Coastal vulnerability assessment for archaeological sites on San Clemente Island and San Nicolas Island, California

PDF

Download

Open document

Flip pages

Summarize

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

Coastal Vulnerability Assessment for Archaeological Sites on
San Clemente Island and San Nicolas Island, California

by

Tori Kristin Oulie

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

Copyright © 2018 by Tori Oulie

To my parents, whose constant support is always my driving factor.

1
Table of Contents
List of Figures ................................................................................................................................. 3
List of Tables .................................................................................................................................. 4
Acknowledgements ......................................................................................................................... 5
List of Abbreviations ...................................................................................................................... 6
Abstract ........................................................................................................................................... 7
Chapter 1 Introduction .................................................................................................................... 9
1.1. Channel Islands Archaeology ...........................................................................................10
1.2. Motivation .........................................................................................................................12
1.3. Research Objectives ..........................................................................................................13
1.4. Thesis Layout ....................................................................................................................14
Chapter 2 Related Work ................................................................................................................ 15
2.1. Coastal Vulnerability ........................................................................................................15
2.1.1. Sea Level Rise ..........................................................................................................15
2.1.2. Coastal Erosion ........................................................................................................17
2.2. Coastal Vulnerability Assessment Methods .....................................................................18
2.3. Tools Adapted for Archaeology and Cultural Resource Management .............................23
Chapter 3 Methods ........................................................................................................................ 25
3.1. Data Acquisition and Preparation .....................................................................................25
3.1.1. Archaeological Site Data .........................................................................................27
3.1.2. Elevation ..................................................................................................................28
3.1.3. Slope ........................................................................................................................28
3.1.4. Inundation ................................................................................................................29
3.1.5. Generalized Rock Types ..........................................................................................31
3.1.6. Vegetation ................................................................................................................32

2
3.1.7. Coastline ..................................................................................................................34
3.2. Coastal Vulnerability Index (CVI) ....................................................................................35
3.3. Cultural Resource Vulnerability Index (CRVI) ................................................................36
3.3.1. Coastline CVI ...........................................................................................................37
3.3.2. Archaeological Site Distance From Coast ...............................................................37
3.3.3. CRVI Equation .........................................................................................................38
Chapter 4 Results and Discussion ................................................................................................. 39
4.1. Vulnerability Rankings for Environmental Variables ......................................................39
4.1.1. Slope ........................................................................................................................40
4.1.2. Inundation ................................................................................................................42
4.1.3. Generalized Rock Type ............................................................................................44
4.1.4. Vegetation ................................................................................................................45
4.2. Coastal Vulnerability Index ..............................................................................................47
4.2.1. CVI for Entire Study Area .......................................................................................47
4.2.2. Coastline CVI ...........................................................................................................50
4.3. Cultural Resource Vulnerability Index (CRVI) ................................................................53
4.3.1. San Nicolas Island ....................................................................................................53
4.3.2. San Clemente Island ................................................................................................55
Chapter 5 Conclusions .................................................................................................................. 59
5.1. Limitations ........................................................................................................................59
5.2. Future Research ................................................................................................................61
5.3. Final Take Away ...............................................................................................................62
References ..................................................................................................................................... 63
Appendix A – A Short History of San Clemente Island and San Nicolas Island ......................... 68

3
List of Figures
Figure 1 Study area including surrounding areas 10
Figure 2 Sea Level Rise Viewer from NOAA’s Office of Coastal Management with view of the
southern Channel Islands including San Nicolas and San Clemente Islands 13
Figure 3 Global sea level change from 400,000 years ago to present (NOAA 2010) 16
Figure 4 Coastal erosion from January 2 – 24, 2010 from storm surges in San Francisco 17
Figure 5 The distribution of Texas coastal risk using CVI in 30-m resolution 21
Figure 6 Archaeological sites ranked according to CRVI 24
Figure 7 The workflow of spatial data preparation for the calculation of CVI 27
Figure 8 Slope in degrees for San Nicolas Island and San Clemente Island 29
Figure 9 CRVI Workflow 36
Figure 10 The area percentage of vulnerability ranking for the CVI variables in the study area 40
Figure 11 Classification of slope from least vulnerable to sea level rise to most vulnerable 41
Figure 12 Classification of water inundation zones based on the predicted sea level rise ranked
from most vulnerable to least vulnerable 43
Figure 13 The vulnerability classification of generalized rock type 45
Figure 14 The vulnerability classification of vegetation 46
Figure 15 The Coastal Vulnerability Index for the entire San Clemente Island 48
Figure 16 The Coastal Vulnerability Index for the entire San Nicolas Island 50
Figure 17 Coastline CVI for San Nicolas Island 51
Figure 18 Coastline CVI for San Clemente Island 52
Figure 19 CRVI for San Nicolas Island with archaeological site rankings 55
Figure 20 The CRVI for archaeological sites on San Clemente Island 57
Figure 21 A photograph of a woman believed to be the Lone Woman of San Nicolas Island 70

4
List of Tables
Table 1 Gornitz et al (1994)’s coastal risk classification scheme 19
Table 2 CVI Equations from Gornitz, Beaty, and Daniels (1997) 21
Table 3 Data sources 25
Table 4 Reclassification of Inundation Levels 30
Table 5 Generalized Rock Type reclassification rankings 31
Table 6 Vegetation ranking for San Clemente Island 32
Table 7 Vegetation ranking for San Nicolas Island 33
Table 8 Coastal Vulnerability Index ranking 35
Table 9 Distance from coastline ranking for archaeological sites 37
Table 10 Ranking of CRVI variables on San Nicolas Island 54
Table 11 Ranking of CRVI variables on San Clemente Island 56

5
Acknowledgements
I am grateful to my loving parents, who have encouraged me through every step of this thesis. I
would like to thank Dr. An-Min Wu, for her constant direction and encouragement. For the
faculty of the Spatial Sciences Institute of the University of Southern California and how they
have inspired this next step in my life. I am also grateful for Dr. Colleen Delaney and Andy
Yatsko, who opened doors and led me to focus my thesis on my two academic passions
(archaeology and GIS) by connecting me to the cultural resource team at NAVFAC. I would like
to thank the U.S. Navy NAVFAC Southwest Cultural Resources team for the opportunity to
work towards a great cause with amazing archaeological data. And finally, I would like to thank
Austin, for his kindness and patience through the stresses of thesis writing.

6
List of Abbreviations
CRVI Cultural Resource Vulnerability Index
CVI Coastal Vulnerability Index
GIS Geographic Information System
NAVFAC Naval Facilities Engineering Command
SCI San Clemente Island
SNI San Nicolas Island

7
Abstract
Archaeology allows us to see our human past and who we are as a people. This is a global
narrative that spans the entirety of human existence. Many archaeological sites are delicate and
are often unknowingly destroyed by human development. Because of this, pristine and protected
islands offer a complete wealth of archaeological information. Stewardship programs and
regulations set in place for protecting these cultural resources have been set into place on
federally owned lands. San Nicolas Island and San Clemente Island, two of the Channel Islands
owned by the United States Navy, are among the most well-documented and protected locations
for archaeological sites in the United States. However, many of these sites are currently at risk
from inundation and erosion. Global sea level rise not only potentially inundate the coastal zones
but also accelerate geological erosion processes.
To help the U.S. Navy understand and protect against the threats from these natural
processes, this study aims to identify the at-risk archaeological sites on San Clemente Island and
San Nicolas Island. A spatial-explicit Coastal Vulnerability Index (CVI) was developed from the
ranked vulnerability score of environmental variables, including slope, inundation, generalized
rock type, and vegetation, using a Geographic Information Systems (GIS). Based on the CVI, a
Cultural Resource Vulnerability Index (CRVI) was developed to rank the coastal vulnerability of
the archaeological sites on the two islands
The results of the CRVI showed that 3.6% of the archaeological sites on San Nicolas
Island and 19.2% of the archaeological sites on San Clemente Island fall within the Highly
Vulnerable to Very Highly Vulnerable categories. The CRVI informs the land managers in the
U.S. Navy an earlier response time to save these at-risk sites that may be completely destroyed in
the next 100 years. With the result from the CRVI, further actions can be taken to mitigate and/or

8
excavate these at-risk sites to preserve the rich cultural resources of San Nicolas Island and San
Clemente Island.

9
Chapter 1 Introduction
Archaeological sites are an important part of our collective human history. These fragile
resources provide a wealth of knowledge that not only affect our understanding of the past, but
also allow us to understand the trajectory of where our human story is headed (Reeder, Rick, and
Erlandson 2012). Archaeology allows us as humans to find the source of who we are as a people.
This is a global narrative that spans the entirety of human existence.
The study of material culture pushes past the written historic literature that focuses on the
world’s great individuals and societies, and allows for researchers to understand the everyday
lives of all cultures, even if their unique histories were thought to be lost. Similar to modern day
living, evidence of ancient cultures is often found in coastal areas due to the bountiful resources
and pleasant climate. However, many of these sites are at risk because continuous erosion and
wave action effectively tear apart these sites and destroy the resources found within. While
archaeological sites might survive under water, the wave actions are incredibly damaging to the
sites and erosion processes (Jones 2017). With global sea level projected to rise as high as two
meters by 2100, sea level rise is expected to further exacerbate these factors and weaken coastal
zone stability within the next 100 years (Gornitz, Beaty, and Daniels 1997; Lindsey 2017;
NOAA 2010).
This study started with the request by the U.S. Navy to identify at-risk archaeological
sites due to the potential sea level rise on the Navy-owned San Nicolas and San Clemente
Islands. San Clemente Island and San Nicolas Island off the coast of Southern California are one
of the most well-documented areas in the United States (Chiles 2015) (Figure 1). These islands
are owned and protected by the United States Navy from typical human threats such as looting
and coastal development. Through archaeological surveys conducted by the Naval Facilities

10
Engineering Command (NAVFAC) Southwest Division, 5,058 archaeological sites have been
found on San Clemente Island and 535 archaeological sites have been found on San Nicolas
Island. With sites dating back to approximately 8,000 to 9,000 years ago on the two islands, a
wealth of cultural knowledge can be studied from early years to today (Jazwa and Perry 2013).

Figure 1 - Study area including surrounding areas.
1.1. Channel Islands Archaeology
The archaeology of the Channel Islands is unique due to the islands’ relative isolation,
lack of development, and generally arid climate (Braje 2010). Due to these variables, the rich

11
archaeological history of these islands provides a wealth of knowledge to archaeologists.
Archaeological evidence has provided researchers with an abundance of data that will influence
the ways we see how Native Americans lived and even arrived in the Americas. In fact, the
Channel Islands are at the center of a debate on the Americas’ early settlement.
Because of the early dates of archaeological sites on these islands and the ancient human
remains that we now call Arlington Springs Man, it is now believed that the earliest human
ancestors to reach the Americas probably followed the coastline down from Alaska to Central
and South America while also spreading eastward instead of the initial theory that these
ancestors migrated between glaciers following game animals through Canada (Chiles 2015; Rick
et al. 2005; Erlandson et al. 2007). Among the sites found that contribute to the new North
American migration theory, Arlington Springs Man (found on Santa Rosa Island with bones
carbon dated to around 13,000 years ago, making the remains one of the oldest in North
America) and numerous sites showing evidence of early mastery of seafaring (shown in very
early fishing technology and shell mounds dating back to over 10,000 years ago) are perhaps the
most notable (Erlandson et al. 2007). However, even before these sites were found,
archaeologists and other researchers were drawn to these islands.
The archaeology of the Channel Islands has been a source of intrigue for scientists since
the 1870s when the first researchers searched the island for artifacts to fill museum shelves
(Yatsko 2000). It was not until the 1950s that professional research and archaeological
excavation was established across San Nicolas, San Clemente, Anacapa, Santa Catalina, and
Santa Barbara Islands through University of California at Los Angeles’ (UCLA) Archaeological
Survey (Yatsko 2000). These surveys focused less on collecting from cemetery sites and more on

12
shell refuse piles (commonly known as shell “middens”) to gain a better understanding of the
native peoples’ day-to-day life (Yatsko 2000).
With the focus less on collecting and more on understanding, archaeology has created a
research boom on these islands to understand many aspects of ancient life that include human
impact on island ecosystems, subsistence strategies, Holocene settlement, marine adaptations,
trade systems, and the emergence of cultural complexity (Rick et al. 2005). Today, the entirety of
San Nicolas Island has been surveyed and most of San Clemente Island has been surveyed with
results that include a vast variety of sites that can contain burials, housing depressions, quarries,
middens, tool production sites, and many more (Rick et al. 2005). More information about the
history of these islands can be seen in Appendix A.
1.2. Motivation
With the concerns of sea level rise, the United States Navy is in pursuit of saving many
archaeological sites through their NAVFAC Cultural Resources Program. Through a partnership
with the program, this study attempts to understand which of the archaeological sites are most at
risk on Navy-controlled San Clemente Island and San Nicolas Island. Although these islands
have been well surveyed, a study of this kind has yet to be accomplished for the two islands. This
study uses geographic information systems (GIS) to examine the sites by exploring spatial-
explicit assessment methods such as the Coastal Vulnerability Index (CVI). Through these
assessment methods, a multitude of factors that could endanger archaeological sites, such as
water inundations and coastal erosion are considered (Anderson et al 2017; Reeder, Rick and
Erlandson 2012).
Although NOAA has conducted many studies on sea level change, their simulation
models have not covered to the spatial extents of San Nicolas Island and San Clemente Island.

13
As seen in Figure 2, NOAA’s sea level rise estimator does not cover the two islands (NOAA,
n.d.). These two islands are exempt from the simulation perhaps due to their close connection
with the Navy and how the Navy has protected these islands and the data attached to them.

Figure 2 - Sea Level Rise Viewer from NOAA's Office of Coastal Management with view of the
southern Channel Islands including San Nicolas and San Clemente Islands.
1.3. Research Objectives
San Clemente Island and San Nicolas Island are unique because of their deep ties to Navy
stewardship and possession. In order to continue their pursuit of stewardship, the NAVFAC
Cultural Resources team needs a viable source of data to protect the archaeological sites at risk
from rising sea levels. Viable data for the Navy’s use is the primary goal of this study. The goal
of this study is to understand which archaeological sites are most at risk and to understand which
portions of the coastline are most vulnerable on San Nicolas Island and San Clemente Island.
The study is aimed to recognize at-risk archaeological sites on the two islands, not only due to
sea level rise threats but more broadly on overall coastal vulnerability.

14
1.4. Thesis Layout
The following chapters will outline previous works done in the field, the methods used to
complete this study, the results of this work, and a conclusion to the works created in this study.
Chapter 2 will examine how previous researchers have designed and understood coastal
vulnerability through assessment processes and how these assessment processes have been
adapted for use in the archaeological field. The methods used in this study are examined in
Chapter 3 and include the introduction of data editing, the Coastal Vulnerability Index (CVI)
creation and use, and the final creation of the Cultural Resource Vulnerability Index (CRVI).
Once methods are established, Chapter 4 describes the results of this study. Finally, Chapter 5
takes a critical look at the results and concludes with the successes and changes that can be made
in future renditions of this study.

15
Chapter 2 Related Work
Coastal assessment is an important step towards future coastal mitigation. Researchers have
attempted to understand how coasts change and what can be done to prevent coastal devastation
for many years. This chapter reviews past studies of coastal vulnerability assessments to bridge
what has been done in the past and what this study accomplishes. It starts with an overview of
the environmental characteristics that contribute to coastal vulnerability (Section 2.1), moves on
to describe the coastal vulnerability assessments (Section 2.2), and finishes with the assessment
methods for cultural resources along the coast (Section 2.3).
2.1. Coastal Vulnerability
Sea level rise and coastal erosion are two major threats to communities and existing
archaeological sites near coastlines. Sea level rise, while gradual, has received much attention for
research in recent years due to its linkage to global climate change. On the other hand, near-coast
areas also experience the danger from eroding by dynamic ocean processes. The causes and
projections of sea level rise are reported in Section 2.1.1, and the environmental characteristics
affecting coastal erosion are presented in Section 2.1.2.
2.1.1. Sea Level Rise
Oceanography research has been attempting to understand the potential sea level rise.
Water expansion from warmed ocean waters and ice/glacier melt are two major contributors to
sea level change (Williams and Gutierrez 2009), and have increased the volume of ocean waters
significantly over the past 200 years (NOAA 2010). While this global sea level change has been
seen during the natural warming and cooling cycles for the last 400,000 years (Figure 3) (NOAA

16
2010), the pattern of sea level change in recent century was found greatly punctuated by the
periods of ice and glacial melt (Church et al. 2008).

Figure 3 - Global sea level change from 400,000 years ago to present (NOAA 2010)
The most notable sea level rise projections can be attributed to the National Oceanic and
Atmospheric Administration (NOAA) and the Intergovernmental Panel on Climate Change
(IPCC) (NOAA 2010 and IPCC 2013). The two agencies have shown slightly differing sea level
projections in the next 100 years. NOAA uses tidal gauges located all over the world and satellite
laser altimeters to measure sea level height and change for over 100 years. Combined with these
past records with temperature data (to factor thermal expansion in), the global mean sea level is
projected to have a very high confidence (greater than 90% chance) to rise between 0.2 meters
and 2.0 meters by 2100 (Lindsey 2017).
On the other hand, the IPCC’s Fifth Assessment Report (AR5) estimates sea level rise
using three different scenarios of global carbon dioxide (CO
2
) emission. If globally high CO
2
emissions continue, sea levels are projected to rise between 52 cm to 98 cm by Year 2100. If
there are aggressive emission reductions instated globally, the sea level rise could be reduced to
28 cm – 61cm by the same year. The best scenario estimated, as of 2014, is a 44cm rise by 2100
(IPCC 2013).

17
The affects of sea level rise can be seen in heightened water levels flooding previously
dry shorelines. Moreover, sea level rise enhances storm intensity, shoreline retreat, rates of
erosion, tidal ranges, and coastal flooding. These environmental changes result in further change
of the physical terrain on coastlines (Reeder, Rick, and Erlandson 2012; Rowland 1992; Westley
et al 2011; Erlandson 2012).
2.1.2. Coastal Erosion
Compared to sea level rise, coastal erosion is innate to all coastal areas. For naturally
caused erosion in coastlines, both water erosion and wind erosion can be involved, but water is
known as the major force. The processes, including wave actions, high tide and storm surge,
result in soil material loss and displacement. An example of this destructive force can be seen in
Figure 4 below. As seen in the Figure, storm surges can easily undercut previously secure
embankments, causing significant damage.

Figure 4 - Coastal erosion from January 2-24, 2010 from storm surges in San Francisco
(https://www.usgs.gov/media/images/coastal-erosion-san-francisco-0)
Erosion can impact various types of coastal geomorphology, but it is especially
prominent in sandy beach environments (Feagin, Sherman, and Grant 2005). In sandy beach
environments, wave directions, wave intensity, and availability of barriers from wave actions are
variables highly correlated to erosion’s effect on coastal vulnerability. On the other hand, for

18
cliffed, rocky coastal areas, coastal erosion is controlled by other sets of variables including the
hardness of rocks, structural weakness, coastal configuration, strength of stone compaction, cliff
height, and wave intensity (Feagin, Sherman, and Grant 2005; Shepard and Grant 1947). The
diversity of these variables makes it difficult to pinpoint the exact cause of major erosion events.
Many of the variables mentioned can be attributed to broader variables including slope,
soil, and geology. Vegetation can also strengthen coastlines against erosion. The severity of
erosion, therefore, not only depends on water force from the ocean. Site variables such as slope,
soil, geology, and vegetation inhibit or advance erosion processes. These site variables have been
studied spatially through geographic information systems (GIS) for better understanding of the
complex nature erosion prevention (Gornitz et al 1994; Özyurt and Ergin 2010; Aboudha and
Woodroffe 2010; Thieler and Hammar-Klose 2000).
2.2. Coastal Vulnerability Assessment Methods
One of the oldest methods to understand how coastal areas are affected by sea level rise is
the Bruun Rule. Based entirely on an erosion equation to calculate the probability of shoreline
retreat and erosion caused by sea level rise, Bruun (1962)’s study sparked more coastal
assessment studies that continue to evolve today. More recent studies have examined erosion,
much like Bruun, but most have included variables such as slope, geology, tide, and significant
wave height, among many more. These methods include the DINAS-Coast Dynamic Interactive
Vulnerability Assessment (DIVA), Coastal Zone Simulation Model (COSMO), and the Coastal
Vulnerability Index (CVI) (Kay and Travers 2008; McFadden, Vafeidis, and Nicholls 2003;
Kotinas et al, n.d.).
To understand how sea level rise will affect coastal regions, Vafeidis et al (2008)
attempted an assessment by instating the Dynamic Interactive Vulnerability Assessment (DIVA)

19
modeling tool through the DINAS-COAST project. This tool is a part of a global coastal
database that is intended to form a reliable, consistent basis for coastal vulnerability evaluations.
It uses a graphical user interface (GUI) that produces scenarios regarding climate change and
socio-economics to understand all aspects of vulnerability to sea level rise. The DIVA tool
allows for the database to be used in a practical way by modeling coastline segments in relation
to parameters provided by the database (Vafeidis et al 2008; DINAS-COAST 2006). DIVA,
unfortunately, is now unavailable due to high running costs.
The Coastal Zone Simulation Model (COSMO) is another modeling tool created to
improve coastal management. Created in Holland, the interactive tool explores the impact of
development for environmental and coastal protection. The data processed from this method
enables developers as a decision support tool. However, any data created from this process
would need to be further studied to ensure accurate results (Kay and Travers 2008).
The Coastal Vulnerability Index (CVI) was first introduced by Gornitz et al. (1990) as an
efficient way to spatially analyze the variables that influence coastline susceptibility. A CVI is a
spatial index of coastal vulnerability created by classifying environmental risk variables from
least to most susceptible using a GIS system. The index ranks variables from least to most
vulnerable and then uses layered spatial data to understand areas of least and most vulnerability.
The classification of the variables is in 5 categories: 1 – very low risk, 2 - low risk, 3 – moderate
risk, 4 - high risk, 5 – very high risk, shown in Table 1. The environmental risk variables include
elevation, lithology, subsidence, tropical storm probability, susceptibility to erosion, and storm
surge, but many variables can be added or subtracted based on coastline type and need.
Table 1 - Gornitz et al's coastal risk classification scheme (1994)
Variable Very Low (1) Low (2) Moderate
(3)
High (4) Very High (5)

20
Elevation
(m)
≥ 30.0 20.1 – 30.0 10.1 – 20.0 5.1 – 10.0 0 – 5.0
Geology
(relative
resistance to
erosion)
Plutonic
Volcanic
(lava)
High-medium
grade
metamorphics
Low-grade
metamorphics
Sandstone
and
conglomerate
(well-
cemented)
Most
sedimentary
rocks
Coarse and/or
poorly sorted
unconsolidated
sediments
Fine
unconsolidated
sediment
Volcanic ask

Landform
(geomorpho-
logy)
Rocky,
cliffed coasts
Fiords
Fiards
Medium
cliffs
Indented
coasts
Low cliffs
Glacial drift
Salt marsh
Coral reefs
Mangrove
Beaches
(pebbles)
Estuary
Lagoon
Alluvial plains
Barrier
beaches
Beaches
(sand)
Mud flats
Deltas
Rising sea
level change
(mm/year)
< - 1.0
Land rising
-1.0 – 0.99 1.0 – 2.0
Within
range of
eustatic rise
2.1 – 4.0
Land sinking
> 4.0
Shoreline
erosion/
Accretion
(m/year)
> 2.0
Accretion
1.0 – 2.0 - 1.0 - +1.0
Stable
- 1.1 - -2.0 < -2.0
Erosion
Mean tide
range (m)
< 1.0
Microtidal
1.0 – 1.9 2.0 – 4.0
Mesotidal
4.1 – 6.0 > 6.0
Macrotidal
Maximum
wave height
(m)
0 – 0.29 3.0 – 4.9 5.0 – 5.9 6.0 – 6.9 > 6.9

The classified environmental risk variables, represented in spatial layers in the GIS, were
next used to calculate a numeric value indicating coastal vulnerability for each unit of analysis,
typically a gridded cell along the coastline, as seen in Figure 5. There have been a variation of
the CVI equations used to calculate this index, many of which were recommended by Gornitz,
Beaty, and Daniels (1997) and Thieler and Hammar-Klose (1999). See Table 2 for the list of CVI
equations recommended by Gornitz, Beaty, and Daniels (1997). All equations involve the
multiplication or addition of all classified coastal risk values.

21

Figure 5 - The distribution of Texas coastal risk using CVI in 30-m resolution (Source: Gornitz
et al 1994)
Table 2 - CVI Equations from Gornitz, Beaty, and Daniels (1997)
Modifications of CVI Equations (Gornitz, Beaty, and Daniels 1997)
Product mean
𝐶𝑉𝐼=
(𝑥
!
∗𝑥
!
∗𝑥
!
∗… 𝑥
!
)
𝑛

Square root of product
mean

𝐶𝑉𝐼=
(𝑥
!
∗𝑥
!
∗𝑥
!
∗… 𝑥
!
)
𝑛

Modified product mean

22
𝐶𝑉𝐼=
[𝑥
!
∗𝑥
!
∗
!
!
𝑥
!
∗𝑥
!
∗𝑥
!
∗
!
!
𝑥
!
∗𝑥
!
]
𝑛−2

Average sum of squares

𝐶𝑉𝐼=
(𝑥
!
!
∗𝑥
!
!
∗𝑥
!
!
∗𝑥
!
!
… 𝑥
!
!
)
𝑛

Modified product mean
(2)

𝐶𝑉𝐼=
(𝑥
!
∗𝑥
!
∗𝑥
!
∗𝑥
!
… 𝑥
!
)
5
(!!!)

Sum of products
𝐶𝑉𝐼= 4𝑥
!
+ 4𝑥
!
+2 𝑥
!
+ 𝑥
!
+ 4𝑥
!
+2(𝑥
!
+ 𝑥
!
)

Where:
n = variables present
x
1
= mean elevation
x
2
= local subsidence trend
x
3
= geology
x
4
= geomorphology
x
5
= mean shoreline displacement
x
6
= maximum wave height
x
7
= mean tidal range

The main difference between these CVI equations is their sensitivity to data change
(Kotinas et al, n.d.). The mean value for the product of all the classified environmental risk
values (product mean) is highly sensitive to small changes in the vulnerability ranking of
variables while the square root of the product mean allows for a more consistent ranking system
(Kotinas et al, n.d.). As the result, the equation used in this study is the square root of product
mean introduced by Thieler and Hammar-Klose’s (1999) study:
𝐶𝑉𝐼=
𝑎 ∗𝑏 ∗𝑐 ∗𝑑 ∗𝑒 ∗𝑓
6

where a was geomorphology, b was coastal slope, c was relative sea-level rise rate, d was
shoreline erosion rate, e was mean tide range, and f was mean wave height.

23
While the DIVA and COSMO systems have been successful in their own right, the CVI
continues to be the most popular method to inform coastal vulnerability. Since Gornitz et al
(1990), the CVI approach has been used for coastal vulnerability assessments not only in the
United States (e.g. Atlantic Coast, Pacific Coast, and the Gulf of Mexico), but broadly around the
world (e.g. Turkey, Israel, Newfoundland, Australia, India, and Sicily) (Özyurt and Ergin 2010;
Lichter and Felsenstein 2012; Westley et al 2011; Abuodha and Woodroffe 2010; Nageswara
Rao et al 2008; Anfuso and Del Pozo 2009).
While Gornitz et al (1994)’s and Thieler and Hammar-Klose (1999)’s CVI allow for
large variations of environmental risk variables, some recent studies believe that weighted
variables create a stronger study. Nageswara Rao et al (2008) created a weighted CVI on the
Andhra Pradesh coast of India. Instead of the original CVI’s un-weighted variables, the authors
placed higher weights on the coast’s largest stressors (geomorphology, slope, and shoreline
change) in the threat of sea level rise. This allowed for a higher variation of variables, but
enabled the most important variables (geomorphology, slope, and shoreline change) to have a
stronger influence on the study. I chose not to use a weighted equation due to inconsistent levels
of data resolution.
2.3. Tools Adapted for Archaeology and Cultural Resource Management
In the past several years, CVI has also been adapted for use in studying the effect of
coastal vulnerability to archaeology sites. Reeder, Rick, and Erlandson (2012) studied on the
vulnerability of coastal archaeological sites for the northern Santa Barbara Channel region of
California, including the northern Channel Islands and the mainland coast ranging from Santa
Barbara to Point Conception, by extending CVI to a Cultural Resource Vulnerability Index
(CRVI). Based on Thieler and Hammar-Klose (1999)’s CVI (the square root of the product

24
mean), Reeder, Rick, and Erlandson (2012) further used GIS to measure an archaeological site’s
risk to coastal vulnerability by a weighted average approach:
𝐶𝑅𝑉𝐼=
4𝑑+3𝑣+2𝑢
3

where d is the distance of the archaeological site to shoreline, v is the CVI of the shoreline the
site is the closest to, and u is the threat of urban development. Similar to CVI, these variables
were firstly classified, with a numeric value from one (least vulnerable) to five (most vulnerable)
before inserting to the equation for the final CRVI (Reeder, Rick, and Erlandson 2012). The final
product of this study can be seen in Figure 6 below.

Figure 6 - Archaeological sites ranked according to CRVI (Reeder, Rick, and Erlandson 2012)

25
Chapter 3 Methods
This chapter describes data sources, data preparation, and analysis approaches used in this study.
This chapter starts with the data source and preparation (Section 3.1). The section thereafter
discusses the calculation of the Coastal Vulnerability Index (CVI) (Section 3.2) and the
subsequent Cultural Resources Vulnerability Index (CRVI) (Section 3.3).
3.1. Data Acquisition and Preparation
The data for this study was compiled from various sources, both within organizations’
ArcGIS Online databases and own hosted websites. The list of data information and their sources
can be found in Table 3.
Table 3 - Data Sources
Source Data: Year:
Data
Type:
Data
Available:
Additional
Information:
Source:
Cultural
Resource
Datums
N/A Points San Clemente
Island
Gathered and
corrected
using
digitized
reports and
site maps
Navy
NAVFAC
SNI_
Archaeological
_Sites_2002
2002 Polygon San Nicolas
Island
Navy
NAVFAC
USGS NED 2013 Raster Southern
Channel
Islands and
Small Portion
of Mainland
Coast
1/3 arc
second
(~10 meter)
Imported as
three DEM
rasters.
USGS
(https://viewer.
nationalmap.g
ov/basic/#prod
uctSearch)
Geologic Map
of California
2010 Polygon California Source media
from
hardcopy
paper.
1:750,000
scale.
Department of
Conservation
(https://maps.c
onservation.ca.
gov/cgs/gmc/)

26
Vegetation –
San Nicolas
Island [ds1501]
2013 Polygon San Nicolas
Island
Created using
National
Vegetation
Classification
Standards
(NVCS) for
Naval Base
Ventura
County
(NBVC)
HDR, Inc.
(https://tiles.ar
cgis.com/tiles/
Uq9r85Potqm
3MfRV/arcgis/
rest/services/bi
osds1501_cpu/
MapServer)
SCI_
Vegetation
Map20180430
2018 Polygon San Clemente
Island
Created for
Navy use.
Created from
4-band,
digital
multispectral
image dataset
with spatial
resolution of
0.15m.
Tested
accuracy of
77% using
fuzzy
methods.
Department of
Geography,
San Diego
State
University
(https://service
s1.arcgis.com/
SIYkiqjmENw
eC50g/arcgis/r
est/services/dr
aftMap201806
18/FeatureSer
ver)
Medium
Resolution
Shoreline
1994 Vector
Line
Contiguous
United States
1:80,000 NOAA
(https://idpgis.
ncep.noaa.gov/
arcgis/rest/serv
ices/NOAA/N
OAA_Medium
_Resolution_S
horeline/MapS
erver)

Through careful research of the coastal assessment methods descried in Chapter 2, the
methods used in this chapter were mainly adapted from Reeder-Myers (2015) and Aboudha and
Woodroffe (2010). The variables investigated in this study included vegetation, generalized rock
type, slope, and inundation levels. Due to the government-protected nature of these islands, very
little high-resolution data was available for use through traditional online public GIS sites.

27
Therefore, data such as geomorphology, sea level change, mean tide range, shoreline
erosion/accretion, and wave height are not used.
The overall workflow of data preparations is presented in Figure 7. Specific processes of
data preparation for individual variables are detailed in the following subsections. Unless noted,
the preparation and analysis were conducted in Esri ArcGIS Pro 2.1. All data was projected into
the Universal Transverse Mercator (UTM) Projection Zone 11N with a North American Datum
of 1983 (NAD 83) prior to analysis.

Figure 7 - The workflow of spatial data preparation for the calculation of CVI
3.1.1. Archaeological Site Data
Archaeological site data was provided by the NAVFAC team. The spatial data for the
archaeological sites for the two islands were collected for different archaeological research

28
purposes, which resulted in different data types (Rich Bark, personal communication). The SCI
archaeological site data was a point shapefile collected from several previous studies. The site
data on SNI, on the other hand, was surveyed as polygons documenting the boundaries of the
sites. The data of the archaeological sites for both islands was ready to use as-is with the
exceptions of projection transformation (to UTM zone 11N) and reclassification (as described in
the CRVI section below).
3.1.2. Elevation
The elevation data comprised of 1/3 arc-second Digital Elevation Model (DEM) rasters
downloaded from the U. S. Geological Survey website. Three raster grids covered the entire
extent of the study area was mosaicked together using the “Mosaic to New Raster” tool in
ArcGIS Pro. An island boundary shapefile was used to only focus on the two islands. This was
done by creating a 200-meter buffer around the island shapefile and merging the buffer to the
polygon. Then, the DEM was clipped to this buffered polygon to account for any possible land
area not accounted for in the coastline polygon. This elevation gridded data was then used to
create slope and inundation zone variable data, described in the subsections below.
3.1.3. Slope
A slope raster was created by the DEM created in the last section for the study area using
the Slope tool (Figure 8). The same resolution as the DEM layer (10 m) was used for the slope
raster. In order to create a CVI, the slope raster was reclassified into five categories ranked from
one (most vulnerable) to five (least vulnerable) in terms of coastal vulnerability: 1 - Cliffed slope
(> 45 degrees), 2 - steep slopes (20.1 – 45 degrees), 3 - moderate slopes (10.1 – 20.0 degrees), 4 -
gentle slopes (6.1 – 10 degrees) and 5 - low plains (0 – 6.0 degrees).

29

Figure 8 - Slope in degrees for San Nicolas Island and San Clemente Island
3.1.4. Inundation
Inundation zone layer was the second dataset produced from the DEM. Inundation
indicates the area extent potentially underwater with various levels of sea level rise projected by
NOAA (2010) and IPCC (2013). In order to create inundation, the DEM raster was reclassified
into five categories using the “Reclassify” tool in ArcGIS Pro. The classification of the
inundation zones layer took into account both the sea level rise projection from the two
trustworthy sources (NOAA and IPCC) and the accuracy of the source data (USGS 10-meter
DEM). The classification of the inundation zones can be seen in Table 4.

30
Table 4 - Reclassification of Inundation Levels
DEM Range
(m)
Classification Description Vulnerability
Ranking
0* - 0.3
Any land area less than 0.3 m; the value was rounded
from both NOAA projection (0.272 m) and the
lowest end of IPCC’s aggressive emission reduction
projection (0.28m)
Very High (5)
0.3 – 0.6
From IPCC’s aggressive emission reduction
projection range (0.28 to 0.61m) and also
encompasses IPCC’s best estimate (0.44m) and
NOAA’s best estimate (0.272m)
High (4)
0.6 – 1.0
From the high end of IPCC’s estimates (0.52 to
0.98m rounded up to 0.6 – 1.0m)
Moderate (3)
1.0 – 2.0
From highest IPCC estimate (0.98 rounded to 1m) to
highest NOAA projection (2m)
Low (2)
> 2.0 Above highest NOAA estimate (2m) Very Low (1)
*Including any DEM reported lower than zero (0)

Any land area less than 0.3 m in elevation, rounded from both NOAA’s best-scenario
projection (0.272 m) and the lowest end of IPCC’s aggressive emission reduction projection
(0.28m), was most likely to be inundated; the area was reclassified into the value of five (5) as
“most vulnerable”. Areas with elevation equal or higher than 0.3 m but lower than 0.6 m were
reclassified to the value of four (4) as moderately high vulnerable; the range of values were
rounded up from IPCC AR5’s aggressive emission reduction projection range (0.28 to 0.61m).
Based on the high end of IPCC’s estimates (0.52 to 0.98m), the inundation zone for the areas
with elevation between 0.6 m to 1.0 m was reclassified to the value of three (3). Similarly, the
“Less Vulnerable” area of inundation was reclassified to the value of (2), with the range from 1.0
m (rounded up from the highest IPCC estimate (0.98 m) to 2.0 m (the highest NOAA projection).
Any areas with elevation equal to or greater than 2.0 m were reclassified as the value of one (1).

31
3.1.5. Generalized Rock Types
Generalized rock types show areas of rock difference on the two islands. Different rock
types erode at unique rates and this data shows the areas where the rock is most susceptible to
erosion. This dataset gives insight to areas that are more vulnerable than others across the two
islands. The generalized rock type data used in this study comes from the 1977 California
Geological Survey conducted by the California Department of Conservation and digitally
updated in 2010. The dataset was digitized from hardcopy maps with an inherent potential error
in the creation of the map for legibility at the 1:750,000 scale.
The data file for this variable included all of California and was clipped to include only
SCI and SNI. A new layer was created from this clip. Once this was done, the vector data was
transformed into a raster for use in the CVI equation using the Polygon to Raster tool in ArcGIS
Pro. Once the raster was created, the generalized rock type variable was reclassified much like
the inundation layer. Table 5 below shows the created rankings.
Table 5 - Generalized Rock Type reclassification rankings
Rock Type Abbreviation
Gornitz et al
(1994)
Relation
Vulnerability
Ranking
Tertiary volcanic rocks

Tv Plutonic,
volcanic, High-
medium grade
metamorphics
Very Low (1)
Eocene marine sedimentary rocks

E Low grade
metamorphics,
Sandstone and
Conglomerates
Low (2)
Miocene marine sedimentary rocks

M Most
sedimentary
rocks
Moderate (3)
Pleistocene marine and non-marine
sedimentary rocks

Qoa Coarse, poorly
sorted
unconsolidated
sediments
High (4)

32
Pleistocene-Holocene marine and non-
marine sedimentary rocks

Q Fine
unconsolidated
sediments,
volcanic ash
Very High (5)
NODATA NODATA

3.1.6. Vegetation
Vegetation data for these islands can only be found for each individual island. Vegetation
for SNI was created for regulatory purposes on Naval Base Ventura County (NBVC) San
Nicolas Island by the company HDR while the vegetation data for SCI was created by a team at
San Diego State University for Navy use which followed California Native Plant Society
protocol (Uyeda et al 2018). The vegetation data used in this study was a strong addition because
the data was created by two respected institutions hired by the Navy to be conducted for each
island. The two vegetation datasets used allowed for a significantly better understanding of how
each island differed and how erosion is naturally prevented. I would feel confident using this
data in future research.
Both datasets give the proper Latin names of each vegetation type, and while there are
slightly different vegetation types for each island, the general types of vegetation are
comparable. The Latin names of these plants and the common names can be seen below in Table
6 for SCI and Table 7 for SNI as well as the generalized vegetation type and the reclassification
CVI vulnerability ranking.
Table 6 - Vegetation ranking for San Clemente Island
Metadata Name Common Name Vegetation Type
Vulnerability
Ranking
Ambrosia
chamissonis –
Ambronia maritima
Silver burr ragweed,
Red sand verbena
Scrub, Shrub Very Low (1)
Artemisia California Shrub Very Low (1)

33
californica sagebrush
Baccharis pilularis Coyote brush Shrub Very Low (1)
Bare - Bare Very High (5)
California Cliff,
Scree, Rock
Vegetation Group
- Bare Very High (5)
Coastal Unvegetated - Bare Very High (5)
Cylindropuntia
prolifera
Association
Coastal Cholla Cactus Moderate (3)
Developed - Developed NODATA
Distichlis spicata Desert saltgrass Grass Moderate (3)
Grassland semi-
natural stands
- Grass Moderate (3)
Iceplant - Iceplant Moderate (3)
Lycium californicum California boxthorn Shrub Very Low (1)
Lyonothamnus
floribundus
(individual)
Catalina ironwood Tree Very Low (1)
Optunia littoralis Coastal prickly pear Cactus Moderate (3)
Prunis ilicifolia ssp.
lyonii
Catalina cherry Tree Very Low (1)
Quercus tomentella
(individual)
Channel Islands oak Tree Very Low (1)
Rhus integrifolia
(individual)
Lemonade berry Shrub Very Low (1)
Salicornia pacifica
(alliance)
Pickleweed Succulent Moderate (3)
Salsola sp.
provisional
Saltwort Succulent Moderate (3)

Table 7 - Vegetation ranking for San Nicolas Island
Metadata Name Common Name Vegetation Type Ranking Number
Ambrosia
chamissonis –
Ambronia maritima
– Cakile maritima
Silver burr ragweed,
Red sand verbena,
European searocket
Scrub Very Low (1)
Ammophila arenaria Marram grass Grass Moderate (3)
Artemisia nesiotica Island sagebrush Shrub Very Low (1)
Baccharis
pilularis/Annual
grass-herb
Coyote brush Shrub Very Low (1)
Barren Barren Very High (5)

34
Cakile maritima –
Ambrosia
chamissonis –
Carpobrotus edulis
European searocket,
Silver burr ragweed,
Iceplant
Succulent Moderate (3)
Carpobrotus edulis
(or other iceplant)
Iceplant Iceplant Moderate (3)
Coreopsis gigantea Giant coreopsis
(Daisy)
Bush Very Low (1)
Deinandra clementia Island tarplant Bush Very Low (1)
Distichlis spicata Desert saltgrass Grass Moderate (3)
Frankenia salina Alkali heath Shrub Very Low (1)
Frankenia
salina/Distichlis
spicata
Alkali heath, Desert
saltgrass
Shrub Very Low (1)
Isocoma menziesii Coastal goldenbush Shrub Very Low (1)
Lupinus albifrons Silver lupine Shrub Very Low (1)
Marine - Marine Very High (5)
Mediterranean
California
Naturalized Annual
& Perennial
Grassland
- Grass Moderate (3)
Optunia littoralis –
Mixed coastal sage
scrub
Coastal prickly pear Cactus Moderate (3)
Saxlis lasiolepis Arroyo willow Tree Very Low (1)
Urban - Urban NODATA

These vegetation layers were created as vector layers. Since this study needs raster data
for the CVI equation, the data was first transformed into rasters by using the Polygon to Raster
tool in ArcGIS Pro. Once the vegetation layers for both islands were transformed into rasters, the
vegetation types were reclassified into the three rankings shown in Tables 6 and 7 and the two
islands were join together for use in the CVI equation described in Section 3.2.
3.1.7. Coastline
While coastline data was not a part of the CVI equation, it was still an important part of
this study due to its use in creating a clipped DEM. This coastline data was created by the USGS

35
to assess shoreline change and is a fairly good resolution for encompassing all of the contiguous
United States.
3.2. Coastal Vulnerability Index (CVI)
In this study, Thieler and Hammar-Klose (2000)’s equation was used due to its
minimizing effect on extreme variables that might be outliers:
𝐶𝑉𝐼=
𝑎 ∗𝑏 ∗𝑐 ∗𝑑
4

In this equation, a is Slope in Degrees, b is Inundation in Meters, c is Generalized Rock Type,
and d is Vegetation Type. All data used in these equations are discussed in the previous sections
and have been edited and reclassified for use within these equations. A summary of these
classifications can be seen in Table 8 below. Using the equation mentioned above, the “Raster
Calculator” tool was used to input all variables as designated. The result of this calculation
allows for a better understanding of vulnerability around the islands, and will be used for the
CRVI analysis in the following section.
Table 8 - Coastal Vulnerability Index ranking
I.D. Variable
Very Low
(1)
Low (2)
Moderate
(3)
High (4) Very High (5)
a
Slope

(in Degrees)

Cliffed
Slopes

(> 45)
Steep Slopes

(20.1 – 45)
Moderate
Slopes
(10.1 – 20.0)
Gentle
Slopes

(6.1 – 10.0)
Low Plains

(0.0 – 6.0)
b
Inundation
(in Meters)
> 2.0

2.0 – 1.0

1.0 – 0.6

0.6 – 0.3

< 0.3

c
Generalized
Rock Type

Tertiary
volcanic
rocks
Eocene
marine
sedimentary
Miocene
marine
sedimentary
Pleistocene
marine and
non-marine
Pleistocene-
Holocene
marine and

36
(Tv)

rocks
(E)

rocks
(M)

sedimentary
rocks
(Qoa)

non-marine
sedimentary
rocks
(Q)

d
Vegetation
Coverage

Ground
Cover,
Shrubs,
Trees
- Grasses,
Cactus, Ice
plant
- Barren, Bare,
Unvegetated,
Marine

3.3. Cultural Resource Vulnerability Index (CRVI)
The final analysis for this study was to create the Cultural Resource Vulnerability Index
(CRVI) to recognize the most at-risk archaeological sites based on CVI. The steps for this task
are outlined below in Figure 9. This was done by creating a cell index around the coastline of
the islands that include the index numbers of the cells calculated from the CVI.

Figure 9 - CRVI workflow
Variables used in CVI had inconsistent coastline polygon boundaries due to the source
data. A new coastline was created from the CVI to ensure uniformity. CVI was converted from a
floating data type into an integer data type raster using the Raster Calculator tool to remove

37
decimals in the CVI by multiplying the raster by a multiple of 10. For the CVI in this study, the
raster needed to only be multiplied by 10 to convert to an integer. Once this was completed, I
used the “Int” tool to convert the CVI raster into a new Integer Raster. This was done to create
the new coastline for use in the CRVI for distance and for a coastline-specific CVI.
3.3.1. Coastline CVI
An index grid along the coastline was necessary for use in the CRVI equation.
Using the newly established coastline, 10-meter squares were placed along the coast. The
coastline polygon squares created allowed for the Zonal Statistic tool to be used. The Zonal
Statistic tool was used to attribute the CVI value to the newly created squares along the coastline.
The output from this tool was used in the Join, which ultimately creates the CRVI.
3.3.2. Archaeological Site Distance From Coast
Before using the Add Join function to create the CRVI, the distance of the archaeological
sites to the coastline was calculated by using the Near tool. The resulting distance was seen in the
archaeological sites’ attribute table and was used to ascribe a vulnerability ranking for use in the
CRVI equation.
For use in the CRVI, distances needed to be reclassified into ranks in order for the CRVI
equation to be successful. This was done by manually reclassifying the distance ranges into five
ranks in the archaeological sites’ attribute tables. The distance ranks can be seen below in Table
9.
Table 9 - Distance from coastline ranking for archaeological sites.
Distance to Coastline
Vulnerability
Rank
< 50 meters Very High (5)
50 – 100 meters High (4)

38
100 – 500 meters Moderate (3)
500 – 1000 meters Low (2)
> 1000 meters Very Low (1)

Once these rankings were completed, the Add Join tool was used to join the data in the
archaeological sites’ attribute tables to the CVI coastline data. Once this tool was run, all data
necessary for the CRVI equation was located in the archaeological sites’ attribute table and a
new layer was created showing the CVI along the coast.
3.3.3. CRVI Equation
Finally, after completing all steps mentioned above, the CRVI equation was used to
calculate the vulnerability of each archaeological site. This was done by opening the newly
produced CVI coastline layer (created from the Add Join step above) and creating a new attribute
table field. This new field was created to use the Field Calculator tool to calculate the CRVI. In
this study’s case, the equation was:
𝐶𝑅𝑉𝐼=
𝑎 ∗𝑏
2

where a was the newly reclassified distance variable and b was the CVI ranking created from the
Join step mentioned in the previous section. All new numeric results from this CRVI equation
could be seen in the CRVI field in the islands’ attribute tables. This new data was visualized and
the result of this visualization can be seen in the next chapter.

39
Chapter 4 Results and Discussion
The analysis results of the coastal vulnerability and the at-risk archaeological sites for
San Nicolas Island and San Clemente Island are reported in three parts in this chapter. First, the
vulnerability ranking results for the relevant variables are presented in Section 4.1, followed by
the results of the CVI and the CRVI in Section 4.2 and Section 4.3, respectively. Discussions
about the results and their implications are also included at the end of individual sections.
4.1. Vulnerability Rankings for Environmental Variables
This section presents the classification results for the four variables that affect coastal
vulnerability. Upon examining the vulnerability rankings for slope, inundation, generalized rock
type, and vegetation, it can be seen that a large percentage of area (44.08%) had a high
vulnerability ranking (or 5) for slope (Figure 10). In contrast, inundation, vegetation, and rock
type had the high percentage of areas with very low vulnerability rankings (1). The percentage of
area for all rankings of individual variables (1-5) are described in each subsection below.

40

Figure 10 - The area percentage of vulnerability ranking for the CVI variables in the study area
4.1.1. Slope
Figure 11 below shows the slope vulnerability rating for San Nicolas Island and San
Clemente Island. More than half of the island areas were relatively flat, either at low plains (0 –
6°) (44.08%) or with a gentle slope (6-10°) degrees slope (17.93%). The areas with gentle slopes
along the coastline were considered vulnerable due to the ease of water inundation. On San
Nicolas Island, the low elevation areas near the coastline with gentle slopes were at northern and
northwestern parts of the island. On San Clemente Island, however, these plains areas are located
on the northern and northwestern regions.
0 (Urban) 1 2 3 4 5
Slope 2.01 17.4 18.58 17.93 44.08
Inundation 86.29 0.46 0.24 0.03 12.98
Rock Type 69.26 14.38 2.14 6.6 7.62
Vegetation 2.27 44.36 0 50.21 0 3.16
0
10
20
30
40
50
60
70
80
90
100
Percent of Area
Percent of Island Area Per Variable
Vulnerability Ranking

41

Figure 11 - The classification of slope from least vulnerable (1) to sea level rise to most
vulnerable (5).
While one could argue that gentle slopes are less vulnerable in regards to erosion, it is
important to state that this study focuses on the coastal risk created from sea level rise. Because
of the bedrock formation on cliffed slope (greater than 45°), the areas with cliffed slopes are
more resistant to erosion as well as permanent water inundation from sea level rise. These cliffed
slopes are less susceptible to erosion than gentle slopes (less than 10 degrees) (Westley et al.
2011 and Aboudha and Woodroffe 2010). With very little elevation change, flooding from storm
surges and inundation from sea level rise are a threat (Gesch, Gutierrez, and Gill 2010 and
Reeder-Myers 2015). Gentle slopes along the coast are also typically indicators of sandy
beaches, which are vulnerable due to frequent erosion and shoreline retreat (Church et al. 2008).

42
Steep slope (20.1 – 45° slope) and cliffed slope (>45° slope) areas, ranked low and very
low in coastal vulnerability, respectively, were seen primarily on the northeastern shoreline and
some narrow ridges in the center of San Clemente Island. The same vulnerability settings (on
steep slope or cliffs) were seen inland around the entire San Nicolas Island. The cliffed slope was
only seen on 2.01% of the two islands, but is indicative of immediate change between the
coastline and inland. The steep slopes, on the other hand, had the highest area percentage of the
overall rankings, amassed 17.40% of the entire study area. Moderate slopes (10 – 20°),
comprising 18.58% of the islands, were found near low to very low vulnerability ranking areas,
as expected.
4.1.2. Inundation
Inundation zones, as expected, were in general consistent around the shorelines of the
two islands. It is, however, important to note that a buffer was created around the islands to
ensure all low-lying land areas were accounted for, as no consistency of shoreline boundaries
were found from source data as the baseline extent. Therefore, some portions of the low-lying
areas might include the oceans; there were about 13.0% of the area ranked most vulnerable in the
inundation classification (< 0.3 meters above sea level) around all of the two islands. Perhaps the
better indicator of vulnerability in coastal inundation is the classification of areas at 0.3 – 0.6 m
above mean sea level (‘High’ vulnerability with a ranking score of ‘4’). However, these high
vulnerability areas of inundation were seen primarily along San Clemente Island’s northern coast
and consisted of less than 1% (0.03%) of the islands’ total area. Both moderate vulnerability
areas of inundation (0.6 – 1 m above mean sea level) and low vulnerable areas of inundation (1.0
– 2m above mean sea level) also comprised less than 1% (0.24% and 0.46%, respectively) of the
two islands. The moderate vulnerability area of inundation were along the northwestern portion

43
of San Nicolas Island and the southern and northern points of San Clemente Island, and the low
vulnerability area of inundation were primarily seen on the northern tip of San Clemente Island
and southeastern quadrant of San Nicolas Island.
Most areas (86.3%) of the two islands, in the use of our vulnerability classification based
on the sea level rise concern, were higher than 2 meters above mean sea level on San Nicolas and
San Clemente Islands. The maps for the vulnerability classification of inundation can be seen in
Figure 12 below.

Figure 12 - The classification of water inundation zones based on the predicted sea level rise,
ranked from most vulnerable (5, in red) to least vulnerable (1, in blue).

44
With NOAA and IPCC sea level rise projections estimating a 2-meter sea level increase
by 2100, approximately 13.71% of land area on San Nicolas and San Clemente Islands were
susceptible to inundation (Lindsey 2017). Inundation can be devastating to archaeological sites.
Inundation also exacerbates shoreline retreat. In fact, 70% of the world’s shorelines have been
experiencing retreating due to erosion caused by inundation (Maio et al. 2012 and Church et al.
2008).
4.1.3. Generalized Rock Type
Figure 13 shows the vulnerability classification of the generalized rock types. A total of
five generalized rock types were seen on the two islands. The most vulnerable rock type, “Q”,
covered 7.6% of total area across the two islands, and the second most vulnerable rock type,
“Qoa”, covers 6.6% of total land area across the two islands.
Over the entire study area, 14.2% of the areas consisted of very highly and highly
vulnerable rock types (“Q” and “Qoa”, respectively). These two rock types were only seen on
San Nicolas Island. The most vulnerable rock type “Q” was located along the northwestern
portion of the island while the “Qoa” rock type was seen directly south and east of “Q.” The
moderate rock type (with a ranking score of 3), “M”, covered only 2.14% of the total islands’
land mass and was located in small portions across San Clemente Island. Finally, the second
lowest vulnerability (with a ranking score of 2) can be seen across 14.38% of the landmass. This
“E” rock type was located in many areas across San Nicolas Island, specifically along the
island’s southern quadrant and along portions of the island’s northern coastline.

45

Figure 13 - The vulnerability classification of generalized rock type.
4.1.4. Vegetation
Existing vegetation plays an important role to prevent the islands from both water and
wind erosion as it creates erosion-inhibiting root systems and wind barriers. This type of erosion
control, with shrubs, trees, and groundcover, were seen across San Nicolas Island and San
Clemente Island. Groundcover, shrubs, and trees covered 44.4% of the two islands (Figure 14).
Most of San Nicolas Island was in this very low vulnerability class, but only seen along the
northeastern and western coastal areas of San Clemente Island. Some of this vegetation class was
also found in the small canyons on San Clemente Island’s southern quadrant.

46

Figure 14 - The vulnerability classification of vegetation.
Grasses, cactus, and ice plants were prolific across the two islands. 50.2% of the islands
total area was in this moderately vulnerable classification of vegetation (Figure 14). Most of San
Clemente Island, especially inland, was covered by grasses, cactus and ice plants. The type of
vegetation was spread sporadically across the middle of San Nicolas Island and the island’s
northern point. Only 3.2% of the total study area was there a lack of vegetation (“barren,
unvegetated, and marine”) and ranked the most vulnerable due to its lack of vegetation erosion
control. Specifically, these areas were seen along most coastlines. 2.3% of the islands was
classified as urban development and was given a score of ‘0’ in this classification.

47
4.2. Coastal Vulnerability Index
The Coastal Vulnerability Index (CVI) combined all classified variables, to rank areas
along the coastline from least vulnerable to most vulnerable. In the following sections, the results
of CVI for San Nicolas and San Clemente Island are described in detail.
4.2.1. CVI for Entire Study Area
The CVI for the entire island of San Clemente Island is presented in Figure 15. The most
vulnerable areas, as expected, were located along the coastline as thin strips. Overall, San
Clemente Island was less vulnerable with the exception of four very highly vulnerable zones on
the southern coast and on the mid portions on the northern half of the island. These vulnerable
zones are displayed in red in Figure 15. The areas were categorized as “very high” in coastal
vulnerability due to the weaker of the two rock types on the island (Miocene sedimentary rocks,
“M”). While the rock type “M” was classified as moderately vulnerable to erosion, it can be seen
that the areas that are within the boundary of this rock type are significantly more vulnerable on
San Clemente Island. Apart from these four areas, the coastline of San Clemente Island was the
most vulnerable portion (“Very High”) of the island – specifically the northeastern portion of the
coastline and the northern most tip of the island. Recall that Figure 14 and Figure 12 showed
some areas of the island with no vegetation and low elevation; these areas were susceptible to sea
level rise impacts. The most vulnerable areas of the San Clemente Island were the southern-most
coast where the rock type was the weakest and the elevations was close to sea level and would
allow excessive water inundation.

48

Figure 15 - The Coastal Vulnerability Index for the entire San Clemente Island
Fortunately, most of the San Clemente Island was ranked low in coastal vulnerability.
This was directly related to two risk variables: rock type and slope. Because San Clemente Island

49
was primarily covered by the “Tv” (Tertiary volcanic) rock type, most of the island was fairly
resistant to erosion. Moreover, since most of the island’s northern side had steep slopes, the
vulnerability was ranked low, and the potential damage caused by inundation in the next 100
years would be low.
On the other hand, many of the areas in San Nicolas Island contained the most vulnerable
ratings of various variables and resulted in more vulnerablility (Figure 16). Specifically, the
island is most vulnerable due to its flat slope, “Q” (Pleistocene-Holocene marine and nonmarine
sedimentary rocks) and “Qoa” (Pleistocene marine and nonmarine sedimentary rocks) rock
types, and larger portions of barren areas without vegetation. The rock types are highly
susceptible to erosion due to its unconsolidated to semi-consolidated states. The slope on San
Nicolas Island was also gentler allowing for more water inundation along the coastlines. This
was seen across most of the island, specifically across the entire northern tip of the island. In
low-plain areas that were near the sea level, the barren landscape is commonly seen across the
island; these areas would be highly vulnerable to inundation.

50

Figure 16 - The Coastal Vulnerability Index for the entire San Nicolas Island
Aside from the urban areas that were rated most vulnerable (in dark green in Figure 16),
the second to most vulnerable areas (in light green in Figure 16) were across the islands. These
areas consisted of the rock type “E” (Eocene marine sedimentary rocks) and had steep slopes.
The combination of these two variables resulted in the areas less vulnerable to sea level rise and
other coastal threats.
4.2.2. Coastline CVI
Based on the results in Section 4.2.1, it is clear that coastlines are not the only vulnerable
areas for inundation and erosion on these two islands. However, CVI near the coastline was
further inspected for the archaeological sites vulnerability. The coastline CVI layer was created

51
for San Clemente Island and San Nicolas Island. Figure 17 shows coastal areas in the northwest
quadrant of San Nicolas Island are the most vulnerable compared to other parts of the coastline
on the same island. For San Clemente Island, the coastal CVI indicated the most vulnerable
coastal areas were on the southern coast and farthest northern point of the island (Figure 18).

Figure 17 - Coastline CVI for San Nicolas Island

52

Figure 18 - Coastline CVI for San Clemente Island
Only 0.3% of the coastline on the two islands was considered most vulnerable. These
areas were scattered throughout the islands, but were primarily seen on San Nicolas Island’s
southwestern coast. The second to most vulnerable class of CVI consisted of 1.7% of coastline

53
grids on the two islands. Next, moderate vulnerability was found to encompass a higher
percentage (9.4%) of coastal areas, primarily along the northern point of San Clemente Island
and scattered along San Nicolas Island’s northwestern side. The low and very low vulnerability
areas covered 33.2% and 55.3% coastal area, respectively, mainly located on the southwestern
side of San Clemente Island and the northeastern side of San Nicolas Island.
The purpose of generating the coastline CVI layer was to calculate the CRVI, but areas of
“No Data” exist due to missing raster points in the original variable layers. Upon the coastline
CVI being conducted, a small amount of “No Data” area could be seen. A fix was conducted by
using the Zonal Statistic tool to fill any areas of no data with the mean of the points and merging
the Zonal Fill with priority set on Zonal Fill. Since the fix, very few errors occurred.
4.3. Cultural Resource Vulnerability Index (CRVI)
Based on the nearest coastal cell’s CVI and its distance to the archaeological site, CRVI
ranked the coastal vulnerability for archaeological sites. The CRVI for each island is presented in
the following sections.
4.3.1. San Nicolas Island
As seen in Table 10, most archaeological sites (75.7%) were more than 100 meters away
from the coastline. These sites (ranked 1-3) were less at risk due to their distance from the ocean
and higher elevation (so less susceptible to water inundation by sea level rise). There were 31
(5.79%) of these sites within 50 – 100 meters from the coast (a ranking of 4) and 99 (18.51%)
archaeological sites were less than 50 meters from the shore (a ranking of 5). These sites would
be most at risk due to their close proximity to rising sea levels and the effects related to this sea
level change. However, upon calculating the coastline CVI for these sites, it can be seen that 5
archaeological sites (0.93%) are close to 7.5 – 10.0 CVI rankings (“4”) and only 4 sites (0.75%)

54
are close to the most at risk CVI (“5”). Moderately vulnerable CVI rankings (“3”) make up
7.29% of archaeological sites’ coastline CVI. The second least vulnerable CVI ranking value
(“2”) can be seen with 263 sites (49.16%) and the least vulnerable ranking value (“1”) can be
seen with 224 sites (41.87%).
Table 10 - Ranking of CRVI variables on San Nicolas Island where n is the number of
archaeological sites found for each ranking.

Cultural Resources Vulnerability – San Nicolas Island
Very Low
(1)
Low (2) Moderate (3) High (4)
Very
High (5)
No. of Samples
in Distance from
Coast
(> 1000 m) (500 - 1000 m) (100 -500 m) (50 - 100 m)
(0 - 50
m)
191 81 133 31 99
No. of Samples
in Coastal CVI*
(0 - 2.6) (2.6 - 5.0) (5.0 - 7.5) (7.5 - 10.0) (> 10.0)
224 263 39 5 4
No. of Samples in
CRVI**
(0 - 4.0) (4.0 - 8.2) (8.2 - 12.2) (12.2 - 16.3) (> 16.3)
344
(64.3%)
126
(23.6%)
46
(8.6%)
16
(3.0%)
3
(0.5%)
*Coastal Vulnerability Index

**Cultural Resources Vulnerability Index

The final map of CRVI for San Nicolas Island can be seen in Figure 19. The CRVI
ranking for San Nicolas Island shows 3 archaeological sites (0.56%) that are most vulnerable
(“5” ranking). These sites are small in area and are located on the northeastern coastline and the
southwestern coastline of the island. More archaeological sites (2.99%) are ranked highly (“4”)
by the CRVI. These 16 sites are fairly scattered across the island’s coastline, but can be seen on
the southwestern coast, northern coast, and southeastern coast. Specifically, these sites are found
in areas of very low slope, slope that was ranked “5” and “4” for the CVI. These areas also
include barren vegetation. Moderately ranked archaeological sites (“3”) can be seen on the
Classification
Parameters

55
northern coast, the southeastern coast, and the northwestern coast and make up 8.60% of the
island’s sites. The least vulnerable archaeological sites in regard to sea level rise make up
23.55% for “2” ranking and 64.30% for “1” ranking. These sites are primarily seen on the inland
of the island.

Figure 19 - CRVI for San Nicolas Island with archaeological site rankings.
4.3.2. San Clemente Island
San Clemente Island is a unique study due to the sheer volume of archaeological sites
across the island. Similar to San Nicolas Island, most archaeological sites (84.80%) are at least
100 meters away from the shoreline, but 769 sites (15.20% of the total sites) still remain within
100 meters from the coast. These sites could still be at risk for its close proximity to the coast,

56
but only 4 sites were within the “high” to “very high” vulnerability in terms of their CVI on-site.
Moderately vulnerable ranking of the CVI (5.0 – 7.5) was dominant of all archaeological sites
(7.89% of the total sites). The second least vulnerable CVI (a ranking score of 2) was seen in
23.1% of the sites across the island. A majority (64.6%) of the archaeological sites has the least
vulnerable CVI ranking (1). A small percentage (4.3%) of the sites received a “0” for its CVI
ranking. This error issue is discussed toward the end of this chapter.
While using CVI, the results showed the majority of the archaeological sites were rated
low to very low vulnerability, the CRVI ranked a larger number of “high” to “very high”
vulnerability for archaeological sites on San Clemente Island, because of the calculation used the
distance and correlating coastline CVI (Table 11). A total of 408 archaeological sites (8%) were
ranked most vulnerable (“very high”, with a ranking of 5). These sites can be seen in red in
Figure 20. Primarily, these most vulnerable sites were located along San Clemente Island’s
northwestern coastline with some scattered along the southwestern and northern tips.
Table 11 - Ranking of CRVI variables on San Clemente Island where n is the number of
archaeological sites found for each ranking.

Cultural Resources Vulnerability – San Clemente Island
Zeros Very Low (1) Low (2) Moderate (3) High (4)
Very High
(5)
No. of Samples
in Distance
from Coast
N/A (> 1000 m) (500 - 1000 m) (100 -500 m) (50 - 100 m) (0 - 50 m)
N/A 1636 1196 1457 300 469
No. of Samples
in Coastal CVI*
= 0 (0 - 2.6) (2.6 - 5.0) (5.0 - 7.5) (7.5 - 10.0) (> 10.0)
216 3269 1170 399 4 0
No. of Samples
in CRVI**
= 0 (0 – 1.3) (1.3 – 2.6) (2.6 – 4.2) (4.2 – 7.2) (> 7.2)
216
(4.3%)
1641
(32.4%)
1427
(28.2%)
801
(15.8%)
565
(11.2%)
408
(8%)
*Coastal Vulnerability Index
**Cultural Resources Vulnerability Index

Classification
Parameters

57

Figure 20 - The CRVI for archaeological sites on San Clemente Island; higher values indicates
higher coastal vulnerability.

58
The second highest ranked CRVI, with a vulnerability ranking of “4”, consisted of 565
sites (11.2% of the total sites) throughout the island. These highly vulnerable sites appeared to be
near those sites that had “very high” vulnerability, and were mostly found on the island’s
northern-most tip. Overall, the “very high” (5) and “high”(4) rankings were in areas with a
combination of gentle slopes, inundation zones (low elevation), and less or lack of vegetation.
Moderately ranked (3) archaeological sites (15.8% of total sites) were located mostly along the
western coast and on the northern tip of the island. The second lowest vulnerability (2) are given
to 28.2% of the archaeological sites and were located inland. The least vulnerable (1)
archaeological sites, consisted of 32.4% of the total archeological sites, were located further
inland.
Collectively across the two islands, 411 archaeological sites were most at risk (“very
high” vulnerability) to coastal instability and inundation. With this in mind, archaeologists can
conduct research more focused on these very highly vulnerable sites before they are to lost to
coastal inundation and erosion. It is also important to keep in mind that 585 of these sites are also
“high” vulnerability
4.3% of archaeological sites on San Clemente Island received a value of zero (0) for the
coastline CVI (the CVI of the closest coastline gridded cell). At this point, it was still unclear
where this error was generated. Possibly, the Add Join function used in linking the
archaeological site data and the nearest coastal cell created areas of No Data when attaching
value to these archaeological sites. It would be possible to recalculate the coastal CVI values for
these sites, but a new method of calculating CRVI in ArcGIS would have to be used.

59
Chapter 5 Conclusions
As this study concludes, it is important to evaluate whether this study was successful and
in what ways it can be improved. This study has found success in understanding which
archaeological sites are most at risk across San Nicolas Island and San Clemente Island off the
coast of Southern California. Using the CVI and CRVI methods, 19 archaeological sites on San
Nicolas Island were found highly or very highly vulnerable due to coastal instability and
potential inundation. In contrast, 973 archaeological sites on San Clemente Island were identified
to be highly or very highly vulnerable.
The second achievement of this study is a better understanding of the spatial locations
that are most vulnerable to inundation and coastal erosion. By creating the spatial layer of CVI
using four environmental variables – slope, inundation, generalized rock type, and vegetation, a
total of 2.07% of the coastline was ranked highly or very highly vulnerable.
5.1. Limitations
This study faced its own set of challenges throughout. The most obvious limitation is data
unavailability to the study area, particularly due to its location and the protected status by the
U.S. Navy. Compared to the CVI variables used in Abuodha and Woodroffe (2010) and the
CRVI variables in Reeder, Rick, and Erlandson (2012), only four variables (slope, inundation,
generalized tock type and vegetation) were used in the study. Several other relevant
environmental variables such as geomorphology and shoreline change were not available for use
at the time of analysis. While the Navy ownership and stewardship of the two islands allows for
research, most of the data from these studies remain unpublished or unavailable for public use.
The data used in this study was secondary. Although the data came from seemingly
reliable sources, it is possible that the data was not accurate in some areas across the two islands.

60
Secondary data are sources collected from studies not conducted by the author. Because the data
is not created by the author, accuracy testing data can only be obtained from metadata when
available. In this study, it is assumed that the generalized rock type data is accurate due to its
creation by the California Department of Conservation. The source layers of vegetation on both
islands were created using California Native Plant Society protocol for the Navy.
Data fitness for use was also a concern during this study. The best available data for
geology was generalized rock type, but it was over generalized for the large spatial scale used in
this study, and very little variation between and within the two islands could be seen. Among the
variety of CVI equations developed, only one equation was adapted for use in this study. It
would be beneficial to test this CVI against the other equations.
Sea level rise data was only referenced globally. Because of the small areas of these
islands, there was very little sea level rise variation across the two islands. The regional sea level
rise projection was not available or could not be found at the time of this study. However, for
better accuracy, sea level rise estimates for the Channel Islands region should be reviewed in
future studies much like this.
Vegetation data was ranked into only three vulnerability classes. This type of ranking
could have caused a skewed weight for the CVI. A better understanding of the complexities of
vegetation erosion control would be beneficial to this type of study in order to rank the
vegetation types into the five classes instead of three.
Finally, there were several cells with the value of zero (0) created from the joining of
archaeological sites to the coastline CVI. These errors came from variables with incomplete
raster coverage where “No Data” was allocated to the raster cells. These 0s might have created
accuracy issues for the resulting archaeological site vulnerability.

61
5.2. Future Research
As missing CVI and CRVI variables limit the result of the study, future research should
consider incorporating these missing variables (e.g. barrier type, shoreline exposure, shoreline
change, mean wave height, and mean tide range) into CVI and CRVI calculations when they
become available and handy. Specifically, king tide and detailed geomorphology data would
need to be added to future renditions of this study. King tide that shows the highest point of the
tide cycle would inform where water will be seen in the future. A more detailed geologic data
that shows variations of rock types among the two islands will remedy the resolution issues. A
more detailed geologic data map was found for San Clemente Island via the USGS National
Geologic Map Database, created in 1958. This geologic data was not used because it did not
include San Nicolas Island. Future studies of this type should use the more detailed geologic
data.
The indices used for this study can be applied to various fields of studies related to sea
level rise and erosion vulnerability in coastal regions. These fields, such as real estate, ecology,
and endangered animal and plant protection, can adapt the Cultural Resource Vulnerability Index
to suit assessment needs. An example in the real estate field would be to understand which
residential communities and houses would be affected by sea level rise. A recent study projected
that 1.9 million homes would be underwater by the year 2100 if the sea level rises six feet (Bretz
2017); a similar index like CRVI used for archaeological sites can be applied to understand
which sites need to protect from the coming sea level rise.

62
5.3. Final Take Away
With the global sea level rise, coastal vulnerability might affect most areas of everyday
life in the next 100 years. This study took steps towards understanding the danger of the potential
sea level rise and coastal instability for a practical reason – protecting archaeological sites.
This study developed the CVI and CRVI layers for San Clemente Island and San Nicolas
Island and will provide the archaeologists in the U.S. Navy a guideline for prioritizing the study
of archaeological sites and mitigating the risks of losing the at-risk archaeological sites.
The CVI and CRVI conducted in this study allow us to quantify and visualize how the
world around us is changing every day. With sea levels continuing to rise, researchers and
scientists continue to work on understanding the threat and creating methods to adapt or mitigate
the inevitable environmental change. It is my hope that this study would be a small stepping-
stone for helping these coastal researchers in the right direction.

63
References
Abuodha, Pamela A. O. and Colin D. Woodroffe. 2010. “Assessing vulnerability to sea level rise
using a coastal sensitivity index: a case study from southeast Australia.” Journal of
Coastal Conservation 14 (3): 189-205. DOI: 10.1007/s11852-010-0097-0
Anderson, David G., Thaddeus G. Bissett, Stephen J. Yerka, Joshua J. Wells, Eric C. Kansa,
Sarah W. Kansa, Kelsey Noack Myers, R. Carl DeMuth, and Devin A. White. 2017.
“Sea-level rise and archaeological site destruction: An example from the southeastern
United States using DINAA (Digital Index of North American Archaeology).” PLoS One
12 (11): 1-25. https://doi.org/10.1371/journal.pone.0188142
Anfuso, Giorgio and José Ángel Martínez Del Pozo. 2009. “Assessment of Coastal Vulnerability
Through the Use of GIS Tools in South Sicily (Italy).” Environmental Management 43:
533-545. DOI: 0.1007/s00267-008-9238-8
Braje, Todd J. 2010. Modern Oceans, Ancient Sites: Archaeology and Marine Conservation on
San Miguel Island, California. Salt Lake City: The University of Utah Press.
Bretz, Lauren. 2017. “Climate Change and Homes: Who Would Lose the Most to a Rising
Tide?” Zillow Research, October 18, 2017. https://www.zillow.com/research/climate-
change-underwater-homes-2-16928/
Bruun, P. 1962. “Sea-Level Rise as a Cause of Shore Erosion.” Journal of the Waterways and
Harbors Division 88 (1): 117-132.
Chiles, Frederic Caire. 2015. California’s Channel Islands: A History. Norman: University of
Oklahoma Press.
Church, John A., Neil J. White, Thorkild Aarup, W. Stanley Wilson, Philip L. Woodworth, Catia
M. Domingues, John R. Hunter, and Kurt Lambeck. 2008. “Understanding global sea
levels: past, present and future.” Sustain Sci 3: 9-22. DOI: 10.1007/s11625-008-0042-4
DINAS-COAST Consortium. 2006. "DIVA 1.5. 5." Potsdam Institute for Climate Impact
Research, Potsdam, Germany, CD-ROM. https://unfccc.int/files/national_reports /non-
annex_i_natcom/cge/application/pdf/diva_print_me_first.pdf
Erlandson, Jon McVey. 2012. “As the world warms: rising seas, coastal archaeology, and the
erosion of maritime history.” Journal of Coastal Conservation 16 (2): 137 – 142. DOI:
https://doi.org/10.1007/s11852-010-0104-5
Erlandson, Jon M., Michael H. Graham, Bruce J. Bourque, Debra Corbett, James A. Estes, and
Robert S. Steneck. 2007. “The Kelp Highway Hypothesis: Marine Ecology, the Coastal
Migration Theory, and the Peopling of the Americas.” Journal of Island & Coastal
Archaeology 2: 161-174. DOI: 10.1080/15564890701628612

64
Feagin, Rusty A., Douglas J. Sherman, and William E. Grant. 2005. “Coastal erosion, global sea-
level rise, and the loss of sand dune plant habitats.” Front Ecol Eviron 3(7): 359-364.
DOI: https://doi.org/10.1890/1540-9295(2005)003[0359:CEGSRA]2.0.CO;2
Gesch, Dean B., Benjamin T. Guiterrez, and Stephen K. Gill. 2010. “Coastal Elevations.” In
Coastal Sensitivity to Sea Level Rise Focusing on the Mid-Atlantic Region, edited by
Melvin C. Urajner, 27 – 61. New York: Nova Science Publishers, Inc.
Gornitz, V. 1990. Vulnerability of the East Coast, U.S.A. to future sea-level rise. Proceedings of
the Skagen Symposium, J. of Coastal Res. Special Issue No. 9.
Gornitz, Vivian M., Richard C. Daniels, Tammy W. White, and Kevin R. Birdwell. 1994. “The
Development of a Coastal Risk Assessment Database: Vulnerability to Sea-Level Rise in
the U.S. Southeast.” Journal of Coastal Research (Special Issue No. 12: Coastal
Hazards): 327-338.
Gornitz, Vivien M., Tammy W. Beaty, and Richard C. Daniels. 1997. A Coastal Hazards Data
Base for the U.S. West Coast. ORNL/CDIAC-81, NDP-043C, Oak Ridge National
Laboratory, Oak Ridge, Tennessee. DOI: https://doi.org/10.2172/661514
IPCC. 2013. “Summary for Policymakers.” In: Climate Change 2013: The Physical Science
Basis. Contribution of Working Group I to the Fifth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge: Cambridge University Press: 1
– 30. DOI: 10.1017/CBO9781107415324.004
Jazwa, Christopher S. and Jennifer E. Perry, eds. 2013. California’s Channel Islands: The
Archaeology of Human-Environment Interactions. Salt Lake City: The University of
Utah Press.
Jennings, Charles W., Carlos Gutierrez, William Bryant, George Saucedo, and Chris Wills. 2010.
“Geologic Map of California.” California Geologic Data Map Series Version 2.0.
California Geological Society, California Department of Conservation.
Jones, Nicola. 2017. “Sea Level Rise Threatens Archaeological Sites.” Sapiens, November 29,
2017. https://www.sapiens.org/news/sea-level-rise-threatens-archaeological-sites/
Kay, R. C. and A. Travers. 2008. “Coastal vulnerability and adaptation assessment.”
Compendium of coastal resources, tools and methodologies. Report to UNFCCC.
University of Wollongong.
Kotinas, Vasilis, Niki Evelpidou, Anna Karkani, and Miltiadis Polidorou. n.d. Modelling Coastal
Erosion. Athens: National and Kapodistrian University of Athens. Accessed August 12,
2018. https://eclass.uoa.gr/modules/document/file.php/ GEOL312/Coastal%20erosion/
Modelling%20coastal%20erosion_theory.pdf
Li, Xingong, Rex J. Rowley, John C. Kostelnick, David Braaten, Joshua Meise, and Kalonie
Hulbutta. 2009. “GIS Analysis of Global Impacts from Sea Level Rise.”

65
Photogrammetric Engineering & Remote Sensing 75 (7): 807-818. DOI:
https://doi.org/10.14358/PERS.75.7.807
Lichter, Michal and Daniel Felsenstein. 2012. “Assessing the costs of sea-level rise and extreme
flooding at the local level: A GIS-based approach.” Ocean & Coastal Management 59:
47-62. DOI: 10.1016/j.ocecoaman.2011.12.020
Lindsey, Rebecca. 2017. “Climate Change: Global Sea Level.” Climate.gov, September 11.
https://www.climate.gov/news-features/understanding-climate/climate-change-global-
sea-level
Maio, Christopher V., Allen M. Gontz, David Elliot Tenenbaum, and Ellen Berkland. 2012.
“Coastal Hazard Vulnerability Assessment of Sensitive Historical Sites on Rainsford
Island, Boston Harbor, Massachusetts.” Journal of Coastal Research 28 (1A): 20-33.
DOI: 10.2112/JCOASTRES-D-10-00104.1
McCoy, Mark D. 2018. “The Race to Document Archaeological Sites Ahead of Rising Sea
Levels: Recent Applications of Geospatial Technologies in the Archaeology of
Polynesia.” Sustainability 10: 1 – 22. DOI: 10.3390/su10010185
McFadden, Loraine, Athanasios Vafeidis, and Robert J. Nicholls. 2003. "A coastal database for
global impact and vulnerability analysis." In Proceedings of 5th international symposium
on coastal engineering and science of coastal sediment processes: coastal sediments ’03:
18-23.
Nageswara Rao, K., P. Subraelu, T. Venkateswara Rao, B. Malini, Ratheesh Ramakrishnan, S.
Bhattacharya, A. S. Rajawat, and Ajai. 2008. “Sea-Level Rise and Coastal Vulnerability:
An Assessment of Andhra Pradesh Coast, India through Remote Sensing and GIS.”
Journal of Coastal Conservation 12 (4): 195-207. DOI: 10.1007/s11852-009-0042-2
NOAA. 2010. “Technical Considerations for Use of Geospatial Data in Sea Level Change
Mapping and Assessment.” NOAA Technical Report NOS 2010-01.
NOAA. n.d. “Sea Level Rise Viewer.” https://coast.noaa.gov/slr/
Olmsted, F. H. 1958. Geologic Reconnaissance of San Clemente Island California. Geological
Survey Bulletin 1071-B. Washington: United States Printing Office.
https://doi.org/10.3133/b1071B
Özyurt, Gülizar and Aysen Ergin. 2010. “Improving Coastal Vulnerability Assessments to Sea-
Level Rise: A New Indicator-Based Methodology for Decision Makers.” Journal of
Coastal Research 26 (2): 265-273. DOI: https://doi.org/10.2112/08-1055.1
Reeder, Leslie A., Torben C. Rick, and Jon M. Erlandson. 2012. “Our disappearing past: a GIS
analysis of the vulnerability of coastal archaeological resources in California’s Santa
Barbara Channel region.” Journal of Coastal Conservation 16 (2): 187-197. DOI:
10.1007/s11852-010-0131-2

66
Reeder-Myers, Leslie A. 2015. “Cultural Heritage at Risk in the Twenty-First Century: A
Vulnerability Assessment of Coastal Archaeological Sites in the United States.” The
Journal of Island and Coastal Archaeology 10 (3): 436-445. DOI:
https://doi.org/10.1080/15564894.2015.1008074
Rick, Torben C., Jon M. Erlandson, René L. Vellanoweth, and Todd J. Braje. 2005. “From
Pleistocene Mariners to Complex Hunter-Gatherers: The Archaeology of the California
Channel Islands.” Journal of World Prehistory 19 (3): 169-228. DOI:
https://doi.org/10.1007/s10963-006-9004-x
Rowland, M. J. 1992. “Climate Change, Sea-Level Rise and the Archaeological Record.”
Australian Archaeology 34 (1): 29-33. DOI: 10.1080/03122417.1992.11681449
Shepard, F. P. and U. S. Grant IV. 1947. “Wave Erosion Along the Southern California Coast.”
Bulletin of the Geological Society of America 58: 919-926. DOI:
https://doi.org/10.1130/0016-7606(1947)58[919:WEATSC]2.0.CO;2
Smith, Charlotte A. and Jennifer Freer Harris. 2001. “Why is Archaeology Important? Global
Perspectives, Local Concerns.” Early Georgia 29 (1): 27-33.
Theodore Payne Foundation. n.d. “Plants for Erosion Control.”
http://www.theodorepayne.org/plants/plants_for_erosion_control.htm
Thieler, E.R. and E.S. Hammar-Klose. 1999. National Assessment of Coastal Vulnerability to
Future Sea Level Rise: Preliminary Results for the U.S. Atlantic Coast. U.S. Geological
Survey, Open-File Report 99-593, 1 sheet.
Thieler, E.R. and E.S. Hammar-Klose. 2000. National Assessment of Coastal Vulnerability to
Future Sea-Level Rise: Preliminary Results for the U.S. Atlantic Coast. U.S. Geological
Survey, Open File Report 00-178, 1 sheet.
Vafeidis, Athanasios T., Robert J. Nicholls, Loraine McFadden, Richard S. J. Tol, Jochen
Hinkel, Tom Spencer, Poul S. Grashoff, Gerben Boot, and Richard J. T. Klein. 2008. “A
New Global Database for Impact and Vulnerability Analysis to Sea-Level Rise.” Journal
of Coastal Research 24 (4): 917-924. DOI: https://doi.org/10.2112/06-0725.1
Westley, Kieran, Trevor Bell, M. A. P. Renouf, and Lev Tarasov. 2011. “Impact Assessment of
Current and Future Sea-Level Change on Coastal Archaeological Resources – Illustrated
Examples from Northern Newfoundland.” The Journal of Island and Coastal
Archaeology 6 (3): 351-374. DOI: https://doi.org/10.1080/15564894.2010.520076
Williams, S. Jeffress and Benjamin T. Gutierrez. 2009. “Sea-level rise and coastal change:
Causes and implications for the future of coasts and low-lying regions.” Shore & Beach
77 (4): 13 – 21.
USGS. 2007. Divisions of Geologic Time – Major Chronostratigraphic and Geochronologic
Units. U.S. Geological Survey. Fact Sheet 2007 – 3015.
https://pubs.usgs.gov/fs/2007/3015/fs2007-3015.pdf

67
USGS. n.d. What is the Difference Between Consolidated and Unconsolidated Sediments?
Frequently Asked Questions. Accessed August 12, 2018.
https://www.usgs.gov/faqs/what-difference-between-consolidated-and-unconsolidated-
sediments?qt-news_science_products=0#qt-news_science_products
Uyeda, Kellie, Kelsey Warkentin, Douglas A. Stow, John F. O’Leary, Thomas Zink, and Julie
Lambert. 2018. Vegetation Mapping at NALF San Clemente Island, Naval Base
Coronado, California. Prepared for U.S. Army Corps of Engineers. Final Report.
Reporting period: 1 October 2015 – 30 June 2018.
Yatsko, Andy. 2000. “From Sheepherders to Cruise Missiles: A Short History of Archaeological
Research at San Clemente Island.” Pacific Coast Archaeological Society Quarterly 36
(1): 18-24. http://www.pcas.org/Vol36N1/2Yatsko.pdf

68
Appendix A – A Short History of San Clemente Island and San Nicolas Island
San Nicolas and San Clemente Islands have long histories spanning thousands of years.
From archaeological data, researchers believe San Nicolas Island was first settled 8,000 years
ago and San Clemente Island was settled around 9,000 years ago, if not earlier (Yatsko 2000).
With many similarities in technology and evidence of abundant trade between the two islands,
archaeologists have attempted to find familiar relationships between the two island populations.
However, because the languages of the native people on two islands were not the same, the
genetic relationships of these islands could not be confirmed. The histories of these cultures may
seem remarkably similar on the surface, the trajectory of the islands and their inhabitants deviate
in the years just before contact (Chiles 2015).
San Clemente Island
With the general isolation of San Clemente Island, the native peoples of the island were
believed to be excellent sea voyagers. Evidence of this can be seen in the Clementeños deep
archaeological record that spans from the Early Holocene to Spanish contact in the mid-sixteenth
century. While the pre-contact history of this island parallels the story of San Nicolas Island, San
Clemente Island’s initial contact with outside forces occurred much earlier than their San Nicolas
neighbors (Chiles 2015).
Juan Rodríguez Cabrillo first spotted San Clemente Island in 1542, but contact with the
islanders was not initiated until 1796 when the Franciscans and Russian-American otter hunters
sought out the island for its people and the resources it held. Following a similar history as the
Nicoleños after contact, by 1803, the approximately 250 Clementeños were reduced to only
eleven. The last of these islanders departed for the mainland around 1829 for Mission San
Gabriel (Chiles 2015).

69
In 1850, the United States government officially owned the title for San Clemente Island
when California was granted statehood. In the years to follow, the island was used as a safe
harbor for whalers, fishermen, hunters, ranchers, and even scientists. Unfortunately, as with San
Nicolas, sheep herding dramatically changed the vegetation of the island. By 1934, leasing of the
land ended and administration of the island was given to the Navy (Chiles 2015).
San Nicolas Island
The Nicoleños of San Nicolas Island were successful traders, sea mammal hunters, and
fishermen. With numbers reaching 200 to 300 individuals, the Nicoleños were successful on the
small island. San Nicolas Island was the only island to evade notice of the Spanish, Mexicans,
and missionaries for more than two centuries. However, by the nineteenth century, Russian-
American otter hunters found the island and made first contact with the native peoples. After this
contact, Franciscan missionaries heard of native peoples on this island and enticed many of the
Nicoleños to come to the missions in Los Angeles and San Gabriel. The earthquake of 1812
convinced many of the native peoples to go to the mainland with the missionaries, but most of
them died due to diseases and poor living conditions. By the end of the second decade of the
nineteenth century, fewer than two-dozen Nicoleño people remained. By 1835, the Franciscans
sent for the remaining Nicoleños and all but one woman left for the mainland. The one woman,
whose picture can be seen in Figure 20, who remained on the island sparked the legend of “The
Lone Woman of San Nicolas” which inspired the classic novel “The Island of the Blue
Dolphins” by Scott O’Dell. While only a legend for eighteen years, the Lone Woman was finally
found in 1853. Her finding sparked scientific interest of the island with archaeologists (Chiles
2015).

70

By 1858, the first modern settlers came to the island that was now under federal
government control. These settlers brought sheep herding to the island, and by the 1870’s, the
vegetation of the island had changed dramatically from the herds. Since 1934, the United States
Navy took over the administration of the island. To this day, San Nicolas Island is a designated
auxiliary air station and training ground for the U.S. Navy (Chiles 2015).

Figure 21 - A photograph of a woman
believed to be the Lone Woman of
San Nicolas Island, 1853 (Source:
Chiles 2015)

Abstract (if available)

Abstract Archaeology allows us to see our human past and who we are as a people. This is a global narrative that spans the entirety of human existence. Many archaeological sites are delicate and are often unknowingly destroyed by human development. Because of this, pristine and protected islands offer a complete wealth of archaeological information. Stewardship programs and regulations set in place for protecting these cultural resources have been set into place on federally owned lands. San Nicolas Island and San Clemente Island, two of the Channel Islands owned by the United States Navy, are among the most well-documented and protected locations for archaeological sites in the United States. However, many of these sites are currently at risk from inundation and erosion. Global sea level rise not only potentially inundate the coastal zones but also accelerate geological erosion processes. ❧ To help the U.S. Navy understand and protect against the threats from these natural processes, this study aims to identify the at-risk archaeological sites on San Clemente Island and San Nicolas Island. A spatial-explicit Coastal Vulnerability Index (CVI) was developed from the ranked vulnerability score of environmental variables, including slope, inundation, generalized rock type, and vegetation, using a Geographic Information Systems (GIS). Based on the CVI, a Cultural Resource Vulnerability Index (CRVI) was developed to rank the coastal vulnerability of the archaeological sites on the two islands. ❧ The results of the CRVI showed that 3.6% of the archaeological sites on San Nicolas Island and 19.2% of the archaeological sites on San Clemente Island fall within the Highly Vulnerable to Very Highly Vulnerable categories. The CRVI informs the land managers in the U.S. Navy an earlier response time to save these at-risk sites that may be completely destroyed in the next 100 years. With the result from the CRVI, further actions can be taken to mitigate and/or excavate these at-risk sites to preserve the rich cultural resources of San Nicolas Island and San Clemente Island.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

A model for emergency logistical resource requirements: supporting socially vulnerable populations affected by the (M) 7.8 San Andreas earthquake scenario in Los Angeles County, California

Asset Metadata

Creator Oulie, Tori Kristin (author)

Core Title Coastal vulnerability assessment for archaeological sites on San Clemente Island and San Nicolas Island, California

School College of Letters, Arts and Sciences

Degree Master of Science

Degree Program Geographic Information Science and Technology

Publication Date 11/09/2018

Defense Date 10/16/2018

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

Tag ArcGIS,ArcGIS Pro,archaeological sites,archaeology,California,Channel Islands,coastal,coastal erosion,coastal vulnerability,Coastal Vulnerability Index,coastline,CRVI,cultural resource management,Cultural Resource Vulnerability Index,cultural resources,CVI,environmental,environmental science,Erosion,Esri,geographic information sciences,geographic information system,GIS,inundation,islands,mitigation,Navy,OAI-PMH Harvest,San Clemente Island,San Nicolas Island,sea level rise

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Wu, An-Min (committee chair), Longcore, Travis (committee member), Oda, Katsuhiko “Kirk” (committee member)

Creator Email oulie@usc.edu,torioulie@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-102332

Unique identifier UC11675692

Identifier etd-OulieToriK-6943.pdf (filename),usctheses-c89-102332 (legacy record id)

Legacy Identifier etd-OulieToriK-6943-2.pdf

Dmrecord 102332

Document Type Thesis

Format application/pdf (imt)

Rights Oulie, Tori Kristin

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