Professional Documents
Culture Documents
Bathymetricmappingusingsatelliteimage PDF
Bathymetricmappingusingsatelliteimage PDF
Chandan Roy
2003
Rajshahi University
MAPPING COASTAL BATHYMETRY OF THE
UPPER BAY OF BENGAL USING SATELLITE’S
OPTICAL RADIANCE
by
Chandan Roy
Approved by :
Professor Dr. Raquib Ahmed
Date : ……………………
Rajshahi University
...To my father and mother…
…with all my pride…
v
ABSTRACT
vi
TABLE OF CONTENTS
Declaimer............................................................................................................................. iii
Acknowledgement ............................................................................................................. iv
Dedication ............................................................................................................................ v
Abstract................................................................................................................................ vi
Table of Contents..............................................................................................................vii
List of Figures ..................................................................................................................... xi
List of Photographs ..........................................................................................................xv
List of Tables.....................................................................................................................xvi
1. INTRODUCTION..................................................................................................... 1
1.1. Research objectives ............................................................................................. 2
1.2. Hypotheses to be tested ..................................................................................... 3
1.3. Data and materials............................................................................................... 3
1.4. Method used......................................................................................................... 7
1.4.1. Introduction................................................................................................ 7
1.4.2. Research stages........................................................................................... 8
1.4.2.1. Preparation ..................................................................................... 8
1.4.2.2. Processing and description........................................................10
1.4.2.3. Mapping and analysis .................................................................10
1.4.2.4. Evaluation and reporting...........................................................10
2. STUDY AREA...........................................................................................................12
2.1. Study Area: Upper Bay of Bengal....................................................................12
2.1.1. Geographical location and settings ......................................................12
2.1.1.1. Hydrological conditions ............................................................16
2.1.1.2. Temperature ................................................................................16
2.1.1.3. Salinity ..........................................................................................18
2.1.1.4. Tides..............................................................................................21
2.1.1.5. Color and water transparency ..................................................22
vii
2.1.1.6. Sea level ........................................................................................23
2.1.1.7. Ocean current .............................................................................23
2.1.2. Bottom topography.........................................................................................24
2.1.2.1. Continental shelf.........................................................................26
2.1.2.2. Swatch of no ground .................................................................27
2.1.2.3. Ninety east ridge.........................................................................28
2.1.2.4. Eighty five ridge..........................................................................29
2.1.2.5. Bengal deep sea fan....................................................................29
viii
4.4.2.2. Atmospheric correction ............................................................57
4.4.2.3. Positional registration ................................................................58
4.4.2.4. Oceanographic sampling for "sea truth" ...............................58
4.4.2.5. Image processing........................................................................60
4.4.2.6. Oceanographic applications of satellite
remote sensing............................................................................60
4.4.2.6.1. Visible wavelength “ocean color”
sensor ..........................................................................60
4.4.2.6.2. Sea surface temperature from
infrared scanning radiometers ................................61
4.4.2.6.3. Passive microwave radiometers .............................61
4.4.2.6.4. Satellite altimetry of sea surface
topography .................................................................62
4.4.2.6.5. Active microwave sensing of
sea-surface roughness................................................62
4.4.3. Marine observing satellites......................................................................63
4.4.3.1. CZCS ............................................................................................63
4.4.3.2. MOS..............................................................................................65
4.4.3.3. SeaWiFS .......................................................................................67
ix
5.3.3.2. Lineariseing the relationship.....................................................91
5.3.3.3. Calculating the ratio ...................................................................92
5.3.3.4. Generation of depth
invariant indices .........................................................................93
5.3.4. Implementation ........................................................................................96
5.4. Data correction ...................................................................................................97
5.5. Satellite data and 3D model............................................................................107
6. CONCLUSION ......................................................................................................117
6.1 Causes of error in the result ...........................................................................117
6.1.1. Turbidity .................................................................................................120
6.1.2. Tide..........................................................................................................121
6.1.3. Seasonal variation of water level ........................................................121
6.1.4. Wave........................................................................................................121
6.1.5. Depth of water sample collection......................................................122
6.1.6. Depth of water ......................................................................................123
REFERENCES...............................................................................................................125
APPENDIX……… . ...................................................................................................129
Appendix A. Summery of image geo registration…………………………….129
x
LIST OF FIGURES
Number Page
Figure 1.1 BIWTA echo sound chart of the Bay of Bengal……………... 6
Figure 1.2 Satellite digital data of Landsat ETM+ (20 jan 2001)
of the Bay of Bengal………………………………………… 7
Figure 1.3 Flowchart of the Research Methodology………………………9
Figure 1.4 Flowchart of the Landsat ETM image and BIWTA
sound chart processing……………………………………… 10
Figure 2.1 Bangladesh, Bay of Bengal and part of the Indian
Ocean………………………………………………………. 13
Figure 2.2 Study area…………………………………………………… 14
Figure 2.3 Upper coastal regions of the Bay of Bengal and major
rivers of Bangladesh………………………………………... 15
Figure 2.4 Vertical distribution of temperature in the Bay of
Bengal……………………………………………………… 17
Figure 2.5 Distribution of the surface salinity of the Bay in
Summer…………………………………………………….. 19
Figure 2.6 Distribution of the surface salinity of the Bay in
Winter………………………………………………………. 19
Figure 2.7 Vertical distribution of salinity in the Bay of Bengal…………. 20
Figure 2.8 Bottom relief of the Bay of Bengal…………………………... 25
Figure 2.9 Hypsographic/hypsometric curves…………………………... 26
Figure 2.10 Depth zones and the Swatch of no ground of the
Bay of Bengal……………………………………………….. 28
Figure 2.11 Location of the Ninety east ridge……………………………. 30
Figure 3.1 Volume reflectance spectra for various suspended
matter concentrations in a water column…………………….. 33
Figure 4.1 Electro magnetic Remote Sensing of earth resources………... 42
Figure 4.2 The electromagnetic spectrum……………………………… 44
xi
Figure 4.3 Atmospheric attenuation of electromagnetic
energy and transmission windows…………………………… 45
Figure 4.4 Basic interactions between electromagnetic energy
and an earth surface feature………………………………… 46
Figure 4.5 Major factors influencing spectral characteristics of
a water body………………………………………………... 47
Figure 4.6 Energy loss in water column depth/attenuation
of light with different wavelengths ………………………….. 49
Figure 4.7 Interaction of water with the spectrum……………………… 50
Figure 4.8 Typical spectral reflectance curves for vegetation, soil,
concrete, asphalt and water………………………………... 50
Figure 4.9 Atmospheric pathways of electromagnetic radiation
between the sea and the satellite sensor……………………… 59
Figure 4.10 Advanced Very High Resolution Radiometer
(AVHRR) image of sea surface temperature…………………. 61
Figure 4.11 Spectral reflectance of different remote sensing objects……… 66
Figure 4.12 Pigment and sediment concentration in the Ganges
estuary region of the Bay of Bengal, viewed with
MOS sensor………………………………………………… 66
Figure 5.1 Work flow chart…………………………………………….. 69
xii
along 89°15′E……………………………………………….. 75
Figure 5.7 Study area divided into 8 sub frames………………………… 76
Figure 5.8 3D view of the sea bottom relief of sub frame 1……………... 77
Figure 5.9 3D view of the sea bottom relief of sub frame 2 ……………. 78
Figure 5.10 3D view of the sea bottom relief of sub frame 3……………... 79
Figure 5.11 3D view of the sea bottom relief of sub frame 4…………….. 80
Figure 5.12 3D view of the sea bottom relief of sub frame 5…………….. 81
Figure 5.13 3D view of the sea bottom relief of sub frame 6…………….. 82
Figure 5.14 3D view of the sea bottom relief of sub frame 7…………….. 83
Figure 5.15 3D view of the sea bottom relief of sub frame 8…………… 84
Figure 5.16a Location of profile in the study area…………………………. 85
Figure 5.16b Pattern of slope in the blue band image
before (upper) and after (lower) contraction
(row and column reduction). Profile drawn
along 90°20′E………………………………………………. 86
Figure 5.17 Differential attenuation of the four wavebands
in the water column…………………………………………. 87
Figure 5.18 Processes of water column correction, showing the
steps involved in creating depth-variant indices
of bottom type for sand and sea grass ……………………….92
Figure 5.19 Bi-plot of log-transformed CASI bands 3 and 4.
Data obtained from 348 pixels of sand with variable
depth from 2-15 meter……………………………………… 95
Figure 5.20 Distribution pattern of the suspended sediments
in the study area. Water column collection sample
locations are also shown in the image using dots…………….. 99
Figure 5.21a Pattern of the distribution of the amount of
suspended load in sea water………………………………... 102
Figure 5.21b Pattern of the distribution of the suspended
load size in sea water……………………………………….. 103
Figure 5.22 Contour lines of total signal decay. The corresponding
xiii
blur figures are representing the total signal decay
(in DN) which has been used in generating continuous
surface of total signal decay ………………………………... 105
Figure 5.23 Relation between image and the actual depth………………. 106
Figure 5.24 3D image of the whole study area………………………….. 108
Figure 5.25 Corrected image divided into 8 sub frames…………………. 109
Figure 5.26 Simulated sea floor relief generated from
satellite image (sub frame 1)………………………………... 109
Figure 5.27 Simulated sea floor relief generated from
satellite image (sub frame 2)……………………………….. 110
Figure 5.28 Simulated sea floor relief generated from
satellite image (sub frame 3)………………………………... 111
Figure 5.29 Simulated sea floor relief generated from
satellite image (sub frame 4)………………………………... 112
Figure 5.30 Simulated sea floor relief generated from
satellite image (sub frame 5)………………………………... 113
Figure 5.31 Simulated sea floor relief generated from
satellite image (sub frame 6)………………………………... 114
Figure 5.32 Simulated sea floor relief generated from
satellite image (sub frame 7)………………………………... 115
Figure 5.33 Simulated sea floor relief generated from
satellite image (sub frame 8)………………………………... 116
Figure 6.1 Location of profile in the study area………………………... 118
Figure 6.2 Pattern of slope in the BIWTA sound
generated DEM and corrected satellite image.
Profile along AB…………………………………………… 118
Figure 6.3 Pattern of slope in the BIWTA sound
generated DEM and corrected satellite image.
Profile along CD…………………………………………... 119
Figure 6.4 Pattern of slope in the BIWTA sound
generated DEM and corrected satellite image.
xiv
Profile along EF…………………………………………… 119
Figure 6.5 Pattern of slope in the BIWTA sound
generated DEM and corrected satellite image.
Profile along GH…………………………………………... 119
Figure 6.6 Effect of turbidity upon spectral properties of water………... 120
Figure 6.7 Spectra of calm and wind-roughed water surfaces…………... 122
Figure 6.8 Pattern of slope in the BIWTA sound
generated DEM and corrected satellite image.
Profile along IJ…………………………………………….. 123
Figure 6.9 Pattern of slope in the BIWTA sound
generated DEM and corrected satellite image.
Profile along KL…………………………………………… 124
LIST OF PHOTOGRAPHS
Number Page
Photograph 5.1 Collection of sea water………………………………….. 98
Photograph 5.2 The vessel used for water collection……………………... 99
Photograph 5.3 Microscopic view of the suspended sediment
at water collection location 21°44/ N 90°05/ E.
Magnified 900x………………………………………… 100
Photograph 5.4 Microscopic view of the suspended sediment
at water collection location 21°40/ N 90°05/ E.
Magnified 900x………………………………………… 100
Photograph 5.5 Microscopic view of the suspended sediment
at water collection location 21°24/ N 90°04/ E.
Magnified 900x ………………………………………... 101
Photograph 5.6 Microscopic view of the suspended sediment
at water collection location 21°20/ N 90°04/ E.
Magnified 900x………………………………………… 101
xv
LIST OF TABLES
Number Page
Table 1.1 Characteristics of different sensors for visible regions………… 5
Table 2.1 Tidal levels at the coastal tide gauging stations……………… 21
Table 2.2 Tidal levels at the coastal tide gauging stations
on 20 january 2001……….………………………………….. 22
Table 4.1 Landsat MSS bands…………………………………………. 53
Table 4.2 Landsat TM bands………………………………………….. 54
xvi
Chapter 1
INTRODICTION
2
¾ To develop a method to use the satellite data to generate a three
dimensional model of the sea bottom and check it’s validity.
The general hypotheses which will be tested in this research are the following:
¾ In the continental shelf region of the Bay of Bengal the slope is gentle
and it is a huge submarine fan
¾ Swatch of no ground is located at the south west part of the study area
¾ The amount and size of the suspended sediment declines from the
coast towards the deep sea
¾ There may be a relation between the depth and the reflectance pattern
Two important and relevant data which have been used in this research are the
BIWTA echo sound chart of the Bay of Bengal (Figure 1.1) and the raw satellite
digital data of Landsat ETM+ (20 Jan 2001) of the Bay of Bengal (Figure 1.2).
The first BIWTA echo sound survey of the Bay of Bengal was carried out in
1980. Of course Bangladesh Navy maintains a similar program of bathymetric
survey of their own since recently but it is not available for general use.
However, it takes a long time to cover such a wider coverage of the bay - from
3
about 150 N up to the coast. The bay is also considered very unfriendly and
hostile during about 9 months from March to November due to high weave.
The big waves are principally because of the funnel shape of the bay’s northern
part. The bay is also known for frequent visit of tropical cyclones originated
from the Indian Ocean.
The potential usable satellite data are collected by various satellites such as
Landsat series, IRS series, SPOT series etc. But what is most important is the
spectral coverage of the satellites as well as the temporal resolution. As the area
coverage is significantly wide. The lower spatial resolution (even up to 1 km)
impacts little to view the features. Whereas, higher spectral resolution may be
better to separate different features more correctly. Temporal resolution will
give better situation in examining time series analysis – which is particularly
important for the present research. The table below gives comparative
characteristics of some satellite sensors for visible parts. It is important to note
that blue spectrum region of Landsat 7 occupies most upper part of the visible
area in compared to other satellite sensors (Table 1.1) TM channel blue having
spectrum width of 0.45 μm to 0.52 μm was found to be the most suitable.
Among other visible spectrums the blue has the maximum water penetration
capacity of up to 20m (Lillesand and Kiefer, 2002) due to its shorter wave
length but susceptible to back scattering (Rayleigh’s effect) due to the presence
of smaller suspended particles. Also, availability of Landsat data is easier and
cheaper than all others. There is of course a better option- the MODIS data. It’s
bandwidth is shorter and much better. Band 10 of MODIS satellite having a
bandwidth between 0.483 and 0.493 µm can provide much better bathymetric
maps (described in detail in chapter 4). Major problem incorporating MODIS in
present research was it’s radiometric resolution of 12 bit, which was unable to
be processed due to software limitation. However for main bathymetric data
generation for the present research the data of Landsat ETM+ blue channel
(single) of Dec 25th 2001 was used. The raw data was atmospherically and
4
radiometrically corrected from its source. For calibration purposes of the
backscattering in the water column 2 liters of water samples were colleted for
pre-selected 10 locations located every 10 km from the Bangladesh coast at 900
05' east longitude. To reach the pre-selected locations hand GPS (Magellan
2000XL) was used. The error rate of ±30 m of the GPS was acceptable because
of such a very big area.
5
6
Figure 1.1: BIWTA echo sound chart of the Bay of Bengal
Figure 1.2: Satellite digital data of Landsat ETM+ (20 jan 2001)
of the Bay of Bengal
7
background variation and back scattering, c) amount of suspended materials in
the sea water etc. A remarkable matter here is that all the relevant researches
were more or less first of its kind because this is a newly flourishing field of
application of remotely sensed data. As the work is first in Bangladesh of it’s
kind so the previous works were used as main reference.
1.4.2.1 Preparation
This stage composed of activities such as literature review, proposal finalization,
collection of satellite images and BIWTA bathymetric chart, and locating the
probable points on the bathymetric chart from where water samples can be
collected. This stage was done at the laboratory of the Department of
Geography and Environmental Studies, Rajshahi University.
Literature review
This activity was done transversally throughout the entire research process. It
includes the bibliographic studies from journals and books concerning the
relevant research topic. Literature review has been carried out in order to
develop the knowledge on scientific and technical aspects. Methodology
development for bathymetric mapping from the satellite image has been the
main subject of this stage. After a systematic review of different literature
8
source, some methods for mapping coastal bathymetry were found.
Collection of the image and BIWTA bathymetric chart
These two elements can be considered as the raw materials of the study so these
are collected from the relevant authorities.
Locating the water sample collection points
Before collecting the water sample from the sea some points have been selected
on the map to get greater advantages at the time of collecting the samples.
Identification of problems
Preparation Stage
Literature Collection of image Locating the water
review and bathymetric chart sample collection points
9
1.4.2.2 Processing and description stage
This stage includes all the stages of processing the image and the BIWTA sound
chart (Figure 1.4). Later sediment size and amount has been measured for
calibration purpose.
3D image generation
Bathymetric map
10
sensing images (Landsat ETM) for studying bathymetry. Except this a
comparative study has been done between the image generated 3D model and
the BIWTA sound chart generated DEM. The present report, including maps is
the final result of this thesis work.
11
Chapter 2
STUDY AREA
The study area covers the upper part of the Bay of Bengal. Basically Bay of
Bengal a northern extended arm of the Indian Ocean and is located between
latitudes 5°N and 22°N and longitudes 80°E and 100°E(Figure2.1). As remotely
sensed data is only suitable for shallow coastal waters only the upper part of the
Bay of Bengal has been selected as the study area. The study area is located
between 20°N and 22°N latitudes and 89°7′E and 91°20′E longitudes
(Figure:2.2) covering an area of about 32400 square kilometers. The Bay of
Bengal is bounded in the west by the east coasts of Sri Lanka and India, on the
north by the deltaic region of the Ganges-Brahmaputra-Meghna river system,
and on the east by the Myanmar peninsula extended up to the Andaman-
Nicobar ridges. The southern boundary of the Bay is approximately along the
line drawn from Dondra Head in the south of Sri Lanka to the north tip of
Sumatra. The Bay occupies an area of about 2.2 million sq km and the average
depth is 2,600m with a maximum depth of 5,258m. Bangladesh is situated at the
head of the Bay of Bengal (Figure: 2.3).
12
Figure 2.1: Bangladesh, Bay of Bengal and part of the Indian Ocean
13
14
Figure 2.2: Study area
Figure 2.3: Upper coastal regions of the Bay of Bengal and major rivers of Bangladesh
15
2.1.1.1 Hydrological conditions
2.1.1.2 Temperature
As the bay is surrounded by land mass from three sides so the land mass has a
great impact upon the water temperature of the bay. The temperature of about
two third water of northern portion of the bay remains between 25°C and 28°C
from December to March. From April the temperature of the bay starts to
increase. The maximum temperature is observed in May (30°C). But in July the
temperature is reduced and remains same till September. In October the
temperature reduces again and in January the lowest temperature is seen and the
minimum temperature is 25°C. The mean annual temperature of the surface
water is about 28°C. But the annual variation in temperature is not large, about
2°C in the south and 5°C in the north. In the bay water has a inverse
relationship with depth of water, that is if the depth increases the temperature
16
decreases. Average vertical distribution of temperature of the bay in given in
Figure 2.4.
Temperature (°C)
Depth in meters
Temperature
INDEX
Summer temperature
Winter temperature
17
2.1.1.3 Salinity
Bay of Bengal is unique in the world in terms of salinity. As some large rivers of
the world have fallen into the bay so the salinity of the surface water of the bay
is less saline than other seas of the world. Except this seasonal variation in
salinity is seen in the bay, this causes mainly due to the variation in rainfall
seasonally. In rainy season when rainfall is highest then the water discharge
from the river increases also and due to this huge amount of discharge in the
monsoon sometimes at the estuary the salinity becomes 0. The surface salinity
in the open part of the Bay oscillates from 32‰ to 34.5‰ (parts per thousand,
i.e. grams per kilogram of sea water) and in the coastal region varies from 10‰
to 25‰. But at the river mouths, the surface salinity decreases to 5‰ or even
less. The coastal water is significantly diluted throughout the year, although the
river water is greatly reduced during winter. Along the coast of the Ganges-
Brahmaputra Delta, salinity decreases to 1‰ during summer (Figure 2.5) and
increases up to 15‰ to 20‰ in winter (Figure 2.6). Salinity gradually increases
from the coast towards the open part of the Bay and near the coast the seasonal
variation in salinity is greatest when in the deep sea this variation is very less.
The surface salinity at the mouths of some large rivers like the Ganges,
Brahmaputra, Irrawaddy and some Indian rivers like the Krishna, Godavari,
Cauvery and Mahanadi varies widely from one day to another, especially in
summer. Salinity of water also changes vertically (Figure 2.7). The influence of
the fresh water is experienced up to depths of 200-300m. From the surface, the
salinity gradually increases downward and at about 200-300m it reaches 35‰
and at about 500m the salinity is more than 35.10‰, but at 1,000m it decreases
slightly and attains 34.95‰. With further increase of depth salinity decreases
and at 4,500m it is close to 34.7‰ (Banglapedia CD-ROM edition, Version-1)
18
Salinity in ‰
Bay of Bengal
Salinity in ‰
Bay of Bengal
19
Salinity (‰)
Salinity
Depth in meters
INDEX
Summer salinity
Winter salinity
20
2.1.1.4 Tides
In the Bay the tide is semi-diurnal in nature, i.e. two high and two low tides
during the period of 24 hours and 52 minutes. The highest tide is seen where
the influence of bottom relief and the configuration of the coast are prominent,
i.e. in shallow water and in the Bay and estuary. The average height of tidal
waves at the coast of Sri Lanka is 0.7m and in the deltaic coast of the Ganges it
is 4.71m (due to funnel effect). In the Bay of Bengal tidal currents specially
develop in the mouths of the rivers, like the Hooghly and the Meghna . Tidal
levels at the coastal tide gauging stations are given in table 2.1 and tidal levels at
those stations on 20 january 2001 are given in table 2.2.
21
Table 2.2: Tidal levels at the coastal tide gauging stations on
20 January 2001
The color of the water in the open part of the Bay is dark blue which gradually
changes to light blue to greenish towards the coast. Transparency is high, 40-
50m in some places. In the central part of the Bay of Bengal, the anticyclone
circulation is generated and the zone of convergence lies in the center of this.
This region is characterized by high water transparency. Regions of low
transparency and turbid water are available in the limited area of the pre-deltaic
part of the rivers Ganges and Brahmaputra. The absorption and scattering of
the light by the water depends upon the suspended and dissolved materials in
the water. These elements may be organic or inorganic in nature.
22
On the basis of transparency of water Bay of Bengal can be divided in the
following three regions:
Due to the influence of water density and wind the seasonal changes of the sea
level in the Bay are remarkable and one of the highest in the world. The range
of sea level change at Khidirpur is 166 cm, at Kolkata 130 cm and at Chittagong
118 cm. But towards the southwestern coast at Madras and Vishakhapatnam
[Vishakhapatnam] the range is small compared to the northern and northeastern
coasts of the Bay. The lowest variation of sea level at the southeastern coast of
India is due to its geographical location at the edge of a comparatively deep sea.
23
Up-welling and down-welling are seasonal, being created by monsoon winds
that blow from the southwest during the summer, then reverse direction and
come from the northeast during the winter. The persistence of the monsoon,
especially from the southwest and the orientation of the coasts cause up-welling
to occur along most of the east coast of India. That is why in the east coast of
India the up-welling takes place in summer and down welling in winter, and in
the eastern part of the Bay of Bengal and in the Myanmar coast, up-welling
occurs in winter and the down-welling in summer. However, the duration and
intensity of vertical movement of water on both sides of the Bay of Bengal is
not as great as on the Somalia or North and South American coasts. But it does
have a profound effect on the food economy of the sea through its influence on
chemical properties and biological populations.
24
25
Figure 2.8: Bottom relief of the Bay of Bengal
Source: United States Geological Survey
Most of the features of the bottom topography of the Bay are similar to other
bays and seas of the world. The overall topography of the bay can be discussed
under the following three headlines:
i. Continental shelf
ii. Continental slope, and
iii. Deep sea plains (Figure:2.9)
As the study area covers only the upper part of the Bay that is the continental
shelf region so description of the continental self region has been given here
only.
The width of the continental shelf off the coast of Bangladesh varies
considerably. It is less than 100 km off the south coast between Hiron Point
26
and the swatch of no ground and more than 250 km off the coast of Cox’s
Bazar. Sediments are fine seaward and westward with the thickest accumulation
of mud near the submarine canyon, the Swatch of no Ground. The shallow part
(less than 20m) of the continental shelf off the coast of Chittagong and Taknaf
is covered by sand and the intertidal areas show well-developed sandy beaches.
The shallower part of southern continental shelf off the coast of the
Sundarbans, Patuakhali and Noakhali is covered by silt and clay; and extensive
muddy tidal flats are developed along the shorelines. It is mainly due to the high
sediment yield from the rivers in this region. Some of the shoals and sand ridges
present on this part of the continental shelf show an elongation pattern pointed
towards the Swatch of no Ground. The over all depth of the continental shelf
region is not more than 30 meters and this character of this shelf has supported
us to map the coastal bathymetry with satellite’s optical radiance.
Except these common features it has some unique features also, those are:
It is the most unique feature of the Bay and also known as Ganges Trough.
Swatch of no Ground has a comparatively flat floor 5 to 7 km wide and walls of
about 12° inclination. At the edge of the shelf, depths in the trough are about
1,200m. The Swatch of no Ground has a seaward continuation for almost 2,000
km down the Bay of Bengal in the form of fan valleys with levees (Figure 2.10).
The sandbars and ridges near the mouth of the Ganges-Brahmaputra delta
pointing toward the Swatch of no Ground showing sediments are tunneled
through this trough into the deeper part of the Bay of Bengal. The Swatch of no
Ground is feeding the Bengal Deep Sea Fan by turbidity currents.
27
2.1.2.3 Sunda Trench
It is also known as Java Trench. Running parallel along the west side of the arc
of the Nicobar and Andaman islands it is extended northward up to 10°N into
the Bay and joins the eastern limit of the Himalayan range. It originated
tectonically at the junction of the Indian and Myanmar plates.
Figure 2.10: Depth zones and the Swatch of no ground of the Bay of Bengal
28
Sunda Trench between the Bengal Fan and the Nicobar Fan (Figure 2.11). The
Ninety East Ridge has existed since early in the formation of the Bay of Bengal.
The ridge represents the trace of a hot spot formed during the northward flight
of India and its associated oceanic lithosphere of the Bay of Bengal.
It is a ridge along 85°E longitude. More than 5 km thick sediments have been
deposited on either sides of the ridge. The main turbidity current channel of the
sub aerial drainage pattern lies immediately east of the buried ridge.
The world's largest submarine fan, also known as Bengal Fan. It is 2,800 to
3,000 km long, 830 to 1,430 km wide and more than 16 km thick beneath the
northern Bay of Bengal (Figure 2.10). Sediments are tunnelled to the fan via a
delta-front trough, the Swatch of no Ground. It can be divided into three parts:
upper fan, middle fan and lower fan. Rapid terrigenous sedimentation on an
incipient Bengal fan began in the Eocene age (58 to 37 million years ago) as a
response to the first intraplate collision and continued to the present, building
the world's largest submarine fan.
29
Figure 2.11: Location of the Ninety east ridge
Source: Geological Survey of India
30
Chapter 3
31
all of the non-chlorophyllous matter, phytoplankton and detritus are referred to
“total suspended matter” (TSM). The inorganic fraction of TSM can be formed
from biological sources (e.g. coccolihs), benthic (re-suspension of bottom
sediment) or fluvial origin from river discharge. It is measured by optical
methods that are often difficult to be quantified accurately in terms of weight or
volume. Some researchers have discussed the relationship between suspended
sediment and reflectance. Ritchie et al. (1996) mentioned that the suspended
sediment increases the radiance from surface water in visible and near infrared
ranges of the electromagnetic spectrum. Laboratory measurements have shown
that the surface water radiance is affected by sediment type, texture, color,
sensor view and sun angles, as well as water depth (Ritchie and Schiebe, 1998).
Since the mid 1970’s, remote sensing studies of suspended matter have been
using the data from satellite platforms such as Landsat, SPOT, IRS, Coastal
Zone Color Scanner (CZCS) and SeaWiFS (Sea-viewing Wide Field of View
Sensor). Those studies have shown a significant relationship between suspended
matter and radiance or reflectance from single band or combination of some
bands in satellite or airborne platforms. Ritchie et al. (1976), concluded that the
wavelength between .7µm and .8µm were the most useful range for determining
suspended matter in surface water. Dekker (1993) described that the remote
sensing of water bodies is restricted to a relatively narrow range of optical
wavelength compared to remote sensing of terrestrial object. This is caused by
low solar irradiance at wavelengths shorter than approximately .4µm and by a
combination of lower solar energy and the sharply increasing absorption of light
beyond approximately .85µm. Therefore, the range of .4µm to .85µm is often
used for research aimed at estimation of water quality parameters. Figure: 3.1,
illustrates the impact of suspended matter on volume reflectance spectra, just
beneath the air water interface (Bukata et al., 1995). The impact of suspended
matter on volume reflectance spectra is clearly evident. Even at small
concentrations, suspended matter can substantially increase the volume
32
reflectance in a manner that becomes more pronounced as the wavelength
becomes longer. The absorption of radiance by suspended sediment is generally
much smaller than that of chlorophyll, but the scattering is much higher. An
increase of sediment concentration results in an increase of the backscattering
and hence, an increase in the emergent radiance leaving the water.
33
factors such as fluctuation of river flow, sediment load and phytoplankton. As a
result, data must be acquired at the same time as the overpass of the satellite.
The most common techniques used for analysis of remote sensing data to
determine water quality concentration are based on the brightness of
reflectance. To obtain the water quality concentration from the water leaving
radiance that is detected by the optical sensor, the retrieval algorithms can be
used. Morel and Gordon (1980) pointed out three different approaches: a)
empirical approach, b) semi-empirical approach and c) analytical approach.
34
¾ The relation between inherent optical properties and the optical water
quality parameters.
There are simple and multiple regression equations. These are the subjects of
research done by Ritchie and Cooper (1988), Baban (1993) and Shimoda et al.
(1986). Linear and multiple regressions were proved useful for the study of the
suspended sediment. They yielded sufficiently accurate concentration
estimations. They gave better accuracy if the measurement is at the same time as
the acquisition date of remotely sensed imagery.
35
obtaining sediment map in coastal regions. Spectral mixture analysis, as a data
analysis tool, is done using a fixed reference (end-members). The end-members
are represented by spectral data from either the purest pixel of a specific
material on an image or the purest material in the laboratory (Metres et al.,
1991). He proved that spectral mixture analysis is a powerful tool for estimating
suspended sediment concentration in the surface waters. The “neural network”
can be applied to define the transfer function between the chlorophyll or
sediment concentration and the satellite receiver radiance (Keiner and Xiao-Hai
Yan, 1997). It was found that a neural network using three visible bands of
Landsat TM as input has been successful in modeling the water quality
parameters.
36
¾ The air water system can be divided into subsystems, for which the
separate model and inversion procedures can be developed and
improved more easily;
¾ It allows the analysis of error propagation, which enables us to predict
errors in retrieved concentration;
¾ It can be adapted to other spectral bands;
¾ Only initial measurement is needed to establish optical properties of the
relevant waters in an area, require little measurement; this approach is
cost effective and optimizes the use of archives images
37
channels for suspended sediment studies showed similar characteristics (Schiebe
et al. 1992, Choubey and Subramaniam 1992). The trend suggests that the non-
linear relation is preferred by the data set. The single band method was found to
be less accurate. Generally the accuracy increased when more spectral bands
and higher order series were included in the regression analysis.
Bathymetric mapping with the satellite data is a very recent field of application
of the satellite data. As it is a quietly new field of application so literatures
regarding this are not so available. A few literatures which are available are not
suitable for all the coastal waters of the world, which have been proved by the
adoption of different calibration techniques for different regions. That is why
any literature could not be followed uniquely.
A valuable work on bathymetric charting was done at the Penang Strait in
Malaysia where the signal reflectance data were corrected (and compared too)
using sound signals by K. Abdullah, M.Z. MatJafri and Z.B.Din. They
conducted a survey to measure the new sounding points using a boat equipped
with an echo sounder and the sounding locations were determined with a GPS
system. Landsat TM and SPOT data acquired between January 1997 and
February 1997 were used by them for the study. Image locations were related to
the map GCP coordinates through the second degree polynomial
transformation equations. The pixel values of the same locations were extracted
and were used as independent variables and the measured sounding points as
dependent variables. In the study multiband water depth algorithm was used in
the calibration analysis. Regression techniques were used for calibration of the
satellite signals for water depth measurement. From the regression equation
they examined the correlation coefficient and root-mean-square deviations for
each data set. Later the accuracy of each calibration algorithm was further
38
verified using other known points. At last the calibration algorithm was applied
to the corresponding image to generate water depth map.
Philpot tried to develop a spectral analysis tool with the hyperspectral image
data that can be used to detect ocean color and water quality, extract
bathymetry, and bottom type information. Their main objectives were to
develop specific algorithm and procedures to classify water type, differentiate
among different bottom types and extract bathymetry from passive
hyperspectral image data. They indicated that, when the water type and bottom
39
reflectance are uniform over the study area, bathymetric mapping with passive
remote sensing data is a relatively straight-forward, one variable problem and
requires a minimum ground information data. But the physical properties of
water is not same everywhere, it differs from region to region. In that case the
depth can not be determined without simultaneously resolving the bottom
reflectance and basic optical water properties. That is why he suggested to use
more than one band to extract bathymetry from the satellite’s optical radiance
and in this case the hyperspectral images are more effective (opl.ucsb).
Satellite remote sensing techniques can be used together with limited water
depth measurements from conventional methods to chart the coastal areas in a
cost-effective manner (Dr. Seeni Mohd, M.I., Ahmad, S, Yem, M.). This paper
reports on a study to obtain water depths in the coasts! Waters of Pulau
Tioman, Malaysia using the Landsat-5, Thematic Mapper data that were
acquired on 1 April 1990. Band 1 (0.45 – 0.52 µm ) of the data was used since it
has the best depth penetration capability in the relatively clear waters of Pulau
Tioman. They corrected the satellite data for atmospheric effects prior to
computation of water depths with a computer program. An algorithm which
expresses the exponential relationship between water depth and pixel intensities
were used together with a few in-situ calibration depths that were taken at the
time of satellite pass. Comparisons of calculated depths with measured depths
at some check points indicate an error of 0.5 – 2.0 m in depths of up to about
50 m of water. The depth accuracy requirement are 30 cm for depths up to 30
m , 1 m for depths from 30 m to 100 m and 1% of the depth for deeper than
100 m according to the accuracy standards recommended for hydrographic
surveys by the International Hydrographic Organization. The results obtained in
this study and other studies (Ibrahim 1989) indicate that these accuracy
requirements are difficult to achieve by remote sensing techniques. However,
the hydrographical chart derived from the Landsat-5, TM satellite data show
40
many similarities with the corresponding hydrographic chart derived from the
Admiralty hydrographic charts despite the large difference in the dates of field
survey and satellite data acquisition (1960 and 1990). This shows that in areas
where the water clarity is good, satellite data can be used to obtain some general
idea on the depth contours.
In most of the literatures stated in the above paragraphs, hyperspectral satellite
images have been used for coastal bathymetric mapping. Beside this Landsat
TM image has been used also. In this research only the Landsat ETM+ blue
band image has been used. One of the most important causes behind the
selection of this band is it’s greater water penetration capacity. Blue band of
Landsat ETM+ having wavelength between 0.45 µm and 0.52 µm penetrates up
to 20 meters in the clear water. But as those hyperspectral satellite data are very
costly and not easily available. So the blue band of Landsat ETM+ has been
used. Except this the study area: The upper part of Bay of Bengal is a region of
active delta building. So here the amount of suspended sediment is greater in
comparison with the study areas of the above mentioned literatures. Because of
this excessive amount of sediment concentrations in the water a different
calibration technique has been used in this research.
41
Chapter 4
4.1 Introduction
Remote Sensing is the science and art of obtaining information about an object,
area, or phenomena through the analysis of data acquired by a device that is not
in contact with the object, area, or phenomena under investigation. The term
"remote sensing" is itself a relatively new addition to the technical lexicon. It
was coined by Ms Evelyn Pruitt in the mid-1950's when she
(geographer/oceanographer) was with the U.S. Office of Naval Research
(ONR) outside Washington, D.C. In much of remote sensing, the process
involves an interaction between incident radiation and the targets of interest.
This is exemplified by the use of imaging systems where the following nine
elements are involved. Note, however that remote sensing also involves the
sensing of emitted energy and the use of non-imaging sensors.
The generalized processes and elements involved in electromagnetic remote
sensing of earth resources are represented in schematically in Figure 4.1. The
two basic processes involved here are data acquisition and data analysis.
42
The elements of the data acquisition process are:
¾ energy sources
¾ presentation of the information to the users so that they can use it for
their decision making process.
43
Figure 4.2: The electromagnetic spectrum
Source: Lillesand, T.M. and Kiefer, 2002.
A narrow range of EMR extending from 0.4 to 0.7 µm, the interval detected by
the human eye, is known as the visible region (also referred to as light but
physicists often use that term to include radiation beyond the visible). White
light contains a mix of all wavelengths in the visible region
The light which our eyes - our "remote sensors" - can detect is part of the
visible spectrum. It is important to recognize how small the visible portion is
relative to the rest of the spectrum. There is a lot of radiation around us which
is "invisible" to our eyes, but can be detected by other remote sensing
instruments and used to our advantage. The visible wavelengths cover a range
from approximately 0.4 to 0.7 μm. The longest visible wavelength is red and the
shortest is violet. Blue, green, and red are the primary colors or wavelengths of
the visible spectrum. They are defined as such because no single primary color
can be created from the other two, but all other colors can be formed by
combining blue, green, and red in various proportions. Although we see
sunlight as a uniform or homogeneous color, it is actually composed of various
wavelengths of radiation in primarily the ultraviolet, visible and infrared
portions of the spectrum. The visible portion of this radiation can be shown in
its component colors when sunlight is passed through a prism, which bends the
light in differing amounts according to wavelength. The whole portion of the
44
spectrum is not suitable for remote sensing. The sun light before falling upon
the earth’s surface and after being reflected from the earth’s surface has to travel
through the atmosphere. And light while traveling through the atmosphere the
suspended particles of varying size present in the atmosphere causes scattering
effect. Except this effect when the light moves through the atmosphere certain
portion of it is absorbed by ozone, carbon dioxide, and water molecules etc.
which are present in the atmosphere. This effect is called absorption. Those
areas of the spectrum which are not severely influenced by atmospheric
absorption and thus, are useful to remote sensors, are called atmospheric
windows (Figure 4.3)
45
incident on the element are reflected, absorbed, and/or transmitted. Applying
the principle of conservation of energy, we can state the interrelationship
between these three energy interactions as:
46
¾ roughness of the surface
¾ angles of observation and illumination, and
¾ in some extent, reflection of light from the bottom (Figure 4.5)
As incident light strikes the water surface, some is reflected back to the
atmosphere; this reflected radiation carries little information about the water
itself. This portion of light can be used to measure the roughness of the surface,
and therefore, about wind and waves. The spectral properties (i.e., “color”) of a
water body are determined largely by energy that is scattered and reflected
within the water body itself, known as volume reflection because it occurs over a
range of depths rather than at the surface. Some of this energy is directed back
toward the surface, where it again passes through the atmosphere, and then is
recorded by the sensor (Figure 4.5). This light sometimes known as underlight, is
the primary source of color of a water body.
47
The light that enters a water body is influenced by:
¾ absorption and scattering by pure water, and
¾ scattering, reflection, and diffraction by particles that may be suspended
in water.
For the deep water bodies, it is expected (in the absence of impurities) that
water will be blue or blue-green in color. Maximum transmittance of light by
clear water occurs in the range 0.44 to 0.54 µm, with peak transmittance at 0.48
µm. Because the color of water is determined by volume scattering, rather than
surface reflection, spectral properties of water bodies are determined by
transmittance rather than surface characteristics alone. In the blue region the
light penetration is not at it’s optimum, but at the slightly lower wavelengths, in
the blue-green region, penetration is greater and at these wavelengths the
opportunity for recording features on the bottom of the water body are greatest
Longer wavelengths, visible and near infrared radiation is absorbed more by
water than shorter visible wavelengths. Thus water typically looks blue or blue-
green due to stronger reflectance at these shorter wavelengths, and darker if
viewed at red or near infrared wavelengths. If there is suspended sediment
present in the upper layers of the water body, then this will allow better
reflectivity and a brighter appearance of the water. But if the Water body is
relatively free of suspended sediments then the light with shorter wavelengths
(like blue) can penetrate easily up to 20 meters (Figure 4.6), basically this
characteristics of the blue band has made it usable in bathymetric mapping.
48
Blue Green Red Near IR
The apparent color of the water will show a slight shift to longer wavelengths.
Suspended sediment can be easily confused with shallow (but clear) water, since
these two phenomena appear very similar. Chlorophyll in algae absorbs more of
the blue wavelengths and reflects the green, making the water appear more
green in color when algae is present. The topography of the water surface
(rough, smooth, floating materials, etc.) can also lead to complications for
49
water-related interpretation due to potential problems of specular reflection and
other influences on color and brightness (Figure 4.7).
Like the water bodies each earth surface features has it’s different reflectance
pattern (Figure 4.8) and this difference in reflectance is the key point to identify
objects through remote sensing.
50
4.4 Observing the earth’s surface through satellites
The earth’s surface features are not unique so to collect data of different earth
surface features different types of satellites have been lunched. Some satellites
are useful for land, some for weather, some for ice and some for the oceanic
and coastal areas. Brief descriptions of some satellites and their comparative
advantages and disadvantages in different fields have been given in the
following paragraphs:
51
polar, sun-synchronous orbits. The first three satellites (Landsats 1-3) are at
altitudes around 900 km and have revisit periods of 18 days while the later
satellites are at around 700 km and have revisit periods of 16 days. All Landsat
satellites have equator crossing times in the morning to optimize illumination
conditions.
The MSS senses the electromagnetic radiation from the Earth's surface in four
spectral bands. Each band has a spatial resolution of approximately 60 x 80
meters and a radiometric resolution of 6 bits, or 64 digital numbers. Sensing is
accomplished with a line scanning device using an oscillating mirror. Six scan
lines are collected simultaneously with each west-to-east sweep of the scanning
mirror. The accompanying table outlines the spectral wavelength ranges for the
MSS (Table 4.1).
52
Table 4.1: Landsat MSS bands
Routine collection of MSS data ceased in 1992, as the use of TM data, starting
on Landsat 4, superseded the MSS. The TM sensor provides several
improvements over the MSS sensor including: higher spatial and radiometric
resolution; finer spectral bands; seven as opposed to four spectral bands; and an
increase in the number of detectors per band (16 for the non-thermal channels
versus six for MSS). Sixteen scan lines are captured simultaneously for each
non-thermal spectral band (four for thermal band), using an oscillating mirror
which scans during both the forward (west-to-east) and reverse (east-to-west)
sweeps of the scanning mirror. This difference from the MSS increases the
dwell time and improves the geometric and radiometric integrity of the data.
Spatial resolution of TM is 30 m for all but the thermal infrared band which is
120 m. All channels are recorded over a range of 256 digital numbers (8 bits).
The accompanying table outlines the spectral resolution of the individual TM
bands and some useful applications of each (Table 4.2).
53
Table 4.2: Landsat TM bands
The latest satellite of Landsat series is Landsat ETM+. The new features on
Landsat ETM+ are:
¾ a panchromatic band with 15m spatial resolution
¾ on board, full aperture, 5% absolute radiometric calibration
¾ a thermal IR channel with 60m spatial resolution
Data from the ETM+, TM and MSS sensors are used for a wide variety of
applications, including resource management, mapping, environmental
monitoring, and change detection (e.g. monitoring forest clear cutting).
4.4.1.2 SPOT
SPOT (Système Pour l'Observation de la Terre) is a series of Earth observation
imaging satellites designed and launched by CNES (Centre National d'Études
Spatiales) of France, with support from Sweden and Belgium. SPOT-1 was
launched in 1986, with successors following every three or four years. All
54
satellites are in sun-synchronous, near-polar orbits at altitudes around 830 km
above the Earth, which results in orbit repetition every 26 days. They have
equator crossing times around 10:30 AM local solar time. SPOT was designed
to be a commercial provider of Earth observation data, and was the first satellite
to use along-track, or pushbroom scanning technology.
The SPOT satellites each have twin high resolution visible (HRV) imaging
systems, which can be operated independently and simultaneously. Each HRV
is capable of sensing either in a high spatial resolution single-channel
panchromatic (PLA) mode, or a coarser spatial resolution three-channel
multispectral (MLA) mode. Each along-track scanning HRV sensor consists of
four linear arrays of detectors: one 6000 element array for the panchromatic
mode recording at a spatial resolution of 10 m, and one 3000 element array for
each of the three multispectral bands, recording at 20 m spatial resolution. The
swath width for both modes is 60 km at nadir. The accompanying table
illustrates the spectral characteristics of the two different modes (Table 4.3).
The viewing angle of the sensors can be adjusted to look to either side of the
satellite's vertical (nadir) track, allowing off-nadir viewing which increases the
satellite's revisit capability. This ability to point the sensors up to 27° from nadir,
allows SPOT to view within a 950 km swath and to revisit any location several
times per week. As the sensors point away from nadir, the swath varies from 60
to 80 km in width. This not only improves the ability to monitor specific
55
locations and increases the chances of obtaining cloud free scenes, but the off-
nadir viewing also provides the capability of acquiring imagery for stereoscopic
coverage. By recording the same area from two different angles, the imagery can
be viewed and analyzed as a three dimensional model, a technique of
tremendous value for terrain interpretation, mapping, and visual terrain
simulations. This oblique viewing capability increases the revisit frequency of
equatorial regions to three days (seven times during the 26 day orbital cycle).
Areas at latitude of 45º can be imaged more frequently (11 times in 26 days) due
to the convergence or orbit paths towards the poles. By pointing both HRV
sensors to cover adjacent ground swaths at nadir, a swath of 117 km (3 km
overlap between the two swaths) can be imaged. In this mode of operation,
either panchromatic or multispectral data can be collected, but not both
simultaneously.
SPOT has a number of benefits over other spaceborne optical sensors. Its fine
spatial resolution and pintable sensors are the primary reasons for its popularity.
The three-band multispectral data are well suited to displaying as false-color
images and the panchromatic band can also be used to "sharpen" the spatial
detail in the multispectral data. SPOT allows applications requiring fine spatial
detail (such as urban mapping) to be addressed while retaining the cost and
timeliness advantage of satellite data. The potential applications of SPOT data
are numerous. Applications requiring frequent monitoring (agriculture, forestry)
are well served by the SPOT sensors. The acquisition of stereoscopic imagery
from SPOT has played an important role in mapping applications and in the
derivation of topographic information (Digital Elevation Models - DEMs) from
satellite data.
56
4.4.2.2. Atmospheric correction
4.4.2.3. Positional registration
4.4.2.4. Oceanographic sampling for "sea truth"
4.4.2.5. Image processing
4.4.2.6. Oceanographic applications of satellite remote sensing
57
The main strategies of atmospheric correction:
¾ No separate attempt of atmospheric correction, instead we calibrate each
scene with ground data.
¾ An universal atmospheric correction based on an average model of
atmospheric effects.
¾ Using different wavelengths, assuming that certain channels are unlikely
to have any upwelling radiation from the sea. In this way we process
each pixel of the image.
¾ An atmospheric (microwave) sounding sensor can be mounted on the
same satellite as an oceanographic sensor.
58
Sensor
4
1 5
6
2
Sea surface
59
Spatial resolution of the sensor is important as compared with spatial variability
of the measured parameter, because the value measured within a point may not
be representative of the average parameter within the whole pixel measured by
the satellite.
60
concentrations of different kinds of matter suspended in seawater, including
phytoplankton cells.
Infrared sensors measure electromagnetic radiation within the band 1-30 µm,
emitted by the ocean surface and resulting from the temperature of the upper
sea layer. The near-infrared and infrared radiation is processed to sea surface
temperature (SST). The most important SST sensors are Advanced Very High
Resolution Radiometer (AVHRR) on NOAA satellites (Figure 4.10), MODIS,
GOES geostationary satellites (csc.noaa, 1999 and marine.usf, 2004)
Figure 4.10: Advanced Very High Resolution Radiometer (AVHRR) image of sea
surface temperature
Source: Internet (obee.ucla)
61
weather devices. This principle advantage is countered by the fact that thermal
emission is very weak at these longer wavelengths. To overcome noise levels a
large field of view must be received; that results in low spatial resolution (25–
150 km). So, these observations are used for studies of heat balance of the
ocean. The emissivity of the sea at microwave frequencies varies with the
dielectric properties of sea water (including salinity) and the surface roughness.
Hence, the development of this technique in future can enable the
measurements of surface salinity.
62
4.4.3 Marine observation satellites
4.4.3.1 CZCS
63
respectively. Channel 5 sensed reflected solar radiance and had a .1 µm
bandwidth centered at .75 µm and a dynamic range which was more suited to
land. Channel 6 operated in the 10.5 to 12.5 µm region and sensed emitted
thermal radiance for derivation of equivalent black body temperature. The
primary objective of this sensor was to obtain observations of ocean color and
temperature, particularly in the coastal zones, which would provide data with
sufficient spatial and spectral resolution for the following applications:
¾ Measure concentrations of chlorophyll-a and phaeophytin.
¾ Map biologically productive areas.
¾ Map suspended sediment distribution and determine the type of
materials suspended in the water.
¾ Map Gelbstoff (yellow substances) as an indicator of salinity.
¾ Detect pollutants in the upper level of the oceans.
¾ Map temperature of coastal waters and the open ocean.
¾ Study the interactions between coastal effluents and open waters.
The accompanying table outlines the spectral ranges of each band and the
primary parameter measured by each (Table 4.4).
64
4.4.3.2 MOS
65
Figure 4.11: Spectral reflectance of different remote sensing objects
Source: German Aerospace Center
66
The characteristics of the two sensors in the visible and infrared are described in
the accompanying table (Table 4.5).
The MESSR bands are quite similar in spectral range to the Landsat MSS sensor
and are thus useful for land applications in addition to observations of marine
environments. The MOS systems orbit at altitudes around 900 km and have
revisit periods of 17 days.
4.4.3.3 SeaWiFS
67
resolution mode of 4.5 km (at nadir) over a swath of 1500 km. Spectral ranges
of the bands of the Sea WiFS satellite are given below in tabular form (Table
4.6)
These ocean-observing satellite systems are important for global and regional
scale monitoring of ocean pollution and health, and assist scientists in
understanding the influence and impact of the oceans on the global climate
system.
68
Chapter 5
Geo
BIWTA referenc Blue band
point data image
Calibration and
value adjustment Image converted
Regression To DEM
in formula
equation
Compare
69
The data analysis and surface modeling stage was divided into six main stages:
BIWTA 3D map generation, Satellite data processing, water column correction,
Collection of water and measurement of suspended particles in the water,
satellite data and 3D model and validity test. The stages mentioned above will
be presented in the following sections:
The available BIWTA sound data of the Bay of Bengal covers a region from the
Meghna River estuary to about 20.00 S that includes the continental shelf, slope
and a very small part of the basin. The BIWTA survey area includes a very deep
trench called ‘Swatch of No Ground’ at the south-west part. But most of the
area is within20 m depth – that is, within the possible range of optical signal
penetration. The echo-sound data generated locations are quite sufficient and
uniformly distributed to give a very reasonable view of the bathymetry. But
however, only the north-western part (Figure 5.2) of the BIWTA survey area
was included in the present analysis due to the limitation of image processing
speed in the computer.
70
km2) and the bottom relief was found to be significantly gentle (figure 5.6a and
5.6b).
71
Figure 5.3a
72
Figure 5.3b
73
Figure 5.4: Relief generated through interpolation of point data
74
Beginning of the
Swatch of no ground
Figure 5.6b: Pattern of slope in the BIWTA sound generated DEM before (upper) and
after (lower) contraction (row and column reduction). Profile along 89°15′E.
75
To have better observational effect the study area has been divided into 8 pieces
(Figure 5.7) and 3D surfaces have been generated from each small pieces. All
the frames are presented here for better viewing (Figure 5.8 to 5.15). To get a
smoother effect the 3D images were filtered in 5X5 pixel window by low pass
filter. The effect of the view was found to be dramatic in case of the Swatch of
No Ground which is partially present in the frames 1, 2, and 5. But however,
the overall gradual increase of depth was found in the southern part, i.e., after
about 30 km to south and the continental slope begins after about 100 km from
the coast. The satellite image also shows a sharp drop of radiance signal
approximately after about 100 km south of the coast at 900 east longitude which
indicates the possible beginning of the continental slope.
76
Figure 5.8: 3D view of the sea bottom relief of sub frame 1
77
Figure 5.9: 3D view of the sea bottom relief of sub frame 2
78
Figure 5.10: 3D view of the sea bottom relief of sub frame 3
79
Figure 5.11: 3D view of the sea bottom relief of sub frame 4
80
Figure 5.12: 3D view of the sea bottom relief of sub frame 5
81
Figure 5.13: 3D view of the sea bottom relief of sub frame 6
82
Figure 5.14: 3D view of the sea bottom relief of sub frame 7
83
Figure 5.15 : 3D view of the sea bottom relief of sub frame 8
84
5.2 Satellite data processing
As spectral aspect was the prime concern over all other satellite and sensor
features the selection procedure indicated that visible portion of 0.45 to 0.52 µm
could be best for bathymetric measurement although this spectral portion is, on
the other hand, susceptible to significant back scattering (Rayleigh’s). Other
optical channels such as green, red, near IR, middle IR, upper IR and thermal
(both emissive and reflective) were initially decided to exclude. The main reason
is the noticeably gradual increase of signal attenuation in water column as the
wave length increases. That means, in the spectral region of 0.76 to 0.90 (Near
IR), the EMR saturation will be close to 100 percent just in a meter depth
(BILKO, UNESCO, 1999).
The study area has been extracted from the raw image using the window
command of Idrisi software. The row and column of the main study area image
were 4308 and 5148 accordingly. For processing purpose the row and column
has been reduced to c718 x r854. This reduction of resolution did not affect the
quality of the image of the study region (Figure 5.16a and 5.16b) because the
region was quite big (about 32400 km²).
85
Figure 5.16b: Pattern of slope in the blue band image before (upper) and after (lower)
contraction (row and column reduction). Profile drawn along 90°20′E
The main problems appeared while processing the satellite image were the geo-
registration and removal of the radiance part due to back scattering in the water.
Here it is mentionable that to resample the satellite image a set of 168 control
points were used and in this case the overall RMS error was 0.035635
(Appendix A). It was an advantage that the image was atmospherically corrected
from its source. To get a smother effect the image was filtered in 3X3 filter
window by low pass filter.
When light penetrates water its intensity decreases exponentially with increasing
depth. This process is known as attenuation and it exerts a profound effect on
86
remotely sensed data of aquatic environments. The severity of attenuation
differs with the wavelength of electromagnetic radiation (EMR). In the region
of visible light, the red part of the spectrum attenuates more rapidly than the
shorter-wavelength blue part (Figure 5.17). As depth increases, the separability
of habitat spectra declines (Figure 5.17).
Figure 5.17: Diagram to show how the spectra for a habitat (such as
macroalgae or seagrass) might change with increasing depth for a four
waveband sensor measuring radiance in the blue, green, red and near infra-red
parts of the electromagnetic spectrum. Differential attenuation of the four
wavebands in the water column results in both a decreased ability
to discriminate between different habitats with increasing depth and different
spectra being recorded for the same habitat at different depths.
87
In practice, the spectra of sand at a depth of 2 m will be very different to that at
20 m – yet the sub-stratum is the same. In fact, the spectral signature of sand at
20 m may be similar to that of sea grass at 3 m. The spectral radiances recorded
by a sensor are therefore dependent both on the reflectance of the substrata and
on depth. These two influences on the signal will create considerable confusion
when attempting to use visual inspection or multispectral classification to map
habitats. Since most marine habitat mapping exercises are only concerned with
mapping benthic features, it is useful to remove the confounding influence of
variable water depth.
The exponential decay of light intensity with increasing depth results from two
processes, absorption and scattering. These processes have been discussed in an
atmospheric context previously but now we will discuss these two processes in
the case of water.
5.3.1.1 Absorption
¾ algae (phytoplankton),
¾ water itself, which strongly absorbs red light and has a smaller effect on
shorter wavelength blue light (hence the blue color of clear water).
88
Absorption is wavelength-dependent (Figure 5.17). The chlorophyll in algae
appears ‘green’ because it reflects in the central portion of the visible spectrum
(i.e. the green) and absorbs strongly at either end. Dissolved organic
compounds absorb strongly at the short wavelength (blue) part of the spectrum
and reflect strongly in the yellow-red end (hence they impart a yellow color to
the water).
5.3.1.2 Scattering
EMR may interact with suspended particles in the water column and change
direction. This process of scattering is largely caused by inorganic and organic
particulate matter and increases with the suspended sediment load (turbidity) of
the water.
The clarity of water bodies varies on many scales. For example, many coastal
areas exhibit a seaward gradient of turbid to clear waters created by increases in
depth (i.e. less resuspension of sediments) and reduced input from terrestrial
sources such as sediment-laden rivers. On a large scale, oceans also vary in their
overall turbidity.
Jerlov (1951) formally classified oceanic water types according to their optical
attenuation properties.
b) Type II: Most clear coastal waters were classified as Type II because
attenuation tends to be greater than that for oceanic waters of low productivity.
c) Between Types I and II: Many water bodies were found to lie between types I
and II and most of the coral reef waters fall into categories I or II.
89
d) Type III: waters are fairly turbid and some regions of coastal upwelling are so
turbid that they are unclassified.
Good digital elevation models of depth are rare, particularly for coral reef
systems where charts are often inaccurate (Zainal 1994). As a compromise,
Lyzenga (1978, 1981) put forward a simple image-based approach to
compensate for the effect of variable depth when mapping bottom features.
Rather than predicting the reflectance of the seabed, which is prohibitively
difficult, the method produces a ‘depth-invariant bottom index’ from each pair
of spectral bands. The technique was tested for water in the Bahamas and is
only appropriate where water clarity is good (i.e. most reef and seagrass areas;
Jerlov water Types I or II) and has been represented in the following sections
as an example. The procedure is divided into four steps.
90
Atmospherically corrected radiance = Li - Lsi
where Li is the pixel radiance in band i and Lsi is the average radiance for deep
water in band i. However, if a full atmospheric correction has already taken
place so that pixel values have been converted to surface reflectance, this
process is unnecessary and values of surface reflectance can be used directly. In
short, Lsi can be ignored if a full atmospheric correction has been undertaken
and this is preferred to using the cruder, dark pixel subtraction method.
In relatively clear water, the intensity of light will decay exponentially with
increasing depth (Figure 5.18). If values of light intensity (radiance) are
transformed using natural logarithms, this relationship with depth becomes
linear (Figure 5.18, step 1). Transformed radiance values will therefore decrease
linearly with increasing depth. If Xi is the transformed radiance of a pixel in
band i, this step is written as:
Xi = ln (Li - Lsi )
Xi = ln ( Li )
91
Figure 5.18: Processes of water column correction, showing the steps involved in
creating depth-variant indices of bottom type for sand and seagrass
Step 1.Exponential attenuation of radiance with depth linearised for bands i and j using
natural logarithms. (Band i has a shorter wavelength, and therefore attenuates less
rapidly, than band j).
Step 2.Plot of (transformed) band i against (transformed) band j for a unique substratum
at various depths. Gradient of line represents the ratio of attenuation coefficients, ki /kj.
The ratio is the same irrespective of bottom type.
Step 3.Plotting of multiple bottom types. Each bottom type has a unique y-intercept
(regardless of its depth). The y-intercept therefore becomes a depth-invariant index of
bottom type.
Li = Lsi + a.r.e(-2kiz)
92
Theoretically, it would be possible to rearrange the equation and generate an
image of bottom type, r (reflectance) which is the measurement we seek.
However, this approach is not feasible because there are too many unknown
quantities – i.e. the value of the constant a, the attenuation coefficient for each
band and the depth of water at each pixel. The method developed by Lyzenga
does not require the actual calculation of these parameters but gets around the
problem by using information from more than one band. All that is required is
the ratio of attenuation coefficients between pairs of spectral bands. Use of
ratios cancels out many of these unknowns and the ratios can be determined
from the imagery itself. Two bands are selected and a bi-plot made of (log
transformed) radiances (or reflectances) for the same substratum at differing
depths (Figure 5.18, step 2). Since the effect of depth on measured radiance has
been linearised and the substratum is constant, pixel values for each band will
vary linearly according to their depth (i.e. points will fall on this straight line).
The slope of the bi-plot represents the relative amounts of attenuation in each
band. In fact, the slope represents the ratio of attenuation coefficients between
bands. Conceptually, the line represents an axis of radiance (reflectance) values
for a unique bottom type. As one moves along the line, the only change is
depth.
If radiance (reflectance) values for another bottom type were added to the bi-
plot (Figure 5.18, step 3), a similar line would be obtained – once again, the only
change between data points would be depth. However, since the second bottom
type will not have the same reflectance as the first, the new line will be displaced
either above or below the existing line (e.g. if line 1 was derived from sand
which generally has a high reflectance, and line 2 was generated from seagrass
with lower reflectance, the latter line would lie below that for sand). The
gradient of each line should be identical because the ratio of attenuation
93
coefficients ki/kj is only dependent on the wavelength of the bands and clarity
of the water.
An index of bottom type can be obtained by noting the y-intercept for each
bottom type (Figure 5.18, step 3). For example, while pixel values lying on the
line for sand show considerable variation in radiance, they all represent the same
bottom type and have the same y-intercept. The y-intercept for pixels of
seagrass is considerably different. The y-axis therefore becomes an axis (or
index) of bottom type.
Of course, not all pixel values for a given bottom type lie along a perfectly
straight line (Figure 5.19). This is because of natural variation in bottom
reflectance, patches of turbid water and sensor noise. Nevertheless, each pixel
can be assigned an index of bottom type once the ratio of attenuation
coefficients has been estimated (ki/kj). This is accomplished by ‘connecting’
each pixel on the bi-plot to the y-axis using an imaginary line of gradient ki/kj.
Pixel values on the bi-plot are then converted to their corresponding positions
on the y-axis (index of bottom type). Using this method, each pixel value is
converted to an index of bottom type, which is independent of depth. These
depth-invariant indices of bottom type lie along a continuum but pixels from
similar habitats will have similar indices.
94
Figure 5.19: Bi-plot of log-transformed CASI bands
3 and 4. Data obtained from 348 pixels of sand with
variable depth from 2-15 m.
y=p+q.x
p=y-q.x
95
Each pair of spectral bands will produce a single depth-invariant band of
bottom type. If the imagery has several bands with good water penetration
properties (e.g. Landsat TM, CASI), multipledepth -invariant bands can be
created. The depth -invariant bands may then be used for image processing or
visual inspection instead of the original bands.
The theory described here scales the depth-invariant index to the y-intercept.
Lyzenga (1978, 1981) provided a further modification which re-scaled the values
to an axis orthogonal to the bi-plot slope. However, such a refinement does not
alter the functionality of the process and it has not been included here for
simplicity’s sake.
The methods described have three main applications for mapping coastal
habitats:
96
5.4 Data correction
97
rivers from the north. It would have been best to collect information of
suspended sediment loads in surface and under water bay water from
sufficiently high number uniform location. But this was extremely difficult to
traverse such a huge area of the Bay of Bengal which is often found to be
rough, even sometimes in winter. This is also difficult as the water collection
survey must be completed in the same time of the year when the satellite passed
the area. So, finally straight traversing of 70 km south from the coast of Kuakata
along 900 05' east longitude was decided as the travel route in a fishing vessel,
mainly for travel convenience. Two liters of water samples were collected from
every 10 km interval both from surface and from 10' (3.33m approximately)
under the surface of the sea (Photo 5.1 and Photo 5.2). The justification of the
methodology was that as the arrival of the sediments from the north is closely
similar (if not uniform due to different amount of water discharge) there will be
an approximately similar decrease of suspended loads as distance increases from
the coast southward ( Figure 5.20).
98
Photograph 5.2 : The vessel used for water collection
Figure 5.20: Distribution pattern of the suspended sediments in the study area.
Water column collection sample locations are also shown in the image using dots
99
FCC image presents the view of suspended sediment distribution in the upper
Bay of Bengal. The collected waters were examined in laboratory to determine
the amount of sediments and their average size for each of the locations both
for the surface and 10' under water (Photograph 5.3, 5.4, 5.5, 5.6).
100
Suspended sand particles
101
Now two graphs are created using the sediment size and sediment amount
values. Both the graphs show a clear decline in the amount of sediments as well
as in the average size of the suspended sediments towards deeper sea. The
general tendency of decrease in size of sediments and in lower proportion of
suspended loads in amount per unit of water as the depth increases. This rate of
decrease of size and amount could be used to deduct the extra reflectance due
to back scattering in the water. There is no easy and shortcut method to adjust
this data to the image processing algorithm. But however, amount of sediments
per unit of water was found most convenient for the purpose. Realistically
sediments size has a clear positive relation with the amount of suspended
sediments (Figure 5.21a and 5.21b).
Figure 5.21a: Pattern of the distribution of the amount of suspended load in sea
water.
102
Figure 5.21b : Pattern of the distribution of the suspended load size in sea water.
103
from the main image a continuous surface has been generated through the
interpolation of the contour lines having the values same as total decay (Figure
5.22). The total decay (absorption + scatter) was subtracted from the actual
available radiance of the water sample collection locations. Now the image
could be considered as corrected from the effect of back scatter and signal
attenuation.
104
Figure 5.22: Contour lines of total signal decay. The corresponding blur figures
are representing the total signal decay (in DN) which has been used in generating
continuous surface of total signal decay
105
The satellite image, which presents DN values are just the reflectance of the
light depends on how far it penetrated into the water (theoretically) – and
maintains a relation with the depth of water. And to indicate the relation
between DN values and depth a regression equation has been developed in
which the corrected satellite image has been considered as a dependent variable
and the BIWTA sound chare generated DEM has been considered as
independent variable (Figure 5.23). Theoretically, the corrected satellite image
and the image of the BIWTA generated DEM must show a negative relation in
general, i.e., where depth in the BIWTA image is more DN reflectance in the
TM image would be low.
106
The equation itself is a mathematical expression of the line. In this case the
equation results as: Y = 22.292295 + 0.164983 X.
Where
r = 0.9138
coefficient of det (r2) = 83.50%
s.d. of X (Sx) = 13.5644894
s.d. of Y (Sy) = 2.4490030
s.e. of estimate = 0.9948812
s.e. of beta = 0.0003717
t stat for r or beta = 443.8184509
t stat for beta <> 1 = -2246.2719727
n = 38915
apparent df = 38913
In effect this equation is saying that it can predict depth at any location when
the dependent variable (the satellite image in this case) is multiplied by
+0.313199 and add 110.819099 to the result. This was the model.
Now the value of the relation line’s interception and height could be fitted in to
the image to convert the image’s reflectance DN to simple bathymetric figures
in meters. The regression does not show a perfect relation but was found to be
reasonably acceptable due to the presence of very high number of observations.
Using the newly converted DN values in the resultant image was used to
generate a 3D model that represents the submarine relief of the upper Bay of
Bengal. But to remove noises and sharp spackles in the images it was run in a
3X3 pixel window for mean pass filteration. The relief map shows (Figure 5.24)
107
some minor but significant rises very close to the coast, particularly south of
Bhola and Kuakata.
Figure 5.24 does not represent the bathymetry of the study area in detailed. So
to have a better observational effect the corrected image has been divided into 8
sub frames (Figure 5.25) and then 3D images have been generated from the sub
frames (Figure 5.26 to 5.33)
108
Figure 5.25: Corrected image divided into 8 sub frames
Figure 5.26 : Simulated sea floor relief generated from satellite image (sub frame 1)
109
Figure 5.27 : Simulated sea floor relief generated from satellite image (sub frame 2)
110
Figure 5.28 : Simulated sea floor relief generated from satellite image (sub frame 3)
111
Figure 5.29 : Simulated sea floor relief generated from satellite image (sub frame 4)
112
Figure 5.30 : Simulated sea floor relief generated from satellite image (sub frame 5)
113
Figure 5.31 : Simulated sea floor relief generated from satellite image (sub frame 6)
114
Figure 5.32 : Simulated sea floor relief generated from satellite image (sub frame 7)
115
Figure 5.33 : Simulated sea floor relief generated from satellite image (sub frame 8)
The speedy approaching water from the south/south-west might inundate little
more than usual expected height which should be due to the small humps
detected in the image.
116
Chapter 6
CONCLUSION
The main objective of this study was to check the usability of satellite data and
identify appropriate channel to be used as an alternative method in
bathymetric mapping. In this research study, the emphasis lies on
development of the processing algorithm for the raw satellite data, developing
a model to calibrate the error generated in the satellite data and at last test the
validity of the satellite image generated bathymetric map through a
comparison with the BIWTA sound chart generated three dimensional surface
of the sea bottom.
The first consideration that can be mentioned is that, the objective of this
study was achieved that is a valid bathymetric map has been generated from
the satellite image. A more detailed view on these conclusions will be given in
this final chapter. It has already been mentioned that, overall satellite image
generated bathymetric map is valid and to test it’s validity several profiles have
been drawn along the same coordinates (Figure 6.1) both in the BIWTA
sound chart generated DEM and corrected satellite image. Later the slopes
along the profiles have been compared with each other. Dramatic similarities
in slope have been found along the profiles AB, CD, EF and GH (Figure 6.2
to 6.5). But however in few locations expected results was not achieved.
117
Figure 6.1: Location of profiles in the study area
Figure 6.2: Pattern of slope in the BIWTA sound generated DEM and
corrected satellite image. Profile along AB.
118
Figure 6.3: Pattern of slope in the BIWTA sound generated DEM
and corrected satellite image. Profile along CD.
119
6.1.1 Turbidity
Turbidity in the sea is a condition in which a lot of suspended particles affect
the clarity of the clear sea water. A lot of causes are responsible for turbidity.
But generally when strong onshore winds pile up water near the sea shore
undercurrents are generated which flows towards the sea. And these
undercurrents bring fine materials in suspension. Thus as the sediment
concentration in the water increases due to turbidity, spectral properties of the
water bodies changes. First, it’s overall brightness in the visible region
increases, so the water body ceases to act as a “dark” object but becomes
more and more of a “bright” object as sediment content increases. Second, as
sediment concentration increases, the wavelength of peak reflectance shifts
from a maximum in the blue region toward the green. The presence of larger
particles means that the wavelength of maximum scattering shifts toward the
blue-green and green regions (Figure 6.6) (Campbell, 1996, p. 523). Not only
this but also a lot of signal is lossed through scattering from the suspended
sand or silt particles. Because all of these things errors are generated in the
satellite image.
120
6.1.2 Tide
Tide has a nearly same effect upon the EMR like turbidity. In addition to
affecting the clarity of the water it decays the signal through movement effect.
Beside this the increasing and decreasing of water level due to tidal variation
affects the satellite signal where the satellite signal has a limited water
penetration capacity. Basically in a region like the Bay of Bengal where the tide
is semi diurnal in nature this effect is more important.
6.1.4 Wave
The surface of the sea is not calm any time. Large or small waves always
propagate through it. Figure 6.7 shows spectra for calm and wave-roughed
surfaces in the visible and the near infrared. Wave-roughed surfaces are
brighter than are smoother surfaces. Calm, smooth water surfaces direct only
volume-reflected radiation to the sensor, but rough, wavy water surfaces direct
a portion of the solar beam directly to the sensor. AS a result, wavy surfaces
are much brighter, especially in the visible portion of the spectrum (Campbell,
1996, P. 526). As a result from the wavy sea surface unexpected pixel values
are gained.
121
Figure 6.7: Spectra of calm and wind-roughed water surfaces
122
have been collected for calibration purpose which do not represent the over
all distribution of suspended sediment of the study area. If water samples
could be collected from more locations (about from 100 locations) then the
error in the satellite image could be corrected properly.
123
Figure 6.9: Pattern of slope in the BIWTA sound generated
DEM and corrected satellite image. Profile along KL
124
REFERENCES
Abdullah, K., MatJafri, M.Z. and Din, Z.B., 2000, Contribution form remote
sensing in updating bathymetric chart, GIS Development Proceedings,
www.gisdevelopment.net/aars/acrs/2000/ps3/ps311pf.htm.
Abdullah, K., MatJafri, M.Z. and Din, Z.B., 2000, Remote sensing of total
suspended solids in Penang coastal waters, Malaysia, GIS Development
Proceedings, www.gisdevelopment.net/aars/acrs/2000/ps3/ps312.shtml.
Abdullah, K., Mohd Dimyata, K., Crackneil, A.P., and Vaghan, R.A., 1991,
Evaluation of TM and SPOT data for shallow water Bathymetry,
GIS Development Proceedings,
www.gisdevelopment.net/aars/acrs/1991/ocean/ocean009.shtml.
Abu Daya, M.I., 2001, Analysis of multitemporal satellite imagery for total
suspended sediments in Coastal waters of Golfito and Golfo Dulce, Costa
Rica. ITC-Coastal Zone Studies report.
Abu Daya, M.I., 2004, Coastal water quality monitoring with remote sensing in
(East Kalimantan) Makassar Strait, Indonesia, (unpublished M. Sc. Thesis)
International Institute for Geo-Information Science and Earth Observation.
125
Bagheri, S. and Zetlin, C., " Bathymetry Mapping with Airborne Imaging
Spectrometer Data". 2nd ERIM Thematic Conf. on Remote Sensing for
Marine and Coastal Environment, New Orleans, LA., Jan. 1994, Vol. II, pp.
104-111.
BILKO, Aug 1999, BILKO for windows module 7, lesson 4, UNESCO, pp.
113-120.
Bukata, R.P., John, H.J., Krill, Y.K. and Dimitry, V.P., 1995, Optical
Properties and Remote Sensing of Inland and Coastal waters, CRC Press,
USA.
Das, S.C., 2002, Modern oceanography (in bangle), 2nd edition, Mrs. Esa Das,
B-16, Jahangirnagar University, Savar, Dhaka-1342, pp. 335-354.
Eastman, R.J., 1992, User’s Guide, IDRISI, Clark Lab, Clark University,
Massachusetts, pp. 119-124.
Edwards, A., 1999, The Remote Sensing Handbook for Tropical Coastal
Management, UNESCO.
126
Hagan, J., 1998, User’s Guide, Cartalinx, Clark Lab, Clark University,
Massachusetts, pp. 92-97.
Lillesand, T.M. and Kiefer, R. W., 2002, Remote Sensing and Image
Interpretation, John Willy & Sons, Inc., new York, p, 318, 396 and 415.
Lo, C.P., and Yeung, A.K.W., 2002, Concepts and Techniques of Geographic
Information System, Prentice-Hall, Inc., New Delhi, pp. 136-182.
Louchard, E.M., Pamela Reid, R. and Carol Stephans, F., 2003, Optical remote
sensing of benthic habitat and bathymetry in coastal environments at Lee
Stocking Island, Bahamas: A comparative spectral classification approach,
Limnol. Oceanogr., 48(1, part 2), pp. 511-521.
Mancebo, F.F., Bruce, R.C., Catalan, Z.B., Lim suan, M.P. and Malayang, B.S.,
1997, Correlation of Total suspended Sediments and Reflectance of Landsat
TM in Laguna De Bay, Philippines, GIS Development Proceedings,
www.gisdevelopment.net/aars/acrs/1997/ts2/ts2003pf.htm.
Narayan, LRA., 1999, Remote Sensing and it’s Application, University Press
(India) Ltd., pp. 1-7.
Philpot, W., Analysis of Hyperspectral Data for Coastal Bathymetry and Water
Quality, www.opl.ucsb.edu/hycode/pubs/onr01.
127
Vathananukij, H., Murai, S., Vibulsresth, S., Katsumi, M., Ryosuke, S.,
Kakiuchi, H., Yasuoka, Y., and Yamakata, Y., 1994, Water Quality Analysis
on the Chao-Phra-Ya Estuary Using Remote Sensing Data, GIS Development
Proceedings, www.gisdevelopment.net/aars/acrs/1994/ts2/ts2006pf.htm.
www.csc.noaa.gov/products/nchaz/htm/ccap5.htm.
www.marine.usf.edu/geology/ccom.html.
www.obee.ucla.edu/test/faculty/nezlin/RemoteSensingOfTheSea.htm.
128
APPENDIX
Appendix A
Summary of image geo registration
Coefficient X Y
b0 -0.4715895102193741 -0.4287562301149599
b1 1.0067683070562811 -0.0005090320561347
b2 -0.0063366441441435 1.0221456685601194
129
89.136850 22.423230 89.136850 22.423230 0.024728
90.308910 22.529090 90.308910 22.529090 0.024394
89.242180 21.571910 89.243650 21.651020 omitted
89.207630 21.579480 89.203130 21.656520 omitted
89.290190 21.619840 89.291690 21.682200 omitted
89.243650 21.630840 89.250530 21.697830 omitted
89.221710 21.665550 89.222920 21.719830 0.061245
89.282950 21.680220 89.285350 21.733040 0.059963
89.250530 21.721300 89.249330 21.759450 0.046314
89.213300 21.712500 89.216910 21.755050 0.050199
89.279350 21.716900 89.286550 21.755050 0.045771
89.189290 21.752110 89.200100 21.778520 0.034901
89.272140 21.769720 89.281750 21.791730 0.030690
89.223790 21.806230 89.232790 21.820540 0.023566
89.239670 21.806070 89.252870 21.826610 0.030622
89.183850 21.822940 89.198850 21.828070 0.016857
89.263670 21.855210 89.275680 21.850810 0.008161
89.255870 21.893340 89.268480 21.880140 0.007160
89.188650 21.897010 89.208460 21.902150 0.020767
89.188050 21.864010 89.199460 21.864010 0.011211
89.272080 21.927080 89.283490 21.901410 0.015783
89.240870 21.932950 89.252870 21.910210 0.013010
89.227070 21.961550 89.202450 21.949080 0.031548
89.283490 21.837610 89.293080 21.846410 0.018809
89.274480 21.839070 89.286480 21.847870 0.019473
89.288890 21.811940 89.290080 21.826610 0.024224
89.273280 21.827340 89.285880 21.837610 0.020864
89.291890 21.825140 89.301490 21.840540 0.025185
89.266680 21.885280 89.280480 21.880880 0.009848
89.282880 21.930020 89.292490 21.906550 0.013041
89.293080 21.876480 89.299690 21.867680 0.001527
89.246870 21.943220 89.258870 21.920480 0.012798
89.321300 21.898480 89.324900 21.880880 0.007546
89.179050 22.062760 89.196450 22.007020 0.043715
89.184450 22.069360 89.201260 22.014350 0.042712
89.199460 22.094300 89.216260 22.026090 0.055415
89.200650 22.073760 89.222260 22.012890 0.049690
89.223470 22.058360 89.239670 21.996020 0.050260
89.215060 22.100160 89.233670 22.036360 0.051241
89.321890 22.003360 89.321890 21.978420 0.013942
89.302090 21.853010 89.315900 21.847870 0.009426
89.309290 21.809740 89.320690 21.814870 0.015293
89.327900 21.768670 89.335700 21.787000 0.026734
130
89.331500 21.806800 89.337500 21.818540 0.020710
89.342300 21.821470 89.350110 21.834670 0.022656
89.324900 21.866940 89.332700 21.861810 0.005314
89.354900 21.891140 89.362710 21.880140 0.002506
89.349500 21.872080 89.354310 21.861080 0.001231
89.336900 21.627860 89.337500 21.693860 omitted
89.334500 21.660860 89.331500 21.718060 0.064381
89.398720 21.663060 89.394520 21.715860 0.060034
89.363910 21.693130 89.359710 21.745200 0.060032
89.423930 21.687260 89.418530 21.732730 0.053334
89.341710 21.784070 89.348300 21.810470 0.035203
89.470750 21.762070 89.468350 21.798000 0.044846
89.359110 21.850810 89.381320 21.877210 0.040364
89.292490 22.089900 89.296680 22.041490 0.034602
89.294290 22.045890 89.289480 22.020960 0.016312
89.308690 22.049560 89.312890 22.008490 0.027996
89.318890 22.008490 89.315290 21.985020 0.014723
89.326100 22.110430 89.344700 22.050290 0.047594
89.343510 22.130230 89.368110 22.098700 0.024231
89.352510 22.127300 89.380110 22.103100 0.022745
89.436530 21.923410 89.446740 21.904350 0.009774
89.445530 21.960820 89.449130 21.930750 0.018751
89.421530 21.976220 89.426330 21.949820 0.014597
89.443730 21.927080 89.458150 21.908010 0.012537
89.471340 21.927080 89.479150 21.912410 0.004765
89.494750 21.928550 89.497150 21.904350 0.013613
89.452140 21.953480 89.460540 21.927810 0.014746
89.439540 21.979150 89.441930 21.946880 0.020730
89.420330 21.991620 89.433530 21.952020 0.028820
89.408930 22.018020 89.427530 21.969620 0.038436
89.423930 22.032690 89.436530 21.984290 0.036510
89.497150 21.951280 89.487550 21.924880 0.020942
89.416130 22.101630 89.429330 22.028290 0.060233
89.429330 22.119960 89.437740 22.064230 0.041452
89.455740 22.098700 89.467740 22.051030 0.034205
89.450940 22.249040 89.467740 22.147100 omitted
89.459340 22.254910 89.476750 22.156630 omitted
89.486950 22.166900 89.494750 22.092090 0.059910
89.429330 22.147100 89.446740 22.075230 0.058398
89.453940 22.121430 89.460540 22.062760 0.044376
89.469540 22.177170 89.482150 22.098700 0.063731
89.486350 22.158830 89.492360 22.087690 0.056304
89.496550 22.074490 89.507350 22.032690 0.028704
131
89.519360 22.045160 89.518170 22.005550 0.027350
89.519360 22.011850 89.513960 22.055430 0.058709
89.520560 22.158100 89.525960 22.086230 0.057084
89.498950 22.198440 89.505550 22.135370 0.047188
89.530770 22.083290 89.528970 22.040760 0.029614
89.563770 22.224840 89.560170 22.146370 0.063066
89.585540 22.212810 89.587940 22.151190 0.045539
89.559130 22.123310 89.560330 22.067560 0.041592
89.538710 22.001540 89.538710 21.969260 0.020650
89.563930 21.969260 89.560330 21.939910 0.019552
89.596350 21.981000 89.599950 21.956050 0.013102
89.621570 21.972190 89.625170 21.951650 0.008797
89.531880 22.346200 89.552890 22.208630 omitted
89.561900 22.360870 89.581410 22.237980 omitted
89.555900 22.395720 89.624950 22.362700 0.063540
89.518360 22.412230 89.591930 22.408560 0.068827
89.527370 22.428740 89.606930 22.412230 omitted
89.555900 22.450750 89.645700 22.399390 omitted
89.573910 22.445240 89.651970 22.373710 omitted
89.567910 22.366370 89.603940 22.313180 0.044586
89.695500 22.527780 89.695500 22.527780 0.025531
89.669980 21.937160 89.669980 21.937160 0.011992
89.644460 21.916980 89.644460 21.916980 0.011594
89.600930 21.973840 89.606930 21.948170 0.014100
89.548390 21.869290 89.548390 21.869290 0.010772
89.614440 21.861960 89.614440 21.861960 0.010411
89.665480 21.696870 89.654980 21.750070 0.061879
89.611440 21.718890 89.608440 21.762910 0.051952
89.700010 21.645520 89.572410 21.713380 0.150224
89.584420 21.717050 89.575410 21.764740 0.056670
89.551390 21.733560 89.545390 21.775750 0.050984
89.600930 21.762910 89.594920 21.795920 0.042367
89.569400 21.801430 89.570910 21.819770 0.027284
89.570910 21.819770 89.585920 21.823440 0.017049
89.823100 22.344360 89.839610 22.265490 0.061436
89.863640 22.353530 89.868140 22.269160 omitted
89.871140 22.234310 89.875650 22.142600 omitted
89.832120 22.131590 89.821600 22.074730 0.044742
89.884650 22.287500 89.898160 22.160940 omitted
89.893660 22.142600 89.895160 22.078400 0.049799
89.932690 22.038040 89.935690 22.994020 omitted
89.952200 21.961010 89.941700 21.933490 0.020455
89.980730 22.049050 89.976230 22.005030 0.031938
132
89.827610 22.098570 89.827610 22.054550 0.030276
89.803590 22.028870 89.814090 21.994020 0.023460
89.796080 22.028870 89.805090 21.990350 0.026661
89.842620 21.979350 89.839610 21.964670 0.006344
89.857630 21.983020 89.851620 21.975680 0.009959
89.802090 21.926160 89.802090 21.926160 0.011429
89.809590 21.926160 89.809590 21.926160 0.011414
89.844120 21.894970 89.844120 21.894970 0.010627
90.075300 21.834440 90.043780 21.860120 0.047693
90.033270 21.959170 90.033270 21.959170 0.011794
90.093320 21.825270 90.082810 21.861960 0.047499
90.198390 21.748230 90.180380 21.801430 omitted
90.262950 21.797760 90.237430 21.838110 0.055139
90.075150 21.995310 90.075150 21.995310 0.012562
90.145420 21.966690 90.145420 21.966690 0.011842
90.204870 21.962290 90.204870 21.962290 0.011698
90.240910 21.986510 90.255320 21.968900 0.015602
90.305760 22.030530 90.305760 22.030530 0.013156
90.338200 22.096570 90.347210 22.072360 0.013600
90.372430 22.138400 90.372430 22.094370 0.029494
90.392450 22.159990 90.385250 22.110110 0.035723
90.423670 22.149720 90.412860 22.091030 0.045464
90.433280 22.136520 90.433280 22.085170 0.037053
90.462100 22.177600 90.462100 22.118910 0.043659
90.493320 22.248020 90.493320 22.171730 0.060104
90.532940 22.324320 90.538940 22.208410 omitted
90.598980 22.450490 90.610990 22.325780 omitted
90.567760 22.611890 90.580970 22.497450 omitted
90.659020 22.626560 90.669830 22.500380 omitted
90.974630 22.500380 90.666220 22.384470 omitted
90.666220 22.363930 90.666220 22.363930 0.020359
90.633800 22.283240 90.663830 22.250960 0.034268
90.629000 22.211340 90.619390 22.187870 0.010971
90.637410 22.019140 90.637410 22.019140 0.012949
90.782700 22.121850 90.782700 22.121850 0.015174
90.837940 22.344860 90.837940 22.344860 0.019914
90.969580 21.772670 90.969580 21.772670 0.009402
90.916150 21.339970 90.916150 21.339970 0.008879
90.483560 21.320150 90.483560 21.320150 0.006323
90.872890 21.026140 90.872890 21.026140 0.013877
90.840450 20.804810 90.840450 20.804810 0.018266
89.929730 21.111120 89.929730 21.111120 0.007755
89.099250 20.744190 89.099250 20.744190 0.014717
133
89.137090 21.186670 89.137090 21.186670 0.005550
89.607320 21.186670 89.607320 21.186670 0.005217
89.230220 21.470880 89.230220 21.470880 0.003930
89.811730 21.632360 89.811730 21.632360 0.004658
90.126970 21.607910 90.126970 21.607910 0.004166
90.540630 21.640950 90.540630 21.640950 0.006011
134