University of Southern California Dissertations and Theses

3D fossil visualization and mapping of the La Brea Tar Pits, Los Angeles, California

(USC Thesis Other)

3D fossil visualization and mapping of the La Brea Tar Pits, Los Angeles, California

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content 1

3D Fossil Visualization and Mapping of the La Brea Tar Pits, Los Angeles, California

by

Kristiane Hill

A Thesis Presented to the
Faculty of the USC Graduate School
University of Southern California
In Partial Fulfillment of the
Requirements for the Degree
Master of Science
(Geographic Information Science and Technology)

December 2018

2

Copyright © 2018 by Kristiane Hill

3

To my thesis advisor, Dr. Jennifer Swift.

4

Table of Contents
List of Figures ................................................................................................................................. 6
List of Tables .................................................................................................................................. 8
Acknowledgments........................................................................................................................... 9
List of Abbreviations .................................................................................................................... 10
Abstract ......................................................................................................................................... 11
Chapter 1 Introduction .................................................................................................................. 12
1.1. Project Overview ..............................................................................................................13
1.2. Motivation .........................................................................................................................14
1.3. Methodological Overview ................................................................................................15
1.4. Thesis Structure ................................................................................................................16
Chapter 2 Background and Related Work .................................................................................... 17
2.1. La Brea Tar Pits Background ............................................................................................17
2.1.1. La Brea Tar Pits Background ...................................................................................17
2.1.2. Fossil Collection Procedures....................................................................................18
2.1.3. Kacey Pham’s Thesis ...............................................................................................19
2.2. Review of Existing 3D Modeling and Visualization Techniques in Archaeology ...........20
2.3. Forensics Background .......................................................................................................26
Chapter 3 Methodology ................................................................................................................ 28
3.1. Method Overview .............................................................................................................28
3.2. Data Sources .....................................................................................................................29
3.3. Data Background ..............................................................................................................29
3.3.1. Box and Grid Data ...................................................................................................29
3.3.2. Fossil Data ...............................................................................................................31
3.4. 3D Visualization Development Approach ........................................................................32
5

3.4.1. Fossil Diagramming .................................................................................................33
3.4.2. Coordinate Transformation in Excel and ArcGIS Pro .............................................34
3.4.3. Box 13a and Grid B-3 Creation using Excel, Adobe Reader and ArcGIS Pro ........45
3.4.4. Fossil Digitization in ArcGIS Pro ............................................................................49
3.4.5. Sources of Error .......................................................................................................51
3.5. Chapter Summary .............................................................................................................52
Chapter 4 Results .......................................................................................................................... 53
4.1. Alternative 3D Modeling Testing and Results..................................................................53
4.1.1. ArcGIS Pro Testing..................................................................................................53
4.1.2. SketchUp Pro, Microsoft 3D Builder and Microsoft 3D Paint Testing ...................55
4.2. 3D Modeling Results ........................................................................................................56
4.3. Chapter Summary .............................................................................................................61
Chapter 5 Discussion and Conclusion .......................................................................................... 62
5.1. Discussion .........................................................................................................................62
5.2. Future Work ......................................................................................................................63
5.2.1. Improving the Import Process of 3D Objects to ArcGIS Pro ..................................63
5.2.2. Museum Exhibit .......................................................................................................64
5.2.3. 3D Data Generation and Archive .............................................................................64
5.2.4. Potential Application for Forensic Archaeology .....................................................65
5.3. Next Steps .........................................................................................................................65
References ..................................................................................................................................... 67
Appendix A ArcGIS Pro Version 2.1. Steps................................................................................. 70
Appendix B Hand-Drawn and PowerPoint Fossil Diagrams ...................................................... 100

6

List of Figures
Figure 1. La Brea Tarpits Location Map ...................................................................................... 14

Figure 2. Project 23 fossil deposits being crated and lifted from the excavation site ................... 18

Figure 3. String laid across grid in early excavation of a single pit at the La Brea Tar Pits ........ 19

Figure 4. North House of Caecilius Iucundus 3D modeled in Esri ArcScene for analysis .......... 22

Figure 5. 3D model of Paliambela Kolindros in Esri ArcScene ................................................... 24

Figure 6. Widget for viewing 3D models of samples in ArcGIS Web App ................................. 26

Figure 7. The grid layout of Box 13a with Grid B-3 highlighted in green ................................... 30

Figure 8. Northing (N), Westing (W), and Below Depth (BD) measurements ............................ 31

Figure 9. Fossil P23 33245 saber-toothed cat coordinate locations.............................................. 32

Figure 10. Example of a hand-drawn diagram and the corresponding PowerPoint diagram of
fossil P23 33245 ............................................................................................................................ 34

Figure 11. Box 13a central coordinate at 0 cm W, 50 cm N on Grid B-3 .................................... 35

Figure 12. Pythagorean theorem diagram ..................................................................................... 36

Figure 13. Distance equation example .......................................................................................... 36

Figure 14. Tangent and arctangent diagram ................................................................................. 37

Figure 15. Atan2 diagram ............................................................................................................. 38

Figure 16. Bearing equation example ........................................................................................... 39

Figure 17. Example of 360° rotation added to negative bearing .................................................. 40

Figure 18. Bearing Distance to Line and Feature Vertices to Points tools results in ArcGIS
Pro ................................................................................................................................................. 44

Figure 19. Transformed coordinates below ground surface in ArcGIS Pro ................................. 45

Figure 20. Project 23 locality map with an arrow indicating Box 13a ......................................... 46

Figure 21. Box 13a rotated and centered over Box 13a’s central coordinate ............................... 47

7

Figure 22. Grid B-3 Level 2 digitized over Box 13a’s central coordinate at 0 cm W,
50 cm N on the grid ...................................................................................................................... 48

Figure 23. Box 13a and Grid B-3 Levels 1 and 2 below ground surface ..................................... 48

Figure 24. Fossil P23 33245 manually added to the 3D Scene .................................................... 50

Figure 25. Fossil P23 33245 manually placed below ground surface using the Move tool
prior to the Scale and Rotate tools ................................................................................................ 50

Figure 26. Fossil P23 33245 manually aligned with corresponding coordinates ......................... 51

Figure 27. All fossils aligned with their respective coordinates ................................................... 57

Figure 28. North-facing view of all fossils in Grid B-3 Levels 1 and 2 ....................................... 57

Figure 29. North-facing view of all fossils in Box 13a................................................................. 58

Figure 30. View from below of the fossils in Box 13a beneath its central coordinate ................. 58

Figure 31. Fossil P23 33242 clicked on to view attribute information ......................................... 59

Figure 32. Fossil P23 33252 coordinate alignment error .............................................................. 60

Figure 33. Fossils extending outside Grid B-3 Levels 1 and 2 ..................................................... 60

8

List of Tables
Table 1. Example data for fossil P23 33245 from CoordinateTransformation.xlsx ..................... 41

Table 2. Example data for fossil P23 33245 from
AllBone_DataForCoordinateTransformation.xlsx to be imported into ArcGIS Pro .................... 43

9

Acknowledgments
I am grateful for the data provided to me by The La Brea Tar Pits and Museum and the
enthusiastic assistance provided by Carrie Howard, Dr. Emily Lindsey and Aisling Farrell. I
would also like to thank Dmitri Bagh from Safe Software, Inc., Brian Sims from Esri and Kacey
Pham. Thanks especially to my thesis advisor, Dr. Jennifer Swift, for guiding me through the
process of developing this thesis project. I would like to thank my thesis committee, Dr. An-Min
Wu and Dr. Laura Loyola, for their knowledge and encouragement. I would also like to thank
my boyfriend, Kevin Brown, as well as my family and friends for their unwavering support.

10

List of Abbreviations
2D Two-dimensional
3D Three-dimensional
3MF 3D Manufacturing Format
COLLADA COLLAborative Design Activity
DPI Dots per inch
DEM Digital elevation model
DSM Digital surface model
GIS Geographic Information System
GIST Geographic Information Science and Technology
JPG Joint Photographic Experts Group
OBJ Wavefront OBJ
P23 Project 23
PDF Portable Document Format
SSI Spatial Sciences Institute
VRML Virtual Reality Modeling Language

11

Abstract
The La Brea Tar Pits and Museum in the middle of Los Angeles, California is a paleontological
marvel containing numerous fossil-rich asphalt deposits. Until very recently, the museum only
recorded their findings in non-spatial databases. As a continuation of work completed by a
former USC Geographic Information Science and Technology student to develop spatial
databases documenting artifacts for the museum, the main objective of this thesis project was to
create a methodology for visualizing fossils as high-resolution 3D objects on a 3D map in their
pre-excavation, in-situ locations. Museum scientists selected nineteen fossils from one asphalt
deposit for mapping. The fossils were laser scanned by museum scientists, and the resulting 3D
objects were provided for this project with accompanying locality data gathered in the manner of
a traditional paleontological dig. The data required extensive processing prior to importing the
3D objects into a GIS, including image file conversion, location and orientation diagramming
and steps for coordinate transformation from paleontological location measurements to real-
world, geographic coordinates. The 3D objects were then imported and manually positioned in a
3D GIS map beneath the earth’s surface. The resulting 3D model provides an interactive, GIS-
enabled visualization of the nineteen fossils in their original locations and orientations prior to
excavation. It is intended that this project support future research efforts of the museum scientists
in spatial analysis and modeling of fossils and substrate (tar pits), ultimately to improve our
understanding of Ice Age animals and the environments in which they lived and died. In
addition, the results of this project serve as an example application of 3D GIS capabilities that
can support forensic archaeology, an important tool in intelligence and criminal investigations.
Lastly, it is anticipated that the georeferenced 3D objects, as well as this 3D fossil visualization,
may become part of an interactive museum exhibit in the future.
12

Chapter 1 Introduction
The La Brea Tar Pits and Museum is home to over three million Ice Age fossils. Located in Los
Angeles, California, it is a unique onsite museum surrounded by active, urban excavation sites.
The museum displays fossil specimens extracted from 10,000 to 40,000-year-old asphalt seeps or
“tar pits” that include a variety of plants and animals from saber-toothed cats to mammoths (La
Brea Tar Pits and Museum 2017a). Until recently, the location of the tar pits had not been
recorded beyond a single hand-drawn, paper map maintained since 1913. In 2015, University of
Southern California Geographic Information Science and Technology (GIST) graduate student,
Kacey Pham, digitally georeferenced the paper map of the tar pits and created fossil
geodatabases as part of her Master’s thesis (Pham 2015, xiv). As a continuation of Pham’s
research, the main objective of this thesis project was to develop a methodology to visualize
fossils as digital high-resolution 3D objects displayed on a 3D map using Geographic
Information System (GIS) technology in their pre-excavation, in-situ locations beneath the
earth’s surface. It is anticipated that a spatial visualization methodology using GIS will support
the museum scientists’ research regarding the interaction between Ice Age animals, humans and
environments (Lindsey, pers. comm.).
This chapter describes this thesis project’s scope, motivation and methodology for
developing a 3D fossil visualization in a GIS environment for the La Brea Tar Pits and Museum.
Section 1.1 provides an overview of the project and section 1.2 discusses the motivation for
creating the 3D visualization. Section 1.3 outlines the methodology for developing the 3D model,
while section 1.4 summarizes the subsequent chapters in this manuscript.
13

1.1. Project Overview
This thesis study consisted of the development of a methodology for mapping high-
resolution 3D objects of fossils excavated from the La Brea Tar Pits (La Brea Tar Pits and
Museum 2017a). Museum scientists selected nineteen fossils for this project from the asphalt
deposit named Box 13a in the excavation site known as Project 23 (Figure 1). The fossils were
manually laser scanned and shared with the author, including the resulting 3D digital objects and
textures as well as supporting documentation regarding the location of the fossils at the time of
extraction (La Brea Tar Pits and Museum 2017b). A texture is herein defined as “an image used
to define the characteristics of a surface” (Graham 2003). With additional data from Pham’s
Master’s thesis, the geodatabase created for this project includes GIS feature classes representing
the fossils, their location coordinates, Box 13a and its grids. In its current stage of development,
the geodatabase can be utilized in Esri ArcGIS Pro version 2.1 (Esri ArcGIS Pro 2.1 2018) and
in other GIS software which includes the capability of reading or importing Esri’s native
geodatabase format (Esri ArcGIS Desktop 2018).
14

Figure 1. La Brea Tarpits Location Map (Pham 2015, 1)
1.2. Motivation
The main motivation for the development of the 3D fossil visualization in a GIS
environment is to support the research conducted by the La Brea Tar Pits and Museum scientists.
Mapping the fossils in 3D will allow the scientists to perform spatial analysis that may inform
how Ice Age animals, humans and environments interacted and impacted each other (Lindsey,
pers. comm., Howard and Farrell, pers. comm.). Spatial patterns found in the fossil distribution
within an asphalt deposit may also provide insight regarding the fluid dynamics of the substrate
or even the cause of death of a given animal.
15

The secondary motivation for this project is to provide museum scientists with a 3D fossil
visualization that may eventually be exported to another software to be viewed interactively in
the museum. An interactive display would include a touchscreen with the 3D model that can be
zoomed and rotated. Ideally, viewers would be able to tap on a fossil on the screen and see its
attribute information such as species and body part (Lindsey, pers. comm.). This style of
interactive viewing would help guests better visualize where and in what orientation the fossils
are found prior to excavation.
1.3. Methodological Overview
The Esri geodatabase containing the 3D model was created using Esri ArcGIS Pro
version 2.1 (Esri ArcGIS Pro 2.1 2018). ArcGIS Pro was chosen for this project because of its
widespread use in industry and its advanced 3D visualization capabilities (G2 Crowd 2018).
Extensive data preparation outside of ArcGIS Pro was performed using Microsoft 3D Builder,
Microsoft PowerPoint 2016, Microsoft Excel 2016 and Adobe Reader XI (Microsoft 3D Builder
2013, Microsoft PowerPoint 2016 2016, Microsoft Excel 2016 2016, Adobe Reader XI 2018).
Prior to constructing the 3D model in ArcGIS Pro, the author created hand-drawn
diagrams to better understand the placement and orientation of the fossils, provided in Appendix
B. The 3D objects representing the fossils were then incorporated into similar diagrams created
in PowerPoint to provide a digital 3D visual. This intermediate visual allowed museum scientists
to validate the spatial location and orientation of each 3D object before proceeding with import
into a GIS (Howard and Farrell, pers. comm.). With this back-to-basics approach, the author was
able to confidently move forward with transforming the location information of the fossils
provided by the museum scientists into geographically referenced coordinates using a
combination of Excel and ArcGIS Pro. Next using ArcGIS Pro, empty 3D feature classes
16

representing Box13a and the grids within it were created and placed in a 3D map. The fossil 3D
objects were then imported for manual placement (location and orientation) in their pre-
excavation, in-situ locations beneath the ground surface.
1.4. Thesis Structure
The remainder of this thesis manuscript is divided into four chapters. Chapter 2,
Background and Related Work, discusses background information regarding the La Brea Tar Pits
and Museum history, current excavation procedures and previous GIS work completed for
Pham’s Master’s thesis (Pham 2015). Chapter 2 also describes existing 3D modeling and
visualization techniques used in archaeology as well as excavation techniques used in forensic
archaeology. Chapter 3, Methodology, outlines the steps to develop the methodology for the 3D
fossil visualization including data preparation, coordinate transformation, and 3D modeling.
Chapter 4, Results, presents the experiences testing alternative methodologies and software to
build the 3D visualization as part of this thesis work as well as the outcomes of the developed
methodology. Chapter 5, Discussion and Conclusion, summarizes the successes and challenges
of the project and proposes future work regarding further development of the 3D model.

17

Chapter 2 Background and Related Work
This chapter provides an overview of the background of the La Brea Tar Pits and Museum and
describes the related work reviewed prior to developing a methodology for constructing the 3D
fossil visualization. Section 2.1 gives an overview of the history of the La Brea Tar Pits and
Museum, its fossil collection protocol and Pham’s previous Master’s thesis work. Section 2.2
describes the related 3D modeling and visualization case studies in GIS that were reviewed to
inform this project’s methodology. Section 2.3 describes excavation practices in the field of
forensic archaeology, which employs investigative field and laboratory practices analogous to
techniques at the La Brea Tar Pits and Museum.
2.1. La Brea Tar Pits Background
This section provides background information on the La Brea Tar Pits and Museum,
current fossil collection procedures and previous GIS Master’s thesis work completed by Kacey
Pham (2015). Section 2.1.1 describes the history of the La Brea Tar Pits including Project 23, of
which Box 13a is a part. Section 2.1.2 reviews current tar pit excavation methods practiced by
museum scientists. Section 2.1.3 summarizes information from Kacey Pham’s Master’s thesis
relevant to this study (Pham 2015).
2.1.1. La Brea Tar Pits Background
The first major excavations at the La Brea Tar Pits and Museum occurred from 1913 to
1915 and resulted in more than 750,000 Ice Age plant and animal specimens of various sizes,
including saber-toothed cats, mammoths and dire wolves. Since initial excavations, the La Brea
Tar Pits and Museum collection has grown extensively as museum scientists continue to
excavate fossils unearthed as recently as 2006 (La Brea Tar Pits and Museum 2017). Known
18

collectively as Project 23, the 23 fossil-rich asphalt deposits unearthed in 2006 were discovered
during the construction of an underground parking lot for the nearby Los Angeles County
Museum of Art. The deposits were boxed up in wooden crates, labeled and relocated to the La
Brea Tar Pits and Museum grounds to be processed by the museum scientists (Figure 2).
Latitude, longitude and depth measurements were recorded prior to deposit extraction (Pham
2015, 11). This thesis project utilized data from Project 23’s deposit Box 13a which is currently
undergoing excavation (Howard and Farrell, pers. comm.).

Figure 2. Project 23 fossil deposits being crated and lifted from the excavation site
(Turner 2006, 9)
2.1.2. Fossil Collection Procedures
From 1913 to the present, fossil locations at the La Brea Tar Pits and Museum have been
recorded manually with pencil and paper. A meter-by-meter grid pattern made of string is spread
across the asphalt deposit and multiple x, y and z locations are recorded for each fossil using a
ruler (Figure 3). This style of data collection may lead to small (cm) errors due to the fact that
19

locations are measured and recorded by hand. The string may also be placed improperly or
loosen over time, further impacting data collection accuracy (Pham 2015, 14-15).

Figure 3. String laid across grid in early excavation of a single pit at the La Brea Tar Pits (Pham
2015, 14, Shaw 1982, 71)
2.1.3. Kacey Pham ’s Thesis
As mentioned in Chapter 1, the inspiration for this thesis was derived from previous work
completed by former GIST program student Kacey Pham in 2015. In her thesis entitled “GIS
Data Curation and Web Map Application for La Brea Tar Pits Fossil Occurrences in Los
Angeles, California,” Pham developed a fossil excavation spatial database by digitizing a
manually maintained historical paper map into a GIS-friendly spatial dataset, also joining data
from the museum’s existing, non-spatial KE EMu database to this to create a feature-rich
geodatabase. She then published the newly created geodatabase to a unique web GIS application
that she customized for the museum as a proof of concept for future GIS work, such as data
sharing in real-time (Pham 2015, xiv-2). As part of the future recommendations, Pham suggested
20

the development of a 3D fossil visualization so that museum visitors could better comprehend
the vast collection of fossils contained in the tar pits (Pham 2015, 67). Pham’s project and
recommendations laid the groundwork for this thesis project.
2.2. Review of Existing 3D Modeling and Visualization Techniques in
Archaeology
Prior to developing the methodology for this study, a thorough literature review of
existing 3D modeling and visualization techniques related to archaeology revealed GIS to be an
up-and-coming software for documenting and representing excavation sites in 3D. In the context
of this thesis, 3D modeling refers to the construction of a 3D visualization and, thus, the terms
“3D model” and “3D visualization” are used interchangeably throughout this document. At the
Pompeii Archaeological area in Italy, researchers from the University of Bologna created a
pipeline for the Superintendence of Pompeii to follow when generating and managing 3D models
of artifacts prior to import into a 3D GIS environment (Apollonio, Gaiani, and Benedetti 2012,
1271). As a case study, they used a combination of close-range photogrammetry and laser
scanning to 3D model 13 artifacts from the area that had been previously excavated (discovered)
but remained in-situ. They then performed quality control testing on their results as well as post-
collection data processing in OpenSceneGraph. Their efforts resulted in a methodology for the
generation of 3D base elements required to develop a large archaeological site like Pompeii in a
3D GIS (Apollonio, Gaiani, and Benedetti 2012, 1271-1275). Their work is similar to this thesis
project through the use of laser scanning to digitally capture and visualize (build a 3D model of)
their artifacts but differs in that they did not attempt to bring their 3D models into a GIS
environment.
21

In the same region but separate from the University of Bologna, researchers from Lund
University working under the Swedish Pompeii Project also performed a case study with the goal
of designing a digital 3D model of the North House of Caecilius Iucundus in Pompeii
(Campanaro, Landeschi, Dell’Unto, and Touati 2016, 321). The artifacts retrieved from this
archeological site were laser scanned, then the resulting 3D model was processed using
AutoDesk 3D Studio Max and split into architectural sub-elements. For example, sub-elements
consisted of planes or façades, which represent structure walls. The researchers then created
horizontal clones, or base maps, of the vertical facades and imported them into Esri ArcMap to
generate 2D thematic maps, analyzing factors such as decay and tilting. Using reference planes
for each façade, the researchers were also able to successfully import vertical façades into Esri
ArcScene while projecting the newly created thematic maps onto the maps vertically (in 3D).
Ultimately, they successfully created a digital 3D visualization of the North House of Caecilius
Iucundus in ArcScene (Figure 4) (Campanaro, Landeschi, Dell’Unto, and Touati 2016, 323-331).
This work is similar to this thesis project because the researchers sought to create a 3D model of
a portion of an excavation site and they used laser scanning to record the 3D artifacts excavated
from the site. However, their work differs in that they focused on 3D modeling and visualizing
larger archaeological, man-made building or construction elements (i.e., structures, walls) up to
3 m tall, while this project focuses on smaller-scale natural, non-human fossils ranging from
10-20 cm in length.
22

Figure 4. North House of Caecilius Iucundus 3D modeled in Esri ArcScene for analysis
(Campanaro, Landeschi, Dell’Unto, and Touati 2016, 330)
In another large-scale archaeological effort, researchers at Ghent University, Belgium
applied image-based 3D modeling in the recording of the Boudelo-2 excavation site as a case
study in recording an entire excavation site in 3D (De Reu, De Smedt, Herremans, Van
Meirvenne, Laloo, and De Clercq 2014, 251-252). Traditional excavation methods were adapted
to accommodate the 3D data collection. High-quality photographs were taken along with ground
control points on site at various times during the excavation. The 3D data were then processed in
Photoscan 0.9.0 and exported in various formats including both horizontal and vertical ortho-
images, digital surface models (DSMs) and Virtual Reality Modeling Language (VRML) files
(De Reu, De Smedt, Herremans, Van Meirvenne, Laloo, and De Clercq 2014, 253-254).
Researchers analyzed ortho-images and DSMs in ArcMap but were unable to import the VRML
files containing the 3D models into ArcScene because the software did not support it. They were
only able to import low-resolution versions of the 3D models which resulted in the loss of the
23

high-resolution information present in the ortho-images and DSMs. The researchers concluded
that image-based 3D modeling could transform archaeological excavation practices in the future,
but that 3D GIS technology limitations at the time of the study would not allow the full digital
potential of 3D models to be realized (De Reu, De Smedt, Herremans, Van Meirvenne, Laloo,
and De Clercq 2014, 260-261). This thesis project mirrors their work in that they attempted to
import VRML 3D object files into Esri ArcGIS software. In fact, one of the major challenges of
this thesis project which still remains is the importing of the fossil 3D objects in VRML format
into ArcGIS Pro at version 2.1, which is covered in detail in Chapters 3 and 4. Their work differs
from this thesis project in that their excavation site was much larger and they collected 3D data
in-situ during excavation. 3D data was not collected as the fossils were excavated from Project
23.
In a separate effort to visualize artifacts both large-scale (m-sized) and small-scale (cm-
sized), researchers from the Aristotle University of Thessaloniki, Greece chose Esri ArcGIS
software to develop a 3D model due to its functional 3D environment, customization abilities,
ability to communicate with external programs (i.e. CAD, SketchUp) and its support of object-
oriented data with linked attribute information at the prehistoric site of Paliambela Kolindros,
Greece (Katsianis, Tsipidis, Kotsakis, and Kousoulakou 2008, 655-658). A combination of point
data collection and photogrammetry was used to create a digital elevation model (DEM) of their
study area with imagery draped over it. In order to represent each archaeologically significant
feature as a single digital object within the GIS environment, SketchUp was used to create
complex 3D objects of artifacts that were then imported into the Esri geodatabase. For smaller
finds such as rocks and fossils, simple 3D point markers (i.e. spheres, tetrahedrons) were used as
symbols while researchers acknowledged that more advanced 3D symbology could be created in
24

other software and then imported to the 3D model (Figure 5) (Katsianis, Tsipidis, Kotsakis, and
Kousoulakou 2008, 659-663). Their creation of complex 3D objects outside of the ArcGIS
software is similar to work done as part of this thesis project in that the fossils from the museum
were laser scanned and processed prior to import into Esri software.

Figure 5. 3D model of Paliambela Kolindros in Esri ArcScene (Katsianis, Tsipidis, Kotsakis, and
Kousoulakou 2008, 663)
The Katsianis, Tsipidis, Kotsakis and Kousoulakou (2008, 659-663) researchers also
programmed several tools in ArcGIS to facilitate archaeological analysis, the most relevant to
this thesis work being a 3D point-to-point distance tool that can calculate the distance between
objects in 3D space, forming the basis for the implementation of spatial statistics. However, the
study in Paliambela Kolindros differs greatly from this thesis project given that future 3D
modeling was taken into consideration in their study when the archaeological data was recorded.
In this thesis project, deposit Box 13a was removed from its original location which has been
25

built over. These circumstances limited the development of this project’s 3D model to the data
that was collected prior to the asphalt deposit’s removal, which only included a single
geographic coordinate (x, y) taken at the center of Box 13a with accompanying depth
measurements.
Also considering small-scale objects, researchers at Queen’s University in Ontario,
Canada used Esri ArcGIS Online, Esri Web App Builder for ArcGIS Developer Edition version
1.3 and ArcGIS API Javascript version 3.15 to incorporate laser scanned hand samples of
mesoscale rocks, minerals and fossils into a GIS (Harvey, Fotopoulos, Hall, and Amolins 2017,
152-162). The researchers built a widget for an ArcGIS Web App that facilitated the 3D
visualization of mesoscale rocks, minerals and fossils in the geographical context of their 2D
macro-scale environment. First, a prototype geological database with samples as georeferenced
point features on a 2D map was developed. Then a widget was customized that allowed the user
to click on a sample on the 2D map to view its attributes as well as the 3D model of the sample
in a pop-up window. The 3D model could be rotated, zoomed in and zoomed out from (Figure 6)
(Harvey, Fotopoulos, Hall, and Amolins 2017, 154-158). The objective of this work was similar
to this thesis project as both studies sought to georeference small-scale samples that were laser
scanned by hand with linked attribute information and included 3D visualization. The study
differed from this thesis project because the widget limits the user to viewing each sample in 3D
individually. The proximity and orientation between samples cannot be viewed, restricting visual
interpretation of spatial patterns between samples to what can be viewed on the 2D map. This
thesis project sought to develop a 3D visualization in which multiple artifacts could be visualized
simultaneously in 3D and at varying scales.
26

Figure 6. Widget for viewing 3D models of samples in ArcGIS Web App (Harvey, Fotopoulos,
Hall, and Amolins 2017, 155)
The review of these studies provided useful insights into the development of the
methodology to construct a 3D fossil visualization, which is described in detail in the next
chapter of this thesis.
2.3. Forensics Background
Forensic archaeology is defined as “the controlled recovery of human remains and other
evidence at forensic scenes (Nawrocki 1996, 1).” In some cases, prior to excavation, a datum or
fixed point is established along with a subdatum at a known distance from the datum near the
site. A reference grid of meter-by-meter grids using materials such as string is then constructed
and assigned Cartesian coordinates based on the relative position of the subdatum. This allows
27

for the recording of “provenience” of remains or evidence from the excavation site. Provenience
refers to the x, y and z coordinate location of an item (Nawrocki 1996, 5-6). Although this is just
one of many methods for recording data in forensic archaeology, the process of establishing
grids with Cartesian coordinates is similar to the excavation practices performed by the La Brea
Tar Pits and Museum. Thus, the methodology outlined in the following chapter for visualizing
excavated fossils may have implications beyond that of paleontology.
Additionally, a lesser-known field referred to as forensic paleontology is newly emerging
as a useful tool in intelligence and criminal investigations. Although the term itself is currently
under debate, forensic paleontology is defined as “the use of paleontology as a scientific tool
available both to law enforcement officers and to judges and defending counsels (Sacchi and
Nicosia 2013, 652).” It appears to overlap with other branches of science such as forensic
archaeology, forensic geosciences and forensic biology, making it less frequently emphasized in
criminal cases and its specific applications less distinct (Sacchi and Nicosia 2013, 651). At
present, it is unclear if the methodology described in the following chapter will have implications
in the field of forensic paleontology.

28

Chapter 3 Methodology
This chapter describes the methodology for the development of the 3D GIS fossil visualization
for the La Brea Tar Pits and Museum. Section 3.1 provides an overview of the methodology and
section 3.2 describes the data sources for this project. Section 3.3 presents additional relevant
background information about the project data and section 3.4 outlines the development of the
3D GIS fossil visualization and potential sources of error.
3.1. Method Overview
As previously stated, the main objective of this methodology was to find a way to
visualize fossils as high-resolution 3D objects displayed on a 3D map in their pre-excavation, in-
situ locations beneath the earth’s surface utilizing 3D GIS technology. The methodology
reported in this thesis is unique in the fields of paleontology and archeology, as well as in the
forensic branch of archaeology, at the time of this writing. Initial steps toward developing this
3D visualization included significant data preparation outside of the GIS environment including
diagramming of fossil orientations, fossil location coordinate transformations and various file
conversion activities. These initial data preparation steps were followed by 3D object creation,
import and placement into 3D maps in ArcGIS Pro. Since access to advanced 3D functionality
within GIS software is relatively new in the field of spatial science, the following sections in this
chapter contain a summary of instructions on utilizing specific tools in GIS software to manually
handle the placement of small-scale (cm-sized) 3D objects below ground surface. Detailed
instructions of the steps performed in ArcGIS Pro are available in Appendix A. At this time,
tools for automatic placement of 3D objects were not available in ArcGIS Pro, and other GIS
software packages with 3D functionality such as Esri CityEngine were not tested due to the
objective of testing ArcGIS Pro as it is heavily used in industry and government, as well as due
29

to thesis project time constraints. Future enhancements in handling high-resolution 3D objects in
ArcGIS Pro are currently under consideration by Esri and solutions are even now being
developed by Safe Software, Inc., all of which are discussed in Chapter 5 in the Future Work
section (Sims, pers. comm., Bagh, pers. comm.).
3.2. Data Sources
Museum scientists provided the author with data from Project 23’s deposit Box 13a from
Grid B-3 Levels 1 and 2. This included nineteen high-resolution 3D objects of the fossils in
VRML format with corresponding imagery (96 dpi) and attribute information (Lindsey, pers.
comm.). Museum scientists also provided images, tables and reports clarifying the locations and
sizes of Box 13a, Grid B-3 Levels 1 and 2 and the locations and orientations of the fossils within
(La Brea Tar Pits and Museum 2018). Pham (2015) provided the author with her thesis
manuscript and an Esri geodatabase that includes a feature class with the location of Box 13a.
3.3. Data Background
This section covers background information regarding the box, grid and fossil data
provided by the museum. Section 3.3.1 describes Box 13a’s grid levels, and section 3.3.2
discusses the nineteen fossils and their corresponding images (textures).
3.3.1. Box and Grid Data
Box 13a was comprised of seven partial or complete 1 x 1 meter grids (Figure 7). The
fossils selected for this project were excavated in Grid B-3 and ranged from Levels 1 to 2. Grid
B-3 Level 1 was located 0 to 25 centimeters below the top of Box 13a, while Grid B-3 Level 2
was located underneath Level 1 at 25 to 50 centimeters below the top of the box. The origin of
each grid was located at its southeastern corner. When the fossils were excavated, three sets of
30

local coordinates were recorded in centimeters for each fossil including northing (x-direction),
westing (y-direction) and below depth (z-direction) with respect to the origin of the grid (Figure
8). Each fossil was marked with three small paint dots representing the three sets of local
coordinates (i.e. Px, MC, LC in Figure 9) (Howard and Farrell, pers. comm.). This information
served as the location and orientation information for the fossils used in this project.

Figure 7. The grid layout of Box 13a with Grid B-3 highlighted in green (La Brea Tar Pits and
Museum 2018)
31

Figure 8. Northing (N), Westing (W), and Below Depth (BD) measurements (Pham 2015, 32,
Shaw 1982, 66)
3.3.2. Fossil Data
Nineteen fossils from Box 13a Grid B-3 Levels 1 and 2 were selected by museum
scientists to be included in the 3D visualization. The fossils included bird (Aves), ancient bison
(Bison antiquus), dire wolf (Canis dirus), coyote (Canis latrans), Western horse (Equus
occidentalis) and saber-toothed cat (Smilodon fatalis) specimens ranging in length from
10 to 20 cm. The fossils were manually laser scanned by museum scientists using an Artec Space
Spider scanner and processed in Artec Studio 12 (Artec Space Spider 2015, Artec Studio 12
2017). The resulting 3D objects were exported into VRML format files. Museum scientists then
provided attribute information in an Excel table format, including species, location and
orientation information within Grid B-3, both levels. This data was also accompanied with
imagery indicating where the local coordinates were taken during excavation on each fossil, as
32

shown in an example in Figure 9. Human error relating to paleontological excavation
measurement methods mentioned in the previous chapter may have impacted the spatial data
accuracy of the fossils. A summary of errors which should be taken into account is provided at
the end of Chapter 3, in section 3.4.5.

Figure 9. Fossil P23 33245 saber-toothed cat coordinate locations (La Brea Tar Pits and Museum
2018)
3.4. 3D Visualization Development Approach
This section outlines the methodology developed to construct the 3D visualization.
Section 3.4.1 describes the fossil scanned image data processing steps taken prior to import into
the GIS in order to understand and validate the orientation of each fossil. Section 3.4.2 describes
the coordinate transformation required to generate geographically referenced coordinates for the
33

fossils using Excel, to support the subsequent required georeferencing steps in ArcGIS Pro.
Section 3.4.3 describes how Box 13a and Grid B-3 Levels 1 and 2 were created in 3D using
Excel and ArcGIS Pro. Section 3.4.4 details how the 3D objects - the scanned images of the
fossils - were imported into ArcGIS Pro and then placed in their pre-excavation, in-situ locations.
See Appendix A for more detailed instructions regarding steps performed in ArcGIS Pro.
3.4.1. Fossil Diagramming
The first step to developing the 3D model of the fossils in Box 13a was to understand and
validate where the fossils were located and how they were oriented within Grid B-3. Hand-drawn
diagrams were made with graph paper and pencil of the top view and north-facing side view of
each fossil within the grid. An example drawing is provided in Figure 10. These simple
visualizations were a tremendous aid in the creation of 3D diagrams of each fossil in PowerPoint.
Although any diagramming software could have been used, PowerPoint was a readily available
tool suitable for this purpose. This second step of producing diagrams in PowerPoint was
deemed necessary in order to produce images clearly depicting the location and orientation of
each fossil for validation by museum scientists (Howard and Farrell, pers. comm.). This step
served as an intermediate check on the validity of the coordinate transformations described in the
following section before moving on to build the visualization in ArcGIS Pro. Real-world,
geographic coordinates were not used in PowerPoint.
Prior to importing the fossil scans into PowerPoint, each 3D object was converted to 3D
Manufacturing Format (3MF) using Microsoft 3D Builder in order to preserve its texture. Using
the 3D model tool, the fossil scans were imported into PowerPoint and manually rotated to their
proper orientation with respect to the grid (Figure 10). Learning how to manually rotate the 3D
objects in PowerPoint inspired later steps in the methodology when building the final 3D
34

visualization in ArcGIS Pro. Diagramming also revealed that some of the fossils’ coordinates
occurred just outside of Grid B-3 Levels 1 and 2 to the east. See Appendix B for the hand-drawn
and PowerPoint diagrams of the 19 fossils.

Figure 10. Example of a hand-drawn diagram and the corresponding PowerPoint diagram of
fossil P23 33245
3.4.2. Coordinate Transformation in Excel and ArcGIS Pro
Before the fossil scans could be imported into ArcGIS Pro, significant coordinate
transformation steps were required in Excel and ArcGIS Pro. The only georeferenced data
provided to the author was the central coordinate of Box 13a, also found in Pham’s geodatabase
(Howard and Farrell, pers. comm., Pham 2015). The grid coordinates accompanying the fossils
were not previously georeferenced, thus, they had to be transformed into real-world, geographic
coordinates. Knowing the central coordinate of Box 13a was taken at the center of the box, a
point was placed at the approximate center of the Box 13a schematic (Figure 11). This point,
representing the central coordinate of Box 13a, landed at 0 cm W, 50 cm N in Grid B-3. This
information was used in conjunction with the fossil coordinates to perform relative distance (eq
1) and bearing (eq 2) calculations between the central coordinate of Box 13a and the fossil
35

coordinates. Bearing and distance values were required for the subsequent coordinate
transformation steps performed in ArcGIS Pro using the Bearing Distance to Line tool.

Figure 11. Box 13a central coordinate at 0 cm W, 50 cm N on Grid B-3
First, the distance (eq 1) equation was written based on principle of the Pythagorean
theorem, which states “if 𝑎 and 𝑏 are lengths of the legs of a right triangle, and 𝑐 is the length of
the hypotenuse, then 𝑎 2
+ 𝑏 2
= 𝑐 2
” (Gelfand and Saul 2001, 7, Figure 12).
36

Figure 12. Pythagorean theorem diagram
In the case of this thesis project, the diagram in Figure 13 is representative of any given
coordinate in centimeters west (𝑊 ) and north (𝑁 ) of the Grid B-3 origin at 0 west, 0 north. Find
the distance (𝐷 ) from the central coordinate of Box 13a at 0 west, 50 north to any given
coordinate.

Figure 13. Distance equation example

37

For the right triangle in Figure 13, this means:
𝐷 2
= (𝑁 − 50)
2
+ 𝑊 2

After square rooting both sides of the equation above:
𝐷 = ±√(𝑁 − 50)
2
+ 𝑊 2

Negative distances do not exist, therefore:
𝐷 = √(𝑁 − 50)
2
+ 𝑊 2

Since the coordinates were given in centimeters but ArcGIS Pro’s Bearing Distance to
Line tool required meter units, the final distance equation was adjusted for meters as follows:
𝐷𝑖𝑠𝑡𝑎𝑛𝑐𝑒 (𝑚 ) = √(𝑁 − 50)
2
+ 𝑊 2
∗
1 𝑚 100 𝑐𝑚
(eq 1)
Next, the bearing (eq 2) equation was developed based on the principles of the arctangent
function. In a right triangle, the tangent of an angle is equal to the length of the opposite side
divided by the length of the adjacent side. While the tangent returns only a ratio, the arctangent is
the inverse function of the tangent and returns an angle (Gelfand and Saul 2001, 35). An example
right triangle with these relationships is shown in Figure 14.

Figure 14. Tangent and arctangent diagram
38

The arctangent function only returns values between -90° and 90° (Microsoft Office
2018a). For this project, an arctangent function that returns values for angles greater than 90°
was desired. The Excel function known as 𝑎𝑡𝑎𝑛 2, a variation of the arctangent function, returns
values between -180° and 180° expressed in radians. As shown in the right triangle in Figure 15,
the inputs of 𝑎𝑡𝑎𝑛 2 are the adjacent side followed by the opposite side, referred to as
(𝑥 _𝑛𝑢𝑚 , 𝑦 _𝑛𝑢𝑚 ) respectively in the Excel syntax (Microsoft Office 2018b).

Figure 15. Atan2 diagram
In the case of this thesis project, Figure 16 illustrates an example of an angle or bearing
(𝐵 ) that needs to be found. Bearings, like those on a compass, are measured clockwise from
north (0°) (United States Marine Corps 2001, 20). This shows the bearing as negative because
the example bearing is counter-clockwise from north. The majority of the fossils’ local
coordinates occurred within Grid B-3, located in the northwest quadrant of the local x, y grid
shown in the example in Figure 16.
39

Figure 16. Bearing equation example
For the right triangle in Figure 16, this means:
−𝐵 = atan 2(𝑁 − 50, 𝑊 )
To solve for 𝐵 , multiply both sides of the equation by -1, therefore:
𝐵 = −atan 2(𝑁 − 50, 𝑊 )
As previously stated, 𝑎𝑡𝑎𝑛 2 returns values in radians. ArcGIS Pro’s Bearing Distance to
Line tool required bearing units in degrees. The conversion factor is 180° per 𝜋 radians, therefore
(Microsoft Office 2018b):
𝐵 = − atan 2(𝑁 − 50, 𝑊 ) ∗
180°
𝜋
The equation above returns values between -180° and 180°. However, bearings between
0° and 360° were preferred for this project for clarity as bearings are conventionally reported as
positive values. When the variable 𝑊 is negative, the equation produces a positive bearing.
Conversely, when the variable 𝑊 is positive, the equation produces a negative bearing. In this
instance, it is necessary to add a full 360° rotation to the equation to generate an equivalent
40

positive bearing (Gelfand and Saul 2001, 90-95). To generate the final bearing equation, let 𝑖 = 1
when 𝑊 is greater than 0 and let 𝑖 = 0 when 𝑊 is less than or equal to 0:
𝐵𝑒𝑎𝑟𝑖𝑛𝑔 (°) = −atan 2(𝑁 − 50, 𝑊 ) ∗
180°
𝜋 + 360°𝑖 (eq 2)
As previously mentioned, the majority of the fossils’ local coordinates occurred in Grid
B-3, west of the Box 13a central coordinate. Coordinates in Grid B-3 had positive 𝑊 values and
required that 𝑖 = 1. However, a few of the fossil coordinates occurred slightly outside of Grid B-
3 to the east of the central coordinate of Box 13a in the northeast quadrant of the x, y grid. These
coordinates had negative 𝑊 values and required that 𝑖 = 0. The example diagram in Figure 17
demonstrates the result of adding a full 360° rotation to a negative bearing to generate an
equivalent positive bearing. All distance and bearing values were calculated in an Excel table
called CoordinateTransformation.xlsx which also contained the fossils’ attribute information
(Table 1).

Figure 17. Example of 360° rotation added to negative bearing

41

Table 1. Example data for fossil P23 33245 from CoordinateTransformation.xlsx

Specimen
Number
Point Taxon Element Deposit Level Grid BD_
cm
N_
cm
W_
cm
i Bearing_deg Distance_m Box 13a
Origin X_m
Box 13a
Origin Y_m
Elevation_
m
P23
33245
Px Smilodon
fatalis
Femur 13 1 B-3 19.5 19.5 37 1 230.5004115 0.479504953 -13175742.2139 4037373.0874 47.244
P23
33245
MC Smilodon
fatalis
Femur 13 1 B-3 30 28.5 44 1 243.9581834 0.489719307 -13175742.2139 4037373.0874 47.139
P23
33245
LC Smilodon
fatalis
Femur 13 1 B-3 22.5 31 46 1 247.5572466 0.497694686 -13175742.2139 4037373.0874 47.214
42

Next, in ArcGIS Pro, the Box 13a central coordinate from Pham’s (2015) geodatabase
was exported to a new geodatabase called LaBreaTarPits_P23Box13a_FossilsModel.gdb as a
point feature class named P23_Box13a.fc. Pham’s data and geodatabase were projected into the
WGS 1984 Web Mercator Auxiliary Sphere. Thus the same projection was used for the new
geodatabase created for this thesis project to best support the existing or current museum
geodatabases. The author used the Esri basemap named ‘Imagery’ as an elevation source in
meters. The Add Geometry Tool was run on point feature class P23_Box13a.fc to generate x and
y coordinates in the associated attribute table. The x and y coordinates were then added to a table
named CoordinateTransformation.xlsx. The correct elevation for each coordinate was then
calculated in the Excel table based on the elevation attribute information found in
P23_Box13a.fc for the top of Box 13a. The fossils’ attribute information included depth below
the top of the box (BD) values in centimeters. Thus, the BD values were subtracted from the
elevation of the top of Box 13a to generate z values in meters. Another Excel table called
AllBones_DataForCoordinateTransformation.xlsx was then created including only information
from CoordinateTransformation.xlsx required for the next step of importing the information into
ArcGIS Pro (Table 2).

43

Table 2. Example data for fossil P23 33245 from AllBone_DataForCoordinateTransformation.xlsx to be imported into ArcGIS Pro

ID Spec_Num Point Taxon Element Level Bearing_deg Distance_m Box13a_OriginX_m Box13a_OriginY_m Elevation_m
31 P23 33245 Px Smilodon
fatalis
Femur 1 230.5004115 0.479504953 -13175742.21386 4037373.087427 47.244
32 P23 33245 MC Smilodon
fatalis
Femur 1 243.9581834 0.489719307 -13175742.21386 4037373.087427 47.139
33 P23 33245 LC Smilodon
fatalis
Femur 1 247.5572466 0.497694686 -13175742.21386 4037373.087427 47.214
44

The AllBones_DataForCoordinateTransformation.xlsx was then imported into the
LaBreaTarPits_P23Box13a_FossilsModel.gdb as a table. With the table as input, the Bearing
Distance to Line tool was used to create lines running from the central coordinate of Box 13a to
the fossils’ coordinates. The Feature Vertices to Points tool was then used with the newly created
polyline feature class to generate vertices at the ends each line (Figure 18). The Join Field tool
was then run using the vertices feature class and the imported table as input to add any attribute
fields to the vertices feature class that were lost in the previous step. The vertices opposite of the
Box 13a central coordinate were then selected according to fossil using the attribute table and
exported into 19 individual point feature classes with z values enabled in the
LaBreaTarPits_P23Box13a_FossilsModel.gdb.

Figure 18. Bearing Distance to Line and Feature Vertices to Points tools results in ArcGIS Pro
Since the transformed coordinates did not have built-in z values, the Move To tool was
utilized to select each coordinate and move it to its exact elevation indicated in its attribute table
(Figure 19). The Add Geometry Attribute tool was then used to generate x, y and z values for
45

each coordinate in the attribute table as well as to crosscheck the z field with the elevation field
for accuracy.

Figure 19. Transformed coordinates below ground surface in ArcGIS Pro
3.4.3. Box 13a and Grid B-3 Creation using Excel, Adobe Reader and ArcGIS Pro
The next step in building the 3D visualization was to create digital representations of Box
13a and Grid B-3 Levels 1 and 2. Using Excel, the height of Box 13a was subtracted from the
known elevation of the top of the boxed substrate prior to extraction in order to determine Box
13a’s elevation below ground. The height of Grid B-3 Level 1 was also subtracted from the
known elevation of the top of the box to determine Level 1’s elevation, while Level 2’s elevation
was determined by subtracting both Level 1 and 2’s total height from the known elevation of the
top of the box. This elevation information was then used later in ArcGIS Pro to position the box
and grids underground accurately.
In Adobe Reader, a hand-drawn vicinity map of Project 23 provided by the museum in
PDF was cropped and converted to JPG format (Figure 20) (La Brea Tar Pits and Museum
46

2018). Next, the image was imported into the LaBreaTarPits_P23Box13a_FossilsModel.gdb in
ArcGIS Pro. The Georeference tool was used to approximately align the map with the point
feature class P23_Box13a.fc as well as with the imagery in the 2D Scene. From there, a new
multipatch feature class named P23_Box13a_obj.fc was created in the geodatabase
LaBreaTarPits_P23Box13a_FossilsModel.gdb. Using the Create tool and selecting the “Cube”
model, the dimensions of Box 13a (width: 1.9812 m, depth: 1.9812 m, height: 1.40208 m) were
entered into the Parameters tab. Prior to digitizing Box 13a, the cube was manually dragged or
hovered over the Project 23 vicinity map and the Rotation was adjusted using the Parameters tab
to match that of the map. The cube was then digitized in the 3D Scene to create Box 13a and the
transparency was adjusted to 75%. The Move To tool was then used to center the box over its
centroid, P23_Box13a.fc (Figure 21). The box symbol was then colored “Electron Gold.”

Figure 20. Project 23 locality map with an arrow indicating Box 13a (La Brea Tar Pits and
Museum 2018)
47

Figure 21. Box 13a rotated and centered over Box 13a’s central coordinate
The next step was to create two new multipatch feature classes to represent Grid B3
Levels 1 and 2. First, a multipatch feature class named P23_GridB3_Level2.fc was created in the
LaBreaTarPits_P23Box13a_FossilsModel.gdb. Using the Create tool and selecting the “Cube”
model, the dimensions for Grid B-3 Level 2 (width: 1 m, depth: 1 m, height: 0.25 m) were
entered in the Parameters tab. The grid was then manually digitized approximately over Box
13a’s central coordinate at the grid’s 0 cm W, 50 cm N location (Figure 22). The grid did not
need to be rotated since its eastern side runs directly north. A copy of P23_GridB3_Level2.fc
was then made and renamed to P23_GridB3_Level1.fc in order to create Grid B-3 Level 1. Grid
B-3 Levels 1 and 2 have identical dimensions.
The final step in creating the 3D representations of the box and grids was to use the Move
To tool to place each of them at their correct elevations below the ground surface. The color of
the symbols for Level 1 and 2 were set to “Mars Red” and “Big Sky Blue,” respectively. The
transparency was set at 50-75% for each level to maximize visualization (Figure 23). Attribute
48

fields including height, width and elevation were then added and manually filled out for
P23_Box13a_obj.fc, GridB3_Level1.fc and GridB3_Level2.fc.

Figure 22. Grid B-3 Level 2 digitized over Box 13a’s central coordinate at 0 cm W, 50 cm N on
the grid

Figure 23. Box 13a and Grid B-3 Levels 1 and 2 below ground surface
49

3.4.4. Fossil Digitization in ArcGIS Pro
Given the fossils’ coordinates were in the proper location in ArcGIS Pro, the fossil scans
were ready to be imported and manually scaled, rotated and placed in their pre-excavation, in-
situ locations. The 3D objects were first converted from VRML format to Wavefront OBJ (OBJ)
format using Microsoft 3D Builder in order to upload them to ArcGIS Pro. During this process,
the textures of the 3D objects were lost, leaving the fossils white in color. Many solutions and
workarounds were tested throughout this thesis work to solve this issue, such as importing using
different 3D GIS software and imagery formats, as well as trying a variety of ArcGIS Pro tools.
At the time of this writing, however, ArcGIS Pro version 2.1 did not support the import of high-
resolution 3D objects in VRML format (Sims, pers. comm.). Current efforts underway to solve
this deficiency are described in Chapter 5, in section 5.2.1. (Bagh, pers. comm.).
Next, the 3D objects in OBJ format were imported into ArcGIS Pro by first creating a
new multipatch feature class for each fossil in the LaBreaTarPits_P23Box13a_FossilsModel.gdb.
Then the Create tool was utilized to upload the 3D objects as Esri Model Files, which were then
manually digitized in the 3D Scene above the ground surface. Using the Symbology template,
the color of each fossil was changed to “Leather Brown” to appear as similar as possible to the
original high-resolution fossil image textures that were lost during file format conversion (Figure
24). Each fossil was then manually positioned using a combination of the Move, Scale and
Rotate tools to align them with their corresponding coordinates as close as possible (Figure 25-
Figure 26). The Join Field tool was then used to link relevant attributes from the fossils’
coordinate point feature classes to their respective multipatch feature classes. Finally, the origin
and function of each dataset were added to each of the feature class’ metadata by right-clicking
on each feature class and selecting “Edit Metadata.”

50

Figure 24. Fossil P23 33245 manually added to the 3D Scene

Figure 25. Fossil P23 33245 manually placed below ground surface using the Move tool prior to
the Scale and Rotate tools
51

Figure 26. Fossil P23 33245 manually aligned with corresponding coordinates
3.4.5. Sources of Error
Beyond the human error inherent in the excavation process, other sources of error exist in
this methodology. It was approximated that the central coordinate of Box 13a occurred at
0 cm W, 50 cm N in Grid B-3 based on the grid’s hand-drawn schematic (Figure 7). For this
reason, the fossils’ geographic coordinates relative to the central coordinate of Box 13a may
have been impacted by a few centimeters. It is important to mention, however, that the fossils’
positions relative to each other were not affected by that approximation. Additionally,
GridB3_Level2.fc was digitized at approximately 0 cm W, 50 cm N on Grid B-3 in ArcGIS Pro
and then copied to create GridB3_Level1.fc. This may have led to error of up to a few
centimeters when digitizing the grids. Lastly, Box 13a was digitized approximately over its
central coordinate and rotated to match its orientation depicted on the georeferenced locality
map, creating more potential for error in the 3D model.
52

3.5. Chapter Summary
The methodology described above includes hand-drawn graphing of the fossils in Box
13a, coordinate transformation of local grid coordinates to real-world, geographic coordinates,
file conversion of 3D objects from VRML to OBJ format and building an Esri geodatabase of 3D
objects representing fossils. Although the main objective of the thesis project was met, due to
limitations in the 3D functionality of the GIS software chosen for this project, the methodology
developed is manual, rather than automated, and the results fell short of the museum scientists’
expectations that the 3D representations of the fossils in the visualization would include the
textures of each individual fossil. The 3D fossil visualization resulting from this methodology is
outlined in the next chapter, Chapter 4 Results. Current efforts are underway to improve upon
both the process of building the visualization as well as retaining the original fossil imagery are
described in Chapter 5 under Future Work.
53

Chapter 4 Results
This chapter describes the results of the development of the methodology for the 3D fossil
visualization for the La Brea Tar Pits and Museum. Section 4.1 discusses the alternative
methodologies and software that were tested during the thesis work as well as the overall
methodology developed, and section 4.2 presents the final 3D model results.
4.1. Alternative 3D Modeling Testing and Results
Several methodologies and software were tested to develop the methodology for the 3D
visualization, including ArcGIS Pro, SketchUp Pro 2017, Microsoft 3D Builder and Microsoft
3D Paint (SketchUp Pro 2017 2017, Microsoft 3D Paint 2016). Section 4.1.1 documents the
results of these efforts in ArcGIS Pro and section 4.1.2 describes the results of testing the other
software.
4.1.1. ArcGIS Pro Testing
The primary approach to building the 3D model in ArcGIS Pro was to utilize the Import
3D Files tool to upload the fossil P23 33245 3D object in VRML format as a test. This step was
the first attempt at importing the 3D object into ArcGIS Pro. The tool was initially run without
indicating a coordinate system, which resulted in distortion of the 3D object in the 3D Scene
such that fragments comprising the object spread across the globe. Prior to running the tool for a
second time, a coordinate system matching that of the 3D Scene was chosen and, although the
tool ran successfully, the object could not be located in the 3D Scene. When zooming to the
generated output feature class, the 3D Scene panned over Africa. It is important to note that the
coordinate system chosen for testing was the same as that used successfully throughout the rest
of the project work and that the x, y (longitude, latitude) were correctly assigned when using the
54

tool. The tool was then run a third time using a selected Placement Points feature class. This
feature class was generated in advance by creating a point feature class in the same coordinate
system as the 3D Scene, digitizing a point in the vicinity of Box 13a, extruding the point below
ground with the Adjust 3D Z tool, and then running the Add Geometry Attributes tool to create
x, y and z values in the feature class’ attribute table. Although a placement point was provided
and the Import 3D Files tool claimed to have run successfully, the imported 3D object could not
be located in the 3D Scene. Again, when zooming to the generated output feature class, the 3D
Scene panned over Africa.
Another approach taken to import the 3D object was to first convert the fossil P23 33245
VRML file into a COLLAborative Design Activity (COLLADA) file. Using the Data
Interoperability Extension FME Workbench 2017.1 in ArcGIS Pro, the P23 33245 VRML file
was converted to a COLLADA file (FME Workbench 2017.1 2017, Bagh, pers. comm.). The
Import 3D Files tool was then run again with the newly created P23 33245 COLLADA file,
which caused ArcGIS Pro to close unexpectedly. Next, a multipatch feature class was created,
and the Create tool was selected. The COLLADA file was selected as the Model File to be
digitized at this stage, and ArcGIS Pro crashed again. It was determined that the combination of
ArcGIS Pro version 2.1 and the USC SSI ONLINE system resources could not support the
import of even a single COLLADA file of such high resolution (Sims, pers. comm.).
This led the author to convert the VRML files into a supported OBJ format using
Microsoft 3D Builder for further testing. In ArcGIS Pro, the OBJ format was not supported by
the Import 3D Files tool, so the author repeated the process above of creating a new multipatch
feature class and using the Create tool to then select the Model File fossil P23 33245 in OBJ
format. The fossil was then manually positioned within the 3D Scene. After testing both the
55

Georeference tool and the Data Interoperability Extension FME Workbench, it became clear that
the P23 33245 3D object could not be georeferenced numerically or automatically to its set of
three location coordinates (FME Workbench 2017.1 2017). This led the author to next test
ArcGIS Pro’s Move, Scale and Rotate tools to manually align the P23 33245 3D object with its
respective coordinates. An additional limitation of ArcGIS Pro discovered during this testing
process was that the Add Geometry Attributes tool did not work with the P23 33245 3D object.
Thus, the multipatch feature classes representing the fossils in the final 3D model do not have
finalized x, y and z coordinates in their attribute tables which could be used to validate
placement within the 3D visualization.
4.1.2. SketchUp Pro, Microsoft 3D Builder and Microsoft 3D Paint Testing
Other software that were tested for 3D visualization capabilities including SketchUp Pro,
Microsoft 3D Builder and Microsoft 3D Paint (SketchUp Pro 2017 2017, Microsoft 3D Builder
2013, Microsoft 3D Paint 2016). The author considered building the 3D visualization in a GIS
other than Esri and then importing it into ArcGIS Pro as a single entity. SketchUp Pro was the
first software that was tested but could not be used because the free version of the software
required the user to publish the museum’s 3D objects to the software’s 3D Warehouse, thus
potentially releasing the fossil images to the public. This would have violated the current
Memorandum of Understanding between the author and the museum (La Brea Tar Pits and
Museum 2017b). The next software tested was Microsoft, including 3D Builder and 3D Paint,
both of which proved to have insufficient capabilities for building the 3D visualization because
they lacked the option to make 3D objects transparent. The lack of transparency would not have
worked for this project for viewing the fossils inside of Box 13a. 3D Builder and 3D Paint also
56

lacked all georeferencing tools making it impossible to recreate the 3D model with any spatial
accuracy.
4.2. 3D Modeling Results
The 3D visualization produced for this project consists of an Esri geodatabase containing
multipatch feature classes representing Box 13a, Grid B3 Levels 1 and 2 and the nineteen fossils’
simplified 3D scanned imagery. The geodatabase also includes point feature classes representing
the three sets of transformed coordinates associated with each fossil. Additionally, the
geodatabase contains the table that was imported for the coordinate transformation as well as two
feature classes produced by the Bearing Distance to Line tool and the Feature Vertices to Points
tool. And finally, three visual, image mensuration feature classes were autogenerated while
building the 3D model. These image mensuration point, line, and polygon feature classes allow
the user to measure ground features from georeferenced imagery but were not utilized in this
project.
Overall, the 3D visualization adequately displays the fossils underground in the position
and orientation in which they were originally excavated (Figure 27-Figure 30). When viewed in
ArcGIS Pro, the user can zoom and rotate the 3D Scene to view the fossils from all angles within
their grids in Box 13a. Transparency of all objects including the fossils, grids and box can be
controlled by the user. The user can also click on the fossils, box and grids to view their basic
attribute information (Figure 31). For more information, the user can view the attribute tables of
the point feature classes containing the fossils’ coordinates (x, y and z) to obtain precise location
information. The origin and function of each dataset were also added to the feature class’
metadata to provide clarity for museum scientists when viewing the 3D model.
57

Figure 27. All fossils aligned with their respective coordinates

Figure 28. North-facing view of all fossils in Grid B-3 Levels 1 and 2
58

Figure 29. North-facing view of all fossils in Box 13a

Figure 30. View from below of the fossils in Box 13a beneath its central coordinate
59

Figure 31. Fossil P23 33242 clicked on to view attribute information
Unfortunately, the 3D objects of the fossils lost their original textures due to the lack of
support of functionality for import of VRML files into ArcGIS Pro version 2.1. The fossil 3D
objects had to be converted into a file type that ArcGIS Pro could handle, OBJ format. During
this process, the connection or link between the 3D objects and their respective textures was
broken. To mitigate the visual impact of the texture loss, the author symbolized the fossils with a
brown color that was similar to the texture.
Although the fossils were manually aligned with their respective coordinates to the best
of the author’s ability within the constraints of the Move, Scale and Rotate tools in ArcGIS Pro,
the fossils did not always fit their anticipated locations (Figure 32). This is partly due to the
limitations of the ArcGIS Pro toolset but may also be due to error in the marking and cataloging
of each fossil’s local coordinates during excavation. While the scale of the fossils was adjusted to
align with their respective coordinates within this visualization, the proportions of the fossils
were not altered by the author in any way in an attempt to preserve their original shape. For these
60

reasons, the fossils were aligned with their coordinates with errors up to +/-1 cm in the x, y or z
direction.

Figure 32. Fossil P23 33252 coordinate alignment error
As anticipated in the previous chapter, evidence of error regarding the location of the
grids also occurred in the final 3D model. Although three of the fossils were expected to extend
outside of Grid B-3, Figure 33 shows one additional fossil (P23 33243) unexpectedly extending
outside the grid, demonstrating the predicted error of up to a few centimeters when digitizing the
grids.

Figure 33. Fossils extending outside Grid B-3 Levels 1 and 2
61

4.3. Chapter Summary
In summary, substantial testing was performed both inside and outside of ArcGIS Pro to
develop a viable methodology for building a 3D visualization of small-scale (cm-sized) objects
beneath the ground. These efforts resulted in an Esri geodatabase containing 2D and 3D spatial
and non-spatial data sets and feature classes, including well-documented visualization issues and
spatial accuracy concerns. The following chapter, Chapter 5 Discussion and Conclusion,
describes the successes and challenges of this thesis project and suggestions about how this work
can be further developed in the future.

62

Chapter 5 Discussion and Conclusion
This chapter describes the successes and failures of the 3D fossil visualization for the La Brea
Tar Pits and Museum and provides recommendations for future work to improve the
methodology. Section 5.1 discusses the limitations of the 3D visualization. Section 5.2 describes
potential future work including improvements in GIS tools for importing 3D objects, creating an
interactive museum exhibit and building an archive of high-resolution scanned images of fossils
as 3D objects. Section 5.3 prioritizes recommendations and provides final thoughts on furthering
this thesis work. Section 5.4 describes potential applications of the developed methodology to
forensic archaeology.
5.1. Discussion
This thesis project achieved its objective to develop a methodology to create a 3D
visualization of fossils in their pre-excavation, in-situ locations at the La Brea Tar Pits and
Museum. The author was successful in developing a methodology for converting traditionally
collected paleontological or archeological artifact coordinates into real-world, geographic
coordinates, making this procedure available and sharable with other researchers as well as the
museum in Microsoft Excel format. The author was also successful in developing the 3D model
in an Esri geodatabase in ArcGIS Pro but was limited to the constraints of the software at version
2.1. ArcGIS Pro could not import the 3D objects of the fossils in their original VRML format, so
the author had to convert the files to a supported OBJ format at the expense of the object
textures. The latter meant that the texture was lost. This, in combination with the fact that the 3D
objects themselves could not be georeferenced in ArcGIS Pro, produced a somewhat slow,
clunky visualization with 3D objects that had to be manually aligned with their respective
coordinates. ArcGIS Pro’s Move, Scale and Rotate tools were limited when making small-scale
63

adjustments to the fossils positions, thus, the fossils were aligned with their coordinates with
errors up to +/- 1 cm. This style of manual placement limited the entire 3D visualization in its
spatial accuracy.
The time-consuming nature of manually digitizing the 3D objects makes the process of
building the 3D fossil visualization far from streamlined. Due to limitations in the 3D
functionality of the GIS software chosen for this project, this thesis project did not meet its
primary purpose of producing a 3D model that could be spatially analyzed using GIS by museum
scientists. However, the resulting 3D visualization fulfilled the secondary purpose of this thesis
project to create a 3D model that could be interactively displayed in the museum using software
and hardware that can support the visualization. While the possibilities for improving the 3D
model are vast, the following section provides some ideas for ongoing work.
5.2. Future Work
This section describes long-term ambitions for the continuation of this project. Section
5.2.1 describes ongoing work being done and proposes future work to improve the process of
importing 3D objects into ArcGIS Pro. Section 5.2.2 discusses the future development of an
interactive display for the 3D visualization at the museum. Section 5.2.3 discusses the potential
for archiving 3D objects that could be used in many different types of projects by the museum.
Lastly, section 5.2.4 discusses the potential applications of this thesis work in the field of
forensic archaeology.
5.2.1. Improving the Import Process of 3D Objects to ArcGIS Pro
The first step to improving the methodology will be to improve the import process of
complex 3D objects into ArcGIS Pro. In July 2018 at the Esri User Conference, the author began
collaborating with developers from Safe Software, Inc., the company that developed the Data
64

Interoperability Extension FME Workbench (Bagh, pers. comm.). They are currently working on
a process to import VRML 3D objects into ArcGIS Pro that would allow automatic
georeferencing with real-world, geographic coordinates using FME Workbench. If successful,
this procedure would eliminate the need to manually position the 3D objects in ArcGIS Pro and,
thus, could radically streamline and improve the efficiency of the 3D fossil visualization
methodology. This procedure could be applied not only to the complex 3D objects like the fossils
but also to the simpler 3D objects such as Box 13a and Grid B-3 Levels 1 and 2. Currently,
ArcGIS Pro version 2.1 cannot georeference complex 3D objects or generate x, y and z values
for multipatch feature classes. Advanced geoferencing techniques developed in FME Workbench
may provide a workaround in the future while ArcGIS Pro continues to improve its own software
(FME Workbench 2017.1 2017).
5.2.2. Museum Exhibit
The 3D visualization created in this project could be displayed interactively in the
museum, but its viewing platform has yet to be determined. In line with the visions of the
museum scientists, future work will need to be done to set up a durable touchscreen display that
visitors can use to view fossil deposits interactively. The museum scientists envisioned that
visitors could click on various features on the 3D map to view information and better understand
the complexity of the tar pits (Lindsey, pers. comm.). This will require a software to be chosen
that both displays the 3D model and tolerates the interactive nature and heavy-handed use of the
exhibit anticipated.
5.2.3. 3D Data Generation and Archive
Another step to further the development of the 3D fossil visualization would be to create
and store an archive of the high-resolution scanned fossils readily accessible to the museum’s
65

research teams, such as in a secured private Google Drive. In addition, it is recommended that
the museum scientists begin laser scanning the fossils by hand immediately after excavation, and
even in-situ immediately after being marked (3 points). It would also be ideal if at least one real-
world, geographic coordinate marked on the fossil was recorded during excavation. This could
lead to the semi-automated generation of a 3D Data Archive that could be used to continue to
build and expand the 3D fossil visualization methodology documented in this thesis. As 3D GIS
functionality improves and usage becomes more commonplace, this archive could be a valuable
asset for the La Brea Tar Pits and Museum when it comes to both visualization and spatial
analysis.
5.2.4. Potential Application for Forensic Archaeology
The methodology developed for this 3D fossil visualization may have an application in
forensic archaeological practices in the future. As previously mentioned in Chapter 2, one
method of recording the locations of remains and evidence at excavation sites involves the
establishment of grids and a local Cartesian coordinate system based off a fixed point. This
practice produces x, y and z field data recordings similar to those taken in the La Brea Tar Pits
and Museum excavations. By applying this methodology developed for the fossil 3D
visualization to data recorded by forensic archaeologists, this thesis project could inform the
process of reconstructing crime scenes. For example, one benefit of having a crime scene
mapped in a GIS would be the access to spatial analysis and modeling tools to support solving
criminal investigations.
5.3. Next Steps
It is recommended that this project is continued through student interns with the same or
similar qualifications as the author and through funded research. In this way, the 3D fossil
66

visualization methodology could continue to evolve and contribute to the proposed 3D artifacts
digital archive, and the La Brea Tar Pits and Museum could move forward more quickly toward
new interactive 3D displays of fossils in-situ for the public to enjoy.
67

References
Adobe Reader XI. 2018. Accessed August 10, 2018. https://get.adobe.com/reader/.
Apollonio, Fabrizio I., Marco Gaiani, and Benedetto Benedetti. 2012. “3D Reality-Based
Artefact Models for the Management of Archaeological Sites Using 3D GIS: A
Framework Starting from the Case Study of the Pompeii Archaeological Area.” Journal
of Archaeological Science 39, no. 5 (May): 1271-1287. Accessed October 29, 2017.
https://doi.org/10.1016/j.jas.2011.12.034.
Artec Space Spider. 2015. Accessed August 12, 2018. https://www.artec3d.com/portable-3d-
scanners/artec-spider.
Artec Studio 12. 2017. Accessed August 12, 2018. https://www.artec3d.com/3d-software.
Bagh, Dmitri. Personal communication with author. Safe Software, Inc.
Campanaro, Danilo Marco, Giacomo Landeschi, Nicoló Dell’Unto, and Anne-Marie Leander
Touati. 2016. “3D GIS for Cultural Heritage Restoration: A ‘White-Box’ Workflow.”
Journal of Cultural Heritage 18 (March-April): 321-332. Accessed October 29, 2017.
https://doi.org/10.1016/j.culher.2015.09.006.
De Reu, Jeroen, Phillipe De Smedt, Davy Herremans, Marc Van Meirvenne, Pieter Laloo, and
Wim De Clercq. 2014. “On Introducing an Image-Based 3D Reconstruction Method in
Archaeological Excavation Practice.” Journal of Archaeological Science 41 (January):
251-262. Accessed October 29, 2017. https://doi.org/10.1016/j.jas.2013.08.020.
Esri ArcGIS Desktop. 2018. “Export Data.” Esri. Accessed August 19, 2018.
https://pro.arcgis.com/en/pro-app/help/data/geodatabases/overview/export-data.htm.
Esri ArcGIS Pro 2.1. 2018. Accessed August 10, 2018. https://pro.arcgis.com/en/pro-app/.
FME Workbench 2017.1. 2017. Accessed August 10, 2018. https://www.safe.com/fme/new-
features/2017.1/.
Gelfand, I.M. and M. Saul. 2001. Trigonometry. Boston, MA: Birkhäuser Boston.
Graham, M. Ian. 2003. “15-462 Computer Graphics Lecture 10: Texture Mapping.” Lecture at
Carnegie Mellon University, Pittsburgh, PA, February 13. Accessed August 19, 2018.
https://www.cs.cmu.edu/~fp/courses/graphics/pdf-color/10-texture.pdf.
G2 Force. 2018. “Best GIS Software.” G2 Force, Inc. Accessed August 18, 2018.
https://www.g2crowd.com/categories/gis.
Harvey, A.S, G. Fotopoulos, B. Hall, and K. Amolins. 2017. “Augmenting Comprehension of
Geological Relationships by Integrating 3D Laser Scanned Hand Samples within a GIS
68

Environment.” Computers & Geosciences 103 (June): 152-163. Accessed October 29,
2017. https://doi.org/10.1016/j.cageo.2017.02.008.
Howard, Carrie and Aisling Farrell. Personal communication with author. La Brea Tar Pits and
Museum Department of Research and Collections.
Katsianis, Markos, Spyros Tsipidis, Kostas Kotsakis, and Alexandra Kousoulakou. 2008. “A 3D
Digital Workflow for Archaeological Intra-Site Research Using GIS.” Journal of
Archaeological Science 35, no. 3 (March): 655-667. Accessed October 29, 2017.
https://doi.org/10.1016/j.jas.2007.06.002.
La Brea Tar Pits and Museum. 2017a. “About the Museum.” La Brea Tar Pits and Museum.
Accessed September 27, 2017. https://tarpits.org/our-story/about-the-page#.
La Brea Tar Pits and Museum. 2017b. Excavation Data and 3D Scan Use Agreement. Los
Angeles County of Museum of Natural History.
La Brea Tar Pits and Museum. 2018. 3D objects, images, maps and tables from Project 23.
Department of Research and Collections.
Lindsey, Emily. Personal communication with author. La Brea Tar Pits and Museum Department
of Research and Collections.
Microsoft 3D Builder. 2013. Accessed August 10, 2018. https://www.microsoft.com/en-us/p/3d-
builder/9wzdncrfj3t6.
Microsoft 3D Paint. 2016. Accessed August 10, 2018. https://www.microsoft.com/en-us/p/paint-
3d/9nblggh5fv99.
Microsoft Excel 2016. 2016. Accessed August 10, 2018. https://www.microsoft.com/en-
us/store/b/excel-2016.
Microsoft Office. 2018a. “Atan.” Microsoft Office. Accessed August 14, 2018.
https://support.office.com/en-us/article/atan-function-50746fa8-630a-406b-81d0-
4a2aed395543.
Microsoft Office. 2018b. “Atan2.” Microsoft Office. Accessed August 14, 2018.
https://support.office.com/en-us/article/atan2-function-c04592ab-b9e3-4908-b428-
c96b3a565033.
Microsoft PowerPoint 2016. 2016. Accessed August 10, 2018. https://www.microsoft.com/en-
us/store/b/powerpoint-2016.
Nawrocki, Stephen P., Ph.D., D.A.B.F.A. 1996. “An Outline of Forensic Archaeology.”
University of Indianapolis Archaeology & Forensics Laboratory. Accessed August 18,
2018. http://archlab.uindy.edu/documents/ForensicArcheo.pdf.
69

Pham, Kacey Johnson. 2015. “GIS Data Curation and Web Map Application for La Brea Tar Pits
Fossil Occurrences in Los Angeles, California.” Master’s thesis, University of Southern
California.
Sacchi, Eva, Ph.D. and Umberto Nicosia, M.Sc. 2013. “Forensic Paleontology: A Tool for
‘Intelligence’ and Investigation.” Journal of Forensic Sciences 58, no. 3 (May): 651-657.
Accessed August 18, 2018. https://doi.org/10.1111.1556-4029.12084.
Shaw, C.A. 1982. “Techniques Used in Excavation, Preparation, and Curation of Fossils from
Rancho La Brea.” Curator: The Museum Journal 25, no. 1 (March): 63-71. Accessed
August 18, 2018. https://doi.org/10.1111/j.2151-6952.1982.tb00583.x.
Sims, Brian. Personal communication with author. Esri.
SketchUp Pro 2017. 2017. Accessed August 10, 2018. https://www.sketchup.com/download/all.
Turner, Robin D. 2006. Archaeological and Paleontological Monitoring Report: LACMA
Transformation Project, Los Angeles, California. Venice: ArchaeoPaleo Resource
Management, Inc.
United States Marine Corp. 2001. SCRS 1201 Sergeants Course Land Navigation Student
Handout. 25 Jan 2001. Marine Corps University.
70

Appendix A ArcGIS Pro Version 2.1. Steps
INITIAL STEPS:
1. Open ArcGIS Pro. Under Create a New Project, select Blank. Name project
LaBreaTarPits_P23Box13a_FossilsModel.aprx. This step creates a default file
geodatabase in the Catalog called LaBreaTarPits_P23Box13a_FossilModel.gdb. Add
Pham’s geodatabase LaBreaTarPitsApp_Backup_20160228.gdb to Databases in Catalog
tab.

2. Under Insert tab, select New Map and clicked New Scene. Under Map tab, select
Basemap and clicked Imagery. Under View tab, click Convert to view a 2D Scene
(Scene_2D) as well as a 3D Scene (Scene).
71

3. Right click on Scene under Contents, click Properties, select Elevation Surface, and
check box “Allow Navigation below ground” and verify that Elevation source is Imagery
basemap (Location: https://elevation3d.arcgis.com/arcgis/rest/services/WorldElevation3D
/Terrain3D/ImageServer. Service Name: WorldElevation3D/Terrain3D. Vertical Units:
Meters). Then select Coordinate Systems and verify Coordinate System is geographic
CGS_WGS_1984.
72

4. In Scene_2D, add LABREA_FOSSILOCALITIES feature class (projection:
WGS_1984_Web_Mercator_Auxiliary_Sphere) from
LaBreaTarPitsApp_Backup_20160228.gdb.

73

5. Open LABREA_FOSSILOCALITIES attribute table and select Box 13a feature. Right
click on the feature class, select Selection and click Make Layer From Selected Features.

6. Right click on the symbol below newly created layer, LABREA_FOSSILLOCATIES
selection, to open the Symbology template. Select Properties and change the symbol
color.
74

7. Then right click LABREAFOSSILOCALITIES selection, select Data and click Export
Features.

75

8. Open Copy Features template and under Parameters tab, name Output Feature Class:
P23_Box13a under LaBreaTarPits_P23Box13a_FossilsModel.gdb. Under Environments
tab, indicate Output Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere
and indicate Output has Z Values: Same as Input. Click symbol under P23_Box13a and
change color in Symbology template.

9. Remove LABREA_FOSSILLOCALITIES selection layer. Run Add Geometry Attributes
tool on P23_Box13a to calculate x and y in meters. Under Add Geometry Attributes
template click on Parameters tab and indicate Coordinate System:
WGS_1984_Web_Mercator_Auxiliary_Sphere. Then click on Environments tab and
indicate Output Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere. Add
the x and y data to the Excel table CoordinateTransformation.xlsx for coordinate
transformation step.
76

COORDINATE TRANSFORMATION:
10. Right click on LaBreaTarPits_P23Box13a_FossilsModel.gdb, select Import and click
Table. Under Table to Table template under Parameters tab, indicate Input Rows:
Sheet1$ of AllBones_DataForCoordinateTransformation.xlsx, Output Location:
LaBreaTarPits_P23Box13a_FossilsModel.gdb and Output Table:
AllBones_DataForCoordinateTransformation. Then run Table to Table tool.
77

11. Run the Bearing Distance to Line tool with Input Table:
AllBones_DataForCoordinateTransformation and Output Feature Class:
AllBones_BearingDistancetoLine under LaBreaTarPits_P23Box13a_FossilsModel.gdb.
In the Bearing Distance to Line template under Parameters tab, match X and Y fields to
origin fields, match bearing and distance fields, indicate Distance Units: meters and
Bearing Units: decimal degrees, match ID to ID field and indicate Spatial Reference:
WGS_1984_Web_Mercator_Auxiliary_Sphere.
78

12. Run Feature Vertices to Points tool: in template under Parameters tab, indicate Input
Features: AllBones_BearingDistanceToLine, name Output Feature Class:
AllBones_FeatureVerticesToPoints under LaBreaTarPits_P23Box13a_FossilsModel.gdb
and indicate Point Type: End vertex. Under Environments tab, indicate Output
Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere and left Output has Z
Values: Same as input.
79

13. Use Join Field tool to add fields to attribute table lost in previous step. Under Join Field
template under Parameters tab, indicate Input Table: AllBones_FeatureVerticesToPoints,
Input Join Field: ID, Join Table: AllBones_DataFor CoordinateTransformation, Output
Join Field: ID and Join Fields: Element, Elevation_m, Level, Point, Spec_Num, and
Taxon.
80

14. Select first set of coordinates (P23_32911) from AllBones_FeatureVerticesToPoints
attribute table, right click on feature class, select Selection, click Make Layer From
Selected Features.

81

15. Right click on symbol below newly created AllBones_FeatureVerticesToPoints selection
layer to change the color in the Symbology template. Right click on
AllBones_FeatureVerticesToPoints selection layer, select Data and click Export Features.

16. Under Copy Features Template under Parameters tab, name Output Feature Class:
P23_32911 under LaBreaTarPits_P23Box13a_FossilsModel.gdb. Under Environments
tab, indicate Output Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere
and Output have Z Values: Enabled. Run tool. Right click on symbol below newly
created P23_32911 feature class to change the color in the Symbology template. Remove
AllBones_FeatureVerticesToPoints selection layer from Contents.
82

17. Repeat the last three steps for the remaining 18 bones’ coordinates.

83

18. In the Scene tab, add P23_Box13a and P23_32911 feature classes. Open the P23_32911
attribute table. Select P23_32911 under Contents and click the Edit tab and then Modify.
Select the three-point features in the attribute table. Click the Move To tool and under
Move To template, select Method: Absolute and for Z: enter the elevation in meters for
the selected point listed in the attribute table. Complete this step for the remaining two
points in the feature class and save edits.

19. Clear the selection and run Add Geometry Attributes tool on P23_32911. Under the Add
Geometry Attributes template, click Parameters tab and indicate Input Features:
P23_32911, Geometry Properties: Point x-, y-, and m- coordinates and Coordinate
System: WGS_1984_Web_Mercator_Auxiliary_Sphere. Click Environments tab and
specify Output Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere. This
step is to verify the that the Move To tool was executed properly by generating a z value
as well as to calculate x and y data for each coordinate.
84

20. Repeat the last two steps for the remaining 18 fossils’ sets of coordinates.

85

IMAGE GEOREFERENCING & BOX AND GRIDS CREATION:
21. In ArcGIS Pro in Scene_2D, right-click Folders under the Catalog tab and click Add
Connection to access Museum Data folder. Drag Project23LocalityMap.jpg into
Scene_2D and adjusted Transparency to 50% under the Appearance tab to ensure the
imagery is facing north so that Box 13a orientation can be deduced. Use the Georeference
tool to approximately scale and align the map with Box 13a and the surrounding
property.

22. Add Project23LocalityMap.jpg to the Scene. Do not adjust transparency. Right click on
LaBreaTarPits_P23Box13a_FossilsModel.gdb in Catalog, select New and click Feature
Class. Under Create Feature Class template under the Parameters tab, indicate Feature
Class Name: P23_Box13a_obj, Geometry Type: Multipatch and Coordinate System:
WGS_1984_Web_Mercator_Auxiliary_Sphere. Under the Environments tab, specify
86

Output Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere and Output
has Z Values: Enabled.

23. Select P23_Box13a_obj in Contents, select Edit tab, click Create and select the Cube in
the Active Template. Under Active Template, click the icon to unlink size dimensions
together to scale proportionally and indicate Height (Z): 1.40208 m, Width (X): 1.9812
m, Depth (Y): 1.9812 m. and Rotation: 38 dd. Digitize cube in Scene approximately in
line with Box 13a on Project23LocalityMap.jpg. Set Transparency to 75% under the
Appearance tab.
87

24. Under Edit tab, select Modify and use Move tool to center box over P23_Box13a point.

25. Right click on LaBreaTarPits_P23Box13a_FossilsModel.gdb in Catalog, select New and
click Feature Class. Under Create Feature Class template under the Parameters tab,
88

indicate Feature Class Name: GridB3_Level2, Geometry Type: Multipatch and
Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere. Under the
Environments tab, specify Output Coordinate System:
WGS_1984_Web_Mercator_Auxiliary_Sphere and Output has Z Values: Enabled.
Selected GridB3_Level2 in Contents and select Edit tab. Click Create and select the Cube
in the Active Template. Under Active Template, click the icon to unlink size dimensions
together to scale proportionally and indicate Height (Z): 0.25 m, Width (X): 1 m and
Depth (Y): 1 m. Do not adjust the Rotation: 0 dd. Position and digitize cube in Scene on
P23_Box13a at approximately 0 cm W, 50 cm N in Grid B3 Level 2. Set Transparency to
50-75% under the Appearance tab. Save edits. GridB3_Level 2 is pictured in blue above
below.

26. Select P23_Box13a_obj in Contents and select Edit tab. Click Modify and select Move
To tool. Select P23_Box13a_obj in Scene. Under Move To template, indicate Method:
89

Absolute and Z: 46.03792 m. This process moves P23_Box13a_obj below ground
surface. Select symbol below P23_Box13a_obj. Under Format Mesh Symbol template
under Symbology, select Properties tab and select Color: Electron Gold.

27. Select GridB3_Level2 in Contents, select Edit tab, click Modify and select Move To tool.
Select GridB3_Level 2 in Scene. Under Move To template, specify Method: Absolute
and Z: 46.94 m. This process moves GridB3_Level2 below ground surface. Select
symbol below GridB3_Level2 and under Format Mesh Symbol template under
Symbology, select Properties tab and select Color: Big Sky Blue.
90

28. Under LaBreaTarPits_P23Box13a_FossilsModel.gdb in Catalog, right click
GridB3_Level2 and click copy. Then right click
LaBreaTarPits_P23Box13a_FossilsModel.gdb and click Paste. Rename copy of
GridB3_Level2 feature class to GridB3_Level1 and add GridB_Level1 to the Scene.
Select GridB3_Level1 in Contents, select Edit tab, click Modify and select Move To tool.
Select GridB3_Level1 in Scene. Under Move To template, indicate Method: Absolute
and Z: 47.19 m. This process moves GridB3_Level1 below ground surface. Select
symbol below GridB3_Level2 and under Format Mesh Symbol template under
Symbology, select Properties tab and select Color: Mars Red. Set Transparency to 75%
under Appearance tab. Transparencies of Box 13a and grids are set at 50-75% as needed
for best viewing below.
91

29. To add attributes to P23_Box13a_obj, GridB3_Level1 and Grid_B3 Level2, right click
each feature class in the Contents pane and select Attribute Tables. Under Data tab, select
Fields and click Add New Field on Fields table. For P23_Box13a_obj., add Fields:
Name, Width_WE_m, Length_SN_m, Height_m and Elevation_m. For GridB3_Level1
and GridB3_Level2, add Fields: Name, Width_m, Length_m, Height_m and
Elevation_m. Save changes in Fields tab and then fill in attribute information in tables.
92

FOSSIL 3D OBJECT IMPORT AND PLACEMENT INSTRUCTIONS:
30. Turn off ProjectLocalityMap.jpg layer. Right click
LaBreaTarPits_P23Box13a_FossilsModel.gdb in Catalog, select New and select Feature
Class. Under Parameters tab, indicate Feature Class Location:
LaBreaTarPits_P23Box13a_FossilsModel.gdb, Feature Class Name: P23_33245_obj,
Geometry Type: Multipatch and Coordinate System:
WGS_1984_Web_Mercator_Auxiliary_Sphere. Under Environments tab, specify Output
Coordinate System: WGS_1984_Web_Mercator_Auxiliary_Sphere and Output has Z
Values: Enabled. Click Run. Click on P23_33245_obj in Contents tab and select Edit tab
and then Create. Under Active Template, select Model File icon and add OBJ file of
fossil P23 33245, change Size to meters and Reset Height (Z) to 3 m as Width (X) and
Depth (Y) scaled proportionally. Digitize fossil above Box 13a centroid P23_Box13a.
Save edits.
93

31. With P23_33245_obj selected in Scene, click Appearance tab. Under Symbology, select
Single Symbol and click on symbol below it. Under Format Mesh Symbol tab, click
Properties tab and change color to Leather Brown. Click Apply.

94

32. Select P23_33245_obj in Scene, click Edit tab, click Modify. Under Modify Features,
select Scale. Click and hold mouse on yellow box in Scene to manually scale down
feature while maintaining proportions. Click box with green check mark to apply
changes.

33. Under Modify Features, click Move and click and hold mouse over green arrow to move
the feature below the ground surface. Click on red and blue arrows to move the feature
side to side.
95

34. Turn off all 3D layers except P23_33245 and P23_33245_obj. Right click on P23_33245
in Contents and select Labeling Properties and then Class tab. Under Fields, double click
Point to add to Expression box. Under Symbol tab, change color to “Fire Red.” Right
click on P23_33245 in Contents and select Label.

96

35. Select P23_33245_obj. Under Modify Features, select Rotate and use green, red and blue
rings to rotate feature to align with its respective coordinates. Use imagery from museum
of coordinate locations on fossil as reference for aligning the feature.

97

36. Utilize Move, Scale and Rotate tools to adjust the position of P23_33245 until it aligns
with its coordinates as closely as manually possible. Turn on GridB3_Level1 and
GridB3_Level2 to see placement.

98

37. To add attributes to P23_33245_obj, run Join Field tool with corresponding coordinate
feature class. Under Parameters tab, indicate Input table: P23_33245_obj, Input Join
Field: OBJECTID, Join Table: P23_33245, Output Join Field: OBJECTID and Join
Fields: Spec_Num, Taxon, Element and Level. Save edits.

99

38. Repeat steps 30-37 for the remaining 18 fossils.

39. Right click on each feature class in the geodatabase, select “Edit Metadata” and add the
function and origin of each dataset. Save all edits.
100

Appendix B Hand-Drawn and PowerPoint Fossil Diagrams
All locations were confirmed by La Brea Tar Pits and Museum scientists (Howard and Farrell,
pers. comm.).
1. P23 32911 Smilodon fatalis rib
101

102

2. P23 32912 Canis dirus femur
103

104

3. P23 33237 Canis dirus dentary
105

106

4. P23 33238 Canis dirus humerus
107

108

5. P23_33239 Smilodon fatalis vertebra
109

110

6. P23 33240 Smilodon fatalis vertebra
111

112

7. P23_33241 Equus occidentalis vertebra
113

114

8. P23 33242 Bison antiquus rib
115

116

9. P23 33243 Aves tarsometatarsus
117

118

10. P23 33244 Smilodon fatalis scapula
119

120

11. P23 33245 Smilodon fatalis femur
121

122

12. P23 33246 Canis dirus dentary
123

124

13. P23 33247 Canis dirus femur
125

126

14. P23 33248 Smilodon fatalis dentary
127

128

15. P23 33250 Canis latrans tibia
129

130

16. P23 33251 Canis latrans humerus
131

132

17. P23_33252 Canis latrans innominate
133

134

18. P23 33253 Smilodon fatalis rib
135

136

19. P23 33254 Equus occidentalis scapula
137

Abstract (if available)

Abstract The La Brea Tar Pits and Museum in the middle of Los Angeles, California is a paleontological marvel containing numerous fossil-rich asphalt deposits. Until very recently, the museum only recorded their findings in non-spatial databases. As a continuation of work completed by a former USC Geographic Information Science and Technology student to develop spatial databases documenting artifacts for the museum, the main objective of this thesis project was to create a methodology for visualizing fossils as high-resolution 3D objects on a 3D map in their pre-excavation, in-situ locations. Museum scientists selected nineteen fossils from one asphalt deposit for mapping. The fossils were laser scanned by museum scientists, and the resulting 3D objects were provided for this project with accompanying locality data gathered in the manner of a traditional paleontological dig. The data required extensive processing prior to importing the 3D objects into a GIS, including image file conversion, location and orientation diagramming and steps for coordinate transformation from paleontological location measurements to real-world, geographic coordinates. The 3D objects were then imported and manually positioned in a 3D GIS map beneath the earth’s surface. The resulting 3D model provides an interactive, GIS-enabled visualization of the nineteen fossils in their original locations and orientations prior to excavation. It is intended that this project support future research efforts of the museum scientists in spatial analysis and modeling of fossils and substrate (tar pits), ultimately to improve our understanding of Ice Age animals and the environments in which they lived and died. In addition, the results of this project serve as an example application of 3D GIS capabilities that can support forensic archaeology, an important tool in intelligence and criminal investigations. Lastly, it is anticipated that the georeferenced 3D objects, as well as this 3D fossil visualization, may become part of an interactive museum exhibit in the future.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

GIS data curation and Web map application for La Brea Tar Pits fossil occurrences in Los Angeles, California

PDF

Comparative 3D geographic web server development: visualizing point clouds in the web

PDF

Filling in the gaps: 3D mapping Arizona's Basin and Range aquifer in the Prescott Active Management Area

PDF

An examination of close-range photogrammetry and traditional cave survey methods for terrestrial and underwater caves for 3-dimensional mapping

PDF

Utilizing existing museum collections and GIS for paleontological site assessment and management

PDF

3D visualization models as a tool for reconstructing the historical landscape of the Ballona Creek watershed

PDF

Geological modeling in GIS for petroleum reservoir characterization and engineering: a 3D GIS-assisted geostatistics approach

PDF

An open-source program to animate and visualize the recorded temperature and relative humidity data from dataloggers including the building 3D geometry

PDF

Assessing the impact of a web-based GIS application to promote earthquake preparation on the University of Southern California University Park Campus

PDF

From ruins to pixels: using remote sensing and GIS to analyze, document, and visualize archaeological sites and ancient roadways in Chaco Canyon, New Mexico

PDF

Mapping native plants: a mobile GIS application for sharing indigenous knowledge in Southern California

PDF

Semantic structure in understanding and generation of the 3D world

PDF

Modeling potential impacts of tsunamis on Hilo, Hawaii: comparison of the Joint Research Centre's SCHEMA and FEMA’s HAZUS inundation scenarios

PDF

GIS-aided production of equipment locator maps for Metro rail maintenance and support

PDF

Spatial and temporal patterns of long-term temperature change in Southern California from 1935 to 2014

PDF

The movement of Mexican migration and its impact based on a GIS geospatial database

PDF

3D vessel mapping techniques for retina and brain as an early imaging biomarker for small vessel diseases

PDF

Structural clustering analysis of CVMS-4.26: a 3D seismic velocity model for southern California

PDF

3D assessment of virtual bracket removal for modern orthodontic retainers: a prospective clinical study

Asset Metadata

Creator Hill, Kristiane Michele (author)

Core Title 3D fossil visualization and mapping of the La Brea Tar Pits, Los Angeles, California

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 09/24/2018

Defense Date 08/23/2018

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag 3D GIS,3D mapping,3D model,3D object,3D visualization,archaeology,coordinate transformation,forensic archaeology,fossil,La Brea Tar Pits,museum exhibit,OAI-PMH Harvest,paleontology,subsurface mapping

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Swift, Jennifer (committee chair), Loyola, Laura (committee member), Wu, An-Min (committee member)

Creator Email hill.kristie@gmail.com,kristimh@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-73387

Unique identifier UC11672296

Identifier etd-HillKristi-6778.pdf (filename),usctheses-c89-73387 (legacy record id)

Legacy Identifier etd-HillKristi-6778.pdf

Dmrecord 73387

Document Type Thesis

Format application/pdf (imt)

Rights Hill, Kristiane Michele

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA