User manual | Interpreting maritime cultural space through the

Interpreting maritime cultural space through the
utilization of GIS: a case study of the spatial
meaning of shipwrecks in the coastal waters of
South Australia
Jun Kimura
Master of Maritime Archaeology
Department of Archaeology
Flinders University
South Australia
2006
Interpreting maritime cultural space through the
utilization of GIS: a case study of the spatial
meaning of shipwrecks in the coastal waters of
South Australia
Jun Kimura
Master of Maritime Archaeology
Department of Archaeology
Flinders University
South Australia
2006
Acknowledgements
Without many people’s contributions, this study would not have been completed.
The author would like to express thanks to Ms Jennifer Mckinnon for supervising
me with the relevant advice.
Her profound insight allowed me to take a logical
approach to the research, and her assistance enabled me to access the relevant
resources for the analysis. My thanks are for her consistent support throughout
the entire process.
My deep appreciation is especially expressed to Dr Mark Staniforth.
The
concept of this study derived from the knowledge and experience that I obtained
from his teaching during the postgraduate program of maritime archaeology.
His
generosity and encouragement helped me complete this study.
Thanks to my senior colleagues for their insightful advice.
Present technical
officer Jason Raupp has greatly contributed to enhance the quality of this paper
with editorial assistant. Thanks to previous technical officer Rick Bullers for
providing pertinent instructions to my questions.
My coworker Amer Khan
assisted me in developing the theoretical and technical aspects of this study.
i
The data used for this study were obtained from several organizations. The
author would like to express his appreciation to the following agencies for
allowing the use of their resources. Thank you to Terry Arnott of the Department
of Environment and Heritage of South Australia. Thank you also to Merridy
Lawlor of the State Library of South Australia.
The School of Geography,
Population and Environmental Management of Flinders University provided the
GIS data, which was essential to this study, and is greatly appreciated.
Thanks
very much to Mark Lethbridge for giving permission to access the database, and
also thanks to Robert Keane for assisting with data analysis.
Thanks are also extended to the following people who contributed to the progress
of my study; Amanda Hale, Bill Welsh, Brandi Lockhart, Debra Shefi, Dianna
Zwart, Dr Joe Flatman, Linda Honey, Peta Knott, Sharan Bhaskar and Dr Susan
Briggs.
work.
More people who are not listed in here helped me with completing this
The author would like to express genuine appreciation to all those people.
ii
Contents
1
Introduction ……………………………………………………………….....1
1-1 Introduction ………………………………………………………………..1
1-2 Research themes …………………………………………………………...4
2
Literature Review ………………………………………………………..…..7
2-1 What is the meaning of space in archaeology? …………………………….7
2-2 The concept of maritime cultural landscape ……………………………...12
2-3 The traditional approach of wreck site formation process ………………..16
2-4 Geographic Information Systems in archaeology ………………………...23
2-5 The case study of the usage of Geographic Information Systems in
maritime archaeology ……………………………………………………30
3
Methodology …………………………………………………………...…...38
3-1 GIS data collection strategy ………………………………………………38
3-2 GIS models of predictive modeling……………………………………….44
3-3 The determination of variables for maritime spatial analysis …………….53
4
The analysis of maritime space of South Australia ………………………73
4-1 Interpreting maritime space ………………………………………………73
4-2 Spatial analysis …………………………………………………………...93
4-3 Calculating the correlation between the location of shipwrecks and water
depth ……………………………………………………………………117
5
Discussion and Conclusion ……………………………………………….130
5-1 Interpreting the resultant analysis ……………………………………….130
5-2 Perspectives to research questions ………………………………………140
5-3 Conclusion ………………………………………………………………148
References ………………………………………………………………….….150
Appendix ………………………………………...…………………………….160
I - GIS resources on the Internet …………………………………………….160
II - Description of South Australian Shipwreck database …………………...161
III - Data of shipwrecks located in the visible area of each navigational light
…………………………………………………………………………..162
IV - Elevation data of shipwrecks ……………………...……………………169
iii
List of Figures
Figure 1-1-1. Location map of study area: The gulf coast of South Australia ……3
Figure 2-3-1. Muckelroy’s original model of shipwreck site formation processes
……………………………………………………………………………………17
Figure 2-3-2. Expanded model of wreck deterioration processes ……………….19
Figure 2-3-3. Recent developed model highlighting cultural factors in shipwreck
site formation …………………………………………………………………21
Figure 2-4-1. Concept of GIS data layer in archaeology ………………………..26
Figure 3-1-1. The image of a risk assessment model ……………………………51
Figure 3-3-1. Development of South Australian Marine Board …………………62
Figure 4-1-1. Locations of wrecking 1800 – 1899 ………………………………75
Figure 4-1-2. Admiralty Charts: Australia. South coast. Gulf of St. Vincent and
Spencer 1855 – 1865 ………………………...………………………………..80
Figure 4-1-3. Admiralty Charts: South Australia – St. Vincent and Spencer Gulfs
1863 – 1918 …………………………………………………………………...81
Figure 4-1-4. Lighthouse map of the province of South Australia 1883 ………..82
Figure 4-1-5. The vector data of coastline of South Australia …………………..91
Figure 4-1-6. An example of Georeferencing …………………………………...92
Figure 4-2-1. Conceptual framework about the spatial analysis of maritime space
based on the distribution of shipwrecks ………………………………………94
Figure 4-2-2. Comparative graph regarding the historical changes of the
shipwrecks and vessels in South Australian waters ………………………..95
Figure 4-2-3. Construction date of navigational lights in South Australia ……...96
Figure 4-2-4. The location information of the navigational lights provided by the
chart “Lighthouse map of the province of South Australia 1883” ……………..102
Figure 4-2-5. The distribution of 218 shipwrecks dating from 1837 to 1899 on the
chart “Lighthouse map of the province of South Australia 1883” from the data
frame of ArcMap …………………………………………………………….103
Figure 4-2-6. The spatial relationship between shipwrecks and navigational lights
in South Australia in the mid-nineteenth century ……………………………111
Figure 4-2-7. The spatial relationship between shipwrecks and navigational lights
in South Australia in the late nineteenth century …………………………….112
Figure 4-2-8. Chronological changes in shipping routes in South Australia during
the latter half of nineteenth century ………………………………………….115
Figure 4-3-1. Bathymetric data for South Australia ……………………………118
iv
Figure 4-3-2. TIN surface modeling demonstrating the change in elevation by
color and relief …………………………………………………………………120
Figure 4-3-3. A dataset comprised of quite a number of geographic contour data
that provide South Australia containing sea level value ……….……………122
Figure 4-3-4. Establishing the boundary of the analytical area from the geographic
feature data containing the geo-coordinate system …………………………123
Figure 4-3-5. Digital Elevation Model for the gulf coast of South Australia ….124
Figure 4-3-6. Specific values for depth in the DEM are extracted and added to the
shipwrecks data ……………………………………………………………...126
Figure 4-3-7. Examples of extracted values regarding water depth from DEM in
the South Australian shipwrecks dataset …………………………………….127
Figure 5-2-1. A shipwreck assessment model focusing on pre- and
post-depositional cultural and natural factors in the macro and micro level
analysis of shipwrecks ……………………………………………………...…..142
v
List of Tables
Table 2-5-1. Fourteen case studies of GIS application in maritime archaeology
presented ………………………………………………………………………...36
Table 3-1-1. Procedure for the establishment of database in GIS analysis ……...39
Table 3-3-1. Equating land and marine variables for GIS ……………………….55
Table 4-1-1. The geographic coordinate system adopted in the datasets of the
South Australian shipwrecks …………………………………………………….86
Table 4-1-2. Historical charts obtained from the State Library of South Australia
for the purpose of analyzing navigational aids ………………………………….89
Table 4-2-1. The construction date of navigational lights expressed on the
“Lighthouse map of the province of South Australia 1883” ……………….……97
Table 4-2-2. The visibility of navigational lights described on the chart
“Lighthouse map of the province of South Australia 1883” ……………….…..101
Table 4-2-3. The number of shipwrecks from the 1830s through the 1890s located
in before and after establishing of the navigational lights ……………………..106
Table 4-2-4. An example of calculation based on the average number of the
vessels wrecked on the light visible area per year ……………….…………….108
Table 4-3-1. Statistical data regarding the number of shipwrecks (1837-1899) and
water depth in South Australian waters ………………….….………………….129
Table 5-1-1. Agencies responsible for lighthouses and jetties until the
establishment of the Harbors Board in1914 ……………………………………136
Table 5-1-2. Data regarding the relationship between water depth and the number
of ancient shipwrecks in the Mediterranean Sea ……………………………….139
vi
1 Introduction
1-1 Introduction
The relationship between space and archaeological remains has been a main
research theme for archaeologists. My research will examine the spatial analysis
of underwater sites using a Geographic Information Systems (GIS) case study of
shipwrecks located in South Australian coastal waters.
The study will pursue the
distributional meaning of 218 shipwrecks dating to the second half of the
nineteenth century in correlation with the cultural and natural factors that affected
their distribution.
Maritime archaeologists have highlighted cultural and natural factors using site
formation processes theory in order to understand the distributional context of
single wreck sites (Muckelroy 1978; Ward et al. 2001; Gibbs 2006).
According
to Clarke (1977, p. 12), however, archaeologists can also extend the perspective of
spatial resolutions from a micro level to a macro level in the way of interpreting
sites.
One of the key concepts of this research is to demonstrate the significance
of a holistic, macro approach through the analysis of distribution of multiple
1
shipwrecks.
The concept of a holistic spatial analysis has improved due to a recent
development of maritime cultural landscape theory (Westerdahl 1992).
This
theory allows for maritime archaeologists to conduct research on not only one
shipwreck but multiple shipwrecks, the natural environment and other maritime
infrastructures within the maritime landscape. Consequently, this analysis will
use maritime cultural landscape theory to highlight the role of navigational lights
as cultural factors and bathymetric data as a natural factor which influenced the
location of South Australian shipwrecks.
This analysis employs the concept of GIS predictive modeling as an analytical
model to investigate the correlation between the location of shipwrecks and
cultural and natural factors identified using landscape theory.
The validity of
predictive modeling in shipwreck spatial analysis is controversial (Mather &
Watts 2002, pp. 693-694), and only a few models exist in assessing maritime
space (Boyd et al. 1996).
The purpose of this analysis is to demonstrate the
potentials for using GIS in maritime archaeology as an analytical model to
2
understand the distribution of multiple shipwrecks.
Figure 1-1-1. Location map of study area: the gulf coast of South Australia
3
1-2 Research themes
Using GIS analysis supported by both a theoretical framework of site formation
processes and maritime cultural landscape, the distribution of shipwrecks in South
Australian coastal waters will be examined.
This thesis will give some insight to
the following ideas:
1.
A focus on the analysis of comprehensive site formation processes in macro
space using multiple shipwrecks, rather than the traditional, particularistic
site formation processes studies of individual shipwrecks.
2.
Site formation processes and maritime cultural landscape theories can be
used to identify factors of past human activity and the physical environment,
which affect the spatial distribution of shipwrecks.
3.
The interpretation of the spatial relationships between the distribution of
shipwrecks and other maritime infrastructures, such as jetties and
lighthouses can demonstrate the significance of established maritime culture
population centers.
4
4.
Although the validity of using GIS as a predictive model to analyse
underwater sites is still disputable, this study may help in determining some
of the natural variables, such as prevailing winds, currents, hydrography,
channels alignment, and bottom sediments that affect shipwreck
distribution.
This thesis consists of five chapters which outline the course of research.
Chapter One introduces the reader to the fundamental concept and structure of this
study.
Chapter Two presents an overview of theoretical aspects in this study.
A
literature review highlights the significance of space in archaeology with a brief
description of its progress in archaeology.
The purpose of this section is to
understand how using GIS can contribute to archaeological research.
The
second and third sections are a review of the theory of maritime cultural landscape
and wreck site formation processes, which support the conceptual framework of
this study. In the forth and fifth sections, the development of using GIS in both
terrestrial and maritime archaeology will be addressed.
5
Chapter Three will introduce the spatial analysis methodology employing
ArcGIS9.
The first section demonstrates the methods used to collect
geo-spatial data in the process of establishing a database for spatial analysis.
The second section pursues the application of existing GIS models to the analysis
of maritime cultural space.
In the third section, the natural and human factors
adopted in this analysis are identified.
Chapter Four will outline the actual process of GIS analysis. This section will
demonstrate the approach used to reveal the distributional meaning of South
Australian shipwrecks in correlation with natural and cultural factors.
In the first
section, the maritime cultural space of South Australia is described based on
historical and archaeological perspectives.
Chapter Five describes the results of spatial analysis.
The first section provides
perspectives for the correlation between shipwrecks and natural and human
factors within the maritime space of South Australia. In the second section,
insights into future research are addressed. Finally, the significance of the results
of this study is summarised in the third section.
6
2 Literature Review
The purpose of this chapter is to provide essential knowledge and outline key
concepts of implementing an integrated approach of spatial analysis and
Geographic Information Systems (GIS) in maritime archaeology in South
Australia (S.A.).
The first three sections of this chapter present an introduction
to spatial archaeology, maritime cultural landscape, and wreck site formation
processes, and provide a theoretical framework for the analysis of shipwrecks.
The latter sections review examples of the utilisation of GIS in both terrestrial and
underwater research.
These multiple approaches will counter criticism about
the insufficiency of theoretical approaches in maritime archaeology today, and
will complement the idea of the adaptation of GIS to the study of maritime
cultural heritage.
2-1 What is the meaning of space in archaeology?
GIS analyses process spatial information that is comprised of material remains
7
and geographical factors.
In terms of an intimate link between the concept of
GIS and the study of space in archaeology, a few publications focusing on GIS
analysis in archaeology emphasise the importance of previous studies of spatial
archaeology as a precursor for the development of present GIS analysis. Some
examples of these include those by Allen, Green and Zubrow (1990), Kaneda,
Tsumura and Niiro (2001) and Wheatley and Gillings (2002).
Space as a static aspect and its precise measurement
The principles of archaeology deal with spatial data on diverse scales, ranging
from the location of sites upon a landmass to the distribution of artefacts within a
single site.
In order to better understand the spatial relationship of these
archaeological components, initial archaeological work records exact distributions
of remains. This data is used to produce precise site plans and elevation maps.
This approach dates back to at least the early formal excavations in the eighteenth
century, including the plan of the Villa of Papyri in Herculaneum drawn up by
Weber and the scaled plans and sections of the Roman earthwork of the Bokerly
dyke produced by Pitt-Rivers (Wheatley & Gillings 2002).
Throughout the
history of the discipline, meticulous measurements used to produce maps in the
field have been the only effective scientific means for archaeologists to evaluate
8
space.
The importance of producing site plans which leads to the comprehension
of horizontal relationships of archaeological features in space, and vertical profiles
which aid in the determination of temporal sequences, is accepted as basic
archaeological practice.
Until archaeologists create such maps, spatial and
temporal information from sites cannot be visualised in their survey results.
Precise measurement leads to the understanding of site structure, and artefact
distribution and chronological order of archaeological sequences.
It is important
to note that these earlier analytical frameworks were only revealing static aspects
of space where the meaning of past human activities and influences lying behind
the space was not the focal point of the study.
Space with dynamism
In recent years archaeology has focused more on the dynamism of cultural space
where the dispersion and spread of archaeological remains is studied in relation to
the patterning of human activities and the development of civilization over time.
Firstly, the distribution of archaeological materials in space was prevailingly
explained as an offspring of historical changes of culture (Trigger 1989), and the
idea was initially formed by Childe’s cultural diffusion model.
Later, as a result
of pursuing further scientific verification, a new movement toward the study of
9
economic and evolutionary aspects of cultural remains occurred in the 1940s.
This movement stressed the direct spatial patterning of cultural remains, and went
beyond simple modeling of cultural diffusion (Wheatley & Gillings 2002, p. 5).
An example of this approach includes the examination of cultural ecology and its
adaptation by Julian Steward.
Wheatley & Gillings (2002, p. 5) state “here the
mapping of archaeological sites was undertaken on a regional scale, with the
express purpose of studying the adaptation of social and settlement patterns within
an environmental context i.e. to find causal linkages.’’ Although the concept was
apparently innovative, Hodder has pointed out that using the simple visual
examination of distribution maps at this stage of spatial analysis was intuitive and
subjective (Hodder 1977, p.223; Wheatley & Gillings 2002, p. 5).
Spatial archaeology
A methodological development of spatial archaeology was promoted through the
occurrence of quantitative approaches which stressed objectivity and verifiability
after the rise of the new approach known as New or Processual Archaeology in the
early 1960s by Louis Binford. Considering present spatial patterns as an on-going
cultural system from past human activities and process, processual archaeology
presumes that archaeological evidence of space could be objectively calculated
10
through the systematic testimony and verification that are employed using
deductive approach and statistical methods (Wheatley & Gillings 2002). Harris
and Lock (1990, p. 47) state that a successful case study of spatial analysis based
on the combination of map-based methods and quantitative approaches was
accomplished by David Clarke at the Iron Age settlement site; Clarke (1977)
consequently defined spatial archaeology. Through the introduction of spatial
analysis and the implication of random aspects of the observed archaeological
patterns (Hodder 1977), considerable spatial methods were outlined at this stage.
These same methods can be utilised by modern GIS spatial analysis.
In brief, a
sequence of statistical and quantitative approaches including mathematical
calculations for the analysis of directions, densities and distances of artefact
distributions is regarded as a precursory form of present GIS approaches
(Kvamme 1995, pp. 1-2).
Interpreting the meaning of space
Apart from the advancement of technical aspects of spatial analysis, the
domination of Processual Archaeology began to be superseded by the
Post-processual Archaeology movement in the mid 1980s.
Post-processual
archaeology presented an alternative interpretation of human behavioural patterns
11
in archaeological space (Wheatley & Gillings 2002).
Currently, spatial
archaeologists highlight distinguished cultural phenomenon generated from
distinctive human activities and actions, rather than stressing a systematic aspect
of culture. Archaeologists acknowledge that space contains multiple structures
and diverse information, and this complexity might reflect some surface traits of
space and produce different results upon its analysis.
Although the function of
GIS enables researchers to integrate a large number of elements objectively, the
fact is that scales and variables for spatial analysis depend on the ability to
determine the meaning of space.
2-2 The concept of maritime cultural landscape
The concept of “landscape” in archaeology assists in understanding the physical
and mental elements underlying cultural space.
In particular, the theory of
“cultural landscape” is used as a means to clarify the interaction of human activity
with the natural environment.
The initial idea of cultural landscape in both
European and American archaeology and anthropology dates back to early studies
12
of historic settlement patterns with concern to environmental variables.
Researchers from Cultural-geography emphasised the role of landscape and
geography to determine the conditions which impact upon cultural activities.
The pioneers of this field, such as the English archaeologist O.G.S. Crawford
demonstrated historical changes in a region through several eras by archaeological
and environmental distribution maps (Wheatley & Gillings 2002).
His approach
on the basis of distribution maps examined the characterisation and spatial
delineation of archaeological evidence not simply using the dots of sites, but also
adopting the lines and the large-scale features for description of field boundaries,
remains of irrigation and agricultural systems dykes (Johnson 2005, p. 156).
In modern archaeological theory, however, apart from the perspective adhered to
by the Crawford tradition, archaeologists have given the term landscape diverse
meanings: landscape as a set of resources, or site catchment/territorial analysis;
landscape as a reflection of society and its relation to theories of the formation of
complex societies and states; and landscape as an expression of people’s ways of
thinking and acting upon the world (Johnson 2005, p. 157).
Knapp and
Ashmore (1999, p. 19) point out “contemporary, etic, archaeological perceptions
13
of the landscape were not those used by prehistoric or early historic societies to
conceptualise their environment.”
Indeed, the usage of cultural landscape
theory in current archaeology is a means of establishing a cognitive framework to
grasp a holistic image of culture and reveal symbolic aspects of society, rather
than to describe physical configurations of the land (Tilley 1994).
Due to
limitations in the archaeological record in which past human experience is not
obviously reflective, archaeologists need to rely on the availability of a conceptual
framework of cultural landscape theory that explains how human mentality
impacts the environment, society, and land.
The term “cultural landscape” was recently introduced into maritime archaeology
in a paper entitled “The maritime cultural landscape” by Christer Westerdahl
(1992).
Westerdahl (1992, p. 6) states that the basic idea of maritime cultural
landscape is aimed at the conceptual unity of maritime cultural remains located on
both land and underwater which resulted from human utilisation of maritime
space.
Within the total topography, maritime features located in the vicinity of a
waterfront are as important as submerged features because the entire range of
maritime economies are indicative of maritime cultural landscapes, including
14
tangible and intangible materials.
A collaborative work by maritime
archaeologists from the Northern Ireland Environment and Heritage Service and
University of Ulster is well known as the case study regarding practicing the
concept of maritime cultural landscape (McErlean et al. 2002).
Apart from the
widespread use of the term maritime cultural landscape among maritime
archaeologists, this theory and its definition are still disputed.
According to
Hunter (1994, pp. 261-262), the use of the phrase maritime cultural landscape
suggested by Westerdahl resulted from his occupational viewpoint as a heritage
manager, rather than a strictly analytical standpoint.
Highlighting the significance of a holistic view of entire maritime infrastructures
and culture is an innovative approach, as compared to the traditional approach of
dealing only with a spatially isolated shipwreck.
More importantly the idea of
maritime cultural landscape is not simply limited to either a narrow dimensional
link or a superficial contemporary relationship among maritime cultural remains
within a perceptible range.
To identify certain components and meaning of the
maritime space is one of many concerns in the conceptual range of maritime
cultural landscape theory.
It is important to demonstrate explicitly the
15
relationship between maritime cultural landscape and its conceptual boundary
from the analytical viewpoint in this study.
2-3 The traditional approach of wreck site formation process
“We must be able to model and understand the archaeological formation
process, both natural and cultural, before we can model where sites might be
found”(Kvamme 2006, p. 7).
Since Keith Muckelroy (1978) first presented the theory of shipwreck site
formation process, the interpretation of shipwrecks that have survived to the
present state has concerned maritime archaeologists.
In comparison with
existing approaches to highlight a series of wreck disintegration processes after a
ship is wrecked (Muckelroy 1978; Ward et al. 2001; Gibbs 2006), there seems to
be room to expand an aspect of pre-deposit formation process of shipwrecks.
The diagram of the evolution of a wreck by Muckelroy (1978) demonstrates the
general deterioration processes of vessels resulting from human and natural
disturbances.
This model was intended to have a potential application to all
16
wreck sites and describes conceivable physical alternations that can modify
important archaeological evidence associated with a sunken vessel and its
associated cargos (Figure 2-3-1). The establishment of the flow chart diagram
assists in the retrieval of past information and helps with the assessment of present
and future conditions of wreck sites.
Ship
PROCESS OF
WRECKING
Material
floated away
SALVAGE
OPERATIONS
Material
salvaged
DISINTEGRATION OF
PERISHABLES
Material
disintegrated
SEABED MOVEMENT
Material
subsequently
CHARACTERISTICS
OF EXCAVATION
OBSERVED SEABED
DISTRIBUTION
Figure 2-3-1. Muckelroy’s original model of shipwreck site formation processes
(Muckelroy 1978, p. 158)
17
Some researchers have made inroads into the development of Muckelroy’s
original model and discussed its concept. For example, the site formation of
HMS Pandora which wrecked on the outer shelf of the Great Barrie Reef was
explained by the four-stage disintegration process with focus on physical,
biological, chemical influences (Ward et al. 1998).
With regard to the existing
model of wreck formation, Ward et al. (2001, p. 213) points out that there is a
paucity of consideration of distinguishing factors between process-related and
produced attributions which result in a more descriptive rather than predictive
method.
From this view, Ward et al. (2001) suggests that the model focuses on
the main processes and influential features that directly affect wreck disintegration
(Figure 2-3-2).
More recently, Martin Gibbs (Gibbs 2006) presents the
developed model of site formation processes emphasising the aspect of human
impacts (Figure 2-3-3).
18
Ship
Material floated
away
PROCESS OF
WRECKING
HYDRODYNAMIC
ENVIRONMENT
SEDIMENT BUDGET
GRAVEL
SAND
MUD
CHARACTERISTICS
OF SEABED
PHYSICALDOMINATED
DETERIORATIO
Wreck
Information
Material
floated away
Material
removed
HIGH Storm
ENERG
LOW
ENERGY
CHEMICAL-ANDBIOLOGICALDOMINATED
DETERIORATION
IN SITU
MODIFICATION
EXCAVATION
OBSERVED
WRECK & SEABED
CHARACTERISTICS
Figure 2-3-2. Expanded model of wreck deterioration processes (Ward et al. 2001,
p. 215)
An underlying idea related to these shipwreck formation theories derives from an
appreciation for the complex cultural and natural processes that could impact the
formation and transformation of archaeological deposits (Schiffer 1987).
In past
the three decades, researchers from the field of behavioural archaeology have
19
developed theories in which the historical nature of archaeological objects, such
as the general sequence of the process of ceramics containing procurement,
manufacture, use, deposition, and decay is seen as a principle research theme
(Lamotta & Schiffer 2005, pp. 121-125).
When integrating the concept of the
historical nature of remains to shipwreck archaeology Muckelroy pointed out that
there is a need to consider the pre-wreck nature of a ship and its content (Gibbs
2006, p. 4). The more comprehensive approach including “pre-depositional”,
“depositional” and “post-depositional” process of wreck is then descriptively
introduced in the analysis of the shipwrecks (O’Shea 2002).
However, Gibbs
(2006, pp. 6-7) states that the concept of pre-depositional, depositional and
post-depositional process is “still primarily oriented towards explanation of the
archaeological deposition and distribution, rather than the cultural processes
behind them.”
20
Ship
DISASTER
AVOIDED
PRE-IMPACT
Threat
PRE-IMPACT
Deliberate
Running
Ashore
Refloating
Warning
Processes
Wreck
IMPACT
Crisis Salvage
RECOIL
Survivor Salvage
Storage
RECONSTRUCTDISASTER
Intentional
Deposition
Systematic
Salvage(s)
Abandoment
Opportunistic
Salvage(s)
Material
Subsequently
Deposited Site
Jettisoning
Salvage of
Jetsam/Lagan
Survivor
Camp
Salvage
Camp
Natural Site
Formation
OBSERVED SEABED
DISTRIBUTION
Figure 2-3-3. Recent developed model highlighting cultural factors in shipwreck
site formation (Gibbs 2006, p. 16)
21
Within the previous approaches of the wreck site formation models, the processes
have been treated solely as an aspect of post-deposit or disintegration which
focuses on the physical influences on a wreck site.
More importantly, the
existing models emphasise the systematic flow of natural and cultural processes
dealing with the site formation processes of a single shipwreck.
Spatial analysis
in this study, however, aims at extracting commonly observed factors from the
distribution of multiple shipwrecks located in the specific areas.
The factors
observed at the multiple shipwrecks will be consequently formulated as dominant
natural and cultural environments that affect the site formation process of
individual shipwrecks within an area. This study pursues the feasibility that
pre-depositional processes will be assessed as factors that determine the
distribution of wrecks.
The concept of shipwreck site formation based on natural
and cultural processes is not a spatially isolated phenomenon.
Apart from the
traditional approaches used to hypothesise the diverse factors that alter a single
shipwreck, the wreck site formation process in this study will highlight the
significance of pre-deposit processes with the identification of natural and cultural
environments affecting at multiple shipwrecks.
22
2-4 Geographic Information Systems in archaeology
Applications of Geographic Information Systems (GIS) are no longer innovative
techniques in archaeology. Indeed, the publication entitled GIS in archaeology: An
annotated bibliography in 1995, provided a substantial number of reports and
articles, and implies a development of this field (Petrie et al. 1995). The purpose
of this section is to provide fundamental knowledge about the role of GIS in
archaeology through its definition, development, utilisation, and recent status in
the discipline.
Despite the growth of recognition about the utility of GIS, dissemination and
familiarisation seem to remain issues among archaeologists. The complexity and
various difficulties such as the high cost of using GIS act as a deterrent for its
widespread use, and understanding its functions.
Researchers have attempted to
define GIS (Savage 1990 pp. 22-23; Kvamme 1999, pp. 157-160; Kaneda et al.
2001, p. 6; Wheatley & Gillings 2002, p. 9).
It is normally regarded as a generic
term to explain a computer-based system to store, manipulate, analyse and present
information about geographic space.
The typical analysis process of geographic
23
information within systems is then summarised as the following (Wheatly &
Gillings 2002, pp. 11-13);
1. Data Entry – Available spatial information is introduced into GIS through
digitisation. This data can be acquired through several methods; for example,
scanning images of aerial photography or regional registers of land, and using
the survey tools such as Total Stations and Global Positioning Systems.
2. Spatial Database – Data within the spatial database is translated into different
layers (or sheets), in which, each layer contains a different theme; examples
include soil types, river networks, or site locations. In addition, at this phase
an internal database can be connected to an external database to deal with real
time data, as well as massive data in the case of large-scale systems.
3. Manipulation and Analysis – A subsystem comprised of various functions
conducts spatial analysis, data transformation, and modeling of data. To take a
brief example of the data manipulation function of GIS, purposive layers are
produced from existing ones.
24
4. Visualisation and Reporting – Visualisation is often considered as the most
representative function of GIS. It creates not simple maps comprised of points
or dots, but visualised maps on the screen.
One evident outcome of GIS is to create a new map resulting from the analysis of
archaeological and geographical data.
In the production of the GIS map,
selective geographic data and databases produce several layers (also known as
“sheets”, “themes”, “coverage” or “images”) to compose the map.
Each layer
belongs to specific geographic or archaeological attributes: for example,
“topography (in the form of contours); hydrology (rivers and streams); modes of
communication (roads, track and pathway); land use (wooded areas, houses, and
industrial areas); boundaries (political or administrative areas); and archaeology
(the point locations of archaeological sites)” (Wheatley & Gillings 2002, p. 25).
Consequently, a typical image of an archaeological GIS map is suggested in the
following diagram (Figure 2-4-1).
25
Figure 2-4-1. Concept of GIS data layer in archaeology (reproduced from
Wheatley & Gillings 2002, p. 26)
In terms of conceptual approaches, the early stages of GIS derive from systematic
spatial analyses and mapping to generalised artefact and site distributions
performed in the late 1970s and the 1980s.
26
Some researchers had begun
employing computer techniques and writing their own computer programmes for
spatial analysis; these include cartography, creation of regional databases, digital
elevation models, and computer simulations in a single region.
A successful case
study was on the Granite Reef archaeological project which collaborated with
computer scientists; it is considered a landmark in computer based spatial analysis
(Kvamme 1995, p. 2). Despite the fact that researchers involved in the project
did not use the term GIS, they established distinct data layers for elevation, soils,
geology, rainfall, temperature, and other environmental factors over a vast area in
order to research environmental suitability for early hunting, travel in the desert,
and prehistoric agriculture.
Indeed, their innovative approaches are little
different from present GIS-based settlement studies (Kvamme 1995, p. 2).
In the
late 1980s, the wide application of GIS was gradually recognised since academic
papers that discussed the usefulness of GIS in the discipline were presented at
annual conferences like the Society for American Archaeology (SAA) in the US
and Computer Applications in Archaeology (CAA) in the UK (Harries & Lock
1990, p. 35).
At present, there is a movement concerning the usage of GIS away from an initial
27
request of predictive models and towards post-research processes consisting of
management and presentation of data (Ghobadi & Tsumura 2006).
From the
beginning, GIS performance was regarded as a beneficial tool for cultural heritage
management.
In reality, the usage of GIS for heritage management is
widespread in the United States, where its ability is not only evaluated for
regional and archaeological data storage, but also for the construction of site
predictive modeling in archaeological methods (Kvamme 1995, p. 3; Kvamme
1999, pp. 171-174; Wheatley & Gillings 2002, p. 19, p. 217; Kvamme 2006, p. 4).
The prevalence of the use of GIS predictive models resulted from the fact that the
US government cultural heritage managers needed to design extensive site
management plans for large areas using limited information based on small-scale
surveys conducted in the 1970s and the 1980s (Wheatley & Gillings 2002, p. 165).
The aims of these predictive modelings were to determine possible locations of
archaeological remains not yet discovered.
The GIS model could theoretically lead to both deductive and inductive analyses.
For example, GIS may enable archaeologists to presume unknown site locations
from geographic features, and include archaeological resources without any
28
particular information about the location of the site itself.
Alternatively,
archaeologists could establish new theories; for instance, the hypothetical location
of sites resulting from observation and interpretation of the relationships between
archaeological remains.
Researchers are likely to pursue the possibility of using GIS to manage and
present archaeological data recovered from site investigations.
The progress of
various technologies, such as remote sensing survey techniques including Global
Positioning System (GPS) and the prevalence of the Internet result in an
enhancement of the quality of geographic data and sharing of geographical
information on regional and global levels. Currently, a large amount of spatial
data, not to mention commercial products, is available through various web sites,
and both local and state governments have contributed to the accumulation of
diverse types of information.
The appearance of leading companies in GIS
software like the Environmental Systems Research Institute (ESRI), play an
important role in dissemination of GIS applications and synthesis of the digital
file format utilised for spatial analysis.
However, it should be noted that the
popularisation of GIS applications based on the growth of the business market
29
might cause an increase in simple non-analytical approaches which pursue only
graphic output and visual effects.
The production of effective visual information
coincides with the outcome of relevant data analysis and is supported by methods
and theories.
The qualitative issues include precise data input and the
acquisition of pertinent information and are inevitably the primary concerns in
GIS analysis.
2-5 The case study of usage of Geographic Information Systems in maritime
archaeology
The use of GIS in maritime archaeology for spatial analysis can be understood
through its applications.
Maritime archaeologists have employed GIS for
different research purposes due to its potential (Mather & Watts 1998; Murphy
1998b).
Due to its various functions and high capacity, the use of GIS is not
bound to only one purpose.
It is suggested that the range of applications for
GIS in maritime archaeology can be classified into four main realms: survey, site
specific investigations, cultural resource management, and predictive modeling
30
and exploratory data analysis (Mather & Watts 2002, p. 681).
Based on this clarification, primarly GIS could be used as an efficient survey tool
in the analysis of collected data created during remote sensing survey work.
This is the most fundamental application of GIS, that is, to distribute the data from
several surveys on a map.
“Post-processed data from magnetometers, side-scan
sonars, and positioning systems can be overlaid in a GIS to permit rapid visual
and mathematical analysis” (Mather & Watts 2002, p. 682).
An early use of
GIS based on this concept was employed during the investigation of the Dry
Tortugas in the Florida Keys.
There the data from magnetic and acoustic
surveys and positioning surveys including proton precession magnetometer, side
scan sonar, subbottom profiler, survey fathometer, RoxAnn bottom classification
equipment, and DGPS (Differential-corrected Geographic Positioning System),
were transformed into a GIS format (Murphy 1993).
Another example is the
case of the Civil War submarine H.L Hunley, located in outer Charleston Harbor,
South Carolina.
Nondestructive surveys for site assessment were conducted
using remote sensing equipment, and combined with comprehensive research
including data collection, post-plotting, analysis and presentation designed in a
31
GIS database (Murphy 1998a).
More recently, researchers employed HPASS
(High Precision Acoustic Surveying System) to produce a site plan. HPASS is a
sort of sound system available to fix particular positions or to plot the main figure
of the site (Green & Souter 1999).
In this process, GIS helped to describe the
configuration of the site from the HPASS data.
The next application of GIS refers to a site specific investigation. According to
Mather & Watts (2002, p. 691), GIS can store all archaeological data generated by
a site-specific investigation including architectural and construction features,
stratigraphic profiles and artefact provenience, photographic documentation, and
historical or literary reference.
This combined data can allow archaeologists to
conduct comparative research about artefacts and features among different sites.
In the case of the USS Monitor National Marine Sanctuary in North Carolina (the
test project to realise the potential of GIS), digital images such as photomosaic
maps and sonar mosaics were combined and adjusted in order to produce an
image of wreck on the screen; in addition, spatial information was developed by
the supplementing data from artefacts and site investigations (Mather & Watts
2002, pp. 691-693).
32
Primary predictive modeling by GIS analysis is expected to be used for the
demonstration of shipwreck potential in uninvestigated waters where all
environmental information, such as ship routes, bottom depths and sediments,
prevailing winds, and currents could be transformed into data.
However, it is
indicated that at present there is difficulty in constructing GIS-based predictive
models for ship losses due to immensely complex, expensive, and limiting factors
(Mather & Watts 2002, pp. 693-694).
Although there is a case study which
tried to locate the German World War II submarine U-559 by GIS analysts and
geophysicists; their approaches only suggest a model focusing on the analysis of
data from the result of remote-sensing survey and marine environmental factors
might be replicated (Cooper et al. 2002).
The procedure of predictive modeling
has not been established yet in maritime archaeology.
The Richmond River shipwreck heritage risk GIS model should be introduced as a
remarkable example being close to predictive modeling which used deductive
approaches.
Although this project was based on the concept of predictive
modeling, Boyd et al. (1996) emphasises that the purpose of the study
fundamentally concentrated on an estimation of the degree of risk for buried
33
remains rather than to predict the actual location of unidentified shipwrecks.
According to the report (Boyd et al. 1996), the GIS model adopted data from
previous predictive modeling research regarding the implementation of risk
assessment to buried remains within the river mouth and coastal area.
In this
case, the usefulness of GIS in maritime archaeology is proven through the
construction of a risk assessment model collaborated with a predicative modeling
approach.
The use of GIS predictive modeling in maritime archaeology
remains a matter to be discussed further, and continued research in this field with
the accumulation of data will clarify the utility of GIS applications to pure
predictive models.
It is expected that the most profitable use of GIS for maritime archaeology is the
application to underwater culture heritage management.
Especially in the
United States, cultural resource managers have attempted to conduct GIS-based
underwater cultural resource management that is equal to terrestrial archaeology.
Mather & Watts (2002, pp. 682-683) state that the legislation in the United States
provides the relevant assessment and the identification of submerged cultural
resources for the registered heritage place, which would face developmental
34
impact.
As a result, heritage managers have adopted GIS to store, organise, and
administer the large amount of archaeological and historical data.
The
following are four examples of GIS based underwater culture heritage
management projects; James River Historic Properties Treatment Plan GIS;
Charleston Harbor Project GIS; U. S. Navy Resource Management Program; Dry
Tortugas and Key Biscayne Shipwreck and Natural Resource Survey System and
GIS (Mather & Watts 2002, pp. 683-691).
The use of GIS in cultural heritage management can provide inclusive means for
manipulating multiple data ranging from primary information like survey results
to secondary information generated by data processed by other computer
programmes.
To determine future research and assessment, it is important to
estimate particular sensitive zones within heritage areas that are classified using
specific criteria such as the degree of threat in heritage management.
Also,
many cases of heritage managements designed by GIS programmes provide
people with access to cultural heritage resources on the internet.
Thus GIS in
heritage management interprets available information for a comprehensive
understanding of data.
35
Case Studies of GIS analysis in maritime archaeology
Dry Tortugas National Preserve
H.L. Hunley Site Assessment
River Richmond, New South Wales
Langstone Harbour, University of Portsmouth
Sea of Marmara, University of Florida
Galle Harbour Project
Table Bay, University of Cape Town
Florida Bay, Florida
Gulf of Marine Info Atlas, Marine
UKDMAP, British Oceanographic Centre, Birkenhead
Crown Estate Commissioners, Edinburgh
Forth Estuary Foram, Scottish Natural Heritage, Edinburgh
Sea Mammals Research Unit, University of St Andrew
Danish National Record of Marine and Maritime Sites
Table 2-5-1. Fourteen case studies of GIS application in maritime archaeology
presented (reproduced from Groom & Oxley 2002 p. 52).
The idea of classification of GIS applications and the introduction of associated
cases are provided in order to clarify the usefulness of GIS in maritime
archaeology. Groom and Oxley (2002) reference fourteen case studies within
maritime archaeology to establish a feasible survey in Fife (Table 2-5-1).
These
fourteen cases demonstrate a diversity of applications of GIS in maritime
archaeology and prove that GIS is highly useful in applications of a direct manner,
a database tool, or its application for heritage management (Christoffersen 1994, p.
287).
By using GIS software, maritime archaeologists can facilitate basic
36
research and data manipulation without sophisticated mathematical ability
(Murphy 1998a).
On the other hand, it should be noted that cultural heritage
agencies might possibly face the administrative issues of GIS programmes such as
the investment and development of systems.
The heritage management based on
GIS should be carefully designed (Wheatley & Gillings 2002, pp. 299-232).
Nevertheless, it is valid to say that the utility of GIS is apparent to maritime
archaeologists through the above cases and therefore associated issues need to be
discussed.
37
3 Methodology
This chapter introduces the methods employed in this study. The first section
will discuss the policy of data collection using digital resources from the Internet.
The next section deals with the establishment of an appropriate GIS framework
for maritime spatial analysis.
This will integrate the concepts of predictive
modeling and risk assessment as effective tools for revealing the meaning of
archaeological material in space.
The final section explains how the variables
for spatial analysis in South Australian coastal waters are selected with
consideration to both natural and social factors.
3-1 GIS data collection strategy
Analyses using GIS require gradual and systematic steps. The purpose of this
section is to reveal the necessary procedures in establishing a GIS model for the
analysis of shipwreck distributions in South Australian coastal waters.
The
project is designed using ArcGIS 9 software produced by the Environmental
38
Systems Research Institute (ESRI).
The first step in producing a coherent
approach requires identifying the aim of a project, and then a creating a database
from the selected datasets.
The methodology in this section specifically
concentrates on methods for collecting the archaeological and environmental
datasets for this project.
The establishment of databases
The establishment of databases is the most difficult part of any project in terms of
its significance and time involved; further, the outcome of the project is dependent
on the quality of data and its accuracy and entirety (Savage 1990, p. 27). The
general process of establishing a database contains the following steps:
ⅰ). Database Planning
Identification of the selected spatial data based on the requirement for analysis
Determination of attributes and variables for analysis
Setting the range and boundary of analysis
Selection of the geo-coordination used for the analysis
ⅱ). Data Collection
Digitizing the data to a proper format
Confirmation and modification of error
ⅲ). Data Administration
Confirmation of geo-coordination
Combination of adjacent layers
Table 3-1-1. Procedure for the establishment of a database in GIS analysis
39
Digital datasets from web-based databases
A limitation of GIS analysis is the insufficiency of available digital datasets for
the purpose of research (Kvamme 1995, pp. 5-6).
One solution is to employ
accessible datasets that have already been produced by organisations (Zubrow &
Green 1990).
An issue of consideration for this approach might lie in the
inherent data quality. Regardless of the importance of uniformity in criterion for
data selection, there is still a gap of permissible ranges among people who carry
out spatial analysis.
In particular, spatial analysis in archaeology frequently
requires setting up a strict dataset of scales and ranges with accuracy that should
be equal to measurements taken in the field.
For example, the value of contours
and grids for a digital elevation model (DEM) is likely to show better qualities in
urban rather than rural areas because the use of GIS is widespread in city planning
and its associated environmental problems.
In comparison to the development
and utilisation of GIS in other disciplines, and to the rapid increase in case studies
using GIS and its products in business and government applications,
archaeologists are still struggling to obtain suitable datasets for their own studies
(Kaneda et al. 2001, p. 18).
40
Nevertheless, it is notable that the development of the Internet contributes to
shared digital datasets, which was forecasted as the globalisation of information of
“corporate metadata” (Kvamme 1999).
Indeed, a large number of web resources
provided for GIS utilisation enables researchers to access extensive digitised
information about geography, topography, hydrography, and cultural phenomenon
(Appendix I).
The Australian Spatial Data Directory (ASDD) website is an
excellent example of the development of open resources of digital data on the
Internet.
The ASDD website was launched in 1988 by the Australian New
Zealand Land Information Council (ANZLIC), an intergovernmental organisation
for the collection, management and use of spatial information.
The ASDD
provides search interfaces, “to discover geospatial dataset descriptions (metadata)
throughout Australia” (Australian New Zealand Land Information Council 2006).
The website is designed for users to search and download substantial amounts of
spatial datasets and geographic information, which are maintained by contributing
organisations.
These organisations include South Australian Spatial Information
Directory, Geoscience Australia, and Australian Hydrographic Service; each of
these has completed geographic surveys beneficial for maritime archaeology and
provides useful spatial information.
41
The website of ASDD produces brief dataset descriptions that explain certain sets
of geospatial data and sometimes the actual raster or vector data.
Regarding
shipwreck data, the ASDD only demonstrates basic information about the context
of the South Australian shipwreck database, and the following information is
provided on the website: Citation; Description; Data currency; Dataset status;
Access; Data quality; Contact information; Metadata information; and Additional
metadata (ASDD).
The Department for Environment and Heritage (DEH), South Australia, maintains
custody and jurisdiction over shipwreck data. A description provides an abstract
which explains that although the data interact with the Commonwealth Historic
Shipwrecks Act 1976 and the South Australian Historic Shipwrecks Act 1981, the
database also includes those shipwrecks that have not been declared under either
of these Acts.
The available information held in the database covers general
information about shipwrecks including historical background, management
policies, location, description, conservation, bibliographic references and
associated artefacts.
Regarding issues of data quality, the contents imply that
potential revisions about historical facts from sources like historic newspapers and
42
physical information from sources like Global Positioning System (GPS) will
occur on the basis of more accurate and reliable data acquisition. The category
of “Access” illustrates information on the type of digital data format and the data
accessibility.
Although access to the database is restricted, the existence of
spatial information about the locations of shipwrecks in South Australia can be
identified. Consequently, the ASDD spatial database provides valid datasets for
this study, and one source of data collection will depend on web-based database.
For data management, the substance of data based either on fieldwork or previous
research is influential to the outcome of analyses.
One of the issues lies in the
fact that since GIS can produce a visual product and a plausible result without
rigour and insightful data analysis, the quality of data and manner of data
collection should be clearly presented in GIS analysis (Kvamme 1999, pp.
161-162; Kaneda et al. 2001, p. 18).
The entire process of data collection related
with the issue of its quality, accuracy, reliability and completeness is a main
concern in this research.
43
3-2.
GIS models of predictive modeling
General ideas regarding the analytical techniques and capabilities of GIS
modeling are summarised in Kvamme’s paper titled “Recent Directions and
Development on Geographic Information Systems” in the Journal of
Archaeological Research (Kvamme 1999). According to precursory insights and
case studies (Allen 1990; Warren 1990a; Warren 1990b; Kvamme 1995; Boyd et
al. 1996; Mardry et al. 2006) locational analysis at the regional level, predictive
modeling and risk assessment analysis focusing on the likelihood of feature
allocations are regarded as useful models for assessing local area analysis. From
this perspective, it is inferred that the incorporation of these models will produce
an analytical GIS model for evaluating the location of shipwrecks in South
Australian waters.
The concept of predictive modeling suggests that archaeologists can view past
human use of space with focus on environmental factors, and that this
information
can
determine
patterns
of
site
and
artefact
locations.
“Archaeological predictive modeling is one of the earliest applications made
44
possible by GIS” (Kvamme 1999, p. 156), and Warren (1990a) states that the first
contribution of predictive modeling was carried out by Kenneth L. Kvamme
through the adaptation of logistic regression.
The gradual development of
empirical predictive modeling occurred in the US within the scheme of cultural
resource management (CRM) and eventually impacted landscape archaeology in
the UK. However, the current direction of GIS in archaeology is apparently
moving away from the analytical domain of predictive modeling and towards
database management and development, and displaying cultural resources as a
tool for the public (Kvamme 1999, p. 162; Ghobadi & Tsumura 2006). As
previously mentioned, this perspective is also dominant in maritime archaeology
where certain achievements result from the application of GIS to underwater
cultural heritage management, rather than deductive frameworks to analyse
unknown shipwreck locations (Mather & Watts 2002, p. 693).
The negative
aspect of predictive modeling emphasises the incompleteness in deductive
methods which cannot lead to the location of archaeological sites.
Many
restrictions related to determining factors for site distribution, including
archaeological, environmental, behavioural, and technical aspects in the modeling
process, cause difficulties in pursuing models of archaeological site location.
45
Issues related to these restrictions include data bias from subjective observation,
inaccuracy of survey data, environmental differences between past and present,
idiosyncratic human behaviour, and the existence of many exceptional sites.
Despite the apparent bias, innovative and indomitable approaches of
archaeological predictive modeling are presented in a recent publication “GIS and
Archaeological
Site
Location
Modeling”
(Mehrer
&
Wescott
2006).
Acknowledging many limitations in archaeological site modeling, the book
provides insight into enhancing the validity of analytical frameworks.
With
regard to defects in the deductive approach of predictive modeling, for example,
the authors suggest that one of the key concepts for a better understanding of
predictive modeling is to discard the idea of an extreme dichotomy between
deductive models and correlative models (Kvamme 2006, pp. 12-13).
The
concept of predictive modeling is not bound to only deductive model based
entirely on theory.
The correlative model based on empirical observation of
archaeological phenomena will also allow for evolving if a location contains
archaeological material or not.
A deductive approach and a correlative approach
are coincident where theory and data are not universally distinguished: “…data is
46
collected within a theoretical context, …while theories are generally based to
some extent on empirical observation” (Wheatley & Gillings 2002, p. 166).
Models focusing on correlation have long been accepted in other disciplines;
biologists for example, have accumulated data lists of variables relevant to a
species distribution and or its habits.
This method has been used more in the
recent methodological development of GIS (Kvamme 2006, p. 12).
While
reexamination of archaeological predictive modeling is an ongoing issue, Mehrer
and Wescott (2006) demonstrate that possible improvement can arise through the
adaptation of new variables and algorithms.
From this view, the idea of
predictive modeling will be one of the methods employed for evaluating the
spatial meaning of shipwrecks in South Australian coastal waters.
Two case studies integrating theory to practice
A case study of Cultural Resource Management (CRM) by the North Carolina
Department of Transportation demonstrates an example of an integrated model
between the adaptation of GIS
predictive modeling and the assessment of
cultural space (Mardry et al. 2006). The project was designed in several steps,
47
and one of the purposes for using GIS predictive modeling was to aid in the
selection of preferred highway alternatives based on an estimation of potential
impacts on archaeological sites.
The project structure initially required
collecting archaeological site data and environmental information from one region
of North Carolina.
These datasets were then integrated into a database for
statistical manipulation of archaeological and environmental data.
Although
these steps were aimed at predicting site location, cultural heritage managers
recognised a lack of appropriate validation procedures for archaeological
predictive modeling.
Furthermore, they referred to the principle that “the
generally accepted method of archaeological site predictive modeling in use today
is an inductive modeling procedure using logistic regression analysis techniques”
(Mardry et al. 2006, p. 326).
Following this principle, they adopted an alterative
method in which the datasets of known archaeological sites are initially analysed
with various environmental factors for the same area.
After determination of
statistical relationships between the data, areas displaying similar patterns of
occurrence are identified as, “predicted to be of higher likelihood to contain
similar archaeological remains” (Mardry et al. 2006, pp. 326-327).
This
inductive predictive modeling seems to be an antipodal approach to deductive
48
predictive modeling at first sight; however, both have a coincidental aspect in the
fact that their goals are equally aimed at identifying sensitive areas containing
archaeological remains.
As mentioned in section 2.5, in terms of risk avoidance in potential disturbance to
unknown archaeological remains, a similar approach to coastal management was
employed in northern New South Wales (NSW) to predict unknown shipwrecks in
the River Richmond mouth (Boyd et al. 1996). In this model, an inductive
approach was adopted to produce an estimate of probable locations of buried
shipwrecks.
These were calculated from multiple variables consisting of
historical distribution, land surface elevation, and geological substrate.
The
estimated numbers indicating the probability of potentially buried shipwrecks
were treated as absolute factors.
These were then multiplied by actual risk
factors such as land usage for the Richmond River mouth.
Consequently, the
risk to space containing potential buried shipwrecks was equalised by the
manipulation of these factors. In this approach, the site location modeling that
was generated from inductive approaches played a significant role in area
assessment for coastal resource management.
49
Risk assessment model
A wrecking event signals the end of a vessel’s life.
Considering the relationship
between maritime space and shipwrecks, questions arise regarding the sort of
spatial factors that determine the longevity of wrecked vessels and how these can
be logically calculated.
Studies conducted in the social sciences regarding
causality between the living environment and longevity may provide some insight
and an analytical framework for the methodology.
With regard to GIS analysis,
researchers logically identify plus and minus environmental factors within an area,
and the information about their locations is transformed into several raster images
with values assigned along density, distances and vicinity (Figure 3-2-1).
50
Figure 3-1-1. The image of a risk assessment model (reproduced from Tanaka
2005, p. 102)
The layers of the raster images are overlapped, and the raster calculation of values
of assigned areas results in the total quality of one’s living environment.
For
example, according to urban sociologists (Tanaka 2005, p. 100) living
environments can impact longevity; this can be assessed through spatial analysis
concerning the location of living-related facilities with their qualitative aspects of
safety, amenity, and convenience.
In the general theory of environmental
assessment, a location model of living-related facilities evaluates the distances
51
and accessibility within living areas. For example, areas adjacent to open spaces
like parks provide comfort for living, as well as safety in terms of their important
roles as places of refuge.
The location of hospitals, their number, vicinity and
distance within a living area is directly related to the quality of life.
On the other
hand, noise and vehicle exhaust from highways and environmental pollution from
industrial plants can cause serious stress and disease.
Therefore, the close
vicinity of these facilities to living areas is regarded as a considerable negative
factor in environmental assessment.
In terms of application to wrecking events, the GIS risk assessment model
identifying plus and minus factors in maritime space can be one method for
comprehension of the historical distribution of shipwrecks in a region. In this
analytical framework, the location of shipwrecks is a focal point for maritime
spatial analysis, and various environmental variables surrounding wreck sites need
to be identified as influential factors for ship losses. Some variables within
South Australian waters and coastal areas will demonstrate positive aspects for
seafaring, but others will have negative influences.
Manipulation of these
factors leads to the spatial meaning of distribution of shipwrecks in coastal waters,
52
which can be archaeologically interpreted as outcomes of historical events in the
maritime space of South Australia.
3-3 The determination of variables for maritime spatial analysis
By adopting an integrated approach of predictive modeling and risk assessment
modeling for maritime spatial analysis, this section seeks variables that impact
wrecking events such as environmental and cultural activities. The combination
of such selected variables and the appropriate GIS model will explain the
distributional meaning of shipwrecks in South Australian coastal waters.
Although no definite guidelines for suitable variables for maritime spatial analysis
exist, potential variables can be identified by referring to the maritime history of
South Australia and to maritime archaeological research relating to shipwrecks in
South Australian waters.
53
Variables for GIS analysis in maritime archaeology
Natural factors of maritime space
Variables determining shipwreck distributions can range over hydrographic,
geographic, and human mental features in maritime space.
Analyzing the
distribution of archaeological phenomena in respect to the natural environment is
a typical approach in GIS analysis and an analytical framework for studying
specific marine environments which yield shipwreck remains needs to be
developed.
Hydrographic aspects include types of seabed, prevailing winds,
storm, tidal currents, and wave action. Highlighting hydrographic impacts is
significant when considering entire wrecking events, including pre- and
post-deposition site formation processes.
Groom and Oxley (2002, p. 51)
present comparable variables for GIS analysis between terrestrial archaeology and
maritime archaeology (Table 3-3-1).
A hydrographic variable will be adopted
to create a thematic layer in this study due to its significance.
54
Marine
Land
Bathymetry (water depth)
Elevation (land height)
Distance to established trade route
Distance to freshwater
Seabed sediment
Soil
exposure to on-shore fetch
Aspect
Seabed topology
Slope
source
Table 3-3-1. Equating land and marine variables for GIS (Groom & Oxley 2002, p.
51)
Cultural factors
Determining environmental variables that affect spatial patterns of shipwrecks is
only one aspect that must be considered; the effects of human behaviour on board
the vessel during the wrecking event must also be considered.
Decisions
regarding shipping routes and anchorage locations, among other things, are based
on the seamanship and navigational knowledge of vessel captains and may be
reflected in wreck patterns.
It is inferred that maritime space may exhibit
randomness of shipwreck distribution, and certain patterns of seafaring are
unlikely to be identified. Human behaviour during seafaring can be vulnerable
and sensible due to changing weather conditions.
Although research that
derives certain distributional pattern of shipwrecks from human behaviour and
activities has not yet evolved, the maritime spatial analysis in this study will
55
attempt to find credible evidence of human influence at the time of wrecking.
Recent developments in GIS modeling have led archaeologists to emphasise the
role of cultural landscape and the social environment, taking into consideration
the relationship between space and human activities (Kvamme 2006; Lock &
Harris 2006).
Lock & Harris (2006, p. 51) suggest, “…in determining site
locations there is a need to move beyond a sole reliance on environmental factors
and to draw upon aspects of landscape archaeology that could provide additional
determinants of site location.”
One of the key concepts demonstrated in this
statement is that exploring the meaning of archaeological distribution through GIS
spatial analysis can be compensated by the juxtaposition of diverse factors, not
depending solely on natural environment.
Other authors also acknowledge the
necessity of considering the social environment for improvement of spatial
analysis (Kvamme 2006, p .21). Kvamme (2006, p. 22) states, “Social variables
typically refer to characteristics of the human-created environment.
In complex
societies it is markets, central places, intervillage (sic) spacing, road networks,
political boundaries and like that drive uses of space.”
This statement focuses on
cultural complexity in society and space and assists in understanding spatial
56
linkage between cultural features.
Human activities in maritime cultural systems and the distribution of shipwrecks
can be related to maritime infrastructures, which are a part of the maritime
landscape.
Engagements of vessels are directly related to the existence of other
maritime features including ports that exist as vessels’ destinations, the
establishment of markers and lighthouses which direct vessels to certain routes,
and jetties that operate for loading and unloading of commodities and people.
Furthermore, behind these maritime infrastructures are certain land-based human
activities with broader economic and social systems.
Maritime infrastructures
are a relay point between vessels and land in their roles in space.
Considering
both the natural and social environment of maritime space will assist in
understanding of the location of shipwrecks.
Obtaining ideas from South Australian maritime and terrestrial history
This section will outline the maritime history of South Australia with focus on the
development of navigational aids such the establishment of lighthouses and
production of charts.
The navigational aids are regarded as human-created
57
factors that play an important role in wreck events in the waters.
A review of the
history of the South Australian coast also provides spatial and temporal scales
with this study.
It is inferred that some Indigenous groups must have been familiar with the
coastline of South Australia for many thousands of years, although the use of
watercraft by these people is regarded as limited to riverine systems (Clark 1990).
The colonisation of South Australia by European people dates back to
approximately 170 years ago, and the total length of South Australian coastline
(including offshore islands) is approximately 4,000 kilometres (Clark 1990, p.1).
Although a Dutch ship named Gulden Zeepard initially sighted the south coast of
Australia in 1627, the topographical features of the coast were not significantly
explored and recorded until the early 1800s. These early nineteenth century
European expeditions were led by Matthew Flinders in Investigator in 1802,
Nicolas Baudin in Le Géographe in 1803, and Louis Freycinet in Casuarina in
1803.
The significance of these voyages is well known and resulted in a
dramatic increase of knowledge about the coast (Sexton 1986, p. 1).
Although
delayed by war between England and France, the chart of Baudin’s expedition
58
was published in 1808. Redrawn on sixteen separate sheets by the Royal Naval
Hydrographer, the charts of Flinders’ expedition in 1814 were published (Griffin
& McCaskill 1986, p. 6). In addition to these government sponsored voyages,
British and American sealing and whaling parties also approached the South
Australian coast and contributed to knowledge of the coastline (Griffin &
McCaskill 1986, p. 8). Since their knowledge and discoveries were not reported
to authorities in many cases (Parsons 1986b, p. 4), the shared information about
seafaring seemed to depend on the results of official reports produced by the
Royal Navy during the colonial periods.
During the 1830s settlement plans led to attempts at safely navigating South
Australian waters based on the accumulated knowledge of explorers, sealers and
whalers.
People who planned to go to the new provinces of South Australia
could obtain guides describing the coastal areas through the “Australian
Directory; Volume I,” printed by the British Hydrographic Office in 1830
(Hydrographical offices 1830). The capital city of Adelaide was fixed in 1836,
and the famous Flinders’ chart entitled, “The Maritime Portion of South Australia"
(published in 1839) covered important sections of the South Australia coast, as
59
well as the precise position of the city (Jeffery 1989, p. 51).
Within a week of
official settlement in 1837 information about sailing routes and anchorage points
at Holdfast Bay was provided to captains who sailed toward South Australia.
More detailed sailing directions were issued for Port Adelaide and Port Lincoln in
1840, and a description of Gulf St Vincent and the approach to Port Adelaide were
published in 1845.
Both of the earlier sailing directions were produced by
Captain Thomas Lipson a Naval Officer and Collector of Customs in South
Australia (Lipson 1853).
Lipson’s booklet, combined with the chart compiled by
Flinders, is regarded as valuable and his guide provided the types of concerns
expressed and a possible factor for less vessels being wrecked (Jeffery 1989, p.
51).
Despite the aforementioned sailing directions, navigational aids were still
restricted during South Australia’s colonial stage.
For a newcomer with no local
knowledge, navigating the coast of South Australia was particularly difficult and
dangerous. Shipmasters at the time used incomplete charts, bearings taken of
coastal features, and the sounding lead, and there were no constructed beacons or
markers to help navigation during the daytime, and no lighthouses for nighttime
60
seafaring either (Parsons 1986b, p. 134).
Only small improvements occurred in
approaches to Port Adelaide; the ship Rapid set up a number of buoys at Port
Creek in 1838 and the lightship Ville de Bordeaux was moored off the entrance to
the port in 1848.
On the other hand, a lack of relevant navigational aids in a
large portion of the South Australian coast caused critical wrecking events,
including the brig Dart and the barque Parses stranding on the Troubridge Shoals,
off the southeast tip of Yorke Peninsula (Parsons 1986a, pp. 6-7).
A colonial government organisation to treat maritime affairs has existed since the
installation of Thomas Lipson as the first Harbor Master in 1836 (Marine &
Harbors Department, n.d.).
The colonial office provided him with the
appointment of Department of Harbor Master, which was later transferred to the
Harbour Department (Figure 3-3-1). However, the establishment of appropriate
navigation aids covering the populated areas of the South Australian coast was not
carried out until 1850 due to the absence of funds and a legal body to build and
maintain maritime infrastructures. From the 1830s to 1840s, even the issues of
maritime affairs were little dealt with by legislative council (Parsons 1986b,
p.136).
61
1836
Harbor Master &
Naval Officer
1840
Department of
Harbor Master
Harbor Department
Trinity Board
1850
Port adelaide
Harbor Trust
Coast and
Harbor Service
Marine Board
1860
1914
Local Marine
Board
Harbor Board
1920
1967
Department of
Maritime and Harbors
Figure 3-3-1. Development of South Australian Marine Board (Marine & Harbors
Department, n.d., p. iii)
In the early 1850s, the prosperity of the colony became apparent.
Lobbying and
debates for improvement of navigational aids and the establishment of ports,
jetties, and lighthouses, as well as other demands concerning the investment of
other public and individual infrastructures such as tramways, railways and better
62
urban water-supplies, increased throughout this decade (Griffin & McCaskill 1986,
p. 12).
The first progress was the establishment of the Trinity House of Port
Adelaide in 1851.
Established with assistance from English institutions, the
functions and jurisdictions of this organisation were comprehensive and took
responsibility for the main navigational aids. These included the erection and
maintenance of lighthouses, lightships, and buoys, as well as harbour pilotage and
other the navigation safety issues.
The records of the Marine and Harbors
Department state, “…from that date the following services received pay from the
corporation: pilots, steam tug, lighthouse, lightship, powder magazine” (Marine &
Harbors Department, n.d., p. ii).
As a result of the establishment of a stable and profitable organisation, the first
lighthouse was constructed in 1852 at Cape Willoughby on Kangaroo Island.
This marked the approach to the Backstairs Passage and the entry to Gulf St
Vincent.
In 1855, a lighthouse was finally placed on the Troubridege Shoal,
known until then as the “black spot”.
In order to assist the approach to
Investigator Straits, in 1858 the Trinity House decided to place a lighthouse at
Cape Borda on Kangaroo Island. Erected in the same year and operational by
63
1859, the lighthouse at Cape Northumberland in the southeast of the colony
provided aid to inter-colonial traders from Victoria, as well as overseas ships
having just passed through the rough and tumble passage of the Roaring Forties
(Parsons 1986b, pp. 136-137). The construction of these four lighthouses is
considered useful for the reduction in navigational hazards and the occurrence of
shipwrecks along the South Australian coast (Griffin & McCaskill 1986, p.16;
Gordon 1988, pp. 55-56).
With the development of Port Adelaide and the increase of coastal trading in
South Australia, several organisations dealt with marine matters during the 1850s.
Formed in 1854, the Port Adelaide Harbor Trust consisted of members from
Trinity House and was tasked with maintaining the channel and entrance to the
harbour of Port Adelaide by undertaking deepening operations to allow passage of
vessels of great tonnage.
In addition, an organisation known as the Local Marine
Board controlled the shipping office and mercantile marine matters.
The
considerable overlapping of responsibilities ended with the creation of the Marine
Board of South Australia under the Marine Board Act of 1860.
The Marine
Board became, “the Department to undertake the general superintendence of all
64
marine maters…” (Marine & Harbors Department, n.d., p. ii).
Thus, before 1860
the tasks related to maritime affairs, including navigational aids, fell under the
influence of a loose cooperation from several colonial government offices,
individual authorities, private enterprise, and a group of local businessmen known
as the Port Adelaide Council.
In the 1860s, the development of better navigational knowledge was derived from
coastal surveys undertaken by Commander John Hutchison and Frederick Howard,
and known as the Admiralty Survey of the South Australian Colony.
The
schooner HMS Beatrice engaged in coastal survey in Northern Territory of South
Australia and the colony of South Australia from 1863 to 1880 (Parsons 1986b,
p.151).
The resulting information was used to produce a series of Admiral charts
of South Australia and the new “Hydrographic Notice” published in 1877 (Marin
Board Office 1877).
In the 1880s, South Australia experienced an apparent economic growth based on
the foundation of primary products such as cooper, grain, and wool. Although
several years of economic depression occurred by 1900 (Griffin & McCaskill
65
1986, p.20), in comparison with the 1860s, the internal network of South
Australian marine activities continued to develop until 1890.
With regard to
traits of the entire coast of South Australia, dispatching farm produce was highly
dependent on shipping transportation, and the number of vessels employed
increased throughout the late 1800s (Jeffery 1989, p. 52). The installation of
thirty navigational aids comprised of fourteen lighthouses, lightships and jetty
lights by 1890 resulted from a large number of vessels using sea-lanes in the
South Australia waters (Griffin & McCaskill 1986, p.20).
Certain cultural and social dynamism occurred from 1840 to 1890 along the entire
coast of South Australia in terms of the development of navigational aids. This
period represents the focal point of this study due to the construction of
lighthouses and increasing knowledge of coastal navigation, which are considered
variables impacting wrecking events.
Insights from maritime archaeological approaches by precursors
Apart from individual site reports, maritime archaeologists have already
conducted thematic and comparative analyses of multiple shipwrecks of South
66
Australia (Jeffery 1989; Coroneos 1997; Coroneos & Mckinnon 1997). Some of
these studies provide considerable insight into factors related to the wrecks which
occurred along the South Australian coast (Jeffery 1989; Mckinnon 1993;
Coroneos 1997; Coroneos & Mckinnon 1997).
A series of regional survey
reports produced by the South Australian Heritage Branch regarding the
shipwrecks in Commonwealth waters in South Australia are considered important
because they represent cohesive approaches based on specific research themes to
multiple shipwrecks, rather than a distinctive approach to a singular site
investigation.
Therefore, through these archaeological and historical studies it is
possible to obtain the variables that should be taken into consideration in
determining the reasons for a vessel to be wrecked on the South Australian coast.
Former State Heritage Branch Officer Bill Jeffery has analysed coastal sailing
vessels wrecked in South Australian waters in terms of their constructions and
types.
In 1982, the archaeological survey carried out on the Tasmanian-built
vessel Water Witch wrecked in the River Murray, and its results provided a new
understanding of early Australian shipbuilding (Jeffery 1989, p. 51). According
to Jeffery (1989, p. 51), the results of the survey contradicted the dominant theory
67
that most Tasmanian-built vessels were constructed from local timbers, which led
to the need for further research about other vessels in South Australia. Jeffery
initially conducted historical research dealing with a selected 84 Australian-built
sailing vessels wrecked in South Australia from 1840 to 1900 (Jeffery 1989).
His research aimed at identifying typical vessel types among the shipwrecks based
on shipbuilding technologies and types of rig, categorised as cutters, ketches, and
schooners. Part of this research also considered the issues of navigational aids as
a factor that played a significant role in the wreck events (Jeffery 1989, pp. 51-52).
Consequently, he emphasised that comparative and typological analysis of the
vessels including design, suitability, construction, and capability of maintenance
needed to be integrated with other environmental factors; this correlative approach
can reveal the causes of wrecks in South Australian waters.
Maritime archaeological surveys have been conducted at the wreck sites in
Encounter Bay, Backstairs Passage, Investigator Strait, and the lower Yorke
Peninsula.
Since 1986 the State Heritage Branch of South Australia has
conducted archaeological surveys with the purpose of documenting the
shipwrecks in South Australia.
Known as the ‘Commonwealth Historic
68
Shipwrecks Program of South Australia,’ the project is based on regional surveys,
but does not include a large portion of Spencer Gulf and Gulf St. Vincent. The
survey areas are divided into three main regions: south-east, central and western.
The central region is sub-divided into the Investigator Strait and Backstairs
Passage.
The first regional report contained shipwreck data from the south-east
region and was published in 1990 (Clark 1990). Next the regional survey of the
shipwrecks of Kangaroo Island was produced by McKinnon in 1993 (McKinnon
1993).
Cosmos Coroneos then published the results of two surveys in the central
region (Coroneos 1997).
Clark produced a database of the 52 shipwreck sites located along the south-east
coast ranging from the mouth of River Murray to the border of Victoria (Clark
1990).
Most of the shipwreck data in this report derives from historical research,
and recommendations for the future management of the sites are proposed for the
wrecks in this area, as well as unidentified wrecks, Australian-built vessels, and
shipwrecks associated with Chinese immigrants bound for the Victorian gold
fields.
Each of these reports provides fundamental information about the
topographical features, historical backgrounds of the regions and shipwreck data,
69
and according to McKinnon are intended to, “…provide a basis for future research
management of the resources” (McKinnon 1993, p. xiii).
In comparison to Clark’s report, McKinnon concentrated on dealing with the
archaeological record and material remains of the vessels wrecked on Kangaroo
Island.
McKinnon reasoned that an archaeological survey was necessary to
supplement the publication Shipwrecks of Kangaroo Island, which provides
historical details of the vessels wrecked on Kangaroo Island (Chapman 1976 cited
in McKinnon 1993, p. xiii). McKinnon (1993) analysed 31 shipwrecks around
the island. It is notable that the author classified these thirty-one vessels based
on five analytical perspectives: Australian-built coastal vessels; international
cargo vessel; iron ships and steamers; foreign-built local traders; and recreational
vessels. In the last two chapters, McKinnon suggests that the outcome of this
survey is intended for use as a tool for underwater cultural resource management
of Kangaroo Island’s submerged resources.
Both of these works provide
coherent shipwreck data within a regional framework and are useful for spatial
analysis.
70
Cosmos Coroneos (1997) employed a more thematic approach when reporting the
results of surveys of Backstairs Passage, the Lower Yorke Peninsula and
Investigator Strait.
The aim of these reports is to reveal, “…the reasons why
vessels were lost and how the shipwrecks fit into the trade dynamics of each
particular area, and of South Australia as a whole” (Coroneos 1997, p. xi). With
regard to the meaning of this examination, Coroneos (1997, p. xi) reports that,
“…most publications are likely to look at shipwrecks as a single and an isolated
event, as well as considering them to be unrelated to each other and out of context
with the world they operated in.”
In his report on the 28 shipwrecks of
Backstairs Passage and Encounter Bay he adopts a statistical approach to analyse
the number and distribution of shipwrecks in correlation to geographic, economic,
and chronological factors in the region. Based on a regional perspective, the
outcome of the research leads to causality between the loss of vessels and the
unsuitable conditions of the harbours at Encounter Bay.
Using the same method,
the 26 vessels wrecked on the Investigator Strait and the lower Yorke Peninsula
were analysed in his second report. The comprehensive and thematic analysis
provided by these surveys produced valuable statistical relationships between the
multiple shipwrecks in the regions and other social environments; this data will be
71
applied to this spatial analysis of the entire South Australian coast.
The purpose this section is to obtain insight into the social environments of South
Australian waters based on the implementation of relevant spatial analysis
composed of not only natural, but also cultural perspectives.
Through the study
of the maritime history of South Australia and the existing maritime
archaeological research, the meaning of the distribution of the shipwrecks will be
interpreted in relation to natural influences and variables from human actions and
activities.
72
4 The analysis of maritime space of South Australia
The following chapter will outline the content of spatial analysis in this study.
The details of processing geo-spatial data including the database of shipwrecks,
the digitalized historical charts, and bathymetric data are presented in the first
section.
The second section clarifies the chronological aspect of wrecking data
and analyses of the spatial correlation between shipwrecks and lighthouses
through the combination of statistical and map-based approaches.
The third
section pursues interrelation between the location of shipwrecks and water depth.
4-1 Interpreting maritime space
The data of the shipwrecks in South Australian coastal waters
According to the Commonwealth there are a total of 702 shipwrecks in South
Australia; of those only 140 have been located (Kenderdine 1997, pp. 10-11).
Although information of these individual South Australian shipwrecks is available
through printed material such as site reports and popular literature, spatial analysis
requires a distinctive dataset covering multiple wrecks rather than a single one
73
describing each individual shipwreck.
This section
presents the content of
the dataset obtained for this study.
The Society for Underwater Historical Research (SUHR) collected details of
vessels lost in the South Australian waters before 1900 and published them in the
booklet “Society for Underwater Historical Research: South Australian
Shipwrecks 1800-1899” (Temme 1975).
The database was derived from
historical archives and presents information about the shipwrecks including: date
of wrecking, name of vessel, rig, tonnage, hull materials, and approximate
wrecking location.
While the dataset provides fundamental information
regarding 191 of the vessels wrecked in South Australian waters, the information
provided about their locations is not suitable for GIS analysis.
Shipwreck
databases need to contain certain spatial information which includes geographic
positions based on easting and northing values.
74
Figure 4-1-1. Locations of wrecking 1800 – 1899 (SHUR)
The exact location of shipwrecks is sensitive in terms of their protection and
management.
The policies regarding public access and site preservation are a
75
controversial issue in underwater cultural resource management (Hannahs 2003).
In the case of South Australia, the Department for Environment and Heritage
(DEH) contributes locational information of almost all known shipwrecks in
South Australia to a web-based atlas known as the “Atlas of South Australia”
(Government of South Australia, 2000).
To demonstrate various geographic
information digitally, the website integrates data into the GIS based-map system.
The original database (including the South Australian Shipwreck Database) to
operate this system must contain longitude and latitude information, but the
website
“Atlas of South Australia” does not supply the database itself or direct
coordinate information of the location of the shipwrecks. The Australian Spatial
Data Directory describes that the accessibility of the database regarding the South
Australian shipwrecks is strictly managed by the DEH.
The shipwreck database provided by the DEH consists of a dataset file of South
Australian shipwrecks and a “shape file,” which is a file format used for ESRI
products. These data files co-operate on the ArcGIS 9 system to create a layer
containing shipwreck data on the spatial map.
There are several different data
columns in the shipwrecks database including historical and archaeological data.
76
The historical data fields include: name of shipwreck, type of vessel, description
of rig, description of hull, tonnage, length, build date, constructed port, date of
loss, location of loss, and cause of loss. The archaeological data include: wreck
location, inspection information, protected status, agency jurisdiction, and
description of wrecking region. It is important to note that a possible limitation
in the description of the shipwreck data could be the lack consistent details for
each wreck.
For example, the information may not reflect the original
construction date because many vessels have several registration records resulting
from structural alterations made during their careers.
The “shape file” is the most popular file format for ESRI products and is vector
type data consisting of point data, line data, or polygon data. The shape file uses
dot features to describe spatial information of shipwrecks on the spatial layer.
With regard to locational accuracy of these dots, the description of the dataset
states, “Accuracy can range from +/- 5 metres when differential GPS readings are
used up to +/- 5 kms when values were derived from charts” (Australian New
Zealand Land Information Council, 2001, see also Appendix II).
From the
viewpoint of archaeological measurement standards, the quality of locational
77
information must be improved.
The database provides information about on 751
vessels wrecked in South Australian waters, and out of those only 149 have been
located and inspected. This indicates that eighty percent of the shipwrecks in
South Australia have imprecise locational information.
Although 751 vessels are
recorded in the government shipwreck database, GIS spatial analysis in this study
focuses on 218 shipwrecks that are dated from 1837 to 1899 in loss date.
The data of navigational aids in South Australian waters
This GIS analysis of maritime space pursues correlation between the cause of
vessels lost and the development of navigational aids. A total of fifty-three
navigational lighthouses exist along the South Australian coast at present
(Ibbotson 2000). However, by the end of the nineteenth century only fourteen
lighthouses assisted sailing vessels and steamships.
Some of those same
lighthouses are still operational, while others have been relocated to different
places in the twentieth century.
Technical developments in the illumination of
Australian lighthouses have also occurred throughout the centuries (Gordon 1988,
pp. 189-206), and the current lighthouses provide greater visibility than the early
lighthouses.
78
In order to assess the correlation between causes of vessels lost and the
development of navigational lights in maritime space, this GIS analysis treats the
location of these lighthouses and the range of light visibility as influential
variables.
The analytical concept highlights differences in the number of
shipwrecks before and after construction of lighthouses in specific areas to assess
the actual effect they had as aids to navigation in South Australian waters in the
end of the nineteenth century.
To achieve this it is necessary to identify the
range of light visibilities produced by these nineteenth century lighthouses.
79
Figure 4-1-2. Admiralty Charts: Australia. South coast. Gulf of St. Vincent and
Spencer 1855 – 1865 (The State Library of South Australia).
80
Figure 4-1-3. Admiralty Charts: South Australia – St. Vincent and Spencer Gulfs
1863 – 1918 (The State Library of South Australia).
81
Figure 4-1-4. Lighthouse map of the province of South Australia 1883 (The State
Library of South Australia).
82
A series of Admiralty Charts produced by British Hydrographic Office provides
information relating to the visible arc of lights from lighthouses and lightships
operating in South Australia (Figure 4-1-2 & 4-1-3).
Furthermore, the Marine
Board of South Australia published a chart entitled “Lighthouse Map of the
Province of South Australia” in 1882 which provided details of the thirty
navigational lights located along the South Australian coast, as well as
distributional information of lighthouses, lightships, and lights placed on jetties
(Figure 4-1-4).
Based on this data, this analysis assesses the influence of the
human created-environment in maritime space by questioning exactly how the
establishment of navigational lights impacted wrecking events.
The data of natural environments in South Australian waters
It is inferred that many types of natural environmental variables in maritime space
can cause vessels to wreck.
This study manipulates spatial correlations between
bathymetric data and the distribution of shipwrecks.
The digital bathymetry data
was provided by the Spatial Information Systems Laboratory of the School of
Geography, Population and Environmental Management at Flinders University.
The original datasets were produced by Geoscience Australia, a government
83
agency which has collected bathymetric data around the Australian coast line
since 1963 (Petkovic & Buchanan 2002, p. 2). The digital datasets were created
using information from various sources, including spot depth data recorded by the
Australian Hydrographic Services, Royal Australian Navy (Petkovic & Buchanan
2002, p. 8).
The bathymetric data held by the Spatial Information Systems
Laboratory covers the central waters of South Australia including Spencer Gulf, St.
Vincent Gulf, and Kangaroo Island.
Whereas there is a certain difficulty
regarding issues of bathymetric data collection, the quality of this data is
equivalent to bathymetric data provided on navigation charts.
Intuitively, it is
acknowledged that shipwrecks might be allocated in the shallow waters.
However, this presumption should be objectively assessed based on this
geographic data.
Manipulation process of spatial data
ArcGIS 9 is composed of multiple programmes; the core applications used for this
study are ArcCatalog, ArcMap, and ArcToolbox.
ArcCatalog allows users to
administrate the collected data, and in this study this application is utilised for
confirmation of the geographic information of the datasets.
84
ArcMap is the
application to conduct mapping, spatial analysis, and output the result of analysis.
Although ArcToolbox provides various functions to process geographic
information, it is only used for explicating spatial data in this study.
For details
about each of these applications refer to the ESRI website (http://www.esri.com/).
Geographic coordinate system and projected coordinate system
Datasets for GIS analysis should normally include spatial information.
The
ArcGIS programme adopts the “geographic coordinate system” and the “projected
coordinate system” to identify the location of objects in space.
The most
universal method of sharing spatial information about an object is the application
of latitude and longitude values through geographic coordinate system.
Various
geographic coordinate systems are employed at regional, state, and global levels,
and the adaptation of different geographic coordinate systems results in different
positioning of northing and easting values. Latitude and longitude values are
determined based on an associated geodetic system such as the World Geodetic
System 1984 (WGS 84).
This is one of the global standard geodetic systems and
is commonly used for GPS data.
The same point and spatial information on the
earth’s surface can be represented by different latitude and longitude, depending
85
on the geographic coordinate system employed.
When producing a GIS
distribution map based on latitude and longitude values, it is necessary to employ
a consistent geographical coordinate system.
The ArcCatalog function allows
the user to confirm information from the geographic coordinate system used to
create spatial datasets, or apply a new coordinate system to the datasets. This is
because the spatial datasets manipulated in the ArcMap programme need a
consistent geographic coordinate system. The shipwreck dataset from the DHE is
based on the following geographic coordinate systems (Table 4-1-1):
Horizontal coordinate system
Projected coordinate system name: GDA_1994_South_Australia_Lambert
Geographic coordinate system name: GCS_GDA_1994
Table. 4-1-1. The geographic coordinate system adopted in the datasets of the
South Australian shipwrecks.
The Geocentric Datum of Australia 1994 (GDA 94) is a coordinate for Australia
that is compatible with coordinates produced by GPS (WGS84).
It is suggested
that, “GDA94 and WGS84 coordinates can be considered the same and no
transformation is required” (Inter governmental Committee on Surveying and
86
Mapping 2001, p. 1). On the other hand, it is important to note that 10 cm
inaccuracy will occur between two geographic coordinate systems expressing
same point. All geographic datasets used for this study are based on the
geographic coordinate systems of GDA94 or WGS84.
Although the ArcGIS programme encourages the use of consistent geographic
coordinate system when setting spatial datasets, once the original datasets have
been established ArcMap, can use its own algorithm to project the data based on
different coordinate systems.
Since the object of the spatial analysis for this
study, ranges over nearly the entire portion of South Australian waters, the quality
of the latitude and longitude data produced by GDA94 and WGS84 are considered
consistent values.
After latitude and longitude values have been determined according to the
geographic coordinate system and the object locations have been identified, the
locational information must be projected on a two-dimensional screen as a digital
image.
Since the earth is a sphere, a map projection (like the Mercator
projection) must be employed to describe two-dimensional space.
87
ArcCatalog
shows that the locations of shipwrecks are projected based on the Lambert
Conformal Conic (LCC) projection, which compared to the Mercator projection
excels in the correction of the distortion of the area.
Due to the spatial
information of the shipwreck datasets based on GDA94, this is adopted as a basic
geographic coordinate system on the data frame of the ArcMap in this study.
Georeferencing
The spatial datasets manipulated as geographic data layers on the ArcMap
programme must include spatial reference information based on a geographic
coordinate system or a projected coordinate system. However, not all raster
images include relevant spatial reference information.
For example scanned
maps and charts often do not include such information, and even some aerial
photographs and satellite images, which normally encompass spatial information,
have inaccurate coordinate information.
“Georeferencing” is an ArcMap
function used to incorporate these sorts of raster data with other spatial data in the
ArcGIS programme.
This is regarded as an “overlay” function meaning that a
raster data layer can be combined with a vector data layer that includes relevant
geographic coordinate system.
88
The following historic navigational charts were obtained from the South Australia
State Library’s map collection and used in this study (Table. 4-1-2):
The maritime portion of South Australia from the survey of Captn. Flinders & of
Col. Light Survey. Genl.
/ by John Arrosmith Feby. 5th. 1839
Flinders chart upon an enlarged scale
/ resurveyed and corrected by Thomas Lipson, Esq., 1851
Australia. South coast. Gulf of St. Vincent and Spencer: surveyed in 1802 by
/ compiled, drawn & engraved by J. Arrowsmith 1855 - 1865
South Australia – St. Vincent and Spencer Gulfs
/ surveyed by Commr. J. Hutchinson R.N. and Staff Cmmr. F. Howard 1863 -1918
Lighthouse map of the province of South Australia 1883
/ Published by order of the Marine Board of South Thos. N. Stephens. Secretary. 1882
Table 4-1-2. Historical charts obtained from the State Library of South Australia
for the purpose of analyzing navigational aids
By comparing these charts ranging from the 1830s to 1880s, it is possible to
obtain a general idea about the gradual development of navigational systems
along South Australian coast.
In particular, three of them were selected and used
as digital datasets for georeferencing. The three charts include: a). Australia.
South coast. Gulf of St Vincent and Spencer; b). South Australia – St. Vincent and
Spencer Gulfs; c). Lighthouse map of the province of South Australia 1883
(Figure 4-1-2, 4-1-3, & 4-1-4).
89
The general procedure for implementing the georeferencing function is to link
several datum points of vector data to several geographical features of raster data
in the data frame of the ArcMap. The digital dataset of the Coastline of South
Australia held by the Corporate Database of the Spatial Information Systems
Laboratory at Flinders University is the underling vector data that provides certain
datum points for conducting georeferencing in this study (Figure 4-1-5).
The
historic charts that cover the central waters of South Australia consisting of
Spencer Gulf, St. Vincent Gulf, and Kangaroo Island are overlapped onto this data.
The initial datum points are established: 1.) at the tip of on the east side of the
Kangaroo Island, 2.) at the tip of Cape Wiles on the top of the Eyre Peninsula, 3.)
at the tip of Point Lowly.
These points on both the vector data of the South
Australian coastline and the digitised navigational charts are linked with each
other through ArcMap (Figure 4-1-6).
Consequently, on the data frame of
ArcMap a new thematic layer comprised of the datasets of the shipwrecks and
historical navigational charts is displayed for assessing the correlation between the
cause of wrecking and the development of navigational aids in South Australian
waters.
90
Figure 4-1-5. The vector data of coastline of South Australia ((provided by the
School of the School of Geography, Population and Environmental Management,
Flinders University).
91
Point Lowly
Cape Wiles
Cape St Albans
Figure 4-1-6. The model of Georeferencing in the ArcMap programme
92
4-2 Spatial analysis
Statistical data from spatial data
Vessels operating in South Australian waters
Studying space in archaeology means analyzing spatial and temporal data of
objects.
A map-based approach does not necessarily excel in analyzing the
chronological changes of shipwreck data, but the use of statistical data focusing
on these changes may provide insight into this map-based approach in order to
understand the distribution of the shipwrecks. The spatial analysis of this study
examines what historical developments occurred in the maritime space of South
Australia.
It accomplishes this by using chronological information based on
statistical analysis from the shipwreck database and reviewing map-based
information regarding the distribution of shipwrecks.
A map-based spatial analysis of South Australian waters will show different
distributions of shipwrecks for each decade of the middle and latter of the
nineteenth century.
Taking into consideration the chronological change in the
total number of vessels engaged in activities in South Australian waters
throughout the second half of the nineteenth century, this spatial analysis provides
more insight into the correlation between the distribution of the wrecks and the
93
development of navigation aids. This concept is represented in the following
vessels & shipwrecks
The simple correlation
of the number of
diagram (Figure 4-2-1):
Vessels engaged in
operation
X Shipwrecks
The passage of time
Figure 4-2-1. Conceptual framework about the spatial analysis of maritime space
based on the distribution of shipwrecks.
In this graph “X” represents the natural, social, and human environmental factors
impacting the loss of vessels.
Each of these played a significant role in
determining the distance between the total number of vessels operating in coastal
waters and the number that wrecked.
The identification of these factors is the
fundamental purpose of this research through the study of the role of the
navigational aids.
94
Chronological distribution of shipwrecks
80
60
The number of
shipwrecks on the SA
coast
The number of
Australian-built
coastal vessels
40
20
0
0
18
9
0
18
8
0
18
7
0
18
6
0
18
5
0
18
4
0
18
3
0
18
2
18
1
18
0
1
0
Figure 4-2-2. Comparative graph regarding the historical changes of shipwrecks
in South Australian waters (original data from shipwreck database administrated
by the Department of Environment and Heritage, South Australia).
The above line graph was produced from data provided in the DEH shipwreck
database in order to obtain insight into the chronological changes of vessels
operating along the South Australian coast in the nineteenth century.
The
establishment of the shipwreck database is an ongoing project and contains
information about the construction dates and dates of the loss of these vessels.
Therefore, this graph does not directly express all of the features of vessels
operating along the South Australian coast at the time. Regarding the issue of
the number of vessels working in South Australia at the time, Bill Jeffery (1989, p.
95
52) suggests, “The number of vessels using the South Australian coast from 1840
to 1900 increased”.
He also mentions the result of Ronald Parsons’ research
(Parsons 1983) that identified the 1880s and the 1890s as the highest usage of
ketches in South Australia (Jeffery 1989, p. 52). Looking at the graph, however,
it is valid to say that there are remarkable concentrations of shipwrecks in the
1850s and the 1870s, which indicates that the usage of vessels increased in each
different decade in the nineteenth century.
When referring to the chronological
data of navigational lights on the South Australian coast, in particular the
construction date of lighthouses in the 1850s and the 1870s, there is an increase in
concentration as well (Figure 4-2-3).
Construction date of navaigational lights on the SA coast
15
Navigational
lightships &
lighthouses
Lighthouses
10
5
0
1830 1840 1850 1860 1870 1880 1890
Figure 4-2-3. Construction date of navigational lights in South Australia.
96
The perspectives from construction dates of navigational lights
On the historical chart “Lighthouse map of the province of South Australia 1883,”
the Marine Board of South Australia showed the location of thirty navigational
lights.
The construction dates of each light were obtained from previous
research and are included in the table below (Gordon 1988 pp. 54-56, pp.
165-171; Parsons 1986b pp. 134-149).
Name of Lighthouse
Cape Willoughby
Troubridge
Cape Borda
Cape Northumberland
Glenelg (Ship)
Glenelg Jetty
Semaphore Jetty
Victor Harbor
Tipara
Meningie Jetty
Port Adelaide
Milang Jetty
Moonta Jetty
Cape Jervis
Cape Jaffa
Edithburgh Jetty
Port Wakefield Wharf
Ardrossan Jetty
Althorpe Island
Point Malcolm
Rivoli Bay, Penguin I.
Germein Bay (Ship)
Germein Bay Jetty
Wallaroo Jetty
Kingston Jetty
Rivoli Bay Jetty
Port Victoria Jetty
Corny Point
Lowly Point
Cape Banks
Date of construction/lit
1852
1855 (lit 56, upgrade in 1882)
1858
1858 (lit 59, improved 1882)
1859
1859
1860
1864, 1875
1866, 1877
1867
1867 (1873 &1977)
1869
1871
1871
1872 (for Amendment 1875)
1872(1879)
1873 (1872)
1876
1877-1879
1878 & 1888
1878
1879
1880
1880
1880
1880
1881/1882
1882
1882(lit 83) but light ships in there
1883
Table. 4-2-1 The construction dates of navigational lights expressed on the
“Lighthouse map of the province of South Australia 1883”.
97
While the South Australia Marine Board continuously established navigational
lights throughout the latter half of the 1800s, as mentioned before, the highest
concentration of construction occurred between the 1850s and 1870s.
The
construction of a lighthouse must have been the greatest campaign in the activities
of the Marine Board, and the reason for establishing lighthouses derived from the
need to assist vessels engaged in shipping operations.
The construction of a
lighthouse deals with not only intrastate, but also interstate issues.
The reason
for establishing navigational aids such as lighthouses may not have been to
support local ships used in the early colonial era as a chief means for
transportation and trade between local markets.
Thomas Lipson’s decision to
establish the first lighthouses on the eastern end of Kangaroo Island at Cape
Willoughby seems to derive from concerns about interstate communication and
transportation with Victoria, rather than considerations for existing intrastate trade
within South Australia.
Parsons points out that despite the fact that South
Australian settlers repetitively insisted upon the need for a lighthouse on or near
Troubridge Shoals, the first lighthouse on the South Australian coast was placed at
Cape Willoughby to create a safe route from the Victoria gold fields and for mail
steamers (Parsons 1986b, pp. 136-137).
98
The role of lighthouses is distinguished
from that of lightships or signal lights placed on the tip of jetties, since the latter
played a more important role in the transportation and navigation of local people
and products.
Issues regarding the establishment of lights for navigation purposes resulted in
several conferences in the early Australian colonies.
In 1856 the first
inter-colonial conference took place in Melbourne to discuss the erection and
maintenance of lighthouses; however, it failed to establish a joint commission
responsible for lighthouses (Gordon 1988, pp. 65-66).
Although some conflicts
of interest existed over sharing responsibility and costs, the second inter-colonial
conference in 1864 resulted in the establishment of highway (principally sea
lanes) lights among states (Gordon 1988, pp. 114-115).
In 1873 a third
conference was held in Sydney which produced a requirement that all state
colonial officers engaged in marine affairs review the state and condition of
coastal lights, policies for their management and maintenance, and the
establishment of more lights.
As a result of this inter-colonial agreement by the
States, South Australia was compelled to establish lights on the coastal sections:
Cape Spencer, Corny Point, Tipara Reef, Lowly Point and Penguin Island (Gordon
99
1988, pp. 115-116).
The statistical analysis of chronological data demonstrates the consistency
regarding the high concentrations of the number of the shipwrecks and the
construction of lighthouses in the 1850s and the 1870s.
With regard to the 1870s,
historical perspective provided the fact that there was a political issue in the
decision regarding the construction of lighthouses in South Australian waters.
The purpose of adopting chronological data analysis was to clarify the temporal
aspect of maritime space, and the perspective obtained from this approach will
assist in understanding spatial aspects of a map-based approach in GIS analysis.
Spatial assessment of lighthouses
A new layer on the data frame of the ArcMap programme completes the
georeferencing for overlaying two different data layers: the distributional data of
South Australian shipwrecks, and the historical navigational chart showing
locational information of thirty navigational lights.
The visibility of these
navigation lights obtained from the chart “Lighthouse map of the province of
South Australia 1883” is summarised in diagrams (Table 4-2-2 & Figure 4-2-4).
Based on this data, it is possible to assess the correlation between the range of
100
visibility in areas where lights are placed and the distribution of the shipwrecks
and ascertain impacts that these lights had on the distribution of these wrecks.
The correlation is revealed through the calculation of the number of vessels lost
within non-lit areas and those within visible lights in South Australia waters.
No.
Name of Lighthouse
DistanceVisible.(Miles)
19
11
10
28
16
17
15
30
5
22
14
20
6
18
24
12
13
29
9
21
25
2
3
4
23
26
7
8
1
27
Cape Willoughby
Troubridge
Cape Borda
Cape Northumberland
Glenelg (Ship)
Glenelg Jetty
Semaphore Jetty
Victor Harbor
Tipara
Meningie Jetty
Port Adelaide
Milang Jetty
Moonta Jetty
Cape Jervis
Cape Jaffa
Edithburgh Jetty
Port Wakefield Wharf
Ardrossan Jetty
Althorpe Island
Point Malcolm
Rivoli Bay, Penguin I.
Germein Bay (Ship)
Germein Bay Jetty
Wallaroo Jetty
Kingston Jetty
Rivoli Bay Jetty
Port Victoria Jetty
Corny Point
Lowly Point
Cape Banks
24
15
30, 17
20
3
8
4
5
30
5
16
5
4
10
18
5
6
5
25/17
10
12
8
4
4
10
5
4
14
10
20
Date of construction/lit
1852
1855 (lit 56, upgrade in 1882)
1858
1858 (lit 59, improved 1882)
1859
1859
1860
1864, 1875
1866, 1877
1867
1867 (1873 &1977)
1869
1871
1871
1872 (for Amendment 1875)
1872(1879)
1873 (1872)
1876
1877-1879
1878 & 1888
1878
1879
1880
1880
1880
1880
1881/1882
1882
1882(lit 83) but light ships in there
1883
Table 4-2-2. The visibility of navigational lights described on the chart
“Lighthouse map of the province of South Australia 1883”.
101
1 3
2
EYRE PENINSULA
4
6 YORKE PENINSULA
5
13
7
29
14
8
15
16
17 FLEURIEU PENINSULA
12
11
9
20
18
10
KAGRAROO
ISLAND
30
21
22
19
24
23
25 26
0
27
100km
28
Figure 4-2-4. The location information of the navigational lights provided by the
chart “Lighthouse map of the province of South Australia 1883”.
102
Figure 4-2-5. The distribution of 218 shipwrecks from 1837 to 1899 on
“Lighthouse map of the province of South Australia 1883”
103
The light visibility data has been described as table data (Table 4-2-2). The map
also illustrates the range of actual light visibility as arcs emanating from the centre
of the light and are adopted to provide the range of visibility from the water.
On
the data frame of the ArcMap, a layer including the location of shipwrecks has
been overlaid onto the historical chart.
The number and details of the
shipwrecks from the 1830s to the 1890s are placed within the areas of visibility
are identified (Figure 4-2-5).
Holistic interpretation of the data
According to analysis of the ArcMap, 91 of the 218 shipwrecks dating from the
1830s to the 1890s were distributed within lighted areas. The coverage of the
South Australian coast provided by the thirty navigational lights in operation at
that time produced a total visible distance of about 330 miles in good weather.
This is not much visibility considering that the total length of the South Australian
coastline is approximately 2000 miles (Parsons 1985, p. 13). By basing the
coverage on the total amount of visible distance of the navigational lights and not
on the calculation of the coverage of water, the ratio of coverage of navigational
lights for South Australia comes to 18% of the total coast. The fact that 41% of
the wrecks within South Australian waters have been located in light visibility is
104
important to note.
This spatial analysis study also deals with the individual data regarding the
number of shipwrecks located within the visible area of each navigational light.
The following table compares the difference in the number of shipwrecks in
specific areas before and after the construction of navigational lights (Table 4-2-3,
see also Appendix III):
105
The number of the shipwreck located
within lights visble area
The construction/lit date of lights
Name of lights
Before
After constructing
constructing lights
lights
Cape Willoughby
1852
0
1
Troubridge
1855 (lig 56, upgrade in 1882)
5
2
Cape Borda
1858
0
4
Cape Northumberland 1858 (lit 59, improved 1882)
4
15
Glenelg (Ship)
1859
0
2
Glenelg Jetty
1859
Semaphore Jetty
1860
1
1
Victor Harbor
1864, 1875
1
1
Tipara
1866, 1877
1
2
Meningie Jetty
1867
0
0
Port Adelaide
1867 (1873 &1977)
4
4
Milang Jetty
1869
0
0
Moonta Jetty
1871
0
0
Cape Jervis
1871
0
0
Cape Jaffa
1872 (for Amendment 1875)
2
0
(Visinity of Cape Jaffa)
8
2
Edithburgh Jetty
1872(1879)
1
0
Port Wakefield Wharf 1873 (1872)
0
1
Althorpe Island
1877-1879
0
1
Visnity of Althorpe Is
2
4
Point Malcolm
1878 & 1888
0
0
Rivoli Bay, Penguin I. 1878
2
1
Germein Bay (Ship)
1879
1
0
Germein Bay Jetty
1880
(Visnity of Germein Bay)
1
1
Wallaroo Jetty
1880
3
2
Kingston Jetty
1880
1
1
Rivoli Bay Jetty
1880
0
0
Port Victoria Jetty
1881/1882
0
0
Corny Point
1882
1
0
Lowly Point
1882(lit 83) but light ships in there
1
1
Cape Banks
1883
4
1
(Visnity Cape Banks)
1
2
44
49
Total
Table 4-2-3. The number of shipwrecks from the 1830s through the 1890s located
before and after establishing the navigational lights (the location of several
shipwrecks is covered by two navigational lights)
106
The data demonstrates that there is no remarkable change in the total number of
ships wrecking after the establishment of navigational lights.
However, this
statistical data has limitations to some degree in accuracy.
The problem with this approach lies in the determination of the time range
established before and after construction of the lights. The calculation of the
former time range is based on the period dating from the first shipwreck to the
construction of navigational lights; on the other hand, the latter calculation is
based on the period beginning with the construction of the lights and the end of
the nineteenth century. The number of shipwrecks was not calculated in the
same time range before and after construction of the lights.
The longer time
range in the latter data will feasibly result in a greater number of the shipwrecks in
the areas with lights. A better way to equate the data might be to determine the
average number of vessels lost per year within a standard time frame both before
and after the construction of the navigational lights.
Table 4-2-4 demonstrates a
feasible method for calculating the average number of wrecks per year by
comparing the number of shipwrecks before and after constructing the
navigational lights.
In the case study of Cape Northumberland, Table 4-2-3
107
initial inspection shows that a large number of vessels wrecked even after the
construction of the lighthouse, whereas taking the average number of shipwrecks
per year demonstrates that there is an equal ratio in the same time period between
before and after lighthouse construction (Table 4-2-4).
Based on this method,
the ratio of the number of the shipwrecks per year is reduced in the areas of
visibility for almost all areas.
The calculation models for the number of shipwrecks per year in specific areas before
and after establishing navigational lights & a case study of Cape Northumberland
a). Before establishment of navigational lights
Time range 1 (T1) = Establishment date of a navigational light - Frist date of a vessel loss
The average of vessel loss per year before constructing navigational lights
= The number of shipwrecks during T1 / T1
b). After establishment of navagiational lights
Time range 2 (T2) = Establishment date of the navigational light + T1
The average of vessel loss per year after constructing navigational lights
= The number of shipwrecks during T2 / T1
e.g. Cape Northumberland (lit in 1859): before lit 4 shipwrecks; after lit 15 shipwrecks
a).
7 (T1) = 1859 - 1852 (The loss date of the J. Lovett )
0.57 = 4 / 7
Before
b).
0.57 = 4 (The number of vessels wrecked from 1859 to 1866) / 7
0. 57 = 4 / 7
Table 4-2-4. An example of calculation based on the average number of vessels
wrecked on the light visible area per year
108
The only exception is the area near Althorpe Island where the number of
shipwrecks after the establishment of the lighthouse exceeded that beforehand.
This approach is also limited by the fact that the calculation can not provide any
comparative ratio in the areas that include 0 values (Table 4-2-3).
Therefore,
more relevant statistical calculations must be adopted to provide more rigorous
results.
Spatial and temporal perspectives
Apart from focusing on the number of shipwrecks in the individual areas of higher
light visibility, the ArcMap programme produces two distinct spatial data layers to
produce a holistic distribution of wrecks and navigational lights.
The spatial
data for these layers derives from Admiralty Charts for the same areas from two
different time periods.
This study allows for comparable and chronological
analysis regarding the distribution of shipwrecks and navigational lights in South
Australian waters. The first data layer is produced from the historical chart
“Australia South Coast: Gulf of St Vincent and Spencer,” which was published
from 1855 to 1865.
The second data layer is based on the chart “South Australia
– St. Vincent and Spencer Gulfs,” published in 1913. Both Admiralty Charts
109
cover the central coast of South Australia and they show the establishment of
navigational aids in different time periods.
The first chart was derived from the
hydrographic survey conducted by Captain Matthew Flinders in 1802, as well as
additional information produced by the Marine Board of South Australia. The
second chart was created based on the hydrographic survey implemented by
Commander John Hutchison and Frederick Howard from 1863 to 1973.
Obviously, there is a gap of some years between the two charts.
The location of navigation lights was extracted from digitalised charts,
georeferencing so that the data layer of the shipwreck locations is overlaid onto
the data frame of navigational charts. This results in the production of two
thematic layers which represent the historical distribution of shipwrecks and
navigational lights in the middle and late nineteenth century.
Areas that include
a high concentration of shipwrecks are complementarily emphasised through
ArcToolbox (Figure 4-2-6 & 4-2-7).
110
Figure 4-2-6. The spatial relationship between shipwrecks and navigational lights
in South Australia in the mid-nineteenth century.
111
Figure 4-2-7. The spatial relationship between shipwrecks and navigational lights
in South Australia in the late nineteenth century.
112
By comparing these two data frames, it is apparent that the distribution of
shipwrecks and navigational lights were constantly altered throughout the latter
half of the nineteenth century.
Based on this analysis, the spatial phenomena of
South Australian’s coast during this period is summarised as follows;
1.
Over time the distribution of shipwrecks and navigational lights expanded
toward the west coast of South Australia.
2. Over time the number of shipwrecks apparently declined in some areas, such
as at Encounter Bay and from Cape Jaffa to Robe, the middle section of
Limestone Coast.
3. Meanwhile, intensive areas of the shipwrecks with high densities of wrecks,
such as the section from Port Adelaide to Cape Jervis on the west coast of
the Fleurieu Peninsula, consistently existed in some throughout the latter
half of nineteenth century.
These perspectives are regarded as underlying ideas and explain the distributional
meaning of South Australian shipwrecks in the 1830s and 1890s. The alteration
and expansion of shipwrecks and navigational lights are easier to understand by
looking at the development of shipping routes throughout the latter half of the
nineteenth century. Figure 4-2-8 presents chronological changes to the main
shipping routes in the waters of South Australia at three stages in the latter half of
the nineteenth century. From 1850 to 1865, the main shipping routes advanced
to Spencer Gulf.
In addition to these main shipping routes, many smaller
113
shipping lanes were established in 1890.
The establishment of shipping
networks in the latter half of the nineteenth century resulted from the development
of transportation of colonial products, such as copper, grain, and wool. These
products were particular to South Australia, which sustained the growth of basal
industries and built an economic infrastructure (Griffin & McCaskill 1986, p. 18).
The colonial government encouraged settlers to construct jetties around the shores
of the Gulf of St Vincent and Spencer Gulf to export these products.
The
prosperity of maritime activities through the use of small vessels and the
construction of jetties resulted in an increase in the establishment of navigation
lights on the jetties and lightships.
114
1850
0
1865
150km
0
Main shipping route
150km
Main shipping route
1890
0
150km
Main shipping route
Shipping route
Figure 4-2-8. Chronological changes in shipping routes in South Australia during
the latter half of nineteenth century (reproduced from Griffin & McCaskill 1986, p.
13, p. 17, p. 21).
115
Spatial phenomenon underlying the distribution of shipwrecks
Spatial analysis also allows for inspecting the meaning of historical shipwreck
distributions based on the perspective of navigational lights regional harbours, and
ports.
Analysis of regional and historical phenomenon makes the increase and
decrease in the number of shipwrecks in specific areas clear.
For example, in
Encounter Bay the disappearance of the high concentration of shipwrecks in the
second data frame of Figure 4-2-7 indicates the decline of the number of vessels
used there during the latter half of the nineteenth century. It is clear that the
importance of ports located around Encounter Bay and the Murray River trade
diminished through time.
The use of Encounter Bay early on an anchorage is
well known and the distribution of shipwrecks proves it.
More importantly, spatial analysis shows that the continual loss of vessels is
spatially linked to the location of harbours and jetties.
For instance, Port
Adelaide shows a high concentration of shipwrecks throughout the late half of the
nineteenth century.
The west coast of the Fleurieu Peninsula saw the
establishment of several jetties, including those at Noarlunga, Willunga, Myponga,
Yankalilla, Second Valley jetty and Rapid Bay jetty.
With regard to the existence
of the large number of shipwrecks in these ports, there is a need to pursue details
116
about their backgrounds.
Nevertheless, it is pointed out that the distributional
correlation between shipwrecks and their arrival points including ports, harbours,
and jetties, is apparent. This maritime spatial phenomenon and the meaning of
the location of navigational lights must be considered.
4-3 Calculating the correlation between the location of shipwrecks and water
depth
This analysis highlights the spatial correlation between shipwrecks and water
depth.
The dataset derived from the School of Geography, Population and
Environmental Management at Flinders University covers 1422 bathymetric
points in the Spencer Gulf, St. Vincent Gulf, and around Kangaroo Island (Figure
4-3-1), and consists of only point data for recording the depth of water.
The
point data itself is unsuitable to determine changes in water depths and
topographic alterations of the seabed.
In order to correlate the location of
shipwrecks and water depths the point bathymetric data must be translated into
elevation data.
ArcToolbox can create a topographic map known as a “Digital
Elevation Model (DEM)” using three-dimensional values (X, Y, Z) in a dataset.
117
To do this a DEM of South Australian waters must be created.
Eyre Peninsula
Yorke Peninsula
Fleurieu Peninsula
Figure 4-3-1. Bathymetric data for South Australia (provided by the School of the
School of Geography, Population and Environmental Management, Flinders
University)
Several methods can be employed to create a DEM, but issues of quality of
elevation values exist with some models.
For example, by using bathymetric
point data ArcToolbox easily allows for creating a contour map and a
“Triangulated Irregular Network (TIN)” surface model, which is a solid
118
topographic map comprised of triangulations (Figure 4-3-2).
Both the contour
maps and TIN models are theoretically formed by connecting data points
containing equal elevation values. However, a limited number of data points can
result in inconsistency when connecting lines and cannot describe the gradual
change of elevation in these models.
In addition, the resultant model is a vector
type consisting of feature data defined as points, lines, and polygons.
In
comparison with raster type of data, these models also generate spatial gaps on the
surface of the digital map that do not include elevation data.
In order to
demonstrate smooth changes in elevation of the seabed, a high resolution raster
image composed of small “cells” is a preferable spatial data model.
The raster
model is also beneficial in calculating the correlation between shipwreck locations
and water depths.
119
Figure 4-3-2. TIN surface modeling demonstrating the change in elevation by
color and relief.
A function of ArcToolbox known as “Interpolation” assists in producing a
raster-based DEM.
Regarding the concept of interpolation, the ESRI web-site
<www.esri.com> explains:
“An interpolation technique that estimates cell values in a raster from a set of
sample points that have been weighted so that the farther a sampled point is from
the cell being evaluated, the less weight it has in the calculation of cell’s.”
Employing this technique for creating raster cells, it is possible to insert weighted
elevation values to spaces that do not contain actual bathymetric data.
Besides depth values, there is a need to set two additional spatial datasets when
120
implementing this interpolation.
boundary data.
These include contour data of the land and area
Since it is essential to produce continuous elevation from the
shallow waters to the land and bathymetric data itself does not contain sea level
data, the creation of a DEM requires that contour data of the land contains a “0”
metre value (Figure 4-3-3).
Furthermore, the spatial boundaries should be
established along the study area; otherwise, the spatial analysis will suffer from
overload during data processing due to an inability to deal with such a massive
amount of spatial data from both bathymetric and contour data.
For this study
bathymetric data, contour data and spatial boundaries were set to cover all of
South Australia’s gulf coast and the waters around Kangaroo Island (Figure 4-3-4).
The resulting quality of elevation values for the DEM were related to the output
cell size. By setting the smaller size of the cells, a continuous elevation can be
pursued. However, the adoption of smaller cells to obtain high resolution can
result in an overloading in data processing, depending on the capability of
computer.
121
Figure 4-3-3. A dataset comprised of quite a number of geographic contour data
that provide South Australia containing sea level value (original data provided by
the School of the School of Geography, Population and Environmental
Management, Flinders University).
122
Figure 4-3-4. Establishing the boundary of the analytical area from the
geographic feature data containing the geo-coordinate system (original data
provided by the School of the School of Geography, Population and
Environmental Management, Flinders University).
The performance of the specific computer employed determines the success of
data processing as it requires adequate memory (such as a 2.0 GHz processer with
1.0 GB RAM memory, and 3079 MB visual memory).
While such performance
capabilities do allow for the creation of a DEM of South Australia with a
resolution of 30 raster cell size, it requires more than a total of 60 hours to process
123
the data.
The quality of this data theoretically enables DEM to describe a change
in elevation within 30 metre grids. Figure 4-3-5 represents the DEM created for
South Australian waters.
Figure 4-3-5. Digital Elevation Model for the gulf coast of South Australia
(original data provided by the School of the School of Geography, Population and
Environmental Management, Flinders University)
The correlation between the distribution of shipwrecks and water depths can be
rigourously calculated through the spatial analyses of DEM and shipwreck
124
datasets using ArcToolbox. The first approach was produced using a function of
ArcToolbox known as “Raster calculation,” which employs only raster datasets.
In this process, the distribution data for shipwrecks must be transformed into
raster data. 218 vector point data were transformed to 212 cells that indicate the
location of shipwrecks in the data frame of ArcMap (some cells included a few
shipwrecks).
The outcome of the calculation of 212 shipwreck raster cells
showed 133 shipwrecks located in less than -20 metres and 9 shipwrecks located
in other depths.
While the result of this raster calculation implies that many
shipwrecks are located in shallow waters, significant inaccuracy occurred due to
inconsistency between original raster cells and resultant cells.
Due to a lack of
rigour regarding the procedure for analyzing the data, this was not considered a
pertinent approach.
The second approach is regarded as both simpler and reliable.
An ArcToolbox
function known as “Extract to Values to Point” enables the point data of
shipwrecks to add the values of depth established in the raster cell of the DEM
(Figure 4-3-7 & 4-3-8).
125
Figure 4-3-6. Specific values for depth in the DEM are extracted and added to the
shipwreck data.
126
Figure 4-3-7. Examples of extracted values regarding water depth from DEM in
the South Australian shipwreck dataset.
In the dataset of shipwrecks, the extracted values of depth from the DEM are
allocated to the data of the total 218 shipwrecks, although for 64 wrecks there are
no relevant depth values since they are out of range of the DEM.
The number of
shipwrecks are clarified by depth values and are summarised (Table 4-3-1, see
also Appendix IV).
The data demonstrates that most shipwrecks that occurred in
South Australian between 1837 and 1899 are distributed in shallow waters.
Of
those shipwrecks the schooner Experiment, which was wrecked six miles off of
127
Althorpe Island in Investigator’s Strait in 1881, presents the maximum depth of
water at approximately 56 metres.
The quality of each depth value is
controversial due to insufficient accuracy of most shipwreck positions.
For
example, the result of this analysis shows that the barque Mars, which wrecked
near Cape Borda on the Kangaroo Island, is located in 36 metres above sea level
in elevation. However, the remains of this vessel have not been identified on the
land.
Robert Mckinnon reports only vessel’s mast ring remaining on the shore
(Mckinnon 1993, p.51).
Considering the 64 shipwrecks that do not contain
relevant depth values, the result of this analysis has obvious limitations. Almost
all these shipwrecks are located out of gulf coastal waters, including the west
coast of the Eyre Peninsula and the southernmost section of Limestone Coast.
Despite the inaccuracy in the locational and elevational values, it should be noted
that the number of South Australian shipwrecks shows a high concentration in
waters less than 20 metres deep.
128
Value of depth
Number of Shipwrecks
Above sea level
38
0 - -9
80
-10 - -19
20
-20 - -29
7
-30 - -39
6
-40 - -49
1
-50 - -59
2
No value
64
Total
218
Table 4-3-1. Statistical data regarding the number of shipwrecks (1837-1899) and
water depth in South Australian waters.
129
5 Discussion and Conclusion
This chapter will summarise the results of this spatial analysis.
The results will
provide insights into the meaning of the distribution of South Australian
shipwrecks with a focus on existing historical and archaeological data. It will
also supply answers to the research questions established in the Chapter 1.
5-1 Interpreting the resultant analysis
Correlation between the location of shipwrecks and navigational lights
This analysis shows the distributional correlation between wrecks and
navigational lights in South Australian waters in the latter half of the nineteenth
century.
Considering the effectiveness of navigational lights, the spatial analysis
and statistical data demonstrate two different aspects regarding their distribution.
First, the navigational lights established on the South Australian coast reduced the
number of shipwrecks. Second, wreck hazards have continually occurred even
after the establishment of the lights. These aspects may contain inconsistencies
in evaluating the effectiveness of navigational lights, however this can be
130
explained through the characterisation of their locations by the following two
perspectives.
First, whereas the purpose of establishing navigational lights is to keep vessels
clear of hazardous positions, the risk potential for sailing vessels will remain in
the light coverage areas.
The natural factors causing ships to wreck are
topographical features in the vicinity of lighthouses.
Cosmos Coroneos and
Robert Mckinnon demonstrate statistical data about causes of vessel loss based on
six criteria.
These include bad luck, deliberate, unknown, weather,
weather/navigation, and weather/other human error (Coroneos & Mckinnon
1997, p. 103). Of these causes weather is the most common hazard for wrecking.
The major cause of loss for the majority of vessels in areas covered by
navigational lights is also heavy weather including squalls, gales, and strong
westerly winds (See Appendix III). Rough weather directly caused anchors to
drag, vessels to be driven ashore, and vessels to be struck on reefs and beaches.
Topographical features and poor weather conditions in the vicinity of navigational
lights and adjacent to lighthouses are not favourable for sailing vessels.
In the
spatial analysis of maritime space, the meaning of the location of navigational
lights should be calculated as hazardous, rather than safe positions.
131
Another perspective regarding the meaning of navigational lights is related to the
number of vessels using the coastal regions. As Bill Jeffery suggests, “…in
some cases as a result of previous wrecking but also because of a significant
number of vessels using a section of the coast” (Jeffery 1989, p. 51). According
to Jeffery establishing navigational lights resulted from taking into account highly
congested areas used by sailing vessels at the time.
There is difficulty estimating
exactly the historical number of vessels that sailed in particular sections of the
coast in the latter half of the nineteenth century, but this issue is important for
reconstructing the meaning of the distribution of South Australian shipwrecks.
Looking at the distributional map from maritime spatial analysis (Figure 4-2-6 &
4-2-7), it is valid to say that an archaeological approach that pursues the
distributional patterns of shipwrecks and navigational lights could inductively
reveal the more historically congested areas.
This perspective is assisted by the
fact that specific sections of heavy vessel traffic needed navigational lights.
Therefore the locations of navigational lights represent the historically congested
area.
132
Looking at maritime space through navigational lights
The correlation between the distribution of shipwrecks, the locations of
navigational lights, and historically congested areas should be regarded as a
principle in the study of maritime space.
On the other hand, in the case of South
Australia, through the process of this analysis the fact remains that continuous
wrecking occurred even after the establishment of navigational lights. This data
raises questions about the effectiveness of navigational lights in the South
Australian navigation system during the latter half of nineteenth century.
Looking at the historical background of the establishment of fourteen lighthouses
by the South Australian Marine Board (SAMB), their decision and campaign were
involved in not only intrastate circumstances but also interstate circumstances.
Although the SAMB was regarded as the government agency authorised to deal
with any maritime incidents during the settlement of South Australia, its capability
regarding navigational aids seems to be fairly limited.
With regard to the issue
of maintenance and administration of the navigation system, it is pointed out that
navigational lights in South Australia at the time were constructed by individual
authorities, architects, contractors and consultants who engaged in varied tasks
(Gordon 1988, p. 73).
Regarding the inconsistency of the administration of
133
navigational lights, it is suggested that:
“With regard to the construction and maintenance of lighthouses, jetties and
harbor works, most of the agencies ostensibly concerned with marine affairs did
not undertake construction and maintenance despite the various sections of
legislation entitling them to do so”（Marine & Harbors Department n.d.）.
It is necessary to take into consideration the capacity of the SAMB for managing
the navigation system in terms of financial and social issues.
While colonial
society in South Australia showed gradual growth during the latter half of
nineteenth century, it was still at an embryonic stage in which government
services were vulnerable to drought, labour shortages and economic declines.
Such unstable financial conditions might have affected the capability of the
SAMB to control all marine affairs.
Obviously establishing a relevant navigation system must have been an important
task for the SAMB, which must have been responsible for a sequence of
agreements about providing light services with other states. Nevertheless, the
responsibility for lighthouses and jetties was shared with other agencies
throughout the late half of nineteenth century (Table 5-1-1).
134
An incoherent
policy for establishing lighthouses on the South Australian coast might have
arisen due to this situation.
Considering the conditions of other navigational
light sources such as lightships, lights on the jetties, light makers, and light buoys,
it is valid to say that the status of the entire navigational aid system in South
Australian waters was unclear.
In addition, it is important to note that until the
enactment of the Commonwealth Navigation Act of 1912, the British government
retained jurisdiction over the navigation of Australian waters due to imperial
perspectives regarding nautical matters including defense, trade, and safety.
Although the colony was granted and well controlled by self-governing offices,
the British government continued to concern itself with the issues of establishing
lighthouses in all Australian waters (Gordon 1988, p. 113).
The result of this spatial analysis did not clarify whether or not there was actual
detriment to the navigational aid system due to the above obscure responsibilities
and interference by various organisations. This could be a specific theme for
future research, which could conduct archaeological and historical research into
the development of navigation systems and their operations.
135
AGENCY
Colonial Engineer
Colonial Architect's Office
Public Works Department
Colonial Architect's Office
Engineer and Architect's Office
Government Architect's Office
Engineer-in-Chief's Office
Engineer if Harbors and Jetties
Engineer-in-Chief's Office
PERIOD UNDER MINISTER
1841-1851
1852-1853
1854-1857
1858-1860
1860-1867
18671867-1876
1876-1880
1880-1914
Table 5-1-1. Agencies responsible for lighthouses and jetties until the
establishment of the Harbors Board in1914.
The spatial meaning of the distribution of shipwrecks in relation to the
historical development of the South Australian colony
According to the results of the spatial analysis, it is possible to relate the spatial
meaning of shipwrecks to the expansion of the colony of South Australia.
The
distribution of both shipwrecks and navigational lights shows an apparent
alteration from the middle nineteenth century to the late nineteenth century.
Figure 4-2-6 shows that from the 1840s to the early 1860s the focal area for
maritime activity was along the west coast of the Fleurieu Peninsula and the east
coast of South Australia.
From the 1860s and 1880s (as indicated in Figure
136
4-2-7) the distribution of shipwrecks and navigational light expanded along all of
the gulf coasts.
This spatial expansion can be explained by the gradual
development of economic and social activity which occurred on the both Yorke
and Eyre Peninsulas in the late nineteenth century (Griffin & McCaskill 1986 pp.
16-21). The use of sailing vessels played an important role in colonizing South
Australia because vessels provided a reliable system that enabled people to
transport materials and communicate with others.
The use of sailing vessels
prevented the colony of South Australia from becoming spatially isolated.
The
South Australian colony same as other colonies, which attempted to sustain an
equalised level of consumer society, could not be consisntent without depending
on contact with other consumer societies via shipping (Staniforth 2003b, pp.
143-151). The distribution of shipwrecks and navigational lights dating to the
late half of nineteenth century provided evidence for the prosperity of the colonial
society which was sustained by both the intrastate and intrastate maritime
activities.
The variables of maritime spatial analysis
With regard to variables impacting the distribution of shipwrecks, this study
evaluated both natural and social aspects.
137
Employing the bathymetric data for
analyzing the correlation between the distribution of shipwrecks and water depths
demonstrates the fact that a large number of shipwrecks are located in the shallow
waters.
This is significant to coastal resource management where cultural
agencies can concentrate on
administrating shallow and coastal sections of
waters.
The concept that shallow waters are likely to contain many shipwrecks has been
commonly discussed in maritime archaeological studies.
For example, based on
Lloyd’s of London records, Willard Bascom (1976, p. 72) suggests that for the
eighteenth and nineteenth centuries, approximately 40 percent of wooden sailing
ships ended their careers on the shore, and another 10 to 20 percent were wrecked
well offshore.
A. J. Parker (1992, p. 5) also demonstrates the relationship
between depth and condition of shipwrecks in his shipwreck database that deals
with the 1,189 Mediterranean shipwreck sites dating to the Roman period (Table
5-1-2).
The data shows that there is the highest concentration of shipwrecks in
shallow waters.
138
Depth
Number of shipwrecks
Silted/dry land
48
Shallow (0-15m)
336
Medium (15-30m)
166
Deep (30-60m)
236
Very deep (60- m)
47
Depth unknown
356
Total
1189
Table 5-1-2. Data regarding the relationship between water depth and the number
of ancient shipwrecks in the Mediterranean Sea (reproduced from Parker 1992, p.
5).
There is an unequalness of the distribution of shipwrecks even within the defined
area of coastal and shallow waters. In the case of the Mediterranean, Parker
points out that one can observe how unequal the distribution of shipwrecks in
coastal waters is when creating distribution maps (Parker 1992, p. 6).
The
correlation between the high concentration of shipwrecks and shallow waters is
not simply defined based on the depth of waters. Looking at the entire maritime
space of South Australia, the location of shipwrecks is not necessary equal to the
location of shallow waters (Figure 4-2-6 & 4-2-7).
More concretely, the
shipwrecks are located adjacent to harbours and jetties, being the end of shipping
routes.
Vessels frequently face a risk when approaching harbours and jetties.
139
The concept that a high distribution of shipwrecks lies in shallow waters must be
integrated into the location of harbours and jetties.
5-2 Perspectives to the research questions
Four research themes were established at the beginning of this spatial analysis
study.
Some perspectives concerning these themes were obtained during this
study and will be addressed in this section.
The research themes include:
1. A focus on the analysis of comprehensive site formation processes in macro
space using multiple shipwrecks, rather than the traditional, particularistic site
formation processes studies of individual shipwrecks.
The study area in this spatial analysis was established as 218 shipwreck sites
dating to the latter half of the nineteenth century and located in South Australian
waters.
A holistic approach was adopted for this study, based on an analytical
framework highlighting the maritime cultural landscape.
Consequently, the
spatial meaning of multiple shipwrecks was demonstrated to correlate with
cultural factors.
To approach multiple shipwrecks using site formation processes
theory is an innovative idea.
This analysis highlighted the significance of
140
pre-deposit wreck site formation processes by focusing on multiple shipwrecks.
This approach can be distinguished from existing theories of “pre-depositional,”
“depositional” and “post-depositional” wreck processes which highlight the
interpretation of an individual shipwreck.
In comparison with traditional flow
charts for wreck site formation processes, which aimed at applying middle-range
theory to maritime archaeology, this approach can be comprehended using a more
inclusive diagram (Figure 5-2-1).
The key concepts of this model are the clarification of “pre-deposition processes”
and the adoption of a holistic, multiple shipwreck approach.
Muckelroy
emphasized that there is “the need to understand the pre-wreck nature of a ship
and its contents”(Muckelroy 1976 cited in Gibbs 2006, p. 4). Several cultural
and natural factors related to the nature of single wrecking events commonly exist
at the pre-depositional stage; these factors also can be understood by observing
multiple shipwrecks.
141
Post-deposition formation processes
(Micro level)
Observed wreck
distribution
Cultural / Natural factors
Cultural / Natural factors
*
*
*
*
Cultural & Natural factors (i.e.maritime
infrastructures, ship construction
techniques, and hydrographic data)
Pre-deposition formation processes
(Macro level)
Cultural / Natural factors
* Selected
C/N factors
Figure 5-2-1. A shipwreck assessment model focusing on pre- and
post-depositional cultural and natural factors in the macro and micro level
analysis of shipwrecks
142
*
The diagram indicates a multiple-level analysis which is demonstrated by
concentric circles of pre- and post-depositional site formation processes.
The
inner circle relates observed wreck distribution included within a specific
geographic area of research. The entire assemblage of shipwrecks in an area is
affected by cultural and natural factors, such as maritime infrastructure, ship
construction techniques, and hydrographic data.
identified
by
maritime
archaeologists
These factors which can be
contribute
to
understanding
the
pre-depositional site formation processes at the macro level of analysis.
The micro level is represented by an outer circle which includes a focus on
individual shipwrecks.
Maritime archaeologists are able to create their own
explainable assemblage of cultural and natural factors in order to identify the final
observed seabed distribution of a single shipwreck.
This study introduces a new model which allows for the analysis of multiple
shipwrecks and pre-depositonal factors not outlined yet by previous researchers.
It also allows for archaeologists to independently choose those factors that may
have affected wrecking events rather than following a processual, steadfast pattern
143
as defined in previous flow charts. Further, maritime archaeologists are able to
move between macro level analysis and micro level analysis which allows for a
more flexible and inclusive approach to the spatial analysis of shipwrecks.
2. Site formation processes and maritime cultural landscape theories can be used
to identify past human activity and the physical environment, which affect the
spatial distribution of shipwrecks.
Understanding both cultural and natural environments in maritime space can help
to explain the historical distribution of wrecks.
Based on this premise, the
shipwrecks located in shallow waters were concluded to be the result of vessels
which had faced risks in approaching habours and jetties in South Australia.
This analysis did not clarify whether the causes of the risk originally derived from
structural defects or placement of South Australian harbours and jetties. Further
analysis must be conducted to address this issue. Additionally, this study did not
address the correlation between location of shipwrecks and other environmental
factors such as rough seas and gale forces in particular areas.
For example,
historical navigational charts and notices to mariners describe the direction of
hazardous winds and seas in some areas.
144
By plotting these data on spatial maps
the location of shipwrecks may be related to further natural variables. There is the
potential in further research to identify these influential variables.
It should be noted that geo-spatial data for GIS obtained from recent surveys does
not precisely reflect past topographical features and environments.
Since it is
inferred that natural environments might have been altered over time, efforts to
make a physical gap between historical environments and current geo-spatial data
are required.
3. The interpretation of the spatial relationships between the distribution of
shipwrecks and other maritime infrastructures, such as jetties, and lighthouses,
can demonstrate the significance of established maritime culture population
centers.
This analysis has proved that the distribution of shipwrecks can be explained in
spatial relation to navigational lights and vessel destinations. It is assumed that
with the establishment of relevant navigation systems and more favourable, safer
destinations the prosperity of specific areas adjacent to maritime facilities these
maritime facilities would increase. It is necessary to focus archaeological and
historical research on individual sets of maritime infrastructure at both regional
145
and state levels.
Maritime cultural landscape is conceptually understood by
identifying relationships between shipwrecks, maritime infrastructures, and land
properties. In many cases shipwrecks should not be approached as spatially and
temporally isolated.
4). Although the validity of using GIS as a predictive model to analyse underwater
sites is still disputable, this study may help in determining some of the natural
variables, such as prevailing winds, currents, hydrography, channels alignment,
and bottom sediments that affect shipwreck distribution.
Predictive modeling procedures are not well established even in terrestrial
archaeology, and few have attempted to create completely deductive models for
predicting the exact location of shipwrecks.
This maritime spatial analysis used
navigational light locations and bathymetric data as variables that could impact
the distribution of shipwrecks.
Correlative GIS modeling shows success in
reconstructing relationships between objects within the framework of a macro
area analysis.
This study has demonstrated that using GIS in maritime
archaeology to understand the spatial analysis of multiple shipwrecks and
maritime infrastructures can be successful.
146
Finally, it is acknowledged that these insights resulted from an inclusive approach
adopted by this study.
Understanding maritime cultural space depends on
geographical, historical, and archaeological perspectives rather than a single
dominant concept.
Without recent technological developments in the field of
GIS spatial analysis in both geography and archaeology, this study would hardly
be complete.
The achievement was sustained by the recurrent growth of theory
and research strategies that have occurred in maritime archaeology for thirty
decades.
Studies tracing the theoretical development of maritime archaeology
would characterise this research as broad-based as it allows for taking a
comprehensive and interdisciplinary approach (Gibbins & Adams 2001, pp.
284-286; Staniforth 2003a).
Therefore, it is inferred that more innovative
approaches resulting from the technical and theoretical developments in maritime
archaeology will provide better solutions for identifying the complexities of
maritime cultural space.
147
5-3 Conclusion
Clarifying the spatial and temporal relationships of cultural remains is a principle
theme in archaeology.
Maritime archaeologists have made inroads into
understanding the spatial and temporal structures of maritime cultural remains.
In particular, research has highlighted spatial relationship within individual
shipwreck sites and their associated cargos.
On the other hand, there is a shift
toward employing holistic and thematic analyses which highlight multiple sites,
including maritime infrastructures. This study examined how maritime space
can be interpreted for distributional meaning through multiple wreck sites.
The
research dealt with wrecks located in South Australian waters which dated to the
latter half of the nineteenth century and were assocated with maritime economic
development occurring at the time.
Although GIS analysis is not common
among maritime archaeologists, its use in this study demonstrated a spatial
correlation between shipwrecks, navigational lights, and shallow waters.
As a
result, this analysis provided new perspectives regarding the theory of wreck site
formation processes by revealing the significance of pre-deposit processes
through the analysis of multiple shipwrecks in a large area.
The utility of
analyzing multiple shipwrecks based on thematically spatial and temporal
148
frameworks was proven through this study; further providing that GIS not only
has great potential for managing and displaying maritime cultural heritage, but
also is useful as a tool for data analysis.
149
References
Allen, KMS 1990, ‘Manipulating space: a commentary on GIS applications’, in
KMS Allen, SW Green & EBW Zubrow (eds), Interpreting space: GIS and
archaeology, Taylor & Francis, London, pp. 197-200.
Allen, KMS, Green, SW & Zubrow EBW 1990, Interpreting space: GIS and
archaeology, Taylor & Francis, London.
Australian New Zealand Land Information Council 2001, Shipwrecks of South
Australia,
viewed
22
March
2006,
<http://www.asdd.sa.gov.au/asdd/ANZSA1022000021.html>.
Australian New Zealand Land Information Council 2006, Australian Spatial Data
Directory SADD, viewed 22 March 2006, <http://asdd.ga.gov.au/>.
Bascom, W 1976, Deep water, ancient ships: the treasure vault of the
Mediterranean, David & Charles publishers, Newton Abbot, London.
Boyd, WE, Pathirana, S & Bell, G 1996, ‘A Contribution to the management of
the buried historical shipwreck heritage: a GIS risk assessment model for the
River Richmond mouth, northern New South Wales’, The Bulletin of
theAustralian Institute for Maritime Archaeology, vol. 20, no. 2, pp. 57-70.
Chapman, GD 1972, Kangaroo Island shipwrecks: an account of the ships and
cutters wrecked around Kangaroo Island, Roebuck Society, Canberra, Australia.
Christoffersen, J 1994, ‘The Danish national record of marine and maritime sites’,
in C Westerdahl (ed.), Crossroads in ancient shipbuilding: proceeding of the Sixth
international Symposium on Boat and Ship Archaeology Roskilde 1991, ISBSA 6,
150
Oxbow Monograph 40, Oxford, pp.283-288.
Clark, P 1990, Shipwreck sites in the south-east of South Australia 1838-1915,
South Australian maritime archaeology series no. 1, Australian Institute for
Maritime Archaeology Special Publication no. 5, Department of Environment and
Planning , Fremantle, Western Australia.
Clarke, DL 1977, ‘Spatial information in archaeology’ in DL Clarke (ed.), Spatial
Archaeology, Academic Press, London, pp. 1-32.
Cooper, M, Crenshaw, R & Penman, T 2002, ‘Applications of GIS in the search
for the German U-5599 submarine: a brief case study’, in DJ Wright (ed.), Under
sea with GIS, ESRI Press, California, USA, pp.155-164.
Coroneos, C 1997, Shipwrecks of Encounter Bay and Backstairs Passage, South
Australian Maritime Heritage Series no. 3, Australian Institute for Maritime
Archaeology Special Publication no. 8, Department of Environment and Natural
Resources, Adelaide, South Australia.
Coroneos, C & Mckinnon, R 1997, Shipwrecks of investigator strait and the lower
Yorke Peninsula, South Australian Maritime Heritage Series no. 4, Australian
Institute for Maritime Archaeology Special Publication no. 9, Department of
Environment and Natural Resources, Adelaide, South Australia.
Ghobadi, A & Tsumura, H 2006, ‘The development of GIS applications in
archaeology: perspectives from both sides of the Pacific.’, paper presented to the
World Archaeological Congress (WAC) Inter-Congress: Osaka, 2006, Osaka,
Japan, 12-15 January.
Gibbins, D & Adams, J 2001, ‘Shipwrecks and maritime archaeology’, World
151
Archaeology, vol. 32, no. 3, pp. 279-291.
Gibbs, M 2006, ‘Cultural site formation process in maritime archaeology: disaster
response, salvage and Muckelroy 30 years on’, The International of Journal of
Nautical archaeology, vol. 35, no. 1, pp. 4-19.
Gordon, R 1988, From dusk till dawn: a history of Australian lighthouses,
Macmillan Company of Australia, South Melbourne, Australia.
Government of South Australia 2000, Atlas South Australia, viewed 22 March
2006, <http://www.atlas.sa.gov.au>.
Green, J & Souter, C 1999, ‘Applications of the HPASS (High Precision Acoustic
Surveying System) to the Pandora archaeological project’, The Bulletin of the
Australian Institute for Maritime Archaeology, vol. 23, pp. 75-78.
Griffin, T & McCaskill, M (comps) 1986, Atlas of South Australia, South
Australian Government Printing Division, South Australia.
Groom, D & Oxley, I 2002, ‘Maritime Fife, managing Fife’s underwater heritage:
a feasibility study for a maritime archaeological GIS’, in D Wheatley, G Earl & S
Poppy (eds), Contemporary themes in archaeological computing: University of
Southampton of Archaeology Monograph No.3, Oxbow Books, Oxford, pp. 50-57.
Hannahs, T 2003, ‘Underwater parks versus preserves: data or access’ in JD
Spirek & DA Scott-Ireton (eds), Submerged cultural resource management:
preserving and interpreting our sunken maritime heritage, Kluwer Academic &
Plenum Publishers, New York, pp. 5-16.
Harris, T & Lock, G 1990, ‘The diffusion on a new technology: a perspective on
152
the adoption of geographic information systems within UK’, in KMS Allen, SW
Green & EBW Zubrow (eds), Interpreting space: GIS and archaeology, Taylor &
Francis, London, pp. 33-53.
Hodder, I 1977, ‘Some new directions in the spatial analysis of archaeological
data at the regional scale (Macro)’, in DL Clarke (ed.), Spatial Archaeology,
Academic Press, London, pp. 223-351.
Hunter, JR 1994, “Maritime culture’: notes from the land”, The International
Journal of Nautical Archaeology, vol. 23, no. 4, pp. 261-264.
Hydrographical offices 1830, The Australia directory; volume I.: containing
directions for the southern shores of Australia, from Cape Leeuwin to Port
Stephens, including Bass’ Strait and Van Diemen’s Land, London.
Ibbotson, J 2000, Lighthouse of Australia: images from the end of an era,
Australian Lighthouse Traders, Victoria, Australia.
Intergovernmental Committee on Surveying and Mapping 2001, What is the
difference between WGS84 and GDA94?, Commonwealth of Australia, Canberra,
Australia, viewed 22 March 2006
<http://www.icsm.gov.au/icsm/gda/wgs84fact.pdf>.
Jeffery, B 1989, ‘Research into Australian-built coastal vessels wrecked in South
Australia, 1840-1900’, The Bulletin of the Australian Institute for Maritime
Archaeology, vol. 13, no. 2, pp. 51-56.
Johnson, M 2005, ‘Thinking about landscape’ in C Renfrew & P Bahn (eds),
Arcaheology: the key concepts, Routledge, Oxon, UK, pp.156-159.
153
Kaneda, A, Tsumura, H, & Niiro, I 2001, Koukogaku-no tameno GIS nyuumonn
(A handbook for the GIS analysis in archaeology), Kokon Shoin, Tokyo [in
Japanese].
Kenderdine, S 1997, Culture and heritage: shipwrecks and associated objects,
Australia: state of environment technical paper series (Natural and Cultural
Heritage), Environment Australia, Canberra, Australia.
Knapp, AB & Ashmore, W 1999, ‘Archaeological landscapes: constructed,
conceptualized, ideational’, in W Ashmore & AB Knapp (eds), Archaeologies of
landscape: contemporary perspectives, Blackwell Publishers, Oxford, pp.1-30.
Kvamme, KL 1995, ‘A view from across the water: the North America experience
in archaeological GIS’, in G Lock & Z Stančič (eds), Archaeology and
geographical information systems: a European perspective, Taylor & Francis,
London, pp. 1-14.
Kvamme, KL 1999, ‘Recent directions and developments in geographical
information systems’, Journal of Archaeological Research, vol. 7, no. 2, pp.
153-201.
Kvamme, KL 2006, ‘There and back again: revisiting archaeological location
modeling’, in MW Mehrer & KL Wescott (eds), GIS and archaeological site
location modeling, CRS Press Taylor & Francis, Boca Raton, USA, pp. 3-38.
Lamotta, VM & Schiffer, MB 2005, ‘Archaeological formation processes’, in C
Renfrew & P Bahn (eds), Arcaheology: the key concepts, Routledge, Oxon, UK,
pp. 121-127.
Lipson, T 1853, The South Australia directory: or sailing directions from Denial
154
Bay, Lat. 32° 15½’ S., Long. 133°26’ E., to Cape Northumberland, Lat. 38° 3’ 15”
S., Long. 140°38’ E., W. C. COX, Government Printer, Adelaide, South Australia.
Lock, G & Harris, T 2006 ‘Enhancing predictive archaeological modeling:
integrating location, landscape, and culture’, in MW Mehrer & KL Wescott (eds),
GIS and archaeological site location modeling, CRC Press Taylor & Francis,
Boca Raton, USA, pp. 41-62.
Mardry, S, Cole, M, Gould, S, Resnick, B, Seibel, S, & Wilkerson, M 2006, ‘A
GIS-based archaeological predictive model and decision support system for the
north Carolina Department of Transportation, in MW Mehrer & KL Wescott (eds),
GIS and archaeological site location modeling, CRC Press Taylor & Francis,
Boca Raton, USA, pp. 317-334.
Marine Board office 1877, Hydrographic notice no. 28: Australia directory, vol. I.
notice no. 15: coast of the colony of South Australia, WC COX, Government
Printer, Adelaide, South Australia.
Marine & Harbors Department n.d., ‘Department of marine and harbors: the
administration of marine and port affairs in South Australia since 1836’, Public
Record Office of South Australia, agency registration sheet, South Australia.
Mather, IR & Watts, GPJr. 1998, ‘Geographic information systems for submerged
cultural resource management and site specific investigation’, Underwater
Archaeology: Society for Historical Archaeology, pp. 3-13.
Mather, IR & Watts, GPJr. 2002, ‘Geographic information systems’, in CV Ruppé
& JF Barstad (eds), International Handbook of Underwater Archaeology, Kluwer
Academic & Plenum Publishers, New York, pp. 679-696.
155
McErlean, T, McConkey, R & Forsythe, W 2002, Strangford Lough: an
archaeological survey of the maritime cultural. Northern Ireland Archaeological
Monographs No. 6. Blackstaff Press and Environment and Heritage Service,
Belfast, Ireland.
Mckinnon, R 1993, Shipwreck sites of Kangaroo Island, State Heritage Branch,
Department of Environment and Land Management, Adelaide, South Australia.
Mehrer, MW & Wescott, KL (eds) 2006, GIS and archaeological site location
modeling, CRC Press Taylor & Francis, Boca Raton, USA.
Muckelroy, K 1976, ‘The integration of historical and archaeological data
concerning an historic wreck site: the “Kennemerland”’, World Archaeology, vol.
7, no. 3, pp. 280-289.
Muckelroy, K 1978, Maritime archaeology, Cambridge University Press,
Cambridge, UK.
Murphy, LE (ed.) 1993, Dry Tortugas National Park: submerged cultural
resources assessment, Submerged Resources Unit, National Park Service,
Department of the Interior, USA.
Murphy, LE (ed.) 1998a, H.L Hunley: a cooperative project of National Park
Service, Naval Historical Center, South Carolina Institute of Archaeology and
Anthropology, Submerged Resources Unit, National Park Service, Department of
the Interior, USA.
Murphy, LE 1998b, ‘Geographic information system (GIS)’ in JP Delgado (ed.),
Encyclopedia of underwater and maritime archaeology, Yale University Press,
New Haven and London, pp. 166-168.
156
O’Shea, JM 2002, ‘The archaeology of scattered wreck-sites: formation processes
and shallow water archaeology in western Lake Huron’, The International
Journal of Nautical Archaeology, vol. 31, no. 2, pp. 211-227.
Parker, AJ 1992, Ancient shipwrecks of the Mediterranean & the Roman provinces,
BAR International Series 580, Tempvs Reparatvm, Oxford.
Parsons, R 1983, Ketches of South Australia Australian Maritime Historical
Society, Adelaide, South Australia.
Parsons, R 1985, Lighthouses of South Australia, Ronald Parsons, Magill, South
Australia.
Parsons, R 1986a, Shipwrecks in South Australia (1836-1875), Ronald Parsons,
Goolwa, South Australia.
Parsons, R 1986b, Southern passages: a maritime history of South Australia,
Wakefield Press, South Australia.
Petkovic, P & Buchanan, C 2002, Australian bathymetry and topography grid,
Geoscience Australia, Canberra, Australia.
Petrie, L, Johnson, I, Cullen, B & Kvamme, KL 1995, GIS in archaeology: an
annotated bibliography, Sydney University archaeological methods series1,
Archaeological Computing Laboratory Archaeology: Prehistoric and Historical
University of Sydney, NSW, Australia.
Savage, SH 1990, ‘GIS in Archaeological research’, in KMS Allen, SW Green &
EBW Zubrow (eds), Interpreting space: GIS and archaeology, Taylor & Francis,
157
London, pp.22-33.
Schiffer, MB 1987, Formation processes of the archaeological record,
Albuquerque: University of New Mexico Press.
Sexton, R 1986, The South Australian coast 1853: extracts from sailing directions
compiled by Captain Thomas Lipson, R.N., Harbour Master and Naval Officer of
South Australia. Sail Training Association of South Australia Inc, Adelaide, South
Australia.
Staniforth, M 2003a, ‘Annales-informed approaches to the archaeology of
colonial Australia’, Historical Archaeology, vol. 37, no. 1, pp. 102-113.
Staniforth, M 2003b, Material culture and consumer society, Kluwer Academic &
Plenum Publishers, New York.
Tanaka, K 2005, ‘Toshi Kankyoo-no Bunseki (An analysis of urban area)’ in S
Takahashi, T Inoue, K Sanjyou & T Takahashi (eds), Zirei-de Manabu GIS-to
TiikiBunseki: using ArcGIS-wo motiite (The case studies of GIS and spatial
analysis: using ArcGIS), Kokon Shoin, Tokyo, Japan, pp. 100-114 [in Japanese].
Temme, AR 1975, Society for underwater historical research: South Australian
shipwrecks 1880-1899, South Australian Museum, Adelaide, South Australia.
Tilley, C 1994, A phenomenology of landscape: places, paths and monuments,
Berg Publishers, Oxford.
Trigger, BG 1989, A history of archaeological thought, Cambridge University
Press, Cambridge, UK.
158
Ward, IAK, Larcomb, P & Veth, P 1998, Towards new process-oriented models
for describing wreck disintegration: an example using the Pandora wreck, The
Bulletin of the Australian Institute for Maritime Archaeology, vol. 22, no. 1, pp.
109-114.
Ward, IAK, Larcomb, P & Veth, P 2001, ‘A new process-based model for wreck
site formation’, in M Staniforth & M Hyde (eds), Maritime archaeology in
Australia: a reader, Southern Archaeology, Blackwood, South Australia, pp.
213-222.
Warren, RE. 1990a, ‘Predictive modeling in archaeology: a primer’, in KMS
Warren, RE 1990b, ‘Predictive modeling of archaeological site location: a case
study in the Midwest’, in KMS Allen, SW Green & EBW Zubrow (eds),
Interpreting space: GIS and archaeology, Taylor & Francis, London, pp. 201-215.
Westerdahl, C 1992, ‘The maritime cultural landscape’, The International Journal
of Nautical Archaeology, vol. 21, no. 1, pp. 5-14.
Wheatley, D & Gillings, M 2002, Spatial Technology and archaeology: the
archaeological applications of GIS, Taylor & Francis, London.
Zubrow, EBW & Green, SW 1990, ‘Coping with space: commentary on GIS data
sources, hardware and software’, in KMS Allen, SW Green & EBW Zubrow (eds),
Interpreting space: GIS and archaeology, Taylor & Francis, London, pp. 129-133.
159
160
http://www.esri.com/data/download/basemap/
http://www.maproom.psu.edu/dcw/
http://www.geographynetwork.com/
http://www.ga.gov.au/
http://www.ga.gov.au/oracle/#mar
http://www.ga.gov.au/oracle/mars/
http://www.usgs.gov/
http://asdd.ga.gov.au/
ESRI World Basemap Data
Digital chart of the world
Geographic Network
Geoscience Australia
Online Mapping & Databases
Marine Sediments Database
US Geological Survey
Australian Spatial Data Directory (ASDD)
http://www.perseus.tufts.edu/
http://ahds.ac.uk/archaeology/index.htm
http://idp.bl.uk/
http://www.abc.net.au/wordmap/
Perseus Digital Library
The Arts and Humanities Data Service
International Dunhuang Project
Australian Word map
http://www.esri.com/software/arcgis/arcims/index.html
http://www.cmcus.com/Products/Vendors/MapInfo/mapxtreme.asp
http://www.empower.com/pages/products_miDiscovery.htm
ArcIMS
Mapinfo MapXtreme
Mapinfo Discovery
Commercial Mapping Servers
http://www.atlas.sa.gov.au
Atlas of South Australia
Web-based GIS Project
http://www.ecai.org/
The Electronic Cultural Atlas Initiative
GIS Web data Resources
AppendixⅠ- GIS resources on the Internet
Appendix II – Description of South Australian Shipwreck database
161
162
LOSSDATE
1860
1860
1881
1892
LOSSCAUSE
DRIVEN ASHORE BY STRONG WIND
DRIFTED TOWARDS SHORE WITHOUT ANCHOR AND STRUCK ROCKS
VESSEL OVERLOADED, STRUCK BY SQUALL, CAPSIZED & SANK
VESSEL SPRANG A LEAK AND SANK
LOSSCAUSE
ACCIDENTALLY STRUCK ROCKS DUE TO TIDAL MOVEMENT
STRUCK SANDBANK, OPENED LEAK AND CARRIED NW DIRECTION 9T]
BLOWN ASHORE DURING HEAVY WEATHER
STRUCK MARION REEF
WENT OFF COURSE WITH WESTERLY CURRENT & RAN ASHORE
FOUNDERED DURING HEAVY SQUALL LOST WITHOUT A TRACE
INCORRECT COMPASS READING & RAN AGROUND
WRECKNAME
J. LOVETT
WITNESS
NENE VALLEY
IRON AGE
ADELAIDE
BANDICOOT
RIG/HULLDESC
SCHOONER
BRIGANTINE
BARQUE
BARQUE
BRIGANTINE
SCHOONER
LOSSDATE
1852
1853
1854
1855
1861
1861
DRIVEN OFF COURSE BY ROUGH WEATHER & WENT ASHORE
CAPT MISTOOK LAND FOR A CLOUD & BLOWN ASHORE IN WSW BREEZE
DUE TO COMPASS ERROR, LIGHTS MISTAKEN & RAN AGROUND ON REEF
DRAGGED ANCHORS, DRIVEN ASHORE DURING GALE & BROKE BACK
INEFFICIENT GOVT MOORINGS, DRIVEN ASHORE & BROKE BACK
LOSSCAUSE
Cape Northumberland 1858 (lit in 59, improved in 1882)
1858
RIG/HULLDESC
CUTTER
BARQUE
SCHOONER
KETCH
Cape Borda
WRECKNAME
ATALANTA
FIDES
EXPERIMENT
ENTERPRISE
LOSSDATE
1838
1838
1849
1851
1854
1867
1873
1855 (lit 56, upgrade in 1882
RIG/HULLDESC
BARQUE
BRIG
SHIP
SHIP
BRIG
CUTTER
SHIP
Troubridge
WRECKNAME
PARSEE
DART
SULTANA
MARION
CHARLES CARTER
AGNETA
IRON KING
1852
RIG/HULLDESC LOSSDATE LOSSCAUSE
CUTTER
1853
MAINSAIL SPLIT IN HEAVY SEA & DRIFTED ONTO ROCKS
Cape Willoughby
WRECKNAME
MIDGE
Appendix III - Data of shipwrecks located in the visible area of each navigational light
163
1864 (improved in 18759
RIG/HULLDESC LOSSDATE LOSSCAUSE
1877
DRAGGED ANCHORS IN HEAVY GALE AND WENT TO PIECES
Victor Harbor
1860
RIG/HULLDESC LOSSDATE LOSSCAUSE
CUTTER
1850
GROUNDED DURING STORM AND BROKE UP
CUTTER
1879
PRESUMED SUNK IN SQUALL WITH ALL SAILS SET
Semaphore Jetty
WRECKNAME
JANE FLAXMAN
MARIA
WRECKNAME
LADY OF THE LAKE
RIG/HULLDESC LOSSDATE LOSSCAUSE
YACHT
1886
DRAGGED MOORINGS DURING GALE & RAN INTO JETTY & BECAME TOTAL WRECK
YACHT
1884
WENT DOWN AT MOORINGS IN ROUGH WEATHER, WENT ASHORE
LOSSCAUSE
BROKE FROM ANCHORAGE DURING WSW GALE, STRUCK REEF & WENT ASHORE
STRUCK BY SQUALL, CAPSIZED, DRIFTED TO CAPE NORTHUMBERLAND & BROKE UP
PARTED FROM MOORINGS, DRIFTED ASHORE & BEACHED
PARTED FROM MOORINGS & LODGED ON REEF DURING HEAVY WEATHER
SPRANG A LEAK, VESSEL RAN ASHORE AND BROKE UP
BROKE CABLE, DRIVEN ASHORE IN HEAVY SEAS & BROKE UP
BROKE FROM MOORINGS DURING STRONG S.E.GALE AND RAN AGROUND
SLIPPED MOORINGS AND DRIFTED ONTO ROCKS DURING SS GALE
PRESUMED TO HAVE BEEN SUNK IN STRONG NNW SQUALL
PENGUIN IS LIGHT MISTAKEN FOR CAPE NORTHUMBERLAND STOVE IN
WENT ASHORE IN FOG, HIT REEF AND BROKE UP
BROKE MOORINGS, WENT ASHORE IN HEAVY SEAS, BROKE UP IN SSW GALE
MISTAKE IN BEARINGS, STRUCK REEF AND BROKE UP BY SEA
STRUCK REEF, WATER ENTERED AND ABANDONED
WRECKNAME
MINNIE
HERO
LOSSDATE
1861
1861
1873
1873
1873
1876
1876
1876
1880
1880
1890
1892
1893
1894
1859
1859
SHIP
BARQUE
BARQUE
KETCH
RIG/HULLDESC
BARQUE
BRIG
SCHOONER
SCHOONER
SCHOONER
BARQUE
SCHOONER
BRIGANTINE
Glenelg (lightship)
Glenelg Jetty
WRECKNAME
MIAME
JOHN ORMEROD
ORWELL
PRINCE OF WALES
FLINDERS
ST. MARC
COUNTESS
GALATEA
PRIMA DONNA
SOUTHERN CROSS
GLENROSA
LOTUS
TENTERDEN
AEOLUS
164
Meningie Jetty
Milang Jetty
1871
RIG/HULLDESC LOSSDATE LOSSCAUSE
BRIGANTINE
1855
DURING HEAVY WEATHER AND THICK FOG RAN UPON REEF
1856
WHILE TOWING, TOW ROPE BROKE, VESSEL TOOK IN WATER & SANK
CUTTER
1858
FOUNDERED AND DRIVEN ASHORE
SCHOONER
1864
RAN ASHORE DURING HEAVY WEATHER
Cape Jervis
1871
1869
LOSSCAUSE
GALE DROVE SHIP ASHORE
SUPPOSEDLY SUNK DURING HEAVY WINDS
RAN ASHORE AND SANK DURING HEAVY SEAS
BAD QUALITY SAILS & GEAR, SANK IN BAD WEATHER
FOUNDERED DURING STORM
SANK BECAUSE OF HEAVY SQUALLS, FILLED WITH WATER AND FOUNDERED
EXPLOSION IN STEAM ROOM OF VESSEL AND TOTAL LOSS
BROKE-UP DURING GALE, BROKE AGAINST JETTY & BECAME TOTAL WRECK
WRECKNAME
SANS PAREILLE
GOULBURN
VENTURE
VANQUISH
NO SHIPWRECKS
Moonta Jetty
NO SHIPWRECKS
CUTTER
LOSSDATE
1850
1853
1854
1865
1871
1875
1883
1890
1867 (improved in 1873 &19
RIG/HULLDESC
BARQUE
SCHOONER
CUTTER
CUTTER
CUTTER
CUTTER
Port Adelaide
1867 (improved in1877)
LOSSCAUSE
STRANDED ON REEF DURING HEAVY WEATHER & BROKE UP
FAILED TO OBSERVE BUOY AND RAN ONTO REEF
COURSE OF VESSEL ALTERED AND STRUCK REEF
WRECKNAME
GRECIAN
ROSE
TRIAL
CORSAIR
ELIZABETH
NIMROD
LITTLE ORIENT
AMY
NO SHIPWRECKS
LOSSDATE
1865
1873
1882
1866 (improved in 1877)
RIG/HULLDESC
BARQUE
SCHOONER
KETCH
Tipara
WRECKNAME
SAN MIGUEL
KANGAROO
YOUNG LION
165
WRECKNAME
SARAH
DURING HEAVY WSW GALE, LOST MAST & DRIVEN ASHORE, STRUCK REEF
DURING NW GALE SHIP LOST CABLE & ANCHORS & STRUCK ROCKS
PARTED CABLE WHEN AT ANCHOR, AND WENT ASHORE
STRUCK REEF ON ENTERING BAY NO PILOT, TOOK IN WATER & RAN ASHORE
STOOD TOO LONG ON ONE TACK, RAN TOO CLOSE TO SHORE & GROUNDED
DURING NW GALE, WINDLASS CUT IN TWO, LOST CABLE & DRIVEN ASHORE
BROKE CABLE IN N.W. GALE, CAME ASHORE AND BROKE UP
DRAGGED ANCHOR AND RAN ASHORE ON ROCKS DURING GALE
RIG/HULLDESC LOSSDATE LOSSCAUSE
YAWL
1879
CIRCUMSTANCES UNKNOWN
Port Wakefield Whar1873
1872
RIG/HULLDESC LOSSDATE LOSSCAUSE
CUTTER
1867
FOUNDERED DURING HEAVY SQUALL LOST WITHOUT A TRACE
Edithburgh Jetty
WRECKNAME
AGNETA
1849
1853
1855
1857
1857
1857
1861
1861
SANK
UNSEAWORTHY CONDITIONS, BLOWN ASHORE IN STORM & BROKE UP
SUNK DUE TO HEAVY WEATHER
MISJUDGING OF DISTANCES & WENT ASHORE ON ROCKS, TOTAL LOSS
UNKNOWN
WESTERLY GALE DROVE VESSEL ASHORE - CAPSIZED IN SHALLOW WATERS
STRUCK BLIND REEF, TRIED TO GET OFF COULDN'T SO ABANDONED
STRUCK REEF DURING HEAVY STORM
LOSSCAUSE
CUTTER
BARQUE
CUTTER
BARQUE
SHIP
SHIP
SHIP
SHIP
1880
1895
RIG/HULLDESC LOSSDATE
SCHOONER
1866
1880
CUTTER
1894
BRIGANTINE
1840
SCHOONER
1846
BARQUE
1852
SCHOONER
1865
WRECKNAME
THISTLE
HOPPER BARGE NO. 3
POLLY
MARIA
VICTORIA
MARGARET BROCK
AGNES
(Near to Kingston)
KINGSTON
KINGSTON
(Near to Robe)
THOMSON
DUILIUS
JOSEPH LEE ARCHER
SULTANA
PHAETON
KONING WILLEM II
ALMA
LIVINGSTONE
166
Ardrossan Jetty
1878 (improved in1888)
Point Malcolm
RIG/HULLDESC LOSSDATE
CUTTER
1851
BARQUE
1874
SCREW STEAMER1881
LOSSCAUSE
BLOWN ASHORE
PARTED FROM CABLES AND DRIVEN ASHORE
RAN AGROUND ON REEF, FOUNDERED AND BROKE UP
WRECKNAME
RIG/HULLDESC LOSSDATE LOSSCAUSE
JAMES AND MARGARECUTTER
1878
BOAT SET AFIRE AND COMPLETELY DESTROYED
Germein Bay (Ship) 1879
Germein Bay Jetty 1880
WRECKNAME
RESOURCE
WAVE QUEEN
EURO
Rivoli Bay, Penguin I1878
NO SHIPWRECKS
BARQUE
1860
SCREW STEAMER1862
BARQUE
1879
SCHOONER
1881
LUGGER
1886
KETCH
1892
DRIFTED TOWARDS SHORE WITHOUT ANCHOR AND STRUCK ROCKS
RAN ASHORE ON ROCKS AFTER MISTAKING POSITION OF ALTHORPE IS
PRESUMED TO HAVE STRUCK REEF AND WENT DOWN
VESSEL OVERLOADED, STRUCK BY SQUALL, CAPSIZED & SANK
WENT ASHORE DURING GALE & VESSEL BROKE UP IN SURF
VESSEL SPRANG A LEAK AND SANK
1879
RIG/HULLDESC LOSSDATE LOSSCAUSE
SCHOONER
1878
PARTED MOORINGS DURING SW GALE AND WENT ASHORE
Althorpe Island
1876(improved in 1883)
WRECKNAME
YOUNG ST. GEORGE
(Outside of Althorpe)
FIDES
MARION
ISMYR
EXPERIMENT
PIONEER
ENTERPRISE
NO SHIPWRECKS
167
RIG/HULLDESC LOSSDATE LOSSCAUSE
LAUNCH
1880
SANK
LAUNCH
1895
UNSEAWORTHY CONDITIONS, BLOWN ASHORE IN STORM & BROKE UP
1880
WRECKNAME
KINGSTON
KINGSTON
Rivoli Bay Jetty
Port Victoria Jetty
1883
RIG/HULLDESC LOSSDATE LOSSCAUSE
SCREW STEAMER1859
HIT A REEF DURING FOG
BRIG
1870
WENT ASHORE IN THICK WEATHER BECAUSE OF MISPLACED BEARINGS
Cape Banks
WRECKNAME
ADMELLA
FLYING CLOUD
1882(lit 83) but light ships in there
RIG/HULLDESC LOSSDATE LOSSCAUSE
BARGE
1876
FOUNDERED DUE TO LEAKS, SANK AND BECAME TOTAL LOS
KETCH
1882
DRIVEN ONTO ROCKS DURING NORTHERLY GALE
Lowly Point
WRECKNAME
SARAH
PARARA
1882
RIGDESC
LOSSDATE LOSSCAUSE
DURING SEVERE WEATHER, CAUGHT IN SQUALL, CAPSIZED & SANK
RIG/HULLDESC 1866
Corny Point
WRECKNAME
OMEO
NO SHIPWRECKS
1881/1882
1880
Kingston Jetty
NO SHIPWRECKS
1880
RIG/HULLDESC LOSSDATE LOSSCAUSE
CUTTER
1865
DRIVEN ASHORE DURING HEAVY WESTERLY WEATHER
CUTTER
1870
DURING HEAVY WESTERLY GALE DRAGGED ANCHOR AND DRIFTED ONTO SLAG
1870
VANDALS CUT PAINTER ROPES & DURING STORM DRIFTED ONTO ROCKS
CUTTER
1880
DRAGGED ANCHOR, GROUNDED & CAPSIZED IN HEAVY WESTERLY GALE
YACHT
1895
PARTED MOORINGS DURING SQUALL, WENT ASHORE AND BROKE UP
Wallaroo Jetty
WRECKNAME
BLANCHE
BEN MOROWIE
JUNO
YOUNG EDITH
BANSHEE
168
WRECKNAME
AGNES
EDITH HAVILAND
AEOLUS
(Near Northem)
NENE VALLEY
SOUTHERN CROSS
GLENROSA
LOSSDATE
1876
1877
1894
1854
1880
1890
RIG/HULLDESC
BARQUE
BRIG
SHIP
BARQUE
BARQUE
BARQUE
CAPT MISTOOK LAND FOR A CLOUD & BLOWN ASHORE IN WSW BREEZE
PENGUIN IS LIGHT MISTAKEN FOR CAPE NORTHUMBERLAND STOVE IN
WENT ASHORE IN FOG, HIT REEF AND BROKE UP
LOSSCAUSE
POOR VISIBILITY AND RAN AGROUND
BEARINGS MISTAKEN, STRUCK ROCKS & CANTED OVER
STRUCK REEF, WATER ENTERED AND ABANDONED
169
VALUE OF ELEVATION
Out of data range
Out of data range
Out of data range
Out of data range
2.10904360
-9.02889442
Out of data range
Out of data range
-6.99997520
-5.09540892
-0.88746768
NO VALUE
-8.00847626
-7.59180450
Out of data range
-0.89047605
-7.74730968
Out of data range
-18.06853485
2.13168097
-7.83031845
-5.87265110
Out of data range
WRECKNAME
ADELAIDE
ADELAIDE
ADMELLA
AEOLUS
AGENORA
AGNES
AGNES
AGNES
AGNETA
ALBATROSS
ALBERT
ALMA
ALPHA
ALTERNATIVE
AMELIA
AMY
APOLLO
ARACHNE
ARK
ATALANTA
ATHENS
ATHOL
BANDICOOT
1831
1849
CUTTER
BRIGANTINE
SCHOONER
CUTTER
KETCH
BARQUE
BARQUE
SCHOONER
SCHOONER
BARQUE
SCHOONER
CUTTER
CUTTER
SCHOONER
SHIP
SCHOONER
CUTTER
CUTTER
CUTTER
SHIP
1853
1838
1873
1809
06/1884
1842
1863
1855
1844
1863
1858
1855
1874
1860
1840
1886
20/09/1887
20/03/1864
15/04/1861
29/11/1860
09/11/1881
06/1848
27/12/1889
14/01/1863
15/09/1876
18/07/1876
13/03/1865
09/07/1867
10/08/1848
07/06/1875
15/12/1861
25/07/1847
24/09/1884
08/02/1883
14/11/1890
01/09/1894
06/08/1859
18/04/1861
1874
BUILDDATE LOSSDATE
SCREW STEAMER 17/09/1857
BRIGANTINE
KETCH
RIGDESC
Appendix IV- Elevation data of shipwrecks
ALONG COORONG, 48KM NORTH OF
LACEPEDE BAY
MACDONNELL BAY, 3 MILES SE OF TOWN
CARPENTERS ROCKS, 20 MILES WEST OF
CAPE NORTHUMBERLAND
HELOS REEF, NEAR CAPE BANKS, OPPOSITE
PELICAN POINT
PORT WILLUNGA
NEAR POINT PEARCE, SOUTH OF MOONTA
CARPENTERS ROCKS
MARGARET BROCK REEF
TROUBRIDGE SHOAL, NEAR SULTANA
ONKAPARINGA
ANTECHAMBER BAY, KANGAROO ISLAND
GUICHEN BAY, NEAR ROBE
ROSETTA HEAD
MEMORY COVE, NEAR PORT LINCOLN
AVOID BAY
LARGS BAY, ADELAIDE
BETWEEN TWO HUMMOCKS POINT, LOWLY
POINT, SPENCER GULF
YANERBY, AT NORTH END OF SCEALE BAY.
CALLED TRIAL BAY AT TIME OF LOSS.
DANGEROUS REEF, SPENCER GULF
EMU GREEN, ?RAVINE DES CASOARS,
KANGAROO ISLAND
PORT LINCOLN
PORT ELLIOT, WEST OF POINT COMMODORE
PORT MACDONNELL, 1 KM EAST OF
LOSSLOCATION
170
VALUE OF ELEVATION
-2.37718773
-1.27078974
-1.04659057
0.85420245
-10.18591499
-6.05278397
-2.14095426
-0.15635492
Out of data range
-5.68584728
-5.56407118
-8.06991577
-9.64151096
-6.37973070
Out of data range
-22.79851151
13.59598064
Out of data range
0.16998011
-1.86895525
Out of data range
-16.77322960
-23.30092239
Out of data range
-7.59251642
-1.33323622
-54.92544556
Out of data range
-56.63610077
-5.24150848
WRECKNAME
BANSHEE
BEN MOROWIE
BLACK DIAMOND
BLANCHE
BREEZE
CHARLES CARTER
COMMODORE
CORSAIR
COUNTESS
COWRY
DARING
DART
DART
DOLPHIN
DUILIUS
ECLAIR
EDITH
EDITH HAVILAND
ELIZABETH
ELIZABETH
ELIZABETH REBECCA
EMILY SMITH
EMMA
EMU
EMU
ENDEAVOUR
ENTERPRISE
EURO
EXPERIMENT
FAIRFIELD
1870
BRIGANTINE
SCHOONER
SCHOONER
SCHOONER
CUTTER
KETCH
SCREW STEAMER
SCHOONER
SHIP
BRIG
CUTTER
CUTTER
BRIG
SCHOONER
CUTTER
SCHOONER
SCREW STEAMER
KETCH
BRIG
CUTTER
CUTTER
BARQUE
SCHOONER
CUTTER
BRIG
CUTTER
CUTTER
1883
06/1874
1874
1846
1851
1847
1849
1828
1840
1849
1871
08/1865
1818
1875
1879
1853
1848
1833
1818
15/05/1877
23/08/1840
12/1861
01/05/1853
10/03/1881
06/03/1892
24/08/1881
09/05/1881
09/08/1874
10/04/1845
25/07/1865
06/07/1863
23/02/1854
29/02/1856
15/05/1865
16/08/1876
06/06/1889
10/08/1885
29/03/1838
30/03/1882
08/1872
15/04/1853
14/10/1875
03/08/1897
20/06/1877
15/12/1878
09/1871
25/05/1872
16/12/1895
27/06/1870
BUILDDATE LOSSDATE
SCREW STEAMER 1864
YACHT
CUTTER
RIGDESC
WALLAROO
PORT WALLAROO
WILBERTA REEF, WALRUS ROCKS, NEAR
MOONTA BAY
PORT WALLAROO, NEAR CUSTOM HOUSE
PELICAN LAGOON, KANGAROO ISLAND
TROUBRIDGE SHOAL
COMMODORE POINT, PORT ELLIOT
PORT ADELAIDE, NEAR SITE OF "GRECIAN"
MACDONNELL BAY, NEAR CRESS CREEK
NORMANVILLE
HOG BAY, KANGAROO ISLAND
SOUTHERN END OF TROUBRIDGE SHOAL
ALDINGA, NEAR MYPONGA JETTY
BOSTON BAY, NEAR PORT LINCOLN
GUICHEN BAY
YANKALILLA, NEAR JETTY
SPILSBY ISLAND, NEAR PORT LINCOLN
CARPENTERS ROCKS
PORT GAWLER
WHITING POINT, 3 MILES BELOW TORRENS
YANERBY, NORTH END OF SCEALE BAY.
CALLED TRIAL BAY AT TIME OF LOSS.
MAUPERTUIS BAY
ALDINGA BAY
BETWEEN PORT MACDONNELL & LACEPEDE
PORT ELLIOT
NORTH ARM, PORT RIVER
BETWEEN ALTHORPES AND CAPE FORBIN
BEACHPORT, 9 MILES FROM
6 MILES SOUTH OF ALTHORPE ISLAND
CAPE CASSINI, KANGAROO ISLAND
LOSSLOCATION
171
VALUE OF ELEVATION
-7.63067055
23.82011986
Out of data range
-15.05861759
Out of data range
-0.00735471
Out of data range
Out of data range
-1.24125659
-2.91834664
-11.07156754
Out of data range
Out of data range
3.21365380
Out of data range
Out of data range
Out of data range
-8.52916527
-7.26332474
-17.85399818
-15.25368881
-3.69748116
8.82927704
-1.04844546
10.26129532
-13.95671844
WRECKNAME
FANNIE M
FANNY
FANNY WRIGHT
FIDES
FIRE FLY
FITZJAMES
FLINDERS
FLYING CLOUD
FLYING DUTCHMAN
FLYING FISH
FRANCES
FREEBRIDGE
GALATEA
GAZELLE
GELTWOOD
GEORGE HOME
GLENROSA
GOLDEN HOPE
GOOD INTENT
GOULBURN
GOVERNOR GAWLER
GRECIAN
GRENADA
GULDAX
HALCYON
HARRIET
BARGE
SCHOONER
BARQUE
SCHOONER
BARQUE
BRIG
CUTTER
BARQUE
BARQUE
BARQUE
CUTTER
SCHOONER
BRIG
SCHOONER
CUTTER
SCHOONER
BRIGANTINE
BRIG
CUTTER
HULK
SCHOONER
BARQUE
SCHOONER
SCHOONER
BARQUE
RIGDESC
1876
1848
1840
1841
01/1876
1809
11/1875
1847
1845
1843
1839
1850
1860
1855
1852
1863
1857
1873
1877
11/05/1887
29/06/1856
01/08/1847
13/10/1850
20/06/1856
02/09/1887
13/11/1857
19/05/1856
06/1876
26/04/1851
18/01/1890
03/1894
09/1848
28/07/1851
03/12/1860
29/08/1840
12/06/1877
31/03/1876
04/04/1870
22/10/1866
1891
29/06/1873
22/05/1860
07/08/1877
21/06/1838
15/06/1885
BUILDDATE LOSSDATE
HOG BAY, KANGAROO ISLAND
REEVESBY IS, JOSEPH BANKS GROUP,
OUTER HARBOR, GULF ST VINCENT
PORT WILLUNGA, NORTH OF JETTY
NORMANVILLE
NEAR MURRAY MOUTH 6 MILES BELOW
NEAR SHEOAK FLAT, ORONTES BANK. 7
MILES NE OF POINT VINCENT
RAPID BAY, SECOND VALLEY
KINGSCOTE, KANGAROO ISLAND, 3 MILES
SE POINT MARSDEN
NEAR CAPE BERNOUILLI, 30M. EAST OF
ENCOUNTER BAY
WATERLOO BAY
HALF A KILOMETRE WEST OF SNUG COVE,
KANGAROO ISLAND
ANXIOUS BAY, NEAR ELLISTON
JERVOIS BASIN
PORT MACDONNELL, NEAR CRESS CREEK
NEAR CAPE NORTHUMBERLAND,
CARPENTER ROCKS
PORT ADELAIDE
PORT ELLIOT
NEPTUNE ISLAND, EASTERLY BAY OF EAST
ELLISTON, WATERLOO BAY
PORT MACDONNELL, NEAR CRESS CREEK
NEAR MURRAY MOUTH, 8 MILES NORTH OF
SALT CREEK
12 KMS OFF RIVOLI BAY, GELTWOOD REEF
150M SSE OF KANGAROO ISLAND
CAPE BANKS, 6 MILES EAST OF
RED BANKS, NEPEAN BAY, KANGAROO
LOSSLOCATION
172
VALUE OF ELEVATION
-13.95671844
-3.58145547
Out of data range
-6.99331999
0.23978174
4.89001560
-11.85559559
4.76362610
Out of data range
-12.02000237
-14.53570557
Out of data range
-0.16984865
25.91961288
-0.21415900
Out of data range
Out of data range
-0.38372228
0.07437171
0.00534076
-5.97850657
-12.00408936
Out of data range
Out of data range
Out of data range
15.72283268
-8.36162186
-0.90709996
0.12685199
-1.42977989
WRECKNAME
HARRIET
HARRY
HELEN
HENRY AND MARY
HERO
HOPPER BARGE NO. 3
IDA
INDUSTRY
IRON AGE
IRON KING
ISMYR
J. LOVETT
JAMES AND MARGARET
JANE AND EMMA
JANE FLAXMAN
JOHN ORMEROD
JOSEPH LEE ARCHER
JOSEPHINE LOIZEAU
JUNO
KADINA
KANGAROO
KATE
KINGSTON
KINGSTON
KONING WILLEM II
LADY FERGUSSON
LADY KINNAIRD
LADY OF THE LAKE
LADY WELLINGTON
LAPWING
CUTTER
LAUNCH
LAUNCH
SHIP
CUTTER
BARQUE
BARGE
BRIG
KETCH
SCHOONER
1852
HULK
1808
1868
02/1877
1855
1878
1855
05/1852
1854
01/1867
01/1868
1846
01/1880
1852
1835
1839
1826
1848
1841
1842
06/1844
14/09/1895
04/02/1880
30/06/1857
16/04/1870
21/01/1880
03/10/1877
31/08/1838
06/09/1856
10/02/1873
09/12/1856
09/02/1882
26/05/1861
20/01/1884
21/10/1880
15/01/1857
18/08/1854
15/02/1855
11/12/1873
24/02/1879
19/09/1852
06/12/1878
05/1852
02/05/1850
22/01/1861
09/06/1855
10/07/1856
12/04/1870
20/04/1879
11/05/1887
BUILDDATE LOSSDATE
BRIG
YACHT
CUTTER
YACHT
BARGE
BRIG
KETCH
BARQUE
SHIP
BARQUE
SCHOONER
CUTTER
CUTTER
CUTTER
BRIG
CUTTER
SCHOONER
RIGDESC
NEAR SHEOAK FLAT, ORONTES BANK. 7
MILES NE OF POINT VINCENT
PORT ELLIOT
CARPENTERS ROCKS
ALDINGA BEACH NEAR WILLUNGA JETTY
GLENELG, NORTH OF PATAWALONGA
CAPE JERVIS
PORT WILLUNGA
YANKALILLA BAY
OFF CAPE NORTHUMBERLAND
TROUBRIDGE SHOAL
NEAR REEF HEAD, YORKE PENINSULA
CAPE NORTHUMBERLAND, 12 MILES NORTH
TELOWIE BEACH
ROSETTA HEAD
NEAR SEMAPHORE
OFF CAPE NORTHUMBERLAND
GUICHEN BAY, 2090 YARDS FROM ROBE
PORT ELLIOT
PORT WALLAROO
ANGAS INLET
OFF CAPE ELIZABETH 3/4 MILE NORTH WEST
OF MOONTA
BOSTON IS., NEAR PORT LINCOLN
KINGSTON
PORT CAROLINE, LACEPEDE BAY
GUICHEN BAY, 3 MILES EAST OF ROBE
OFF ALDINGA, NEAR PORT WILLUNGA
CAPE BURR, SPENCER GULF
VICTOR HARBOR
PORT ADELAIDE
PORT ELLIOT
LOSSLOCATION
173
-5.39315844
0.80256867
9.57371521
-4.20231009
Out of data range
MINNIE
MOZAMBIQUE
NASHWAUK
NENE VALLEY
Out of data range
MARGARET BROCK
MIMOSA
-1.78783238
Out of data range
19.23831940
Out of data range
-2.98528600
-31.74847794
-7.28493023
6.07608366
-26.03065491
0.61575049
LITTLE ORIENT
LIVINGSTONE
LOCH SLOY
LOTUS
LOUISE
LUCRETIA
MAID OF AUSTRALIA
MAID OF THE MILL
MAID OF THE VALLEY
MALVINA MAUD
-11.70101547
Out of data range
10.01821995
-12.16607666
1.38229549
36.54886627
-9.30442619
-1.30871296
8.15898132
-0.00368428
Out of data range
-2.56013942
0.02299722
LETTY
MARIA
MARIA
MARINER
MARION
MARION
MARS
MARY
MARY ANN
MELBOURNE
MERCURY
MIAME
MIDGE
VALUE OF ELEVATION
WRECKNAME
1848
1875
1879
1857
1877
1874
1869
1885
1869
1849
1863
YACHT
BARQUE
SHIP
BARQUE
KETCH
1833
1853
1852
1874
22/04/1886
19/08/1854
12/05/1855
19/10/1854
14/04/1884
10/06/1879
28/06/1840
07/11/1845
29/07/1851
11/07/1862
16/06/1885
11/02/1859
24/11/1885
16/11/1859
19/06/1878
24/05/1861
29/06/1853
23/11/1852
20/11/1883
16/12/1861
24/04/1899
23/06/1892
11/1878
08/09/1899
08/07/1899
27/02/1856
13/11/1857
20/02/1876
11/11/1866
BUILDDATE LOSSDATE
CUTTER
BRIGANTINE
1823
SCHOONER
1839
SHIP
1850
SCREW STEAMER 1854
BARQUE
05/1877
CUTTER
1856
CUTTER
PADDLE STEAMER 1852
KETCH
1864
BARQUE
1848
CUTTER
1840
BARQUE
SCREW STEAMER
SHIP
BARQUE
KETCH
SCHOONER
SCHOONER
SCHOONER
KETCH
BRIGANTINE
SCHOONER
SCHOONER
RIGDESC
REDCLIFFS, 15 MILES FROM PORT AUGUSTA
- REMOVED AND GROUNDED NEAR THE
CUSTOMS HOUSE, PORT AUGUSTA
LARGS BAY, ADELAIDE
GUICHEN BAY, NEAR ROBE
MAUPERTIUS BAY KANGAROO ISLAND
MACDONNELL BAY, EAST OF CRESS CREEK
PORT RICKABY, HARDWICKE BAY
BUFFALO REEF, SPILSBY ISLAND, SPENCERS
WARDANG ISLAND
MOUTH OF THE ONKAPARINGA RIVER
PORT WILLUNGA 1/2 MILE NE OF MARTINS
POINT JARROLD, NEAR PORT PIRIE
MARGARET BROCK REEF, 5 1/2 KMS WEST OF
CAPE JAFFA
3.5 MILES OFF LIGHTHOUSE, NEAR
MARGARET BROCK REEF, LACEPEDE BAY
COORONG, 40 MILES BELOW MURRAY
TROUBRIDGE SHOAL 8.5 KMS SE OF
CABLE HUT BAY, 2 NM NE OF CAPE SPENCER
WEST BAY, NEAR CAPE BORDA
PORT WILLUNGA
15 MILES SOUTH-WEST OF MOUNT YOUNG
MURRAY MOUTH, NEAR SPIT AT PORT
PORT PIRIE
MACDONNELL BAY
NEAR CAPE WILLOUGHBY LIGHTHOUSE
STOKES BAY, NORTH COAST OF KANGAROO
ISLAND
GLENELG
COORONG, BELOW SEA MOUTH OF MURRAY
MOANA, NEAR MOUTH OF PEDLARS CREEK
CAPE NORTHUMBERLAND, 8 MILES NORTH
LOSSLOCATION
174
VALUE OF ELEVATION
-8.76093102
Out of data range
-3.27179456
-36.45262527
-4.73591948
Out of data range
-21.62951469
-1.97030735
-8.00708961
Out of data range
-3.16043830
-5.85067463
Out of data range
Out of data range
0.12685199
-0.52787703
Out of data range
Out of data range
-33.89080811
-21.31656265
-32.54788208
-2.18287897
-7.04980230
0.05325653
-0.13758214
-0.00303450
-0.00030583
WRECKNAME
NIMROD
NORA CREINA
O.G.
OMEO
ORIANA
ORWELL
OSMANLI
PARARA
PARSEE
PHAETON
PIONEER
POLLY
PRIMA DONNA
PRINCE OF WALES
PSYCHE
RAMBLER
RED ROVER
RESOURCE
ROSE
ROSE
RUBY
SAN MIGUEL
SANS PAREILLE
SARAH
SARAH
SATURN
SCUD
CUTTER
CUTTER
KETCH
SCHOONER
KETCH
BARQUE
BRIGANTINE
BARGE
YAWL
BARQUE
YACHT
SCHOONER
SCHOONER
BARGE
LUGGER
CUTTER
SHIP
CUTTER
SCHOONER
SCREW STEAMER
SCHOONER
SCREW STEAMER
KETCH
BARQUE
BRIGANTINE
CUTTER
RIGDESC
1874
1877
1839
1842
01/1875
1853
1871
07/1864
1871
1845
1870
1855
1871
1846
1874
1831
1840
1858
1834
1866
17/09/1887
28/06/1851
10/12/1875
15/04/1858
20/07/1890
08/05/1865
30/01/1855
21/12/1876
1879
07/01/1888
23/04/1886
17/01/1877
05/1865
20/06/1873
08/08/1880
24/07/1886
03/11/1894
02/02/1857
23/05/1854
07/1866
29/09/1885
28/04/1873
25/11/1853
10/07/1882
11/11/1838
01/01/1859
26/07/1875
BUILDDATE LOSSDATE
OFF PORT ADELAIDE, OUTSIDE LIGHTHOUSE
ABOUT 17 NAUTICAL MILES WEST OF CAPE
MARTIN (BEACHPORT)
POOLES FLAT, 1.5 MILES NORTH OF SECOND
NEAR CORNY POINT, YORKE PENINSULA
ALDINGA REEF
MACDONNELL BAY 1/4 MILE FROM
D'ESTREES BAY, KANGAROO ISLAND
POINT LOWLY, NEAR PORT AUGUSTA
TROUBRIDGE SHOAL, 4 MILES SE OF ISLET
GUICHEN BAY, 3 MILES EAST OF
BOATSWAINS POINT
MARION BAY
BETWEEN RAPID HEAD AND CAPE JERVIS
BETWEEN PORT MACDONNELL AND PORT
ADELAIDE
MACDONNELL BAY, NEAR PINCHGUT REEF,
EAST OF CRESS CREEK
PORT ADELAIDE ?
BETWEEN WOOL BAY & STANSBURY, 3
MILES NORTH OF WOOL BAY
BETWEEN COFFIN BAY AND PORT LINCOLN
RIVOLI BAY
CARRICKALINGA, 4 MILES AT SEA
BETWEEN BLACK POINT & ADELAIDE
BETWEEN STANSBURY AND PORT
TIPARRA REEF, OFF MOONTA BAY
CAPE JERVIS
POINT LOWLY, NEAR PORT AUGUSTA
CLINTON, NEAR PORT WAKEFIELD
COCKLE SPIT, PORT PIRIE
TORRENS ISLAND
LOSSLOCATION
175
VALUE OF ELEVATION
-0.69637120
-12.12137890
9.04912376
-7.24192190
Out of data range
Out of data range
7.65634727
0.00305263
-32.24257660
-10.14759636
-3.57078409
-5.51053381
Out of data range
0.15950248
Out of data range
-8.54137421
Out of data range
-10.26211929
-6.15517139
-32.21574402
18.87435722
5.86376143
-6.86501312
Out of data range
Out of data range
Out of data range
Out of data range
Out of data range
Out of data range
-41.29790115
WRECKNAME
SECRET
SOLWAY
SOPHIA JANE
SOUTH AUSTRALIAN
SOUTHERN CROSS
ST. MARC
ST. VINCENT
STAG
STAR
STAR OF GREECE
STRANGER
SULTANA
SULTANA
TARNA
TENTERDEN
THISTLE
THOMSON
THREE SISTERS
TIGRESS
TIPARA
TRADER
TREASURE TROVE
TRIAL
TROAS
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
UNNAMED
UNKNOWN
UNKNOWN
UNKNOWN
UNKNOWN
CUTTER
SCREW STEAMER
SCHOONER
CUTTER
KETCH
SNOW
SLOOP
SCHOONER
CUTTER
CUTTER
SHIP
CUTTER
BARQUE
BARQUE
CUTTER
HULK
CUTTER
SHIP
CUTTER
SHIP
BARQUE
BARQUE
CUTTER
SHIP
CUTTER
RIGDESC
1857
1854
05/1883
1843
1837
1874
1840
1854
09/1868
1884
1837
1849
1840
1851
1819
1859
1829
1840
23/12/1893
03/03/1866
10/09/1849
13/03/1899
26/09/1848
02/09/1877
28/09/1866
21/05/1884
01/04/1854
14/05/1865
1841
16/09/1849
1875
1845
08/05/1889
05/10/1839
06/1948
08/01/1880
28/09/1876
09/1844
09/03/1883
1860
13/07/1888
14/06/1898
28/09/1849
27/04/1857
08/12/1837
20/07/1873
21/12/1837
09/1844
BUILDDATE LOSSDATE
POINT BOLINGBROKE, 3 MILES NORTH OF
ROSETTA HEAD,ENCOUNTER BAY
COORONG, 60 MILES FROM RIVER MURRAY
ROSETTA HARBOUR 3 MILES WEST OF
VICTOR HARBOR
POINT DOUGLAS, 4 MILES WEST OF
PORT MACDONNELL, WEST OF FRENCH
MURRAY MOUTH
NORTH ARM, PORT ADELAIDE
SPENCER GULF
BETWEEN PORT WILLUNGA AND BLANCHE
NORTH SIDE OF WEDGE ISLAND
TROUBRIDGE SHOAL
CAPE DOMBEY, OFF CAPE LANNES, NEAR
YORKE PENINSULA OFF DALY HEAD, NEAR
CORNY POINT
BREAKSEA REEF, CAPE NORTHUMBERLAND
CAPE JERVIS
GUICHEN BAY, 2-3 MILES SOUTH OF
LIPSON COVE, 10 MILES FROM TUMBY
SEAFORD, 2 KM SOUTH OF ONKAPARINGA
15 MILES SW OF WARDANG ISLAND
WILLUNGA, GULF ST. VINCENT
KINGSCOTE, KANGAROO ISLAND
LIGHT SHIP NEAR ADELAIDE, NORTH SAND
OFF LAKE BONNEY, NEAR NORTHERN END
HEAD OF THE BIGHT, EAST OF TWIN ROCKS
AVOID BAY
ST PETER ISLAND
FLINDERS ISLAND
FENELON ISLAND
GULF ST. VINCENT, PRECISE LOCATION
LOSSLOCATION
176
VALUE OF ELEVATION
-7.82193947
14.65034389
NO VALUE
-17.41556549
Out of data range
Out of data range
Out of data range
Out of data range
-4.55417633
Out of data range
-22.13594055
-0.42140368
-18.65062714
4.22675514
-10.35175514
Out of data range
-5.76257038
-7.48703766
0.56546605
-4.92666149
-4.82924223
-8.92486763
-25.57737350
-9.53716278
WRECKNAME
UNNAMED LIGHTER
VANQUISH
VAROON
VENTURE
VICTORIA
VULCAN
VULCAN'S CANVAS BOAT
WAITEMATA
WALTER AND JOHN
WAVE QUEEN
WELLING
WILDFLOWER
WILLIAM
WILLIAM
WILLIAM HENRY
WITNESS
YOU YANGS
YOUNG E.B.
YOUNG EDITH
YOUNG HEBE
YOUNG LION
YOUNG ST. GEORGE
YOUNG SURVEYOR
ZANONI
BARQUE
CUTTER
CUTTER
CUTTER
CUTTER
CUTTER
BRIGANTINE
SCREW STEAMER
KETCH
CUTTER
SCHOONER
KETCH
SCHOONER
KETCH
BARQUE
CUTTER
SCHOONER
SHIP
CUTTER
SCHOONER
SCHOONER
CUTTER
SCHOONER
CUTTER
RIGDESC
1865
1874
1856
1871
1833
1842
1850
1856
1882
07/1861
1845
1852
1870
1837
1846
1853
11/02/1867
19/07/1892
11/11/1877
23/08/1847
08/1838
06/1844
01/05/1853
14/06/1890
30/06/1888
21/10/1880
03/1849
18/10/1882
03/01/1878
02/11/1899
06/09/1874
06/07/1892
05/11/1864
01/1856
14/09/1858
09/06/1846
22/04/1845
15/09/1845
20/05/1860
27/02/1871
BUILDDATE LOSSDATE
NORMANVILLE
CAPE JERVIS, 2.5 M WEST OF TALISKER MINE
20 MILES WEST OF CAPE
SNAPPER POINT, KANGAROO ISLAND
CAPE JAFFA
FLINDERS ISLAND
COFFIN BAY
PETREL BAY, ST. FRANCIS ISLAND
POINT BOLINGBROKE
2 MILES NORTH EAST OF JETTY AT
SOUTHEND IN RIVOLI BAY
ALTHORPE ISLAND
WHITING POINT
HOG BAY, NEAR PENNESHAW, KANGAROO
YANKALILLA BAY
BOSTON IS., NEAR PORT LINCOLN
CAPE NORTHUMBERLAND, 1 MILE SE OF
PELORUS ROCKS NEAR CAPE GANTHEAUME
MARINO ROCKS
WALLAROO
20 MILES EAST OF CAPE SPENCER
CAPE ELIZABETH REEF, SOUTH OF MOONTA
ALTHORPE ISLAND
5 MILES WEST OF LONG SPIT BUOY
GULF ST. VINCENT, NEAR LONG SPIT NEAR
ARDROSSAN
LOSSLOCATION

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project