REDUCTION OF REFRACTION EFFECTS DUE TO INADEQUATE
SOUND VELOCITY PROFILE MEASUREMENTS IN
MULTIBEAM ECHOSOUNDER SYSTEMS
M.D.E.K. GUNATHILAKA
UNIVERSITI TEKNOLOGI MALAYSIA
REDUCTION OF REFRACTION EFFECTS DUE TO INADEQUATE
SOUND VELOCITY PROFILE MEASUREMENTS IN
MULTIBEAM ECHOSOUNDER SYSTEMS
M.D.E.K. GUNATHILAKA
A thesis submitted in fulfilment of the
requirement for the award of the degree of
Master of Science (Hydrography)
Faculty of Geoinformation Engineering and Sciences
Universiti Teknologi Malaysia
JULY 2008
iii
DEDICATION
To my beloved parents and my loving Biyani …
iv
ACKNOWLEDGMENT
First of all, I would like express my sincere thanks to Professor Dr. Mohd
Razali Mahmud for his assistance and support throughout my study at Universiti
Teknologi Malaysia. This work would never be successful without his valuable
advices, guidance and encouragements.
My gratefulness to the lectures and the staff of the faculty of the
Geoinformation Science and Engineering in UTM, those who helped directly or
indirectly during my studies. Especially to Mr. Bustami and Mr. Gazali for their
logistic assistance during the data collection, and to the undergraduate students who
helped in many ways in the data collection.
Special thanks to the Vice Chancellor of the Sabaragamuwa University of
Sri Lanka, the Dean of the faculty of Geomatics, Heads of the departments of
Surveying and Geodesy and the department of CPRSG and my fellow staff
members at the University of Sabaragamuwa, for their full support on this study.
I also convey my gratitude to Dr. Othman, Mr. Joseph, Mr. Kelana and my
colleagues of Hydrographic Research and Training Office (HRTO), who have
helped me in many ways, especially giving valuable suggestions regarding to this
study. Last, but not least, I thank my family and my friends for their
encouragement, love and moral support, provided during my study.
v
ABSTRACT
The single most important acoustical variable in the water is its speed. The
average speed of sound in the ocean is about 1500 m/s, but its precise value in a
location is strongly depends on temperature, pressure and salinity of that particular
location. These factors change rapidly in time and space due to various reasons. In
data acquisition, the collection of these denser sound speed data becomes critical.
These inadequate sound speed measurements cause unknown propagation through
the water column that adds a major uncertainty to the multibeam echosounder
measurements (MBES). There are two types of sound speed measurements made in
the flat array multibeam sonars. Surface Sound Speed (SSS) is measured at the face
of the transducer and Sound Velocity Profile (SVP) is measured through the water
column. SSS is used to determine the beam pointing angle (beam steering) and SVP
is used to determine the depth and position (ray-tracing) of each beam. From these,
the SSS is measured almost at each ping vies and SVP may be once or twice a day
depending on the situation. When it comes to ray tracing, one has the options of
using either the SSS or the surface value of SVP (SSVP). Some multibeam software
manufacturers use the SSS in Snell’s refraction constant determination while others
use the surface value of the last performed SVP. In this study, both methods of
refraction constant determination are evaluated. The results clearly showed that the
use of SSS for Snell’s refraction constant determination gives about 25% to 30%
better results in multibeam bathymetry against refraction than the use of SSVP. A
combined solution of SSS and SSVP provide a better, simpler and cost effective
method of reduction of refraction effects in MBES. The results also demonstrated
that the effects of inadequate sound speed measurements in each phase of
bathymetric calculations would result in both depth and positional errors.
vi
ABSTRAK
Salah satu pemboleh ubah akustik bagi air yang penting adalah kelajuan air.
Purata kelajuan bunyi dalam laut adalah 1500 m/s, akan tetapi nilai kejituannya bagi
sesuatu kawasan bergantung kepada suhu, tekanan dan ketumpatan bagi kawasan
tersebut. Kesemua faktor ini berubah dengan cepat mengikut masa dan keluasan
disebabkan oleh pelbagai punca. Pengumpulan data kelajuan bunyi yang banyak
menjadi kritikal ketika kutipan data dilakukan. Pengukuran kelajuan bunyi yang
kurang berupaya untuk menghasilkan perambatan yang tak diketahui melalui lapisan
air telah menambahkan ketidakpastian kepada pengukuran pemerum gema berbilang
alur (MBES). Terdapat dua jenis kelajuan bunyi yang dilakukan dalam susunan
mendatar sonar berbilang alur. Kelajuan permukaan bunyi (SSS) adalah diukur pada
paras muka tranduser manakala profil halaju bunyi (SVP) diukur melalui lapisan air.
SSS digunakan untuk menentukan sudut arah alur dan SVP pula digunakan untuk
menentukan kedalaman dan kedudukan setiap alur. Dari sini, SSS diukur pada hampir
setiap ping dan SVP diukur sekali atau dua kali dalam satu hari bergantung pada
keadaan. Apabila hendak mengesan sinar, kaedah yang boleh dilakukan adalah sama
ada menggunakan SSS atau nilai permukaan bagi SVP (SSVP). Beberapa pengeluar
perisian pemerum gema berbilang alur menggunakan lebih banyak SSS dalam
penentuan angkatap biasan Snell, manakala pengeluar lain menggunakan nilai
permukaan bagi SVP terakhir yang diperolehi. Dalam kajian ini, penilaian terhadap
kedua-dua kaedah penentuan angkatap biasan dilakukan. Hasil kajian yang diperolehi
jelas menunjukkan penggunaan lebih banyak SSS bagi penentuan angkatap biasan
Snell memberi keputusan yang lebih baik bagi batimetri berbilang alur berbanding
dengan penggunaan SSVP iaitu antara 25% ke 30%. Gabungan penyelesaian SSS dan
SSVP menghasilkan kaedah lebih baik, mudah dan penjimatan untuk mengurangkan
kesan pembiasan dalam MBES. Hasil kajian juga menerangkan tentang kesan kurang
upaya bagi kelajuan bunyi dalam setiap fasa pengiraan MBES akan menyebabkan
selisih kedalaman dan penentududukan.
vii
TABLE OF CONTENTS
CHAPTER
1
TITLE
PAGE
TITLE PAGE
i
DECLARATION
ii
DEDICATION
iii
ACKNOWLEDGEMENTS
iv
ABSTRACT
v
ABSTRAK
vi
TABLE OF CONTENTS
vii
LIST OF TABLES
xii
LIST OF FIGURES
xiii
LIST OF SYMBOLS
xviii
LIST OF ABBREVIATIONS
xx
LIST OF APPENDICES
xxii
INTRODUCTION
1
1.1 Background
1
1.2 Research Problem
3
1.3 Aim of the Research
6
1.4 Research Objectives
7
1.5 Research Scope
7
1.6 Significance and Contributions of the Study
8
1.7 Review of the Relevant Literature on Refraction Issue
in MBES
1.8 Summary
9
13
viii
2
PRINCIPLE OF MULTIBEAM ECHOSOUNDING
14
2.1 Characteristics of the Acoustic Wave
14
2.2 Sound Wave in the Hydrographic Medium
15
2.2.1 Properties of Seawater Affecting Speed of Sound
15
2.2.1.1 Temperature
15
2.2.1.2 Salinity
16
2.2.1.3 Pressure
17
2.2.1.4 Density
17
2.2.2 Sound Speed Measurements in Water
18
2.2.3 Sound Speed Variability in the Ocean
18
2.2.3.1 Sound Speed Layers in the Oceans
19
2.3 Equation for Speed of Sound in the Water
22
2.4 Multibeam Echosounder Systems
27
2.4.1 Introduction
27
2.4.2 Principle of MBES Operation
28
2.4.3 Transducer
29
2.4.4 Transducer Arrays
30
2.4.4.1 Flat Array Transducers
31
2.4.4.2 Curved array Transducers
33
2.5 Beam Steering in MBES
34
2.6 Beam Steering in Flat Arrays
34
2.6.1 Mechanical Beam Steering
35
2.6.2 Electronic Beam Steering
35
2.6.2.1 Time Delay Method
37
2.6.2.2 Phase Delay Method
38
2.6.2.3 Fast Fourier Transformation Method
38
2.7 Beam Steering in Curved Arrays
39
2.8 Ray Tracing
41
2.8.1 Introduction
41
2.8.2 Vertical Incidence
42
2.8.3 Oblique Incidence
42
2.8.3.1 Layers with Constant Sound Speed
44
2.6.3.2 Layers with Constant Sound Speed Gradient
45
ix
2.9 Sound Speed Measurements in MBES
48
2.9.1 Surface Sound Speed (SSS)
49
2.9.2 Sound Velocity Profile (SVP)
49
2.10 Errors in Multibeam Systems
50
2.10.1 Introduction
50
2.10.2 What are the Largest Errors?
51
2.10.3 Does our Sound Speed Measurements
Adequate Enough?
2.10.4 Refraction in Multibeam Echosounders
3
51
52
2.10.4.1 Introduction
52
2.10.4.2 Effects during the Beam Steering
53
2.10.4.3 Effects Through the Water Column
54
2.11 Summary
55
FIELD DATA COLLECTION
56
3.1 Introduction
56
3.2 Survey Instrumentation
56
3.2.1 The MBES system
56
3.2.2 The Single Beam Echosounder (SBES)
57
3.2.3 The Positioning System
58
3.2.4 Sound Speed Measurements
59
3.2.4.1 SSS Measurements
59
3.2.4.2 SVP Measurements
60
3.2.5 Motion (Attitude) Sensor
61
3.2.6 Tide Gauge
61
3.3 Survey Software
62
3.4 Survey Platform
63
3.5 Field Data Collection
64
3.6 Methodology for Determination of Inadequate
Sound Speed Measurements in MBES
65
3.6.1 The Effects of Inadequate SSS on MBES
65
3.6.1.1 Simulated Data Case for SSS
65
3.6.1.2 Real Data Case for SSS
66
x
3.6.2 Determination of Inadequate Sound Velocity
Profile (SVP) Effects on MBES
67
3.6.2.1 Simulated Data Case for SVP
68
3.6.2.2 Real Data Case for SVP
69
3.7 Comparison of SSS and SSVP in Determination of
4
5
6
Snell’s Refraction Constant for Refraction Reduction
70
3.7.1 Data Collection for Refraction Reduction
70
3.7.2 Raw Data Extraction
71
3.7.2.1 MBES Data
71
3.7.2.2 Transducer Position Data
72
3.7.2.3 Vessel Attitude Data
73
3.7.2.4 SBES DTM Data
74
COMPUTER PROGRAM DEVELOPMENT
76
4.1 Introduction
76
4.2 The SSS Program
76
4.3 The Algorithm of the SSS Program
89
4.4 The SSVP Program
92
4.5 The Algorithm of the SSVP Program
92
DATA PROCESSING
94
5.1 Programme Validation
94
5.2 Data Processing
94
5.2.1 MBES Data Processing
94
5.2.2 SBES Data Processing
97
5.2.3 Final Comparison
99
RESULTS AND ANALYSIS
102
6.1 Introduction
102
6.2 Results of Program Validation
102
6.2.1 Northing Comparison
103
6.2.2 Easting Comparison
104
6.2.3 Depth Comparison
106
xi
6.2.4 Summary of Program Validation
6.3 Inadequate SSS Effects on Flat Array MBES Transducers
107
108
6.3.1 Synthetic Data Results
108
6.3.2 Real Data Results
110
6.3.3 Summary of Inadequate SSS Effects on
Flat Array MBES Transducers
6.4 Inadequate SVP Effects on Flat Array MBES Transducers
111
112
6.4.1 Synthetic Data Results
113
6.4.2 Real Data Results
114
6.4.3 Summary of Inadequate SVP Effects on
Flat Array MBES Transducers
6.5 Refraction Reduction Results
7
115
116
6.5.1 Nadir Comparison
116
6.5.2 Outer Comparison
118
6.5.3 Summary of Refraction Reduction Results
120
CONCLUSION AND RECOMMENDATIONS
121
7.1 Conclusion
121
7.2 Recommendations
122
REFERENCES
124
Appendices A-E
128-145
xii
LIST OF TABLES
TABLE NO.
TITLE
PAGE
2.1
Table of coefficients (UNESCO Equation)
25
2.2
Table of coefficients (Grosso’s Equation)
27
3.1
Sound speed configurations to determine the SSS effects
in the simulated data case
3.2
SSS and SVP configuration to determine the SSS effects
in the real data case
3.3
67
Sound speed configurations to determine the SVP effects
in the simulated data case
3.4
66
69
SSS and different SVP configurations to determine the SVP
effects in the real data case
70
xiii
LIST OF FIGURES
FIGURE NO.
1.1
TITLE
PAGE
Illustration of how refraction degrade the accuracy of
MBES data
5
1.2
Observe the parallel ridges and valleys due to sound speed errors
5
2.1
Variation of water temperature with depth in Labrador Sea, Canada
16
2.2
Variation of water salinity with depth in Labrador Sea, Canada
17
2.3
Example of sound speed profiles and it’s diurnal variation
19
2.4
Oceanic water layers and example deep-sea SVP
20
2.5
Typical temperature and salinity variations as a function of depth
21
2.6
The complexity of the oceanography of coastal water masses
22
2.7
MBES beam footprint and swath coverage
29
2.8
Beam forming in flat transducer arrays
30
2.9
Beam footprints resulting from the intersection of transmission
and reception in RESON SeaBat 8124 MBES
32
2.10
Example for flat transducer arrays
32
2.11
Curved or Barrel type transducer array
33
2.12
Typical examples of curved transducer arrays
34
2.13
Electronic Beam Steering
36
2.14
Applied delays to individual transducer elements to detect
oblique beams
37
2.15
Stave selection for beam steering in curved transducer array
39
2.16
Weights added to the neighbourhood and outermost beams
have to be steered
2.17
2.18
40
Outer beams steered using the physical shape of the transducer
combined with electronic steering
40
Ray tracing in MBES
41
xiv
2.19
Illustration of oblique incidence
43
2.20
Modelling the sound speed profile in the water
44
2.21
Ray path in a constant sound speed gradient layer
46
2.22
Cross-section of the sound speed structure on the edge of
Georges Bank
52
2.23
Refraction effects in each phase of the MBES
52
2.24
Effect of change in SSS in beam pointing angle in a flat array
transducer: In the case of true SSS is greater than the measured value 54
3.1
The MBES System
57
3.2
The SBES System
58
3.3
The DGPS System
58
3.4
The SSS Probe
59
3.5
The SVP Probe
60
3.6
MAHRS Attitude Sensor
61
3.7
Tide Gauge
62
3.8
QINSy console
63
3.9
Survey Platform
64
3.10
Survey areas
65
3.11
Altering the SSS value in the sonar processor
67
3.12
Synthetic two-layered SVP
68
3.13
SVPs used to determine the SVP effects in the real data case
69
3.14
Selected data items in each MBES beam
71
3.15
Exported raw MBES data string
72
3.16
Selected raw data items in transducer positions
72
3.17
Exported Transducer Position data string
73
3.18
System selection (MRU) in analyse
74
3.19
Exported raw attitude data string
74
3.20
Selected data source and parameters in SBES DTM
75
3.21
SBES DTM data string
75
4.1
Conversion of total samples to travel time
77
4.2
Interpolation of roll, heave and pitch with respect to each ping time
77
4.3
Flowchart for the calculation of effective beam angle
78
4.4
Flowchart for the calculation of net pitch angle
79
xv
4.5
Flowchart for the calculation of final beam direction
4.6
Flowchart for the calculation of the Snell’s refraction constant
using surface sound speeds
4.7
80
Flowchart for the calculation of the sound speed layer number
and travel time up to (N-1) sound speed layer
4.8
79
81
Flowchart for the calculation of the travel time in the last
sound speed layer
82
4.9
Calculation of the range distance in the last sound speed layer
82
4.10
Flowchart for the calculation of the depth in the last
sound speed layer
4.11
Flowchart for the calculation of the final reduced depth
of each beam
4.12
5.1
91
Flowchart for the calculation of the Snell’s refraction constant
using SSVP
4.21
89
Algorithm for bathymetric calculations using the SSS in
refraction constant
4.20
88
Flowchart for the calculation of the final Easting and
Northing for each beam
4.19
87
Flowchart for the calculation of the Easting and Northing
differences with respect to the sonar head position for each beam
4.18
87
Flowchart for the calculation of the corrected total across track with
respect to the corrected beam direction for each beam of each ping
4.17
86
Flowchart for the calculation of the corrected beam direction
with respect to the sonar head position for each beam of each ping
4.16
85
Flowchart for the calculation of the total across track for
each beam of each ping
4.15
84
Flowchart for the calculation of the across track distances
for each beam at the last sound speed layer
4.14
83
Flowchart for the calculation of the total across track distance up to
(N-1) sound speed layer for each beam from the sonar head position
4.13
83
92
Algorithm for bathymetric calculations using SSVP
in the refraction constant
93
Final MBES coordinate conversion
95
xvi
5.2
Processed MBES bathymetric data from the program output
loaded in to AutoCAD as a multiple point script file
96
5.3
Quicksurf software loaded in AutoCAD R14
96
5.4
Generated DTM for the first MBES data set using Quicksurf
97
5.5
SBES Script (SCR) generation
98
5.6
SBES profile after running the script file in AutoCAD
98
5.7
Loaded data sets to AutoCAD
99
5.8
Generated profiles are saved as blocks with a reference (base) point
100
5.9
Loaded profile blocks in to a single drawing for the comparison
101
5.10
All the blocks are overlaid each other using the common
base point in the final comparison
6.1
Northing coordinate comparison between QINSy vs. SSVP
programmes for the first ping
6.2
104
Easting coordinates comparison between QINSy vs. SSVP
for the first ping
6.4
103
Northing coordinate comparison between QINSy vs. SSVP
programmes for the second (200th) ping
6.3
101
105
Easting coordinates comparison between QINSy vs. SSVP
for the second (200th) ping
105
6.5
Depth comparison between QINSy vs. SSVP for the first ping
106
6.6
Depth comparison between QINSy vs. SSVP for the
second (200th) ping
6.7
107
Variation of the magnitude of the angular error with respect to the
beam-pointing angle for different SSS variations
108
6.8
Across track errors for 100m flat sea bottom for different SSS errors
109
6.9
Depth errors for 100m flat sea bottom for different SSS errors
109
6.10
Impact on the shape of the swath for different SSS errors
on a flat sea floor from a flat MBES
110
6.11
Real examples for SSS variation effects on a flat array MBES swath
111
6.12
IHO error budgets for different levels of surveys
112
6.13
Depth errors due to 10m/s SVP variation at the first 10m layer
of the SVP for a 100m deep, flat sea bottom
113
xvii
6.14
Across track errors due to 10m/s SVP variation at the first
10m layer of the SVP for a 100m deep flat sea bottom
6.15
113
Impact of the sound velocity profile errors on the swath shape
of a flat 100m deep sea floor due to 10m/s SVP variation at the
first 10m layer of the SVP
6.16
114
True examples for SVP variation effects on swath
in a flat array MBES
115
6.17
SSS and SSVP profiles at the nadir from the MBES line 01
117
6.18
SSS and SSVP profiles at the nadir from MBES line 02
117
6.19
SSS, SSVP and corresponding SBES profile comparison
at the outer edge of the swath of MBES line 01
6.20
118
SSS and SSVP outer beam profiles for MBES line 01 and
corresponding SBES and adjacent MBES nadir (line 02)
profile comparison
6.21
119
SSS and SSVP outer beam profiles for MBES line 02 and
corresponding SBES and adjacent MBES nadir (line 01)
profile comparison
119
xviii
LIST OF SYMBOLS
B
-
Bulk modules
C, C, c -
Speed of sound
C0
-
Sound speed at the transducer face
C1
-
Incorrect sound speed measured at the transducer face
Ci
-
Sound speed at the ith layer
D
-
Depth
d
-
Element spacing
E
-
Easting
f
-
Frequency
gi
-
Gradient of the sound speed
h
-
Depth of the sound speed layer
N, N
-
Northing
n
-
Number of elements in the transducer array
Ri
-
Radius of the curvature at the sound speed layer
P
-
Pressure
p
-
Density
R
-
Range
S
-
Salinity
T
-
Temperature
t
-
Time
v
-
Sound speed of each layer
x
-
Horizontal distance
Δi
-
Layer thickness
Δϕ s
-
Phase shift for the ith element
λ
-
Wave length
μ
-
Harmonic mean speed of sound
xix
θ
-
Beam angle
θs
-
Steering angle
ρ
-
Snell’s refraction coefficient
xx
LIST OF ABBREVIATIONS
ASCII
-
American Standard Code for Information Interchange
AutoCAD
-
Automatic Computer Aided Design
CoG
-
Centre of Gravity
CTD
-
Conductivity Temperature Density
CSV
-
Comma Separated Values
Db
-
Database
DGPS
-
Differential Global Positioning System
DTM
-
Digital Terrain Model
DWG
-
Drawing
DXF
-
Data Exchange Format
EEZ
-
Exclusive Economic Zone
GPS
-
Global Positioning System
IHO
-
International Hydrographic Organisation
MAHRS
-
Meridian Attitude and Heading Reference System
MATLAB
-
Matrix Laboratory
MBES
-
Multibeam Echosounder System
MRU
-
Motion Reference Unit
MVP
-
Moving vessel Profiler
NPL
-
National Physics Laboratory
OMG
-
Ocean Mapping Group
ppt
-
parts per thousand
pps
-
pulse per second
QINSy
-
Quality Integrated Navigation System
QPS
-
Quality Positioning Service
Qsurf
-
QuickSurf
RTKGPS
-
Real Time Kinematic Global Positioning System
SBES
-
Single Beam Echosounder
xxi
SCR
-
Script file
SSS
-
Side Scan Sonar
SSVP
-
Surface value of the Sound Velocity Profile
SVP
-
Sound Velocity Profile
TIN
-
Triangular Irregular Network
TWTT
-
Two Way Travel Time
UNB
-
University of New Brunswick
UNESCO
-
United Nations Educational, Scientific and Cultural
Organization
USACE
-
United States Army Crops of Engineers
XLS
-
Microsoft Excel file
3D
-
Three-dimensional
xxii
LIST OF APPENDICES
APPENDIX
TITLE
PAGE
A
Database Settings
128
B
Synthetic data for SSS case
136
C
Synthetic data for SVP case
137
D
Program validation Results
138
E
Publications
144
CHAPTER 1
INRODUCTION
1.1 Background
One of the most impressive hydrographic survey technique developed
during the past few decades is the Multibeam Echosounder System (MBES). It is a
rapid and more automated depth measurement system, guaranteeing the full bottom
coverage.
Therefore it has become the number one choice for most of the
hydrographic surveys. Multibeam sonars uses sound as a remote sensing tool. The
fundamental data collected by these sonars are the two way travel time of the short
acoustic pulse travelling between the transducer and the bottom surface and the
direction from which the echo is reflected. A typical MBES (eg. RESON Seabat
8124) has some 80 separate beams, spanning 120 degrees are sounded across the
ship’s track on each acoustic ping, which will normally covers an area of 3.5 times
of the depth.
Use of MBES for accuracy–critical applications has now become wide
spread with the improvement in acoustic transducer design and digital data
processing. Now MBES have become a cost effective, reliable tool and being
increasingly employed in ocean mapping, dredging operations, route surveys and
various other underwater engineering works (Dinn et al., 1995). Along with the
adoption of the MBES as the instrument of choice of the most hydrographic
applications, has come the challenge of minimising any associated errors
(Cartwright and Clarke, 2002).
2
Final sounding data from the MBES system is a result of processing
information from several data sources.
These include the ship’s heading and
attitude data from the gyrocompass and the motion sensor; vertical reference data
from the tide gauge; positional data from the Global Positioning System (GPS) unit
and sound velocity data from the Conductivity Temperature Density (CTD) or
Sound Velocity Profile (SVP) probe in addition to the basic MBES data itself. Data
from each source is subject to individual errors contributing to overall data quality.
To limit these, system planners often have established error budgets for various
components of the system.
The International Hydrographic Organization (IHO) recommends accuracy
limits for the type of hydrographic surveys and the depth of water in which a survey
is conducted.
These accuracies are divided in to two categories, horizontal
accuracy and depth accuracy.
Horizontal accuracy refers to the horizontal
positioning accuracy of each sonar beam and depth accuracy includes amongst
other things like tidal measurement errors, data processing errors and measurement
system and sound velocity errors (Batton, 2004).
Thanks to the intensive researches carried over the last decade, system
manufactures have introduced equipments advertised to achieve positional
uncertainties of 2 cm or better (Real Time kinematics GPS), tidal measurement
uncertainties less than 2 cm (Real Time Tide) and vessel altitude uncertainties of
0.02 degrees. The uncertainties of these instruments contribute to the surveys are
within or if not better, than the accuracy suggested by the IHO (Batton, 2004).
Therefore with the advent of the new technologies, the last remaining obstacles to
absolute precision are sound speed variance and roll biases (William and Capell,
1999).
3
This chapter outlines the core areas of this study including research problem
statement, aim and objectives, research scope, significance of the study as well as
the discussion of related previous works. Chapter 2 provide a detailed theoretical
overview of the marine acoustical environment, sound wave propagation, MBES
system and need of sound speed in MBES. The next chapter (Chapter 3) discusses
the field data collection. The computer program development part is explained in
Chapter 4. Here the algorithms used and the flowcharts of each program are
discussed in detail. In Chapter 5, data processing techniques are presented. The
results and data analysis are discussed in Chapter 6 and finally, Chapter 7 concludes
the results obtained and the recommendations for the future studies also presented.
1.2 Research Problem
The nature of the sea environment is the most fundamental factor, which
separates land and sea surveying. The sea is fluid and dynamic. It is coronial and
full of living organisms that changes the structure.
The characteristics of the
medium through which measurements are made are always subjected to variation
(Ingham and Abbott, 1992). These variations must be understood and corrections
to be applied in order to achieve precise results.
When it comes to acoustic depth measurements in the oceans, the dominant
character is speed of sound. The speed of sound in the oceans is subjected to
significance changes caused by rapid changes in temperature, pressure and salinity
over a short period of time. These changes are more prominent in continental shelf
regions; as a result of rapid heating and cooling of the water surface due to solar
heating, interactions with fresh waters carried by rivers, tidal and current mixing
and so on (William and Capell, 1999).
4
Measuring these physical properties that control the speed of sound in the
ocean (using CTD probe) or direct sound speed measurements (using SVP probe) is
the standard procedure for collecting the sound speed information. These physical
oceanographic variables have clearly demonstrated temporal and spatial scale
variation during common hydrographic surveys that are usually extending from
days to weeks and survey lines from kilometers to tens of kilometers. As a result,
in most hydrographic operations one must take discrete measurements of sound
speeds at periods of more than once a day; bringing a survey vessel to a halt,
lowering a sensor several hundred meters and then taking care of all the data quality
assurance and data transfer protocols necessary. This would commonly involve at
least 30 minutes of ship time.
Because of this, agencies are reluctant to take more frequent observations
and thereby implicitly assumed that the space and time variability of the ocean
could adequately be described using these sparse observations. Even more, now the
swath of the multibeam sonars have moved to ever wider angular sectors in order to
achieve even wider coverage, as the other sources of uncertainties have been
gradually eliminated, which means they are more sensitive to refraction (Clarke et
al. 2000). When applying these SVPs, there are two principal limitations exists:
a) The water mass really does change over time scales much
shorter than the standard sampling period.
b) The application of SVP is almost universally done based on
the prior observations only.
This inadequate sound speed measurements cause an unknown propagation
(refraction errors) that adds a major source of uncertainty to depth measurements,
resulting artifacts can create short-wavelength topographic features that may be
misinterpreted as sea floor relief (Gardner, et al., 2001) as shown in Figures 1.1 and
1.2.
5
Figure 1.1 Illustration of how refraction degrade the accuracy of MBES data
(OMG-UNB, Canada)
(a)
(b)
Figure 1.2 Observe the Parallel ridges and valleys due to sound speed errors
(a) Exhibits an artificial wave-like pattern in DTM (Jeroen, 2007) (b) Exhibits how
contours are altered by these artificial features (OMG-UNB)
Almost all of the flat array MBES (eg. RESON SeaBat 8124) measures two
types of sound speed measurements. The surface sound speed (SSS) measured
using the probe near the sonar head is used for beam steering purpose and the sound
6
velocity profile (SVP) measured through the water column used for depth and
position calculation of each beam (ray tracing). The SSS is measured continuously
through out the survey period, while the SVP is only measured in discrete of times.
Therefore the dominant uncertainty remaining to be solved is caused by the fact that
we have an imperfect knowledge of the water column and accompanying changes
in sound speed with depth (Cartwright and Clarke, 2002).
In this case, to address this imperfect knowledge on SVP, some multibeam
system manufacturers use more frequent (real-time) SSS measurements, measured
at the sonar head along with spares SVP in ray tracing. Here, they use SSS in
refraction (Snell’s) constant determination for each beam, measured almost at each
ping vies (about >10Hz). While other manufacturers use SSS in beam steering
purpose only and the SVP is used alone in ray tracing (here they use the surface
value of the SVP for refraction constant determination). This seems that, still there
is no agreement in the hydrograplic community, which one gives better results
against refraction.
1.3 Aim of the Research
The aim of this research is to evaluate, the most appropriate value in the
determination of refraction coefficient for the ray tracing purpose to perform the
refraction calculations. That is, either the surface sound speed (SSS) or the surface
value of the sound velocity profile (SSVP) giving the best results in ray tracing.
7
1.4 Research Objectives
The objectives of this research are;
1) To study the effects caused by inadequate sound speed measurements in each
phase of the multibeam echosounder system. The effects from:
a) The surface sound speed.
b) The sound speed through the water column.
2) To develop two computer programmes for MBES bathymetric calculations
using SSS and SSVP as refraction constant.
3) To perform a comparative test between the above two approaches to identify
any significance difference between the two methods of refraction constant
determination.
1.5 Research Scope
Unlike in the open oceans, where the sound velocity profile has a
predictable and stable shape, in coastal and shallower areas, (continental shelf
regions) the SVPs are irregular and unpredictable. Therefore, for this study the
fieldwork is carried out in shallow coastal waters in Lido beach, Johor Bahru,
Malaysia. In addition to that, the effects are simulated for a 100m deep synthetic
flat seabed for each case.
Over the years various types of multibeam echosounder system
configurations have been designed and developed for various purposes. Curved
array, flat array, dual flat array are some of them. Each individual system behaves
differently in refraction. This study is limited to the Mill’s cross type, flat array
multibeam configuration. RESON SeaBat 8124 is a typical system of that kind.
8
The SSS is measured using the surface sound speed-measuring probe
located at the face of the transducer and SVP-15 probe is used to measure the sound
velocity profile through the water column.
QINSy version 7.5 software was used to collect, extract raw data and
process the multibeam data. AutoCAD R14 and QuickSurf 5.1 software are used in
visualization of bathymetric data and Digital Terrain Model (DTM) generation.
MATLAB- R2006a is used to develop the computer programmes. Here the
ray tracing is performed assuming that each sound speed layer has a constant sound
speed. The bathymetric calculation procedures used in the developed programs are
the same as the QINSy software procedures, except the refraction constant
determination method. The bathymetric results from the QINSy software are used
to validate the results from the developed programs.
Nadir beams are least affected by refraction, therefore in this study the nadir
beams were used for benchmarking or as reference depths (true depths) in
comparison of refraction effects. SBES data is also used for this purpose.
Corresponding profiles from SSS and SSVP DTMs, SBES lines and
adjacent MBES nadir area are compared to each other in the final comparative test
to determine the significance between the two approaches.
1.6 Significance and Contributions of the Study
Since MBES is a recent development, very few researches have been carried
out in the issue of refraction. For Malaysia, MBES is even newer. There is hardly
any proper study carried out in Malaysian waters of this kind.
9
This study will completely address how the variation of the sound speed
affects in multibeam echosounder bathymetric measurements, both in beam forming
and in ray path calculations. This is very much important in equatorial waters
where the sound speeds are more critical due to solar heating, tidal and current
mixing.
This knowledge will be very much useful to the survey planners to make
extra measures to overcome the effects caused and hence improve the efficiency
and accuracy of the works.
Finally, this will give more insight to MBES system and software
developers to come up with advanced systems and software that will suffer less
effects from sound speed variation (refraction) in future.
1.7 Review of Relevant Literature on Refraction Issue in MBES
Over the years hydrographers and oceanographers have faced greater
challenges when they dealt with oceanic parameters, especially when they began to
use the acoustic techniques. Because of this, many researchers have done much
research and experiments on these matters. Some tried to understand how this
really affect the measurements, while other researchers tried to come up with a
solution for it. These solutions can be discussed in two phases. The first one is in
post processing content like applying ray tracing techniques, while the next
approach is addressing the roots (in data collection stage) of the problem; that is to
collect the near continuous sound speed profiles.
Badiey et al. (2002) try to understand the correlation between the
oceanographic features and the high-frequency acoustic wave propagation. Their
results clearly showed a direct relationship between salinity and temperature
changes with acoustic wave propagation in shallow waters. Furthermore, the
hydrodynamic parameters such as surface waves, tides and current can influence
amplitude and travel time of signal transmissions.
10
Gardner et al. (2001) have highlighted that the refraction is the single
biggest limitation on the quality of bathymetric data, and strong water stratification
causes problems for the beam steering and ray tracing in MBES. They suggested
measuring sound speed profiles more frequently to minimize these effects.
But measuring highly variable and dynamic oceanographic components is
not that easy. Clarke (2002) illustrated how fast water masses changes in oceans, in
time and space, using observed sound speed cross-sections. He also stated that
when beams become less vertical, the affects get worse. As a result one could see
parallel ridges in MBES data, along the ship-track direction where neighbouring
lines get overlapped.
Tonchia and Bisquay (1996) and Dinn et al. (1995) have shown that the
inadequate sound speed measurements effects in two phases in MBES. That is, the
surface sound speed affects the beam forming and the sound velocity profile affects
the ray path. Beam forming depend on the transducer configurations (flat-level,
flat-tiled, circular-faced), and they mathematically illustrate the effects in each
transducer configuration. Furthermore, they have shown that when the vessels roll
is significant, the roll modulate the depth errors contributed by sound speed
uncertainty. Finally, they suggested measuring the surface sound speed continually
and to use adaptive modeling of the error regime coupled with deliberately
introduced redundancy in the depth data in an effort to enable interpolation between
temporally and spatially sparse SVPs.
Kammerer et al. (1998) faced the same problem while they try to monitor
the temporal changes in seabed morphology, using multibeam sonars in Saguenay
River in France. The local mixing of fresh and salt water has introduced more
uncertainties than they first expected, due to the refraction. They dealt with this by
separating the different lines corresponding to the different sound velocity profiles
(SVP) taken during the survey and distinguished them geographically within each
of these sets, assuming that the water masses are affected differently and requiring
different refraction coefficients. Then, they applied estimated refraction corrections
to each of these groups of lines; hence reduce the curvature of the swath.
11
Batton (2004) found out that the sound velocity formula used to compute the
speed of sound in the water column is also a source of uncertainty related to the
horizontal position of the chart depths. She measured the temperature, conductivity
and pressure in North Atlantic Ocean and used Chen and Millero, Meckenzie and
Medwin formulae for the estimation of sound speed. Then, she performed ray
tracing to compute the horizontal distances of refraction for the beams through the
water column. Through this, she concluded that the sound velocity formula used to
compute the sound speed also contribute to uncertainties associated with outer
swath of MBES.
William et al. (1999) described a method to determine the magnitude of the
SVP errors using the MBES data itself, by running cross lines. These crossing
swaths are obtained from the check lines used in most hydrographic surveys. Here,
in addition to refraction errors they observed the roll bias and tidal differences also.
Beaudoin et al. (2004) demonstrated that it is possible to correct soundings
corrupted by incorrect surface sound speed in post-processing.
During their
multibeam survey in Amundsen Gulf, Canada; their surface sound speed probe has
failed in several occasions.
measurements.
This caused a greater uncertainty in their
Then they interpolated SSS from the measured SVP’s and
recalculated the beam steering angles. Through this they were able to improve the
accuracy of the data.
Furlong et al. (1997) had come up with a solution to measure oceanographic
parameters in real-time using a computer-controlled winch and a davit. The winch
deploys a ‘free-fall’ fish that can be instrumented with a sound velocity sensor (like
CTD). They named this as “Moving Vessel Profiler” (MVP). The initial system
was capable of profiling down to 100 meters even at the vessel speeds up to 12
knots and the entire procedure from the launch to recovery take about 4 minutes.
This technique improved the accuracy of the MBES data and do not interrupt the
survey process as it operates while the vessel is underway.
12
Clarke et al. (2000) and Clarke and Parrott (2001) had used the above
technique (MVP) to study the sound speed variability of the oceans and with the use
of MVP along with MBES, frequent water column information allowed a much
better control of sounding errors due to the spatial and temporal variations in the
water column; making the wider swath ( 160o ) MBES more reliable.
Cartwright and Clarke (2002) also faced serious problems with refraction
when they carried out a survey in Fraser River delta, Canada. This River deltaic
area was considered being an extreme refraction environment with strong sound
speed anomaly. Even with the MVP, it was not possible to collect those large
number of spatially dense sound velocity profiles. There they recalculated the
departure angles and ray tracing using the spatially interpolated SVPs in order to
increase the accuracy of the data in post processing.
Kammerer and Clarke (2000) presented another method of removing
refraction effects in MBES using the MBES data itself. They tried to develop a
systematic analysis and correction software package for multibeam in postprocessing context. The methodology consists of the estimation of the variation in
the water sound speed distribution by using the information given by the MBES
dataset. This was done by the evaluation of appropriate modelized SVPs, which
was added to an already existing SVP or applied directly to the raw data. Here they
considered two methods, the first one was using two neighboring parallel lines and
the second method was cross-line method. In both cases they assumed that the
nadir beams are unaffected by refraction.
Beaudoin et al. (2004) developed a sound speed decision support system for
multibeam sonar operations in the Canadian Arctic. This helps hydrographers make
better decisions by integrating the various types of information relating to sound
speed into a single software application.
13
Jeroen (2007) used a method called ‘sound velocity profile inversion’ to
correct the refraction errors in MBES data. The method was based on the overlap
difference of the swaths, the measured SVP and the measured SSS at the sonar
head. By that he defined a linear SVP (linear parameterized SVP) for each ping and
then performed the bathymetric calculations. This way he managed to achieve
promising results against refraction affects.
Furthermore, he proved that the
measuring of SVP could be completely eliminated by adopting this method.
1.8 Summary
In dynamic water environments with considerable variation of sound speeds
in the water column, it is important to adequately correct bathymetric data for
refraction effects in the case of limited SVP information.
The aim of this study is to evaluate the use of SSS in refraction constant
determination for reduction of refraction effects. The best thing about the SSS is, it
is freely available in all flat multibeam systems and can be considered as continues
longitudinal section of sound speed drown across the water surface along the survey
line.
Therefore, no additional measurements (observations) are needed and
computational procedures are also less complicated.
CHAPTER 2
PRINCIPLE OF MULTIBEAM ECHOSOUNDING
2.1 Characteristics of the Acoustic Wave
An acoustic wave is a mechanical pressure disturbance in a medium
generated through a mechanical vibration of some surface. As the surface vibrate
forward and backward it exerts high and low pressure on particles of the medium,
and the wave is gradually transmitted forward (Tucker, 1966). In reality sound
waves propagates as spherical wave fronts. But when it reaches far from the
source, they can be approximated as plane waves. This approximation enables us to
interpret sound wave more easily.
The velocity of propagation of the sound wave through a medium is a
function of bulk modulus of elasticity and density of that particular medium
(Burdic, 1991).
C=
B
p
(2.1)
The bulk modulus B is a measure of the ratio between the stress and the
strain. It is the capacity of the material to be deformed by an external force. The
density p is controlled by the amount of material per unit of volume. The sound
speed is directly proportional to the ability of the medium to be deformed and
inversely proportional to the amount of material per unit of volume.
15
Also the speed of sound ( C ) is a function of its frequency ( f ) and its
wavelength ( λ )
C = f ×λ
(2.2)
2.2 Sound Wave in the Hydrographic Medium
The single most important acoustical variable in the water is the speed of
sound. The distribution of sound speed in the water influences all other acoustic
phenomena (Etter, 2003).
2.2.1 Properties of Seawater Effecting Speed of Sound
The average speed of sound in the seawater is approximately 1500 ms-1, but
its precise value is strongly depending up on temperature, pressure and salinity in
that particular location (Caruthers, 1977; Dera, 1991; Etter, 2003).
2.2.1.1 Temperature
The temperature of the water varies with depth because of the weaker solar
energy penetration, with seasonal cycle and on daily basis corresponding to weather
conditions (Figure 2.1).
Also the currents, tides and underwater geothermal
phenomenon influence the water temperature locally (Berdic, 1991). Temperature
is the easiest measurable parameter in water and indeed is one of the earliest
parameter to be studied. This is the primary dependent of sound speed through the
water column (Etter, 2003). Temperature ranges from 0 to 30 degrees Celsius
throughout the most of the world’s oceans. Typically, a change in temperature in
16
one degree Celsius would correspond an approximately 3 ms-1 change in sound
speed (Schmidt, et al., 2006).
Figure 2.1 Variation of water temperature with depth in Labrador Sea, Canada
(OMG, 2005)
2.2.1.2 Salinity
Salinity is the amount of dissolved materials in water.
Seawater is a
complex solution containing large number of compounds, primarily in their ironic
forms. Among them chloride, sodium, sulphate, magnesium and potassium are the
most abundant. The measurement units of salinity are usually termed as grams of
dissolved salts per kilogram of water and written as % (percentage) or ppt (parts per
thousand). Salinity ranges from 0 to 40 ppt through out most of the world’s oceans.
The salinity of the water is highly variable near the surface and becomes more
constant with increasing depth as shown in Figure 2.2.
It also exhibits both
seasonal and diurnal variations, especially when there is an influx of fresh water
with salt due to river or melting icebergs, and even sometimes due to rain (Horne,
1969). Typically, a change in salinity of one ppt would correspond to an
approximate change in sound speed of 1.2 ms-1 (Schmidt, et al., 2006).
17
Figure 2.2 Variation of water salinity with depth in Labrador Sea, Canada (OMG,
2005)
2.2.1.3 Pressure
The pressure of the water column also affects the speed of sound. It is
related with depth. Water is compressible and density of the water increases with
the depth (Horne, 1969). For hydrodynamic studies, the important parameter is the
density of the water, which is a function of pressure, temperature and salinity. For
underwater acoustics, the primary concern is the pressure, which is a function of
depth, along with atmospheric pressure and latitude. The rate of change of sound
velocity is approximately 0.5 ms-1 for every changes of one atmosphere; i.e.
approximately 10 meters of water depth (Schmidt, et al., 2006). The pressure has a
major influence on the sound velocity in deep water.
2.2.1.4 Density
Water density is dependent upon the previous parameters, i.e. temperature,
salinity and pressure. Fifty percent of the ocean waters have a density between
1027.7 and 1027.9 kgm-3. The largest influence on density is its compressibility
with depth. Water with a density of 1028 kgm-3 at the surface would have a density
of 1051 kgm-3 at a depth of 5000 meters (IHO, 2005).
18
2.2.2 Sound Speed Measurements in Water
Sound speed measurement has a long history.
But the accurate
determination of the speed of sound in water began in 1827 when Colladon, Sturn
and Wood made measurements on Lake Geneva, Switzerland (Burdic, 1991).
There are two primary methods being used for sound speed measurements. The
first one is the indirect method, where the sound speed is calculated from its
measured components like temperature, salinity and pressure or depth. The second
one is to measure the speed of sound directly through the medium using an accurate
transducer (Cartwright, 2003).
2.2.3 Sound Speed Variability in the Ocean
It is clear that the factors affecting the sound speed vary through out the
seawaters. As a result the sound velocity profiles also varies. In the open ocean the
main factors are time of day, season and latitude. Here, the primary determinants
are temperature and depth. The salinity is considered stable and predictable with
only a small variation on the surface due to evaporation and precipitation. The
temperature profile of the ocean can be roughly divided into two layers, the surface
and deep layer, with in the boundary at approximately 100m. The deep layer has
relatively constant or decreasing temperature gradient that remains in place
throughout the year. The surface is subjected to changes in the temperature profile
with the depth because of the influence of solar heating, wind influence and wave
action (Dera, 1991). Therefore, it is quite predictable and stable shape.
Figure 2.3 shows the typical range of sound speeds on a single day of a
survey. The surface sound speed varies from approximately 1460 to 1500 ms-1 over
the duration of the survey.
(Cartwright, 2003).
The data is taken in Fraser River Delta, Canada
19
Figure 2.3 Example of sound speed profiles and its diurnal variation (Cartwright,
2003)
2.2.3.1 Sound Speed Layers in the Oceans
The surface layer extends from the surface to perhaps 150m, and it is this
layer that mostly effected by local weather conditions and even the time of the day.
Even in calm waters the top 10m or so exhibits a changing sound speed changes
during the cause of a day. The surface acquires heat from the sun, resulting in
negative temperature gradient. This is resulting a negative sound speed gradient by
late afternoon. During the night there is some mixing action caused by normal
wave activity as well as heat lost by radiation from the surface. These effects cause
the negative temperature gradient to weaken considerably or possibly disappear
completely. In stormy weather, there is a strong mixing action in this layer that
ends to reduce the temperature gradient to zero. The result is a positive sound
speed gradient. Once thoroughly mixed, the surface layer may retain the isothermal
condition for an appreciable time period following the storm (Urick, 1983; Burdic,
1991).
20
Below the surface layer, the next layer is called seasonal thermocline,
extending approximately up to 300m. Here the water temperature is affected less
by transient effects such as storms or the day-night cycle. But still, there are
significant changes with seasons. Normally this layer has a negative temperature
gradient (Figure 2.4).
The third layer “main thermocline” has more stable temperature verses
depth characteristic with a negative gradient. This layer extends up to 1000m. The
last layer is called the deep isothermal layer, because of its nearly uniform
temperature (Figure 2.5). Here, the sound speed increases gradually with depth
(Urick, 1983; Burdic, 1991).
Figure 2.4 Oceanic water layers and example deep sea SVP (Jacops, 2002)
21
Figure 2.5 Typical temperature (left) and salinity (right) variations as a function of
depth (Jacops, 2002)
In coastal and shallow water areas (continental shelf), the water column
consists entirely of the surface layer. In some areas where there is fresh water
influx like river mouths, salinity becomes much more variable in addition to
temperature. In addition, the influence of the tides interacting with the shoreline and
the sea floor, in combination with wind forces, results in mixing along shore
currents, and upwelling of water bodies. As in the open ocean evaporation and
precipitation play a role in the variability of the surface salinity while solar heating
will vary the surface temperature on a daily scale. However in the case of coastal
waters these factors represent a much larger percentage of the entire water (Figure
2.6). Therefore, it is irregular and unpredictable. Temperature has long been
considered the dominant cause of change in sound speed throughout the world’s
oceans, with salinity as a secondary source (Burdic, 1991; Cartwright, 2003).
22
SOLAR HEATING
PERSIPITATION
EVAPORATION
TIDAL MIXING
FRESH WATER
RUNOFF
WIND INDUCED
WAVE MIXING
ESTUARINE
CIRCULATION
LONGSHORE
CURRENTS
UP-WELLING
Figure 2.6 The complexity of the oceanography of coastal water masses. Many
external force mechanisms influence the velocity structure (Cartwright, 2003)
2.3 Equation for Speed of Sound in the Water
Over the past years, researches have carried out many studies in
measurements of the sound speeds in water. In a particular case, changes in the
sound velocity ‘C’ caused separately by changing salinity ‘S’, temperature ‘T’ and
pressure ‘P’ are highly dependent on the absolute values of all these three
parameters simultaneously. A theoretical solution to this complicated relationship
C (T,S,P) has not yet been found in the form of an analytical function (Dera, 1991).
However, experimental work carried by many researches has provided a
number of formulae that establish this relationship with sufficient accuracy. These
equations are developed by very accurate measurements of sound speed combined
with associated measurement of temperature, pressure and salinity.
These
measurements and their associated equations have been refined over the years by
researchers like Mackenzie, Coppens, Chen and Millero, Wong and Zhu, and
Grosso (NPL, 2000). Each equation has its own range of temperature, pressure and
salinity for which they are considered valid.
sound speed in ocean waters are as follows.
The three renowned equations for
23
The most easily utilised equation is Mackenzie Equation. It is a nine-termed
equation (Mackenzie, 1981).
C ( D, S , T ) = 1448.96 + 4.591T − 5.304 ×10−2 T 2 + 2.374 × 10−4 T 3 + 1.340( S − 35) +
1.630 × 10−2 D + 1.675 × 10−7 D 2 − 1.025 × 10−2 T ( S − 35) − 7.139 × 10−13 TD 3
(2.3)
T = temperature in degrees Celsius
S = salinity in parts per thousand
D = depth in meters
This equation is valid for the temperature ranging from –20 to 300 C, salinity
ranging from 30 to 40 ppt, and depth ranging from 0 to 8000m.
The Coppen’s Equation is as follows;
C ( D, S , T ) = C (0, S , t ) + (16.23 + 0.213t ) D + (0.213 − 0.1t ) D 2 +
[0.016 + 0.0002( S − 35)]( S − 35)tD
(2.4)
C (0, S , t ) = 1449.05 + 45.7t − 5.21t 2 + 0.23t 3 + (1.333 − 0.126t + 0.009t 2 )( S − 35)
(2.5)
t = T/10 where T = temperature in degrees Celsius
S = salinity in parts per thousand
D = depth in kilometers
The range of validity of this equation is, temperature ranging from 0 to 35
°C, salinity ranging from 0 to 45 ppt and depth ranging from 0 to 4000m (Coppens,
1981). These equations use temperature, salinity and depth. The use of depth
rather than pressure introduces a small error that is accounted for in other, more
accurate equations.
24
Two equations that are most accepted by the scientific community are, the
Chen-Millero and the Del Grosso’s Equation (NPL, 2000). Both equations use
pressure rather than depth for increased accuracy. The Chen and Millero equation
has a the wider range of validity, based on the oceanographic measurements from
which it is derived temperature of 0 to 40 °C, salinity 0 to 40 ppt and pressure of 0
to 1000 bar (NPL, 2000). Currently, the Chen and Millero equation has being
accepted by the United Nations Educational Scientific and Cultural Organization
(UNESCO) as their standard equation for sound speed measurements (NPL, 2000).
C ( S , T , P) = Cw (T , P ) + A(T , P) S + B (T , P) S 3/ 2 + D(T , P) S 2
(2.6)
Cw (T , P) = (C00 + C01T + C02T 2 + C03T 3 + C04T 4 + C05T 5 ) +
(C10 + C11T + C12T 2 + C13T 3 + C14T 4 ) P + (C20 + C21T + C22T 2 + C23T 3 + C24T 4 ) P 2 +
(C 30 + C 31T + C 32T 2 ) P 3
(2.7)
A(T , P) = ( A00 + A01T + A02T 2 + A03T 3 + A04T 4 ) +
( A10 + A11T + A12T 2 + A13T 3 + A14T 4 ) P + ( A20 + A21T + A22T 2 + A23T 3 ) P 2 +
( A30 + A31T + A32T 2 ) P 3
B (T , P) = B00 + B01T + ( B10 + B11T ) P
D(T , P) = D00 + D10 P
T = temperature in degrees Celsius
S = salinity in parts per thousand
P = pressure in bar
The coefficients of the UNESCO equation are shown in Table 2.1.
(2.8)
(2.9)
(2.10)
25
Table 2.1 Table of coefficients (UNESCO Equation)
Coefficient
Numerical values
Coefficient
Numerical values
C00
1402.388
A02
7.166E-6
C01
5.03830
A03
2.008E-6
C02
-5.81090E-2
A04
-3.21E-8
C03
3.3432E-4
A10
9.4742E-5
C04
-1.47797E-6
A11
-1.2583E-5
C05
3.1419E-9
A12
-6.4928E-8
C10
0.153563
A13
1.0515E-8
C11
6.8999E-4
A14
-2.0142E-10
C12
-8.1829E-6
A20
-3.9064E-7
C13
1.3632E-7
A21
9.1061E-9
C14
-6.1260E-10
A22
-1.6009E-10
C20
3.1260E-5
A23
7.994E-12
C21
-1.7111E-6
A30
1.100E-10
C22
2.5986E-8
A31
6.651E-12
C23
-2.5353E-10
A32
-3.391E-13
C24
1.0415E-12
B00
-1.922E-2
C30
-9.7729E-9
B01
-4.42E-5
C31
3.8513E-10
B10
7.3637E-5
C32
-2.3654E-12
B11
1.7950E-7
A00
1.389
D00
1.727E-3
A01
-1.262E-2
D10
-7.9836E-6
26
The Del Grosso’s equation is given as:
C ( S , T , P) = C000 + ΔCT + ΔCS + ΔCP + ΔCSTP
(2.11)
ΔCT (T ) = CT 1T + CT 2T 2 + CT 3T 3
(2.12)
ΔCS ( S ) = CS 1S + CS 2 S 2
(2.13)
ΔCP ( P) = CP1 P + CP 2 P 2 + CP 3 P 3
(2.14)
ΔCSTP ( S , T , P) = CTPTP + CT 3 PT 3 P + CTP 2TP 2 + CT 2 P 2T 2 P 2 + CTP 3TP 3 + CST ST +
CST 2 ST 2 + +CSTP STP + CS 2TP S 2TP + CS 2 P 2 S 2 P 2
(2.15)
T = temperature in degrees Celsius
S = salinity in parts per thousand
P = pressure in kg/cm2
The coefficients of the Del Grosso’s equation are given in Table 2.2.
The range of validity of the equation is: the temperature from 0 to 30°C,
salinity from 30 to 40 ppt, and pressure from 0 to 1000 kgcm-2. This is considered
as an alternative equation for the UNESCO equation, which has more restricted
range of validity (NPL, 2000).
27
Table 2.2 Table of coefficients (Grosso’s Equation)
Coefficients
Numerical Values
Coefficients
Numerical Values
C000
1402.392
CTP
-0.1275936E-1
CT 1
0.5012285E1
CT 2 P 2
0.2656174E-7
CT 2
-0.551184E-1
CTP 2
-0.1593895E-5
CT 3
0.221649E-3
CTP 3
0.5222483E-9
CS 1
0.1329530E1
CT 3 P
-0.4383615E-6
CS 2
0.1288598E-3
CS 2 P 2
-0.1616745E-8
CP1
0.1560592
CST 2
0.9688441E-4
CP 2
0.2449993E-4
CS 2TP
0.4857614E-5
CP 3
-0.8833959E-8
CSTP
-0.3406824E-3
CST
-0.1275936E-1
2.4 Multibeam Echosounder Systems
2.4.1 Introduction
Hydrographic surveying has evolved with increasing capabilities in realtime computing and in data storage. Single beam echosounder has been replaced by
high resolution swath mapping systems.
MBES is one of these high-density
mapping tools, which uses sound waves as remote sensing tool for the
measurements (USACE, 2004).
The development of deep-water swath systems began in the 1970's. These
systems, which permit effective and accurate bathymetric surveys over extensive
areas, can also be used for other oceanographic applications such as geological
mapping and other scientific investigations, Exclusive Economic Zone (EEZ)
surveys and surveying for cable laying. Shallow water swath MBES have devolved
28
rapidly during the 1990's and they are being increasingly used for shallow water
surveys, such as harbor and narrow waterway surveys where 100% coverage and a
high accuracy are required. Adoption of the more strict 1998 IHO Standards for
hydrographic surveys has further accelerated the use of MBES systems for shallow
water applications (Jong et al., 2002).
2.4.2 Principle of MBES Operation
MBES systems measure a series of depths in an across-ship track direction
simultaneously. Each system is composed of a transducer, a transceiver and a
processing unit. The transducer generates a fan of beams (each equivalent to that of
a narrow single beam) that are sent towards the sea floor (Figure 2.7). The same
transducer receives the reflected echo coming from the collision of these beams
with the bottom. The transceiver generates the signals sent to the transducer and
gathers the signals received by the same transducer. The processing unit computes
the depth and position of each sounding (bottom detection) using the two-way
travel time of the wave and the beam angle.
But the final result will be an
integration solution with external data such as position, orientation and heading of
the ship and the tidal measurements (Jong et al., 2002).
29
Vessel Heading
Beam Footprint
Figure 2.7 MBES beam footprint and swath coverage
2.4.3 Transducer
Transducers are made of piezoelectric materials, which have the capability
to convert electric energy into mechanical energy (vibration) and vice versa. When
an electric current is applied to the transducer, it transmit an acoustic pulse into
water and when the echo is returned, the same transducer can convert the
mechanical stress (compress) into electric charge (Coates, 1990; Ingham and
Abbott, 1992).
30
2.4.4 Transducer Arrays
The transducer is constructed in a way to produce the sound beam in a
particular shape, which is called beam forming (Urick, 1983). The beams are in a
spherical form, so that the acoustic pulse travels out equally in all directions (Figure
2.8(a)). This is the case of an ideal point source. But, through careful design, a
transducer can be constructed in such a way to restrict its sensitivity into a
particular angular sector and direction. Transducer array being developed by using
this technique, having a string of transducer elements combined into a single
rectangular array.
Figure 2.8 schematically explains the beam forming in
transducer arrays.
Maximum response
axis defined by a cone
around the long axis of
the transmit array
(a). Transmit beam pattern
Maximum response axis
defined by the
intersection of the two
cones
(c). TR-RC product beam pattern
Maximum response
axis defined by a cone
around the long axis of
the receive array
(b). Receive beam pattern
Resultant beam aligned
along the intersection
of the two cones
(d). Final beam vector
Figure 2.8 Beam forming in flat transducer arrays (OMG-UNB)
31
By combining transducer elements in various configurations, various types
of multibeam sonar arrays being designed. The flat array MBES and the curved
array MBES are the main transducer configurations (RESON Inc, 2005).
Depending on the configuration, the beam forming is accomplished in different
manners.
2.4.4.1 Flat Array Transducers
Normally, flat array transducers are composed of two rectangular transducer
arrays arranged in different manner. Mill’s cross is a typical flat array transducer
configuration. Here the two rectangular transducer arrays arranged in orthogonal.
The transmission array is narrow in across track direction and wide in the along
track (Urick, 1983). This has the effect of a transmitted beam that is narrow in the
along track and wide in the across track. The receiver array is narrow in the along
track and long in across track. Therefore the receive array is “listening” only inside
this narrow across track beam. Once in operation, the entire transmit array
simultaneously transmitting and the entire receiver array simultaneously receiving
and a narrow resultant beam would be formed directly beneath the transducer
(Figure 2.9). The two arrays can be arranged to form a ‘T’ shape, as in the RESON
SeaBat 8124, or they can be mounted in tilted shape (Figure 2.10); however the
same operating principle is applied (RESON Inc, 2005).
32
Figure 2.9 Beam footprints resulting from the intersection of transmission and
reception in RESON SeaBat 8124 MBES (OMG-UNB)
RESON SeaBat 8124
EM300D
Figure 2.10 Example for flat transducer arrays
33
2.4.4.2 Curved Array Transducers
The curved array uses multiple line array staves that are aligned along track
of the ship and arranged in an upward curving arc. Each stave is composed of a
number of elements as illustrated in Figure 2.11.
Vessel Heading
Individual Transducer Staves
Figure 2.11 Curved or Barrel type transducer array (OMG-UNB).
In transmitting, some or all of this transducer staves transmit to make a
wider across track beam. One advantage of this type of transmission is that, while a
single element is capable of forming the narrow acoustic beam, the combination of
multiple elements enables more power to be transmitted within the same narrow
beam (in the along-track). A second advantage is that, in contrast to the Mill’s
Cross configuration, the transmitting staves in the curved array are relatively long in
the along-track direction, resulting in a narrow along-track beam. In order to make
a narrow beam on receive; a number of staves are selected such that their addition
makes an array with enough across track length to make a narrow across track beam
(RESON Inc, 2005). When this is combined with the narrow along-track beam, the
product is relatively narrow in both the along and across track directions. In order
to account for the slight curvature of the arrangements of the staves, slight time
delays are added to the outer staves. The larger the number of staves used (the
longer the effective receive array length) the narrower the receive beam. If one was
only concerned with the beam that is directly tangent to the base of the curved
array, it is possible to receive on all elements (with time delays to account for the
curvature) that would result in a very narrow beam in the across-track. An example
34
of this type of transducer is the Simrad EM1002 (Figure 2.12). This type is often
referred to as a “barrel” arrays (Beaudoin et al., 2004).
RESON SeaBat 9001
Simrad EM1002
Figure 2.12 Typical examples of curved transducer arrays (RESON Inc and Simrad
AS)
2.5 Beam Steering in MBES
Beam steering is the process that enables a beam to be received from a
desired angle, which is oblique to the transducer array (Burdic, 1991). The two
principle methods of beam steering are physical and electronic beam steering. The
methods of beam steering vary, depending on the transducer configuration. The
beam steering methods used in the flat array and the curved array is explained
bellow in detail.
2.6 Beam Steering in Flat Arrays
There are two types of beam steering adopted in the flat array transducers.
Those are the mechanical and the electronic beam steering techniques.
35
2.6.1 Mechanical Beam Steering
This is the simplest method to steer beams in a flat array transducer. Here it
is required to mechanically (physically) move the entire transducer through the
range of angles desired (Urick, 1983) or build the transducer composed of multiple
transducers, each individual transducer pointing in the desired direction (Caruthers,
1977). The limitation of this approach is the compromise between the swath width
and transducer size (USACE, 2004).
2.6.2 Electronic Beam Steering
Electronic steering enables the formation of a complete array of beams with
every transmit-receive cycle of the transducer. Electronic beam steering methods
take advantage of the fact that transducers are not one single element but are
composed of many individual elements that can be controlled and monitored
individually.
Electronic steering is accomplished by digitizing the signal and
computing beams at the desired angles and this is controlled only by the electronics
and the algorithms used (USACE, 2004).
These algorithms are based on the
wavelength of the acoustic wave, the frequency of the transducer and the spacing
between individual transducer elements. For a flat array transducer the two primary
methods of beam steering are time or phase delay and Fast Fourier Transformation
method (USACE, 2004; RESON Inc, 2005).
Without beam steering, all the return echoes from the seabed will be parallel
to the flat transducer array (Figure 2.13(a)). To receive oblique beams, delays are
being applied accordingly to each transducer element; more delays to the direction
of the beam and less delay to the other side (Figure 2.13(b)). The delays are applied
to all beams except for nadir (Figure 2.13(c)). By this way all the beams in the
entire swath are simultaneously formed (Figure 2.13(d)).
36
Delay
Transducer Array
Extra path
Travelled
25 degrees starboard
(a)
(b)
-45 degrees
25 degrees
Applied
delays
Transducer Array
--10 degrees
- 45 degrees
25 degrees
- 10 degrees
(c)
(d)
Figure 2.13 Electronic Beam Steering (a) No delays - Parallel beams (b) Applied
delays to each element to achieve desired steering direction (c) Applied delays for
different beams from both side of the transducer (d) How all the beams in the
swath are generated simultaneously by beam steering (OMG-UNB)
Figure 2.14 shows how graduated delays introduce to each individual
transducer elements to virtually steer the array. In the actual transducer system, all
the received waveforms are digitized and placed in a buffer where it is possible to
simultaneously calculate all the required angles to result in many narrow receive
beams (Cartwright, 2003).
37
Figure 2.14 Applied delays to individual transducer elements to detect oblique
beams (Kammere and Clarke, 2000)
2.6.2.1 Time Delay Method
The time delay method introduces graduated time delays at each individual
element to virtually “steer” the array. Equation 2.16 defines the relationship of the
acoustic and physical parameters that need to be considered in determining the time
delays to be applied.
Time delay at nth element =
n×d
× Sinθ
f ×λ
Where θ = angle steered
λ = Wavelength
d = element spacing
f = frequency
(2.16)
38
2.6.2.2 Phase Delay Method
Phase delay method is similar in concept to time delay, however rather than
time delays phase shifts are added to each element before they are summed. After
adding the phase shifts, the desired steered beam will result in all of the elements
receiving the wave fronts at the same time, or in phase. While this is a similar
method to time delay, the steering directions are not limited by the sampling
frequency, however the number of beams that can be produced is limited by the
number of individually monitored staves (Cartwright, 2003).
According to Burdic (1991) the phase shift Δψ s for the nth element is;
Δψ s =
2π nd sin θ s
λ
(2.17)
Where Δψ s = phase shift
n = number of elements
d = element spacing
θ s = angle steered
λ = wavelength
2.6.2.3 Fast Fourier Transformation Method
Unlike the phase and time delay methods in which the angle is assumed and
the time to bottom detection is sought, this beam forming technique consists of the
determination of the angle of arrival of the echoes, the time of arrival is assumed to
be known. Fourier transform is a mathematical method of breaking up a signal in to
its set of sine and cosine components (Coates, 1990). This method dramatically
speeds up the computation process (Cartwright, 2003). The angular spacing is
given by;
39
⎧λ n ⎫
θ = sin −1 ⎨ × ⎬
⎩d
Where
(2.18)
N⎭
θ = angle steered
λ = wavelength
d = element spacing
n = element number
N = number of elements
2.7 Beam Steering in Curved Arrays
In curved arrays, the beam steering is done by taking advantage of the
physical shape of the transducer combined with an appropriate selection of
transducer elements. The steered beam will be orthogonal to the tangent of the
curve as shown in the Figure 2.15 below. Here, the beam widths are consistent
throughout the across track angular range of the transducer (Cartwright, 2003).
Virtually no beam steering is required.
θ
Selected staves for
steering angle θ
Individual
transducer staves
Figure 2.15 Stave selection for beam steering in curved transducer array
(Cartwright, 2003)
40
To generate the narrow beam, numbers of staves are selected in the local
neighbourhood and a weighting function is added to control the sidelobes (Figure
2.16).
But there are two practical limitations for this case, when one moved
towards the edge of the transducer arc for the outer most beams. Here, there are no
neighbouring elements on one side (towards outside) for the weighting function.
And to increase the number of outer beams (after a certain limit), one has to
compensate for the amount of curvature in the selected portion of the array.
Therefore minor delays being applied to the outer staves (Figure 2.17).
Weighting
Function
Extreme
beams need
to be steered
No Steering
Figure 2.16 Weights are added to the neighbourhood and outermost beams has to
be steered (Kammere and Clarke, 2000)
θ
Electronic
Steering
Figure 2.17 Outer beams steered using the physical shape of the transducer
combined with electronic steering (Cartwright, 2003)
41
2.8 Ray Tracing
2.8.1 Introduction
Once the beam pointing angle is known in beam steering, the next step is to
convert the measured travel time of each beam into depth and position. For this
case, the sound speed of each layer where the ray was being crossed must be
known. By knowing the sound speed cross-section (SVP) of that water column,
then one can trace it back where it came (Figure 2.18). Here range and across track
are calculated for each sound speed layer until it reach the half of the two way
travel time.
v1
R
v2
2
h1
θ1
R1
θ2
X1
h2
X2
TWTT / 2
vn
Rn
θn
hn
Xn
Figure 2.18 Ray tracing in MBES
During ray tracing, ocean can be described as a horizontally layered medium
(homogeneous) in terms of sound speed (Burdic, 1991). This means that vertical
beam of the MBES transducer array is orthogonal to the layers and all other beams
are oblique with respect to the layers. Therefore ray tracing can be discussed
further in each case as follows.
42
2.8.2 Vertical Incidence
Each layer of the water has its own local speed of sound. When the nadir
beam of a MBES is emitted vertically from the transducer, it travels through the
medium at each local sound speed, which varies with the layer. Therefore the
harmonic mean speed of sound ( μ ) can be used to get the respective value of sound
speed of the water column, which is the ratio of the total distance traveled by the
total time of travel. It can be expressed mathematically as follows:
z − z0
μ=
N Zi +1
∑∫
i =1 Zi
(2.19)
dz
ci ( z )
Where ( z − z0 ) is the total distance traveled, ( zi , zi +1 ) is the layer traveled at
the sound speed ci ( z ) and N is the number of speed layers (Cartwright, 2003). The
final depth is travel time times this harmonic mean speed.
2.8.3 Oblique Incidence
For the oblique beams, one has to consider two parameters.
First the
distance traveled through the water column, then the deviation of the actual travel
path due to the refraction of the beam through each different sound speed layer.
This can be achieved by applying Snell’s law (Figure 2.19). It states that the ratio
of the sine of the angle of incidence of the ray through a layer over the sound speed
in the layer remains constant as the ray transits through another layer with different
sound speed. It is illustrated as follows;
43
θ1
Sound Speed Layer 1
θ2
Sound Speed Layer 2
Figure 2.19 Illustration of oblique incidence
sin θ1
sin θ 2
=
= Snell’s constant ( ρ )
sound speed 1 sound speed 2
(2.20)
To calculate the final bathymetry, one has to find the horizontal and vertical
components of the ray-path through each layer for the entire water column. By
knowing the starting depth, departure angle, two way travel time and the sound
velocity profile; one can trace it back. To achieve this, the Snell’s constant has to
be determined in the first place.
ρ=
sin θ 0
C0
(2.21)
Where: ρ = Snell’s constant
θ 0 = Departure angle
C0 = Sound speed at transducer
Then, one has to calculate the horizontal and vertical components of the raypath till it reaches the bottom (end of the travel time). But the actual sound velocity
profile is more complex one. Therefore, it has to be generalized into a simpler
format. There are two standard approaches for it. That is by subdividing the water
column into layers of constant sound speed or using constant sound speed gradient
(Figure 2.20) (Burdic, 1991; Urick, 1983).
44
True sound
speed profile
in the water
Layers with
constant
sound speed
Layers with
constant
gradient
sound speed
Figure 2.20 Modelling the sound speed profile in the water
2.8.3.1 Layers with Constant Sound Speed
This is a much simple and straightforward method. Here the sound speed in
each layer is considered as constant. Final across tracks and depths are calculated
by applying Snell’s law at each layer and summing up all the horizontal deflections
(in distance) and individual travel times until the halfway of two-way travel time
(Burdic, 1991; Urick, 1983).
45
N
Δi
i =1
ci 1 − (ci ρ ) 2
t=∑
ci ρΔ i
N
x=∑
i =1
1 − (ci ρ ) 2
(2.22)
(2.23)
Where: t = time of traveled
x = horizontal distance traveled
N = number of layers
ρ = Snell’s constant
ci = sound speed in layer
Δ i = layer thickness
2.8.3.2 Layers with Constant Sound Speed Gradient
Here, each speed layer is considered having a constant sound speed gradient.
In this case the ray travels in a curved path than that of a straight-line path in
constant speed layer case. Therefore this method can be considered equivalent to
fitting a smooth curve with constant radius to the ray path to each layer (Figure
2.21) having a constant sound speed gradient (Burdic, 1991; Urick, 1983). This
method considered being very close to the actual ray path.
46
Ray path with Constant Sound
Speed Gradient
xi
Start of Layer
Ci
Ri
Ci +1
θi
Δi
End of Layer
θi +1
Figure 2.21 Ray path in a constant sound speed gradient layer
Where:
Ri = radius of curvature at layer
ci = sound speed at start of later
ci +1 = sound speed at end of layer
Δ i = layer thickness
xi = horizontal distance
θi = ray angle at star of layer
θi +1 = ray angle at end of layer
The radius ( Ri ) can be calculated from the following equations.
1
Ri = −
ρ gi
gi =
ci +1 − ci
Δi
Where: Ri = radius of curvature at layer
ρ = Snell’s constant
gi = sound speed gradient of layer
ci = sound speed at the start of later
ci +1 = sound speed at the end of layer
Δ i = layer thickness
(2.24)
(2.25)
47
The horizontal distance in each layer can be calculated based on radius and
Snell’s constant ρ ;
xi = Ri (cos θ i +1 − cos θi ) =
cos θi − cos θi +1
ρ gi
(2.26)
Where: xi = horizontal distance
θi = ray angle at the start of layer
θi +1 = ray angle at the end of layer
ρ = Snell’s constant
gi = gradient in layer
The travel time for each layer can be calculated by using harmonic sound speed;
ti =
⎡c ⎤
Ri (θi − θ i +1 ) θi +1 − θ i
Ln ⎢ i +1 ⎥
=
2
ρ gi Δ i
cH i
⎣ ci ⎦
Where: ti = time in layer
θi = ray angle at the start of layer
θi +1 = ray angle at the end of layer
ρ = Snell’s constant
gi = gradient in the layer
ci = sound speed at the start of layer
ci +1 = sound speed at the end of layer
cHi = harmonic sound speed to end of layer
Δ i = layer thickness
(2.27)
48
But in practice, we only have two-way-travel-time, Snell’s constant, sound
speed at the start and the end of each layer and thickness of each layer. Then each
component can be compute using the following formulas;
ti =
xi =
a sin ⎣⎡ ρ ( ci + gi Δ i ) ⎦⎤ − a sin [ ρ ci ]
ρ g Δi
2
i
⎡ gΔ ⎤
Ln ⎢1 + i i ⎥
ci ⎦
⎣
1 − ( ρ ci ) 2 − 1 − {ρ (ci + gi Δ i )}2
ρ gi
(2.28)
(2.29)
Where: ti = time in layer
xi = horizontal distance
ρ = Snell’s constant
gi = gradient in layer
ci = sound speed in start layer
ci +1 = sound speed at end of layer
Δ i = layer thickness
This has to be calculated for each layer of water column and by summing up
each together one can get the depth and total across track for each beam.
2.9 Sound Speed Measurements in MBES
It is clear that sound speed is an important factor of the accuracy of the
MBES measurements. Depending on the transducer configuration; two types of
sound speed measurements are needed. That is the surface sound speed and the
sound velocity profile through the water column (Dinn et al., 1996; Schmidt et al.,
2006).
49
2.9.1 Surface Sound Speed (SSS)
The surface sound speed (SSS) means the speed of sound at the face of the
transducer. MBES uses SSS in the process of electronic beam steering.
There are different methods of getting surface sound speed:
1.
By direct and real-time SSS measurements, using a sound velocity
probe near the face of the transducer.
2.
By real-time temperature measurements near the transducer face,
assuming the salinity value.
3.
By getting the SSS at the transducers depth from the frequent or less
frequent SVP measurements.
4.
By getting the value at the transducer depth from the frequent or less
frequent temperature and salinity profile measurements.
5.
By applying a constant value for SSS.
2.9.2 Sound Velocity Profile (SVP)
The SVP gives a representation of the change of sound speed through the
water column. All bathymetric sonar systems calculate water depths by measuring
the time it takes a sound pulse to travel to the bottom and back to the receiver. To
translate these time measurements in to depth and distance, one must know the
speed of the echo traveled through the water and the traveled direction (ray tracing).
As discussed at the beginning of this chapter, the velocity structure of the water
varies much. The difference in sound speed across the water column acts as a lens
bending the path that sound travels. For these reasons, it is a must to have accurate
SVPs for any data set (Schmidt et al., 2006).
50
The followings are the most popular SVP measurement techniques:
1.
By direct SVP measurements. In this case the vessel usually has to
stop and a probe has to lower down.
2.
By some temperature and salinity profile measurements. In this case
the vessel has to be still.
3.
By using a database value.
4.
By using a constant value for SVP.
2.10 Errors in Multibeam Systems
2.10.1 Introduction
As stated earlier in Section 2.4.2, all bathymetric measurements in MBES
are based on two-way-travel-time and beam pointing angle. These measurements
are made with respect to the frame of reference of the ship on which the transducer
is mounted (Dinn et al., 1995; Ingham and Abbott, 1992). A number of parameters,
many of which are measured at the time of the ping, are required to transform the
measured slant range travel times and angles into accurate georeferenced depths (x,
y, z or latitude, longitude, depth).
These parameters include;
•
The position of the ship or more precisely, the position of some points
on the ship, e.g. the GPS antenna.
•
The pitch, roll and heading angles of the ship relative to the above
position.
•
The vertical position of the transducer with respect to the average
water level, i.e., the ship's draft and heave at the transducer.
•
The changes in water level due to tides and atmospheric effects; the
profile of sound velocity vs. depth.
•
The vector distance (x, y, z) in the ship frame of reference between the
transducer and the positioning sensor (offset measurements).
•
Mounting offsets of the sonar head.
51
2.10.2 What are the Largest Errors?
By virtue of their error propagation characteristics, roll angle, sound speed
measurements and vertical control are the more significant parameters affecting
depth accuracy (Dinn et al., 1995). But with the latest system improvements the
single biggest limitation on the quality of bathymetric data is the refraction of the
sound wave in the water column (Gardner et al., 2001).
2.10.3 Does Our Sound Speed Measurements Adequate Enough?
In most MBES systems SSS is continuously being measured. But SVP is
known only at discrete time. There is no rule that will tell the surveyor when and
where to take a SVP. It depends on ones own experience. The surveyor should be
aware of the factors that influence the sound velocity. If surveying in a small area
with little variability, possibly only one cast per survey day is required. If the
surveyor finds a large variation of sound velocities, within the survey area, it would
be wise to take this into account when designing the layout for the survey. One
should avoid running a long line that would have two or more applicable sound
velocities. It would be advisable to seek an alternative layout where each line
would have only one applicable sound velocity profile to be applied. Even that
sometimes it is critical due to the high variability of ocean waters. Clarke (2002)
has shown how water masses changes rapidly both in time and space, using crosssection of sound speeds in Georges Bank, Boston. Here, even during a single survey
line the SVP changes rapidly (Figure 2.22). This clearly illustrates the
inadequateness of conventional sound speed measurements in MBES surveys.
52
Colour range 10m/s, 45 km long profile from 0-100m
Figure 2.22 Cross section of the sound speed structure on the edge of Georges
Bank (Clarke, 2002)
2.10.4 Refraction in Multibeam Echosounders
2.10.4.1 Introduction
The refraction effects the bathymetric measurements in MBES at two places
(depending on the transducer shape); at the face of the transducer during the beam
steering and during the progression (ray tracing) of the sound wave through the
water column (Figure 2.23).
Vessel
Positional error
Water Surface
Beam Steering
angle error
Ray path with the
Measured SVP
Ray path with the
True SVP
Depth error
Figure 2.23 Refraction effects in each phase of the MBES
53
2.10.4.2 Effects During the Beam Steering
Multibeam sonars use beam steering both for transmission as well as
reception of the sonar pulse. Moreover, beam steering is done, not only in the
athwartship direction to create the swath, but fore and aft to increase measurement
resolution. SSS is very important in MB sonars because any changes must be
accounted and corrected in the first place. Because MB sonar systems do not save
beam angle from individual hydrophones, so one cannot go back and apply
corrected SSS to beam forming calculations. It is not recoverable (Schmidt et al.,
2006).
SSS directly affects the directivity of the beams produced by the sonar. Any
defects will results that the sonar is not exactly looking in the direction one would
expect. As discussed in section 2.6.2, all electronic beam steering methods in flat
MBES depend on the wavelength (Equations 2.16, 2.17, 2.18). And this λ is
varying with the speed of sound. This can be further explained schematically using
the Figure 2.24. In this case we assume that true sound speed is C0 and the
measured is C1 at the face of the transducer ( C0 > C1 ).
Here, the estimated
wavelength based on the measured SSS is shorter than the true wavelength (which
is failed to measure). This wrongly measured speed ( C1 ) adds more delays to the
elements than the true speed case and resulted in incorrect beam angle.
54
Water Surface
Water Surface
Virtual Array
Virtual Array
Delay
Delay
Δθ
Incorrect
Beam Angle
True Beam
Angle
True Beam
Angle
(a)
(b)
Figure 2.24 Effect of change in SSS in beam pointing angle in a flat array
transducer: in the case of true SSS is greater than the measured (a) Virtual array
facing correct direction with true SSS. (b) Virtual array pointing towards wrong
direction due to incorrect SSS
2.10.4.3 Effects Through the Water Column
If the measured SVP is not the true one or it differs much from the actual
SVP, the refraction calculations are not correct. The refraction angle, the estimated
travel time and the across track calculations for each speed layer is not true
anymore (Section 2.8, Equations 2.19 to 2.29). Therefore the estimated ray path is
incorrect with the true path causing depth and positional errors in the final
bathymetry.
55
2.11 Summary
James et al. (2001) said that if there is strong water stratification, it would cause
problems for the beam steering and ray tracing of individual beams in MBES.
According to Batton, 2004, Tonchia and Bisquay, 1996 and Dinn et al., 1995, any
failure to take into account of these sound speed changes in the water column can
result in significant errors in MBES bathymetric measurements.
The errors
associated with refraction in the water column are larger, especially with wider
sector multibeam sonars. These errors depend on following factors;
1. The shape of the transducer array.
2. Type of the beam steering being used.
3. The mounting angle of the sonar array.
4. The actual sound velocity near the transducer face.
5. The actual SVP.
CHAPTER 3
FIELD DATA COLLECTION
3.1 Introduction
This chapter discuss the methodology used to collect data to find the best
approach in determination of the Snell’s refraction coefficient for ray tracing
purpose. Firstly, a brief description about each of the survey instrumentation used
for the data collection is presented. Then the raw data extraction from each sensor
is presented.
3.2 Survey Instrumentation
3.2.1 The MBES System
The RESON SeaBat 8124 is used as the MBES system. This is a flat Mill’s
T cross type MBES system (Figure 3.1). It is a 200 kHz, 80 beams system; which
generates a 120-degree across track and 1.5 degree along track swath coverage.
Each beam covers a footprint of 1.5 × 1.25 degrees. This system is capable of
measuring depths from 0.5m to 750m, depending on the bottom backscatter strength
and water column attenuation (RESON Inc., 2003).
57
(a)
(c)
(b)
(d)
Figure 3.1 The MBES System (a) The RESON SeaBat 8124 multibeam system
with transducer and 81P processing units (RESON Inc.) (b) The MBES mounted
over the side of the boat (c) The 81P processor inside the boat (d) The MBES data
collecting using QINSy software
3.2.2 The Singlebeam Echosounder (SBES)
The Odom Hydrotrac is used as the SBES (Figure 3.2). Its acoustical
frequency is 210 kHz (Odom Hydrographic Systems Inc., 2005). Usually, SBES
are least affected by the refraction; therefore it used for comparison of the MBES
data in this study.
58
(a).
(b).
(c)
Figure 3.2 The SBES System (a) Odom Hydrotrac SBES (b) The SBES mounted
on starboard side (c) Hydrotrac unit inside the boat.
3.2.3 The Positioning System
The Trimble DSM 212H DGPS receiver is used as the positioning system
(Figure 3.3). The study area is quite closer to the Singapore navigation radio
beacon (298) and therefore the positioning obtained from this unit is well within 1
to 2m and it is good enough for this study.
(a)
Figure 3.3 The DGPS System
(b)
(a) Trimble DSM 212H DGPS unit (Trimble
Navigation Ltd. 2002) (b) DGPS antenna location on top of the boat
59
3.2.4 Sound Speed Measurements
Since the MBES system used is a flat array system and it is intended to
study the affects caused by the each sound speed measurements, it is necessary to
measure both SSS and SVP separately.
3.2.4.1 SSS Measurements
Since the RESON SeaBat 8124 MBES is a flat transducer, it has a surface
sound speed (SSS) measuring probe at the face of the transducer, which measures
real-time surface sound speed at a frequency of about 10 Hz (Figure 3.4.a). To
activate this real-time SSS measurement, there is an option, which has to be
selected during the database set-up in QINSy (Figure 3.4.b).
(a)
(b)
Figure 3.4 The SSS Probe (a) Surface sound velocimetre (mounted just above the
transducer unit) (b) Real-time SSS option in QINSy survey database setting
60
3.2.4.2 SVP Measurements
The SVP-15 velocimeter is used to measure the sound velocity profile
through the water column (Figure 3.5), which uses UNESCO equation to calculate
sound speed at each depth. The depth range for SVP-15 is up to 200m (Navitronic
System AS, 1998). SVP control software was used to read the data from the unit
and then applied this SVP to the QINSy for refraction calculations.
(a)
(b)
(c)
Figure 3.5 The SVP Probe (a) SVP-15 sound velocity profiler with data logger
(b) Launching the probe in the middle of the survey area to get the sound speed
profile (c) SVP Control software reading the probe.
61
3.2.5 Motion (Attitude) Sensor
Due to the wider swath coverage in all MBES surveys, vessel attitude data is
crucial and must be measured and applied to the survey software for the necessary
corrections. For this study, TSS MAHRS motion sensor was used (Figure 3.6). It
has an inbuilt gyrocompass and a motion sensor, which measures heading, heave,
pitch and roll simultaneously.
(a)
(b)
Figure 3.6 MAHRS Attitude Sensor (a) TSS MAHRS unit (VT TSS Ltd., 2003)
(b) MAHRS mounted inside the boat
3.2.6 Tide Gauge
The Valeport 740 pressure sensor tide gauge (Figure 3.7(a)) is set up at the
marine department jetty, Johor at a close vicinity to the study areas (Figure 3.7(b)
and 3.7(c)).
Tidal data are logged every 10 minutes interval in the internal data
logger and later transferred for processed with QINSy >Tidal Manager.
62
(a)
(b)
(c)
Figure 3.7 Tide Gauge (a) Valeport 740 tide gauge unit with data logger and
transducer (Valeport Ltd., 2004) (b) Tide gauge installed at the Marine departments
jetty (c) Tide gauge house and the tide gauge benchmark on the jetty
3.3 Survey Software
The QINSy version 7.5 is used as the data collection software. QINSy is a
modular build programme. Which means that it is not just a single programme, but
a suite of applications linked together (QPS BV., 2004). The entry point for QINSy
is a program called the “Console”. From this console all other programs can be
started.
Figure 3.8 shows the ‘QINSy Console’ and its icons (sub
programs/modules). The top part always has four icons. The ‘Setup’ icon is to
make and configure the database. The ‘Online’ icon is used as the base for raw data
collection. ‘Replay’ icon allows the user to change the database configurations and
63
replay the data. Processing is used to clean and process the data. The bottom half
also contain icons that user can add on, like ‘Line Data Manager’, which can use to
design survey lines.
Figure 3.8 QINSy console
3.4 Survey Platform
A 14m-fibreglass boat “PG19” is hired for the purpose and converted as a
survey launch (Figure 3.9(a)). RESON SeaBat 8124 MBES was mounted portside
and Hydrotarc SBES was mounted in starboard side. TSS MAHRS was set up
closer to the boats centre of gravity (CoG) point. Figure 3.9(b) shows the vessel
configuration diagram.
64
(a)
(b)
Figure 3.9 Survey Platform (a) Survey vessel (PG-19) (b) Sensor locations in
vessel configuration diagram
3.5 Field Data Collection
Field Data was collected at the Lido Beach, Johor, Malaysia during 1st to 4th
of March 2008. The study areas are marked in the chart MAL5128 (Figure 3.10).
Data was collected in two stages. Study area-1 was used to test the inadequate
sound speed measurement effects and study area-2 was to test the refraction
constant determination methods. Study area-1 is a flat, about 10m deep and the
study area-2 is a crater about 45m deep. QINSy software is used as the data
collection software. The project template and database settings are shown in the
Appendix A. Proper calibration procedures were followed for both SBES (Bar
Check) and MBES (Patch Test), prior to the data collection.
65
Study Area 1
Study Area 2
Figure: 3.10 Survey Areas
3.6 Methodology for Determination of Inadequate Sound Speed Measurements
in MBES
As discussed in Chapter 2, the effects are studied both in beam steering
(SSS effects) and in ray tracing (SVP effects).
3.6.1 The Effects of Inadequate SSS on MBES
For this case, simulated and real data were used. Firstly, depths and
positions are simulated for a single ping of RESON SeaBat 8124 MBES for 100m
deep flat synthetic seafloor for different SSS variations having a correct SVPs.
Then, data collected at study area-1 are used to justify the effects.
3.6.1.1 Simulated Data Case for SSS
Here, a constant single layered 1500 ms-1 SVP is assumed as the true SVP
and 1500 ms-1 assumed as the true SSS. 1505 ms-1 and 1510 ms-1 are used as
positive erroneous SSS and 1495 ms-1 and 1490 ms-1 are used as negative erroneous
SSS. Table 3.1 shows the sound speed configuration used in each data set.
66
Table 3.1 Sound speed configurations to determine the SSS effects in the simulated
data case
Data Set
SSS Error
SSS value (ms-1)
SVP
1
No error (True SSS)
1500
Constant 1500 ms-1
2
+10 ms-1
1510
Constant 1500 ms-1
3
+5 ms-1
1505
Constant 1500 ms-1
4
-5 ms-1
1495
Constant 1500 ms-1
5
-10 ms-1
1490
Constant 1500 ms-1
In the 1st data set, all the beams are correct in steering direction because of
the correct SSS. Using this information, true travel times are calculated for each of
the 80 beams of the ping using constant 1500 ms-1 SVP for the synthetic flat seabed.
In other cases (all erroneous SSS), beam-pointing angles are recalculated for the
respective SSS value using the Equation 2.16. Then the final depths and positions
are calculated based on the above true travel time and SVP (data set-1). Then the
beam angles, depths and across track positions are compared in each case.
3.6.1.2 Real Data Case for SSS
Here, the same 50m long survey line is run on a flat area (study area-1) with
different SSS values entered in to the RESON SeaBat sonar processor (Table 3.2).
As the true SVP, the SVP collected at the middle of the site is used and real-time
SSS is used as the correct SSS. Since the area is shallow, a higher SSS difference is
used as erroneous SSS to obtain significant effects. These erroneous SSSs are set
manually in the sonar processor (81p) after switching off the real-time SSS unit
(Figure 3.11).
67
Table 3.2 SSS and SVP configuration to determine the SSS effects in the real data
case
SSS Error
SSS value
SVP
1
No error
Real time SSS
True
2
- Ve error
1450 ms-1 (< real time SSS)
True
3
+ Ve error
1600 ms-1 (>real time SSS)
True
(a)
(b)
Figure 3.11 Altering the SSS value in the sonar processor (a) SSS switch that can
on/off for real time SSS (b) Manual SSS value entered to the Sonar Processor in
the Ocean menu
3.6.2 Determination of Inadequate Sound Velocity Profile (SVP) Effects on
MBES
Here also, both the simulated and the real data are used. First, the depths
and positions are simulated for a synthetic flat 100m deep seabed for different SVPs
and then real data are used to justify the effects as in the above SSS case.
68
3.6.2.1 Simulated Data Case for SVP
Here, the correct SSS is assumed to be 1510 ms-1 in all the cases and for the
SVPs; two-layered synthetic SVPs are used for simplicity in ray tracing
calculations. First layer is 10m deep and second one is 90m deep (Figure 3.12).
Each layer is considered being of constant sound speed.
1500
0
1510
Layer 1
Sound Speed
10m
10
Layer 2
90m
100
Depth
Figure 3.12 Synthetic two-layered SVP
The SSS and SVP configurations are shown in Table 3.3. Since the SSS is
correct, all the beam angle directions are correct. By using this true SVP in data
set-1; true travel times and across track distances are calculated for the synthetic flat
seabed. Then using these travel times, depths and across tracks are calculated for
the faulty SVP data sets and finally the results are compared.
69
Table 3.3 Sound speed configurations to determine the SVP effects in the simulated
data case
Data set
SVP Error
SSS value (ms-1)
SVP (ms-1)
1
No error (True SVP)
1510
1510, 1500
2
+ Ve error
1510
1520, 1500
3
- Ve error
1510
1500, 1500
3.6.2.2 Real Data Case for SVP
For this case, one survey line in both study areas are used. SSSs are
measured real time and as true SVP, the SVP collected at the middle of the line is
used. Then for the erroneous SVPs, SVPs are collected at different time/day at the
same site are used (Figure 3.13).
SVPs
+VE SVP
-VE SVP
TRUE SVP
Sound Speed
1520
0
1525
1530
1535
1540
1545
-5
Depth
-10
-15
-20
-25
Figure 3.13 SVPs used to determine the SVP effects in the real data case
70
To apply the faulty SVPs to the survey lines in the database, same line is
replayed with the respective faulty SVPs using QINSy > Replay Manager. The SSS
and SVP configurations adopted are shown in the Table 3.4. Same ping is selected
in each case to compare the effects. Finally, the shape of the swath is compared
with the respective simulated data sets.
Table 3.4 SSS and different SVP configurations to determine the SVP effects in the
real data case
Set
SVP Error
SSS
SVP
1
No error (True SVP)
True SSS from the unit
True (at the site)
2
+ Ve SVP error
True SSS from the unit
Different time (noon)
3
- Ve SVP error
True SSS from the unit
Different time (after rain)
3.7 Comparison of SSS and SSVP in Determination of Snell’s Refraction
Constant for Refraction Reduction
To compare the effects in each technique in refraction coefficient
determination, two computer programs were developed using MATLAB R2006a. In
the ‘SSS’ program, SSSs are used to calculate the refraction coefficient. To use
surface value of the SVP in refraction coefficient ‘SSVP’ program was developed.
This program development is discussed in the next chapter (Chapter 4) in detail.
3.7.1 Data Collection for Refraction Reduction
For this case, two multibeam survey lines were run in study area-2 at the
same direction, same speed (3-4 knots) with 50% overlap using RESON SeaBat
8124 MBES, so that the nadir beams of the first line and the outermost beams of the
second line are overlapped. Then two single beam echo sounder survey lines were
also run along the same line as MBES lines. A proper MBES and SBES calibration
are performed, before carrying out the survey lines. Real time SSS are collected
71
throughout the survey for the MBES and SVP is collected at the middle of the site
for true refraction calculations (ray tracing).
3.7.2 Raw Data Extraction
Different types of raw data collected from the MBES and various other
sensors are necessary to feed the developed computer programs (SSS and SSVP).
Each data is extracted in different export stages using QINSy software. All the data
extracted as ASCII format, so that they can be easily used in other applications
(Notepad/Excel).
3.7.2.1 MBES Data
The raw data collected from the MBES system is extracted from QINSy >
Replay > Raw Data Manager > Generic Export in each beam vies. Figure 3.14
shows each raw observation item selected for export. Figure 3.15 shows exported
data string in Notepad. In SeaBat 8124 MBES, one-way travel time of each beam is
logged in total number of samples.
Figure 3.14 Selected data items in each MBES beam
72
Figure 3.15 Exported raw MBES data string (travel time, beam angle, SSS, beam
number, ping number and system time)
3.7.2.2 Transducer Position Data
The vessel heading and transducer position (E, N) data are extracted from
QINSy > Processing Manager > Export > User Defined ASCII in each ping vies.
Figure 3.16 shows each of the selected parameters and Figure 3.17 shows sonar
head position and vessel heading in the exported data string.
Figure 3.16 Selected raw data items in transducer positions
73
Figure 3.17 Exported transducer position data string (system time, Northing,
Easting and vessel heading)
3.7.2.3 Vessel Attitude Data
The Pitch, Roll and Heave data from the motion sensor unit are extracted
from QINSy > Replay > Raw data manager > Analyze > Export > ASCII. Figure
3.18 shows the system selection to be exported and Figure 3.19 shows the exported
raw attitude data in Notepad.
74
Figure 3.18 System selection (MRU) in analyse
Figure 3.19 Exported raw attitude data string (system time, pitch, roll and heave)
3.7.2.4 SBES DTM Data
SBES data are least affected by the refraction. Therefore in this study, SBES
data is used for to verify the bathymetry. Because of that, SBES data is no need to
be processed using the developed programs. But to generate the profiles along the
DTM for the final comparison, SBES depth and the position (E, N) are needed. For
75
that purpose, the SBES DTM data are extracted form QINSy > Processing Manager
> Export > User defined ASCII. Figure 3.20 shows the category and the parameter
selection and Figure 3.21 shows the exported ASCII data string in Notepad.
Figure 3.20 Selected data source and parameters in SBES DTM
Figure 3.21 SBES DTM data string (time, Easting, Northing and depth)
CHAPTER 4
COMPUTER PROGRAM DEVELOPMENT
4.1 Introduction
In this chapter, the program flow and the algorithms used to develop the two
computer programs are explained. The only difference between the two (SSS vs
SSVP) programs is the value used to calculate the Snell’s refraction coefficient, the
rest of the bathymetric calculation procedures are same.
4.2 The SSS Program
Here, real-time Surface Sound Speed (SSS) is used to compute the
refraction coefficient for each of the beam at each ping. First of all, all the raw data
are read to MATLAB. The measured travel times of each beam of each ping from
the MBES are in total number of samples. By using the pulse sample frequency
(13125Hz), these are converted back into travel time (TT) in seconds for all the
beams of the selected 200 pings. The program flowchart for this step is shown in
Figure 4.1.
77
Initialisation
i = 1, j = 1
i=i+1
j=j+1
Update calculation
for each beam of each ping
TT (i, j ) = TSAMP(i, j ) /13125
j<=80
i<=200
Resultant Travel Time matrix
TT
Figure 4.1 Conversion of total samples in to travel time
The raw motion sensor data (MR, MP, MH) are in different time intervals
than from the MBES ping time. Therefore MRU data are interpolated with respect
to the MBES ping timing (ER, EP, EH), so that all the vessel attitude data can apply
directly into the MBES data (Figure 4.2).
Initialisation
i =1
i=i+1
Update calculation
for each beam of each ping
ER(i ) = Interp1[ MT (i ), MR(i ), ET (i )]
EH (i ) = Interp1[ MT (i ), MH (i ), ET (i )]
EP(i ) = Interp1[ MT (i ), MP(i ), ET (i )]
i<=200
Resultant Interpolated Roll, Heave
and Pitch matrices
ER, EH , EP
Figure 4.2 Interpolation of roll, heave and pitch with respect to each ping time
78
Then the effective beam angles (EBA) are calculated for each beam to
correct for the vessel roll (ER) and the MBES mounting angle (R) from the patch
test (Figure 4.3). Here, the interpolated roll and the sonar head mounting roll angle
are subtracted from the measured beam angle by the MBES.
Initialisation
i = 1, j = 1
i=i+1
j=j+1
Update calculation
for each beam of each ping
EBA(i, j ) = BA(i, j ) − ER(i) − R
j<=80
i<=200
Resultant effective beam angle matrix
EBA
Figure 4.3 Flowchart for the calculation of effective beam angle
After that the net pitch angles (NP) are computed to correct each beam for
MBES mounting pitch angle (P) and real time vessel pitching angles, which are
interpolated with respect to the MBES timing (EP). This step is shown in the
Figure 4.4.
79
Initialisation
i =1
i=i+1
Update calculation each ping
NP (i, j ) = EP(i ) + P
i<=200
Resultant effective pitch angle matrix
NP
Figure 4.4 Flowchart for the calculation of net pitch angle
Then the correct beam directions (BD) are determined by using the
calculated effective beam angle (EBA) and the net pitch angle (NP) for each beam
(Figure 4.5). This is the true direction of the beam after applying all the vessel
movements.
Initialisation
i = 1, j = 1
i=i+1
j=j+1
Update calculation
for each beam of each ping
BD(i, j ) = tan −1 (tan( NP(i )) 2 + tan( EBA(i, j )) 2 )
j<=80
i<=200
Resultant beam direction angle matrix
BD
Figure 4.5 Flowchart for the calculation of final beam direction
80
The Snell’s refraction constants (SNCT) are determined for each beam of
each ping using the above computed beam direction (BD) and the Surface Sound
Speeds (SSS) recorded by the MBES at each ping (Figure 4.6).
Initialisation
i = 1, j = 1
i=i+1
j=j+1
Update calculation
for each beam of each ping
SNCT (i, j ) =
sin( BD(i, j ))
SSS (i )
j<=80
i<=200
Resultant refraction constant matrix
SCNT
Figure 4.6 Flowchart for the calculation of the Snell’s refraction constant using
surface sound speeds
Then the travel times up to n-1 sound speed layer (up to the layer before the
last) for each beam (ttt) are computed using the Equation 2.22. Here the sound
speed layer thickness is 0.5m (depth interval of the SVP). In this same step, the
total number of sound speed layers that a particular beam had travelled also
determined (N). The flowchart for this stem is shown in the Figure 4.7.
81
Initialisation
i = 1, j = 1
i=i+1
j=j+1
k = 1,& tt = 0
k=k+1
Update calculation for each beam
until total travel time (TT)
t=
0.5
(1 − ( SVP (k ) × SNCT (i, j )) 2
SVP(k )
tt = tt + t
tt<=TT
yes
Update sound speed layer number
and travel time counter
no
N (i, j ) = k − 1
ttt (i, j ) = tt − t
j<=80
i<=200
Resultant matrixes for
(N-1) sound speed layer counter and
travel time up to (N-1) speed layer
N , ttt
Figure 4.7 Flowchart for the calculation of the sound speed layer number and travel
time up to (N-1) sound speed layer
Then the travel times of last sound speed layers (TN) are determined by
subtracting the travel time up to n-1 layer (ttt) from the total travel time of each
beam (TT) for the data set (Figure 4.8).
82
Initialisation
i = 1, j = 1
i=i+1
Update calculation
for each beam of each ping
j=j+1
TN (i, j ) = TT (i, j ) − ttt (i, j )
j<=80
i<=200
Resultant matrix for last sound speed layer
TN
Figure 4.8 Flowchart for the calculation of the travel time in the last sound
speed layer
The range (RN) and the depth (DN) for the last sound speed layer are also
calculated for each of the beam with respect to the travel times (TN) of the last
layer (Figure 4.9 and Figure 4.10).
Initialisation
i = 1, j = 1
i=i+1
Update calculation
for each beam of each ping
j=j+1
RN (i, j ) = SVP( N (i, j ) + 1) × TN (i, j )
j<=80
Resultant matrix for range distance
in the last sound speed layer
i<=200
RN
Figure 4.9 Calculation of the range distance in the last sound speed layer
83
Initialisation
i = 1, j = 1
i=i+1
Update calculation
for each beam of each ping
j=j+1
DN (i, j ) = RN (i, j ) × (1 − ( SVP (( Ni, j ) + 1) × SNCT (i, j )) 2
j<=80
Resultant matrix for depth in the
last sound speed layer
i<=200
DN
Figure 4.10 Flowchart for the calculation of the depth in the last sound speed layer
The final depths (TOTD) are computed using the layer depth (0.5m),
number of layers travelled (N-1), last speed layer’s depth (DN), transducer draft,
tide and heave (EH) for each beam (Figure 4.11).
Initialisation
i = 1, j = 1
i=i+1
Update calculation
for each beam of each ping
j=j+1
TOTD(i, j ) = N (i, j ) × (−0.5) − DN (i, j )
− Draf + Tide(i ) + EH (i )
j<=80
i<=200
Resultant matrix for final depth
TOTD
Figure 4.11 Flowchart for the calculation of the final reduced depth of each beam
84
After that with the above information the total across track distances
(ACCTK) for each beam up to n-1 layer are calculated using the Equation 2.23.
This process is shown in the Figure 4.12.
Initialisation
i = 1, j = 1
i=i+1
j=j+1
n = 1, x = 0 , d = 0
n=n+1
Summation of across track in each
sound speed layer up to (N-1) layer
d=
SVP(n) × SNCT (i, j )
2 (1 − ( SSVP( n) × SNCT (i, j )) 2
x = x+d
n<=N(i,j)
yes
no
Across track distance
ACCTK (i, j ) = x
j<=80
i<=200
Update calculation for each beam
until (N-1) sound speed layer
Resultant matrix for across track up to
(N-1) sound speed layer
ACCTK
Figure 4.12 Flowchart for the calculation of the total across track distance up to
(N-1) sound speed layer for each beam from the sonar head position
85
The across tracks for the last speed layer (ACCTN) are also calculated using
the last layer’s range distance (RN), Snell’s constant (SNCT) and layer speed for
each beam (Figure 4.13).
Initialisation
i = 1, j = 1
i=i+1
j=j+1
Update calculation
for each beam of each ping
ACCTN (i, j ) = RN (i, j ) × SVP ( N (i, j ) + 1) × SNCT (i, j )
j<=80
i<=200
Resultant matrix for across track
at the last sound speed layer
ACCTN
Figure 4.13 Flowchart for the calculation of the across track distances for each
beam at the last sound speed layer
Then by summing up the across track up to n-1 layer (ACCTK) and the
across track of the last sound speed layer of each beam (ACCTN), the total across
track of each beam is determined. Figure 4.14 shows the process of calculating the
total across track (TOTACCTK).
86
Initialisation
i = 1, j = 1
i=i+1
Update calculation
for each beam of each ping
j=j+1
TOTACTK (i, j ) = ACCTK (i, j ) + ACCTN (i, j )
j<=80
i<=200
Resultant matrix for total across track
TOTACCTK
Figure 4.14 Flowchart for the calculation of the total across track for each beam of
each ping
The final beam position (BPOS) resulting with the net pitch (NP) and
effective beam angle (EBA) is then determined for each beam (Figure 4.15). This is
the actual direction that each beam had emitted from the sonar head. With this
information and the total across track of each beam (TOTACCTK), the final
corrected across track distance of each beam footprint (COTOTACCTK) is
determined from the sonar head. This is illustrated in Figure 4.16.
87
Initialisation
i = 1, j = 1
i=i+1
Update calculation
for each beam of each ping
j=j+1
⎧ tan( NP(i )) ⎫
BPOS (i, j ) = tan −1 ⎨
⎬
⎩ tan( EBA(i, j )) ⎭
j<=80
i<=200
Resultant matrix for
corrected beam direction
BPOS
Figure 4.15 Flowchart for the calculation of the corrected beam direction with
respect to the sonar head position for each beam of each ping
Initialisation
i = 1, j = 1
i=i+1
Update calculation
for each beam of each ping
j=j+1
COTOTACCTK (i, j ) = TOTACCTK (i, j ) × cos( BPOS (i, j ))
j<=80
i<=200
Resultant matrix for corrected
total across track distance
COTOTACCTK
Figure 4.16 Flowchart for the calculation of the corrected total across track with
respect to the corrected beam direction for each beam of each ping
88
After that, Easting and Northing differences (dx, dy) for each beam is
calculated with respect to the sonar head position using the corrected total across
track distance (COTOTACCTK) and effective vessel heading (Figure 4.17).
Effective vessel heading is the sum of the interpolated vessel heading (VH) from
the gyrocompass and the sonar mounting Yaw angle (Y).
Initialisation
i = 1, j = 1
i=i+1
j=j+1
Update calculation
for each beam of each ping
dx(i, j ) = COTOTACCTK (i, j ) × cos(VH (i ) + Y ))
dy (i, j ) = COTOTACCTK (i, j ) × sin(VH (i ) + Y ))
j<=80
i<=200
Resultant matrix for dE and dN
dx, dy
Figure 4.17 Flowchart for the calculation of the Easting and Northing differences
with respect to the sonar head position for each beam
Then the final position of each beam (BX, BY) is computed using the sonar
head’s Easting (SHPOSE) and Northing (SHPOSN) with the above determined dx
and dy values. Here for all the starboard beams (EBA>0), the calculated dx values
are positive and the dy’s are negative. For the port side beams (EBA<0), dx’s are
negative and dy values are positive. This is shown in the Figure 4.18.
89
Initialisation
i = 1, j = 1
i=i+1
Update calculation for each beam of each ping
j=j+1
j<=80
i<=200
EBA(i, j ) ≤ 0
Yes
BX (i, j ) = SHPOSX (i ) − dx(i, j )
BY (i, j ) = SHPOSY (i ) + dy (i, j )
Easting and Northing calculation
for each beam of each ping
BX
BY
Resultant matrix
For final E and N
No
BX (i, j ) = SHPOSX (i ) + dx(i, j )
BY (i, j ) = SHPOSY (i ) − dy (i, j )
Figure 4.18 Flowchart for the calculation of the final Easting and Northing for each
beam
4.3 The Algorithm of the SSS Program
Figure 4.19 schematically shows the overall algorithm used in the SSS
program development.
It summarized each processing steps that are being
discussed in the above flowcharts.
Firstly, all the raw data measured by each sensor are extracted. From the
MBES calibration, the sonar head mounting offset values are also obtained (R, P,
Y). The measured travel times of each beam of the MBES are in number of samples
(Total Samples). This is then translated in to travel time in seconds using the pulse
sample frequency 13125 pps. The vessel attitude data measured (VR, VP, VH) by
the MRU are in different times than the MBES ping timing. To apply the vessel
90
attitude data to the MBES data, the attitude data also needed to be interpolated with
respect to the ping time (ER, EP, EH).
Measured beam directions are modulated by the vessel roll. Therefore, the
effective beam angles (EBAng) are calculated by applying the sonar mounting roll
angle (R) and the interpolated vessel roll angles (ER) to the each of the measured
beam angle (BAng). Net pitch angles (NP) are also calculated for each ping using
the sonar mounting pitch angle (P) and the interpolated vessel pitch angles (EP).
Then, using these two values (effective beam directions and the net pitch angles)
the final beam directions (BD) are determined.
Then, using the SSSs and the above calculated beam directions; the Snell’s
refraction coefficients are determined for each of the beams of the MBES data set.
Now using the refraction constant, SVP, transducer draft, tidal value and the
interpolated heave (EH), the depth of each beam is calculated. The across track
distance from the sonar head position to the each beam footprint is computed using
the SVP and the refraction constant. After that, the Easting and the Northing
differences (dE, dN) from the sonar head to each beam footprint are determined
using the vessel heading and the sonar mounting yaw angle (Y). With the sonar
head’s Easting and Northing and the above calculated dE and dN, the final Easting
and the Northing of each beam footprint is calculated.
This way the final
bathymetry is computed for the data set, using the SSS in refraction constant
determination.
91
SSS, SVP, Beam Angle, V-Heading,
V-Attitude (VR,VH,VP), Draft,
Sonar head Position (E,N), Tide,
MB Calibration (R,P,Y), Total Samples,
Sample Frequency
BAng,
R
Total Samples,
Sample frequency
Effective B.Ang,
(EBAng)
Travel Time
VR, VP
VH
Interpolated Roll,
Pitch, Heave
(ER, EP, EH)
ER
P
SSS
EP
Net Pitch
(NP)
Snell’s Refraction
Constant
Final Beam
Direction
(BD)
EH
Depth
SVP,
Tide, Draft
SVP
Across Track
from Sonar Head
Vessel
Heading, Y
dE and dN
from Sonar Head
Sonar head
Position (E, N)
Final N and E
of Beams
Northing, Easting
and Depth of all
Beams
Figure 4.19 Algorithm for bathymetric calculations using the SSS in refraction
constant
92
4.4 The SSVP Program
Here, the surface value of the sound velocity profile (SSVP) is used to
compute the refraction coefficient (SNCT) for each beam in each ping using the
beam direction of each beam (BD). Unlike in SSS case, this SSVP is one value and
it is constant for the whole data set.
Figure 4.21 shows the flow chart for
calculating the refraction coefficient using SSVP. In the SSVP program, all other
processing steps are same as discussed in the Section 4.2.
Initialisation
i = 1, j = 1
i=i+1
j=j+1
Update calculation for each beam of each ping
SNCT (i, j ) =
sin( BD(i, j ))
SSVP(1)
j<=80
i<=200
Resultant refraction constant matrix
SNCT
Figure 4.20 Flowchart for the determination of the Snell’s refraction constant using
SSVP
4.5 The Algorithm of the SSVP Program
The only different in this SSVP algorithm from the above SSS algorithm is
the method of computing the refraction coefficient. The rest of the bathymetric
calculation procedures are same as discussed in the Section 4.3. Figure 4.21 shows
the overall algorithm used for SSVP program development.
93
SSS, SVP, Beam Angle, V-Heading,
V-Attitude (VR,VH,VP), Draft,
Sonar head Position (E,N), Tide,
MB Calibration (R,P,Y), Total Samples,
Sample Frequency
BAng,
R
Total Samples,
Sample frequency
Travel Time
Effective B.Ang,
(EBAng)
VR, VP
VH
Interpolated Roll,
Pitch, Heave
(ER, EP, EH)
ER
P
SSVP
EP
Net Pitch
(NP)
Snell’s Refraction
Constant
Final Beam
Direction
(BD)
EH
Depth
SVP,
Tide, Draft
SVP
Across Track
from Sonar Head
Vessel
Heading, Y
dE and dN
from Sonar Head
Sonar head
Position (E, N)
Final N and E
of Beams
Northing, Easting
and Depth of all
Beams
Figure 4.21 Algorithm for bathymetric calculations using SSVP in the refraction
constant
CHAPTER 5
DATA PROCESSING
5.1 Programme Validation
It is necessary to validate the developed computer programs to ensure that
the adopted methodology for bathymetric calculations is giving correct results. For
this purpose, output bathymetry from the SSVP program for the 1st MBES line is
compared with corresponding QINSy final output for the same survey line. QINSy
is widely used commercial MBES software and it also uses the surface value of
SVP (SSVP) in refraction constant determination (QPS BV.). Here, the Eastings
Northings and depths of the beams from the 1st and the last (200th) pings of the data
set are compared with the corresponding values from the QINSy results.
5.2 Data Processing
5.2.1 MBES Data Processing
Raw MBES data (beam angle, total samples, SSS) are selected for each of
80 beams from 200 pings from the first MBES line and 253 pings from the second
MBES line, for the same overlapping area. The corresponding MAHRS, DGPS,
tidal data were extracted for the same corresponding time period (with respect to the
selected MBES data time periods) for the both MBES lines. All the raw data are
then copied to Excel files. Then the bathymetric calculations are done, separately
95
using the developed two computer programmes (SSS and SSVP). The final outputs
(E, N and Depth) are again copied back to Excel files separately for each MBES
line from each program. Then these Excel files are converted to CSV files (Figure
5.1a) and later converted to SCR file format using Notepad (Figure 5.1b).
(a)
(b)
Figure 5.1 Final MBES coordinate conversion (a) Opened CSV file in Notepad (b)
Converting the CSV file into a SCR file in Notepad
These SCR files are made as ‘multiple point’ type, so that all these points
can directly run as script files in ‘AutoCAD’ (figure 5.2). Then TIN based DTMs
are generated for each data set using the ‘Quicksurf’ software loaded into the
AutoCAD as shown in Figure 5.3. Figure 5.4 shows the generated DTM for the 1st
MBES line data set (200 pings) using Quicksurf.
96
Figure 5.2 Processed MBES bathymetric data from the program output, loaded in
to AutoCAD as a multiple point script file
Quicksurf tool bar
Figure 5.3 Quicksurf software loaded in AutoCAD R14
97
DTM Surface
Boundary
Figure 5.4 Generated DTM for the 1st MBES data set using Quicksurf
5.2.2 SBES Data Processing
SBES data (E, N, Depth) are directly exported to Excel from QINSy
software for the corresponding area with MBES data sets. Then, ‘poly line’ SCR
files are created using Easting and Northing values of the lines in Notepad (Figure
5.5a), to be used as the reference lines to generate subsequent cross sections on
MBES DTMs. In order to generate the SBES profiles, ‘line’ SCR files are created
using the distance and the depth of the SBES (Figure 5.5b). By this way, all the
profiles (MBES and SBES) are assured along the same reference lines. Figure 5.6
shows the SBES profile generated in AutoCAD after running the SCR file.
98
(a).
(b)
Figure 5.5 SBES script (SCR) generation (a) Poly line script file to create the
reference line (SBES) (b) SBES profile generation using distance and depth in the
script file
Figure 5.6 SBES profile after running the script file in AutoCAD
99
5.2.3 Final Comparison
The main objective of this study is to determine which refraction constant
determination technique giving better results in ray tracing. Here, two things are to
be determined; how each refraction constant determining technique performs at the
nadir and how each performs at the edge of the swath. These analyses are
performed based on profile comparison.
For the nadir comparison, DTMs are generated form SSS and SSVP
program outputs for each MBES lines and pairs of nadir profiles are generated on
each SSS and SSVP DTMs for each MBES line. Then each nadir profile pairs (SSS
and SSVP) from same line are compared. In the outer swath profile comparison
case, outer swath pairs (SSS and SSVP) from the same MBES line are compared
against the corresponding SBES profile and the adjacent MBES lines’ nadir profile
(from SSVP DTM). Since two MBES lines are being used, a common DTM
boundary is used to restrict the generated DTM. This way, the generated DTMs are
always over the same area for all the MBES lines for all the cases (SSS and SSVP).
Also to ensure all the profiles are made along the same line on DTMs, the SBES
line is used as the reference line for all the profiles (Figure 5.7).
SBES line
2nd MBES line
Common Boundary
for DTMs
1st MBES line
Figure 5.7 Loaded data sets in to AutoCAD (two MBES lines, SBES line in the
middle and the DTM boundary)
100
Then, each profile is generated after turning off the other DTM layers to
deactivate the required DTM surfaces. After that each profile is converted as
‘blocks’ in AutoCAD drawing (Figure 5.8). The lower left corner of the each
profile is given as the reference point for each block. By doing so, all the profiles
need to compare can inserted in to a single AutoCAD drawing as shown in the
Figure 5.9. This way, the generated profiles can exactly overlay each other using
this common reference point (Figure 5.10).
Base/Reference point
of the block
Figure 5.8 Generated profiles are saved as blocks with a reference (base) point
101
SSS profile
Block (MBES)
SSVP profile
Block (MBES)
SBES profile
Block
Figure 5.9 Loaded profile blocks in to a single drawing for the comparison
Common reference point of all the blocks
Figure 5.10 All the blocks are overlaid each other using the common base point in
the final comparison
CHAPTER 6
RESULTS AND ANALYSIS
6.1 Introduction
This chapter discusses the results obtained from both simulated and real data
in SSS and SVP cases and the comparison results from refraction constant
determination techniques. Firstly, the validation results of the developed computer
programs are presented. Secondly, the effects of inadequate SSS and SVP are
presented. The IHO standards are used for this analysis. Finally, the results
obtained from the proposed refraction reduction methodology is analysed and
inferences drawn.
6.2 Results of Program Validation
The validation is performed to test that the adopted bathymetric calculation
procedures used in the program development are correct. The only difference
between the developed SSVP and the SSS program is the method of refraction
constant determination. However, all the other computational procedures used are
same in the both programs.
103
The refraction constant determination technique and the computational
procedures in QINSy software and the developed SSVP program are similar.
Hence; the SSVP program is used in the validation with QINSy software. The
validation was done using two pings (1st and the 200th ping) from the MBES line-1
dataset. Corresponding Northing, Easting and the depths are compared in this case.
The full validation results are shown in the Appendix D.
6.2.1 Northing Comparison
A comparative analysis of the northing coordinates of the QINSy software
and the SSVP program were carried out. The Northing comparison results are
presented in Figures 6.1 and 6.2 for the corresponding two pings respectively.
Result shows that the northing differences in the first ping ranges from 0.13m to
0.24m while in the second ping, range between 0.09m and 0.19m. The SSVP
program’s Northings are always slightly greater than QINSy Northings.
QINSy
Qinsy vs SSVP (Northing) - Ping 01
SSVP
160030.00
160020.00
Northing (m)
160010.00
160000.00
159990.00
159980.00
159970.00
1
11
21
31
41
51
61
71
81
Beam No
Figure 6.1 Northing coordinate comparison between QINSy vs. SSVP programmes
for the first ping
104
QINSy
Qinsy vs SSVP (Northing) - Ping 200
SSVP
160080.00
Northing (m)
160070.00
160060.00
160050.00
160040.00
160030.00
160020.00
1
11
21
31
41
Beam No
51
61
71
81
Figure 6.2 Northing coordinate comparison between QINSy vs. SSVP programmes
for the second (200th) ping
6.2.2 Easting Comparison
The Easting coordinates of the same two pings are compared between
QINSy software and SSVP program. Easting in the first ping ranges between 0.15m
and 0.30m, while in the second ranges from 0.13m to 0.31m. Easting comparisons
are shown in the Figures 6.3 and 6.4 respectively. Here also the differences are
consistent.
105
Easting Comparison
Ping - 01
Qinsy
SSVP
1296880
1296870
1296860
Easting (m)
1296850
1296840
1296830
1296820
1296810
1296800
1296790
1296780
1
11
21
31
41
Beam No
51
61
71
Figure 6.3 Easting coordinates comparison between QINSy vs. SSVP for the first
ping
Easting Comparison
Ping - 200
Qinsy
SSVP
1296920
1296910
1296900
1296890
Easting (m)
1296880
1296870
1296860
1296850
1296840
1296830
1296820
1296810
1
11
21
31
41
Beam No
51
61
71
Figure 6.4 Easting coordinates comparison between QINSy vs. SSVP for the
second (200th) ping
106
6.2.3 Depth Comparison
In depth comparison, the first ping gave 0.08m to 0.14m differences and the
second ping gave 0.08m to 0.13 m differences. SSVP program depths are always
slightly deeper than the corresponding QINSy software depths and consistent
Figures 6.5 and 6.6 are showing the depth comparison results of the two pings.
Qinsy
Depth Comparison
Ping - 01
SSVP
-24
-26
Depth (m)
-28
-30
-32
-34
-36
-38
-40
1
11
21
31
41
Beam No
51
61
71
Figure 6.5 Depth comparison between QINSy vs. SSVP for the first ping
107
Qinsy
Depth Comparision
Ping - 200
SSVP
-27
-29
Depth (m)
-31
-33
-35
-37
-39
-41
-43
-45
1
11
21
31
41
Beam No
51
61
71
Figure 6.6 Depth comparison between QINSy vs. SSVP for the second (200th) ping
6.2.4 Summary of Program Validation
The above results show a slight, but consistence differences between the
QINSy software and the SSVP programme outputs (SSVP results are always
slightly greater than QINSy results), but the original bathymetric features are
preserved. Even more, the difference falls well within the IHO special order (which
is the highest order). This shows that, there is agreement between QINSy software
results and SSVP program results.
The validation of the SSVP program confirmed that, the bathymetric
calculation procedures used in the SSVP program development is correct. Since the
developed SSVP and SSS programmes have the same computational procedures,
except the method of refraction constant determination; the validation results do not
cause any effect to the final analysis.
108
6.3 Inadequate SSS Effects on Flat Array MBES Transducers
The results of the inadequate SSS effects on flat array MBES transducers
are analysed based on the synthetic data and real data. The analysis considered the
depth and positional effects in each case. The IHO standards (IHO, 2008) are used
as the benchmarks in this analysis.
6.3.1 Synthetic Data Results
Figure 6.7 presents the variation of the magnitude of the angular error in the
beam-pointing angle with different SSS variations in a flat array MBES. For +10
ms-1 SSS variation, the angular error at nadir is 0o. For 30o it gives 0.22 degree
error, while at 60o the error is 0.67o. This indicates that the error increases with the
beam pointing angle. The pattern is similar for the other SSS variations also.
SSS affects in Beam Pointing Angle
0.8
Anguler Error (degrees)
0.6
0.4
+10 m/s
0.2
+5 m/s
0
-0.2
0
10
20
30
40
50
60
-5 m/s
-10 m/s
-0.4
-0.6
-0.8
Beam Angle (degrees)
Figure 6.7 Variation of the magnitude of the angular error with respect to the beam
pointing angle for different SSS variations
109
To determine the effects of the variation of the SSS errors in a flat array
MBES in terms of across track and depth errors; the across track and depth errors
with beam angles are analysed for 100m flat seabed. According to the Figure 6.8
and the Figure 6.9 respectively, for +10 ms-1 SSS variation, 60o beam is giving
1.15m across track error and 2.03m depth error; while at 30o, the across track error
is 0.38m and the depth error is 0.22m. However, no errors are witnessed at the
nadir. The pattern is similar for the other SSS variations as well. The full SSS
variation results are presented at the Appendix B.
Across Track Errors with Beam Angle
Across Track Error (m)
1.5
1
0.5
+10 m/s
+5 m/s
0
0
10
20
30
40
50
60
-5 m/s
-10 m/s
-0.5
-1
-1.5
Beam Angle ( degrees)
Figure 6.8 Across track errors for 100 m flat sea bottom for different SSS errors
Depth Errors with Beam Angle
2.5
2
Depth Error (m)
1.5
1
-10 m/s
0.5
-5 m/s
0
-0.5 0
10
20
30
40
50
60
+5 m/s
+ 10 m/s
-1
-1.5
-2
-2.5
Beam Angle (degrees)
Figure 6.9 Depth errors for 100 m flat sea bottom for different SSS errors
110
The Figure 6.10 is showing the result of the swath shape of a flat seabed
seen by a flat array MBES (SeaBat 8124) due to erroneous SSS measurements.
From the figure, it appears as a parabola directed upward or downward depend on
the sign of the SSS variation; virtually there are no distortions at nadir. But the
effects are more prominent towards the outer edge of the swath.
SSS Effects in Swath Shape
Beam Angle (degrees)
60
50
40
30
20
10
0
-10
-20
-30
-40
-50
-60
97
Depth (m)
98
99
+10 m/s
+5 m/s
100
TRUE
-5 m/s
101
-10 m/s
102
103
Figure 6.10 Impact on the shape of the swath for different SSS errors on a flat sea
floor from a flat MBES
6.3.2 Real Data Results
Here, the results of the real data observed in the study area-1 as discussed in
Section 3.6.1.2 are analysed. The corresponding survey lines are opened form
QINSy> Processing Manager>Validator to visualize the effects of SSS variation on
swath shape. The shapes of the swaths are shown in the Figure 6.11. The typical
parabolic shapes are observed here, as seen in the case of the synthetic data analysis
section. The flat seabed is seen as a curved seabed due to the SSS error. For positive
SSS error, it curls upward and for negative error the same flat seabed tends to curl
downward.
111
(a) Swath shape with Positive SSS (1600ms-1)
(b) Swath shape with correct SSS (1533 ms-1)
(c) Swath shape with Negative SSS (1450ms-1)
Figure 6.11 Real examples for SSS variation effects on a flat array MBES swath
6.3.3 Summary of Inadequate SSS Effects on Flat Array MBES Transducers
Changes in the SSS at water surface with respect to the assumed or
measured SSS at the face of the transducer causes deviation from the direction to
which the beams supposed to steer. Thus leading to depth and positional errors in
MBES bathymetry. The magnitude of the error depends on the magnitude of the
beam pointing angle and the magnitude of the sound speed error. The errors
induced all over the swath except at the nadir beams, the effects are symmetric
about the nadir and non linear. For positive SSS errors, the swath curled up and
curled down for negative SSS errors.
112
From the above analysis, it is clear that with 10 ms-1 SSS variation, beams
greater than 50o will induce high errors not satisfying the IHO special order and
beams greater than 60o will not satisfy the order 1a and 1b survey requirements
(IHO, 2008). In the case of 5 ms-1 SSS variation, beams greater than 60o will not
satisfy the IHO special order standards (Figure 6.12).
Special Order
1st Order 1a and 1b
2nd Order
3
Maximum Error (m)
2.5
2
1.5
1
0.5
0
0
10
20
30
40
50
60
70
80
90
100
Depth (m)
Figure 6.12 IHO error budgets for different levels of surveys
The IHO standards for positioning requirements; 2m in special order, 5m
+5% of water depth in first order and 20m +10% of water depth in second order
(IHO, 2008). From the analysis, it is clear that the SSS variation does causes
positional errors in the bathymetry. However, it does not exceed the any IHO limits
even for 10ms-1 SSS variation. Therefore what is most concerned is depth effects
rather than horizontal positional errors.
6.4 Inadequate SVP Effects on Flat Array MBES Transducers
In this case also, the results are discussed for both simulated data and real
data cases separately in terms of both depth and positional errors. The effects are
analysed against IHO standards.
113
6.4.1 Synthetic Data Results
Figure 6.13 and Figure 6.14 respectively show the effects due to 10ms-1
SVP variation at the first 10m layer of the SVP for a 100m deep flat sea bottom in a
flat array MBES (SeaBat 8124) in terms of depth and position. Both depth and
positional errors increase with the grazing angle. It causes 1.8m depth error and
0.89m across track position error at 60o beam angle, while the same SVP error
causes 0.90m depth and 0.63m across track error at 50o. But, no errors sighted at the
nadir. The errors are increasing with the beam-pointing angle. Appendix C displays
the full SVP variation results in the bathymetry.
Depth Errors with Beam Angle
2
Depth Error (m)
1.5
1
0.5
+Ve SVP
0
-0.5 0
10
20
30
40
50
60
-Ve SVP
-1
-1.5
-2
Beam Angle (degrees)
Figure 6.13 Depth errors due to 10 ms-1 SVP variations at the first 10m layer of the
SVP for a 100m deep, flat sea bottom
Across Track Errors with Beam Angle
Across Track Error (m)
1
0.5
+Ve SVP
0
0
10
20
30
40
50
60
-Ve SVP
-0.5
-1
Beam Angle (degrees)
Figure 6.14 Across track errors due to 10 ms-1 SVP variation at the first 10m layer
of the SVP for a 100m deep flat sea bottom
114
Figure 6.15 shows the resultant swath shape of a flat seabed, seen by a flat
array MBES due to SVP variations. It also shows the typical parabolic shape
curved up or down depend on the sign of the variation. At nadir, no affects, but
effects gets larger with the grazing angle. Here, the effects are completely opposite
to the SSS case (Figure 6.10 and Figure 6.15).
SVP Effects inSwath Shape
Accross Track (m)
-200
98
-150
-100
-50
0
50
100
150
200
98.5
Depth (m)
99
99.5
TRUE
100
+SVP
100.5
-SVP
101
101.5
102
Figure 6.15 Impact of the sound velocity profile errors on the swath shape of a flat
100m deep sea floor due to 10 ms-1 SVP variation at the first 10m layer of the SVP
6.4.2 Real Data Results
Figure 6.16 (a) shows a comparison of the same ping (from the study area-1)
processed with different SVPs. Since the SVP variation was not much and the area
is a shallow, one cannot observe larger effects (only cm level differences).
However, result from the study area 2 gives greater effects for the same SVP
variations (Figure 6.16 (b)). In this case, the difference between the true and the
negative SVP swath is about 1.1m at the outermost edge of the swath while the
difference between positive SVP and the true SVP swath is about 0.85m, but no
errors at the nadir.
115
(a) Ping comparison (Study area-1)
Neg. SVP (Red)
True SVP (Blue)
Pos. SVP (Green)
(b) Swath comparison in QINSy - Validator (Study area-2)
Figure 6.16 True examples for SVP variation effects on swath in a flat array MBES
6.4.3 Summary of Inadequate SVP Effects on Flat Array MBES Transducers
Variation in the SVP in the water that is assumed or measured causes
unknown propagation of the beams through the water column. Depth and positional
(horizontal) errors are results of this matter. Errors induced all over the swath,
except at nadir. Effects are symmetric about the nadir and non linear. For positive
SVP errors the swath curl down (frown) and for negative errors swath curl up
(smile).
116
From the above results, it’s clear that, with 10 ms-1 variation at the first 10m
layer of the SVP, beams greater than 50o will not satisfy the IHO special order
depth requirements. The beams greater than 60o will not satisfy the IHO order 1a
and 1b survey requirements.
The horizontal positional errors induced by the
variation of SVP, do not exceed any IHO limits, even with 10 ms-1 variations at the
first 10m layer for 100m sea bottom.
6.5 Refraction Reduction Results
The main objective of this research is to evaluate the SSS and SSVP values
in the refraction constant determination. For this purpose two computer programs
called ‘SSS’ and ‘SSVP’ were developed. The same data sets were processed using
these programs. In this section, the results from the SSS and SSVP programs are
discussed. The corresponding profiles from both SBES and adjacent MBES line’s
nadir area are used as benchmarks for the comparison.
6.5.1 Nadir Comparison
Usually, the nadir beams are least affected by the refraction (as seen in the
Sections 6.3 and 6.4). Figures 6.17 and 6.18 show the nadir profile comparison
results of the SSS and SSVP programs for the two MBES lines respectively. The
SSS profile is in blue colour and the SSVP profile is in brown colour.
In the comparison, both profiles tally well almost everywhere of the profiles,
except the places where there were random errors in DTM. This proves that, both
SSS and SSVP values in refraction constant determination giving almost the same
the results at the nadir.
117
Both SSS and SSVP
Profiles
(merging together)
1m
Figure 6.17 SSS and SSVP profiles at the nadir from the MBES line 01
Both SSS and SSVP
Profiles
(merging together)
1m
Figure 6.18 SSS and SSVP profiles at the nadir from MBES line 02
118
6.5.2 Outer Comparison
According to the results obtained from the Sections 6.3 and 6.4, the outer
beams are seriously affected by the refraction. Here, the corresponding outer edge
swath profiles from SSS and SSVP program DTMs are compared against
corresponding SBES and the adjacent MBES SSVP nadir profiles. Figure 6.19
shows the SSS and SSVP outer edge swath profile comparison with SBES profile.
Red coloured profile is the SBES profile, while the green and the blue are the SSS
and the SSVP profiles, respectively.
SSVP outer Profile -L1
SSS outer Profile - L1
1m
SBES Profile
Figure 6.19 SSS, SSVP and corresponding SBES profile comparison at the outer
edge of the swath of MBES line 01
Figures 6.20 and 6.21 show the outer edge swath profile comparisons for the
two MBES lines using both SSS and SSVP DTMs, with the corresponding SBES
and adjacent MBES nadir profiles. The adjacent MBES nadir profiles are in dark
ash colour.
119
SSVP outer Profile - L1
SSVP nadir Profile -L2
SSS outer Profile -L1
1m
SBES Profile
Figure 6.20 SSS and SSVP outer beam profiles for MBES line 01 and
corresponding SBES and adjacent MBES nadir (line 02) profile comparison
SSVP nadir Profile –L1
1m
SSVP outer Profile – L2
SSS outer Profile –L2
SBES Profile
Figure 6.21 SSS and SSVP outer beam profiles for MBES line 02 and
corresponding SBES and adjacent MBES nadir (line 01) profile comparison
120
The results clearly show that, at the outer edge of the swath, the SSS and
SSVP profiles do not match together as at the nadir. They are separated. The SSS
profiles are always deeper than the corresponding SSVP profiles. None of the
profiles matches with the corresponding SBES or the adjacent MBES nadir profiles.
But the corresponding SBES profiles and the adjacent MBES nadir profiles are
matching quite well. This difference between outer and nadir profiles gives an
indication of the used (true) SVP is not fully relevant to the dataset; even though it
is collected at the same locality of the survey area, prior to the data collection.
Therefore, it is clear that there are refraction effects exist in the data set. But in the
final comparison, it is obvious that the SSS profiles are closer to the both nadir
MBES and SBES profiles than the corresponding SSVP profiles. The differences
between the SSVP outer profiles and the corresponding nadir profiles in the both
data sets are around 0.10m to 0.15m. At the same time SSS profile differences are
about 0.04m to 0.06m.
The above results shows that, the SSS technique in
refraction constant determination is giving better results in refraction artifacts (25%
to 30%).
6.5.3 Summary of Refraction Reduction Results
The results show that at the nadir, both approaches are giving the similar
results. It does not make any difference, whether the SSS or the SSVP used in
refraction constant determination.
However, at the outer edge of the swath, the SSS profiles always giving
closer profiles to the corresponding SBES and MBES nadir profiles than that of the
SSVP. This proves that the use of the more frequent SSS in refraction constant
determination giving better results than the use of SSVP.
CHAPTER 7
CONCLUSION AND RECOMMENDATIONS
7.1 Conclusion
Sound speeds measurements are clearly a critical component of the
multibeam survey. The measurements, applications and the effects of this critical
parameter must take into consideration prior to conducting of the survey. The
hydrographer must have a thorough knowledge of the area, when and how each
comes in to play and the affects generated; and the procedures in mitigating these
effects. However, manufacturers of multibeam systems do not fully present the
implications of spatially under-sampling of the sound velocity in a quantitative
manner. This study has provided a broad spectrum and the nature of the refraction
artifacts that occurs due to inadequate monitoring of the sound speeds (in SSS and
SVP). In addition to that, the study compared the SSS and SSVP values in
refraction constant determination. Based on the results obtained, the following
conclusions are made.
Surface sound speed measurements are required in flat array multibeam
transducers to determine the correct delays to be applied to each beam in the beam
steering process. If the SSS in not measured in real-time or the applied SSS is
different from the original SSS, the beams are steered in a different direction as
against the calculated; thus leading to depth and positional errors in the MBES
bathymetry.
The SSS effects are less at the nadir area and maximum at the
outermost beams. Proper care must be taken during the SSS measurements to avoid
the outermost beams of the MBES not satisfying the IHO standards.
122
The SVP determines the ray propagation through the water column in
MBES. If the measured SVP is not the correct SVP of a location, the calculated ray
paths are incorrect. This also results to depth and positional errors in the multibeam
bathymetry. The artefacts are minimum at the nadir and maximum at the outer edge
of the swath. If one does not be careful enough with the SVP measurements, the
outermost beams of the MBES will not satisfy the IHO standards.
In almost all flat array multibeam systems, both SSS and SVP
measurements are made.
However, most of the MBES data processing software
does not utilize these real-time collected SSS data for the refraction purpose. This
study clearly shows that, the use of more frequent real-time SSS in refraction
constant determination gives 25% to 30% better results than the SSVP.
This
indicates that, the SSS is correlated with the SVP, hence in the absence of a better
method to measure the rapid time and space varying SVPs, this method would
provides a better, cost effective and simpler means of reducing such effects in
multibeam data.
Reduction of refraction effects increases the accuracy and reliability of the
bathymetric data. This method provides an improved image of the seafloor for
engineers and geoscientists and also cost effective means of handling non-critical
survey projects (pipe-lines, cable route, fisheries habitats). The method also makes
the results of successive survey comparisons are much more meaningful (in
dredging surveys).
7.2 Recommendations
Tropical seas (like in Malaysia) do not exhibits much variation in the sound
speeds both temporally and diurnally, unless in river mouth areas where there is
continues mix of fresh and salt waters. This study was carried out at the Lido
Beach, Johor Bahru, Malaysia, where the environment in not that highly refractive.
123
Therefore a similar study should be carried out in a high refractive area preferably,
Northern or Southern seas or a river deltaic area.
The maximum water depth of the study area is about 40m. Here, one cannot
expect large refraction effects. Therefore, it is proposed to carry out a similar study
in the deeper waters as well.
In this study, single beam and adjacent multibeam nadir profiles were used
for the comparison of the two techniques. Here, the obtained profile mismatch
between the outer profiles vs. the corresponding nadir and SBES profiles might not
completely due to the refraction effects. Other errors inherent in the systems may
have being involved. Hence, it is recommended to carryout a through study using
known seabed or a simulated dataset.
The validation results showed that the developed SSVP program is giving
acceptable bathymetric results with commercial MBES processing software
(QINSy). Therefore based on this, an alternative MBES processing software can be
develop as an in-house multibeam processing software for research purposes. Most
commercial software use for MBES data processing does not make provision for
the user to customise the process; makes it difficult for using such for research
purposes.
124
REFERENCE
Badiey, W., Wong, K. and Lenain, L. (2002). High Frequency Acoustic
Propagation in the Presence of Oceanographic Variability. Proceeding
NATO Scalant Undersea Research Center. Roma.
Batton, D. (2004). The Effect of Refraction on Oblique Angles of Multibeam
Echosounder Due to Sound Speed Changes Through the Water Column. The
Hydrographic Journal, July 2004:15-20.
Beaudoin, J., Clarke, J. E. H. and Bartlett, J. (2004). A Sound Speed Decision
Support System for Multibeam Sonar Operation in the Canadian Arctic.
University of New Brunswick, Department of Geodesy and Geomatis
Engineering.
Online: www.omg.unb.ca/omg/papers/beaudoin_ASM2004_poster.pdf
Visited: November, 2006.
Beaudoin, J., Clarke, J. E. H. and Bartlett, J. (2004). Raytracing (And Reraytracing) Amundsen's Journey Through the Northwest Passage.
Proceedings of the Canadian Hydrographic Conference 2004, Ottawa.
Burdic, W. S. (1991). Underwater Acoustic System Analysis. 2nd ed. New Jersey:
Prentice Hall.
Cartwright, D. S. and Clarke, J. E. H. (2002). Multibeam Survey on the Fraser
River Delta, Coping with Extreme Refraction Environment. Proceedings of
Canadian Hydrographic Conference, 2002.
Cartwright, D. S. (2003). Multibeam Bathymetric Surveys in the Fraser River Delta,
Managing Severe Acoustic Refraction Issues. University of New Brunswick:
M.Eng. Report.
Caruthers, J. W. (1977). Fundamentals of Marine Acoustics. Amesterdam,
Netherlands: Elsdvier scientific publishing Company
Charls W. Brennen, (2002). Multibeam Lecture Series. Reson Inc. September, 2002
Clarke, J. E. H., Lamplugh, M. and Kammerer, E. (2000). Intergration of NearContinuous Sound Speed Profile Information. Canadian Hydrographic
Conference, 2000. Montreal.
Clarke, J. E. H. and Parrott, R. (2001). Integration of Dense, Time-Varying Water
Column Information with High-Resolution Swath Bathymetric Data. U.S.
Hydrographic Conference, 2001, Norfolk VA.
125
Clarke, J. E. H. (2002). Visualization of Seafloor: Mapping Submarine Landscapes.
Symposiun on Visualization and Looking Beyond Earth. American
Association of Advencement of Sciences.
Online: www.omg.unb.ca/AAAS/UNB_Seafloor_Mapping.html
Visited: October, 2006.
Coates, R. F. W. (1990). Underwater Acoustic Systems. London: Macmillan
Education LTD.
Coppens, A. B. (1981). Simple equation for the speed of sound in Neptunian
waters. The Journal of the Acoustical Society of America. 69(3): 862-863.
Dera, J. (1991). Marine Physics. Amsterdam; New York; Elsevier; Warszawa:
PWN-Polish Scientific Publishers.
Dinn, D. F., Loncarevic, B. D. and Costello, G. (1995). The Effect of Sound
Velocity Errors on Multi-beam Sonar Depth Accuracy. Challenges of Our
Changing Global Environment, IEEE Ocean’95 Conference Proceedings. 2:
1001-1010.
Etter, P. C. (2003). Underwater Acoustic Modeling and Simulation. 3rd ed. London:
Spon Press
Furlong, A., Beanlands, B. and Chin-Yee, M. (1997). Moving Vessel Profiler
(MVP) Real Time Near Vertical Data Profiles at 12 Knots. IEEE Xplore,
1997.
Gardner, J. V., Mayer, A. L., Clarke, J. E. H., Dartnell, P. and Sulak, K. (2001). The
Bathymetry and Acoustic Backscatter of the Mid Shelf to Upper Slope Off
Panama City, Florida, Northeastern Gulf of Mexico. CURISE Report.
Horne, R. A. (1969). Marine Chemistry: the structure of water and the chemistry of
the hydrosphere. New York: Wiley Interscience.
Huibert J.L., Robert V. D.V., Tim Haycock, Pieter Jansen, Rein de Vries, Paul V.
W. and Cor Beemster, (2006). Handbook of Offshore Survey-Acquision and
Processing. London: Clarkson Research Services Ltd.
Ingham, A. E. and Abbott, V. J. (1992). Hydrography for Surveyor and Engineer.
3rd ed. New Jersey: Blackwell Science.
International Hydrographic Organization (2008). IHO standards for hydrographic
Surveys, 5th ed. Monaco. Special Publication No.44.
International Hydrographic Organization (2005). Manual on Hydrography. 1st ed.
Monaco: International Hydrographic Bureau.
Jacops, M. (2002). Analysis of High Resolution Bathymetric Data in the Eltanin
Impact Area. Delft University of Technology, Germany: Masters Thesis.
126
James V.G., Larry A.M., Clarke J.E.H., Peter D. and Kenneth J.S., (2001). The
Bathymetry and Acoustic Backscatter of the mid shelf to upper slope of
Panama City, Mexico. CRUISE Report, USGS. M1-01-GM September 2001
Jeroen Van D.A. (2007). Improving MBES measurements. Hydro International.
Vol 11 November 02: 27-29
Jong, C. D., Lachapelle, G., Skone, S. and Elema, I. A. (2002). Hydrography. 1st
ed.
BM Delft, The Netherlands: Vereniging voor Studie- en
Studentenbelangen te Delft.
Kammerer, E., Clarke, J. E. H., Doucet, L. N. and Godin, A. (1998). Monitoring
Temporal Changes in Seabed Morphology and Composition Using
Multibeam Sonars: a Case Study of the 1996 Saguenay River Floods.
Proceedings of the Canadian Hydrographic Conference 1998. Victoria,
Canada: 450-461.
Kammerer, E. and Clarke, J. E. H. (2000). New Method of Removal of Refraction
Artifacts in Multibeam Echosounder Systems. Canadian Hydrographic
Conference 2000. Montreal.
Mackenzie, K. V. (1981). Nine -term equation for sound speed in the oceans. The
Journal of the Acoustical Society of America. 70(3): 807-8112.
National Physics Laboratory (2000). Technical Guidelines-Speed of Sound in Sea
Water. National Physical Laboratory, Middlesex, UK: Technical Report.
Navitronic System AS. (1998). SVP-14/15 Operator’s Manual, Revision 1.00.
Navitronic Systems Association, Denmark.
Odom Hydrographic Systems Inc. (2005). Hydrotrack Installation/Operation
Manual, Version 2.22. ODOM Hydrographic Systems Incorporation,
Louisiana, USA.
OMG. (2005). Y2005 Research Status Report. Chair, Ocean Mapping Group, UNB,
Canada. Online: www.omg.unb.ca/omg/papers/Chair_Report_2005.pdf
Visited: December 2006.
OMG-UNB. Lecture Notes. Ocean Mapping Group- University of New Brunswick,
Canada.
QPS BV. (2004). Drivers and Interfacing Manual QINSy Version 7.x. Quality
Positioning System BV, The Netherlands.
RESON Inc. (2005). RESON Training CD for Universiti Tecknologi Malaysia,
RESON Incorporation, The Netherlands.
RESON Inc. (2003). SeaBat 8124 Operator’s Manual. Version 4, RESON
Incorporation, California, USA.
127
Rodney, F. W. (1990). Underwater Acoustics Systems. London: Macmillan
Education LTD.
Schmidt, V., Chayes, D. and Caress, D. (2006). MB-System Cookbook. Online,
www.mbari.org/data/mbsystem/mb-cookbook/. Visited: December, 2007.
Simrad AS. (2003). MB Stave Display - Instruction Manual. Base Version.
Kongsburg Simrad Association, Horten, Norway.
Tonchia, H. and Bisquay, H. (1996). The Effect of Sound Velocity on Wide Swath
Multibeam System Data. IEEE Oceans’96 “Prospects for the 21st Century”
Conference Proceedings. 2: 969-974.
Trimble Navigation Ltd (2002). DSM 21/212 Receiver User Guide, Version 1.73,
Trimble Navigation Limited, California, USA.
Tucker, D.G. (1966). Underwater observation using sonar. London: Fishing News
(Books)
Urick, R. J. (1983). Principles of Underwater Sound. 3rd ed. Los Altos, Calif:
Peninsula Publishing
USACE (2004). Engineering Design and Hydrographic Survey. Engineering
Manual EM1110-2-1003. U.S. Army Crops of Engineers, Washington DC,
USA.
Valeport Ltd (2004). Valeport 740 Tide Gauge user’s Manual. Valeport Limited,
Devon, UK
VT TSS Ltd (2003). MAHRS Surface Product System Manual. Issue 1.1,VT TSS
Limited, Herts, UK
William, J. and Capell, Sr. (1999). Determination of Sound Velocity Profile Errors
Using Multibeam Data. IEEE Oceans' 99 “Riding the Crest into the 21st
Century” Conference Proceedings. 3: 1144-1148.
128
Appendix A- Database Settings
129
Appendix A- Database Settings
130
Appendix A- Database Settings
131
Appendix A- Database Settings
132
Appendix A- Database Settings
133
Appendix A- Database Settings
134
Appendix A- Database Settings
135
Appendix A- Database Settings
136
Appendix B - Synthetic Data for SSS Case
Beam Angle Variation
Beam Ang
0
10
20
30
40
50
60
Deltd Ang +10
0
0.07
0.14
0.22
0.32
0.46
0.67
Delt Ang +5
0
0.03
0.07
0.11
0.16
0.23
0.33
Delt Ang -5
0
-0.03
-0.07
-0.11
-0.16
-0.23
-0.33
Delt Ang -10
0
-0.07
-0.14
-0.22
-0.32
-0.45
-0.66
Depth Variation
True-depth d+10depth d+5depth d-5depth d-10depth
100
-2.03
-1
1
1.99
100
-0.96
-0.48
0.48
0.93
100
-0.47
-0.24
0.23
0.47
100
-0.22
-0.11
0.11
0.22
100
-0.09
-0.04
0.05
0.09
100
-0.04
-0.01
0.01
0.02
100
0
0
0
0
Across Track Differences
B Ang
0
10
20
30
40
50
60
dAcc+10
0
0.12
0.24
0.38
0.56
0.79
1.15
dAcc+5
0
0.05
0.12
0.19
0.28
0.39
0.57
dAcc-5
0
-0.05
-0.12
-0.2
-0.28
-0.41
-0.58
dAcc-10
0
-0.12
-0.25
-0.39
-0.56
-0.8
-1.17
137
Appendix C - Synthetic Data for SVP Case
Depth Errors
B Ang
0
10
20
30
40
50
60
Depth Error
+Ve SVP -Ve SVP
0.07
-0.07
0.12
-0.11
0.14
-0.15
0.26
-0.26
0.47
-0.49
0.87
-0.9
1.73
-1.8
Across Track Errors
Across Track Error
B Ang +Ve SVP -Ve SVP
0
0
0
10
-0.08
0.09
20
-0.18
0.19
30
-0.29
0.32
40
-0.43
0.44
50
-0.62
0.63
60
-0.9
0.89
Swath Shape
Swath Shape of SVP Errors
B Ang True
+Ve SVP
-Ve SVP
60
100
101.73
98.2
50
100
100.87
99.1
40
100
100.47
99.51
30
100
100.26
99.74
20
100
100.14
99.85
10
100
100.12
99.89
0
100
100.07
99.93
-10
100
100.12
99.89
-20
100
100.14
99.85
-30
100
100.26
99.74
-40
100
100.47
99.51
-50
100
100.87
99.1
-60
100
101.73
98.2
138
Appendix D - Program Validation Results
Northing Comparison
Beam
No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Ping -01
QN
My-N
160026.99 160027.23
160025.90 160026.13
160024.87 160025.10
160023.87 160024.10
160022.90 160023.12
160021.97 160022.19
160021.07 160021.31
160020.21 160020.43
160019.41 160019.63
160018.69 160018.91
160017.96 160018.18
160017.25 160017.46
160016.54 160016.75
160015.91 160016.12
160015.28 160015.51
160014.66 160014.87
160014.06 160014.27
160013.47 160013.69
160012.90 160013.11
160012.33 160012.53
160011.78 160011.98
160011.26 160011.46
160010.73 160010.93
160010.22 160010.42
160009.73 160009.93
160009.24 160009.44
160008.76 160008.95
160008.29 160008.49
160007.83 160008.02
160007.38 160007.57
160006.94 160007.13
160006.49 160006.68
160006.06 160006.25
160005.63 160005.81
160005.22 160005.40
160004.80 160004.99
160004.37 160004.55
160003.96 160004.14
160003.54 160003.72
160003.13 160003.31
d-N
-0.24
-0.23
-0.23
-0.23
-0.22
-0.22
-0.24
-0.22
-0.22
-0.22
-0.22
-0.21
-0.21
-0.21
-0.23
-0.21
-0.21
-0.22
-0.21
-0.20
-0.20
-0.20
-0.20
-0.20
-0.20
-0.20
-0.19
-0.20
-0.19
-0.19
-0.19
-0.19
-0.19
-0.18
-0.18
-0.19
-0.18
-0.18
-0.18
-0.18
Beam No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Ping-200
QN
My-N
d-N
160075.33 160075.52 -0.19
160074.50 160074.69 -0.19
160073.73 160073.91 -0.18
160072.97 160073.16 -0.19
160072.22 160072.41 -0.19
160071.45 160071.62 -0.17
160070.70 160070.89 -0.18
160069.97 160070.18 -0.21
160069.29 160069.48 -0.18
160068.56 160068.74 -0.18
160067.82 160068.01 -0.19
160067.18 160067.36 -0.18
160066.53 160066.71 -0.18
160065.85 160066.03 -0.18
160065.20 160065.39 -0.19
160064.53 160064.71 -0.18
160063.87 160064.03 -0.16
160063.19 160063.36 -0.17
160062.57 160062.74 -0.17
160061.93 160062.11 -0.18
160061.29 160061.46 -0.17
160060.66 160060.83 -0.17
160060.05 160060.22 -0.17
160059.43 160059.60 -0.17
160058.83 160058.98 -0.15
160058.24 160058.40 -0.16
160057.65 160057.81 -0.16
160057.07 160057.23 -0.16
160056.50 160056.66 -0.16
160055.94 160056.10 -0.16
160055.39 160055.55 -0.16
160054.84 160054.98 -0.14
160054.30 160054.46 -0.15
160053.75 160053.90 -0.15
160053.24 160053.38 -0.14
160052.71 160052.86 -0.15
160052.19 160052.34 -0.15
160051.67 160051.82 -0.15
160051.13 160051.28 -0.15
160050.54 160050.70 -0.16
139
Appendix D - Program Validation Results
Northing Comparison
Beam
No
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Ping -01
QN
My-N
160002.71 160002.89
160002.30 160002.47
160001.88 160002.06
160001.47 160001.64
160001.04 160001.20
160000.59 160000.76
160000.17 160000.34
159999.77 159999.94
159999.34 159999.50
159998.84 159999.00
159998.43 159998.59
159997.97 159998.12
159997.49 159997.65
159997.02 159997.18
159996.55 159996.71
159996.06 159996.21
159995.54 159995.69
159994.97 159995.12
159994.48 159994.63
159993.97 159994.12
159993.44 159993.59
159992.86 159993.01
159992.30 159992.46
159991.72 159991.86
159991.15 159991.29
159990.52 159990.66
159989.85 159989.98
159989.23 159989.37
159988.56 159988.70
159987.87 159988.01
159987.18 159987.31
159986.48 159986.61
159985.62 159985.75
159984.71 159984.84
159983.84 159983.94
159982.93 159983.06
159981.91 159982.04
159980.89 159981.01
159979.67 159979.81
159978.51 159978.64
d-N
-0.17
-0.17
-0.18
-0.17
-0.16
-0.17
-0.17
-0.17
-0.16
-0.16
-0.16
-0.15
-0.16
-0.16
-0.16
-0.15
-0.15
-0.15
-0.15
-0.15
-0.15
-0.15
-0.16
-0.14
-0.14
-0.14
-0.13
-0.14
-0.14
-0.14
-0.13
-0.13
-0.13
-0.13
-0.10
-0.13
-0.13
-0.12
-0.14
-0.13
Beam No
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Ping -200
QN
My-N
d-N
160050.11 160050.26 -0.14
160049.59 160049.73 -0.14
160049.08 160049.22 -0.14
160048.58 160048.72 -0.14
160048.04 160048.18 -0.14
160047.53 160047.67 -0.14
160047.02 160047.20 -0.17
160046.49 160046.63 -0.14
160045.97 160046.11 -0.14
160045.43 160045.56 -0.13
160044.87 160045.00 -0.13
160044.39 160044.53 -0.14
160043.82 160043.95 -0.13
160043.33 160043.46 -0.13
160042.81 160042.96 -0.15
160042.20 160042.33 -0.13
160041.56 160041.69 -0.13
160040.93 160041.05 -0.12
160040.25 160040.37 -0.12
160039.61 160039.74 -0.13
160038.92 160039.04 -0.12
160038.23 160038.35 -0.12
160037.55 160037.67 -0.12
160036.85 160036.97 -0.12
160036.14 160036.26 -0.11
160035.42 160035.54 -0.12
160034.59 160034.72 -0.13
160033.80 160033.91 -0.11
160032.96 160033.07 -0.11
160032.05 160032.16 -0.11
160031.14 160031.25 -0.11
160030.13 160030.24 -0.11
160029.08 160029.19 -0.11
160028.06 160028.18 -0.11
160026.99 160027.10 -0.11
160025.63 160025.72 -0.09
160024.60 160024.70 -0.10
160023.34 160023.45 -0.11
160022.05 160022.15 -0.10
160020.54 160020.64 -0.10
140
Appendix D - Program Validation Results
Easting Comparison
Beam No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Ping -01
QE
My-E
d-E
1296784.57 1296784.72 -0.15
1296786.58 1296786.72 -0.14
1296788.47 1296788.61 -0.14
1296790.31 1296790.45 -0.14
1296792.09 1296792.22 -0.13
1296793.82 1296793.97 -0.15
1296795.49 1296795.64 -0.15
1296797.12 1296797.27 -0.15
1296798.59 1296798.73 -0.14
1296799.90 1296800.07 -0.17
1296801.24 1296801.40 -0.16
1296802.57 1296802.73 -0.16
1296803.93 1296804.09 -0.16
1296805.05 1296805.21 -0.16
1296806.23 1296806.40 -0.17
1296807.37 1296807.53 -0.16
1296808.50 1296808.67 -0.17
1296809.60 1296809.77 -0.17
1296810.67 1296810.84 -0.17
1296811.74 1296811.91 -0.17
1296812.81 1296812.97 -0.16
1296813.73 1296813.90 -0.17
1296814.78 1296814.95 -0.17
1296815.78 1296815.96 -0.18
1296816.67 1296816.85 -0.18
1296817.60 1296817.78 -0.18
1296818.50 1296818.68 -0.18
1296819.42 1296819.61 -0.19
1296820.28 1296820.47 -0.19
1296821.14 1296821.33 -0.19
1296822.02 1296822.21 -0.19
1296822.85 1296823.05 -0.20
1296823.68 1296823.87 -0.19
1296824.51 1296824.72 -0.21
1296825.32 1296825.52 -0.20
1296826.13 1296826.33 -0.20
1296826.93 1296827.13 -0.20
1296827.72 1296827.94 -0.22
1296828.52 1296828.72 -0.20
1296829.31 1296829.51 -0.20
Beam No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Ping-200
QE
My-E
d-E
1296816.36 1296816.49 -0.13
1296817.71 1296817.84 -0.13
1296818.94 1296819.08 -0.14
1296820.14 1296820.29 -0.15
1296821.37 1296821.52 -0.15
1296822.61 1296822.76 -0.15
1296823.85 1296824.00 -0.15
1296825.05 1296825.21 -0.16
1296826.12 1296826.28 -0.16
1296827.35 1296827.51 -0.16
1296828.60 1296828.77 -0.17
1296829.59 1296829.75 -0.16
1296830.66 1296830.83 -0.17
1296831.78 1296831.95 -0.17
1296832.85 1296833.02 -0.17
1296833.98 1296834.15 -0.17
1296835.09 1296835.26 -0.17
1296836.27 1296836.44 -0.17
1296837.31 1296837.49 -0.18
1296838.40 1296838.58 -0.18
1296839.52 1296839.71 -0.19
1296840.59 1296840.77 -0.18
1296841.66 1296841.84 -0.18
1296842.75 1296842.94 -0.19
1296843.80 1296843.99 -0.19
1296844.84 1296845.04 -0.20
1296845.89 1296846.09 -0.20
1296846.95 1296847.14 -0.19
1296847.93 1296848.13 -0.20
1296848.94 1296849.15 -0.21
1296849.95 1296850.15 -0.20
1296850.93 1296851.14 -0.21
1296851.90 1296852.11 -0.21
1296852.86 1296853.07 -0.21
1296853.83 1296854.03 -0.20
1296854.78 1296855.00 -0.22
1296855.72 1296855.94 -0.22
1296856.66 1296856.87 -0.21
1296857.61 1296857.83 -0.22
1296858.58 1296858.80 -0.22
141
Appendix D - Program Validation Results
Easting Comparison
Beam No
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Ping -01
QE
My-E
d-E
1296830.10 1296830.32 -0.22
1296830.89 1296831.10 -0.21
1296831.68 1296831.90 -0.22
1296832.48 1296832.69 -0.21
1296833.29 1296833.50 -0.21
1296834.12 1296834.33 -0.21
1296834.92 1296835.13 -0.21
1296835.72 1296835.94 -0.22
1296836.54 1296836.76 -0.22
1296837.45 1296837.67 -0.22
1296838.26 1296838.48 -0.22
1296839.13 1296839.36 -0.23
1296840.03 1296840.25 -0.22
1296840.93 1296841.16 -0.23
1296841.83 1296842.06 -0.23
1296842.76 1296842.99 -0.23
1296843.75 1296843.98 -0.23
1296844.80 1296845.03 -0.23
1296845.76 1296845.99 -0.23
1296846.74 1296846.96 -0.22
1296847.77 1296848.01 -0.24
1296848.87 1296849.11 -0.24
1296849.96 1296850.20 -0.24
1296851.08 1296851.32 -0.24
1296852.20 1296852.44 -0.24
1296853.40 1296853.66 -0.26
1296854.69 1296854.94 -0.25
1296855.90 1296856.15 -0.25
1296857.21 1296857.46 -0.25
1296858.56 1296858.81 -0.25
1296859.91 1296860.15 -0.24
1296861.30 1296861.56 -0.26
1296862.96 1296863.21 -0.25
1296864.71 1296864.97 -0.26
1296866.41 1296866.67 -0.26
1296868.19 1296868.45 -0.26
1296870.16 1296870.42 -0.26
1296872.16 1296872.43 -0.27
1296874.51 1296874.79 -0.28
1296876.79 1296877.09 -0.30
Beam No
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Ping-200
QE
My-E
d-E
1296859.47 1296859.70 -0.23
1296860.40 1296860.63 -0.23
1296861.34 1296861.57 -0.23
1296862.26 1296862.48 -0.22
1296863.22 1296863.46 -0.24
1296864.16 1296864.39 -0.23
1296865.09 1296865.33 -0.24
1296866.05 1296866.29 -0.24
1296867.00 1296867.24 -0.24
1296867.98 1296868.21 -0.23
1296868.99 1296869.24 -0.25
1296869.91 1296870.15 -0.24
1296870.94 1296871.19 -0.25
1296871.87 1296872.12 -0.25
1296872.84 1296873.10 -0.26
1296873.93 1296874.19 -0.26
1296875.07 1296875.33 -0.26
1296876.21 1296876.47 -0.26
1296877.42 1296877.68 -0.26
1296878.59 1296878.84 -0.25
1296879.82 1296880.09 -0.27
1296881.08 1296881.36 -0.28
1296882.32 1296882.59 -0.27
1296883.61 1296883.89 -0.28
1296884.92 1296885.19 -0.27
1296886.24 1296886.51 -0.27
1296887.75 1296888.03 -0.27
1296889.21 1296889.49 -0.28
1296890.74 1296891.01 -0.27
1296892.39 1296892.64 -0.25
1296894.07 1296894.36 -0.29
1296895.90 1296896.19 -0.29
1296897.82 1296898.12 -0.30
1296899.70 1296899.99 -0.29
1296901.68 1296901.99 -0.31
1296904.13 1296904.42 -0.29
1296906.08 1296906.38 -0.30
1296908.41 1296908.71 -0.30
1296910.82 1296911.14 -0.31
1296913.59 1296913.90 -0.31
142
Appendix D - Program Validation Results
Depth Comparison
Beam No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Q-D
-39.60
-39.66
-39.74
-39.75
-39.72
-39.65
-39.54
-39.40
-39.35
-39.46
-39.44
-39.37
-39.17
-39.29
-39.25
-39.20
-39.12
-39.03
-38.96
-38.79
-38.56
-38.68
-38.35
-38.08
-38.12
-37.98
-37.86
-37.58
-37.48
-37.40
-37.06
-37.00
-36.75
-36.59
-36.31
-36.09
-36.06
-35.89
-35.77
-35.61
Ping-01
My- D
-39.68
-39.75
-39.83
-39.84
-39.81
-39.75
-39.64
-39.49
-39.45
-39.56
-39.54
-39.46
-39.27
-39.39
-39.35
-39.30
-39.22
-39.13
-39.05
-38.88
-38.65
-38.77
-38.44
-38.17
-38.21
-38.06
-37.94
-37.67
-37.57
-37.48
-37.14
-37.08
-36.83
-36.68
-36.40
-36.18
-36.14
-35.98
-35.85
-35.70
d-D
0.08
0.09
0.09
0.09
0.09
0.10
0.10
0.09
0.10
0.10
0.10
0.09
0.10
0.10
0.10
0.10
0.10
0.10
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.08
0.08
0.09
0.09
0.08
0.08
0.08
0.08
0.09
0.09
0.09
0.08
0.09
0.08
0.09
Beam No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Q-D
-37.11
-37.67
-38.30
-38.90
-39.44
-39.91
-40.35
-40.78
-41.33
-41.64
-41.87
-42.44
-42.87
-43.20
-43.55
-43.77
-43.99
-44.01
-44.29
-44.43
-44.44
-44.52
-44.55
-44.49
-44.48
-44.43
-44.31
-44.04
-44.16
-44.00
-43.72
-43.65
-43.45
-43.52
-43.10
-42.88
-42.74
-42.58
-42.57
-42.83
Ping-200
My-D
-37.24
-37.80
-38.42
-39.03
-39.57
-40.04
-40.48
-40.90
-41.46
-41.77
-41.99
-42.56
-42.99
-43.32
-43.67
-43.89
-44.10
-44.12
-44.40
-44.54
-44.55
-44.62
-44.65
-44.59
-44.58
-44.53
-44.41
-44.14
-44.25
-44.09
-43.81
-43.74
-43.54
-43.61
-43.19
-42.97
-42.83
-42.67
-42.66
-42.92
d-D
0.13
0.13
0.12
0.13
0.13
0.13
0.13
0.12
0.13
0.13
0.12
0.12
0.12
0.12
0.12
0.12
0.11
0.11
0.11
0.11
0.11
0.10
0.10
0.10
0.10
0.10
0.10
0.10
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
0.09
143
Appendix D - Program Validation Results
Depth Comparison
Beam No
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Q-D
-35.48
-35.33
-35.18
-35.04
-34.92
-34.90
-34.69
-34.45
-34.28
-34.33
-34.05
-33.93
-33.82
-33.65
-33.45
-33.28
-33.17
-33.11
-32.84
-32.58
-32.33
-32.13
-31.87
-31.59
-31.26
-30.96
-30.70
-30.29
-29.93
-29.53
-29.06
-28.56
-28.19
-27.79
-27.27
-26.70
-26.15
-25.51
-24.93
-24.20
Ping-01
My- D
-35.57
-35.42
-35.27
-35.13
-35.02
-34.99
-34.79
-34.55
-34.38
-34.44
-34.16
-34.04
-33.93
-33.76
-33.56
-33.39
-33.29
-33.23
-32.97
-32.70
-32.46
-32.26
-31.99
-31.72
-31.39
-31.10
-30.83
-30.42
-30.06
-29.65
-29.18
-28.68
-28.31
-27.91
-27.38
-26.81
-26.26
-25.62
-25.05
-24.31
d-D
0.09
0.09
0.09
0.09
0.10
0.09
0.10
0.10
0.10
0.11
0.11
0.11
0.11
0.11
0.11
0.11
0.12
0.12
0.13
0.12
0.13
0.13
0.12
0.13
0.13
0.14
0.13
0.13
0.13
0.12
0.12
0.12
0.12
0.12
0.11
0.11
0.11
0.11
0.12
0.11
Beam No
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
Q-D
-42.08
-41.91
-41.71
-41.44
-41.36
-41.15
-40.86
-40.67
-40.43
-40.25
-40.10
-39.68
-39.51
-39.07
-38.71
-38.55
-38.43
-38.23
-38.10
-37.85
-37.64
-37.39
-37.06
-36.73
-36.36
-35.96
-35.68
-35.26
-34.85
-34.47
-34.03
-33.61
-33.17
-32.59
-31.98
-31.56
-30.71
-29.96
-29.13
-28.33
Ping-200
My-D
-42.17
-42.00
-41.81
-41.54
-41.46
-41.24
-40.96
-40.78
-40.53
-40.35
-40.20
-39.79
-39.62
-39.18
-38.83
-38.66
-38.54
-38.35
-38.22
-37.97
-37.76
-37.51
-37.18
-36.85
-36.48
-36.08
-35.79
-35.38
-34.96
-34.58
-34.13
-33.71
-33.26
-32.67
-32.06
-31.74
-30.92
-30.16
-29.34
-28.43
d-D
0.09
0.09
0.10
0.10
0.10
0.09
0.10
0.11
0.10
0.10
0.10
0.11
0.11
0.11
0.12
0.11
0.11
0.12
0.12
0.12
0.12
0.12
0.12
0.12
0.12
0.12
0.11
0.12
0.11
0.11
0.10
0.10
0.09
0.08
0.08
0.08
0.11
0.10
0.11
0.10
144
Appendix E - Publications
1). Joint International symposium and exhibition on geoinfromation and
International Symposium on GPS/GNNS 2007. November 2007, Johor Bahru,
Malaysia
AN APPRAISAL OF MULTIBEAM ECHO-SOUNDER CALIBRATION
Mohd Razali Mahmud
Gunathilaka M.D.E.K.
Kelvin Tang Kang Wee
Hydrographic Research and Training Office
Department of Geomatic Engineering
Faculty of Geoinformation Science and Engineering
Universiti Teknologi Malaysia
81310 UTM Skudai, Johor
Malaysia
ABSTRACT
One of the most impressive Hydrographic technique developed over the past
few decade is Multibeam sonar systems. Sounding data from these systems is a
result of processing information from several data sources. Among them, positional
data from Global Positioning System (GPS), vessels heading and attitude data from
gyro and motion sensor systems, vertical reference data from tide gauge and sound
speed data, in addition to the multibeam data itself. There must be a good
coordination between these systems in order to obtain reliable data. To determine
this, a proper and thorough field calibration procedure has to be carried out on the
system as a whole. This process begins with measurement of static offsets between
each sensor system with reference to a fixed point on the vessel. Preferably the
point of centre of gravity (CoG). Then the patch test is carried out to determine the
mounting offsets and GPS latency and lastly a performance test to verify whether
the data meet the accuracy requirements for the survey. This is achieved through a
comparison of data with a reference surface. This paper discusses the theoretical
aspects, steps involved and results of the calibration procedures for Multibeam
sonars, using RESON SeaBat 8124 multibeam system. Finally a summary of
multibeam sonar calibration criteria is also presented showing the methodology
involve which include when to perform each test and applying corrections.
145
Appendix E - Publications
2). Joint International symposium and exhibition on geoinfromation and
International Symposium on GPS/GNNS 2007. November 2007, Johor Bahru,
Malaysia
THE EFFECTS OF INADEQUATE SURFACE SOUND SPEED
MEASUREMENTS IN MULTIBEAM ECHOSOUNDER SYSTEMS
Mohd Razali Mahmud
Gunathilaka M.D.E.K.
Hydrographic Research and Training Office
Department of Geomatic Engineering
Faculty of Geoinformation Science and Engineering
Universiti Teknologi Malaysia
81310 UTM Skudai
Johor, Malaysia
ABSTRACT
In any type of survey work, variation of the characteristics of the medium
through which the measurements are made is a challenge, thus; having serious
effects on the accuracy of the measurements. In hydrographic surveying the effects
is even greater when using sonar techniques. The single most important acoustical
variable in the water is its speed. Average speed of sound in the ocean is 1500m/s.
But its precise value in a location is strongly depending on temperature, pressure
and salinity of that particular location. These factors changes rapidly in time and
space due to various reasons like solar heating, evaporation, precipitation, fresh
water inflow etc… and water movements like tides, currents and wave actions. In
data acquisition, the collection of these dense sound speed data becomes critical.
These inadequate sound speeds create unknown propagation through the water
column that adds a major uncertainty to Multibeam echo sounder measurements
(MBES). There are two types of sound speed measurements made in Multibeam
sonars. Surface Sound Speed (SSS) measured at the face of the transducer and
sound speed profile (SSP) through the water column. SSS used to determine the
beam pointing angle and SSP used to determine the depth and position of each
beam. This paper explains the necessity of the surface sound speed in multibeam
sonars and effects generated by inadequate SSS measurements using real data from
RESON SeaBat 8124 Multibeam system. When the vessel roll is significant, the roll
modulate the errors induced by erroneous SSS measurements. These errors are
illustrated in relation to the IHO standards.