UNIVERSITY OF MIAMI SINGLE-BEAM ACOUSTIC SEABED...
Transcript of UNIVERSITY OF MIAMI SINGLE-BEAM ACOUSTIC SEABED...
UNIVERSITY OF MIAMI
SINGLE-BEAM ACOUSTIC SEABED CLASSIFICATION IN CORAL REEF ENVIRONMENTS WITH APPLICATION TO THE ASSESSMENT OF GROUPER
AND SNAPPER HABITAT IN THE UPPER FLORIDA KEYS, USA
By
Arthur C. R. Gleason
A DISSERTATION
Submitted to the Faculty of the University of Miami
in partial fulfillment of the requirements for the degree of Doctor of Philosophy
Coral Gables, Florida
May 2009
©2009 Arthur C. R. Gleason All Rights Reserved
UNIVERSITY OF MIAMI
A dissertation submitted in partial fulfillment of the requirements for the degree of
Doctor of Philosophy
SINGLE-BEAM ACOUSTIC SEABED CLASSIFICATION IN CORAL REEF ENVIRONMENTS WITH APPLICATION TO THE ASSESSMENT OF GROUPER
AND SNAPPER HABITAT IN THE UPPER FLORIDA KEYS, USA
Arthur C. R. Gleason Approved: ________________ _________________ R. Pamela Reid, Ph.D. Terri A. Scandura, Ph.D. Associate Professor of Marine Geology Dean of the Graduate School and Geophysics ________________ _________________ Eugene C. Rankey, Ph.D. Thomas Hahn, Ph.D. Assistant Professor of Marine Geology Assistant Professor of Applied and Geophysics Marine Physics ________________ _________________ G. Todd Kellison, Ph.D. Jonathan M. Preston, Ph.D. Chief, Fisheries Ecosystems Branch Senior Scientist NOAA Southeast Fisheries Science Center Quester Tangent Corporation
GLEASON, ARTHUR C. R. (Ph.D., Marine Geology and Geophysics)
Single-Beam Acoustic Seabed Classification in (May 2009) Coral Reef Environments with Application to the Assessment Of Grouper And Snapper Habitat in the Upper Florida Keys, USA Abstract of a dissertation at the University of Miami. Dissertation supervised by R. Pamela Reid. No. of pages in text. (173)
A single-beam acoustic seabed classification system was used to map coral reef
environments in the upper Florida Keys, USA, and the Bahamas. The system consisted of
two components, both produced by the Quester Tangent Corporation. A QTCView Series
V, operating with a 50 kHz sounder, was used for data acquisition, and IMPACT
software was used for data processing and classification. First, methodological aspects of
system performance were evaluated. Second, the system was applied to the assessment of
grouper and snapper habitat. Two methodological properties were explored:
transferability (i.e. mapping the same classes at multiple sites) and reproducibility (i.e.
surveying one site multiple times). The transferability results showed that a two-class
scheme of hard bottom and sediment could be mapped at four sites with overall accuracy
ranging from 73% to 86%. The locations of most misclassified echoes had one of two
characteristics: a thin sediment veneer overlying hard bottom or within-footprint relief on
the order of 0.5 m or greater. Reproducibility experiments showed that consistency of
acoustic classes between repeat transects over the same area on different days varied, for
the most part, between 50% and 65%. Consistency increased to between 78% and 92%
when clustering was limited to two acoustic classes, to between approximately 70% and
100% when only echoes acquired within two degrees of nadir in the pitch direction were
used, and to between 81% and 87% when a limited set of features was used for
classification. The assessment of grouper and snapper habitat addressed the question
whether areas of high fish abundance were associated with characteristic acoustic or
geomorphological signatures. The results showed, first, that the hard bottom / sediment
classification scheme was a useful first step for stratifying survey areas to increase
efficiency of grouper census efforts. Second, an index of acoustic variability
complemented the hard bottom / sediment classification by further targeting areas of
potential grouper habitat. Finally, five grouper and snapper spawning aggregation sites
were all found to have similar associations with drowned shelf edge reefs in the upper
Florida Keys.
iii
ACKNOWLEDGEMENTS
Funding for this work was provided by Office of Naval Research grant
N000140110671 to Pamela Reid and grants from the NOAA Coral Reef Initiative to
Anne-Marie Eklund, Todd Kellison, and Margaret Miller. Financial support was also
provided by a University of Miami Fellowship and the Yamaha Contender Miami Billfish
Tournament Circle of Friends Fellowship. Charlie and Lisa Evans generously donated the
use of their vessel GRITS for much of the work in the Florida Keys.
The Florida Keys National Marine Sanctuary and Biscayne National Park issued
permits FKNMS-2007-008 and BISC-2007-SCI-0021.
Diving, boat operations, data acquisition, and data analysis were conducted with
the assistance of the following. From UM/RSMAS: Mike Anderson, Grant Basham,
Christine Bauer, Chris Boynton, Marilyn Brandt, Albert Chapin, Cassie Clark, Manuel
Collazo, Vanessa Damoulis, Meghan Dick, Daniel Doolittle, Megan Fairobent, Ryan
Freedman, Mike Feeley, Jack Fell, Brooke Gintert, Rick Gomez, Jimmy Herlan,
Veronique Koch, Phil Kramer, Shawn Lake, Bob Loos, Eric Louchard, Miguel
McKinney, Julie Mintzer, John Parkinson, Dave Powell, Laura Rock. From
NOAA/SEFSC: Heather Balchowsky, Neil Baertlein, Sean Cimulluca, Joe Contillo, Leah
Harman, Doug Harper, Tom Jackson, Jack Javech, Lindsey Kramer, Dave McClellan,
Mark Miller, Jen Schull, Mark Vermeij, Dana Williams. From the Perry Institute of
Marine Science / Lee Stocking Island: Craig & Tara Dahlgren, Brian Kakuk. From the
U.S. Navy's Atlantic Undersea Testing and Evaluation Center: Matt Accordino, Henry
Buerkert, Marc Ciminello, Tom Szlyk. From NOAA/AOML: Paul Dammann and Jules
Craynock. The crew of the Coral Reef II: Mike Fielder, Keith Pamper, Lou Roth, John
Rothchild
iv
Roberto Torres provided the locations of the Ocean, Watson's, and Davis reef
sites described in Chapter 6. Roberto's willingness to share the experiences of decades of
fishing the Florida Keys was both illuminating and crucial to testing the wider
applicability of results from the initial surveys at Carysfort reef.
I enjoyed many conversations about seabed mapping with Bernhard Riegl, of the
National Coral Reef Institute at Nova Southeastern University. Descriptions of the seabed
classes he mapped near Cabo Pulmo, Mexico helped provide context for some of the
results in Chapter 3. Bernhard’s generous loan of a National Instruments data acquisition
board enabled simultaneous operation of multiple Quester Tangent systems during a
survey in November 2006.
The contributions and hospitality of Ben Biffard and Steve Bloomer, of the
University of Victoria, are gratefully acknowledged. Ben conducted the BORIS model
runs that were used in Section 4.7. Steve participated in the Andros Island survey. Both
provided constructive comments on drafts of Chapter 4 and insight into single-beam
seabed classification techniques.
The assistance of the Quester Tangent Corporation marine division is also greatly
appreciated. Karl Rhynas and Rick Pearson provided detailed training and
troubleshooting for hardware and software glitches. Glenda Wyatt provided crucial
support in organizing the Second Acoustic Seabed Classification Workshop at RSMAS in
April 2006. The patience and insight of Bill Collins and Jon Preston proved essential.
Both Bill and Jon offered constructive criticism, instruction, and guidance many times
over the duration of this project, but two particular instances are worth mentioning. Bill
provided on-site training and guidance for and assistance with data collection at Lee
v
Stocking Island (Chapter 3). Jon provided libraries for accessing the IMPACT data
structures, which were necessary to create Figures 4.12, 6.2, to perform the analysis in
Section 4.5.2.2, and, more generally, to acquire an intuitive insight into the echo
classification process.
The support of my committee members Pamela Reid, Gene Rankey, Thomas
Hahn, Todd Kellison, and Jon Preston could not have been greater. I have enjoyed
interacting with each of them and deeply appreciate their assistance and availability,
especially when consulted in the face of a looming deadline. Particular thanks are due to
Pam for exhibiting tremendous trust and unflagging support while letting me explore my
own topics of interest. It has been a great pleasure to work with her.
Finally, I cannot express enough gratitude for the support of my family, including
my parents, Paul and Phyllis, my grandmother, Isabelle, and my wife Louise.
vi
TABLE OF CONTENTS
LIST OF FIGURES ......................................................................................................... ix
LIST OF TABLES .......................................................................................................... xii
Chapter 1: Introduction .................................................................................................1 1.1 What can be mapped with single-beam acoustic seabed classification in
the coral reef environment? ................................................................5 1.2 What applications can utilize single-beam acoustic seabed classification
in the coral reef environment?............................................................7 1.3 Outline ................................................................................................................8
Chapter 2: Introduction to the QTC acoustic seabed classification system............13
Chapter 3: Consistency of single-beam acoustic seabed classification among multiple coral reef survey sites .........................................................16
3.1 Background .......................................................................................................16 3.2 Previous Work...................................................................................................18 3.3 Methods ..............................................................................................................22
3.3.1 Acoustic surveys .........................................................................................23 3.3.2 Acoustic classification ................................................................................24 3.3.3 Identification of acoustic classes.................................................................27 3.3.4 Accuracy assessment...................................................................................27
3.4 Results ..............................................................................................................29 3.5 Analysis: where were the errors? ....................................................................36
3.5.1 Lee Stocking Island.....................................................................................36 3.5.2 Andros Island Area .....................................................................................43 3.5.3 Carysfort Reef and Fowey Rocks ...............................................................46
3.6 Discussion ..........................................................................................................46 3.7 Conclusions........................................................................................................55
Chapter 4: Reproducibility of single-beam acoustic seabed classification under variable survey conditions.................................................................58
4.1 Background .......................................................................................................58 4.2 Previous work....................................................................................................59 4.3 Methods ..............................................................................................................62
4.3.1 Survey site and data acquisition..................................................................62 4.3.2 Vessel attitude and grazing angle computation...........................................63 4.3.3 Acoustic data processing.............................................................................66 4.3.4 Classification reproducibility ......................................................................67 4.3.5 Echo, FFV, and Q-value correlation with survey parameters.....................71
4.4 Results ..............................................................................................................74 4.4.1 Environmental conditions, vessel attitude and grazing angle .....................74
vii
4.4.2 Classification reproducibility ......................................................................77 4.4.3 Echo, FFV, and Q-value correlation with survey parameters.....................86
4.5 Attempts to improve reproducibility ..............................................................91 4.5.1 Cluster into fewer classes............................................................................92
4.5.1.1 Merge baseline classes four through six into a single class .................92 4.5.1.2 ACE-2...................................................................................................93
4.5.2 Keep only near-nadir echoes.......................................................................95 4.5.2.1 Within 5 degrees of vertical (echoes stacked by 5) ..............................96 4.5.2.2 Pitch within 2 degrees of vertical (echoes stacked by 5)......................97 4.5.2.3 Within 5 degrees of vertical (echoes stacked by 1) ..............................99 4.5.2.4 Pitch only within 2 degrees of vertical (echoes stacked by 1) ...........101 4.5.2.5 Incidence angle less than 5 degrees (echoes stacked by 1) ................102
4.5.3 Stability of principal components .............................................................104 4.5.3.1 Robust PCA........................................................................................104 4.5.3.2 Consistent PCA ..................................................................................107 4.5.3.3 Dataset Size ........................................................................................108
4.5.4 Features least subject to ping-to-ping variability ......................................110 4.6 Summary and Conclusions ............................................................................116
Chapter 5: Acoustic signatures of the seafloor: tools for predicting grouper habitat ...............................................................................................119
5.1 Background .....................................................................................................119 5.2 Methods ............................................................................................................121
5.2.1 Acoustic Survey ........................................................................................123 5.2.1.1 Data Collection and Seabed Classification.........................................123 5.2.1.2 Acoustic Variability Index .................................................................125
5.2.2 Diver Survey .............................................................................................126 5.2.3 Comparison of Acoustic and Diver Surveys.............................................127
5.2.3.1 Acoustic Classification Accuracy Assessment...................................127 5.2.3.2 Grouper Abundance vs. Acoustic Classification and Variability.......128
5.3 Results ............................................................................................................129 5.3.1 Acoustic Survey ........................................................................................129 5.3.2 Diver Survey .............................................................................................129 5.3.3 Acoustic Classification Accuracy Assessment .........................................129 5.3.4 Grouper Abundance vs. Acoustic Classification and Variability .............130
5.4 Discussion ........................................................................................................131 5.5 Conclusions......................................................................................................135
Chapter 6: Geomorphology of grouper and snapper spawning aggregation sites in the upper Florida Keys, USA .............................................137
6.1 Background .....................................................................................................137 6.2 Methods ............................................................................................................139
6.2.1 Acquisition and acoustic classification .....................................................140 6.2.2 Identification of acoustic classes...............................................................142 6.2.3 Diver-based assessment of classification at Carysfort Reef .....................143
viii
6.2.4 Locations of FSAs relative to seabed features ..........................................144 6.2.5 Geomorphologic “signatures” of FSAs.....................................................145
6.3 Results ............................................................................................................145 6.3.1 Acquisition and acoustic classification .....................................................145 6.3.2 Identification of acoustic classes...............................................................146 6.3.3 Diver-based assessment of classification at Carysfort Reef .....................148 6.3.4 Locations of FSAs relative to seabed features ..........................................148 6.3.5 Geomorphologic “signatures” of FSAs.....................................................152
6.4 Discussion ........................................................................................................153 6.5 Conclusions......................................................................................................156
Chapter 7: Conclusions ..............................................................................................158
References ............................................................................................................163
ix
LIST OF FIGURES
Figure 2.1: Flowchart of QTCV processing. ..................................................................14 Figure 3.1: Map showing locations of Lee Stocking Island (LSI), Carysfort Reef
(CF), Fowey Rocks (FR), and Andros Island (AI) study sites. ...............23 Figure 3.2: Underwater photographs of pole-mounted transducer, video camera
housing, and sample frames grabbed from underwater video. ................28 Figure 3.3: Sediment grain size distributions for samples from the survey areas. .........32 Figure 3.4: LSI acoustic classification and video classification plotted on top of a
true-color IKONOS image of the LSI study area. ...................................33 Figure 3.5: Andros acoustic classification and video classification plotted on top of a
true-color IKONOS image of the study area. ..........................................33 Figure 3.6: Fowey rocks acoustic classes and diver estimated substrate........................35 Figure 3.7: Within-frame accuracy for the LSI dataset. .................................................37 Figure 3.8: Depth-frequency histogram for sediment-dominated video frames in the
LSI dataset. ..............................................................................................37 Figure 3.9: Classified acoustic tracks, video frames, and dive sites in the Adderly Cut
portion of LSI study area. ........................................................................38 Figure 3.10: Selected underwater photographs from the Adderly Cut portion of the LSI
study area.................................................................................................41 Figure 3.11: Within-frame overall accuracy histogram for the Andros dataset. ............43 Figure 3.12: Oblique underwater photographs of four seabed types in the Andros Island
survey area...............................................................................................45 Figure 3.13: Within-frame overall accuracy histograms for the Carysfort and Fowey
Rocks datasets. ........................................................................................46 Figure 3.14: Plot of overall accuracy as a function of the number of acoustic classes. .51 Figure 3.15: Example of the utility of rock / sediment seabed classification in
interpreting bathymetry to predict fish habitat. .......................................52 Figure 3.16: Cross shelf profile of part of the Navassa Island insular shelf. ..................54 Figure 4.1: Fowey Rocks survey site and depth profiles. ...............................................63 Figure 4.2: Wind speed and water temperature for the periods of the six surveys.........74 Figure 4.3: Boxplots of daily mean and maximum within-stack attitude
measurements. .........................................................................................75 Figure 4.4: Transducer attitude data from May 28, 2007. ..............................................76 Figure 4.5: Q-space of each of the six daily datasets clustered independently...............79 Figure 4.6: All 18 replicates of both the northern and southern transects. .....................80 Figure 4.7: Independently classified replicate track lines lines, Q-space, and pitch
measurements along the northern transect on May 2, 2007. ...................83 Figure 4.8: Classified track lines, Q-space, and pitch measurements along the transect
for independent clustering of each pass along the northern transect on two days...................................................................................................84
Figure 4.9: Renumbering of daily classes.......................................................................85
x
Figure 4.10: Plots illustrating the locations of four stations from which echoes, FFVs, Q-values, and survey parameters were extracted. ........................87
Figure 4.11: Plot of multidimensional angle between echo envelopes versus the distance between echoes for all pairwise comparisons at all four test stations. ....................................................................................................89
Figure 4.12: Plot of multidimensional angle between echo envelopes versus the minimum grazing angle at the times of center echo of each stack. .........89
Figure 4.13: Plot of the magnitude of the difference between echoes versus the minimum grazing angle at the times of center echo of each stack. .........90
Figure 4.14: Plot of the magnitude of the difference between FFVs versus the minimum grazing angle at the times of center echo of each stack. .........90
Figure 4.15: Plots of daily Q-space (left) and renumbering of daily classes (right) for datasets classified by ACE with just two clusters. ..................................94
Figure 4.16: Plot of Q-space for each day and renumbering of daily ACE classes after filtering out all stacks with maximum transducer pointing vector greater than 5 degrees off vertical. ..........................................................96
Figure 4.17: Plot of Q-space for each day and renumbering of daily ACE classes after filtering out all stacks with max pitch greater than 2 degrees off vertical. ....................................................................................................98
Figure 4.18: Plot of Q-space for each day and renumbering of daily ACE classes after filtering out all unstacked echoes with max transducer pointing greater than 5 degrees off vertical. ....................................................................100
Figure 4.19: Plot of Q-space for each day and renumbering of daily ACE classes after filtering out all unstacked echoes with pitch greater than 2 degrees. ..................................................................................................101
Figure 4.20: Plot of Q-space for each day and renumbering of daily ACE classes after filtering out all unstacked echoes with incidence angle greater than 5 degrees. ..................................................................................................103
Figure 4.21: Plot illustrating the concept of robust PCA..............................................106 Figure 4.22: Plots illustrating the concept of robust PCA with more points. ...............107 Figure 4.23: Echogram created from the BORIS dataset. ............................................111 Figure 4.24: BORIS dataset FFVs, FFV coefficient of variation, and loadings of the
first principal component computed from the FFVs. ............................112 Figure 4.25: Q-space for the BORIS dataset using seven different subsets of features
input to the PCA. ...................................................................................113 Figure 4.26: FFVs of the May 1, 2007 Fowey rocks dataset. .......................................114 Figure 5.1: Acoustic survey track lines superimposed on an IKONOS satellite
image of Carysfort Reef and surroundings............................................122 Figure 5.2: Overview of acoustic processing................................................................124 Figure 5.3: Illustration of the computation of acoustic variability. ..............................125 Figure 5.4: The three main acoustic classes and diver-estimated substrate at Carysfort
Reef........................................................................................................130
xi
Figure 5.5: Grouper abundance at each dive site relative to acoustic classification and acoustic variability.................................................................................132
Figure 5.6: Acoustic classification and acoustic variability computed from the echoes closest to each dive site and grouped by the presence / absence of groupers. ................................................................................................133
Figure 5.7: Depth, acoustic class, and acoustic variability along transect A-A’, shown in Figure 5.5...............................................................................................134
Figure 6.1: Map of the study area. ................................................................................139 Figure 6.2: Mean echoes for the six acoustic classes at Davis Reef.............................146 Figure 6.3: Acoustic classification and satellite image for the Davis Reef survey
area. .......................................................................................................147 Figure 6.4: Visualization of processed acoustic data surrounding the Carysfort Reef
survey area.............................................................................................149 Figure 6.5: Visualization of processed acoustic data surrounding the Watson’s Reef
survey area.............................................................................................150 Figure 6.6: Visualization of processed acoustic data surrounding the Davis Reef survey
area. .......................................................................................................151 Figure 6.7: Visualization of processed acoustic data surrounding the Ocean Reef survey
area. .......................................................................................................152
xii
LIST OF TABLES
Table 1.1: Technologies for surveying coral reef environments at different scales. ........2 Table 1.2: Summary of systems, frequencies, sediment mineralogy, and seabed classes
used in previous single-beam ASC studies. ............................................10 Table 3.1: Seabed classes reported by previous efforts to map coral reef environments
with single-beam acoustic seabed classification systems........................21 Table 3.2: Characteristics and settings of the QTCV system used in this study. ...........24 Table 3.3: Values of tunable parameters used when processing each survey with the
IMPACT software. ..................................................................................25 Table 3.4: Acoustic class labels and sizes (by percent of total echoes) in each survey
area and aggregation of classes at fine descriptive resolution to coarse descriptive resolution...............................................................................30
Table 3.5: Error matrices for acoustic hard bottom / sediment classification at multiple sites............................................................................................34
Table 3.6: Diver descriptions of Adderly Cut dives from July 2002..............................39 Table 3.7: Probe depths and rugosity for dive sites in Adderly Cut. ..............................40 Table 3.8: Error matrix for the LSI survey excluding hard bottom sites covered with
a thin sediment veneer. ............................................................................42 Table 3.9: Error matrix for the Andros survey excluding hard bottom sites covered
with a thin sediment veneer. ....................................................................44 Table 4.1: Optimum number of clusters identified by ACE for each daily dataset
clustered separately. ................................................................................77 Table 4.2: Optimum number of clusters identified by ACE for each transect when
clustered separately. ................................................................................78 Table 4.3: Overall accuracy and Kappa coefficient between pairs of daily classified
datasets. ...................................................................................................85 Table 4.4: Percent AMI for the reclassed ACE-best and the original ACE-best
dataset. .....................................................................................................86 Table 4.5: Summary of visual assessment of correlation between envelopes, FFVs,
or Q-space and survey variables..............................................................88 Table 4.6: Overall accuracy and Kappa coefficient between pairs of daily classified
datasets reclassed to 6 classes but then with classes 4-6 merged to a single class...............................................................................................93
Table 4.7: Percent AMI for daily classified datasets reclassed to 6 classes but then with classes 4-6 merged to a single class. ...............................................93
Table 4.8: Overall accuracy and Kappa coefficient between pairs of daily datasets classified by ACE with just two clusters. ................................................94
Table 4.9: Percent AMI for pairs of daily datasets classified by ACE with just two clusters. ....................................................................................................94
Table 4.10: Summary of off-nadir experiments..............................................................96
xiii
Table 4.11: Overall accuracy and Kappa coefficient between pairs of datasets clustered by day after filtering out all stacks with maximum transducer pointing vector greater than 5 degrees off vertical. ...............................................97
Table 4.12: Percent AMI between pairs of datasets clustered by day after filtering out all stacks with maximum transducer pointing vector greater than 5 degrees off vertical. .................................................................................97
Table 4.13: Overall accuracy and Kappa coefficient between pairs of datasets clustered by day after filtering out all stacks with maximum pitch greater than 2 degrees off vertical. .................................................................................99
Table 4.14: Percent AMI between pairs of datasets clustered by day after filtering out all stacks with maximum pitch greater than 2 degrees off vertical. ........99
Table 4.15: Overall accuracy and Kappa coefficient between pairs of datasets clustered by day after filtering out all unstacked echoes with maximum transducer pointing vector greater than 5 degrees off vertical. ...............................100
Table 4.16: Percent AMI between pairs of datasets clustered by day after filtering out all unstacked echoes with maximum transducer pointing vector greater than 5 degrees off vertical. ....................................................................100
Table 4.17: Overall accuracy and Kappa coefficient between pairs of datasets clustered by day after filtering out all unstacked echoes with pitch greater than 2 degrees. ..................................................................................................102
Table 4.18: Percent AMI between pairs of datasets clustered by day after filtering out all unstacked echoes with pitch greater than 2 degrees. ........................102
Table 4.19: Overall accuracy and Kappa coefficient between pairs of datasets clustered by day after filtering out all unstacked echoes with incidence angle greater than 5 degrees. ...........................................................................103
Table 4.20: Percent AMI between pairs of datasets clustered by day after filtering out all unstacked echoes with incidence angle greater than 5 degrees. .......103
Table 4.21: Overall accuracy and Kappa coefficient between daily classified datasets created from a single PCA using the median eigenvector of all six days........................................................................................................108
Table 4.22: Percent AMI for daily classified datasets created from a single PCA using the median eigenvector of all six days. .................................................108
Table 4.23: Optimum number of classes as determined by ACE as a function of dataset size for the entire merged Fowey rocks dataset.....................................109
Table 4.24: Optimum number of classes as determined by ACE as a function of dataset size for the Watson’s Reef dataset.........................................................109
Table 4.25: Subsets of features and their colors as shown in Figure 4.24C, and Figure 4.25. .......................................................................................................114
Table 4.26: Overall accuracy and Kappa coefficient between daily classified datasets using the ACE clustering after computing Q space with only features 1-15........................................................................................................115
Table 4.27: Percent AMI for the datasets using ACE clustering after computing Q space with only features 1-15 ............................................................116
xiv
Table 4.28: Summary of experiments and range of the majority of OA, Kappa, and AMI values in each................................................................................117
Table 5.1: Characteristics and settings of the QTCV system used in this study. .........131 Table 6.1: Survey areas and the spawning aggregations in each area. .........................140 Table 6.2: Characteristics and settings of the QTCV system used in this study. .........141 Table 6.3: Values used for tunable parameters in IMPACT software for processing
QTCV echoes. .......................................................................................142 Table 6.4: Clustering results for the four survey areas. ...............................................145 Table 6.5: Agreement between proposed FSA site model criteria and observations of
the sites surveyed...................................................................................154
1
Chapter 1: Introduction Darwin (1842) classified coral reefs around the world into three
geomorphological types: barrier, fringing, and atoll, and in so doing pioneered the use of
mapping to help understand the development of coral reefs. Maps depicting thematic
seabed classes have been important tools for coral reef research at least since Agassiz
(1885). Today, maps of coral reef habitats are commonly used for developing marine
protected areas, planning field surveys, and educational outreach, among other
applications (U.S. Coral Reef Task Force Mapping and Information Synthesis Working
Group (USCRTFMISWG) 1999; Green et al. 2000). Despite their importance, however,
until recently habitat maps had been produced for only a minute fraction of the world’s
coral reefs. Due to the combined utility and lack of availability of habitat maps in coral
reef environments, mapping and monitoring was the first of four duties assigned to the
U.S. Coral Reef Task Force by Executive Order 13089 (63 FR 32701, 1998 WL 313072).
In conjunction with advances in technology and research (Mumby et al. 2004), the high
priority placed on mapping by the Coral Reef Task Force has enabled mapping of
shallow coral reefs to proceed rapidly over the past several years (see
http://eol.jsc.nasa.gov/Reefs for a review of shallow reef mapping initiatives).
A multi-resolution paradigm has evolved for mapping coral reefs with airborne or
satellite optical imagery (Mumby and Harborne 1999; USCRTFMISWG 1999). Under
the multi-resolution strategy, large areas are mapped at low thematic and spatial
resolution then selected smaller areas of interest are filled in using technologies with
higher thematic and spatial resolution. Deep or turbid water limits the utility of seabed
mapping with overhead imagery, however, to those shallow reefs located in clear water.
The multi-resolution approach, which is a sensible compromise between the needs of
2
global products (large areas, low cost) and local products (high spatial and thematic
resolution), has been enacted at a programmatic level for optical satellite and aerial
imagery (USCRTFMISWG 1999) and could serve as a model for mapping deeper
environments with acoustic technologies (Table 1.1).
Table 1.1: Technologies for surveying coral reef environments at different scales. A multi-resolution strategy of mapping with optical imagery exists for shallow reefs. A logical approach would be to implement a similar hierarchical approach for deep coral ecosystems with acoustic technologies. This dissertation focuses on the question of whether single-beam classification methods are appropriate for mapping deep reefs at the global / regional scale (upper right box).
Areas that cannot be mapped with satellite or aerial imagery are both extensive
and ecologically important. For example, over 55% of the Florida Keys National Marine
Sanctuary (about 1540 square nautical miles) has not been mapped due to water depth or
clarity limitations (FMRI 1998). The Tortugas Bank, Pulley Ridge, and Flower Garden
Banks are three examples from the Gulf of Mexico that illustrate the potential of
luxuriant communities of shallow-water (zooxanthellate) corals to exist at “mesophotic”
depths of 30-75 m (Miller et al. 2001; Hickerson and Schmahl 2005; Jarrett et al. 2005).
In addition, coral communities can exist below the photic zone, where deep-water
(azooxanthellate) corals form mounds up to several hundred meters high. Recent ocean
exploration initiatives indicate that such deep corals are much more extensive than
previously thought (Roberts et al. 2006). Deep corals provide important habitat for fishes,
and shallow coral species may potentially find refuge from warming surface waters at
mesophotic depths (Riegl and Piller 2003).
3
In order to improve the understanding of reef resources in optically deep water,
which in this context means deeper than can be mapped with overhead imagery,
alernative mapping technologies must be used in place of aerial or satellite imagery.
Acoustic mapping systems are a natural solution to mapping optically deep water.
Existing acoustic technologies that could be employed for mapping deep reefs differ in
their cost, resolution, and, therefore, potential coverage, however. One challenge is that
many areas need mapping, and no single acoustic technology will be cost effective on all
spatial scales. A logical approach would be to employ acoustic survey methods in a
hierarchical manner similar to that used for optical imagery. With this approach, a single-
beam system would fill the equivalent role in optically deep water that moderate
resolution satellite imagery fills in optically shallow water (Table 1.1).
Single-beam acoustic seabed classification (ASC) systems are potentially
attractive for mapping large areas at relatively low descriptive resolution. ASC systems,
also known as acoustic ground discrimination systems (Foster-Smith and Sotheran 2003;
Kenny et al. 2003), are relatively inexpensive to purchase and to operate, portable, and
easy to use (Anderson et al. 2008). Thus, ASC has appealing advantages for mapping or
exploration of deep-water coral ecosystems. The long-term vision implied by the multi-
resolution mapping paradigm is that a network of ships deploying commercially available
single-beam ASC systems would, at minimal marginal cost, acquire the data necessary to
produce coral ecosystem habitat maps on a regional to global scale in optically deep
water. As attractive as the long-term vision sounds, there are many challenges, both
methodological and applied, to address before ships of opportunity could systematically
contribute to regional-scale mapping of deep coral ecosystems (Anderson et al. 2008).
4
The optical remote sensing community has, for some time, explicitly
distinguished between development or evaluation of remote sensing methodology and use
of remotely sensed data for particular applications. Green et al. (1996), for example,
reviewed state-of-the-art in remote sensing of tropical resources, including coral reefs,
with satellite and aerial imagery. One of the conclusions of Green et al. (1996) was that
research to that point had heavily emphasized the development and evaluation of novel
techniques with few examples of transitioning these techniques for use in real world
problems. Eight years later, Andrefouet and Riegl (2004) edited a special issue of the
journal “Coral Reefs” in which the number of papers was evenly split between
methodological development and application to science or management. Andrefouet and
Riegl (2004) acknowledged the progress that had been made in a short time transitioning
remote sensing technology from “the tool without application to the mandatory tool.” At
the same time, however, Andrefouet and Riegl (2004) noted that none of the papers in
their special issue had used acoustic remote sensing technology, despite specific effort to
solicit such submissions. Andrefouet and Riegl (2004) pointed out that additional effort
was needed to develop and apply acoustic remote sensing for coral reef science and
management.
The purpose of this dissertation research was to assess the utility of single-beam
ASC for mapping coral reef environments, thereby beginning to address the gap
identified by Andrefouet and Riegl (2004). Acknowledging the importance of both
methodology and application, the overall goal of the research was to make a contribution
in both areas. Thus, the chapters of this dissertation fall into two groups, addressing two
basic questions: What can be mapped with ASC in the coral reef environment? And, what
5
are some example applications of ASC in the coral reef environment? Addressing
questions of both methodology and application is important because the utility of a
technology is defined by its application to real world problems, yet to meaningfully apply
a technological solution one first must have a detailed understanding of the strengths and
weaknesses of the approach.
1.1 What can be mapped with single-beam acoustic seabed classification in the coral reef environment?
The question “what can be mapped” captures the spirit of the methodological
section of this study, but it is overly simplistic, as ASC systems have been available for
about 15 years, and many areas have been mapped using them at several frequencies
between 12-200 kHz (Table 1.2). Two things distinguish this study from previous work
with regard to the question "what can be mapped": first, the geology of the areas being
mapped, and second, an emphasis on standardizing methodology among multiple
surveys. Focusing on coral reef environments and on standardizing methodology refines
the simple question of “what can be mapped” to define specific objectives for this
dissertation.
The focus on coral reef environments distinguishes this study from most other
investigations of ASC performance, which have, for the most part, considered siliciclastic
environments (87% of papers in Table 1.2) and exclusively soft sediment communities
(67% of papers in Table 1.2). Only a handful of previous ASC studies have focused on
carbonate reef environments (Murphy et al. 1995; Hamilton et al. 1999; White et al.
2003; Moyer et al. 2005; Riegl and Purkis 2005; Riegl et al. 2007). One reason to expect
ASC might perform differently in reef environments is that carbonate and siliciclastic
sediments typically differ in depositional texture, fabric, and postdepositional alteration
6
(diagenesis). It is not surprising, therefore, that the geoacoustic properties of surficial
carbonate sediments have been shown to differ from those of siliciclastics (Richardson et
al. 1997; Brandes et al. 2002). Many maps made with ASC in soft sediment communities
have distinguished sediment facies primarily based on grain size (see citations in Freitas
et al. 2003b). In contrast, coral reef ecologists will tend to want subdivisions of rocky
facies, grouping all sediment together as a single class (Mumby and Harborne 1999;
Franklin et al. 2003). The few ASC studies that have included rocky seabeds (Sotheran et
al. 1997; Pinn and Robertson 1998; Bax et al. 1999; Bornhold et al. 1999; Anderson et al.
2002; Pinn and Robertson 2003; Brown et al. 2005) generally consider rock as a single
class. Only Moyer et al (2005) and Riegl and Purkis (2005) have assessed the accuracy of
ASC maps that included multiple hard bottom classes, suggesting that opportunities exist
for further technological progress in this area.
Anderson et al. (2008) proposed a list of ten priorities for research that would
advance the field of acoustic seabed classification. At least five of these priority research
areas fall under the general topic of standardization of instruments and methods.
Transferability, and reproducibility are two aspects of standardization of instruments and
methods that are investigated in this dissertation. First, ASC should be transferable in the
sense that the classes mapped should have the same meaning regardless of location or
instrumentation used for data collection. Second, ASC should be reproducible such that
successive measurements by the same sensor over a short time period produce stable
results. Transferability, and reproducibility have not been well quantified for single-beam
ASC in general, and have not been addressed at all for ASC in coral reef environments.
7
Chapter 3 addresses the question of single-beam ASC transferability. Chapter 4 addresses
the question of single-beam ASC reproducibility.
1.2 What applications can utilize single-beam acoustic seabed classification in the coral reef environment?
In the context of distinguishing between methodology and application (Green et
al. 1996; Andrefouet and Riegl 2004), as discussed above, all previous studies using
single-beam ASC in coral reef environments (Murphy et al. 1995; Hamilton et al. 1999;
White et al. 2003; Moyer et al. 2005; Riegl and Purkis 2005; Riegl et al. 2007) have been
purely methodological. The products of these studies were maps of the distribution of
sedimentary and outcropping rock facies. An application of remotely sensed data, in
contrast, would use facies or habitat maps as a starting point or as an intermediate step
toward some other objective. In this dissertation, the assessment of grouper and snapper
habitat in the upper Florida Keys, USA, was chosen as an example application of single-
beam ASC in coral reef environments.
Eklund et al. (2000) observed a black grouper aggregation just outside of a
protected no-take marine reserve in the Florida Keys. If the existence of this aggregation
had been known or suspected during the reserve planning process, it could have been
considered for inclusion within the no-take reserve boundaries. One conclusion by
Eklund et al. (2000) was that seabed habitat maps for depths greater than 20 m in the
Florida Keys were needed to assist the stratification of reef fish census effort. A second
conclusion by Eklund et al. (2000) was that rapid means for identifying grouper habitat,
in particular possible sites of grouper spawning aggregations, would be beneficial for
conservation efforts. Chapters 5 and 6 begin to address the needs identified in Eklund et
al. (2000) by investigating the potential of a commercial single-beam ASC system to a)
8
create seabed substrate maps for depths greater than 20 m in the upper Florida Keys, and
b) associate seabed geomorphology with grouper and snapper habitat.
1.3 Outline This dissertation focuses on the question of whether single-beam classification
methods are appropriate for mapping deep reefs at the global to regional scale (upper
right box of Table 1.1). Several commercial-off-the-shelf, single-beam ASC systems are
available that could be used for this purpose (QTC1, RoxAnn2, BioSonics VBT3,
Kongsberg SEABEC4). All of the work in this dissertation was completed using a system
produced by the Quester Tangent Corporation (QTC; Sidney, BC, Canada). The QTC
system consisted of two complementary components. Data acquisition was performed
with a QTCView Series V (QTCV) operating with a 50 kHz sounder. Data processing
and classification were preformed with QTC IMPACT software. Chapter 2 introduces
terminology, parameters, and processing steps associated with the QTC system.
Chapters 3 and 4 address the methodological questions related to producing
transferable, and reproducible classification schemes in a coral reef environment (Section
1.1). Chapter 3 addresses single-beam ASC transferability. An abridged version of
Chapter 3, not including the Andros results or Section 3.5, has been published (Gleason
et al. 2009) and a version containing the omitted sections is being prepared for
publication. Part of the Chapter 3 discussion refers to a survey performed at the Navassa
Island National Wildlife Refuge. The details of the survey are not included in this 1 http://questertangent.com/m_prod_view.html
2 http://www.sonavision.co.uk/pages/seabed_classification_menu.html
3 http://www.biosonicsinc.com/vbt.shtml
4 http://www.sonavision.co.uk/pages/seabed_classification_menu.html
9
dissertation but have been published by Miller et al. (2008). Chapter 4 concerns single-
beam ASC reproducibility. The results from Chapter 4 are in preparation for publication.
Chapters 5 and 6 explore the application of single-beam ASC to the mapping and
assessment of grouper and snapper habitat in the upper Florida Keys, USA (Section 1.2).
Chapter 5 presents the results of mapping the aggregation site observed by Eklund et al.
(2000) and its surroundings. Grouper abundances were measured to assess whether the
seabed classes identified by the ASC could stratify the survey area in a way that would
efficiently allocate grouper census effort. Chapter 5 was published as Gleason et al.
(2006). The acoustic variability calculation described in Chapter 5 has been incorporated
into the Classification and Mapping Software (CLAMS; v. 1.1, Quester Tangent Corp.,
Sidney, B.C. Canada, 2004) where it is called the "complexity index."
In Chapter 6 the sites of five known historical grouper and snapper spawning
aggregations in the upper Florida Keys were surveyed using the same commercial single-
beam ASC system. The resulting maps of seabed classes for each site were interpreted,
and similar geomorphological features were found to be associated with each
aggregation. Chapter 6 has been submitted to The Professional Geographer and is
currently in review.
10
Table 1.2: Summary of systems, frequencies, sediment mineralogy, and seabed classes used in previous single-beam ASC studies. Abbreviations used in the System column were: “QTCIII”, “QTCIV” and “QTCV” for QTCView Series III, IV and V, respectively; “Rox” for RoxAnn; “Echo+” for Echoplus, and "ISAH-S" for the precursor to the QTC systems. Entries under the Frequency column are in kHz. Abbreviations under the “Si/CO3” column indicate the predominant mineralogy of sediments in the study area: “Clastic” for siliciclastic, and “CO3” for carbonate. Some of the data for these columns were not explicitly stated in the cited reference but were assumed based on the study’s location or date. Citation System Freq. Si/CO3 Classes used or identified (Anderson et al. 2002) QTCIV 38 Clastic 4 classes of training data (Sand,
gravel, rock, rock with macroalgae), additional 4 classes found in post processing due to expert guidance.
(Bax et al. 1999) Rox 120 Clastic Hard, soft, rough. (Bornhold et al. 1999) QTCIV 200 Clastic Bedrock, sediment veneer,
sand/gravel with boulders, sand/gravel and muddy sand.
(Bloomer et al. 2007) QTCIV 50 Clastic some CO3
3 case studes. Classes mostly based on grain size.
(Brown et al. 2005) Rox 200 Clastic 6 classes based on grain size and rock.
(Collins and McConnaughey 1998)
QTCIV 38 / 120
Clastic Classes not specified.
(Ellingsen et al. 2002) QTCIV 50 Clastic some CO3
6 classes mostly related to grain size, but “sediment characteristics could not alone explain the diversity of acoustic classes.”
(Foster-Smith and Sotheran 2003)
Rox 38 / 200
Clastic 12 biotypes plus rocky.
(Foster-Smith et al. 2004) QTC(IV?) and Rox
200 / 200
Clastic 4 classes based mostly on grain size; 8 classes based on grain size and bedforms.
(Freeman et al. 2002) QTCIV 120 Clastic 5 acoustic classes correlated with their habitat complexity index, which was derived from grain size + ss texture + mud fraction + burrows + wt. of stones.
(Freeman and Rogers 2003) QTCIV 200 Clastic 11 classes: grain size and bedforms. (Freitas et al. 2003a) QTCIV 50 Clastic 3 classes: coarse, fine, very fine. (Freitas et al. 2003b) QTCIV
QTCV 50/50 w/diff width
Clastic 3 classes: fine with silt, fine without silt, mud.
(Freitas et al. 2008) QTCV 50 / 200
Clastic 50: 3classesmed sand, fine sand, mud; 200: not correlate with sediment
(Galloway and Collins 1998) QTCIV 38 / 200
Clastic Mud, sand, gravel - qualitative evaluation
(Gleason et al. 2006) QTCV 50 CO3 7 classes, but 3 main ones. Hardbottom, two sediment classes.
(Greenstreet et al. 1997) Rox 38 Clastic 6-7 based on grain size
11
Table 1.2 (continued): Summary of systems, frequencies, sediment mineralogy, and seabed classes used in previous single-beam ASC studies. (Hamilton et al. 1999) QTCIV
Rox 38 / 50 Clastic
and CO3 5 QTC, 10 Rox. QTC classes based on grainsize (supervised).
(Hetzinger et al. 2006) QTCV 200 Mostly CO3, some clastic
4 classes: fine-med sand, med-coarse sand, coarse sand-granule, v.coarse sand-granule.
(Hewitt et al. 2004) QTCIV 200 Mixed? 5 classes based on biotic communities.
(Hutin et al. 2005) QTCIV QTCV
38 / 50 Clastic 3-5 classes correlated with depth. Once depth was removed via regression, the residuals were able to map scallop beds.
(Magorrian et al. 1995) Rox 50 Clastic Mud/silt (soft) and mussels, scallops. (Morrison et al. 2001) QTCIV 200 Mixed? Mostly grain size, but only
qualitatively assessed via video ground truth
(Moyer et al. 2005) QTCV 50 Mixed 5 classes, mostly hard/soft. ~60% accuracy hard / sediment
(Murphy et al. 1995) Rox Not given
CO3 7 classes, not specified exactly. Some were coral, sand, patchreef, fine sand, seagrass on sand.
(Pinn et al. 1998) Rox 38 Clastic 24 classes in three surveys based on grain size, rock, and “unknown.”
(Pinn and Robertson 1998) Rox Not given
Clastic No actual classes, just E1, E2.
(Pinn and Robertson 2001) Rox Not given
Clastic No actual classes, just E1, E2.
(Pinn and Robertson 2003) Rox 38 Clastic Between 4-9 classes based on grain size, rock, and “unknown” depending on interpolation and track spacing.
(Preston et al. 1999) QTC(IV?) 38 / 200
Clastic Classes not labeled, but instead correlated with linear combinations of geotechnical variables. 38 kHz correlated with volume variables and 200 kHz correlated with surface vars.
(Preston et al. 2004b) QTCIV 38 / 120
Clastic Unspecified, exactly, but some combination of sands.
(Riegl et al. 2005b) QTCV Echo+
50 / 200
Mixed? Sand, seagrass, (dense/sparse) algae. Experiments with algae over fixed spots / dropped baskets.
(Riegl et al. 2005a) QTCV 50 / 200
Mixed? Sand, seagrass, algae.
(Riegl and Purkis 2005) QTCV 50 / 200
CO3 50: Rock and sediment; 200: high and low relief; combined them for 4 classes
(Riegl et al. 2007) QTCV / Echo+
50 Mixed 4 classes: two hardbottom, two sediment. Sediment appears to be divided by sorting.
(Sotheran et al. 1997) Rox Not given
Clastic 14 classes mostly based on bedrock, cobbles / boulders and sand.
(Tsemahman et al. 1997) ISAH-S 50 Clastic Primarily based on grain size. Highest correlation with geotech was corer penetration.
12
Table 1.2 (continued): Summary of systems, frequencies, sediment mineralogy, and seabed classes used in previous single-beam ASC studies. (von Szalay and McConnaughey 2002)
QTCIII QTCIV
38 / 38 Clastic 6 at site one, 13 at site two based on grain size
(White et al. 2003) Rox 200 CO3 Mud, sand, everything else (rock). (Wilding et al. 2003) Rox 200 Clastic No actual classes, just E1, E2.
13
Chapter 2: Introduction to the QTC acoustic seabed classification system
The Quester Tangent Corporation (QTC; Sidney, BC, Canada) produces several
commercial-off-the-shelf acoustic seabed classification (ASC) systems comprised of both
hardware and software components. The surveys described in Chapters 3 to 6 all used a
QTCView Series V (QTCV) for data acquisition and the QTC IMPACT (version 3.4)
software for data processing and classification.
The QTCV data acquisition system consists of an echo sounder, transducer, head
amplifier, analog to digital (A/D) signal capture card and software to record the digitized
echoes. A personal computer is required to house the A/D card, run the software and
store the digitized data. A wide variety of echo sounder/transducer combinations can be
used with the system. Examples in the literature include surveys at 24, 38, 50, 120, and
200 kHz (Table 1.2). The system used in this work operated at 50 kHz.
The QTC IMPACT software processes pre-recorded echoes using three general
steps (Fig. 2.1). In step (A) the Hilbert transform computes the echo envelope from the
full waveform (bipolar) data recorded during data acquisition. All echoes are scaled to a
maximum amplitude of one, and a depth compensation routine is applied to normalize the
echo length by subsampling the echo envelope at different rates depending on the depth
at which it was acquired (Preston 2004). A user-selectable number (5 by default) of
consecutive echoes are averaged (“stacked”) to reduce stochastic ping-to-ping variability.
Finally, a series of algorithms (Fig. 2.1) are used to compute features characterizing the
echo shape; echo amplitude is not used in the QTCV analysis.
Data analysis step (B) reduces the dimensionality of the dataset. Usually the
features generated in step (A) are highly correlated, so principal components analysis
14
(PCA) is used to transform the data set. IMPACT retains the first three principal
components regardless of the percent of dataset variance explained. In the system
documentation, these principal components are given the shorthand names “Q1”, “Q2”,
and “Q3”, thereby forming three-dimensional “Q-space”. For consistency with the
documentation and with the majority of existing literature using QTC IMPACT, these
“Q” names will be retained below, but their meaning should be clear; each point in Q-
space corresponds to a single, possibly stacked, echo in geographic space. Points close to
one another in Q-space have similar shapes, and points far apart in Q-space have different
shapes, but these shapes cannot be recreated from Q-space, nor does a point in Q-space
have an a priori physical interpretation (e.g. gravel vs. sand).
Figure 2.1: Flowchart of QTCV processing. QTCV analysis consists of three steps A) feature extraction, B) data reduction, C) clustering. See text for discussion.
The data are clustered in analysis step (C) to form discrete classes corresponding
to bottom types that differ acoustically. The clustering algorithm used is based on a
15
simulated annealing routine (Preston et al. 2002; Preston et al. 2004a), which produces
the statistically optimum cluster membership for a given number of clusters. The usual
procedure is to determine the optimum cluster arrangement for a range of numbers of
clusters, then choose which number of clusters is best based on the Bayesian Information
Criterion (Preston et al. 2004a; Preston et al. 2004b). The class assignments can be
mapped in geographic space because each echo (point in Q-space) has geographic
coordinates associated with it. Finally, the parameters of the data analysis steps (B) and
(C) are stored in a file known as a QTC “catalog”. The catalog contains the PCA
projection matrix and the mean/covariance matrix for each cluster. Additional data sets
can be processed using this catalog, or subjected to their own PCA/clustering steps.
No single comprehensive description of the IMPACT system has been published,
which hinders communication of results to those who have not personally used the
system. The following brief list of sources should help readers unfamiliar with the QTC
system to gain insight into the processing without reviewing the large literature on single-
beam seabed classification in general. Hamilton et al. (1999; section 1.1) have a concise
description of the underlying principle behind classification based on echo shape. ICES
(2007) provides background material on many aspects of acoustic seabed classification.
The system described by van Walree et al. (2005) is similar to the IMPACT processing
although different features are generated to describe the echo shape. Preston (2004)
described the depth compensation route used in IMPACT. Preston et al. (2004a) provided
an overview of the features calculated by IMPACT and the simulated annealing
clustering routine.
16
Chapter 3: Consistency of single-beam acoustic seabed classification among multiple coral reef survey sites
3.1 Background Coral reef-associated habitats that cannot be mapped with airborne or satellite
imagery are both extensive and ecologically important. For example, over 55% of the
Florida Keys National Marine Sanctuary (about 1540 square nautical miles) has not been
mapped due to water depth or clarity limitations (FMRI 1998). Not all of this unmapped
area contains reefs, of course, so the most basic mapping goal is to search this large area
in a cost effective way for the purposes of discriminating ecologically distinct habitats
such as reef, hard bottom, seagrass, and bare sediment, and prioritizing areas of interest
for detailed surveys.
Acoustic systems are a natural solution for mapping optically deep water, which
in this context means anywhere overhead imagery cannot effectively map the seabed.
Commercial single-beam acoustic seabed classification (ASC) systems could potentially
contribute to the problem of mapping large areas of the seabed in optically deep water
because they are inexpensive relative to other acoustic mapping tools, portable, easy to
operate, and, at least conceivably, operable from ships of opportunity thereby enabling
coverage of certain large areas at minimal marginal cost (Anderson et al. 2008).
Even though single-beam ASC systems have been commercial products since the
early-1990s (Chivers et al. 1990) and have been used in a variety of mapping projects
around the world (Table 1.2), the methods of data acquisition, processing, classification,
and validation continue to improve. Indeed, a recent review of acoustic seabed
classification technologies identified the standardization of methodologies and the
measurement of seabed variability at multiple spatial scales as two priority areas for
17
future research (ICES 2007; Anderson et al. 2008). Standardizing ASC methods would
permit comparisons of data collected in different areas or within the same area at
different times, potentially by different systems (Anderson et al. 2008). As discussed in
Section 1.1, the improvements to ASC methodologies suggested by Anderson et al.
(2008) fall into three categories: objectivity, transferability, and reproducibility. This
chapter concerns the transferability of ASC results.
Previous evaluations of single-beam acoustic seabed classification systems in
coral reef environments have used data from just one study site (Section 3.2).
Consequently, the classification schemes employed have been different in every case
(Table 3.1). In contrast, this study compared results across multiple study sites to assess
what seabed types could be mapped reliably by a particular single-beam ASC system in
coral reef environments in general, rather than at a single specific site.
The overall objective of this study was to determine what acoustic seabed classes
were consistently distinguished among survey areas in coral reef environments. The
methods employed were, first, to survey multiple sites using the same 50 kHz single-
beam ASC system, second, to cluster the acoustic data from each site independently,
third, to associate the clusters of acoustic data with seabed types, and finally, to quantify
the accuracy of the resulting classified map (Section 3.3). The results showed that hard
bottom (rocky) and sediment substrate were common classes identified at all sites
(Section 3.4). An analysis of misclassified portions of the study areas revealed that thin
sediment veneers and within-footprint macrorelief were common sources of errors in the
acoustic classes (Section 3.5).
18
The conclusions of this study should be broadly applicable to mapping coral reefs
and reef-associated habitats. Even though a simple hard bottom /sediment classification
scheme discriminates only two types of seabed, the fact that this scheme is transferable
among sites proves useful in several ways, as discussed in Section 3.6 and exploited in
Chapter 6. Furthermore, documenting the transferability of a single, albeit simple, single-
beam acoustic seabed classification scheme is a step toward the standardization of ASC
methods recommended by Anderson et al. (2008).
3.2 Previous Work Even though this is the first study to compare single-beam ASC maps created at
multiple sites, previous works have investigated the performance of single-beam ASC
systems within single coral reef survey areas (Murphy et al. 1995; Hamilton et al. 1999;
White et al. 2003; Moyer et al. 2005; Riegl and Purkis 2005; Riegl et al. 2007).
Reviewing the classes found in these previous surveys suggests what seabed classes
might be common to all coral reef sites.
Murphy et al. (1995) surveyed Biscayne National Park, in South Florida, USA,
using a RoxAnn system. The sonar frequency was not specified. Murphy et al. (1995)
proposed that seven seabed communities could be mapped with the system, including
different densities of seagrass. Only five specific classes were actually described,
however, by Murphy et al. (1995): coral, patch reef / hard bottom, seagrass on sand, sand,
and fine sand. No accuracy assessment was provided.
Hamilton et al. (1999) used two single-beam ASC systems, a Quester Tangent
Series IV (QTCIV) and RoxAnn, to map a portion of the Great Barrier Reef lagoon, near
Cairns, Australia. The RoxAnn system operated at 50 kHz, while the QTCIV operated at
19
38 kHz. Five classes were mapped with QTCIV: soft mud, plastic mud, silty mud, muddy
gravel, and muddy, coarse, sandy gravel. The RoxAnn data were divided into 16 classes,
merged to eight classes, and ultimately five classes: coarse sand and gravel, muddy sand
and gravel, rough plastic muds, fine sand/sand/mud, and muds. Qualitative assessment
with ground truth suggested that the QTCIV class boundaries were consistent with
changes in grain size, but that the RoxAnn classes were not. No quantitative accuracy
assessment was provided. Hamilton et al. (1999) felt that both systems produced erratic
signals over hard bottom, but they did not include any hard bottom classes in their final
maps.
White et al. (2003) used a RoxAnn system at 200 kHz to map an area offshore
Negros Occidental, Philippines, from 2.6 to 70 m depth. They produced four habitat maps
with different numbers of classes, ranging from 10 classes with 28% overall accuracy to
three classes with 86% overall accuracy. The classes for the coarsest level were mud,
sand, and coral dominated.
Riegl and Purkis (2005) used a Quester Tangent Series V (QTCV) at both 50 and
200 kHz to map coral communities from 1-8 m depth in a ramp setting off the coast of
Dubai. They found that the 50 kHz data discriminated hard bottom from sediments and
that the 200 kHz data discriminated high and low rugosity. They compared their results
with a habitat map made from satellite imagery and found 66% agreement. Comparison
against ground truth yielded 56% agreement.
Moyer et al. (2005) used QTCV at 50 kHz to map drowned Holocene reefs and
surrounding facies between 3 to 35 m depth off the coast of Broward County, South
Florida, USA. After clustering their data into six acoustic classes representing rubble, two
20
types of sediment, and three types of reef they found only 39% accuracy. Moyer et al.
(2005) also consolidated acoustic clusters into two coarser classification schemes with
only three or two classes and found 61% and 64% accuracy, respectively.
Riegl et al. (2007) used two single-beam acoustic seabed classification systems,
QTCV and Echoplus, both operating at 50 kHz, on the same survey in a mixed carbonate-
siliciclastic setting offshore of Cabo Pulmo, Mexico. The QTCV data were divided into
two hard bottom classes, “rocky ridges” and “rock and hardground”, and two sediment
classes, “less sorted sand” and “well sorted sand”. The Echoplus data were divided into
two classes, hardgrounds and sand. Accuracy relative to video images of the seabed was
reported as 90% but no error matrix was provided.
No single seabed class was identified in all of the previous single-beam acoustic
surveys in coral reef environments (Table 3.1). Only a few seabed classes, such as
“sand”, “coral”, and “hard substrate” were identified by more than one of the previous
studies (Table 3.1). “Sand,” found in four of the previous surveys, was the most
commonly identified class, but only when subclasses identified by Murphy et al. (1995)
and Riegl et al. (2007) were combined. The lack of consistency between surveys suggests
that acoustic classes derived from single-beam echo sounders are not transferable among
sites. Fusing the classes found in previous studies into just two coarse-level classes,
however, reveals that five of the previous studies found at least one class that could be
described as “hard bottom” substrate and all of the previous studies had at least two
classes that, taken together, could be described as “sediment” substrate (Table 3.1).
Previous results suggest, therefore, that aggregating classes to a coarser level of detail
might be required to generate consistency between surveys.
21
Table 3.1: Seabed classes reported by previous efforts to map coral reef environments with single-beam acoustic seabed classification systems. The first row contains headings corresponding to the descriptive resolution (DR) of each study. The second row lists the studies, which were all at a fine level of DR, as well as headings for two columns of aggregated classes constructed from the previous studies. The last row contains the overall accuracy (OA) if available or N/A if not reported.
DR Coarse Medium Fine Study Aggregation
of previous results
Aggregation of previous results
Murphy et al. (1995)
Hamilton et al. (1999)
White et al. (2003)
Riegl & Purkis (2005)
Moyer et al. (2005)
Riegl et al. (2007)
Coral Coral Dominated
High Rugosity Hard
Hard Substrate
Rocky Ridges
Hard Bottom Patch
Reef / Hard Bottom
Low Rugosity Hard
Rock and Hardground
Seagrass Seagrass on Sand
Muddy Coarse Sandy Gravel
Gravel
Muddy Gravel
Sand Sand High Rugosity Soft
Sand Less Sorted Sand
Sand Fine Sand
Low Rugosity Soft
Well Sorted Sand
Silty Mud
Pastic Mud
Mud
Classes
Sediment
Mud
Soft Mud
OA N/A N/A N/A N/A 86% 56% 64% “90%”
Green et al. (1996) used the term “descriptive resolution” to refer to the capability
of a sensor to discriminate habitats, but it more generally refers to the level of detail (i.e.
number of distinct classes) provided by a classified map (Mumby et al. 1997; Mumby
and Harborne 1999). In the terrestrial remote sensing literature, common synonyms for
“descriptive resolution” are “thematic resolution” (e.g. Castilla et al.) or simply “level of
detail” (Congalton and Green 1999). The observations that no acoustic class mapped in
22
any previous studies were common to more than a few sites but that “hard bottom” and
“sediment” classes, created as supersets of the classes in each study, were common to all
sites suggest that single-beam acoustic seabed classifications may be transferable
between sites only at a coarse level of descriptive resolution.
Aggregating similar seabed classes produces a hierarchical classification scheme
with multiple levels of descriptive resolution ranging from fine (many subtly
distinguished classes) to coarse (a few general classes). Given the expectation, based on
results of previous studies, that the level of descriptive resolution limits transferability of
classification schemes between sites, the classes found in each of the survey areas of this
study were aggregated to form a hierarchical classfication scheme with two levels: a
coarse level composed of “hard bottom” and “sediment” classes and a fine level
composed of whichever classes were statistically discriminated from the data by the
clustering process.
3.3 Methods Four sites were mapped with the same QTCV system operating at the same
frequency and classified with QTC IMPACT software, using the same overall procedure.
First, an acoustic survey was conducted (Section 3.3.1). Second, the acoustic data were
clustered into groups based on echo shape (Section 3.3.2). Third, each of the clusters was
labeled (Section 3.3.3). Fourth, classification was assessed with respect to ground truth
data (Section 3.3.4). The details of each of these steps, which are described in the
following four sections, varied among the surveys, but all of the surveys contained each
of the four basic steps.
23
3.3.1 Acoustic surveys
The 50 kHz QTCV system described in Chapter 5 was used to acquire the
acoustic data for all four surveys (Table 3.2). Two of the sites investigated were in the
Bahamas, in shallow water on the Great Bahama Bank (Fig. 3.1). The other two sites
were located in the reef crest and forereef environments of the upper Florida Keys (Fig.
3.1).
Survey 1 was conducted in the vicinity of Lee Stocking Island (LSI), Bahamas, on
June 16-20, 2001 (Fig. 3.1). Approximately 145 km of track lines were acquired along
the bank top in water depths ranging from one meter to just over 10 m.
Figure 3.1: Map showing locations of Lee Stocking Island (LSI), Carysfort Reef (CF), Fowey Rocks (FR), and Andros Island (AI) study sites. Dashed line is the 200 m bathymetric contour.
Survey 2 was conducted east of Andros Island, Bahamas (Fig. 3.1) on October 16-
18, 2001. Approximately 73 km of track lines were acquired in depths from 1 to about 8
m. One portion of the bank was covered with a grid of 19 approximately 1500 m long
transects spaced 100 m apart. A second portion of the bank top was covered by nine
24
widely-spaced, cross-shelf transects that, in turn, were connected by segments of an
along-shelf transect formed while transiting between the cross-shelf areas.
Survey 3 was conducted on March 14, 28, and April 4, 2002 offshore of Carysfort
Reef, in the Florida Keys (Fig. 3.1). Fifty-two parallel transects, each about 2 km long,
were run across the upper shelf between depths of 3 to 35 m. The total track length,
including tie lines was approximately 124 km. Further details of this survey are described
in Chapter 5.
Survey 4 was conducted at Fowey Rocks, also in the Florida Keys, approximately
45 km north-northeast of Carysfort (Fig. 3.1). The Fowey survey was conducted October
12 and 20, 2003 and included forty-one parallel transects across the upper shelf between
depths of 3 to 40 m. The total track length, including tie lines, was approximately 72 km.
Table 3.2: Characteristics and settings of the QTCV system used in this study. Parameter Value Sounder model Suzuki 2025 Frequency 50 kHz Power 500 W Echo pulse length 0.3 ms Ping rate recorded 1.5 Hz (approx) Transducer model Suzuki TGN60-50B-12L Beam width (cross track) 42 degrees Beam width (along track) 16 degrees
3.3.2 Acoustic classification
Acoustic data were processed with the QTC IMPACT software package (version
3.4, QTC, Sidney, BC, Canada, 2004). Details of the IMPACT processing were described
in Chapter 2 and have been previously published (e.g. Preston et al. 2004a; Gleason et al.
2006; Freitas et al. 2008), so only a review of the essentials will be given here.
25
First, the raw, bi-polar waveforms were converted to echo envelopes (the
instantaneous amplitude of the time series as computed by the Hilbert transform).
Second, the echoes were depth compensated using the algorithm described by Preston
(2004). The implementation of the Preston (2004) depth compensation algorithm within
IMPACT uses the sounder / transducer characteristics and two main parameters, the
standard echo length (SEL) and the survey depth (Tables 3.2, 3.3).
Table 3.3: Values of tunable parameters used when processing each survey with the IMPACT software. Parameter (units) LSI AI CF FR Standard Echo Length (samples) 100 100 170 170 Echoes deeper than Critical Depth (N/A)
No No Yes Yes
Survey depth (m) Equal to Critical Depth
Equal to Critical Depth
60 60
Stack size (# echoes) 5 5 5 5 Auto cluster iterations (N/A) 30 15 30 40 Auto cluster class range (N/A) 2-10 2-15 2-12 2-10
Third, echoes were stacked (i.e. averaged) to smooth some of the natural ping-to-
ping variability in echo shape. All of the data from surveys for this paper were stacked by
five (Table 3.3), which means that echoes one through five were averaged to form stack
one, echoes six though ten were averaged to form stack 2, and so on. In the remainder of
the description of the methods, the results, and the discussion below, all of which concern
processed data, the term “echoes” actually implicitly means “stacked echoes.”
Fourth, 166 features were computed for each (stacked) echo. IMPACT uses five
types of measurements to generate the 166 features: 1) cumulative amplitude, 2)
amplitude quantiles, 3) amplitude histogram, 4) power spectrum, and 5) wavelet packet
transform (Preston et al. 2004a). Although the exact algorithms used to compute these
26
features remain proprietary, they are metrics of “the shape and spectral character of the
echo” (Preston 2004). Chapter 2 contains further discussion of the features.
The features computed by IMPACT are always highly correlated, so the fifth
processing step was to reduce the dimensionality of the feature dataset using principal
components analysis (PCA). The PCA algorithm converts the [N x 166] feature matrix,
where N is the number of echoes and 166 the number of correlated features, into a [N x
166] principal components matrix, where the 166 components are mutually orthogonal
(see e.g. Davis 1986 for details on transforming data matrices with PCA). The 166
principal components are ordered by the amount of variance each explains in feature
space, with the first principal component explaining most of the variance, the second
explaining most of the remaining information, and so on. IMPACT arbitrarily retains
only the first three principal components, thereby resulting in a truncated principal
components matrix. QTC calls this [N x 3] data matrix “Q-space,” and the three
dimensions, the first three principal components, are called Q1, Q2, and Q3 (Preston
2004; QTC 2004).
Sixth, echoes were clustered into classes based on their distribution in Q-space.
IMPACT’s autocluster function performs clustering using a simulated annealing routine
(Preston et al. 2004a). The outputs of IMPACT’s autocluster function are: a) the optimum
number of clusters into which the echoes should be split, as defined by the Bayesian
Information Criterion (Preston et al. 2004a), b) the mean vector and covariance matrix
defining each cluster in Q-space, and c) a class number for each echo assigning it to one
of the defined clusters.
27
3.3.3 Identification of acoustic classes
IMPACT’s auto-clustering routine divides a dataset into distinct clusters based on
echo shape, but, like any unsupervised classification routine, it cannot give these clusters
descriptive names (e.g. reef, rubble, seagrass etc..). The clusters output from IMPACT,
therefore, must be labeled by reference to other data sources.
Cluster labeling for these surveys employed comparison to satellite imagery,
notes taken while snorkeling or drift diving, statistics of the echoes themselves (i.e.
looking at individual echoes and mean echo shapes for each cluster), bathymetric cross
sections, sediment grain size measurements, and reference to previous seabed
classifications at these sites (Gonzalez and Eberli 1997; FMRI 1998; Lidz et al. 2003;
Louchard et al. 2003; Mobley et al. 2004).
Once the individual classes were labeled, similar classes were aggregated to form
a second classification at each site consisting only of hard bottom and sediment classes.
Thus, the classes identified for each survey area fell into a hierarchy with two levels of
descriptive resolution: a fine level, within which the number of classes could vary among
sites, and a coarse level, consisting of just two classes per site.
3.3.4 Accuracy assessment
Once the acoustic clusters had been labeled, they were quantitatively compared
with “ground truth.” Two types of data were collected for assessing the accuracy of the
acoustic classification: video images and diver-based observations. Downward looking
video images were acquired during the LSI and the Andros surveys where water was
shallow and clear enough that the seabed was always visible from the surface. In contrast,
the seabed was not visible from the surface at all times during the Carysfort and Fowey
28
rocks surveys, due to deeper and less clear water. Therefore, at Carysfort and Fowey
diver-based observations were acquired.
Video data from the LSI and Andros surveys were acquired with a Sony TRV 900
camera in an underwater housing that was mounted to the same pole that supported the
transducer used for the survey (Fig. 3.2). The camera was set to time-lapse mode, so that
an entire day’s worth of surveying could fit on one videotape. In time-lapse mode, the
camera acquired video (at full frame rate) for two seconds and then paused for 28
seconds. At the beginning of each survey day, the camera’s clock was synchronized with
a GPS unit. Using the time code embedded with each frame, the locations of the frames
were determined from the GPS tracks recorded for each day.
The LSI and Andros video data sets consisted of 1502 and 558 two-second long
clips, respectively. Each clip was viewed and a single frame was extracted for analysis.
For each of these frames, an analyst visually estimated what percent of the frame
consisted of rocky substrate, what percent was covered by sandy substrate, and what
percent was covered by rubble.
Figure 3.2: Underwater photographs of pole-mounted transducer, video camera housing, and sample frames grabbed from underwater video. Left: Pole-mounted transducer and video camera housing, as used for the LSI and Andros surveys. Left middle: Sample frame grabbed from the video over a sandy seabed. Right middle and right: Sample frames grabbed over low relief and high relief hard bottom, respectively.
The diver-based observations collected at Carysfort Reef and Fowey Rocks
followed the Bohnsack and Bannerot (1986) stationary reef visual census (RVC) method.
29
The RVC protocol focuses on the collection of fish population data, but habitat data,
including estimates of the percent of the seabed covered by rock, rubble, and sediment,
are also collected (McClellan and Miller 2003). Estimates of substrate were the portion of
the RVC dataset used for this analysis.
Comparison of the acoustic classification with the video / diver estimates of
substrate was accomplished by computing the overall accuracy from an error matrix
constructed for each survey site (Congalton and Green 1999). The comparison was made
between each ground truth sample and the closest acoustic echo to that point. One
refinement of the standard error matrix technique was necessary because the video / diver
data was expressed as a fraction; the substrate at each point was X% sediment, Y% hard
bottom, and Z% rubble. The acoustic classes, on the other hand, were discrete, so each
entry in the error matrix was divided proportionally by the video / diver-estimated
substrate. Chapter 5 has a sample calculation (Table 5.1).
3.4 Results The LSI data clustered into nine acoustic classes. Echoes were not evenly
distributed among the nine classes, however. Four of the classes contained 96.3% of the
echoes, and the other five classes comprised just 3.7% of the echoes (Table 3.4).
Furthermore, the five small (“minor”) classes did not exhibit any spatial coherence; they
were scattered throughout the study area. Efforts to label the classes, assess their
accuracy, and analyze the causes of errors therefore focused on the most populous
(“major”) classes; the minor classes were excluded from further analysis.
The cluster labeling process revealed that the four major classes at LSI
corresponded to one hard bottom and three sediment classes. The three sediment classes
30
were distinguished by the presence or timing of a second echo captured within the 256-
sample IMPACT analysis window. After aggregating the IMPACT-derived acoustic
classes to a coarse level of descriptive resolution, the hard bottom and sediment classes
contained 25.6% and 70.7% of the echoes, respectively. Comparison of the coarse-level
classes with the LSI video dataset indicated that the hard bottom / sediment acoustic
classification had an overall accuracy of 74% (Fig. 3.4; Table 3.5A).
Table 3.4: Acoustic class labels and sizes (by percent of total echoes) in each survey area and aggregation of classes from fine descriptive resolution (DR) to coarse descriptive resolution.
Coarse DR Fine DR
Lee Stocking Island Andros Carysfort Reef Fowey Rocks Aggregated Class Class % Class % Class % Class %
Hard Bottom 25.6 Low Relief Hard Bottom 5.8 Hard Bottom 45.8
Low Relief Hard Bottom 53.4 Hard
Bottom
High Relief Hard Bottom 3.2
High Relief Hard Bottom 9.9
Sediment 59.3 Sediment 78.7 Coarse Sand 9.9 Coarse Sand 25.6
Double Echo 1 6.3 Double Echo 1 5.2 Fine Sand 38.2 Fine Sand 7.9 Sediment
Double Echo 2 5.1 Double Echo 2 3.8 Other 5 Minor classes 3.7 3 Minor classes 3.3 4 Minor classes 6.1 2 Minor classes 3.2
The Andros data clustered into eight acoustic classes. Five of these classes
contained 96.7% of the total number of echoes (Table 3.4), so analysis focused on these
five major classes. The cluster labeling process identified three sediment classes
distinguished by the presence or timing of a second echo and two hard bottom classes
distinguished by relief (Fig. 3.2 has examples of low and high relief hard bottom).
After aggregating the IMPACT-derived acoustic classes to a coarse level of
descriptive resolution, the hard bottom and sediment classes contained 9.0% and 87.7%
of the echoes, respectively. Comparison of the coarse-level classes with the Andros video
31
dataset indicated that the hard bottom / sediment acoustic classification had an overall
accuracy of 73% (Fig. 3.5; Table 3.5B).
The Carysfort data clustered into seven acoustic classes, three of which contained
93.9% of the total number of echoes (Table 3.4). Cluster labeling of these three major
classes identified one corresponding to hard bottom and two corresponding to sediment.
Unlike LSI and Andros, where a second echo differentiated the sediment classes, at
Carysfort all of the echoes were too deep to record a second echo given the settings used
(Tables 3.2 and 3.3). Instead, at Carysfort the sediment classes differed by grain size (Fig.
3.3).
The mode of each of the sediment-sample grain size distributions at LSI (Fig.
3.3A; N=42) and at Andros (Fig. 3.3B; N=7) was in the range 0 (coarse sand) < φ < 3
(fine sand) where φ is the grain size expressed as -log2(grain size in mm). No acoustic
classes were discriminated based on grain size at LSI or Andros. In contrast, two groups
emerged from the sediment-sample grain size distributions at Carysfort and Fowey Rocks
(Fig. 3.3C; N=17). One set of samples had a mode in the range -1 (very coarse sand) < φ
< 1 (coarse sand). The other set of samples was in the range 2 (fine sand) < φ < 4 (very
fine sand). Eight of the ten samples that were closest to an echo labeled as “coarse
sediment” (Fig. 3.3C grey lines with circle markers) had a grain size distribution with a
mode between coarse to very coarse sand (0.5 - 2 mm). Five of the seven samples that
were closest to an echo labeled as “fine sediment” (Fig. 3.3C black lines without
markers) had a grain size distribution with a mode between fine or very fine sand (0.062 -
0.25 mm).
32
Figure 3.3: Sediment grain size distributions for samples from the survey areas. Volume percent of the sample is plotted in equally-spaced φ intervals (φ = -log2(mm)). A) Lee Stocking Island (N=42); B) Andros (N=7); C) Carysfort and Fowey samples plotted together (N=17). LSI and Andros acoustic data were not separated by grain size whereas Carysfort and Fowey acoustic data split into two classes corresponding to coarse and fine sand-sized sediment. Line styles reflect the acoustic class of the closest echo to each sediment sample.
After aggregating the IMPACT-derived acoustic classes at Carysfort Reef to a
coarse level of descriptive resolution, the hard bottom and sediment classes contained
45.8% and 48.2% of the echoes, respectively. Comparison of the coarse-level classes
with the RVC substrate dataset indicated that the Carysfort hard bottom / sediment
acoustic classification had an overall accuracy of 86% (Fig. 5.4, Table 3.5C).
33
Figure 3.4: LSI acoustic classification (left) and video classification (right) plotted on top of a true-color IKONOS image (copyright GeoEye.com) of the LSI study area.
Figure 3.5: Andros acoustic classification (left) and video classification (right) plotted on top of a true-color IKONOS image (copyright GeoEye.com) of the study area.
34
Table 3.5: Error matrices from four survey sites comparing coarse descriptive resolution acoustic classes (Table 3.4) with ground truth obtained from video analysis or diver observations. No acoustic classes corresponded to rubble, so that row, which would be filled with zeros, has been left out of these tables. This omission does not change any of the accuracy calculations. Entries in the table are fractional because ground truth data were acquired as percentages of sediment, rubble, and hard bottom. Overall accuracy ranges from 73% to 86% at these four sites.
A) Lee Stocking Island Video Classes Acoustic Classes Sediment Rubble Hard Bottom User's accuracy
Sediment 890.02 2.08 176.9 0.83
Hard bottom 188.27 2.52 189.21 0.50
Producer's accuracy 0.83 0.52 Total number of video frames 1502 Number of video frames matching minor acoustic classes 53 Number of video frames matching major acoustic classes 1449 Overall accuracy: 0.74 B) Andros Video Classes Acoustic Classes Sediment Rubble Hard Bottom User's accuracy
Sediment 349.36 9.88 123.76 0.72
Hard bottom 15.47 1.4 48.13 0.74
Producer's accuracy 0.96 0.28 Total number of video frames 558 Number of video frames matching minor acoustic classes 10 Number of video frames matching major acoustic classes 548 Overall accuracy: 0.73 C) Carysfort Diver Classes Acoustic Classes Sediment Rubble Hard Bottom User's accuracy
Sediment 7.3 0 0.7 0.91
Hard bottom 1.1 0.7 8.2 0.82
Producer's accuracy 0.87 0.92 Total number of dive sites 22 Number of video frames matching minor acoustic classes 4 Number of video frames matching major acoustic classes 18 Overall accuracy: 0.86 D) Fowey Rocks Diver Classes Acoustic Classes Sediment Rubble Hard Bottom User's accuracy
Sediment 4.22 0 1.78 0.70
Hard bottom 1.34 0.19 7.48 0.83
Producer's accuracy 0.76 0.81 Total number of dive sites 15 Number of video frames matching minor acoustic classes 0 Number of video frames matching major acoustic classes 15 Overall accuracy: 0.78
35
The Fowey Rocks data clustered into six acoustic classes, four of which made up
96.8% of the total number of echoes (Table 3.4). As with the other sites, analysis focused
on these four major classes. The cluster labeling process indicated that two of the
acoustic classes corresponded to hard bottom, and two classes corresponded to sediment.
The two hard bottom classes were discriminated by relief. The two sediment classes were
discriminated by grain size (Fig. 3.3) as at Carysfort Reef.
Figure 3.6: Fowey rocks acoustic classes (track lines) and diver estimated substrate (pie charts).
After aggregating the IMPACT-derived acoustic classes at Fowey Rocks to a
coarse level of descriptive resolution, the hard bottom and sediment classes contained
63.3% and 33.5% of the echoes, respectively. Comparison with the RVC substrate dataset
indicated that the Fowey Rocks hard bottom / sediment acoustic classification had an
overall accuracy of 78% (Fig. 3.6; Table 3.5D).
36
3.5 Analysis: where were the errors? The overall accuracies for hard bottom / sediment classification in the four study
sites ranged from 73 to 86% (Table 3.5). The error matrix alone, however, does not
provide any explanation for the classification errors. In particular, were there any features
of the seabed in areas consistently misclassified that might explain the error?
Understanding the cause of misclassification may help guide future improvements to the
survey system or survey design.
3.5.1 Lee Stocking Island
The first step in analyzing the errors at LSI was to find the areas where the
acoustic classification least agreed with the estimates of substrate from the video. To do
this, the overall accuracy was computed for each individual frame. Traditionally, overall
accuracy for a single validation site (video frame in this case) may take on values of only
zero or one because in most studies the ground truth represents a single class at each
location. As noted in Section 3.3.4, however, the ground truth for these surveys contained
continuous measurements of the fraction of each site covered by hard bottom, rubble, or
sediment. Therefore, just as the overall accuracy computed from the entire error matrix
(Table 3.5) could vary from zero to one, the overall accuracy for each video frame could
vary from zero to one. The distribution of the frame-by-frame overall accuracies at LSI
was bimodal (Fig. 3.7), indicating that most video frames either nearly perfectly agreed
or nearly perfectly disagreed with the class of the nearest echo.
In the LSI dataset, there were 188.3 hard bottom sites classified by the acoustics
as sediment and 176.9 sediment sites classified by the acoustics as hard bottom (Table
3.5a). Most of the misclassified sediment sites occurred in the deepest parts of the study
area (Fig 3.8). In fact, almost all of the sites deeper than six meters that were classified on
37
the video as sediment dominated were classified as hard bottom by the acoustics (Fig
3.8). In contrast, most of the misclassified hard bottom sites were found along the edges
of islands (Fig. 3.7), which were the areas with shallowest water.
Figure 3.7: Within-frame accuracy for the LSI dataset. Left: Within-frame overall accuracy histogram (N=1502). Right: Map illustrating locations of video frames with less than 10% within-frame accuracy. Black box outlines the Adderly Cut area shown in Fig. 3.9; islands and boat track lines are shown in grey.
Figure 3.8: Depth-frequency histogram for sediment-dominated video frames in the LSI dataset. Red bars represent all sediment-dominated video frames (N=1114); blue bars represent sediment-dominated video frames in which the closest acoustic echo was misclassified as hard bottom (N=214).
38
Diver-based observations from July 2002 helped characterize the nature of the
misclassified LSI video frames. Due to time constraints, diving effort focused in Adderly
Cut, the channel just to the north of LSI (boxed area in Fig. 3.7) because it had a large
fraction of the misclassified sites (Fig. 3.7), and had been mapped previously (Gonzalez
and Eberli 1997; Louchard et al. 2003), which assisted the choice of dive sites.
Between July 6-11, 2002, 31 dives were made in the Adderly Cut area (Fig. 3.9).
Target locations for dives were taken from the acoustic tracks. The actual location of
each site was recorded as divers entered the water using a Garmin GPSMap 162 with
WAAS. Data acquired at each site included: sediment depth, rugosity, underwater
photographs, and a qualitative site description. Sediment depth was measured by probing
with a 3/16” stainless steel rod that had been etched at 5 cm intervals up to 40 cm.
Figure 3.9: Classified acoustic tracks, video frames, and dive sites in the Adderly Cut portion of LSI study area. Tracks are colored red or blue based on acoustic class hard bottom or sediment, respectively. Triangles are plotted at the locations of video frames and colored red or blue based on the dominant cover type being hard bottom or sediment, respectively. Dive sites A-D are labeled according to Tables 3.6, 3.7. Sites labeled with an E were too far from track lines or video frames to be useful.
39
Rugosity was measured by draping a 188 cm long chain with 1 cm links along the seabed
and measuring the straight-line distance covered. The actual rugosity index was
calculated from each chain measurement as the total length of the chain divided by the
linear distance covered when draped over the seabed (Luckhurst and Luckhurst 1978).
The chain measurements were repeated six times at each dive site, with the chain laid
three times parallel in one direction and then three times in a perpendicular direction.
There were features observed on the dives, such as large coral heads and sand waves, that
were too large to measure with the 188 cm chain. An effort was made to qualitatively
capture these features, which are hereafter referred to as “macrorelief,” as part of the site
description and with scale bars placed in underwater photographs.
Descriptions of the sites revealed common features within each group (Tables 3.6,
3.7). All of the sites in group A, the video hard bottom sites that were also classified as
hard bottom by the acoustics, contained gravel-sized (> 2mm) but little to no finer-
Table 3.6: Diver descriptions of Adderly Cut dives from July 2002 grouped according to the class of the closest video frame (columns) and acoustic echo (rows). Group letters A-D are consistent with Table 3.7 and Figures 3.9, 3.10.
40
grained sediment. Three of the sites in group A contained corals, and three also contained
“macrorelief” in the form of seagrass swales or hummocky hard bottom. The common
feature among sites in group B, the video hard bottom sites that were classified as
sediment by the acoustics, was the presence of sediment overlying hard bottom. At four
of the sites in group B, the sediment formed a thin veneer on top of an otherwise
relatively flat hard bottom. At the fifth site in group B the sediment was collected in up to
25 cm thick packages forming in the lee of 20-30 cm high hard ground hummocks. Five
of the six sites in group C, the video sediment sites that were classified as hard bottom by
the acoustics, contained seagrass swales, which were 0.5 to 1.5 m high seagrass-stabilized
sand dunes. The sides of the swales were often nearly vertical, formed by large blowouts.
All of the sites in group D, the video sediment sites that were classified as sediment by
the acoustics, were characterized by low relief sediment with seagrass of varying density.
Table 3.7: Probe depths (cm), mean and standard deviation (Std) of rugosity (unitless) in Adderly Cut measured by divers in July 2002. Group average rugosity values were computed on the entire set of raw data for each group, not averaging the average values for each site. Group letters A-D are consistent with Table 3.6 and Figures 3.9, 3.10.
41
The diver data (Tables 3.6, 3.7) revealed common characteristics of misclassified
sites. The most common cause for misclassification of the hard bottom sites was the
presence of a thin veneer of sediment (Group B in Tables 3.6, 3.7, Fig. 3.10). The
dominant cause of misclassification of the sediment sites was the presence of macrorelief
created by large seagrass swales (Group C in Tables 3.6, 3.7, Fig. 3.10). Underwater
photographs captured the archetype of each group (Fig. 3.10).
Knowing that sediment veneers on hard bottom and macrorelief in seagrass were
the two most apparent sources of confusion between the video and acoustic
Figure 3.10: Underwater photographs illustrating the four groups of sites in the Adderly Cut portion of the LSI study area. Groups A and D represent agreement between the video and acoustic datasets; groups B and C represent errors. Divisions on the scale bar are 10 cm. The pictures are from sites A4, B4, C2, and D6. Group letters A-D are consistent with Tables 3.6, 3.7 and Figure 3.9.
42
classifications, the video frames were reviewed to flag those that contained either of these
features. Three hundred thirty-six video frames had been identified with less than 10%
agreement between the video and acoustic classifications (Fig. 3.7). Each of these video
frames was inspected. The hard bottom sites with sediment veneers were readily
identifiable on the video frames. Of the 163 video hard bottom sites that had been
misclassified acoustically as sediment, 130 of them were flagged as having a shallow
surficial layer of sediment. The seagrass sites containing swales with significant relief, on
the other hand, were not easy to reliably identify, owing, in part, to the fact that they were
all located in the deepest part of the survey area (Fig 3.8) where distance to the seabed
and water column attenuation most obscured the details in the video images.
After reviewing the video frames, a new error matrix was computed that excluded
the 130 hard bottom with sediment veneer sites; no seagrass sites were excluded because
it was too difficult to assess local relief from the video frames. Overall accuracy
improved from 74%, with the sediment veneer sites included, to 84% with these sites
excluded (Table 3.8). Thus, 38% of the overall error in the LSI can be attributed to hard
bottom with sediment veneer.
Table 3.8: Revised LSI error matrix computed by excluding hard bottom sites with a veneer of sediment.
Video Classes Acoustic Classes Sediment Rubble Hard Bottom User's accuracy
Sediment 888.92 1.13 49.95 0.95
Hard bottom 188.27 2.52 188.21 0.50
Producer's accuracy 0.83 0.79 Total number of video frames 1372 Number of video frames matching minor acoustic classes 53 Number of video frames matching major acoustic classes 1319 Overall accuracy: 0.82
43
3.5.2 Andros Island Area
Analysis of the misclassified video frames for the Andros datasets followed the
approach used at LSI. Mismatches between the video and acoustic datasets were located
by computing overall accuracy on a frame-by-frame basis, as at LSI. Also like LSI, most
of the video frames either nearly perfectly agreed or disagreed with the nearest acoustic
class (Fig. 3.11). Unlike LSI, which had approximately the same number of hard bottom
sites classified as sediment as sediment sites classified as hard bottom (Table 3.5A), in
the Andros dataset the dominant error type was video hard bottom classified by the
acoustics as sediment (Table 3.5B). Of the 126 video frames with less than 10%
agreement with the class of the nearest echo (Fig. 3.11), 94 were classified by the video
as hard bottom, but clustered acoustically with sediment. Only 27 were classified by the
video as sediment but clustered acoustically with hard bottom.
Reviewing the video frames revealed that 90 of the 94 hard bottom-dominated
frames that were closest to echoes classified as sediment contained thin veneers of
sediment over low relief hard bottom. Acoustically, these low relief hard bottom sites
with sediment veneer clustered with areas covered by thicker sediment. Ecologically,
however, low relief hard bottom with sediment veneer provides a habitat more similar to
Figure 3.11: Within-frame overall accuracy histogram for the Andros dataset (N=558). Most sites exhibited nearly perfect agreement (OA=1) or disagreement (OA=0) between the acoustic and video datasets.
44
hard bottom than to sediment. Sediment sites were either bare (Fig. 3.12A) or covered
with seagrass and macroalagal forms that do not require a holdfast. In contrast, the hard
bottom sites with sediment veneer were covered with hydrocorals, octocorals, small stony
corals, and algae attached to firm substrate (Fig. 3.12B). The two acoustic classes in the
Andros dataset that did correspond to hard bottom (Table 3.4) contained the same types
of benthic organisms found in the hard bottom sites with sediment veneer but in different
proportions. Specifically, meter-scale relief, a metric controlled by the proportion of
stony corals, differentiated the two acoustic hard bottom sites from each other and from
the hard bottom sites with sediment veneer (Fig. 3.12B, C, D).
Reviewing the video frames of the 27 sites classified acoustically as hard bottom
and by the video as sediment did not reveal any common characteristics among the sites
that might explain the misclassification. At LSI, the presence of high relief, seagrass-
stabilized sand waves explained the echoes from sediment sites clustering with hard
bottom, but no such features were observed within the Andros survey area.
Recalculating the error matrix after eliminating sites determined by a review of
the video frames to have sediment veneers over low relief hard bottom resulted in overall
accuracy improving from 71% to 85% (Table 3.9). Thus, 48% of the overall error in the
Andros survey was caused by hard bottom with sediment veneer.
Table 3.9: Error matrix for the Andros survey excluding hard bottom sites covered with a thin sediment veneer. Sediment veneer accounts for 48% of the overall error at Andros.
Video Classes Acoustic Classes Sediment Rubble Hard Bottom User's accuracy
Sediment 348.21 8.04 42.75 0.87
Hard bottom 15.37 1.35 46.28 0.73
Producer's accuracy 0.96 0.52 Total number of video frames 470 Number of video frames matching minor acoustic classes 8 Number of video frames matching major acoustic classes 462 Overall accuracy: 0.85
45
Figure 3.12: Oblique underwater photographs of four seabed types in the Andros Island survey area. Each row contains two pictures of each seabed: A) sediment, B) low-relief hard bottom with sediment veneer, C) low-relief hard bottom, D) high-relief hard bottom. Acoustically, seabed type B clustered with seabed type A, but ecologically seabed type B is closer to seabed type C.
46
3.5.3 Carysfort Reef and Fowey Rocks
Analysis of the acoustic classification for Carysfort Reef and Fowey Rocks
followed the approach used at LSI and Andros. Mismatches between the diver-assessed
substrate and acoustic datasets were located by computing overall accuracy on a dive-by-
dive basis, as they had been computed frame-by-frame for LSI and Andros. No dive sites
had with zero percent accuracy relative to the closest acoustic echo (Fig. 3.13). This
differed markedly from the LSI (Fig. 3.7) and Andros (Fig. 3.11) surveys in which the
majority of the non-matching video frames were complete errors. No systematic
misclassifications were evident in the Carysfort and Fowey Rocks datasets. Therefore, no
reclassification was attempted with either of these datasets.
Figure 3.13: Within-frame overall accuracy histograms for the Carysfort (left) and Fowey Rocks (right) datasets. Unlike the LSI and Andros datasets (Figs. 3.7, 3.11) no dive sites had zero percent accuracy relative to the closest acoustic echo.
3.6 Discussion The overall objective of this study was to determine what seabed classes were
consistently distinguished by a single-beam acoustic classification system among
multiple survey areas in coral reef environments; in other words, “How transferable are
acoustic seabed classes?” The results showed that at coarse descriptive resolution
47
acoustic seabed classes were transferable. In addition, the results in two of the four study
areas revealed specific seabed characteristics that lead to systematic classification errors.
After commenting on descriptive resolution and commonly observed classification errors,
the following discussion considers how these results relate to previous studies and how
single-beam acoustic classification could contribute to mapping coral reef environments,
given the strengths and limitations observed in this study.
The transferability of single-beam acoustic seabed classification schemes was
assessed quantitatively at a coarse descriptive resolution. Hard bottom was well
discriminated from sediment using the QTC system. Overall accuracy for the two-class
hard bottom / sediment classification ranged from 73 to 86% in four study sites separated
by as much as several hundred km (Table 3.5). Quantitatively assessing transferability at
a fine level of descriptive resolution was not possible because the clustering process
identified a unique set of seabed classes at each study site (Table 3.4). Qualitatively,
therefore, the results showed that even though some individual classes were identified at
more than one site, single-beam acoustic seabed classification schemes as a whole were
not transferable between sites at a fine level of descriptive resolution. That acoustic
seabed classes should be transferable at a coarse but not fine level of descriptive
resolution is consistent with expectations based on the classes identified in previous
studies (Table 3.1).
Based on diver observations and reevaluation of the video ground truth data, the
most common classification error was caused by a thin (< 10 cm) veneer of sediment on
top of hard bottom, which caused the hard bottom to be classified by the acoustics as
sediment (Figs. 3.10, 3.12). From a purely physical point of view, one might reasonably
48
argue that classifying a thin surficial layer of sediment as “sediment” does not actually
constitute an error. Ecologically, however, the biological community inhabiting a thin
veneer over hard bottom differs greatly from one inhabiting deep sediment (Cf. Figs.
3.10B vs. 3.10D and Fig. 3.12A vs. 3.12B). Most applications, such as management,
stratifying sampling effort, scaling demographic measurements up by habitat area, and so
forth, would want to distinguish areas with sediment veneer from areas with thick
sediment.
The second commonly misclassified seabed type was found at LSI only and
included large (~1m or higher) seagrass-stabilized sand waves being classified as hard
bottom. These areas frequently, though not always, were intermingled with a pavement
hard bottom and gravel. The physical basis for this type of misclassification was probably
echo lengthening due to macroroughness (large change of depth within the echo
footprint) of the large waves. The relevant roughness length scale for 50 kHz sonar is on
the order of a wavelength, or about 3 cm. Bare hard bottom in any of these survey sites
will have many surface facets on this length scale (Figs. 3.2, 3.10A, 3.12C, D). In
contrast, the hard bottom areas with thin sediment veneer were much smoother (Figs.
3.10B, 3.12B). The roughness length of the seagrass-stabilized sand waves was much
longer, on the order of meters, but the waves were large enough to create significant
depth variation within a single echo. The effect of such macroroughness on echo shape is
the same as wavelength-scale seabed roughness because both lengthen the duration of the
returned echo.
The misclassified sediment veneer and seagrass dune habitats support the
conclusion that the QTC classification in these study areas was more sensitive to seabed
49
roughness than to sediment penetration. Oblique underwater photographs provided
further illustration that habitats clustered by roughness. Two ecologically different but
similarly physically smooth habitats (Fig. 3.12A, B) clustered together, distinct from a
third slightly rougher seabed (Fig. 3.12C) and a fourth much rougher seabed (Fig. 3.12D).
Overall accuracy of the coarse-level classification at the four survey sites was
comparable to previous values for coarse-level habitat mapping reported by studies using
satellite or airborne imagery. Mumby and Edwards (2002), for example, reported overall
accuracies ranging from 68% to 75% using satellite imagery and 89% using airborne
imagery for a four-class scheme (coral, macroalgae, seagrass, sand). Andrefouet et al.
(2003) reported, for four or five habitat classes, overall accuracies of 77%, 78%, and 81%
at Glovers, Heron, and Shiraho reefs using Iknonos imagery. Capolsini (2003) et al.
reported overall accuracies ranging from approximately 73-85% (exact figures not given)
for a single study area classified into three, four, and five classes. Maeder et al. (2002)
reported an overall accuracy of 89% for a scheme with five classes (four benthic classes
plus deep water). In such studies, overall accuracy commonly decreased as the number of
image-derived habitat classes increased (Andrefouet et al. 2003).
Overall accuracy at the four survey sites was also comparable to values reported
by previous single-beam acoustic seabed classification studies in coral reef environments
(Table 3.1; Fig. 3.14) with the possible exception of Riegl et al. (2007). The findings of
this study appear to contradict Riegl et al. (2007) who found that hard bottom with
sediment veneer classified reliably as hard bottom. The present study, in contrast, found
that hard bottom with sediment veneer classified as sediment in both survey areas where
it was present. Riegl et al. (2007) did not discuss sediment veneer in detail but did note
50
that it was interspersed between near-shore rocky ridges. The terrain mapped by Riegl et
al. (2007) near Cabo Pulmo includes an undulating bedrock plain with a patchy thin
veneer of sediment such that bedrock does protrude through the overlying sediment (B.
Riegl, personal communication Dec. 2008). Given this description, there are two
potential explanations for the discrepancy between Riegl et al. (2007) and the results at
LSI and Andros. First, patchy versus unbroken veneer may be a crucial difference. If the
(rough) underlying rocky surface protrudes over even a fraction of the ensonified area, or
over a fraction of the echoes within a stack, the echo might be lengthened sufficiently to
appear as a hard ground. The difference between groups A and B (Tables 3.6, 3.7, Fig.
3.10) was not that group A had no sediment; it was that group B had complete sediment
cover. A second potential explanation for sediment veneers in Cabo Pulmo classifying as
hard bottom may be the macrorelief provided by the nearby rocky ridges. As evidenced
by groups C and D (Tables 3.6, 3.7, Fig. 3.10), sufficient local relief appears the same
acoustically as a rough seabed. Without further quantitative data, it is not possible to
resolve which of these explanations dominates in the case of Cabo Pulmo. It does seem
plausible, however, that the discrepancy between Riegl et al. (2007) and the results of this
study may result more from different definitions of “veneer” than from conflicting
performance of the acoustic classification systems.
What are some potential uses of single-beam acoustic classification in coral reef
environments, given the strengths and limitations illustrated above? As just mentioned,
the overall accuracy of the simple two-class acoustic seabed maps was comparable to
coarse-level seabed maps derived from overhead optical imagery. Even coarse-level
habitat maps derived from overhead images identify from three to five classes, however
51
(Maeder et al. 2002; Mumby and Edwards 2002; Andrefouet et al. 2003; Capolsini et al.
2003). It is worth, therefore, considering whether a simple two-class rock / sediment
classification scheme is useful. An important point to remember is that with acoustic
data, even single-beam data, the classification may be supplemented by utilizing
bathymetry, which is inherently part of the data collection.
Figure 3.14: Plot of overall accuracy as a function of the number of acoustic classes. Open symbols represent the values reported in this study. Closed symbols represent values reported in earlier studies. Generally, accuracy decreases with increased numbers of acoustic classes.
One example of the benefits of adding a simple rock / sediment classification to
traditional bathymetric data for habitat mapping is the ability to discriminate outcropping
parts of the seabed. Hard bottom is known to be important habitat for many types of fish.
Sometimes substrate can be inferred from topographic profiles, but in other cases
interpreting outcrops from bathymetry alone can be misleading. An example from the
Florida Keys (Fig. 3.15) illustrates a case where bathymetry alone provided a misleading
picture of habitat. Based on bathymetry alone, one portion of this example survey area
appeared to be promising fish habitat because it contained two parallel ridges with steep
slopes (orange area highlighted in Fig. 3.15). In contrast, another portion of this example
52
survey area (purple area highlighted in Fig. 3.15) appeared to be a flat, featureless plain
based on bathymetry alone. Considering seabed type in addition to bathymetry, a
different interpretation became apparent; the area in orange was sediment-covered,
providing little shelter for fish, whereas the area in purple was covered with small patch
reefs, which would, in general, provide excellent habitat for reef fish. Augmenting
bathymetry with hard bottom / sediment seabed mapping will be exploited in Chapters 5
and 6 for the interpretation of grouper and snapper habitat.
Figure 3.15: Example of the utility of rock / sediment seabed classification in interpreting bathymetry for fish habitat. Top panel: sunshaded bathymetry from interpolated single-beam echosounder depths along the upper slope seaward of Carysfort and Watson’s Reefs, Florida Keys, USA. Bottom panel: Survey track lines superimposed on the same sunshaded bathymetry as the top panel. The track lines are colored red for hard bottom (rocky) and gray for sediment substrate. In both panels, oblique view is from the ESE, total along-shelf distance is about 12 km, and depth ranges from 3 to 60 m.
53
A second benefit to habitat mapping resulting from the integration of a simple
rock / sediment classification with bathymetry is the potential to create a single, objective
habitat classification scheme based on patchiness and relief. Franklin et al. (2003)
proposed such a habitat classification scheme for coral reef environments. Patchiness was
defined as the percent of the seabed within a certain radius of the point being classified
that was covered with sediment. Relief was defined as the depth range within a certain
radius of the point being classified. Franklin et al. (2003) showed how this classification
scheme effectively stratified survey efforts in the Dry Tortugas, USA. The actual maps
used by Franklin et al. (2003), however, were produced by subjective interpretation and
integration of multiple data sources. Using a single-beam ASC system to map the
Franklin et al. (2003) classes would capture the benefits of this classification scheme and
simultaneously improve both the automation and objectivity of the method.
Miller et al. (2008) used QTCV data and the Franklin et al. (2003) classification
scheme to produce a benthic habitat map for the Navassa National Wildlife refuge. One
of more than 100 transects used to create the Navassa map illustrates the utility of a
classification scheme based on patchiness and relief as well as the ability to implement
such a scheme using QTCV data (Fig. 3.16). Two types of patch reefs were observed
along the transect (Fig. 3.16). Reefs labeled “A” and “B” had similar cross-shelf width,
but those labeled “B” had higher relief. Considering substrate only (Fig. 3.16, top
profile), these reefs were all placed in the same class (hard bottom), but considering both
patchiness and relief (Fig. 3.16, bottom profile), these reefs were discriminated (yellow
vs. blue classes).
54
Figure 3.16: Cross shelf profile of part of the Navassa Island insular shelf. Top: Echoes colored by substrate only, red for hard bottom and green for sediment. Bottom: The same transect with echoes colored by a combination of local patchiness and relief. Low and high relief patch reefs, marked A and B respectively, are all classified as hard bottom, but can be differentiated using a combination of local patchiness and relief.
Franklin et al. (2003) produced a habitat map of the Dry Tortugas using a
classification scheme based on patchiness and relief, but the map was based on
qualitative interpretation only, due to the available datasets: primarily side-scan sonar,
aerial imagery, and diver assessment. In contrast, the classes demarked with the single-
beam ASC approach (Fig. 3.16) were objectively determined. Increased automation and
objectivity of seabed classification is one goal of the ASC approach (Anderson et al.
2008), which is here illustrated by implementing the Franklin et al. (2003) classification
scheme using QTCV data.
One of the most useful attributes of a simple hard bottom / sediment classification
scheme is that the classes can be consistently identified across multiple areas. The seabed
classes mapped in previous studies in coral reef environments were not consistent from
site to site (Table 3.1). Indeed the fine descriptive resolution seabed classes identified in
55
this study were also not consistent from site to site (Table 3.4). All of the classes mapped
in previous studies and all of the classes mapped in this study, however, could be
collapsed to a coarser descriptive level into hard bottom and sediment classes. Having
even just two classes that mean the same thing in different survey areas, when mapped by
different systems and different operators, is potentially extremely useful. One of the
biggest challenges to remotely-sensed inventories in general, and reef assessments in
particular, is that different classification systems, data processing, and assessment
protocols mean that comparing results from different sensors, investigators, or locations
is not straightforward (Green et al. 1996; Mumby and Harborne 1999). Tools and
techniques that yield reliable and objective discrimination between hard bottom and
sediment, such as those developed and illustrated here, are an important first step toward
producing uniform maps across regions and operators.
3.7 Conclusions Transferability of results refers to both the consistency of classes mapped in
different places as well as the consistency of classes mapped by different sensors or
processing methods. This study addressed the first aspect of transferability by using a
single commercial single-beam acoustic seabed classification system to map multiple
sites in coral reef environments. The QTCV / IMPACT acoustic seabed classification
system discriminated rocky from sediment substrate with 73% to 86% accuracy at four
sites using the same hardware and different methods for ground truth.
Further work to improve the ability of single-beam classification systems to
discriminate deep sediment from hard bottom covered with a thin sediment veneer would
improve classification accuracy. For example, hard bottoms with a thin sediment veneer
56
were responsible for 38% and 48% of the total error at LSI and Andros, respectively. As
evidenced at both LSI and Andros, these habitats can cover large areas in carbonate bank
top environments and should therefore be a high priority for improvements to ASC
methods.
Evidence from two systematically misclassified seabed types, hard bottom
covered with sediment veneer and sediment with macrorelief, suggests that variability in
echo shape depended more strongly on surface roughness than volume scattering in these
survey areas. Future surveys over hard bottom with sediment veneer might therefore
benefit from lower frequency data, which should penetrate the seabed deeper than the 50
kHz system used here. Future surveys over seabeds with high macrorelief may benefit
from a narrower beam width transducer, to reduce footprint size, or new post-processing
algorithms incorporating bathymetry.
The ability to accurately discriminate hard bottom from sediment appears to be a
robust property of single-beam acoustic seabed classification systems. These two classes
were reliably distinguished at four sites in this study and at five additional sites in
previous studies, after considering which previously identified classes could likely be
consolidated. A classification scheme that can be objectively applied using multiple
systems, in multiple locations, by different people would be useful for efficiently
exploring large areas of potential coral reef habitat that have not been mapped due to
water clarity or depth.
Single-beam acoustic seabed classification systems have the potential to play an
important role in coral reef mapping efforts due to the capability to extract bathymetry
and distinguish rock from sediment substrate, in combination with their low cost and
57
portability. By documenting class transferability at coarse descriptive resolution, this
study supports the concept that single-beam acoustic seabed mapping systems can
efficiently providing rapid reconnaissance and moderate resolution habitat maps of large
areas, thereby complementing other more detailed, but more costly, survey methods.
58
Chapter 4: Reproducibility of single-beam acoustic seabed classification under variable survey conditions
4.1 Background Maps depicting thematic seabed classes are important tools for both research and
management. Where the water is relatively clear, for example in many parts of the
tropics, satellite and aerial imagery play an important role for classifying the coastal zone
(Green et al. 1996). Most of the global seabed cannot be mapped using overhead
imagery, however, due to water clarity or depth limitations. In deep or turbid water,
acoustic remote sensing is an efficient method for classifying the seabed.
A recent review of acoustic seabed classification (ASC) identified the
standardization of instruments and methods as one category of “burning issues” for future
research (Anderson et al. 2008). A standardized method for ASC would be based on
statistical, objective procedures that reduce the variability inherent to subjective
interpretation (i.e. visually by an analyst). One aspect of a standardized method, which
was discussed in Chapter 3, would be a single classification scheme applicable to
multiple survey areas, different equipment, and different operators. A second aspect of a
standardized method is the reproducibility of results. ASC must be reproducible under
different survey conditions because the goal is to create a map of the seabed, not a map of
uncontrolled survey parameters. Any correlation between echo features and survey
parameters such as sea state will decrease the accuracy of the resulting seabed classes.
The overall objective of this study was to assess the reproducibility of a
commercial single-beam acoustic seabed classification system. The two specific
questions asked were, first, “How reproducible are the results of a particular off-the-shelf
survey system?” Second, “Can reproducibility be improved by altering the standard usage
59
of the off-the-shelf software?” Six surveys performed with a single-beam sounder over a
pair of transects near Miami, FL, USA, between 1 May and 13 August 2007 were used to
answer these questions. The strategy of repeat surveys over the same area follows the
suggestion of Anderson et al. (2008) that verification of instrument performance over
reference areas is one way to assess the reproducibility of entire end-to-end ASC systems,
as opposed to calibrating individual components of the system.
The particular single-beam ASC system used to evaluate questions of
reproducibility consisted of two components, both produced by the Quester Tangent
Corporation (QTC; Sidney, BC, Canada). A QTCView Series V (QTCV) was used for
data acquisition and the QTC IMPACT software was used for data processing. Together,
these components comprise the "QTC system". Using a single commercial system might
appear limiting, and certainly the measured absolute values of reproducibility only apply
to this system and indeed this survey area. The framework described for measuring
reproducibility, however, could be applied to data from any other system, or from another
QTC system in another environment. Furthermore, the conclusions about what factors
affect reproducibility are not necessarily confined to the QTC system. Methodological
improvements that increase the reproducibility of QTC results will likely benefit other
single-beam ASC systems as well. Thus, these experiments should be of interest to the
broader acoustic seabed classification community in addition to existing QTC users.
4.2 Previous work Several aspects of single-beam ASC system reproducibility have been examined
in previous studies, but there has not yet been a systematic evaluation of the
reproducibility of QTCV data. Foster-Smith and Sotheran (2003) surveyed Loch Maddy,
60
Scotland, on two different days with different 200 kHz RoxAnn systems and found 57%
similarity between the resulting maps. Greenstreet et al. (1997) mapped the inner Moray
Firth, Scotland, with a 38 kHz RoxAnn system and found systematic differences when
comparing to a repeat survey of the same area about four months later. Brown et al.
(2005) conducted four surveys of the same area in the Firth of Lorn, Scotland, varying
vessel speed, 200 kHz RoxAnn system, track spacing, and track orientation among the
four surveys. Agreement between pairs of surveys ranged from 36% to 43%. Wilding et
al. (2003) revisited a set of eleven stations off the west coast of Scotland on six days over
the course of ten months using a 200 kHz RoxAnn. Wilding et al. (2003) found that E1
and E2 values, the two parameters on which the RoxAnn system bases its classifications,
drifted over the time scales of hours, days, and months along their repeat transect. They
also found E1 and E2 to vary with vessel speed. Von Szalay and McConnaughey (2002)
performed systematic repeat transects, on the same day, at two sites and found no effect
of vessel speed on classification reproducibility using a 38 kHz QTC Series III (QTCIII)
and 38 kHz Series IV (QTCIV).
The choice of ASC system is one difference between this study and previous
work. Most previous investigations of reproducibility used the RoxAnn ASC system
(Greenstreet et al. 1997; Foster-Smith and Sotheran 2003; Wilding et al. 2003; Brown et
al. 2005). Reproducibility characteristics may differ between RoxAnn and QTC systems,
however, for several reasons. First, the two systems exploit different portions of echoes.
RoxAnn systems analyze the tail of the first echo and the entire second echo whereas
QTC systems analyze the entire first echo. Second, the performance of RoxAnn and QTC
systems has been shown to differ in other ways besides reproducibility. For example,
61
QTCIV and RoxAnn have produced different results in the same study area (Hamilton et
al. 1999). Furthermore, vessel speed has been shown to affect RoxAnn classifications
(Hamilton et al. 1999; Wilding et al. 2003) whereas QTCIV classifications have proven
consistent under variable vessel speed (Hamilton et al. 1999; von Szalay and
McConnaughey 2002).
This study differs from previous work in other ways beyond simply the choice of
ASC system. Some previous studies have interpolated E1 and E2 values to a regular grid
before comparison (Greenstreet et al. 1997; Foster-Smith and Sotheran 2003; Brown et
al. 2005) whereas this study compared individual points without interpolation. There is
no problem with interpolation in general, but it is another confounding step when trying
to understand why maps might differ. Therefore, understanding the limitations to
reproducibility will be simpler without interpolation. Some previous studies compared
the output of different sounders (Foster-Smith and Sotheran 2003; Brown et al. 2005)
whereas this study used one system for all data collection to limit potential confounding
factors to survey conditions. Some previous studies have compared the output of
supervised classification (von Szalay and McConnaughey 2002; Foster-Smith and
Sotheran 2003; Brown et al. 2005), whereas unsupervised classification was used here.
One study used the overall percentages of classes within repeat transects as a metric for
reproducibility (von Szalay and McConnaughey 2002). In this study, all comparisons
were performed point-by-point. This study used 50 kHz data, whereas all previous repeat
surveys have used 200 kHz or 38 kHz (Greenstreet et al. 1997; von Szalay and
McConnaughey 2002; Foster-Smith and Sotheran 2003; Wilding et al. 2003; Brown et al.
2005). Finally, the biggest difference between this study and previous work is that an
62
effort has been made here to improve reproducibility, not just to measure it. Efforts to
improve reproducibility provided insight into how the QTC system in particular, but also
single-beam ASC systems in general, might be improved in the future.
4.3 Methods The methods described in this Section and the results presented in Section 4.4
apply to the first question addressed by this Chapter, namely the reproducibility of an off-
the-shelf ASC system. The second question addressed by this Chapter, how to increase
reproducibility, employed multiple independent approaches. The methods and results for
each of the efforts to increase reproducibility are described in Sections 4.5.1-4.5.4.
4.3.1 Survey site and data acquisition
Surveys were conducted at Fowey Rocks, a reef chosen because its proximity to
Miami made it easy to survey repeatedly and because its geomorphology was typical of
the major upper Florida Keys reefs. A pilot survey performed in 2003 was used to pick
two parallel transects running across the outer shelf and upper slope from depths of 5 to
60 m. Each transect was about 2 km long, taking 12 minutes to run at a nominal survey
speed of 6 knots (Fig. 4.1).
Acoustic data were collected with a 50 kHz QTCV system (Table 3.2 lists system
characteristics). The QTCV system was mounted on the gunwale of an 8m long survey
vessel each day. Repeat transects were run on six days in 2007: May 1, 2, 9, 28, August
3, and 13. On each day, both transects were traversed three times. In addition to the
acoustic data, pitch and roll of the survey vessel were recorded at a rate of 100 Hz.
Environmental conditions during the six surveys were extracted from NOAA
National Buoy Data Center records (Fig. 4.2). The three closest stations to the survey
area were (Fig. 4.1): A) Fowey rocks lighthouse, located at the survey site; B) Virginia
63
Key, located approximately 15 km to the north of the survey site; and C) Molasses reef
lighthouse, located approximately 45 km to the south of the survey site.
Figure 4.1: Fowey Rocks survey site and depth profiles. A) Location of Fowey Rocks (FR), Virginia Key (VK) and Molasses Reef (MR) in relation to South Florida. B) Survey tracks plotted in a different color for each day. Profiles along NW-NE and SW-SE are plotted in panels C and D. C) North transect depth profile (20x exaggeration) colored by classes from the May 1 dataset. D) South transect depth profile (20x exaggeration) colored by classes from the May 1 dataset. The main features are a series of north-south trending linear reefs (three on N transect, two on S transect) that are separated by relatively flat sediment in shallow water and sloping sediment (presumably) in deeper water.
4.3.2 Vessel attitude and grazing angle computation
Pitch and roll of the survey vessel were used along with depth to compute the
incidence angle and its complement, the grazing angle, for each echo. The incidence
angle is the angle between the local seabed normal and the direction of propagation of the
center of the incoming acoustic energy, taken here as a vector normal to the face of the
transducer. Two processing steps were needed to enable the calculation of incidence
64
angle from pitch and roll for each stack of echoes. First, pitch and roll, which were in a
boat coordinate frame, had to be converted to a transducer pointing vector in a world
coordinate frame. Second, the slope of the seabed was computed by fitting a line or plane
to a moving window of depth measurements surrounding each stacked echo.
Pitch and roll were converted by computing the boat heading from the GPS track
and using the heading to rotate the unit vector in boat coordinates to UTM world
coordinates. For each echo, heading was computed as the average of the heading from the
coordinates of that stack to the following and previous echo.
Slope was computed for each echo by using all the depth measurements within a
window of 40 m acquired on a given day. The UTM coordinates of these points were
placed in a Nx2 matrix, where N was the number of points found in the 40 m window,
column one was the UTM eastings and column 2 was the UTM northings. The condition
number of this matrix, which indicates the accuracy of results from matrix inversion, was
computed with the MATLAB ‘cond’ command (R2008a, the Mathworks, Natick, MA) to
assess whether the points could adequately define a plane or if they were aligned along a
line. If the coordinate matrix had condition less than or equal to two, then a plane was fit
to the eastings, northings, and depth data and the surface normal to the plane computed. If
the condition was greater than 2, that indicated that the points fell mostly along a line, so
principal components were used to project the eastings, northings data onto a single
dimension, and then least squares regression was used to fit a line to the projected
coordinate / depth data in order to compute the surface normal. In the latter case, when
fitting a line rather than a plane, the assumption was that there was zero slope component
perpendicular to the track line.
65
The slope calculation produced a seabed normal vector (in UTM coordinates) and
the pitch, roll, and heading data produced a transducer pointing vector (also in UTM
coordinates). The angle between the two, which was computed from the dot product of
these vectors, gave the incidence angle of the pulse onto the seabed, and 90 minus the
incidence angle gave the grazing angle.
A final challenge in the computation of grazing angle was addressing the
procedure of echo stacking. As described in Section 3.3.2, echo stacking, or averaging,
was used when processing the acoustic data to suppress ping-to-ping variability casued
by incoherent scattering from the seabed. For these data, five echoes were averaged to
produce one stacked echo (Section 3.3.2, Table 3.3). Stacking echoes resulted in an
ambiguous definition of grazing angle because the roll and pitch varied for each of the
five echoes in the stack. To address this challenge, grazing angle was computed in three
ways: A) using the roll and pitch at the time of the center ping of the stack; B) using the
maximum roll and pitch observed during the time spanned by the stack; C) using the
minimum roll and pitch observed during the time spanned by the stack. Note that only in
method (A) can the total grazing angle be computed, but that for all three methods both
the pitch and roll components of grazing angle can be computed. This is due to the fact
that the maximum or minimum roll and pitch did not necessarily occur at the same time
during the stack span. In the case of the transducer used for this study it was
advantageous to look at roll and pitch separately because the transducer beamwidths were
quite different in those directions; full beam width angles were 16 degrees in the pitch
direction and 42 degrees in the roll direction (Table 3.2).
66
4.3.3 Acoustic data processing
The acoustic data were processed using QTC IMPACT software (version 3.4; see
Chapter 2 for software overview) using the same settings described for the Fowey Rocks
survey in Chapter 3 (Table 3.3). As described in Section 3.3.2, data processing in
IMPACT consists of six steps. The first three steps, envelope generation, depth
compensation, and echo stacking, produce bottom-aligned, depth-compensated, smoothed
echoes. The fourth step, feature extraction, generates metrics of echo shape for each
stacked echo. IMPACT organizes the feature dataset into a [Nx166] matrix, where N is
the number of echoes and 166 the number of features generated for each echo. The
features for each echo, therefore, may be considered a vector, called the “full feature
vector” (FFV) by IMPACT. In the fifth step, the full [Nx166] matrix is decorrelated and
reduced in size by principal components analysis. IMPACT retains the first three
principal components, labeling them “Q1”, “Q2”, and “Q3”. The reduced [Nx3] matrix is
called “Q-space”. The sixth step in IMPACT processing is clustering Q-space. The final
output from the entire process is an assignment of each echo to a cluster comprised of
echoes with similar shapes. Clusters are also known as “acoustic classes.”
Clustering in IMPACT (v. 3.4) uses the “automatic cluster engine” (ACE), which
employs a modified k-means algorithm embedded within a simulated annealing routine
(Preston et al. 2004a). ACE outputs are: a) the optimum number of clusters into which
the echoes should be split, as defined by the Bayesian Information Criterion (Preston et
al. 2004a), b) the mean vector and covariance matrix defining each cluster in Q-space,
and c) a class number for each echo assigning it to one of the defined clusters.
Acoustic data were clustered in three sets. First, a single merged dataset
comprised of all echoes from all six days was created and clustered. Second, each day’s
67
data were clustered separately. Finally each individual transect on each day was clustered
separately. Thus there were a total of 43 clustered datasets each composed of a different
temporal subset of the same raw data: A) a single merged dataset from all six transects on
all six days; B) six daily datasets, each containing the six transects done on that day; C)
36 individual transects, six for each day.
Analysis of reproducibility was divided into two parts based on the type of data.
First, consistency of acoustic classes, which are nominal data, was investigated (Sections
4.3.4, 4.4.2). Second, correlations between echo envelopes, FFVs, and Q-values, which
are continuous vector data, with survey parameters were investigated (Sections 4.3.5 and
4.4.3).
4.3.4 Classification reproducibility
Classification reproducibility, from a user’s perspective, means consistently
assigning any particular parcel of seabed to the same acoustic class. Two qualitative and
two quantitative methods were used to assess classification reproducibility.
One qualitative indicator of reproducibly was simply the number of classes
discriminated during the clustering process. If the input data, for example on two
different days, were identical, the clustering process would identify the same classes in
those two datasets. Finding the same optimum number of classes in two repeat datasets
does not guarantee that two datasets are identical, but different optimum numbers of
classes is one sign that two repeat datasets are not identical.
A second qualitative assessment of classification reproducibility consisted of
visual assessment of the consistency of classes plotted in both geographic space and Q-
space. Simply put, if the differences between two classified repeat transects are hardly
noticeable by eye, then the reproducibility of the method is better than if the differences
68
are obvious just looking at the maps. By itself, qualitative assessment is not a rigorous or
objective way to evaluate agreement between maps, (Congalton and Green 1999; Foody
2006), but it provides an intuitively understandable complement to quantitative metrics of
map agreement.
Two methods were used to quantitatively assess classification reproducibility
between pairs of daily classified datasets: the error matrix and shared information. Both
techniques were derived from contingency table analysis (Finn 1993; Legendre and
Legendre 1998). The two methods of analysis produced three metrics of dataset
agreement: overall accuracy (OA), the Kappa statistic, and Average Mutual Information
(AMI).
The error matrix is a common tool for evaluating the accuracy of classified maps
with “ground truth” data (Congalton and Green 1999). The error matrix is a square
contingency table in which the class labels of the rows correspond to those of the
columns. Many statistics derived from error matrices can serve as accuracy metrics, but
overall accuracy (OA) and the Kappa statistic are two of the most basic and most
common (Congalton and Green 1999; Foody 2002).
OA is computed as the trace of an error matrix (the number of points which had
the same class in both datasets) divided by the total number of elements in the error
matrix (the number of points compared between the datasets). OA varies from 0,
corresponding to no agreement, to 1, reflecting perfect agreement between the two
datasets being compared. In this case, results are reported as 100 * OA so the scale ranges
from 0-100.
69
One limitation of OA is that some of the positive matches can be attributed to
chance. Kappa is a modified OA designed to correct for this positive bias due to chance
agreement (Congalton and Green 1999). There are several ways to weight or otherwise
modify Kappa (Foody 2004; Sim and Wright 2005). This analysis used the basic Kappa
statistic (e.g. Finn 1993):
€
Κ =OA − Pr1− Pr
(1)
where K is Kappa, OA is overall accuracy, and Pr is the random probability of a correct
match in the confusion matrix. Pr was computed as the inner product of the error matrix
row sum and column sum vectors divided by the square of the number of points
compared between the datasets (Finn 1993).
Error matrices are traditionally square, and they require class names to be the
same for both rows and columns (i.e. class one on map A must mean the same thing as
class one on map B). The daily ACE-clustered QTCV data do not necessarily meet these
criteria; different numbers of clusters are possible on different days, and the cluster
numbers from day to day do not mean the same thing. Therefore the first step was to
renumber the daily classes to a common scheme. To do this, there needed to be at least as
many new (renumbered) classes as the day with the most clusters. Consequently, days
with fewer than the maximum number of clusters had at least one empty new class.
Once the clusters for each day had been renumbered, OA and Kappa were
computed for each pair of datasets as follows. First, a day was chosen as the “reference”
dataset and a second day chosen as the “comparison” dataset. Five hundred random
echoes were selected from the reference dataset and each was paired with the closest echo
in the comparison dataset. Any pairs of points separated by greater than 10 meters were
70
discarded, which usually left somewhere between 350 - 450 matched points. These
remaining matched points were used to create an error matrix, from which OA and Kappa
were calculated for this reference / comparison dataset pair. This process was repeated for
each possible pair of reference and comparison datasets.
It is worth emphasizing that OA is typically used to evaluate a classified map,
assumed to contain errors, against some validation dataset, assumed to contain no error.
In such a situation the term overall accuracy reflects the fact that one of the datasets is
inherently more trustworthy than the other. As used here, however, to evaluate
reproducibility between datasets, it is completely arbitrary which dataset is considered the
reference and which is used for comparison. In fact, as described above, OA is reported
both ways, with each dataset serving as reference and comparison. Therefore, even
though the term overall accuracy is consistent with the error matrix literature, the term
overall agreement better describes what the OA metric means in the context of
reproducibility.
Average mutual information (AMI) is a measure of the information common to
two nominal data sets. AMI was derived from information theory, specifically the
concept of entropy (Finn 1993). AMI was computed following Finn (1993). Like OA and
Kappa, AMI is expressed as a percentage of similarity between two maps, ranging from
0, indicating that map A predicts nothing about map B, to 100, indicating that map A is a
perfect predictor of map B. Unlike OA and Kappa, AMI can be computed for rectangular
contingency tables (i.e. the number of classes in map A does not have to be the same as
the number of classes in map B), and the class labels in map A and B do not have to be
the same.
71
AMI has been used much less frequently to assess remotely sensed data than OA
or Kappa, so examples may help readers with its interpretation. The limiting cases are
simplest to explain, assuming that the classes in map A are the same as those in map B. If
AMI equals 100%, then the two maps are identical, corresponding to a diagonal error
matrix. If AMI equals 0% then the classes of the pixels on map B are randomly assigned
relative to the classes of the pixels on map A, corresponding to an error matrix filled with
equal elements. If AMI equals 50% then 50% of the pixels will be the same class on map
A and map B. Finn (1993) provides a more complete explanation of AMI, including both
idealized and real examples as well as examples of the more general case when the
classes on maps A and B are not the same.
4.3.5 Echo, FFV, and Q-value correlation with survey parameters
An end user might be interested only in the reproducibility of acoustic classes
(Sections 4.3.4, 4.4.2). Quantifying the reproducibility of intermediate IMPACT
processing quantities (stacked echoes, FFVs and Q-values), however, is important to
understand the factors affecting reproducibility.
One way to quantify reproducibility was to look at the differences between echoes
as a function of variables that presumably affect them. In other words, can the differences
between echoes, or their derivatives, FFVs, and Q-values, be attributed to survey
parameters such as wind speed, or other confounding factors such as seabed slope? If so,
then removing those echoes so affected should increase reproducibility. For example, if
echoes were found to be less similar with a low grazing angle than with a high grazing
angle, eliminating off-nadir echoes should increase the reproducibility of repeat transects.
On the other hand, if there were no correlation between survey parameters and echo /
FFV / Q space reproducibility, then the source of the variability between echoes would be
72
either due to some unmeasured survey parameter, ping-to-ping variability inherent to
random scatterers on the seabed, or within-footprint heterogeneity of the seabed.
Four points, hereafter called “stations,” were chosen along the northern transect.
Station one was along the main reef, in an area of high local relief (Fig. 4.1 shows the
“main reef”). Station two was between the main reef and the outlier reef, in a flat area of
sediment. Station three was along the steeply sloping front of the outlier reef. Station four
was on the sloping forereef. For each of the four stations, the closest echo from each of
the 18 repeat transects was identified. For each echo, three dependent variables were
extracted: echo envelopes, FFVs, and Q-values. Fifteen independent variables were also
extracted: depth, date, vessel speed, sound speed, location, total grazing angle, pitch
grazing angle, and roll grazing angle. Sound speed was computed from actual water
temperature (Fig. 4.2) and nominal salinity. The three grazing angles (total, pitch, and
roll) were extracted both at the time of the center ping in the stack and for the maximum
values observed during the time interval of the stack, as described in Section 4.3.2.
Pairwise differences between the independent variables were computed for each
station. At each station there were 18 observations for each variable (six days times three
transects per day), which resulted in 153 comparisons between pairs of observations. As
noted above, the independent variables were all scalar quantities, so the differences
between them were computed as the absolute value of one measurement subtracted from
the other. In addition, the minimum value within each pair was retained for the grazing
angle variables. The logic behind using the minimum grazing angle of each pair rather
than just the difference between grazing angles was that highly off-nadir echoes were
expected to exhibit higher variance. If two echoes both had low grazing angles, the
73
difference between their grazing angles might be small even if the differences between
the echoes were large.
In contrast to the independent variables, which were all scalars, the dependent
variables were vector quantities. Therefore, simple differences between them could not
be directly correlated with the independent variables. To address this complication, three
methods were used to compute scalar differences between the vector quantities: the
correlation coefficient (2), the multidimensional angle (3), and the magnitude of the
difference vector produced by subtracting one from the other (4). For two vectors, A and
B, these quantities are defined as:
€
rAB =covABσ Aσ B
(2)
€
θAB = cos−1 A * BA B
(3)
€
dAB = A − B (4) where covAB was the covariance between vectors A and B, σA was the standard deviation
of vector A, A * B was the dot product of vectors A and B, and ||A|| was the magnitude of
vector A.
Plots were generated for all combinations of the nine dependent metrics (three
variables times three difference methods) and 15 independent variables. The resulting
135 plots were inspected for correlations between echos, FFVs, or Q-values and the
environmental variables.
74
4.4 Results
4.4.1 Environmental conditions, vessel attitude and grazing angle None of the three NOAA sensor packages had logged both wind speed and water
temperature on each of the repeat survey days, but wind speed and water temperature
were available from at least one of the stations during each survey (Fig. 4.2). May 2, and
August 13, 2007 were the calmest days. May 28 was the windiest. Water temperature
increased about 5C over the span of days surveyed.
Figure 4.2: Wind speed (left) and water temperature (right) for the periods of the six surveys. Missing bars indicate no data for that time period. For example, the Virginia Key station does not record water temperature. May 2, and August 13, 2007 were the calmest days. May 28 was the windiest. Water temperature increased about 5C over the span of days surveyed.
Two notable features were apparent in the transducer attitude data (Figs. 4.3 and
4.4). First, the median values of pitch, roll, and total angle off vertical were not zero on
any day. For example, on May 2 the daily median value of the mean within-stack roll
measurements was about -5 degrees (Fig. 4.3C). If the transducer had been perfectly
aligned with the survey vessel, a negative bias of -5 degrees would indicate that the
vessel listed to port during the survey. A more likely explanation is that the transducer
was mounted to the vessel on May 2 such that it pointed 5 degrees to starboard.
75
Figure 4.3: Boxplots of daily mean and maximum within-stack attitude. Each box shows, for all measurements on one day, the median (horizontal red line), 25th-75th percentile (blue box), data within 1.5 times the range between the 25th-75th percentile (black “whiskers”), and individual data points falling outside the whiskers (red ‘+’ markers). A) Mean within-stack pitch, B) maximum within-stack pitch, C) mean within-stack roll, D) maximum within-stack roll, E) Mean within-stack angle off vertical, F) maximum within-stack angle off vertical. Green horizontal lines in (A, B, E, F) are drawn at +/- 8 degrees, and blue horizontal lines in (C, D, E, F) at +/- 21 degrees, which are half the beamwidth in the pitch and roll directions respectively. Note daily biases.
76
Figure 4.4: Transducer attitude data from May 28, 2007. Roll (A) and pitch (B) measurements made during all six transects May 28. C) Pitch component of the grazing angle during all six transects May 28. The horizontal black line in the fourth panel marks 1/2 of the pitch beamwidth. Note that the times when the boat turned at the end of each transect are visible (marked here with vertical black dashed lines) because vessel motion was larger when headed into the wind and waves on the first, third, and fifth transects than when headed downwind on the second, fourth, and sixth transects.
Furthermore, the mounting angle biases changed from day to day (Fig. 4.3). Biases were
also visible in the raw data for a given day (Fig. 4.4).
Second, sea state affected variability of pitch and roll on both daily and transect
time scales. Pitch and roll were most variable on May 28, the windiest day (Fig. 4.3). In
particular, the maximum within-stack values were largest on May 28 (Fig. 4.3).
Maximum within-stack values were low on both May 2 and Aug 13, as expected based
77
on the low wind speeds those days (Fig. 4.2). The data from Aug 3, however, which was
windier than May 2 or Aug 13 (Fig. 4.2), had similar variability in pitch and roll to May 2
and Aug 13 (Fig. 4.3). The relatively low pitch and roll variability on Aug 3 highlighted
the fact that stability of a given vessel is a function of wind direction, fetch, duration, and
vessel heading in addition to wind speed. Pitch and roll variability changed within each
day also, as the survey vessel headed into and away from the wind (Fig. 4.4).
4.4.2 Classification reproducibility
Classification reproducibility was assessed both qualitatively and quantitatively
(Section 4.3.4). Qualitative assessment included tables of the number of classes identified
in repeat datasets and visual assessment of the consistency of classes in both geographic
and Q-space. Quantitative assessment employed the error matrix and shared information
approaches.
IMPACT identified 8 classes as the optimum split level for clustering of the entire
dataset, which totaled 9,562 echoes from all days. Clustering each of the six days
separately (independent PCA and clustering on each day) returned a variable number of
optimum classes (Table 4.1). IMPACT suggested an optimum split level that clustered
four of the six daily datasets into five classes, one day into four classes, and one day into
six classes (Table 4.1).
Table 4.1: Optimum number of clusters identified by ACE for each daily dataset clustered separately. The optimum number varied from 4-6.
Date # Classes # Stacks May 1 5 1822 May 2 6 1583 May 9 5 1983 May 28 5 1397 Aug 3 5 1393 Aug 13 4 1307
78
When processed separately, independent PCA and clustering were performed on
each of the three replicates for both north and south transects each day, resulting in 36
clustered datasets. Most individual replicate transects contained three or four classes
when clustered separately, though one leg clustered into five classes, and one leg resulted
in only two classes (Table 4.2). Thus, the optimum number of classes generally varied by
dataset, with smaller datasets, the individual transects, clustering into fewer classes and
larger datasets, the daily and all days-merged datasets, clustering into more classes.
Table 4.2: Optimum number of clusters identified by ACE for each transect when clustered separately. The “Leg” column numbers the three replicate transects within each survey day. The optimum number varied from 2-4.
North Transect South Transect Date Leg # Classes # Stacks Date Leg # Classes # Stacks
May 1 1 4 264 May 1 1 4 212 May 1 2 3 252 May 1 2 3 212 May 1 3 4 433 May 1 3 3 255 May 2 1 3 380 May 2 1 4 210 May 2 2 3 250 May 2 2 4 249 May 2 3 3 243 May 2 3 4 259 May 9 1 4 235 May 9 1 4 243 May 9 2 4 254 May 9 2 4 233 May 9 3 3 354 May 9 3 4 231 May 28 1 3 256 May 28 1 4 220 May 28 2 4 238 May 28 2 4 238 May 28 3 4 246 May 28 3 3 228 Aug 3 1 5 457 Aug 3 1 4 223 Aug 3 2 3 314 Aug 3 2 2 218 Aug 3 3 4 471 Aug 3 3 3 212 Aug 13 1 3 232 Aug 13 1 3 215 Aug 13 2 3 252 Aug 13 2 4 226 Aug 13 3 3 257 Aug 13 3 4 213
The distribution of points in Q-space was broadly similar among days (Fig. 4.5),
but the number of clusters and the locations of their boundaries shifted between days.
Generally classes with extreme values in Q-space had more stable mean values in the
different daily datasets than classes formed from points in the middle of Q-space. For
example, the mean values for the black, grey, and, yellow classes, which have the
79
smallest and largest values of Q1, were more consistent than the red and blue classes,
which contained echoes with intermediate Q1 values (Fig. 4.5).
Plotting daily classified echoes in geographic space reveals broad similarities, but
many differences (Fig. 4.6). The main and inner reefs, for example (Fig. 4.1), appear
consistently in the yellow class on both north and south transects (arrows labeled “M”
and “I” in Fig. 4.6). The transition (arrow labeled “T” in Fig. 4.6), occurring on both
Figure 4.5: Q-space of each of the six daily datasets after clustering independently. Classes have been colored so that black and grey are the first two colors for classes with the smallest Q-values, and yellow, red and blue are the first three colors for classes with the largest Q-values. August 13 had only four classes, so no yellow was used. May 2 had six classes so green was used for the least populous one. Note that the mean values for the black, grey, and, yellow classes are more consistent than the red and blue classes.
80
Figure 4.6: All 18 replicates of both the northern and southern transects. The left two panels plot all northern (A) and southern (C) transects using their actual coordinates. The right two panels plot all northern (B) and southern (D) transects with 50 m north-south offset between them. Transects in (B) and (D) are grouped in date order, so, for example, the top three are from 1 May and the bottom three are from 13 Aug. Colors in both plots correspond to the Q-spaces shown previously (Fig. 4.5). Y-axis is severely stretched on both plots to space points out for clarity. Note that transects are broadly similar, but not identical. The inner (I) and main (M) reefs appear consistently in the yellow class and the transition (T) between the shelf and slope also appears consistently as a border between either red or blue and grey classes.
81
transects just seaward of the outlier reef, between the red or blue classes on the shelf and
the grey or black classes on the slope also appears consistently defined (compare profile
from Fig. 4.1 with plan view of Fig. 4.6). The distribution of red and blue classes on the
shelf, however, varied day to day and even from transect to transect within the same day
(Fig. 4.6).
Plotting echoes from the individually classified transects in geographic space
reveals broad patterns of agreement between replicates, but many differences in the
details. As observed for the daily datasets, the transition between the shelf and slope
consistently appeared as a border between classes in the individual transects (arrows
labeled “T” in Figs. 4.6, 4.7, 4.8). Unlike the daily datasets, which generally identified a
separate class for the inner and middle reefs (arrows labeled “M” and “I” in Fig. 4.6), the
transects classified individually did not segment these reefs as separate classes (Figs. 4.7,
4.8).
Differences between successive transects that had been classified separately were
often apparent even for transects with the same number of clusters (Figs. 4.7, 4.8). On
May 2, 2007, for example, the first two northern transects each had three clusters, but the
spatial distribution of these classes was not the same between transects (Fig. 4.7A, B, D,
E). The gray class was consistent between the first two replicates of the northern transect
on May 2, but the red and blue classes were not (Fig. 4.7A, B, D, E). The third replicate
of the north transect on May 2, 2007 was clustered into four classes, so differs from the
first two transects in that respect. Nevertheless, the west end of the third transect (Fig.
4.7C, F) appeared similar to the first transect (Fig. 4.7A, B, D, E).
82
The differences in vessel motion between successive transects, caused by heading
into and then away from the wind, were significant on some days (Fig. 4.4), so one might
expect to find greater similarity between transects with similar vessel motion. For
example, reproducibility might be expected to be greater on a calm day than a windy day
because on a calm day vessel pitch and roll will be minimal regardless of heading during
data acquisition. Likewise, on a windy day transects acquired when headed in the same
direction relative to the wind might be more similar than transects acquired when headed
in opposite directions. The classified individual transects do not necessarily support these
hypotheses, however. May 2 was the most calm survey day, yet the three replicates of the
northern transect did not appear particularly reproducible (Fig. 4.7). May 28 was the
windiest survey day, exhibiting notable differences in pitch and roll on upwind and
downwind legs, yet the two transects acquired downwind (Fig. 4.8A, D, G and C, F, I)
were not noticeably more or less similar to each other than to the transect acquired going
upwind, when vessel motion was much larger (Fig. 4.8B, E, H). These observations
suggest that either the pitch and roll of the vessel do not affect classification
reproducibility or event the small pitch and roll measured on May 2 was large enough to
reduce reproducibility. Section 4.5.2 attempts to resolve which of these conclusions is
correct.
The first approach to quantifying reproducibility employed the error matrix to
derive overall accuracy and Kappa statistics. As the daily datasets had clustered into
different numbers of classes (Table 4.1), the first step required was to renumber the
clusters on each day into a common classification scheme. Renumbering of classes was
done by inspecting the positions of the class means in Q-space (Fig. 4.9). IMPACT had
83
identified clusters with class mean Q1 values near -2, -1.4, -0.8 on all six of the daily
datasets (Fig. 4.9). Clusters on each day with mean values closest to these points in Q-
space were renumbered 1-4 and assigned colors black, grey, and green, respectively (Fig.
4.9). IMPACT had identified clusters with class mean Q1 values near 0.2 on two of the
six daily dataset and clusters with class mean Q1 values near -0.4 and +0.4 on five of the
Figure 4.7: Independently classified replicate track lines, Q-space, and pitch measurements along the northern transect on May 2, 2007. A, B, C) All transects on May 2 in light gray, and the actual classified data from a particular line in color (including dark gray). D, E, F) Clustered Q-space for each transect. G, H, I) Pitch measured during each transect. Green lines mark +/- eight degrees, which is the half angle of the transducer in the pitch direction, in green. The arrow labeled “T” marks the transition between the shelf and slope (as in Fig. 4.6). Note clear differences among the transects even though vessel motion was small on this day.
84
Figure 4.8: Independently classified replicate track lines, Q-space, and pitch measurements along the northern transect on May 28, 2007. A, B, C) All transects on May 28 in light gray, and the actual classified data from a particular line in color (including dark gray). D, E, F) Clustered Q-space for each transect. G, H, I) Pitch measured during each transect. Green lines mark +/- eight degrees, which is the half angle of the transducer in the pitch direction, in green. The arrow labeled “T” marks the transition between the shelf and slope (as in Fig. 4.6). Note that transects one (A, D) and three (C, F), acquired when headed downwind, were not much different than transect two (B, E) acquired headed into the wind, even though vessel motion was much greater on transect two (H) than on transects one or three (G, I).
six daily datasets (Fig. 4.9). Clusters on each day with mean values closest to these points
in Q-space were renumbered 5, 4, and 6 and assigned the colors red, blue, and yellow,
respectively (Fig. 4.9).
Once the classes had been renumbered, an error matrix was constructed and OA
and Kappa were computed. OA and Kappa showed some consistency between daily
85
classified datasets, but generally these values were fairly low (Table 4.3). Most of the OA
values fell in the range of about 50% to 65%, with the exception of the comparisons
between August 13 and any of the other days, which were in the range of 30% - 48%.
Figure 4.9: Illustration of the method and results of renumbering of daily classes. Top left: mean +/- one standard deviation of each cluster from each daily dataset. All of the clusters on a given day have the same color in the top plot. Bottom left: same cluster means +/- one standard deviation as the top panel, but colors correspond to the new class numbers. Right: All of the points for each day plotted in their respective Q-spaces using the new class colors from the bottom left panel.
Table 4.3: Overall accuracy (left) and Kappa (right) between pairs of daily classified datasets (reclassed to 6 classes).
For the repeat QTCV datasets, AMI proved an appealing alternative to OA and
Kappa because, in general, the number of classes is not constant across datasets.
Nevertheless, a reclassed version of the QTCV datasets was created above for the OA
86
and Kappa calculations, so the approach taken was to compute AMI on both the
reclassified daily datasets and the original ACE output with differing numbers and labels
of classes each day (Table 4.4). AMI values for the daily classified datasets ranged from
about 40% to 50%, with a few pairs as low as 35%. The difference between AMI
computed with the reclassified data and AMI computed on the original ACE output was
only a few percentage points in each case.
Table 4.4: Percent AMI for the reclassed ACE-best (left) and the original ACE-best dataset (right).
4.4.3 Echo, FFV, and Q-value correlation with survey parameters
The four stations chosen as locations from which to extract echoes, FFVs, and Q-
values for correlation with survey parameters spanned the range of classes identified by
IMPACT on the daily classified datasets (Fig. 4.10). Plots of differences between pairs of
echoes extracted at each station were prepared for each for all combinations of the nine
dependent metrics (three variables times three difference methods) and 15 independent
variables. Numerical correlations between all pairs of variables were low. The resulting
135 plots were also visually inspected for any apparent relationships between the
variables. Rather than including all the plots here, the results were summarized (Table
4.5) and a few of the more interesting plots are included as examples (Figs. 4.11, 4.12,
4.13, 4.14).
87
Figure 4.10: Plots illustrating the locations of four stations from which echoes, FFVs, Q-values, and survey parameters were extracted. Top panel shows all 18 transects grouped and colored by day. The transects were on top of one another (Fig. 4.6), but are here plotted offset to the south by 50 m each day to reduce clutter. The four stations, numbered 1-4, are plotted on each group of transects. Bottom panel shows the locations of the four stations along the August 13 cross shelf profile. Note that stations spanned the range of classes.
Most of the pairs of variables showed no apparent correlation (blue boxes in
Table 4.5; see Fig. 4.11, for example). For some pairs of variables, the values from
station 3 were sufficiently distinct from the values for the other stations that a general
trend was evident when comparing station 3 against the others (red boxes in Table 4.5;
88
see Fig. 4.12, for example). Four of the pairs of variables exhibited correlation when only
stations 3 and 4 were considered (orange boxes in Table 4.5; see Fig. 4.13, for example).
Finally, two pairs of variables showed a correlation when stations 2, 3, and 4 were
considered, ignoring station 1 (yellow boxes in Table 4.5; see Fig. 4.14, for example).
When differences between the dependent variables were measured with the
correlation coefficient (r), no relationships were observed for any variables (Table 4.5).
The only independent variable that appeared to have any affect on survey results was
Table 4.5: Summary of visual assessment of correlation between envelopes, FFVs, or Q-space and survey variables. The dependent variables, envelopes, FFVs, and Q space, are ordered across the top, each with three columns corresponding to the three methods for computing the vector differences. Independent variables are listed in the rows, divided into three groups. The top group contains general survey parameters. The middle group contains parameters related to the grazing angle at the time of the center ping. The bottom group contains parameters related to the minimum grazing angle (most off-nadir) over the duration of the stack span. The method for comparing independent variables is shown in parenthesis: (Δ) indicates the absolute value of the difference; (min) indicates the lower of the two values being compared. The color of each entry in the table reflects which of the four stations were found to have a correlation between each pair of variables (see text and Figs. 4.11, 4.12, 4.13, 4.14 for detailed explanation of the legend). Grazing angle in the pitch direction had the most apparent effect on envelopes, FFVs or Q-space.
89
Figure 4.11: Plot of multidimensional angle (θ) between echo envelopes versus the distance between echoes for all pairwise comparisons at all four test stations. There was no apparent relationship between the similarity of echoes at the same station and the separation of those echoes, even for large distances. The lack of apparent correlation is a typical example of the plots marked with blue squares in Table 4.5.
Figure 4.12: Plot of multidimensional angle (θ) between echo envelopes versus the total minimum grazing angle at the times of center echo of each stack. Taken altogether as an ensemble there was a negative relationship between echo similarity and grazing angle, but this was entirely due to the differences between station 3 and the other stations. Removing station 3 or considering any one station individually, there was no apparent relationship. This was a typical example of the plots marked with red squares in Table 4.5.
90
Figure 4.13: Plot of the magnitude of the difference between echoes versus the minimum grazing angle in the pitch direction at the times of center echo of each stack. Considering stations 3 (blue) and 4 (cyan) only, there was a negative relationship between echo similarity and grazing angle, but this was entirely due to the overall differences between stations. Removing station 3 or 4, or considering any one station individually, there was no apparent relationship. This is an example of the plots marked with orange squares in Table 4.5.
Figure 4.14: Plot of the magnitude of the difference between FFVs versus the minimum grazing angle in the pitch direction at the times of center echo of each stack. Considering stations 2 (green), 3 (blue), and 4 (cyan) only, there was a negative relationship between echo similarity and grazing angle, but this was entirely due to the overall differences between stations. Removing station 3, or considering any one station individually, there was no apparent relationship. This is an example of the plots marked with yellow squares in Table 4.5.
91
grazing angle, specifically the grazing angle in the pitch direction. Furthermore, this
effect of grazing angle generally appeared only when using the minimum value between
the two stacks being compared, not when using the difference, indicating that the
problem was with high grazing angles relative to beamwidth, as others have shown (von
Szalay and McConnaughey 2002; Biffard et al. 2005).
In summary, the data suggest that the only survey-related variable affecting
reproducibility of the echo envelopes, FFVs, and Q-space values was the grazing angle.
In these experiments, the grazing angle in the pitch direction was more important than in
the roll direction. This sensitivity to pitch was due to the different transducer beam
widths in pitch (12 degrees) and roll (42 degrees). On one hand, this result was
encouraging, suggesting that QTCV classification results were not particularly sensitive
to survey parameters. On the other hand these results were discouraging because they do
not point to any obvious way to improve the classification reproducibility with the
exception of using echoes restricted to large grazing angles (Section 4.5.2).
4.5 Attempts to improve reproducibility Classification of the repeat Fowey Rocks transects using the standard IMPACT
processing was not particularly reproducible (Section 4.4), motivating the investigation of
the second question posed by this study, “Can reproducibility be improved by altering the
standard usage of the off-the-shelf software?” This Section describes several alternate
methods of data processing that were attempted to improve the results from the baseline
processing. Four changes were investigated: using fewer classes than suggested by ACE
(Section 4.5.1), discarding off-nadir echoes (Section 4.5.2), using a robust PCA (Section
4.5.3), and using a subset of IMPACT features (Section 4.5.4).
92
4.5.1 Cluster into fewer classes
The ACE clustering routine finds a statistically optimum set of clusters based on
the Bayesian Information Criterion (BIC), but the clusters found do not necessarily have
physical meaning. It may be that ACE segmentation, although optimum from a statistical
point of view, over-segments the feature space from a physical perspective. If this were
the case then an alternate clustering using fewer classes could improve reproducibility.
Two alternate groups of clusters were investigated. The first started with the ACE-
defined optimum number of clusters and merged certain classes together; the second used
just two ACE-defined clusters.
4.5.1.1 Merge baseline classes four through six into a single class The baseline processing (Sections 4.3-4.4) resulted in four to six classes per day
(Fig. 4.5). The three classes on each day with class mean Q1 values near -2, -1.5, and -
0.75 were more consistently identified than the one to three classes each day with larger
class mean Q1 values (Fig. 4.9). All six days had classes with mean Q1 values near -2, -
1.4, and -0.8, but only five of the six days had classes with mean Q1 values near -0.4 and
0.4, and only two of the six days had classes with mean Q1 values near 0.2 (Fig. 4.9).
These observations suggested that classes one through three (see Fig. 4.9 for numbering)
were more reproducible than classes four through six. In this experiment, classes one
through three were retained exactly as in the baseline case, but classes four through six
were merged so that each day had four classes. The OA, Kappa (Table 4.6) and AMI
(Table 4.7) values were computed for this modified configuration of classes. OA and
Kappa and AMI increased slightly with the merged classes relative to the baseline
processing.
93
Table 4.6: Overall accuracy (left) and Kappa (right) between pairs of daily classified datasets reclassed to 6 classes but then with classes 4-6 merged to a single class.
Table 4.7: Percent AMI for daily classified datasets reclassed to 6 classes but then with classes 4-6 merged to a single class.
4.5.1.2 ACE-2 Multiple QTCV surveys in areas containing coral reefs have consistently
discriminated two broad seabed types (see Chapter 3): sediment and hard bottom (rock).
In the “ACE-2” experiment, the IMPACT autocluster results for two classes were used,
even though splitting the dataset into more classes resulted in a lower BIC. Classes were
renumbered to be consistent among days (Fig. 4.15), an error matrix was tabulated, and
the OA, Kappa and AMI (Table 4.9) values were then computed for this modified
configuration of classes. OA values of the datasets with just two ACE clusters ranged
between 80% to 92% (Table 4.8), which was greater than the baseline 35% - 67% (Table
4.3). Kappa values of the datasets with just two ACE clusters ranged between 57% to
84% (Table 4.8), which was greater than the baseline 19% - 55% (Table 4.3). AMI
values of the datasets with just two ACE clusters ranged between 27% to 60% (Table
4.9), which was both better and worse than the baseline 36% - 51% (Table 4.4).
94
Table 4.8: Overall accuracy (left) and Kappa (right) between pairs of daily datasets classified by ACE with just two clusters.
Table 4.9: Percent AMI for pairs of daily datasets classified by ACE with just two clusters.
Figure 4.15: Plots of daily Q-space (left) and renumbering of daily classes (right) for datasets classified by ACE with just two clusters.
95
4.5.2 Keep only near-nadir echoes
The observation that grazing angle was the only survey parameter related to the
consistency of envelopes, FFVs, and Qspace (Table 4.5) is consistent with other results
that have shown seabed slope to be one factor that can degrade the accuracy of single-
beam echosounder seabed classification (von Szalay and McConnaughey 2002; Biffard et
al. 2005). It was logical, therefore, to predict that echoes with high grazing angles might
lower the reproducibility of QTCV surveys.
A series of experiments were conducted in which echoes were filtered based on
their roll and pitch to keep only those that were near-nadir. Two filtering criteria were
used: A) echoes greater than five degrees off vertical were discarded, and B) echoes
greater than two degrees off vertical in the pitch direction were discarded. These criteria
removed so many echoes, however, that most of the days had few to no echoes left for
comparison. Therefore, the entire dataset was reprocessed with a stack size of one (i.e.
not stacked) and the following filters were applied: A) Echoes greater than five degrees
off vertical were discarded; B) Echoes greater than two degrees off vertical in the pitch
direction were discarded; C) Echoes with incidence angle greater than five degrees were
discarded.
In all cases, the method for processing was to export the FFVs for each echo (or
stacked echo), discard the ones too far off nadir, then compute the Q-values of the
remaining echoes using the MATLAB “princomp” command. The resulting Q-values
were then formatted as an IMPACT “.dat” file and clustered in ACE. The number of
echoes remaining after clustering, optimum number of clusters found in ACE, and
number of ACE iterations are summarized in Table 4.10. Results are summarized in
Sections 4.5.2.1-4.5.2.5.
96
Table 4.10: Summary of off-nadir experiments. N/A means that after filtering there were an insufficient number of echoes remaining to cluster the dataset.
4.5.2.1 Within 5 degrees of vertical (echoes stacked by 5) Only three of the six days had enough points remaining after filtering to attempt
clustering (Fig. 4.16). These days had different numbers of optimum clusters (Table 4.10)
so the classes were renumbered to be consistent among days (Fig. 4.16), an error matrix
was tabulated, and the OA, Kappa and AMI (Tables 4.11, 4.12) values were then
computed for this modified configuration of classes.
Figure 4.16: Plot of Q-space for each day (left) after filtering out all stacks with maximum transducer pointing vector greater than 5 degrees off vertical, clustering with ACE, and renumbering the classes to be consistent from day to day. From top to bottom these days were 5/9, 8/3, and 8/13. Renumbering of daily ACE classes (right).
97
Overall accuracy values of the datasets after discarding echoes greater than 5
degrees off vertical ranged between 53% to 66% (Table 4.11), which was about the same
as the baseline 35% - 67% (Table 4.3). Kappa values of the datasets after discarding
echoes greater than 5 degrees off vertical ranged between 37% to 52% (Table 4.11),
which was about the same as the baseline 19% - 55% (Table 4.3). AMI values of the
datasets after discarding echoes greater than 5 degrees off vertical ranged between 48%
to 73% (Table 4.12), but AMI values for only one daily pair was better than the baseline
36% - 51% (Table 4.4).
Table 4.11: Overall accuracy and Kappa coefficient between pairs of datasets clustered by day after filtering out all stacks with maximum transducer pointing vector greater than 5 degrees off vertical.
Table 4.12: Percent AMI between pairs of datasets clustered by day after filtering out all stacks with maximum transducer pointing vector greater than 5 degrees off vertical.
4.5.2.2 Pitch within 2 degrees of vertical (echoes stacked by 5) Only four of the six days had enough points remaining after filtering echoes
greater than 2 degrees off vertical in the pitch direction to attempt clustering (Fig. 4.17).
These four days had different numbers of optimum clusters (Table 4.10) so the classes
98
were renumbered to be consistent among days (Fig. 4.17), an error matrix was tabulated,
and the OA, Kappa and AMI (Tables 4.13, 4.14) values were then computed for this
modified configuration of classes.
Figure 4.17: Plot of Q-space for each day (left) after filtering out all stacks with maximum pitch greater than 2 degrees, clustering with ACE, and renumbering the classes to be consistent from day to day From top to bottom these days were 5/1, 5/2, 5/9, and 5/28.Renumbering of daily ACE classes (right).
Overall accuracy values of the datasets after discarding echoes greater than 2
degrees off vertical in the pitch direction ranged between 67% to 90% (Table 4.13), an
improvement over the baseline 35% - 67% (Table 4.3). Kappa values of the datasets
after discarding echoes greater than 2 degrees off vertical in the pitch direction ranged
between 17% to 85% (Table 4.13), which, except for the comparison between May 2 and
May 28, represented an improvement over the baseline 19% - 55% (Table 4.3). AMI
values of the datasets after discarding echoes greater than 2 degrees off vertical in the
pitch direction ranged between 46% to 79% (Table 4.14), an improvement over the
baseline 36% - 51% (Table 4.4).
99
Table 4.13: Overall accuracy (left) and Kappa (right) between pairs of datasets clustered by day after filtering out all stacks with maximum pitch greater than 2 degrees off vertical.
Table 4.14: Percent AMI between pairs of datasets clustered by day after filtering out all stacks with maximum pitch greater than 2 degrees off vertical.
4.5.2.3 Within 5 degrees of vertical (echoes stacked by 1) After filtering unstacked echoes greater than 5 degrees off vertical, all six days
had enough points remaining to attempt clustering (Fig. 4.18). The optimum number of
clusters was not constant across all days (Table 4.10), so the classes were renumbered to
be consistent among days (Fig. 4.18), an error matrix was tabulated, and the OA, Kappa
and AMI (Tables 4.15, 4.16) values were then computed for this modified configuration
of classes.
Overall accuracy values of the datasets after discarding unstacked echoes greater
than 5 degrees off vertical ranged between 54% to 73% (Table 4.15), a slight
improvement over the baseline 35% - 67% (Table 4.3). Kappa values of the datasets
after discarding unstacked echoes greater than 5 degrees off vertical ranged between 33%
to 52% (Table 4.15), about the same as the baseline 19% - 55% (Table 4.3). AMI values
of the datasets after discarding unstacked echoes greater than 5 degrees off vertical
ranged between 26% to 37% (Table 4.16), a decrease relative to the baseline 36% - 51%
(Table 4.4).
100
Figure 4.18: Plot of Q-space for each day (left) after filtering out all echoes with maximum transducer pointing vector greater than 5 degrees off vertical, clustering with ACE, and renumbering the classes to be consistent from day to day. Renumbering of daily ACE classes (right).
Table 4.15: Overall accuracy (left) and Kappa (right) between pairs of datasets clustered by day after filtering out all unstacked echoes with maximum transducer pointing vector greater than 5 degrees off vertical.
Table 4.16: Percent AMI between pairs of datasets clustered by day after filtering out all unstacked echoes with maximum transducer pointing vector greater than 5 degrees off vertical.
101
4.5.2.4 Pitch only within 2 degrees of vertical (echoes stacked by 1) After filtering unstacked echoes with pitch angles greater than 2 degrees, all six
days had enough points remaining to attempt clustering (Fig. 4.19). The optimum number
of clusters was not constant across all days (Table 4.10), so the classes were renumbered
to be consistent among days (Fig. 4.19), an error matrix was tabulated, and the OA,
Kappa and AMI (Tables 4.17, 4.18) values were then computed for this modified
configuration of classes.
Figure 4.19: Plot of Q-space for each day (left) after filtering out all unstacked echoes with pitch greater than 2 degrees, clustering with ACE, and renumbering the classes to be consistent from day to day. Renumbering of daily ACE classes (right).
Overall accuracy values of the datasets after discarding unstacked echoes with
pitch angles greater than 2 degrees ranged between 49% to 71% (Table 4.17), about the
same as the baseline 35% - 67% (Table 4.3). Kappa values of the datasets after
discarding unstacked echoes with pitch angles greater than 2 degrees ranged between
28% to 52% (Table 4.17), about the same as the baseline 19% - 55% (Table 4.3). AMI
values of the datasets after discarding unstacked echoes with pitch angles greater than 2
102
degrees ranged between 26% to 37% (Table 4.18), a decrease relative to the baseline 36%
- 51% (Table 4.4).
Table 4.17: Overall accuracy (left) and Kappa (right) between pairs of datasets clustered by day after filtering out all unstacked echoes with pitch greater than 2 degrees.
Table 4.18: Percent AMI between pairs of datasets clustered by day after filtering out all unstacked echoes with pitch greater than 2 degrees.
4.5.2.5 Incidence angle less than 5 degrees (echoes stacked by 1) After filtering unstacked echoes with incidence angles greater than 5 degrees, all
six days had enough points remaining to attempt clustering (Fig. 4.20). The optimum
number of clusters was not constant across all days (Table 4.10), so the classes were
renumbered to be consistent among days (Fig. 4.20), an error matrix was tabulated, and
the OA, Kappa and AMI (Tables 4.19, 4.20) values were then computed for this modified
configuration of classes.
Overall accuracy values of the datasets after discarding unstacked echoes with
incidence angles greater than 5 degrees ranged between 58% to 74% (Table 4.19), a
slight improvement over the baseline 35% - 67% (Table 4.3). Kappa values of the
datasets after discarding unstacked echoes with incidence angles greater than 5 degrees
103
ranged between 31% to 50% (Table 4.19), about the same as the baseline 19% - 55%
(Table 4.3). AMI values of the datasets after discarding unstacked echoes with incidence
angles greater than 5 degrees ranged between 28% to 37% (Table 4.20), a decrease
relative to the baseline 36% - 51% (Table 4.4).
Figure 4.20: Plot of Q-space for each day (left) after filtering out all unstacked echoes with incidence angle greater than 5 degrees, clustering with ACE, and renumbering the classes to be consistent from day to day. Renumbering of daily ACE classes (right).
Table 4.19: Overall accuracy (left) and Kappa (right) between pairs of datasets clustered by day after filtering out all unstacked echoes with incidence angle greater than 5 degrees.
Table 4.20: Percent AMI between pairs of datasets clustered by day after filtering out all unstacked echoes with incidence angle greater than 5 degrees.
104
4.5.3 Stability of principal components
One of the characteristics of the traditional principal components transform is that
it is susceptible to outliers (e.g. Jackson and Chen 2004). Since each day (or transect) was
processed independently, thereby generating independent sets of principal components,
an important question affecting reproducibility is whether differences between the days
could affect the eigenvectors enough to destabilize the resulting clusters. Three
approaches were taken to investigate this possibility: use of a robust PCA, use of a single
set of eigenvectors for all days, and consideration of dataset sample size.
4.5.3.1 Robust PCA A number of modifications to the traditional PCA have been proposed that are
designed to detect, filter, or otherwise improve the stability of the transform. Collectively,
these techniques may be called “robust PCA” even though there are several ways to
implement the concept (Rousseeuw 1984; Chen et al. 1994; Chen and Jackson 1995;
Kosinski 1998; Filzmoser 1999; Locantore et al. 1999; Marden 1999; Visuri et al. 2000;
Jackson and Chen 2004).
Robust PCA seemed like a good way to stabilize repeat QTCV datasets, but
actually it may be counterproductive because robust PCA techniques are designed to
operate on datasets with single multivariate normal populations whereas the datasets used
in seabed classification should contain samples from multiple multivariate populations.
Sample calculations with synthetic data illustrate this point (Fig. 4.21). Consider two
multivariate normal populations with the same covariance matrix [1 0.4; 0.4 1]. The only
difference between the two is the locations of their means. Population one has mean [0
0], and population two has mean [-2 2]. Draw twenty random points from population one
(the green dots in Fig. 4.21) and five random points from population two (red triangles in
105
Fig. 4.21). Compute the PCA on this set of 25 points and plot the first eigenvector (red
line in Fig. 4.21). The direction of the first eigenvector reflects the fact that the main
source of variance in this dataset is the difference between the two population means.
Consider a second random sample taken from population two (blue circles in Fig. 4.21),
and the PCA computed on the dataset formed by the green dots from population one and
this second sample from population two. The first eigenvector from this computation is
plotted as a blue line on Figure 4.21. Note that the blue eigenvector and the red
eigenvector on Figure 4.21, although broadly aligned, are not identical. If robust PCA
could stabilize this variation in the directions of the principal components from day to
day it would be potentially useful for increasing consistency of the eigenvectors
computed for repeat transects.
Robust PCA, however, was designed to work with single multivariate normal
populations. In other words, it treats a second population as noise and ignores it. In our
example, since the sample drawn from the first population was larger, it would discard
the points from the second population and orient the first eigenvector along the principal
axis of population one, not along the principal axis of the entire dataset. In Figure 4.21
this is illustrated by the green line, which, it should be noted, is not actually a robust PC
computed on the full dataset, it is in fact the first eigenvector computed using traditional
PCA on just the data from the first population (green dots). The green eigenvector is,
however, what the robust methods would be attempting to estimate.
106
Figure 4.21: Plot illustrating the concept of robust PCA. Green dots are random points selected from one multivariate normal population. Red triangles and blue circles are two sets of random points selected from a second multivariate normal population. Red line is the first eigenvector (direction of the first PC) computed for the set of points comprised of the green dots and the red triangles. Blue line is the first eigenvector (direction of the first PC) computed for the set of points comprised of the green dots and the blue circles. Green line is the first eigenvector (direction of the first PC) computed for just the green dots. Axes are arbitrary units. Regular PCA orients the primary axis to highlight the variance between the two populations (red and blue lines). Robust PCA treats the smaller population as noise and orients the primary axis to highlight the variance in the larger population (green line).
The conclusion from this simple example is that for a dataset drawn from multiple
populations, standard PCA is more likely to respond to the differences in mean values
between the populations, whereas robust PCA is more likely to respond to the variance in
the largest subpopulation. When processing acoustic data in IMPACT, the goal of the
PCA is to enhance the separation between populations while minimizing the within
population variance, therefore the robust PCA may well be a disadvantage.
Finally, the stability of the standard PCA should increase with the number of
points. Since the acoustic datasets usually have thousands of points, the effects of day-to-
day variation should be much less than the simple example above (Fig. 4.22).
107
Figure 4.22: Plots illustrating the concept of robust PCA with more points. Same as figure 4.21 except with 400 and 100 points (left) and 4000 and 1000 points (right) from populations one and two, respectively. Note how the difference between the red and blue vectors decreases with increasing numbers of points.
4.5.3.2 Consistent PCA Another way to test the importance of day-to-day variability in PCA was to use a
single set of eigenvectors (catalog) for all six days. Two methods were tested: the catalog
from May 1, 2007 was applied to all days, and the median eigenvectors computed from
the set of eigenvectors from all six days were applied to all days. Both methods produced
similar results. Therefore, only those for the median eigenvector are shown (Tables 4.21,
4.22).
OA values of the datasets processed with the median eigenvectors for all six days
ranged between 45% to 69% (Table 4.21), which was similar to the baseline 35% - 67%
(Table 4.3). Kappa values of the datasets processed with the median eigenvectors for all
six days ranged between 33% to 59% (Table 4.21), which was similar to the baseline
19% - 55% (Table 4.3). AMI values of the datasets processed with the median
eigenvectors for all six days ranged between 37% to 53% (Table 4.22), which similar to
the baseline 36% - 51% (Table 4.4).
108
Table 4.21: Overall accuracy (left) and Kappa (right) between daily classified datasets created from a single PCA using the median eigenvector of all six days.
Table 4.22: Percent AMI for daily classified datasets created from a single PCA using the median eigenvector of all six days.
4.5.3.3 Dataset Size The number of echoes being submitted to ACE appears to affect the optimal
number of clusters found by ACE, as evidenced by the results of the daily and transect
clustering (Fig. 4.5, Table 4.2). The entire dataset with all six days merged had ~9500
stacks and ACE identified eight classes as the optimal split level. The daily datasets each
had on the order of 1500 stacks and ACE identified four to six classes as the optimal split
level. The individual transect datasets had about 250-300 stacks and ACE identified three
to four classes as the optimal split level. If the number of classes were a strong function
of the number of echoes in the dataset, then reproducibility would be compromised
because neither every day, nor every transect within a day, had the same number of
echoes.
The first test was to subset the original dataset starting with all transects on all
days merged together. The optimum number of classes decreased with number of records
(Table 4.23). The decrease was smooth, however, except for the subset by a factor of 30
109
(Table 4.23). In the range of the number of stacks that corresponded to the daily datasets
(on the order of 1300-2000) the optimum number of classes was consistent for all the
skip factors from 3-10. In the range of the number of stacks that corresponded to the
individual transects, however, (on the order of 200-400) there did appear to be a bit more
variability with number of stacks.
Table 4.23: Optimum number of classes as determined by ACE as a function of dataset size for the entire merged Fowey rocks dataset.
Skip # Usable records (stacks) ACE optimum # 0 9562 8 1 4784 6 2 3183 6 3 2391 5 4 1909 5 5 1589 5 6 1358 5
10 865 5 20 454 4 30 307 5 40 234 4
A second test was to take a different, even larger, dataset to see if these patterns
held. The dataset used was acquired at Watson’s Reef, FL, in two pieces on August 7 and
8, 2007. The data were processed using the same parameters as the Fowey rocks data
(Chapter 6), and were in the same depth range. There were 18090 stacks in the complete
dataset. Subsets of this dataset were fairly consistent in size (Table 4.24).
Table 4.24: Optimum number of classes as determined by ACE as a function of dataset size for the Watson’s Reef dataset.
Skip # Usable records (stacks) ACE optimum # 0 18090 4 1 9041 4 2 6025 3 3 4526 4 4 3615 4
110
It appears that the different number of echoes submitted to ACE clustering may
be at least partially responsible for the overall differences between individual transects,
daily data, and the entire merged dataset, but this is probably not the reason for
discrepancies within each of those subsets. The numbers of classes found were not
monotonic with number of echoes and were more variable with small differences in
numbers of echoes than the subset experiments would suggest (Tables 4.1, 4.2).
4.5.4 Features least subject to ping-to-ping variability
IMPACT computes 166 features for each (stacked) echo. These are then reduced
to three principal components with PCA. Perhaps some of the features are more sensitive
to ping-to-ping variability than others. If this were true, then using only that subset of
features from the full FFV that were least sensitive to ping-to-ping variability might
improve the reproducibility of the classification.
To test the susceptibility of IMPACT FFV elements to inherent random
variability of scattering elements within the seabed, a synthetic dataset was generated
using the Bottom Response from Inhomogeneities and Surface (BORIS) numerical model
(Bergem et al. 1999; Pouliquen et al. 1999). The advantage of the BORIS dataset for this
experiment was that the source of variability between echoes of the same seabed type is
known to be noise due to incoherent backscattering.
In the BORIS dataset, three seabed types were specified that differed from one
another in acoustic impedance and surface roughness. BORIS was used to generate 500
simulated echoes for each of the three seabed types. Within each group of echoes from
the same bottom type, the only variable was the horizontal (x,y) position of the transducer
with respect to the seabed; the angle of the transducer was constant (normal to the
seabed), the water depth was constant, and the seabed was flat. The simulated echoes
111
were processed in IMPACT, generating a total of 300 stacked echoes: 100 stacks of 5 for
each of the three seabed types (Fig. 4.23).
A boxplot of all 300 FFV values shows that most of the variability in the dataset
was contained in the first 38 features (Fig. 4.24A). For each of the three seabed types the
coefficient of variation (CV) was computed for each feature as the standard deviation of
that feature for that seabed type divided by the mean of that feature for that seabed type.
Plotting the coefficient of variation for each seabed type shows that some features have
low CV (the first 31 features, numbers 41-44, and number 102) whereas other features
have higher CV (Fig. 4.24B).
Figure 4.23: Echogram created from the BORIS dataset. Each column is a stacked echo envelope averaging five simulated echoes. Each row is an IMPACT sample. Echoes are grouped by bottom type. The first 100 columns are from a simulated sandy seabed. The second and third 100 columns are from simulated low-relief and high-relief rocky seabeds, respectively. Note that all of the variability between echoes of the same seabed type is caused by incoherent backscattering, representing noise in the clustering process.
112
Figure 4.24: BORIS dataset FFVs, FFV coefficient of variation, and loadings of the first principal component computed from the FFVs. A) Boxplot of features for all BORIS echoes (N=300 echoes). Blue line is the median, grey boxes cover the middle 50% of the data, grey “whiskers” extend to the lesser of 1.5 times the range of the grey box, or the extremes of the dataset, and red crosses mark values outside the range of the whiskers. B) FFV coefficient of variation for the three seabed types modeled. C) Loadings of the first principal component computed with all 166 features (blue) and six subsets of the features (other colors; see Table 4.25 for colors). Note that most of the features in the range 45-166 have high CV but contribute little to the first principal component.
In the BORIS dataset, all variation in echoes of the same seabed type was due to
ping-to-ping variability caused by the stochastic arrangement of scattering elements on
the seabed. The robust features, therefore, are the ones with low CV. An ideal feature for
this dataset would be constant within a seabed type but would vary among seabed types,
resulting in a CV of zero for the features within each seabed type. Features with high CV
113
have large intraseabed variability with respect to their mean values, so are more sensitive
to the random variability in the dataset than the features with low CV.
Principal components were computed using the matlab “princomp” function using
the entire 166 features (equivalent to IMPACT) and six subsets of the full 166 features
(Table 4.25; Fig. 4.24C; Fig. 4.25 left). In general, as fewer features were submitted to
the PCA, the internal variability of the resulting clusters decreased, which was the goal.
Of course, decreasing the number of features sent to the PCA also decreased the
separation between the clusters, an undesirable side effect. Of the six subsets tested, only
feature sets two and three (Table 4.25) retained separation between all three of the
BORIS seabed types (Fig. 4.25 left). Relative to feature set two, feature set three had
smaller within-class variability (tighter clusters) but also slightly less separation between
clusters (Fig. 4.25 left).
Figure 4.25: Q-space for the BORIS dataset using seven different subsets of features input to the PCA. Left: the first two components of Q-space for the BORIS dataset using seven different subsets of features input to the PCA. Each point represents one stacked echo. The colors of the points are determined by the feature set input to compute the PCA (Table 4.25). The marker shape for each point represents the seabed type: squares for sediment, dots for low relief hard bottom (rock), and triangles for high relief hard bottom. Right: the first three components of Q-space for the Fowey Rocks dataset from May 1, 2007. Colors are the same as the left panel, but the actual seabed type was not known for each point so all stacked echoes were plotted as dots.
114
Table 4.25: Subsets of features and their colors as shown in Figure 4.24, bottom panel, and Figure 4.25.
Feature set Features included in subset N features Color 1 1-166 166 Blue 2 1-30 30 Green 3 1-15 15 Red 4 17-23 7 Cyan 5 24-30 7 Magenta 6 41-46 5 Yellow 7 17-30 14 Black
Figure 4.26: Fowey Rocks dataset FFVs, FFV coefficient of variation, and loadings of the first principal component computed from the FFVs. A) Boxplot of features for all echoes of the May 1, 2007 Fowey rocks dataset (N=1822 echoes). Blue line is the median, grey boxes cover the middle 50% of the data, grey “whiskers” extend to the lesser of 1.5 times the range of the grey box, or the extremes of the dataset, and red crosses mark values outside the range of the whiskers. B) FFV coefficient of variation of features for all echoes of the May 1, 2007 Fowey rocks dataset. Note values of many of the features between 45-100 go off scale, which has been set the same as Figure 4.24 for comparison. C) Loadings of the first principal component computed with all 166 features (blue) and six subsets of the features (other colors; see Table 4.25 for colors).
115
The seven feature sets were then computed for the May 1, 2007 Fowey Rocks
dataset (Fig. 4.25 right; Fig. 4.26). In the real dataset, the classes were not as distinctly
defined as in the BORIS simulations. In fact there were no obvious class boundaries for
the real data using any of the feature subsets. Based on this experience it was not clear
that any of the subsets would help stabilize the reproducibility problem noted with the
Fowey repeat surveys. Nevertheless, one experiment was run using feature set three
(Table 4.25). Using just feature set three, the Q-space was computed for each of the daily
Fowey datasets and then clustered in IMPACT using ACE (15 iterations, 2-7 classes).
ACE results suggested that four classes were optimal for each of the six days. The OA,
Kappa, and AMI analysis were then performed on these daily datasets.
OA values of the datasets processed with only features 1-15 (feature set three in
Table 4.25) for all six days ranged from 81% to 88% (Table 4.26), which was an
improvement relative to the baseline 35% - 67% (Table 4.3). Kappa values of the
datasets processed with only features 1-15 (feature set three in Table 4.25) for all six days
ranged from 71% to 81% (Table 4.26), which was an improvement relative to the
baseline 19% - 55% (Table 4.3). AMI values of the datasets processed with only features
1-15 (feature set three in Table 4.25) for all six days ranged from 51% to 64% (Table
4.27), which was an improvement relative to the baseline 36% - 51% (Table 4.4).
Table 4.26: Overall accuracy (left) and Kappa (right) between daily classified datasets using the ACE clustering after computing Q space with only features 1-15 (feature set three in Table 4.25).
116
Table 4.27: Percent AMI for the datasets using ACE clustering after computing Q space with only features 1-15 (feature set three in Table 4.25).
4.6 Summary and Conclusions The reproducibility of the baseline processing was on the order of 50-65%
measured with overall accuracy and 50% or less measured with Kappa or average mutual
information (Tables 4.3, 4.4). None of the modifications to the baseline processing
perfected reproducibility, but three of the approaches resulted in improvements relative to
the baseline (Table 4.28). The three methods with the best improvement were 1) using
just two classes output from ACE, 2) retaining only echoes with low pitch, and 3) using a
subset of features.
Reducing the number of classes in the ACE-2 approach achieved a high OA
score, but the Kappa and AMI values for this method were not as high as for the low
pitch and reduced feature set methods. The relatively low Kappa and AMI values arise
because datasets with a small number of classes have a greater chance for random correct
matches.
Filtering by pitch resulted in the biggest improvements over the near-nadir
experiments (Section 4.5.2), which is consistent with the results of Section 4.4.3 in which
it was shown that echo, FFV, and Q-space reproducibility were sensitive only to pitch
grazing angle out of the survey parameters investigated. Nevertheless, this result needs to
be viewed with caution due to the low sample size for May 1 and 28 in this experiment
117
(Fig. 4.17). The observations that small day-to-day mounting biases were unavoidable
(Fig. 4.3) and that results improved when off-nadir echoes were filtered out suggest that
Greenstreet et al. (1997) may have been correct that different mounting angles of the
transducer contributed to the differences found in their repeat surveys.
Table 4.28: Summary of experiments and range of the majority of OA, Kappa, and AMI values in each. The values for the baseline and three alternate methods with the highest OA scores are shown in bold. Processing method OA Kappa AMI 4.4: Baseline ~50-65 20-60 40-50 4.5.1.1: Merge classes 4-6 ~60-75 30-60 35-55 4.5.1.2: ACE-2 80-95 60-80 35-50 4.5.2.1: Within 5 degrees of vertical (stacked by 5) ~40-60 25-50 30-75 4.5.2.2: Pitch within 2 degrees of vertical (stacked by 5) ~70-100 55-100 50-90 4.5.2.3: Within 5 degrees of vertical (unstacked) ~55-65 35-55 30-35 4.5.2.4: Pitch within 2 degrees of vertical (unstacked) ~50-65 30-50 25-40 4.5.2.5: Incidence angle less than 5 degrees (unstacked) ~60 40-50 30-40 4.5.3.2: Consistent PCA ~50-65 40-60 40-55 4.5.4: Features least subject to ping-to-ping variability 80-90 70-80 50-65
Arguably, the best improvement in reproducibility came from using a reduced set
of features that exhibited low sensitivity, relative to the other IMPACT features, to ping-
to-ping variability due to seabed roughness and volume heterogeneity. The reduced
feature set experiment had about the same OA as the ACE-2 experiment, but higher
Kappa and AMI. The reduced feature set experiment had a higher OA than the low pitch
experiment, and about the same Kappa and AMI, particularly considering only the days
with the most points in the low pitch experiment.
The results of this study suggest two ways to improve single-beam ASC in the
future. First, logging the pitch and roll of the transducer as data are collected will enable
filtering off-nadir echoes and thereby improve reproducibility. Second, searching for new
118
features that are either less sensitive to ping-to-ping variability or specifically exploit
ping-to-ping variability should also improve reproducibility. Some ideas for improved
features include:
• Focus on the early part of the echo, from the bottom pick to the first peak.
This is counter to the observation that the tail of the first echo contains the
most power to discriminate sediments (ICES 2007), but the tail of the echo is
also the most variable part due to incoherent stochastic scattering, particularly
over rocky seabeds.
• Incorporate intra-stack variability as features. The results of Section 4.4.3
suggest that stack-to-stack variability may also be present at the echo-to-echo
level within each stack. If true then this would be a way to recover the potent
power to discriminate sediments from the tail of the first echo (ICES 2007).
This study focused on the QTC system, but several other single-beam ASC
systems are commercially available (Michaels 2007). The methods described in this study
could be used to assess reproducibility using any of them. Furthermore, the two
suggestions on how to increase reproducibility using transducer attitude and ping-to-ping
variability should be applicable to other systems single-beam acoustic seabed
classification systems.
119
Chapter 5: Acoustic signatures of the seafloor: tools for predicting grouper habitat
5.1 Background Several species of groupers (family: Serranidae) are important components of
recreational and commercial fisheries. These fish also contribute to healthy coral reef
ecosystems and are commonly a focus of recreational diving and photography. As such,
preserving healthy populations of groupers is desirable for economic, ecological, and
aesthetic reasons.
The life history and behavior of groupers make them especially susceptible to
overexploitation (Coleman et al. 1999). Groupers are top predators in the coral reef
ecosystem, with long life spans and a low natural mortality rate. When predation by man
decreases their abundance, however, groupers are slow to recover because they do not
begin to reproduce until late ages (Polovina and Ralston 1987; Sadovy 1994). Many
species of grouper, such as goliath grouper (Epinephelus itajara), Nassau grouper (E.
striatus) and red grouper (E. morio), are unwary of divers and are easily caught in traps
or by angling. Furthermore, many groupers form predictable, seasonal, and site specific
aggregations, which are easy to eradicate once located by fishermen (Polovina and
Ralston 1987; Sadovy 1994; Coleman et al. 1999; Sadovy and Eklund 1999). For these
reasons, groupers are a family of fishes that are likely to benefit from marine protected
areas (MPAs; areas of no take).
For MPAs to be useful in grouper conservation, they must incorporate appropriate
habitat. Currently, however, essential grouper habitat is poorly defined. Like most reef
fishes, groupers prefer hard bottom (e.g. coral reef) to unconsolidated substrate (e.g.
seagrass or bare sediment). Beyond this, knowledge of grouper habitat is largely
120
anecdotal. The National Oceanic and Atmospheric Administration (NOAA) National
Marine Fisheries Service (NMFS) Southeast Fisheries Science Center (SEFSC) has been
monitoring grouper density near Carysfort Reef since 1994 (Eklund et al. 2000) and more
recently at other reefs of the Florida Keys. Through experience, the NOAA divers have
developed a qualitative “feel” for good grouper habitat, which often includes features
such as high relief and the presence of caves or crevices, especially on steeply sloping
surfaces.
Maps showing the distribution of potential grouper habitat are limited. In the
Florida Keys, for example, an aggregation of 70-100 black groupers (Mycteroperca
bonaci) was observed just 100 m outside the protected area at Carysfort Reef less than a
year after the preserve opened (Eklund et al. 2000). Discovery of the first known
aggregation of any grouper species in the Florida Keys (Eklund et al. 2000) just outside
the largest MPA in the Keys is ironic. Information on the distribution of fish habitat is
highly relevant to MPA design, yet often such critical information is unavailable. The
experience at Carysfort underscores the need for efficient methods of (1) seabed mapping
and (2) prioritizing limited dive time for fish census.
Diver-based grouper census surveys could potentially benefit from improved
methods of remotely sensed seabed classification. Optical mapping products, such as the
Benthic Habitats of the Florida Keys (FMRI, 1998), are useful in some applications;
however, much important grouper habitat, including the area of the large aggregation
observed by Eklund et al. (2000) outside the Carysfort MPA, is located in deeper water
where optical mapping techniques are not useful. Acoustic mapping systems are a
121
promising technology for mapping areas where the bottom can not be detected by optical
methods.
Acoustic methods have been successfully used to discriminate substrate classes in
many areas around the world (e.g. Hamilton et al. 1999; Morrison et al. 2001; Anderson
et al. 2002; Ellingsen et al. 2002; Freitas et al. 2003a). To date, however, applications of
this methodology in carbonate reefal environments are limited. The overall goal of the
present study was to evaluate the potential of a commercial acoustic mapping system,
QTC View System V (QTCV; Quester Tangent Corporation, Sidney, BC, Canada), to (a)
identify potential grouper habitat and (b) prioritize sites for diver surveys.
Specifically, this project addressed the question: Do areas of high grouper
abundance have characteristic acoustic signatures? Results demonstrate two effective
predictors of grouper presence or absence: (1) simple acoustic seabed classification,
which distinguishes hard bottom from sediment substrate and (2) a newly developed
index of acoustic variability. Incorporating these results in future surveys could assist in
devising sampling strategies for grouper census efforts or in assessing potential MPA
sites.
5.2 Methods The study focused on Carysfort Reef, Florida Keys (Fig. 5.1). An acoustic survey
was performed and the resulting data processed in two ways. First, clusters of
acoustically distinct echoes were segmented using commercially available software.
Second, a new index of acoustic variability was developed. Acoustic variability was
designed to measure seabed heterogeneity by quantifying the degree to which the echo at
a particular location is similar to other nearby echoes.
122
The acoustic survey was complemented with diver surveys, which collected
“ground truth” data on bottom type and grouper abundance. Correlations between
acoustic and diver surveys were conducted to test the value of using acoustic signatures
for identifying potential sites for grouper habitat and prioritizing sites for diver surveys.
Details of the methods are presented below.
Figure 5.1: Acoustic survey track lines superimposed on an IKONOS satellite image (copyright GeoEye.com) of Carysfort Reef and surroundings. The Carysfort lighthouse (star) and protected area (bold rectangular box) are also shown. The arrow in the inset shows the location of the IKONOS image in the wider context of South Florida (solid black), the Florida Keys National Marine Sanctuary (dashed line), and the area that has been mapped by FMRI (1998) from aerial photography (grey). The track lines extend from near the reef crest to deeper water where the bottom is no longer visible.
123
5.2.1 Acoustic Survey
5.2.1.1 Data Collection and Seabed Classification The acoustic survey at Carysfort Reef (Fig. 5.1) was conducted using a QTCV
acoustic mapping system. Acoustic data were recorded using a Suzuki 50 kHz echo
sounder (model 2025). A wide area augmentation system (WAAS) enabled global
positioning system (GPS), mounted with its antenna directly over the acoustic transducer,
provided vessel positioning. The survey, conducted on March 14, 28, and April 4, 2002,
consisted of transects spaced 100 m apart running perpendicular to the reef crest from an
inshore depth of 3 m to a maximum offshore depth of 42 m.
Data processing for seabed classification involved four steps (Fig. 5.2).
Processing used the software QTC IMPACT (version 3.4, QTC, Sidney, BC, 2004).
During the first step, the data acquisition phase, the signal generated by an echo sounder
is passed to a head amplifier that applies both time-varying gain, to compensate for beam
spreading and water depth, and auto gain control, to compensate for variable bottom
reflectance. Individual echoes are then digitized using a 5 MHz analog to digital card and
recorded by a computer.
In the second step, the data reduction phase, the raw bi-polar waveforms are
converted to echo “envelopes” (essentially echo amplitude only). The echo envelopes are
stacked (averaged) in groups of five to reduce ping-to-ping variability. The stacked
echoes are characterized by a number of algorithms that respond to features of the echo
shape. The ensemble of features is reduced using principal components analysis (PCA) to
the first three principal components. The end result is that each stacked echo is
represented by a single point in three-dimensions (“Q-space”; QTC, 2004). The shape of
the stacked echo determines the coordinates of this point.
124
Figure 5.2: Overview of acoustic processing. A) The full waveform for each point is acquired, georeferenced, and recorded. B) Each echo is converted to three coordinates in “Q-space” based on the shape of the echo “envelope” (echo amplitude). C) Q-space is partitioned into distinct groups defined by their means and covariances. D) Each point is assigned a class based on its location in Q-space.
In the third step, the clustering phase, the “cloud” of points in Q-space is
partitioned into clusters using a simulated annealing clustering procedure (Preston et al.
2004a). The statistical descriptions (mean, covariance) of these clusters comprise a
“catalog” (QTC, 2002).
Finally, in the classification stage, the catalog is used to assign a class to all points
in a dataset. The catalog can be applied to the original data used to create the catalog (via
clustering) or it can be applied to another data set acquired with the same hardware
configuration.
The four steps of acquisition, reduction, clustering, and classification are
fundamental to the IMPACT processing procedure. The overview above is similar to
previous descriptions of data processing using QTC View System IV and older versions
of the IMPACT software (e.g. Hamilton et al. 1999; Morrison et al. 2001; Anderson et al.
2002; Ellingsen et al. 2002; Freitas et al. 2003a; Preston et al. 2004a).
125
5.2.1.2 Acoustic Variability Index Standard QTC analysis, as described above, characterizes the acoustic response of
a position on the seafloor relative to all others in the survey area based solely on echo
shape. The geographic location of the echo is irrelevant in the clustering process; location
is used only to plot the classification results. An additional way to characterize acoustic
response at a point is to quantify the degree to which a particular echo is similar in shape
to its geographic neighbors (as opposed to its neighbors in Q-space). Such a measure,
designated acoustic variability, was developed as part of this study (Fig. 5.3). The
computation of acoustic variability, described below, complements the standard QTC
classification to more fully characterize the acoustic signature of any given location on
the seafloor.
Figure 5.3: Illustration of the computation of acoustic variability. A moving window sliding over the dataset in geographic space (left plot) selects neighboring points. Variability is the sum of the lengths of the principal axes of the cloud formed by the selected points in Q-space. If these points have low acoustic variability, they will cluster close together in Q-space (upper center plot). If they have high acoustic variability, they will form a scattered cloud in Q-space (lower center plot). The spatial pattern of variability is apparent when displayed using the geographic location of each echo (right plot).
126
Acoustic variability was computed point-by-point across the data set by
considering a small moving window applied around each echo in the survey (Fig. 5.3).
For each point, all the echoes within 40 m of that point were identified. In Q-space this
subset of the data produces a small cloud of points (typically between 10-20 points,
depending on vessel speed). Variability was defined as the sum of the standard deviations
along the three principal axes of this cloud of points and was computed by taking the
square root of the trace of the covariance matrix computed for each subset of data (Davis
1986). A window of data that includes echoes that are all very similar will have points
very close to one another in Q-space and will therefore have low acoustic variability.
Conversely, a window of data containing echoes that are all very dissimilar will have
points spread across Q-space and will have high acoustic variability.
5.2.2 Diver Survey
Twenty-two dives were conducted near Carysfort during August of 2002 and
October of 2003 to acquire “ground truth” for the acoustic measurements. The locations
of the dives were chosen based on the maps of seafloor classification and acoustic
variability. Since only a limited number of dives were possible, the sites were chosen to
ensure that multiple dives were placed in 1) homogenous areas within each acoustic class,
and 2) areas of high and low acoustic variability.
Diver surveys followed NOAA/SEFSC procedures for conducting fish census
(Bohnsack and Bannerot 1986) and benthic habitat assessment (Franklin et al. 2003), as
described by McClellan and Miller (2003). At every site, two divers each surveyed non-
overlapping, 7.5 m radius cylinders; results from the two divers were averaged to produce
a single set of values for each dive site. Diver collected data that were compared with the
acoustics were: 1) the number of groupers (Epinephelus and Mycteroperca spp.) observed
127
in a five minute interval and 2) estimated percent cover of three substrate classes
(sediment, hard bottom and rubble).
5.2.3 Comparison of Acoustic and Diver Surveys
Results from acoustic and diver surveys were compared to 1) assess the accuracy
of the acoustic classification and 2) correlate grouper abundance with acoustic
classification and variability. The general strategy in both cases was to compare a diver-
estimated parameter with the closest acoustically derived class or index of variability.
5.2.3.1 Acoustic Classification Accuracy Assessment Assessing the accuracy of acoustic classification involved two steps. First, diver-
estimated bottom cover was overlain on the acoustic classification map to visually
determine which acoustic classes corresponded with which bottom types. Second, the
accuracy of the classification was assessed using an error matrix.
The error matrix is a common method of quantifying the accuracy of a thematic
map by comparing “ground truth” for a sample of points on the map with the predictions
made by the map (Congalton and Green 1999). Ground truth is often acquired by visiting
sites and visually determining what is there (e.g. by divers). A matrix can then be
constructed with one column per ground truth class, one row per map class, and entries in
the appropriate row and column for each ground truth point visited. The sum of all the
elements in the matrix equals the total number of ground truth points, and the sum of the
elements in the matrix for which the ground truth class is the same as the map class is the
total number of “correct” points visited on the map. The overall accuracy is the latter
divided by the former. This technique was used here with one modification. Usually each
point visited is assigned one ground truth class and one map class. In this study, however,
each ground truth site (a single dive) was evaluated as a mixture of three classes
128
(sediment, hard bottom, and rubble) covering fractions of the seabed at the dive site, but
an echo had only one acoustic class. To accommodate mixed bottom types, the entry in
the error matrix for the closest echo to a particular dive site was divided among the
columns of the matrix in proportion to the diver-estimated bottom cover for that site.
5.2.3.2 Grouper Abundance vs. Acoustic Classification and Variability Acoustic classes and variability were compared with grouper abundance using
analysis of variance (ANOVA) and multiple comparison tests to determine the
significance of any correlation. Dive sites were grouped into treatments by the number of
groupers at the site, and two ANOVA procedures were run. The first tested the null
hypotheses that the mean percent of a given acoustic class was the same for sites with
different number of groupers. The second tested the null hypothesis that mean variability
was the same for sites with different numbers of groupers.
The MATLAB statistics toolbox (Version 4.0:R13; The MathWorks, Inc., Natick,
MA 2002) was used to perform the tests. First, the null hypothesis that the variables
being compared followed a normal distribution was evaluated using a Lilliefors test
("lillietest" command; see also Conover 1980). Based on the output of the Lilliefors test,
the parametric (“anova1”) or non-parametric (“kruskalwallis”) MATLAB
implementations of ANOVA were used to test the significance of differences between the
group means. Finally, if the null hypothesis that all group means were equal was rejected
by the ANOVA, the “multcompare” function (based on procedures from Hochberg and
Tamhane 1987) was used to determine which pairs of means were significantly different
from one another. A 95% confidence interval (p < 0.05) was used for all statistical tests.
129
5.3 Results
5.3.1 Acoustic Survey Clustering the acoustic survey data discriminated seven acoustic classes. The
three major classes comprised 94% of the echoes in the survey area (Fig. 5.4). The four
minor classes were dispersed widely across the study area. Due to limited dive time, the
sites chosen for ground truth focused on the three largest classes.
5.3.2 Diver Survey
Diver estimates of substrate at the twenty-two sites selected to ground truth the
three major acoustic classes were dominated by sand or hard bottom substrate (Fig. 5.4).
Seven sites had small amounts of rubble substrate, with only one site having > 10%
rubble (Fig. 5.4).
Grouper abundance at each of the dive sites varied from zero to six (Fig. 5.5). At
ten sites, no groupers were observed. Groupers were observed at 12 sites, with higher
numbers corresponding to decreasing frequency of sites. The maximum number of
groupers observed at any site was six (n=1).
5.3.3 Acoustic Classification Accuracy Assessment
Bottom types were assigned to acoustic classes based on visual observation of
diver survey results overlain on the acoustic classification map (Fig. 5.4). Visually,
Acoustic Class 1 corresponds with the dive sites dominated by hard bottom and Acoustic
Classes 2 and 3 both correspond with the dive sites dominated by sediment (Fig. 5.4).
The overall accuracy of the acoustic classification considering only hard bottom
and sand classes was 86% (Table 5.1), which is comparable to the accuracy of optical
sensors for mapping coarse bottom types (Mumby and Edwards 2002).
130
Figure 5.4: The three main acoustic classes, which comprise 94% of echoes in the survey, and diver-estimated substrate at 22 sites. Visual observation of this figure suggests that Acoustic Class 1 corresponds to hard bottom and Acoustic Classes 2 and 3 both correspond to sediment.
5.3.4 Grouper Abundance vs. Acoustic Classification and Variability
Visual inspection of the grouper abundance data (Fig. 5.5) suggests that sites with
high grouper abundance were associated with hard bottom and had higher acoustic
variability than sites with fewer groupers. The differences between group means were
not, however, statistically significant when the data were tested with ANOVA using
seven categories (one each for sites with number of groupers from zero to six); this
negative result may be due to the small number of sites in most categories. The ANOVA
analysis was repeated with sites grouped into only two categories based on the presence
131
(n=12) or absence (n=10) of groupers. Sites with groupers had both a significantly higher
percentage of hard bottom relative to sand (p = 0.006) and significantly higher acoustic
variability (p < 0.001) than sites without groupers (Fig. 5.6).
Table 5.1: Error matrix comparing acoustic classification with diver-based “ground truth.” Acoustic Class 1 was interpreted as a hardbottom class, and Acoustic Classes 2 and 3 were combined to form a single sediment substrate class. Fractional values are possible because the entry for each point was divided proportionally by the diver-estimated substrate at that site. The sum of all entries is 18, indicating that the closest echoes to 18 of the 22 dive sites were classified as Acoustic Classes 1, 2, or 3.
Acoustic Classes Diver-Estimated Substrate Overall Accuracy Hardbottom Sediment Rubble Class 1 (Hardbottom) 8.2 1.1 0.7 Class 2+3 (Sediment) 0.7 7.3 0.0
0.86
5.4 Discussion Results from this study demonstrate that acoustic signatures consisting of a simple
substrate classification and an index of local heterogeneity were distinct for dive sites
with and without groupers at Carysfort Reef. In general, sites where groupers were
present had hard substrate with high local heterogeneity, and sites without groupers had
sediment substrate with low local heterogeneity.
QTC systems have previously distinguished outcropping rock from sediment (e.g.
Anderson et al. 2002); most of these studies, however, have focused on siliciclastic
environments. Moreover, previous work with an older QTCIV system suggested that
rough terrain could adversely affect system accuracy (Hamilton et al. 1999). It is
therefore noteworthy that results from this study show that acoustics can be used to
distinguish hard bottom and sediment with high accuracy in a high relief, carbonate reef
environment.
132
Figure 5.5: Grouper (Epinephelus and Mycteroperca spp.) abundance at each of the dive sites relative to (A) acoustic classification and (B) acoustic variability. Groupers were found primarily in areas with hardbottom substrate and high acoustic variability. Note that the area of high variability is much smaller than the area of hardbottom. Transect A-A’ is plotted in Figure 5.7.
The reason that sediment in the Carysfort Reef area maps as two distinct acoustic
classes (Fig. 5.4) is uncertain, but is likely related to differences in physical properties,
such as sediment grain size. Clustering of sediment with different grain sizes as distinct
acoustic classes would be consistent with previous QTC-derived classification schemes
(Anderson et al. 2002; Ellingsen et al. 2002; Freitas et al. 2003a; Freitas et al. 2003b).
133
Figure 5.6: Acoustic classification and acoustic variability computed from the echoes closest to each dive site and grouped by the presence (number of sites = 12) and absence (number of sites = 10) of groupers (Epinephelus and Mycteroperca spp.). A) Bars indicate the percent of dive sites for which the closest echo classified as hardbottom (Acoustic Class 1) or Sediment (Acoustic Classes 2 or 3). B) Box plots of acoustic variability near dive sites. Horizontal lines mark the lower quartile, median, and upper quartile values. The lines extending from each end of the box show the range of all data. Note that grouper presence correlates with hardbottom and high acoustic variability.
The acoustic classification results at Carysfort Reef demonstrated that the location
of a point in Q-space was related to physical characteristics of the bottom. A set of points
that were spread out in Q-space was therefore more likely to represent different bottom
types than a set of points that was tightly clustered in Q-space. Areas where different
bottom types were located close together, such as patchy environments or along edges,
have high acoustic variability, and areas where the bottom did not change rapidly had low
acoustic variability (Fig. 5.7). Like the Berger-Parker index (Morrison et al. 2001),
acoustic variability highlights transitions between classes (edges) and heterogeneous
areas with mixed classes. Acoustic variability, however, is computed directly from the
reduced acoustic echo features (Q-space) as opposed to operating on classified data.
Operating directly on the Q-values may be advantageous because differences between
echoes are measured continuously, rather than in discrete classes, and because the results
are not dependent on the classification scheme used.
134
Figure 5.7: Depth, acoustic class, and acoustic variability along transect A-A’, shown in Figure 5.5. Acoustic variability within large areas of a single class ranges from high to low, but the highest variability is found at the transitions between classes. Patchy areas, such as the deep outcropping hard bottom can form extended regions of high acoustic variability.
Observations by NMFS divers that groupers are often found over “complex”
bottoms (caves, crevices, ledges) led to the idea of testing acoustic variability. It should
be noted, however, that topographic complexity as observed by divers is not the same as
acoustic variability as defined in this study. Topographic complexity occurs on the scale
of meters and might be thought of as a rough or steep bottom. Acoustic variability, on the
other hand, is measured on the scale of tens of meters and reflects the proximity of
acoustically distinct bottom types.
The observation that sites with groupers had higher acoustic variability than sites
without groupers does not mean that acoustic variability is a measure of essential grouper
habitat. Acoustic variability does not measure what a diver might perceive as important
variables for grouper habitat. Acoustic variability could, however, help prioritize diving
135
effort for grouper population surveys. Acoustic variability might also contribute to a
better understanding of grouper habitat. For example, it is not clear why aggregations are
so site-specific. From a diver’s point of view, the bottom at the site of an aggregation can
appear very similar to the bottom just a few hundred meters away. Measurements of
acoustic variability may help to interpret diver observations by providing context on a
larger spatial scale
The differences in acoustic signatures of sites with and without groupers (Fig. 5.6)
suggest that acoustic classification and acoustic variability are potentially useful tools for
stratifying diver sampling effort for grouper census. A simple map distinguishing hard
bottom from sediment, which can be easily produced with acoustics, is a substantial
improvement over a lack of any bottom type information in optically deep water. A map
of acoustic variability may further refine the location of potential grouper habitat, thereby
increasing the efficiency of divers to conduct fish census surveys.
5.5 Conclusions The results of this study showed:
1) A commercial acoustic seafloor classification system (QTC View V) was
successfully used to discriminate hard bottom from sediment in a carbonate reef
environment.
2) A simple map of hard bottom versus sediment was a useful first step in
discriminating potential grouper habitat.
3) An index of acoustic variability, which measures heterogeneity of bottom
types, complemented the simple bottom classification map to further target areas of
potential grouper habitat.
136
The results of this study indicte that, the acoustic signature of the seafloor, as
measured with acoustic classification and acoustic diversity, is a useful tool for
stratifying sampling effort for diver-based grouper census surveys. Both acoustic
classification and acoustic variability can be rapidly and inexpensively acquired when
needed by fisheries and park managers around the world because they are easily
measured with a single-beam echo sounder.
137
Chapter 6: Geomorphology of grouper and snapper spawning aggregation sites in the upper Florida Keys, USA
6.1 Background Fish spawning aggregations (FSAs) are a vital part of the life cycle of many reef
fishes. Globally, 120 reef fish species are known to reproduce in aggregations (SCRFA
2008); the resulting juveniles replenish reef fish populations, sustain fisheries, and
support livelihoods in coastal communities. The sustainable management of grouper,
snapper and other reef fish fisheries, from both a fisheries and ecosystem perspective is,
in large part, dependent on the protection and conservation of FSAs, as well as the
ecological processes (e.g., migration to FSA sites, larval dispersal, and population
connectivity) associated with spawning events.
In many cases, a lack of knowledge of the location of FSAs prohibits protection
and effective management of these sites. In 1997, for example, following six years of
public comment and draft management plans, the Florida (USA) Keys National Marine
Sanctuary (FKNMS) established a series of 23 MPAs along the Florida Keys coral reef
tract (Fig. 6.1). In January 1998, six months after the MPAs were established, divers from
the National Oceanic and Atmospheric Association’s (NOAA’s) Southeast Fisheries
Science Center (SEFSC) documented an aggregation of over 100 black groupers
(Mycteroperca bonaci) less than 100 m seaward of the Carysfort Reef MPA (Eklund et
al. 2000). The proximity of this aggregation to the MPA was particularly unfortunate
because even though fishing pressure is eliminated within no-take MPAs, fishing
intensity is often concentrated along no-take MPA borders (Murawski et al. 2005;
Kellner et al. 2007).
138
Locations of FSAs are typically documented from interviews with local fishers
(e.g. Lindeman et al. 2000; Heyman et al. 2005). With support from management and
stakeholders, FSAs located in this way may be protected through fishery closure or no-
take marine protected areas (MPAs; Heyman 2004; Burton et al. 2005; Whaylen et al.
2006). Given such protection, aggregations, even at previously intensively exploited sites,
can potentially reform and increase in size (Burton et al. 2005). Many Florida and
Caribbean protected areas, including the Carysfort MPA, were, however, not established
specifically to conserve spawning aggregations (Appeldoorn and Lindeman 2003;
Heyman 2004). In such cases, interviews with appropriate local fishers are less likely to
occur during the process of determining MPA boundaries. Additionally, given the rapid
speed with which FSAs can be decimated using modern fishing techniques (Sala et al.
2001; Burton et al. 2005), it is likely that many unprotected FSA sites are already fished
out (i.e., no longer active spawning sites) or soon will be. Knowledge of the location of
fished out sites is likely to be lost within a generation once they cease to be fished.
Finally, FSA sites may exist that are unknown to fishers. Thus, managers would benefit
from the development of alternative methods to identify the likely location of FSA sites.
The goal of the present research was to investigate the potential of using
geomorphological features to predict the locations of FSA sites. The specific question
addressed was whether FSA sites in the upper Florida Keys, USA, were associated with
particular geomorphological features (i.e. geomorphological or habitat “signatures”) on
the seabed that could be reliably mapped and characterized with an off-the-shelf single-
beam echo sounder.
139
6.2 Methods A former commercial fisherman from the Florida Keys provided geographical
coordinates of four sites that historically had supported spawning aggregations of one or
more of the following snapper and grouper species: mutton snapper (Lutjanus analis),
black grouper (Mycteroperca bonaci) and yellowtail snapper (Lutjanus chrysurus). These
historical FSAs were heavily fished until the 1980s and no longer support spawning
aggregations (Roberto Torres, personal communication January 2007). The seabed at
these sites, however, has likely not changed significantly so conclusions may still be
drawn about the FSA-seabed association even thought these FSAs are no longer active.
Together with the black grouper FSA at Carysfort documented by Eklund et al. (2000),
five FSA sites in the upper Florida Keys were studied, one of which was used by multiple
species (Fig. 6.1; Table 6.1).
Figure 6.1: Map showing the study area. Land mass of South Florida and the Florida Keys is shown in gray. Thin black lines mark the boundaries of the FKNMS and the no-take MPAs within the FKNMS. Letter codes locate reefs specifically mentioned in the text: Ocean Reef (OR), Carysfort Reef (CR), Watsons Reef (WR), Davis Reef (DR), Maryland Shoals (MS), Sand Key (SK), Western Dry Rocks (WDR), and Cosgrove Shoals (CS; off the map to the west). Extents of acoustic surveys at OR, CR, WR, and DR are shown in gray over their respective reefs. Inset shows North America in gray and the location of this map at the southern tip of Florida, USA.
140
Table 6.1: Survey areas and the spawning aggregations in each area. The Watson’s Reef survey area contained two FSA sites, one of which was used by two species. Thus there were six FSAs at five sites within four survey areas. Survey Area FSA sites for: Ocean Reef Mutton Snapper (Lutjanus analis) Carysfort Reef Black Grouper (Mycteroperca bonaci) Watson’s Reef 1) Black Grouper (Mycteroperca bonaci) &
Yellowtail Snapper (Lutjanus chrysurus) 2) Mutton Snapper (Lutjanus analis)
Davis Reef Yellowtail Snapper (Lutjanus chrysurus)
The methods for mapping and analyzing the five FSA sites comprised five steps.
First, single-beam echo sounder data were acquired over four survey areas encompassing
the five FSA sites, and the echoes were processed to classify the seabed into areas of
similar acoustic response. Second, the acoustic seabed classes generated in Step One
were identified as one of two general substrate types: hard bottom (including reef) and
sediment. Third, the accuracy of the hard bottom / sediment classification was assessed
with diver data at the Carysfort Reef site. Fourth, maps and graphics were generated from
the classified single-beam data to facilitate description of the seabed features near the
FSA sites. Finally, common features that were observed in multiple survey areas were
synthesized into a model linking FSA sites to geomorphological / habitat characteristics
at those sites. These steps are described in the following five sections.
6.2.1 Acquisition and acoustic classification
A commercial single-beam acoustic seabed classification (ASC) system produced
by the Quester Tangent Corporation (QTC; Sidney, BC, Canada) was used to map the
four survey areas. Data acquisition was performed using the QTCView Series V (QTCV)
software and hardware (version 2.1, Quester Tangent Corp., Sidney, BC, Canada, 2002).
Data processing was performed using QTC IMPACT software (version 3.4, Quester
141
Tangent Corp., Sidney, BC, Canada, 2004). The process of acquiring and processing
QTCV data was described in Chapters 2, and 5 and has been published elsewhere
(Preston et al. 2004a; Freitas et al. 2008).
Data Acquisition with QTCV: Small, open vessels (24 - 30 feet long) were used
to perform the acoustic surveys. Different boats were used on different dates, but the
same QTCV (Table 6.2) was used for all the surveys. The system was easily mounted on
different boats by using an aluminum pole clamped to the vessel gunwale to support the
transducer.
Table 6.2: Characteristics and settings of the QTCV system used in this study. Parameter Value Sounder model Suzuki 2025 Frequency 50 kHz Power 500 W Echo pulse length 0.3 ms Ping rate recorded 1.5 Hz (approx) Transducer model Suzuki TGN60-50B-12L Beam width (cross track) 42 deg Beam width (along track) 16 deg
The original Carysfort Reef survey was conducted on March 14, 28, and April 4,
2002 and consisted of 53 parallel transects, spaced 100 m apart, running from the reef
crest across the upper slope from nominal depths of 5 to 35 m (Chapter 5). On July 25,
2005 every fourth line of the 2002 survey was extended to an outer depth of 60 m to
capture data across the entire upper slope terrace as defined by Lidz et al. (2003). The 18
transects of the July 2005 survey had wider line spacing (400 m) than the 2002 survey but
covered a larger area, extending another 1.6 km to the south for a total distance along
shelf of 6.6 km.
142
The other survey areas (Table 6.1) were completed in August 2007 using 100 m
line spacing with transects running across the upper slope terrace (Lidz et al. 2003)
between nominal depths of 10 to 60 m. Davis Reef was surveyed on August 4, 6, and 10,
Watsons Reef on August 7 and 8, and Ocean Reef on August 9 and 12. Each of these
surveys consisted of 50 parallel transects for a total along shelf distance of 5 km. The
transects were situated to yield 2.5 km of survey data on either side in the along-shelf
direction of the aggregation sites.
Data Processing with QTC IMPACT: Values used for the tunable IMPACT
parameters (Table 6.3) are provided to assist other QTCV users who may wish to
replicate the methods used here. These parameters are defined in the IMPACT users
manual (QTC 2004), Preston (2004), and Preston et al. (2004a).
Table 6.3: Values used for tunable parameters in IMPACT software for processing QTCV echoes. Parameter Value Standard Echo Length 170 samples Echoes deeper than Critical Depth Yes Survey depth 60 m Stack size 5 echoes Auto cluster iterations 30 Auto cluster class range 2-12
6.2.2 Identification of acoustic classes
The acquisition and acoustic classification step ended with an unsupervised
classification, a process of clustering echoes with similar shapes into unlabeled groups
called “acoustic classes.” The next processing step, therefore, was to label the acoustic
classes with a physically meaningful interpretation. We reduced the number of acoustic
classes in each survey area to just two general substrate types: hard bottom, which
143
includes all rocky substrates, and sediment. Two approaches were used for this step.
First, the mean echoes for each acoustic class were calculated. Second, the acoustic
classes were overlaid on top of IKONOS satellite data.
Mean echoes for each class were computed by averaging the IMPACT-computed
echo envelopes within each class. IMPACT version 3.4 did not directly provide the
functions to do this, but QTC provided libraries to enable extraction of the echo
envelopes. Custom software was written to combine the echo envelope vector, from the
extra QTC libraries, with the echo class information, from the standard IMPACT
“.seabed” files, and then to average the echo envelopes for all echoes of each class. The
expectation was that echoes from rough hard bottom surfaces would have slower rise
times and longer decay times than echoes from smoother sediment surfaces (APL-UW
1994).
The National Oceanic and Atmospheric Administration (NOAA) purchased
IKONOS satellite imagery for the Florida Keys (NOAA CCMA 2008). Portions of this
dataset were used as a backdrop on top of which the classified echoes were displayed to
help interpret what type of seabed corresponded with which acoustic class. The blue
channel for each image was displayed in grayscale and interactively contrast enhanced
with a logarithmic stretch to provide detail in deep water without saturating shallow
portions of the surveyed area.
6.2.3 Diver-based assessment of classification at Carysfort Reef
Diver data were available at Carysfort Reef. Twenty-two stationary reef visual
census (RVC) dives (Bohnsack and Bannerot 1986) were conducted in August 2002 and
October 2003 to acquire “ground truth” data for the original portion of the Carysfort
survey (Chapter 5). Diver estimates of the percent of the seabed substrate consisting of
144
hard bottom, sediment, and rubble were compared with the closest acoustic class to
generate an error matrix as described in Chapters 3 and 5.
A single drift dive using tri-mix was conducted for a portion of the 2005 extended
Carysfort Reef survey along a transect at about 45 m depth. At this deeper depth, the
RVC method was not practical, so video was acquired to qualitatively document bottom
type.
6.2.4 Locations of FSAs relative to seabed features
Surfer software (version 8.08, Golden Software, Golden, CO, 2007) was used to
visualize the processed QTCV data. Bathymetric grids with 30 m cell size were generated
from the depth recorded for each echo using the minimum curvature algorithm (Smith
and Wessel 1990). The grids were used, first, to generate a three-dimensional (3-D) view
of each survey area, and, second, to compute seabed slope. The classified echoes were
draped over the 3-D views and locations of the FSAs were plotted on each figure.
Seabed features were visually interpreted from the 3-D and slope figures. Steep
slopes and linear hard bottom ridges were two features of particular interest based on
previous experience at Carysfort Reef. Eklund et al. (2000) noted that the Carysfort FSA
occurred along a steep portion of the forereef slope and speculated that such habitat may
be favorable for large fishes such as the black grouper. The highest abundance of
groupers found in Chapter 5 was at RVC sites located on a pair of linear rocky ridges.
Ridges could be interpreted from the 3-D views based simply on perspective and
shading, but they were further described as “exposed” or “buried” if the overlying
acoustic classification showed hard bottom or sediment, respectively. Some ridges were
“partially exposed” if the overlying classification was mixed with some hard bottom and
some sediment echoes.
145
Once the seabed features had been located in each survey area, the FSAs were
described relative to those features by reference to the 3-D and slope maps.
6.2.5 Geomorphologic “signatures” of FSAs
To determine if there were any common FSA-seabed relationships across multiple
sites, the final step in the analysis was to look for recurring patterns in the locations of
FSAs relative to nearby seabed features. These patterns were then generalized into a
model “signature” of geomorphology that may help predict the locations of upper Keys
FSAs.
6.3 Results
6.3.1 Acquisition and acoustic classification In total, 68,013 echoes were clustered across the four survey sites (Table 6.4). The
optimum number of clusters determined by IMPACT’s auto cluster algorithm ranged
from four, at Watson's Reef, to seven, at Ocean Reef. Cluster sizes were not even; each
survey area typically contained two large classes and several smaller ones (Table 6.4).
The largest cluster in each survey area contained approximately half of the total number
of echoes in the area.
Table 6.4: Clustering results for the four survey areas (Carysfort reef was performed in two parts). Values under “N” are the number of echoes in each acoustic class. Values under “% total” are the corresponding “N” value divided by the sum of the appropriate “N” column. Letters under “Label” identify each acoustic class as “H” for hard bottom and “S” for sediment.
146
6.3.2 Identification of acoustic classes
The mean class echoes and IKONOS images were used together to label the
unsupervised acoustic classes produced by the IMPACT clustering routine with an
interpretation of bottom type corresponding to each acoustic class. As expected, hard
bottom echoes were longer and had slower rise times than sediment echoes (Fig. 6.2).
Figure 6.2: Mean echoes for the six acoustic classes at Davis Reef (Table 6.4). The average hard bottom echo is shown as a dashed line. The average sediment echoes are shown as grey lines. All of the sediment echoes have similar rise and decay times that are faster than the hard bottom echo. This example is from Davis Reef, but results from other survey areas were similar.
The distribution of hard bottom and sediment acoustic classes agreed with their
delineation on IKONOS satellite imagery except in the shallowest portions of each
survey area (Fig. 6.3). At Davis Reef, for example, areas shallower than 13 m identified
as sediment on IKONOS imagery were frequently misclassified as hard bottom using
acoustics. The acoustic hard bottom / sediment delineation agrees with a visual
interpretation of the IKONOS imagery for water depths between 13 m and approximately
30 m, the limit at which the bottom can be seen at Davis Reef (Fig. 6.3).
147
Figure 6.3: Acoustic classification and satellite image for the Davis Reef survey area. Thick black lines are depth contours at 10 m intervals. The dashed yellow line is the 13 m depth contour. Greyscale picture is the blue channel of an IKONOS satellite image (copyright GeoEye.com) with a logarithmic contrast stretch to enhance deeper areas. As depth increases from the upper left towards the lower right the image generally gets darker due to water attenuation, but at any given depth hard bottom areas are darker than sediment areas. Below about 30 m the seabed is not visible in this image. Shallower than 13 m, almost all echoes are classified as hard bottom, including the sediment within the area “A”. Deeper than 13 m, there are no obvious errors in the acoustic classification based on visual comparison with the imagery. In particular, the sediment areas adjacent to misclassified shallow areas are correctly classified (“B”).
The pattern of excellent discrimination between hard bottom and sediment except
for the shallowest portions of the survey area was evident at all survey areas. The
minimum water depth for accurate acoustic classification varied slightly from survey area
to survey area. At Ocean Reef this misclassification depth was 13 m, at Carysfort Reef 10
m, at Watsons Reef 16 m, and at Davis Reef 13 m.
148
6.3.3 Diver-based assessment of classification at Carysfort Reef
Based on comparison with the 22 RVC dives, the overall accuracy of the hard
bottom / sediment acoustic classification was 86% (see Chapter 5). The difference
between the previously-discussed work and that described here is that in Chapter 5 only
the largest hard bottom class and the two largest sediment classes were used, discarding
the minor classes that made up only about 5% of the dataset and the four RVC dives that
were closest to those minor classes. Here, all the data have been collapsed into either
sediment or hard bottom and all available dives were used to compute overall accuracy.
The drift dive conducted on August 10, 2005 followed a steep slope along the 45
m isobath for over 600 m (Fig. 6.4). Video collected during the drift dive indicated that
the seabed along this slope was patchy hard bottom. Most of the slope was comprised of
outcropping rock, though portions were covered with patches of sediment. The acoustic
classes along this slope were mixed, indicating a partially exposed, or patchy, hard
bottom in agreement with the video.
6.3.4 Locations of FSAs relative to seabed features
At Carysfort Reef, four ridges were visible (Fig. 6.4). Three of the ridges (“O1”,
“O3”, and “O4” in Fig. 6.4) were traced for over 5 km along shelf based on their
bathymetric expression. The fourth ridge, “O2”, was recognizable for about 1.5 km. The
steepest slopes in the area were on the order of six degrees along two segments of the O1
ridge near 20 m depth and along two shorter segments in the trough landward of the O1
ridge (Fig. 6.4). The black grouper aggregation site at Carysfort Reef was located along
one of the steep-slope O1 ridge segments (Fig. 6.4). Exposed portions of the O2, O3, and
O4 ridges were found directly offshore from the black grouper aggregation site (Fig. 6.4).
149
Figure 6.4: Visualization of processed acoustic data for the Carysfort Reef survey area. In both top and bottom images view is from the southeast looking northwest and depth contours are shown at 10 m intervals. Top: An oblique 3-D shaded view of gridded bathymetry with classified track lines superimposed. The interpreted locations of the crests of ridges O1-O4 are traced with green lines. Bottom: An oblique 2-D view of slope computed from the gridded bathymetry. The grey, yellow, red colormap indicates the slope in degrees. Contours are of the depth to help provide spatial reference with respect to the top figure. The location of the aggregation site is marked on both images by the intersection of the vertical blue line with the image surface. Track of the deep drift dive is marked in purple between points “A” and “B”.
Four ridges also were visible at Watson’s Reef (Fig. 6.5). The O1 and O4 ridges
could be traced for the entire 5 km along-shelf length of the survey area, whereas the O2
and O3 ridges were recognized over only about half that distance (Fig. 6.5). The steepest
slopes in the area were on the order of eight degrees along a 2 km long portion of the O4
ridge. The steepest portion of the O1 ridge was in the north central portion of the survey
150
area with a slope on the order of four degrees. Both aggregation sites in the Watson’s
Reef survey area (Table 6.1) were offshore of the steepest portion of the O1 ridge. The
yellowtail snapper / black grouper aggregation site at Watson’s Reef was located at the
southern end of the exposed portion of the O3 ridge and inshore of a partially exposed
section of the O4 ridge. The mutton snapper aggregation site was located on the partially
exposed section of the O4 ridge, offshore of the exposed section of the O3 ridge (Fig.
6.5).
Figure 6.5: Visualization of processed acoustic data for the Watson’s Reef survey area. Image descriptions are the same as Figure 6.4.
Three ridges were visible at Davis Reef (Fig. 6.6). The inner, O1, ridge consisted
of fully exposed hard bottom for its entire 3 km length, exhibited over 15 m of vertical
relief, and had the steepest slope observed in all survey areas, about 10 degrees, along its
seaward face (Fig. 6.6). In contrast, the O2 and O4 ridges were subtle features buried by
151
sediment across the entire Davis survey area (Fig. 6.6). The yellowtail snapper FSA in
the Davis survey area was located along the steepest part of the O1 ridge, inshore of the
buried O2 and O4 ridges (Fig. 6.6).
Figure 6.6: Visualization of processed acoustic data for the Davis Reef survey area. Image descriptions are the same as Figure 6.4.
Four ridges were visible in the Ocean Reef survey area (Fig. 6.7), but their
expression here was the most complicated of all the survey areas because the upper slope
is so narrow at this site. The O1 ridge was exposed along the entire 5 km length of the
survey area (Fig. 6.7). The O2 ridge was visible in two partially exposed segments, each
about 0.8 km long (Fig. 6.7). The O3 ridge could be traced over 2.5 km, with the southern
2 km exposed and the northern 0.5 km only partially exposed hard bottom. The O4 ridge
was partially exposed and only distinct over a 0.8 km section in the southern portion of
152
the survey area, but past its northern extent it appeared to merge into the O3 ridge. Slopes
in this area were higher on average than the other survey areas. The steepest slopes of
about 8 degrees were found along a 0.4 km long section of the O1 ridge in the northern
portion of the survey area. The southern 3 km of the O1 ridge also had a steep slope of
about 6 degrees (Fig. 6.7). The mutton snapper FSA at Ocean Reef was located on the O3
ridge at the point where it transitioned from fully exposed to partially exposed hard
bottom. The FSA was located offshore of the portion of the O1 ridge with the 6 degree
slope.
Figure 6.7: Visualization of processed acoustic data for the Ocean Reef survey area. Image descriptions are the same as Figure 6.4.
6.3.5 Geomorphologic “signatures” of FSAs
The above description of FSA locations relative to seabed features demonstrates
that a series of margin-parallel ridges along the upper slope terrace was one factor in
common to these sites and that a high slope along the inner most ridge was a second
153
common factor. Based on these recurring patterns of FSA locations in the upper Keys, we
observed that, generally, the FSAs surveyed were found at upper Keys sites meeting three
conditions. FSAs were found at sites that were
1) on the face of, or offshore of, the steepest part of the O1 ridge observed within
a ~5 km window along the shelf margin;
2) within 100 m in the along shelf direction of an exposed O2 or O3 ridge; and
3) on the face of or inshore of exposed or partially exposed portions of the O4
ridge.
Together, these three conditions comprise a geomorphological signature for FSA
sites in the upper Keys. Three of the five surveyed FSA sites meet all three conditions,
and, individually, the three criteria were each met by four of the five FSA sites (Table
6.5).
6.4 Discussion The analysis above suggests that FSA sites in the upper Florida Keys share
common geomorphic features: they are associated with outlier reefs. The proposed
geomorphologic “signature” for upper keys FSA sites differs markedly from
geomorphology-FSA associations in other locations, such as the Bahamas and Belize,
where aggregation sites tend to form on promontories that protrude into deep water and
terminate with dramatic escarpments (e.g. Sadovy and Eklund 1999; Heyman et al. 2005;
Heyman and Kjerfve 2008). The FSAs in the upper Keys were not found on
promontories, which are generally lacking along the FL Keys reef tract. Thus, it is
important to consider whether the upper Keys sites mapped in this study are unique or
154
whether the outlier reef FSA model might be applicable to other areas generally lacking
promontories.
Table 6.5: Agreement between proposed FSA site model criteria (numbered 1-3 as in the text) and observations of the sites surveyed. “Y” means the site does meet the criterion, and “N” means the site does not meet the criterion. FSA Site (species) 1 2 3 Carysfort (black grouper) Y Y Y Watsons (Black grouper & yellowtail snapper) Y Y Y Davis (yellowtail snapper) Y N N Watsons (mutton snapper) Y Y Y Ocean (mutton snapper) N Y Y
The first step toward assessing the wider applicability of the upper Keys FSA
geomorphology signature is to understand the origin of the upper slope ridges that
coincide with the FSA sites. We interpret the ridges as outlier reefs described by Lidz et
al. (1991; 2003). Lidz (2006) used aerial photographs and seismic data to map margin-
parallel outlier reef tracts along more than 200 km of the Florida Keys upper slope. Lidz
(2006) found that four outlier reefs were readily mapped in the lower Keys, but that
usually only one or two were visible from the aerial photographs in the middle and upper
Keys. A simple explanation for this may be that the deeper outlier reefs, O2-4 in our
terminology, are simply so small in the upper Keys that their crests are too deep to be
seen from aerial images. Interpreting the ridges observed in our study areas as outlier
reefs suggests that there are indeed at least two reasons to expect other FSA sites may be
associated with outlier reefs.
The first reason to expect that geomorphologic “signatures” for upper keys FSA
sites are probably not anomalous is that Florida Keys outlier reefs are regional features
(Lidz et al. 2006), so other sites in the Florida Keys potentially could fit this model.
155
Lindeman et al. (2000), for example, listed eight potential snapper FSAs in the Key West
area based on interviews with commercial fishers. The eight FSAs occurred in four areas:
Western Dry Rocks, southeast of Sand Key, Maryland Shoals, and Cosgrove Shoal. Lidz
(2006) mapped outlier reefs in three of these areas, and two of them, Western Dry Rocks
and Sand Key, are located in the area with maximum development of all four outlier reefs
(Lidz et al. 2006).
The second reason to expect that geomorphologic “signatures” of upper Keys
outlier reef sites are probably not anomalous is that the outlier reefs are eustatic features.
Hence, Florida Keys outlier reefs are a specific instance of a more general class of
drowned shelf-edge reefs, of which there are many around the world. Macintyre (1972)
documented drowned shelf-edge reefs at seven islands in the eastern Caribbean. Beaman
et al. (2008) mapped drowned reefs along portions of the Great Barrier Reef and suggest
they may extend up to 900 km along the shelf-edge. Since the detailed local history of
drowned reef development will vary around the world, it is likely that the exact
conditions favorable for FSAs in the proposed geomorphological “signature” will also
vary from region to region. For example the Keys outliers are four parallel ridges,
whereas most of Macintyre’s (1972) examples appear to be single reef tracts and those of
Beaman et al. (2008) were two parallel ridges. The important point is that drowned reefs
in general could provide a geomorphologically suitable alternative to promontories for
FSA habitat. A significant finding of this study is that drowned reefs can be rapidly
mapped with commercial single-beam ASC systems with adequate resolution for
associating these features with FSA locations.
156
6.5 Conclusions The objective of this study was to determine if any common geomorphological
features linked FSA sites in the upper Keys. Results showed that drowned, margin-
parallel, rocky ridges, known locally as outlier reefs, were features found in proximity to
all the mapped FSA sites. In particular, three conditions related to the specific
configuration of these outlier reefs were observed to create a “signature” of the
geomorphological features likely to attract FSAs in the upper Keys. These results have
implications for future FSA management and research.
From a management perspective, results indicate the potential benefit of
employing acoustic mapping technologies for reconnaissance in deep water when
considering the placement of MPAs. Single-beam ASC systems can provide sufficient
information at low cost for this purpose. In the case of the Carysfort SPA, a one-day
survey offshore of the potential site would have revealed the spatial extent of three outlier
reefs in water too deep for optical mapping, thereby providing potential justification for
extending the protected area seaward. Although QTCV and similar ASC technologies
were not available in the early 1990s when the FKNMS SPAs were being planned, these
technologies could be used in assessing future protected areas, especially those for which
FSA conservation might not be the primary priority.
From a research perspective, results suggest that drowned shelf-edge reef habitat
should be considered as potentially viable for attracting FSAs. Recognizing that (1) our
sample size was limited to six aggregation sites; (2) variability in results could have been
due to inter-species variation in preference for FSA site characteristics; and (3) other
drowned shelf-edge reefs around the world will not necessarily have four parallel reef
tracts like the Florida Keys, the specific conditions of the proposed geomorphological
157
FSA “signature” must remain tentative. This suggests a logical next course of action is to
investigate other similar sites along the FL Keys to determine if the upper Keys FSA-
geomorphology signature is consistent across wider areas.
158
Chapter 7: Conclusions The purpose of this dissertation research was to assess the utility of single-beam
ASC for mapping coral reef environments. The rationale for the research is that areas of
the seabed that cannot be mapped with satellite or aerial imagery due to depth or turbidity
are both extensive and ecologically important. A logical approach for mapping these
areas would incorporate acoustic survey tools in a hierarchical manner, such that large
areas could be mapped at low resolution and low cost per unit area, thereby helping to
focus more detailed and expensive surveys of smaller areas. Single-beam acoustic seabed
classification (ASC) systems could potentially contribute to such an approach due to their
portability, low cost, small data volumes, and track record of producing useful sediment
maps. Both methodological and applied challenged need to be addressed in order to
effectively implement this vision, however.
The decision to focus on both ASC methodology and application was influenced
by the history of image-based satellite and aerial remote sensing of coral reef habitats.
The transition of satellite remote sensing technologies from research to application in the
field of coral reef science was marred, and perhaps delayed to some degree, by the
general overselling of capabilities to the user community (Mumby et al. 1997) and by the
focus of research effort on developing new algorithms to the exclusion of integrating
existing technology into practical applications (Green et al. 1996). Fortunately, the
balance between technology development and practical application appears to have
improved for aerial and satellite-based remotely sensed data (Andrefouet and Riegl
2004). Thus a balance was sought between realistic assessment of available ASC
technology and the application of an off-the-shelf ASC system to a practical problem in
coral reef conservation.
159
Two chapters each pertained to methodology and real world application. Chapters
3 and 4 addressed the general question “ what can be mapped with ASC in the coral reef
environment?” in a specific way by focusing on the standardization of methodology,
which had been identified by Anderson et al. (2008) as a crucial “burning issue” for
advancing the adoption of ASC technology by a wider user community. Chapters 5 and 6
addressed the general question “ what are some example applications of ASC in the coral
reef environment?” in a specific way by characterizing the habitat of grouper and snapper
species in the upper Florida Keys, USA.
The multisite experiment described in Chapter 3 is the first documentation that
single-beam acoustic seabed classification systems can map the same seabed classes in
multiple locations. The single-beam ASC discriminated rocky from sediment substrate at
four sites with 73% to 86% accuracy. These results were consistent with classes
identified in previous studies at single coral reef sites, suggesting that the ability to
distinguish hard bottom from sediment is a robust capability of single-beam ASC
systems. Thin layers of sediment over hard bottom were identified as a major source of
error in the two bank top sites mapped, pointing to one area requiring improvement.
The utility of a two-class, rock / sediment classification scheme, which appears
simplistic at first glance, was reviewed in Chapter 3. An objective and transferable
classification scheme such as the one used in this chapter has been identified as a priority,
not just for ASC research (Anderson et al. 2008), but also for satellite remote sensing as
well (Mumby and Harborne 1999). The simple rock / sediment classification scheme was
shown to provide a solid foundation for more elaborate schemes that incorporate
coincident bathymetry, such as the one proposed by Franklin et al. (2003). In fact, the
160
habitat mapping approach incorporating both bathymetry and hard bottom / substrate
classes that was suggested at the end of Chapter 3 has already begun to be used for
mapping the Navassa Island National Wildlife Refuge (Miller et al. 2008).
The repeat transect experiment in Chapter 4 showed that the reproducibility of
QTCV classes in a coral reef study area was only on the order of 50% to 65%, but that
reproducibility could be improved with fewer classes, eliminating off-nadir echoes, and
using a subset of features for classification. Although a few aspects of reproducibility had
been investigated before (von Szalay and McConnaughey 2002; Wilding et al. 2003), this
experiment extended previous work by identifying that grazing angle was the survey
factor that most affected reproducibility and suggesting practical steps that could increase
reproducibility. In the short-term, surveys with pitch and roll sensors should increase
reproducibility. In the long term, identifying echo features that are more robust to ping-
to-ping variability, or perhaps exploiting ping-to-ping variability, should also increase
reproducibility.
The overall objective of Chapters 5 and 6 was to associate acoustic signatures of
the seabed with grouper and snapper habitat, as measured by diver census and the
locations of known historical spawning areas. This was a pragmatic objective because if
an acoustic signature that correlated with high abundance or spawning habitat could be
identified, then single-beam ASC would be useful for mapping potential areas for
conservation. Chapters 5 and 6 addressed two research priorities identified by Eklund et
al. (2000). Chapter 5 concerned the need for seabed habitat maps for the Florida Keys
that extended deeper than 20 m to prioritize reef fish census effort. Chapter 6 considered
the question of whether there were particular acoustic signatures associated with grouper
161
and snapper fish spawning aggregations (FSA). Results from Chapter 3 were directly
applied in Chapters 5 and 6. All of the areas surveyed in Chapters 5 and 6 were mapped
using a rock / sediment classification scheme.
The initial project at Carysfort reef (Chapter 5) provided two pieces of evidence
that single-beam ASC could be useful for stratifying unmapped deep areas for grouper
census effort. First, the simple hard bottom / sediment classification scheme was shown
to segment the survey area into regions with high and low abundance of groupers.
Second, an index of acoustic variability was developed that could also segment the
survey area into regions with high and low abundance of groupers. Acoustic variability
improved on the hard bottom / sediment classification because it identified a smaller area
that was likely to have high abundance. Therefore acoustic variability may prove even
more efficient than rock / sediment as a habitat map for grouper census.
The fish spawning aggregation study (Chapter 6) built on the observation from
Chapter 5 that high grouper abundance was found in deep, shelf edge-parallel ridges
offshore of Carysfort Reef. After surveying four other reefs in the upper Florida Keys
with known historical FSAs, a consistent pattern was identified suggesting that drowned
shelf edge reefs may be important habitat controlling, or at least coincident with,
spawning aggregations. Drowned shelf edge reef features are found worldwide, so the
pattern observed in the Keys may have wider application for FSA conservation.
Of course, neither the needed technological advances nor the potential
applications of single-beam ASC to coral reefs were exhausted by this work (Green et al.
2000; Anderson et al. 2008). Nevertheless, the goal achieved has been to show what is
realistic and possible in coral reef environments with today’s single-beam ASC
162
technology and to provide suggestions for the future. Indeed there are two obvious
immediate next steps to extend this work. On the methodology side, pitch and roll should
be routinely collected with all single-beam ASC surveys, and work should proceed on
improved features. On the application side, sites in the lower Florida Keys with known
historical FSAs should be surveyed to see if the relationships between the FSA sites and
outlier reefs are the same as had been identified in Chapter 6 for the upper Keys. In the
longer run, other known drowned, shelf-edge reefs in the Caribbean should be surveyed
for the identification of potential FSA sites.
163
References
Agassiz, A. (1885). The Tortugas and Florida reefs. Memoirs of the American Academy of Arts and Sciences 11: 107-134.
Anderson, J. T., R. S. Gregory and W. T. Collins (2002). Acoustic classification of marine habitats in coastal Newfoundland. ICES Journal of Marine Science 59: 156-167.
Anderson, J. T., D. V. Holliday, R. Kloser, D. G. Reid and Y. Simard (2008). Acoustic seabed classification: current practice and future directions. ICES Journal of Marine Science 65: 1004-1011.
Andrefouet, S., P. Kramer, D. Torres-Pulliza, K. E. Joyce, E. J. Hochberg, R. Garza-Perez, P. J. Mumby, B. Riegl, H. Yamano, W. H. White, M. Zubia, J. C. Brock, S. R. Phinn, A. Naseer, B. G. Hatcher and F. E. Muller-Karger (2003). Multi-site evaluation of IKONOS data for classification of tropical coral reef environments. Remote Sensing of Environment 88: 128-143.
Andrefouet, S. and B. Riegl (2004). Remote sensing: a key tool for interdisciplinary assessment of coral reef processes. Coral Reefs 23(1): 1-4.
APL-UW (1994). APL-UW High-Frequency Ocean Environmental Acoustic Models Handbook, Technical Report APL-UW TR9407 AEAS 9501. Applied Physics Laboratory, University of Washington, Seattle, Washington, 193 pp.
Appeldoorn, R. S. and K. C. Lindeman (2003). A Caribbean-wide survey of marine reserves: Spatial coverage and attributes of effectiveness. Gulf and Caribbean Research 14(2): 139-154.
Bax, N., R. Kloser, A. Williams, K. Gowlett-Holmes and T. Ryan (1999). Seafloor habitat definition for spatial managment in fisheries: a case study on the continental shelf of southeast Australia. Oceanologica Acta 22(6): 705-720.
Beaman, R. J., J. M. Webster and R. A. J. Wust (2008). New evidence for drowned shelf edge reefs in the Great Barrier Reef, Australia. Marine Geology 247: 17-34.
Bergem, O., E. Pouliquen, G. Canepa and N. G. Pace (1999). Time-evolution modeling of seafloor scatter. II. Numerical and experimental evaluation. Journal of the Acoustical Society of America 105(6): 3142-3150.
Biffard, B., S. Bloomer and R. Chapman (2005). Single beam seabed classification: Direct methods of classification and the problem of slope. Proceedings of Boundary Influences in High Frequency Shallow Water Acoustics, N. G. Pace and P. Blondel, Eds., University of Bath, UK, 5-9 September 2005: pp. 227-232.
164
Bloomer, S. F., B. R. Biffard, N. R. Chapman and J. M. Preston (2007). QTC DEEP - A ROV-mounted single beam acoustic seabed classification system for high-resolution mapping. Proceedings of MTS/IEEE Oceans 2007, Vancouver, BC, Canada, 1-4 October 2007.
Bohnsack, J. A. and S. P. Bannerot (1986). A stationary visual census technique for quantitatively assessing community structure of coral reef fishes, NOAA Tech. Rep., NMFS 41. National Oceanographic and Atmospheric Administration / National Marine Fisheries Service, 15 pp.
Bornhold, B. D., W. T. Collins and L. Yamanaka (1999). Comparison of seabed characterization using sidescan sonar and acoustic classification techniques. Proceedings of the Canadian Coastal Conference, Victoria, B.C., Canada.
Brandes, H. G., A. J. Silva and D. J. Walter (2002). Geo-acoustic characterization of calcareous seabed in the Florida Keys. Marine Geology 182(1-2): 77-102.
Brown, C. J., A. Mitchell, D. S. Limpenny, M. R. Robertson, M. Service and N. Golding (2005). Mapping seabed habitats in the Firth of Lorn off the west coast of Scotland: evaluation and comparison of habitat maps produced using the acoustic ground-discrimination system, RoxAnn, and sidescan sonar. ICES Journal of Marine Science 62(4): 790-802.
Burton, M. L., K. J. Brennan, R. C. Munoz and R. O. Parker (2005). Preliminary evidence of increased spawning aggregations of mutton snapper (Lutjanus analis) at Riley's Hump two years after establishment of the Tortugas South Ecological Reserve. Fishery Bulletin 103(2): 404-410.
Capolsini, P., S. Andrefouet, C. Rion and C. Payri (2003). A comparison of Landsat ETM+, SPOT HRV, Ikonos, ASTER, and airborne MASTER data for coral reef habitat mapping in South Pacific islands. Canadian Journal of Remote Sensing 29(2): 187-200.
Castilla, G., K. Larkin, J. Linke and G. Hay The impact of thematic resolution on the patch-mosaic model of natural landscapes. Landscape Ecology.
Chen, Y. and D. A. Jackson (1995). Robust estimation of mean and variance in fisheries. Transactions of the American Fisheries Society 124(3): 401-412.
Chen, Y., D. A. Jackson and J. E. Paloheimo (1994). Robust regression approach to analyzing fisheries data. Canadian Journal of Fisheries and Aquatic Sciences 51(6): 1420-1429.
Chivers, R. C., N. Emerson and D. Burns (1990). New acoustic processing for underway surveying. Hydrographic Journal 42: 8-17.
165
Coleman, F. C., C. C. Koenig, A.-M. Eklund and C. B. Grimes (1999). Management and conservation of temperate reef fish in the southeastern United States. in Life in the Slow Lane - Ecology and Conservation of Long-Lived Marine Animals. J. A. Musick, Ed. Bethesda, Maryland, American Fisheries Society: pp. 233-242.
Collins, W. T. and R. A. McConnaughey (1998). Acoustic classification of the sea floor to address essential fish habitat and marine protected area requirements. Proceedings of the Canadian Hydrographic Conference, March 1998, Victoria, Canada.
Congalton, R. G. and K. Green (1999). Assessing the accuracy of remotely sensed data: Principles and practices, Boca Raton, Lewis Publishers, 137 pp.
Conover, W. J. (1980). Practical Nonparametric Statistics, New York, Wiley, 493 pp.
Darwin, C. (1842). The structure and distribution of coral reefs. Being the first part of the geology of the voyage of the Beagle, under the command of Capt. Fitzroy, R.N. during the years 1832 to 1836., London, Smith, Elder and Co., 214 pp.
Davis, J. C. (1986). Statistics and data analysis in geology. Second Edition Edition, New York, John Wiley & Sons, 646 pp.
Eklund, A.-M., D. B. McClellan and D. E. Harper (2000). Black Grouper aggregations in relation to protected areas within the Florida Keys National Marine Sanctuary. Bulletin of Marine Science 66(3): 721-728.
Ellingsen, K. E., J. S. Gray and E. Bjornbom (2002). Acoustic classification of seabed habitats using the QTC VIEW system. ICES Journal of Marine Science 59: 825-835.
Filzmoser, P. (1999). Robust principal component and factor analysis in the geostatistical treatment of environmental data. Environmetrics 10(4): 363-375.
Finn, J. T. (1993). Use of the average mutual information index in evaluating classification error and consistency. International Journal of Geographical Information Systems 7(4): 349-366.
Florida Marine Research Institute (FMRI) (1998). Benthic habitats of the Florida Keys, FMRI Technical Report TR-4. Florida Marine Research Institute / Florida Department of Environmental Protection and the National Oceanic and Atmospheric Administration, St. Petersburg, FL, 53 pp.
Foody, G. M. (2002). Status of land cover classification accuracy assessment. Remote Sensing of Environment 80(1): 185-201.
Foody, G. M. (2004). Thematic map comparison: Evaluating the statistical significance of differences in classification accuracy. Photogrammetric Engineering and Remote Sensing 70(5): 627-633.
166
Foody, G. M. (2006). What is the difference between two maps? A remote senser's view. Journal of Geographical Systems 8(2): 119-130.
Foster-Smith, R. L., C. J. Brown, W. J. Meadows, W. H. White and D. S. Limpenny (2004). Mapping seabed biotopes at two spatial scales in the eastern English Channel. Part 2. Comparison of two acoustic ground discrimination systems. Journal of the Marine Biological Association of the United Kingdom 84(3): 489-500.
Foster-Smith, R. L. and I. S. Sotheran (2003). Mapping marine benthic biotopes using acoustic ground discrimination systems. International Journal of Remote Sensing 24(13): 2761-2784.
Franklin, E. C., J. S. Ault, S. G. Smith, J. Luo, G. A. Meester, G. A. Diaz, M. Chiappone, D. W. Swanson, S. L. Miller and J. A. Bohnsack (2003). Benthic habitat mapping in the Tortugas region, Florida. Marine Geodesy 26: 19-34.
Freeman, S., M. Bergmann, H. Hinz, M. J. Kaiser and J. Bennell (2002). Acoustic seabed classification: identifying fish and macro-epifaunal habitats. International Council for the Exploration of the Seas Annual Science Conference: Session CM2002/K:08, Copenhagen, Denmark. 28 pp.
Freeman, S. M. and S. I. Rogers (2003). A new analytical approach to the characterisation of macro-epibenthic habitats: Linking species to the environment. Estuarine, Coastal and Shelf Science 56: 749-764.
Freitas, R., A. M. Rodrigues, E. Morris, J. L. Perez-Llorens and V. Quintino (2008). Single-beam acoustic ground discrimination of shallow water habitats: 50 kHz or 200 kHz frequency survey? Estuarine, Coastal and Shelf Science 78(4): 613-622.
Freitas, R., A. M. Rodrigues and V. Quintino (2003a). Benthic biotopes remote sensing using acoustics. Journal of Experimental Marine Biology and Ecology 285-286: 339-353.
Freitas, R., S. Silva, V. Quintino, A. M. Rodrigues, K. Rhynas and W. T. Collins (2003b). Acoustic seabed classification of marine habitats: studies in the western coastal-shelf area of Portugal. ICES Journal of Marine Science 60: 599-608.
Galloway, J. L. and W. T. Collins (1998). Dual frequency acoustic classification of seafloor habitat using the QTC VIEW. Proceedings of IEEE Oceans 98, Nice, France, September 28 - October 1.
Gleason, A. C. R., A.-M. Eklund, R. P. Reid and V. Koch (2006). Acoustic seabed classification, acoustic variability, and grouper abundance in a forereef environment. NOAA Professional Papers NMFS 5: 38-47.
Gleason, A. C. R., R. P. Reid and G. T. Kellison (2009). Single-beam acoustic remote sensing for coral reef mapping. Proceedings of the 11th International Coral Reef Symposium, Ft. Lauderdale, Florida, 7-11 July 2008.
167
Gonzalez, R. and G. P. Eberli (1997). Sediment transport and bedforms in a carbonate tidal inlet; Lee Stocking Island, Exumas, Bahamas. Sedimentology 44(6): 1015-1030.
Green, E. P., P. J. Mumby, A. J. Edwards and C. D. Clark (1996). A review of remote sensing for the assessment and management of tropical coastal resources. Coastal Management 24: 1-40.
Green, E. P., P. J. Mumby, A. J. Edwards and C. D. Clark (2000). Remote sensing handbook for tropical coastal management. A. J. Edwards, Ed., Coastal Management Sourcebooks 3. UNESCO, Paris, 316 pp.
Greenstreet, S. P. R., I. D. Tuck, G. N. Grewar, E. Armstrong, D. G. Reid and P. J. Wright (1997). An assessment of the acoustic survey technique, RoxAnn, as a means of mapping seabed habitat. ICES Journal of Marine Science 54: 939-959.
Hamilton, L. J., P. J. Mulhearn and R. Poeckert (1999). Comparison of RoxAnn and QTC-View acoustic bottom classification system performance for the Cairns area, Great Barrier Reef, Australia. Continental Shelf Research 19: 1577-1597.
Hetzinger, S., J. Halfar, B. Riegl and L. Godinez-Orta (2006). Sedimentology and acoustic mapping of modern rhodolith facies on a non-tropical carbonate shelf (Gulf of California, Mexico). Journal of Sedimentary Research 76(3-4): 670-682.
Hewitt, J. E., S. E. Thrush, P. Legendre, G. A. Funnell, J. Ellis and M. Morrison (2004). Mapping of marine soft-sediment communities: Integrated sampling for ecological interpretation. Ecological Applications 14(4): 1203-1216.
Heyman, W. D. (2004). Conservation of multi-species spawning aggregations. Proceedings of the Fifty-Fifth Annual Gulf and Caribbean Fisheries Institute: pp. 521-529.
Heyman, W. D. and B. Kjerfve (2008). Characterization of Transient Multi-Species Reef Fish Spawning Aggregations at Gladden Spit, Belize. Bulletin of Marine Science 83(3): 531-551.
Heyman, W. D., B. Kjerfve, R. T. Graham, K. L. Rhodes and L. Garbutt (2005). Spawning aggregations of Lutjanus cyanopterus (Cuvier) on the Belize Barrier Reef over a 6 year period. Journal of Fish Biology 67(1): 83-101.
Hickerson, E. L. and G. P. Schmahl (2005). Flower Garden National Marine Sanctuary: Introduction. Gulf of Mexico Science 23(1): 2-4.
Hochberg, Y. and A. C. Tamhane (1987). Multiple comparison procedures, Wiley, 450 pp.
Hutin, E., Y. Simard and P. Archambault (2005). Acoustic detection of a scallop bed from a single-beam echosounder in the St. Lawrence. ICES Journal of Marine Science 62: 966-983.
168
ICES (2007). Acoustic seabed classification of marine physical and biological landscapes, ICES Cooperative Research Report No. 286, 183 pp.
Jackson, D. A. and Y. Chen (2004). Robust principal component analysis and outlier detection with ecological data. Environmetrics 15(2): 129-139.
Jarrett, B. D., A. C. Hine, R. B. Halley, D. F. Naar, S. D. Locker, A. C. Neumann, D. Twichell, C. Hu, B. T. Donahue, W. C. Jaap, D. Palandro and K. Ciembronowicz (2005). Strange bedfellows - a deep-water hermatypic coral reef superimposed on a drowned barrier island; southern Pulley Ridge, SW Florida platform margin. Marine Geology 214(4): 295-307.
Kellner, J. B., I. Tetreault, S. D. Gaines and R. M. Nisbet (2007). Fishing the line near marine reserves in single and multispecies fisheries. Ecological Applications 17(4): 1039-1054.
Kenny, A. J., I. Cato, M. Desprez, G. Fader, R. T. E. Schuttenhelm and J. Side (2003). An overview of seabed-mapping technologies in the context of marine habitat classification. ICES Journal of Marine Science 60: 411-418.
Kosinski, A. S. (1998). A procedure for the detection of multivariate outliers. Computational Statistics & Data Analysis 29(2): 145-161.
Legendre, P. and L. Legendre (1998). Numerical ecology. Second Edition, New York, Elsevier, 853 pp.
Lidz, B. H. (2006). Pleistocene corals of the Florida Keys: Architects of imposing reefs - Why? Journal of Coastal Research 22(4): 750-759.
Lidz, B. H., A. C. Hine, E. A. Shinn and J. L. Kindinger (1991). Multiple outlier-reef tracts along the south Florida bank margin: Outlier reefs, a new windward-margin model. Geology 19: 115-118.
Lidz, B. H., C. D. Reich, R. L. Peterson and E. A. Shinn (2006). New maps, new information: Coral reefs of the Florida Keys. Journal of Coastal Research 22(2): 260-282.
Lidz, B. H., C. D. Reich and E. A. Shinn (2003). Regional Quaternary submarine geomorphology in the Florida Keys. Geological Society of America Bulletin 115(7): 845-866.
Lindeman, K. C., R. Pugliese, G. T. Waugh and J. S. Ault (2000). Developmental patterns within a multispecies reef fishery: Management applications for essential fish habitats and protected areas. Bulletin of Marine Science 66(3): 929-956.
169
Locantore, N., J. Marron, D. Simpson, N. Tripoli, J. Zhang, K. Cohen, G. Boente, R. Fraiman, B. Brumback, C. Croux, J. Fan, A. Kneip, J. Marden, D. Peña, J. Prieto, J. Ramsay, M. Valderrama, A. Aguilera, N. Locantore, J. Marron, D. Simpson, N. Tripoli, J. Zhang and K. Cohen (1999). Robust principal component analysis for functional data. TEST 8(1): 1-73.
Louchard, E. M., R. P. Reid, F. C. Stephens, C. O. Davis, R. A. Leathers and T. V. Downes (2003). Optical remote sensing of benthic habitats and bathymetry in coastal environments at Lee Stocking Island, Bahamas: A comparative spectral classification approach. Limnology and Oceanography 48(1, part 2): 511-521.
Luckhurst, B. E. and K. Luckhurst (1978). Analysis of the influence of substrate variables on coral reef fish communities. Marine Biology 49: 317-323.
Macintyre, I. G. (1972). Submerged reefs of eastern Caribbean. Bulletin of the American Association of Petroleum Geologists 56(4): 720-738.
Maeder, J., S. Narumalani, D. C. Rundquist, R. L. Perk, J. Schalles, K. Hutchins and J. Keck (2002). Classifying and mapping general coral-reef structure using Ikonos data. Photogrammetric Engineering and Remote Sensing 68(12): 1297-1305.
Magorrian, B. H., M. Service and W. Clarke (1995). An acoustic bottom classification survey of Strangford-Lough, Northern-Ireland. Journal of the Marine Biological Association of the United Kingdom 75(4): 987-992.
Marden, J. I. (1999). Some robust estimates of principal components. Statistics & Probability Letters 43(4): 349-359.
McClellan, D. B. and G. M. Miller (2003). Chapter 4: Reef fish abundance, biomass, species composition and habitat characterization of Navassa Island. in Status of Reef Resources of Navassa Island: Nov 2002; NOAA Tech. Memo. NMFS-SEFSC-501. M. W. Miller, Ed. Miami, FL, U.S. Department of Commerce/NOAA/NMFS/SEFSC: pp. 24-42.
Michaels, W. L. (2007). Review of acoustic seabed classification systems. in ICES Cooperative Research Report No. 286: Acoustic seabed classification of marine physical and biological landscapes. J. T. Anderson, Ed.: pp. 94-115.
Miller, M. W., R. B. Halley and A. C. R. Gleason (2008). Reef geology and biology of Navassa Island. in Coral Reefs of the USA. B. Riegl and R. E. Dodge, Eds., Springer: pp. 407-434.
Miller, S. L., M. Chiappone, D. W. Swanson, J. S. Ault, S. G. Smith, G. A. Meester, J. Luo, E. C. Franklin, J. A. Bohnsack, D. E. Harper and D. B. McClellan (2001). An extensive deep reef terrace on the Tortugas Bank, Florida Keys National Marine Sanctuary. Coral Reefs 20: 299-300.
170
Mobley, C. D., L. K. Sundman, C. O. Davis, T. V. Downes, R. A. Leathers, M. Montes, J. H. Bowles, W. P. Bissett, D. D. R. Kohler, R. P. Reid, E. M. Louchard and A. C. R. Gleason (2004). A spectrum-matching and look-up-table approach to interpretation of hyperspectral remote-sensing data. Applied Optics 44(17): 3576-3592.
Morrison, M. A., S. F. Thrush and R. Budd (2001). Detection of acoustic class boundaries in soft sediment systems using the seafloor acoustic discrimination system QTC VIEW. Journal of Sea Research 46: 233-243.
Moyer, R. P., B. Riegl, K. Banks and R. E. Dodge (2005). Assessing the accuracy of acoustic seabed classification for mapping coral reef environments in South Florida (Broward County, USA). Revista De Biologia Tropical 53 Suppl. 1: 175-184.
Mumby, P. J. and A. J. Edwards (2002). Mapping marine environments with IKONOS imagery: enhanced spatial resolution can deliver greater thematic accuracy. Remote Sensing of Environment 82: 248-257.
Mumby, P. J., E. P. Green, A. J. Edwards and C. D. Clark (1997). Coral reef habitat-mapping: How much detail can remote sensing provide? Marine Biology 130: 193-202.
Mumby, P. J. and A. R. Harborne (1999). Development of a systematic classification scheme of marine habitats to facilitate regional management and mapping of Caribbean coral reefs. Biological Conservation 88(2): 155-163.
Mumby, P. J., W. Skirving, A. E. Strong, J. T. Hardy, E. F. LeDrew, E. J. Hochberg, R. P. Stumpf and L. T. David (2004). Remote sensing of coral reefs and their physical environment. Marine Pollution Bulletin 48(3-4): 219-228.
Murawski, S. A., S. E. Wigley, M. J. Fogarty, P. J. Rago and D. G. Mountain (2005). Effort distribution and catch patterns adjacent to temperate MPAs. ICES Journal of Marine Science 62(6): 1150-1167.
Murphy, L., T. Leary and A. Williamson (1995). Standardizing seabed classification techniques. Sea Technology 36(7): 15-19.
National Oceanic and Atmospheric Administration (NOAA) Center for Coastal Monitoring and Assessment (CCMA). (2008). "Benthic habitat mapping of Florida coral reef ecosystems." Retrieved 1 July, 2008, from http://ccma.nos.noaa.gov/ecosystems/coralreef/fl_mapping.html.
Pinn, E. H. and M. R. Robertson (1998). The effect of bioturbation on Roxann (R) a remote acoustic seabed discrimination system. Journal of the Marine Biological Association of the United Kingdom 78(3): 707-715.
Pinn, E. H. and M. R. Robertson (2001). Further analysis of the effect of bioturbation by Nephrops norvegicus (L.) on the acoustic return of the RoxAnn (TM) seabed discrimination system. ICES Journal of Marine Science 58(1): 216-219.
171
Pinn, E. H. and M. R. Robertson (2003). Effect of track spacing and data interpolation on the interpretation of benthic community distributions derived from RoxAnn (TM) acoustic surveys. ICES Journal of Marine Science 60(6): 1288-1297.
Pinn, E. H., M. R. Robertson, C. W. Shand and F. Armstrong (1998). Broad-scale benthic community analysis in the Greater Minch Area (Scottish west coast) using remote and nondestructive techniques. International Journal of Remote Sensing 19(16): 3039-3054.
Polovina, J. J. and S. Ralston (1987). Tropical snappers and groupers: biology and fisheries management, Boulder, CO, Westview Press, 659 pp.
Pouliquen, E., O. Bergem and N. G. Pace (1999). Time-evolution modeling of seafloor scatter. I. Concept. Journal of the Acoustical Society of America 105(6): 3136-3141.
Preston, J. M. (2004). Resampling sonar echo time series primarily for seabed sediment classification. United States Patent and Trademark Office, Patent Number US 6,801,474 B2, 12 pp.
Preston, J. M., A. C. Christney, L. S. Beran and W. T. Collins (2004a). Statistical seabed segmentation - from images and echoes to objective clustering. 7th European Conference on Underwater Acoustics, Delft, Netherlands, 5-8 July 2004. 6 pp.
Preston, J. M., A. C. Christney, L. S. Beran, W. T. Collins and R. A. McConnaughey (2002). Objective measures of acoustic diversity for benthic habitat characterization. Symposium on Effects of Fishing Activities on Benthic Habitats: Linking Geology, Biology, Socioeconomics, and Management, Tampa, FL, November 12-14, 2002.
Preston, J. M., A. C. Christney, W. T. Collins, R. A. McConnaughey and S. E. Syrjala (2004b). Considerations in large-scale acoustic seabed characterization for mapping benthic habitats. Proceedings of the ICES Annual Science Conference 22-25 September 2004, Vigo, Spain.
Preston, J. M., W. T. Collins, D. C. Mosher, R. H. Poeckert and R. H. Kuwahara (1999). The strength of correlations between geotechnical variables and acoustic classifications. Proceedings of Oceans '99, MTS/IEEE, Seattle, USA.
Quester Tangent Corporation (QTC) (2004). QTC IMPACT User Guide Version 3.40, Sidney, BC, Canada, 126 pp.
Richardson, M. D., D. L. Lavoie and K. B. Briggs (1997). Geoacoustic and physical properties of carbonate sediments of the Lower Florida Keys. Geo-Marine Letters 17(4): 316-324.
Riegl, B., R. P. Moyer, L. Morris, R. Virnstein and R. E. Dodge (2005a). Determination of the distribution of shallow-water seagrass and drift algae communities with acoustic seafloor discrimination. Revista De Biologia Tropical 53 Suppl. 1: 165-174.
172
Riegl, B. and W. E. Piller (2003). Possible refugia for reefs in times of environmental stress. International Journal of Earth Science 92: 520-531.
Riegl, B. M., J. Halfar, S. J. Purkis and L. Godinez-Orta (2007). Sedimentary facies of the eastern Pacific's northernmost reef-like setting (Cabo Pulmo, Mexico). Marine Geology 236(1-2): 61-77.
Riegl, B. M., R. P. Moyer, L. J. Morris, R. W. Virnstein and S. J. Purkis (2005b). Distribution and seasonal biomass of drift macroalgae in the Indian River Lagoon (Florida, USA) estimated with acoustic seafloor classification (QTCView, Echoplus). Journal of Experimental Marine Biology and Ecology 326(1): 89-104.
Riegl, B. M. and S. J. Purkis (2005). Detection of shallow subtidal corals from IKONOS satellite and QTC View (50, 200 kHz) single-beam sonar data (Arabian Gulf; Dubai, UAE). Remote Sensing of Environment 95(1): 96-114.
Roberts, J. M., A. J. Wheeler and A. Freiwald (2006). Reefs of the deep: The biology and geology of cold-water coral ecosystems. Science 312: 543-547.
Rousseeuw, P. J. (1984). Least median of squares regression. Journal of the American Statistical Association 79(388): 871-880.
Sadovy, Y. (1994). Grouper stocks of the western central Atlantic: the need for management and management needs. Proceedings of the Gulf and Caribbean Fisheries Institute 43: 43-64.
Sadovy, Y. and A.-M. Eklund (1999). Synopsis of biological data on the Nassau grouper, Epinephelus striatus (Bloch, 1792) and the jewfish, E. itajara (Lichtenstein, 1822), NOAA Technical Report NMFS, 146. U.S. Department of Commerce, Seattle, WA, 65 pp.
Sala, E., E. Ballesteros and R. M. Starr (2001). Rapid decline of Nassau grouper spawning aggregations in Belize: Fishery management and conservation needs. Fisheries 26(10): 23-30.
Sim, J. and C. C. Wright (2005). The Kappa statistic in reliability studies: Use, interpretation, and sample size requirements. Physical Therapy 85(3): 257-268.
Smith, W. H. F. and P. Wessel (1990). Gridding with Continuous Curvature Splines in Tension. Geophysics 55(3): 293-305.
Society for the Conservation of Reef Fish Aggregations (SCRFA). (2008). "Frequently asked questions (FAQs) about spawning aggregations." Retrieved 2 August, 2008, from http://www.scrfa.org/server/spawning/doc/faqs.pdf.
Sotheran, I. S., R. L. Foster-Smith and J. Davies (1997). Mapping of marine benthic habitats using image processing techniques within a raster-based geographic information system. Estuarine, Coastal and Shelf Science 44(Supplement A): 25-31.
173
Tsemahman, A. S., W. T. Collins and B. T. Prager (1997). Acoustic seabed classification and correlation analysis of sediment properties by QTC View. Proceedings of Oceans '97, MTS/IEEE, Halifax, Canada: pp. 921-926.
U.S. Coral Reef Task Force Mapping and Information Synthesis Working Group (1999). Coral reef mapping implementation plan (2nd draft). NOAA, NASA, and USGS (Work Group Co-Chairs), Washington, DC, 17 pp.
van Walree, P. A., J. Tegowski, C. Laban and D. G. Simons (2005). Acoustic seafloor discrimination with echo shape parameters: A comparison with the ground truth. Continental Shelf Research 25: 2273-2293.
Visuri, S., V. Koivunen and H. Oja (2000). Sign and rank covariance matrices. Journal of Statistical Planning and Inference 91(2): 557-575.
von Szalay, P. G. and R. A. McConnaughey (2002). The effect of slope and vessel speed on the performance of a single beam acoustic seabed classification system. Fisheries Research 54: 181-194.
Whaylen, L., P. Bush, B. Johnson, K. E. Luke, C. McCoy, S. Heppell, B. Semmens and M. Boardman (2006). Aggregation dynamics and lessons learned from five years of monitoring at a Nassau grouper (Epinephelus striatus) spawning aggregation in Little Cayman, Cayman Islands, BWI. Proceedings 58th Gulf and Caribbean Fisheries Institute.
White, W. H., A. R. Harborne, I. S. Sotheran, R. Walton and R. L. Foster-Smith (2003). Using an Acoustic Ground Discrimination System to map coral reef benthic classes. International Journal of Remote Sensing 24(13): 2641-2660.
Wilding, T. A., M. D. J. Sayer and P. G. Provost (2003). Factors affecting the performance of the acoustic ground discrimination system RoxAnn (TM). Ices Journal of Marine Science 60(6): 1373-1380.