the sample, therefore analyzing the pore

structure in these two directions may explain their contribution to the overall

velocity and material properties.

Visual analysis of the pore shapes from the four basalt samples from SFVF

can give a general idea of the ratio of spherical pores to narrow cracks (Figure

38). Data from Table 10 shows that although there is little pressure dependence on

the P-wave velocity for any of the basalts, there is a pressure dependence on the

S-wave velocity from all samples but sample SFVF BPQ 5. This may correspond

to pore shape between samples. Figure 39 shows that the SFVF BPQ 5sample has

a high amount of larger spherical pores while the other three samples that have

tomographic data have smaller more irregular shaped pores combined with some

elongate pores.

89



90



Visual analysis of the pore shapes from the nine basalt samples from

LBNM can also give a general idea of the proportion of spherical pores to more

irregular pores and narrow cracks (Figure 38). Data from Table 12 shows that the

sample with the most significant pressure dependence on both the P-wave and

S-wave velocity is sample SKU IN 2. SKU IN 2 has the smallest visible pores and

may experience more crack closure than the basalts with pores of greater

diameter. SKU IN 1, SKU IN 3, SKU LB1 3, and SKU LB2 2, have a slight

pressure dependence lower than that of sample SKU IN 2 but higher than SKU

LB1 1, SKU LB1 2, and SKU LB2 3. Figure 39 shows that these four basalts have

a high density of pores but appear to have fewer pores with a large diameter. The

samples with the lowest pressure dependence, SKU LB1 1, SKU LB1 2, and SKU

LB2 3 have pores with significantly larger diameters and a higher density of these

large pores than the other basalts. A possible explanation for this variation in

pressure dependence is this difference in pore size and shape between samples.

The samples with irregular shaped pores of smaller diameter may experience

more crack closure and therefore have a higher pressure dependence on their

velocity measurements. Samples with pores of larger diameters may experience

less crack closure and have an overall lower pressure dependence on their velocity

measurements.

91



92



5.2: Clay Content

Although porosity is likely the parameter with the greatest control on

velocity, clay content in basalts may also play a significant role. In sandstones,

small amounts of clays can soften the solid rock matrix which in turn will

decrease velocities (Han et al., 1986). However, clays found in basalt will most

likely be a result of weathering and therefore not affect the overall structure of the

material. Clays will likely play a less significant role in the seismic signature of

basalt. Nevertheless, characterizing alterations could provide insight into some

sample variation. Clay content can be estimated by the density of a material. A

basalt with higher clay content should have a lower solid density than a basalt

with low clay content. Clay minerals are also measured by The Planetary

Environments Laboratory at NASA’s Goddard Space Flight Center (Table 17).

Density of basalts from the three study sites is calculated. There is a clear

relationship between density and porosity for all basalt samples (Figure 40). The

higher the density the lower the porosity and vice versa. The denser samples have

a higher volume fraction of solid rock to pore space and therefore the velocity of

the solid rock will have a greater contribution to the overall velocity. The less

dense samples will have a higher volume fraction of pore space and therefore

have a lower overall velocity.The linear regression of density against porosity for

basalts from the SFVF and LBNM have similar slopes while basalts from KH

have a steeper slope. This difference may be due to differences in the composition

of the individual basalts.

93



The density of basalts from all three study sites have a positive correlation

with P-wave velocity. For basalts from the SFVF and from LBNM, the correlation

is relatively weak, with R-squared values of 0.04 and 0.14. Basalts from KH have

a stronger correlation with an R-squared value of 0.74 (Figure 41). The density of

basalts from KH and LBNM have a positive correlation with the S-wave velocity

as well. The R-squared value for basalts from KH is 0.07 and for LBNM is 0.12.

Basalts from the SFVF have no correlation between density and S-wave velocities

with a low R-squared value of 0.02 (Figure 42).

94



95



The basalt sample, SFVF BPQ 5, from the SFVF is the only sample with

confirmed clays in its composition, but other basalts contain amorphous material

in their composition that may include clays (Table 17). The higher overall

densities associated with basalts from KH may be associated with the low clay

content found in some samples from this study site. The tighter clustering of

densities from KH may also be a result of lower compositional variations as

opposed to the greater spread of densities from basalts from the SFVF and

LBNM. Clay content in basalts at LBNM may also contribute to the large spread

of P-wave velocities and the overall lower S-wave velocities.

This study focused on general information about possible differences in

clay content due to sample locale. Because of that, not all samples were included

in the clay content quantification. More intensive clay content analysis may bring

further insights into the discrepancies between study locale.

96



5.3: Extrapolation of Apollo Velocity Data

The Apollo missions provide seismic data for the near subsurface of the

Moon. Although lunar seismic data is limited, seismometers placed on the lunar

surface during the Apollo missions provide estimates of the velocity of the near

lunar subsurface. The P-wave velocity ranges from 0.1 - 1.2 km/s at a depth

ranging from 0 - 300 meters (Watkins and Kovach, 1973; Heffels et. al., 2017).

The S-wave velocity ranges from 0.04 - 0.4 km/s at a depth ranging from 0 - 160

meters (Horvath et al., 1980; Imazato et al., 2023).

The P-wave velocities for basalt data from the three study sites plot in a

cluster, each with a distinct porosity relationship (Figure 43). The S-wave

velocities for basalt data from the three study sites plot in a similar manner. The

overall trend is consistent with data from this study- velocity decreases as a

function of porosity in basalts. This negative correlation is present in both P-wave

and S-wave velocities and the velocity-porosity relationship for each study site is

as follows:

A relationship between velocity and porosity can be determined using a

linear regression (Figure 43).

For the P-wave velocity:

SFVF: 𝑉
𝑝
 =  − 47ϕ +  6000;       𝑟2 = 0. 97 

KH: 𝑉
𝑝
 =  − 99ϕ +  5600;       𝑟2 = 0. 65 

LBNM: 𝑉
𝑝
 =  − 29ϕ +  5100;       𝑟2 = 0. 05 

LBNM (calc.): 𝑉
𝑝
 =  − 56ϕ +  4600;       𝑟2 = 0. 51 

97



Where is the P-wave velocity and is porosity.𝑉
𝑝

ϕ

For the S-wave velocity:

SFVF: 𝑉
𝑠
 =  − 26ϕ +  2900;       𝑟2 = 0. 66 

KH: 𝑉
𝑠
 =  − 69ϕ +  3400;       𝑟2 = 0. 48 

LBNM: 𝑉
𝑠
 =  − 22ϕ +  2400;       𝑟2 = 0. 14

LBNM (calc.): 𝑉
𝑝
 =  − 27ϕ +  2200;       𝑟2 = 0. 51 

Where is the P-wave velocity and is porosity.𝑉
𝑝

ϕ

Using these relationships, the Apollo seismic data, and lunar porosities, an

estimation of the amount of fracturing in the near lunar surface can be

extrapolated. Some estimates of lunar porosities range from ~35-55% in the near

surface (Heffels et. al., 2017). P-wave velocities for the near lunar subsurface

ranges from 0.1 - 1.2 km/s. Using the relationships derived from this study, the

porosity would be largely overestimated. One explanation for this discrepancy is

the presence of large-scale fracturing on the lunar surface. This fracturing, or

large-scale porosity, may contribute to lowering the overall seismic signature of

the lunar subsurface. For the shallow subsurface of the Moon an overall porosity

is estimated using the linear regression for basalts from KH and the calculated

values from LBNM. At depths of tens of meters, the overall porosity ranges from

47 - 78 %. At depths of a few hundred meters, the overall porosity ranges from

42 - 65 %. An average porosity at shallow depths for the lunar surface can be

estimated to be ~ 60%.

98



99



5.4: Field-scale Porosity Relationship

For the three study sites: the SFVF, KH, and LBNM, field data was

conducted by members of the GEODES field team in 2019, 2021, 2022, and

2023. The field data collected through active seismic source techniques indicates

a lower overall field-scale velocity for all three sites. For the SFVF the seismic

line provides a P-wave velocity of 0.5 - 3 km/s depending on depth. For KH the

P-wave velocity ranges from 0.5 - 2 km/s and for LBNM the P-wave velocity

ranges from 0.5 - 4 km/s. Using an average of these velocity values and the

relationship between velocity and porosity derived from the laboratory data in this

study, an estimate of the macro-scale porosity can be determined.

For the SFVF, the average field velocity is 1.8 km/s. Using the

velocity-porosity relationship from the SFVF laboratory data, the estimated

fracturing may be up to ~84%. The seismic line conducted at SFVF spans 100

meters. Based on these porosity estimates, the total amount of solid rock across

that region would be ~16 meters and the total amount of large-scale fracturing is

~84 meters (Figure 44).

For KH, the average field velocity is 1.3 km/s. Using the velocity-porosity

relationship from the KH laboratory data, the estimated fracturing is ~42%. The

seismic line conducted at KH spans 50 meters. Based on these porosity estimates,

the total amount of solid rock across that region would be ~30 meters and the total

amount of large-scale fracturing is ~20 meters (Figure 45).

For LBNM, the average field velocity is 2.2 km/s. Using the

velocity-porosity relationship from the calculated LBNM laboratory data, the

100



estimated fracturing is ~39%. The seismic line conducted at LBNM spans 80

meters. Based on these porosity estimates, the total amount of solid rock across

that region would be ~50 meters and the total amount of pore space/large-scale

fracturing is ~30 meters (Figure 46).

101



102



103



Chapter 6: Conclusions

Hydrostatic and axial deformation experiments were conducted in tandem

to velocity measurements for basalt samples from three terrestrial volcanic

regions used as lunar analog study sites: The San Francisco Volcanic Field,

Kilbourne Hole in New Mexico, and Lava Beds National Monument. For all three

sites, the P-wave velocity and S-wave velocity have a strong negative correlation

to porosity.

For the SFVF basalts P-wave and S-wave velocities range from ~5 km/s to

~6 km/s +/- 0.1 km/s and ~2.1 km/s to ~3 km/s +/- 0.1 k m/s, respectively. For the

KH basalts P-wave and S-wave velocities range from ~3.5 km/s to ~4.4 km/s +/-

0.1 km/s and ~1.8 km/s to ~2.3 m/s +/- 0.1 km/s, respectively. For LBNM basalts

P-wave and S-wave velocities range from ~3.1 km/s to ~6 km/s +/- 0.1 km/s and

~1.2 km/s to ~2.4 km/s +/- 0.1 km/s, respectively.

Relationships between porosity and both P-wave and S-wave velocities

were determined for basalts at all three sample locales. Those relationships are as

follows:

SFVF: sam𝑉
𝑝
 =  − 47ϕ +  6000;       𝑟2 = 0. 97 

KH: 𝑉
𝑝
 =  − 99ϕ +  5600;       𝑟2 = 0. 65 

LBNM: 𝑉
𝑝
 =  − 29ϕ +  5100;       𝑟2 = 0. 05 

LBNM (calc.): 𝑉
𝑝
 =  − 56ϕ +  4600;       𝑟2 = 0. 51 

Where is the P-wave velocity and is porosity.𝑉
𝑝

ϕ

104



SFVF: 𝑉
𝑠
 =  − 26ϕ +  2900;       𝑟2 = 0. 66 

KH: 𝑉
𝑠
 =  − 69ϕ +  3400;       𝑟2 = 0. 48 

LBNM: 𝑉
𝑠
 =  − 22ϕ +  2400;       𝑟2 = 0. 14

LBNM (calc.): 𝑉
𝑝
 =  − 27ϕ +  2200;       𝑟2 = 0. 51 

Where is the P-wave velocity and is porosity.𝑉
𝑝

ϕ

These relationships may be used in future analysis of lunar velocity data in

order to constrain the amount of fracturing in the subsurface of a region.

The field-scale seismic lines conducted in the three study sites indicate a

lower velocity of the subsurface, likely due to large-scale fractures and voids,

such as faults and caves.. Using the relationship found between velocity and

porosity, we estimate that the subsurface fractures and voids is ~84 vol% at the

San Francisco Volcanic Field, ~42 vol% at Kilbourne Hole, and ~39 vol% at Lava

Beds National Monument.

This study may provide physical properties of materials to use in future

modeling work as well as a tangible relationship between velocity and porosity to

be used in understanding the large-scale structure of a lunar study site. Data from

this study suggests velocities of basaltic lunar analog material will depend largely

on the structural properties of the sample locale, and characterization of these

properties will aid in a better understanding of the lunar subsurface.

105



Appendix

A1: Table of all nomenclature and symbols used in this study.

106



A2: Table of porosity measurements for all samples.

107



A3: Table of all calculated elastic moduli using a range of Poisson Ratio (from
0.1 to 0.45).

108



A4: Metadata for CT scans.

University of Texas High-Resolution X-ray CT Facility:

Specimens scanned by Matt Colbert.

16bitTIFF Scan parameters:

● NSI scanner
● Fein Focus High Power source
● 140 kV
● 0.11 mA
● aluminum foil filter
● Perkin Elmer detector
● 0.25 pF gain
● 1 fps
● 1x1 binning
● source to object
● 148.49 mm
● source to detector 1456.85 mm
● continuous CT scan
● 2 frames averaged
● 0 skip frames
● 3000 projections
● 5 gain calibrations
● 5 mm calibration phantom
● data range [-10, 151] adjusted grayscale values
● beam-hardening correction = 0.2
● Voxel size = 20.4 μm
● Total slices = 1894.
● 8bitJPG_FOR_VIEWING_ONLY: 8bit jpg version of the above images

109



A5: Metadata for XRD analysis.

XRD information:

● Bruker D8 Discover 2-70 °2theta
● 40kV
● 40mA

110



A6: Additional velocity data for basalts at the SFVF.

111



112



113



114



A7: Additional velocity data for basalts from KH.

115



116



117



A8: Additional velocity data for basalts from LBNM.

118



119



120



A9: Unloading data for basalts from all three study sites.

121



122



123



A10: Weight and lithologies of samples collected during the duration of the
Apollo missions. Data collected from Taylor, 1975.

Apollo
11

Apollo
12

Apollo
14

Apollo
15

Apollo
16

Apollo
17

Total weight
collected (kg)

21.6 34.3 42.3 77.3 95.7 110.5

Lithology
(% of total weight)

Rocks (> 10 mm) 44.9 80.6 67.3 74.7 72.3 65.9

Anorthosite - - - 0.4 10.5 0.5

Basalt 19.9 52.2 9.1 37.9 - 29.1

Dolerite (coarse
basalt)

- 26.3 - - - -

Other igneous rocks 2.1 0.2 - - - 4

Breccia 22.9 1.9 58.2 34.1 36.8 32.3

Impact melt - - - 2.3 25 -

Fines (<10 mm) 54.6 16.8 30.6 17 19.3 26.7

Cores 0.4 1.2 0.9 6 7.4 6.6

Other - 1.4 1 2.4 1.2 0.9

Total 99.9 100 99.8 100.1 100.2 100.1

124



A11: Additional data

Additional measurements are available online at the Digital Repository at the
University of Maryland (DRUM).

125



Bibliography

Angelis, D.G., Wilson, J.W., Clowdsley, M.S., Nealy, J.E., Humes, .H., Clem,

J.M. (2002), Lunar Lava Tube Radiation Safety Analysis, Journal of

Radiation Research, Volume 43, Issue Suppl, Pages S41–S45,

https://doi.org/10.1269/jrr.43.S41.

Bell, E. (2021) Geophysical Exploration of Terrestrial and Lunar Volcanic

Fields. [Doctoral Dissertation, The University of Maryland].

Bell, E., Schmerr, N., Porter, R., Bleacher, J., Young, K., Huang, M. H., Lekic,

V., Pettit, D. (2022). Active seismic exploration along a human lunar

mission traverse analogue in the San Francisco volcanic field. The

Leading Edge. 41. 690-699. doi:10.1190/tle41100690.1.

Bender, P.L., Currie, D.G., Dicke, R.H., Eckhardt, D.H., Faller, J.E., Kaula,

W.M., Mulholland, J.D., Plotkin, H.H., Poultney, S.K., Silverberg,

E.C., Wilkinson, D.T., Williams, J.G., Alley, C.O. (1973). The Lunar

Laser Ranging Experiment: Accurate ranges have given a large

improvement in the lunar orbit and new selenophysical information.

Science, 182(4109): 229-238, doi: 10.1126/science.182.4109.229.

PMID 17749298. S2CID 32027563.

Benz, W., Cameron, A.G.W., Melosh, H.J. (1989), The origin of the Moon and the

single-impact hypothesis III. Icarus, (81)1, 113-13,

https://doi.org/10.1016/0019-1035(89)90129-2.

126

https://doi.org/10.1269/jrr.43.S41
http://dx.doi.org/10.1190/tle41100690.1
https://doi.org/10.1126%2Fscience.182.4109.229
https://en.wikipedia.org/wiki/PMID_(identifier)
https://pubmed.ncbi.nlm.nih.gov/17749298
https://en.wikipedia.org/wiki/S2CID_(identifier)
https://api.semanticscholar.org/CorpusID:32027563
https://doi.org/10.1016/0019-1035(89)90129-2


Besserer, J., Nimmo, F., Wieczorek, M., Weber, R., Kiefer, W., McGovern, P.,

Andrews-Hanna, J., Smith, D., and Zuber, M. (2014), GRAIL gravity

constraints on the vertical and lateral density structure of the lunar crust,

Geophys. Res. Lett., 41, 5771–5777, doi:10.1002/2014GL060240.

Borgomano, J. V. M., Fortin, J., Guéguen, Y. (2019). Cracked, Porous Rocks and

Fluids: Moon and Earth Paradox. Minerals 9, 693.

https://doi.org/10.3390/min9110693.

Briaud, A., Ganino, C., Fienga, A., Memin, A., Rambaux. (2023), The lunar solid

inner core and the mantle overturn. Nature, (617), 743–746,

https://doi.org/10.1038/s41586-023-05935-7.

Canup, R., Asphaug, E. (2001), Origin of the Moon in a giant impact near the end

of the Earth's formation. Nature, 412, 708–712,

https://doi.org/10.1038/35089010.

Canup, R.M. (2012), Forming a Moon with an Earth-like Composition via a Giant

Impact. Science, (338)6110, 1052-1055, DOI:10.1126/science.1226073.

Carlson, R. L. (2014). The effects of alteration and porosity on seismic

velocities in oceanic basalts and diabases, Geochem. Geophys.

Geosyst., 15, 4589–4598, https://doi.org/10.1002/2014GC005537.

Chappell, B.W., Green, D.H. (1973). Chemical composition and petrogenetic

relationships in Apollo 15 mare basalts. Earth and Planetary Science

Letters. Vol. 18, 2, 237-246,

https://doi.org/10.1016/0012-821X(73)90062-9.

127

https://doi.org/10.1002/2014GL060240
https://doi.org/10.3390/min9110693
https://doi.org/10.1038/s41586-023-05935-7
https://doi.org/10.1038/35089010
https://doi.org/10.1126/science.1226073
https://doi.org/10.1002/2014GC005537
https://doi.org/10.1016/0012-821X(73)90062-9


Chung, D.H. (1973). Elastic wave velocities in anorthosite and anorthositic

gabbros from Apollo 15 and 16 landing sites. Lunar and Planetary

Science Conference Proceedings. Vol. 4, 2591,

https://ui.adsabs.harvard.edu/abs/1973LPSC....4.2591C.

Coombs, C.R., and Hawke, R.B. (1992), A search for Intact Lava Tubes on the

Moon: Possible lunar base habitats. The Second Conference on Lunar

Bases and Space Activities of the 21st Century (2nd ed., Vol 1). NASA.

Johnson Space Center.

Cruikshank, D.P., and Wood, C.A. (1972). Lunar rilles and Hawaiian Volcanic

Features: Possible Analogues. The Moon 3, 412–447

https://doi.org/10.1007/BF00562463.

Esmaeili, S., Kruse, S., Jazayeri, S., Whelley, P., Bell, E., Richardson, J.,

Garry, W. B., Young, K. (2020). Resolution of Lava Tubes with ground

penetrating radar: The TubeX project. Journal of Geophysical

Research: Planets, 125, e2019JE006138.

https://doi.org/10.1029/2019JE006138.

Gawronska A.J., McLeod, C.L., Blumenfeld, E.H., Hanna, R.D., Zeigler, R.A.

(2002). New interpretations of lunar mare basalt flow emplacement

from XCT analysis of Apollo samples. Icarus, Volume 388, 115216,

ISSN 0019-1035, doi: https://doi.org/10.1016/j.icarus.2022.115216.

Gornitz, V. (1973). The Origin of Sinuous Rilles. The Moon, 6, 337-356.

https://doi.org/10.1007/BF00562210.

128

https://ui.adsabs.harvard.edu/abs/1973LPSC....4.2591C
https://doi.org/10.1007/BF00562463
https://doi.org/10.1029/2019JE006138
https://doi.org/10.1016/j.icarus.2022.115216
https://doi.org/10.1007/BF00562210


Greeley, R. (1971). Lava Tubes and Channels in the Lunar Marius Hills. The

Moon, 3(3), 289-314. https://doi.org/10.1007/BF00561842.

Greeley, R. (1987) The Role of Lava Tubes in Hawaiian Volcanoes. In Decker

RW, Wright TL, Stauffer PH (eds) Volcanism in Hawaii. US Geol Surv

Prof Pap 1350:1589-1602.

Groemer, G., Ozdemir, S. (2020). Planetary Analog Field Operations as a

Learning Tool. Frontiers in Astronomy and Space Sciences. Vol. 7.

https://www.frontiersin.org/articles/10.3389/fspas.2020.00032.

Gueguen, Y., Palciauskas, V. (1994). Introduction to the Physics of Rocks.

Princeton, New Jersey. Published by Princeton University Press.

Han, D., Nur, A., Morgan, D. (1986). Effects of porosity and clay content on

wave velocities in sandstones. Geophysics 1986;; 51 (11): 2093–2107.

doi: https://doi.org/10.1190/1.1442062.

Hargreaves, R.B., Buddington, A.F. (1970). Analogy between anorthosite

series on the Earth and Moon. Icarus. Vol. 13, Issue 3, 371-382, doi:

https://doi.org/10.1016/0019-1035(70)90087-4.

Harold, C.U. (1969). Origin and History of the Moon, Bulletin of the Atomic

Scientists. 25:7, 45-51, doi: 10.1080/00963402.1969.11455251.

Hartmann, W.K., Davis, D.R. (1975), Satellite-Sized Planetesimals and Lunar

Origin. Icarus, (24)4, 504-515,

https://doi.org/10.1016/0019-1035(75)90070-6.

129

https://doi.org/10.1007/BF00561842
https://www.frontiersin.org/articles/10.3389/fspas.2020.00032
https://doi.org/10.1190/1.1442062
https://doi.org/10.1016/0019-1035(70)90087-4
https://doi.org/10.1080/00963402.1969.11455251
https://doi.org/10.1016/0019-1035(75)90070-6


Haruyama, J., Hioki, K., Shirao, M., Morota, T., Hiesinger, H., van der Bogert,

C.H., Miyamoto, H., Iwasaki, A., Yokota, Y., Ohtake, M., Matsunaga,

T., Hara, S., Nakanotani, S., Pieters, C.M. (2009). Possible Lunar Lava

Tube Skylight Observed by SELENE Cameras, Geophys. Res. Lett., 36,

L21206, https://doi.org/10.1029/2009GL040635.

Haruyama, J., Morota, T., Kobayashi, S., Sawai, S., Lucey, P.G., Shirao, M.,

Nishino, M.N. (2012). Lunar Holes and Lava Tubes as Resources for

Lunar Science and Exploration. In: Badescu, V. (eds) Moon. Springer,

Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-27969-0_6.

Hawley, J. C. (1981). Pleistocene and Pliocene history of international

boundary area, southern New Mexico, in Hoffer, J.M., and Hoffer,

R.L., eds., Geology of the border, southern New Mexico and northern

Chihuahua: El Paso Geological Society, p. 26-32.

Heffels, A., Knapmeyer, M., Oberst, J., Haase, I. (2017). Re-evaluation of Apollo

17 Lunar Seismic Profiling Experiment data. Planetary and Space

Science, Volume 135, Pages 43-54, ISSN 0032-0633,

https://doi.org/10.1016/j.pss.2016.11.007.

Heiken, G.H., Vaniman, D.T., French, B.M. (1991), Lunar Sourcebook, A User's

Guide to the Moon. Cambridge University Press.

Horvath, P., Latham, G. V., Nakamura, Y., and Dorman, H. J. (1980). Lunar

near-surface shear wave velocities at the Apollo Landing Sites as inferred

from spectral amplitude ratios. J. Geophys. Res., 85(B11), 6572–6578,

https://doi.org/10.1029/JB085iB11p06572.

130

https://doi.org