An introduction to ocean remote sensing (2nd ed.) [s. martin, 2014]

Trim: 247mm × 174mm Top: 14.762mm Gutter: 23.198mm
CUUK2533-FM CUUK2533/Martin ISBN: 978 1 107 01938 6 November 26, 2013 12:15
AN INTRODUCTION TO OCEAN REMOTE SENSING
Second edition
Fully updated, with signiﬁcant new coverage of advances in satellite oceanography and
results from new satellite missions, the second edition of this popular textbook intro-
duces students to how remote sensing works, how to understand observations from Earth-
observing systems, and the observations’ importance to physical and biological oceanog-
raphy. It provides full explanations of radiative transfer, ocean surface properties, satellite
orbits, instruments and methods, visible remote sensing of biogeochemical properties,
infrared and microwave retrieval of sea surface temperature, sea surface salinity retrieval,
passive microwave measurements, scatterometer wind retrieval, altimetry and SAR. This
new edition also includes descriptions of the online archives where data can be obtained,
and where readers can obtain online tools for working with the data – enabling hands-on
engagement with real-world observations.
This is an ideal textbook for graduate and advanced undergraduate students taking
courses in oceanography, remote sensing and environmental science, and provides a prac-
tical resource for researchers and Earth science professionals working with oceanographic
satellite data.
seelye martin is an Emeritus Professor in the School of Oceanography at the University
of Washington. He has been involved with passive microwave, visible/infrared and radar
ice research since 1979, and has made many trips to the Arctic for research on sea ice
properties and oceanography. Professor Martin has served on a number of NASA and
NOAA committees and panels involving remote sensing and high latitude processes. From
2006–2008, he worked at NASA Headquarters as Program Manager for the Cryosphere,
where he also served as program scientist for the ICESat-1 and ICESat-2 missions. From
2009–2012, he worked in a variety of roles for the NASA high-latitude IceBridge remote
sensing aircraft program. For this work, in 2012 he was awarded the NASA Exceptional
Public Service Medal.

AN INTRODUCTION TO OCEAN
REMOTE SENSING
second edition
SEELYE MARTIN
School of Oceanography, University of Washington

University Printing House, Cambridge CB2 8BS, United Kingdom
Published in the United States of America by Cambridge University Press, New York
Cambridge University Press is part of the University of Cambridge.
It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning and research at the highest international levels of excellence.
www.cambridge.org
Information on this title: www.cambridge.org/9781107019386
First edition c Cambridge University Press
Second edition c Seelye Martin 2014
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First edition published 2004
Paperback edition published 2011
Second edition published 2014
Printed in the United Kingdom by MPG Printgroup Ltd, Cambridge
A catalogue record for this publication is available from the British Library
Library of Congress Cataloguing in Publication data
ISBN 978 1 107 01938 6 Hardback
Additional resources for this publication at www.cambridge.org/oceanremotesensing
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication,
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.

To the memory of my mother
Lucy Gray Martin
April 19, 1915–June 13, 2002

Contents
Preface page xi
List of chemical symbols xiv
List of mathematical symbols xv
List of abbreviations and acronyms xxi
1 Background 1
1.1 Introduction 1
1.2 Deﬁnition of remote sensing 3
1.3 Satellite orbits 4
1.4 Geosynchronous satellites 12
1.5 Sun-synchronous satellites 13
1.6 Imaging techniques 15
1.7 Processing levels, archives, data records and processing 22
1.8 Past, present and pending satellite missions 26
2 Ocean surface phenomena 35
2.1 Introduction 35
2.2 Ocean surface winds and waves 35
2.3 Ocean currents, geostrophy and sea surface height 46
2.4 Sea ice 50
3 Electromagnetic radiation 53
3.1 Introduction 53
3.2 Descriptions of electromagnetic radiation 53
3.3 Ways to describe EMR 61
3.4 Radiation from a perfect emitter 66
3.5 The ideal instrument 71
4 Atmospheric properties and radiative transfer 79
4.1 Introduction 79
4.2 Description of the atmosphere 79
4.3 Molecular absorption and emission 86
4.4 Scattering 90
vii

viii Contents
4.5 Atmospheric attenuation 96
4.6 Application to the ideal instrument 99
4.7 The radiative transfer equation 101
4.8 Specific solutions of the radiative transfer equation 105
4.9 Diffuse transmittance and skylight 110
5 Reflection, transmission and absorption at the atmosphere/ocean interface 113
5.1 Introduction 113
5.2 The interface 115
5.3 Transmission across an interface 122
5.4 Absorption and scattering properties of seawater 126
5.5 Reflection from foam 135
6 Ocean color 136
6.2 Absorption and scattering by phytoplankton, particulates and
dissolved material 139
6.3 Ocean color satellite instruments 147
6.4 SeaWiFS, MODIS, VIIRS and their calibrations 152
6.5 Atmospheric correction and retrieval of the water-leaving radiance 159
6.6 Surface validation data sets and the vicarious calibration 169
6.7 Chlorophyll reflectance and fluorescence 171
6.8 The empirical, semi-analytic and biogeochemical algorithms 174
6.9 The Pre-Aerosol, Clouds and ocean Ecosystem (PACE) mission 192
7 Infrared observations of sea surface temperature (SST) 194
7.2 What is SST? 197
7.3 Properties of AVHRR, MODIS and VIIRS bands used in the SST
retrieval 200
7.4 Atmosphere and ocean properties in the infrared 203
7.5 SST algorithms 208
7.6 Cloud-detection and masking algorithms 221
7.7 Error and bias of the data sets 227
7.8 Other GHRSST data sets and merged products 229
7.9 Illustrations and examples 231
8 Introduction to microwave imagers 236
8.2 General antenna properties 237
8.3 Measurement of a surface radiance with an antenna 242
8.4 Conical scanners and microwave surface emissivity 244
8.5 Antenna pattern correction (APC) 245
8.6 Passive microwave imagers 248

Contents ix
9 Passive microwave observations of the atmosphere and ocean surface 260
9.2 Atmospheric absorption and transmissivity in the microwave 260
9.3 Radiative transfer in the microwave 266
9.4 Dependence of the emissivity on surface waves and foam 273
9.5 Temperature and salinity 285
9.6 Open ocean algorithms 288
9.7 WindSat retrieval of wind speed and direction 295
9.8 Sea ice algorithms 300
10 Introduction to radars 308
10.2 Radar equation 309
10.3 Determination of σ◦
within an FOV 313
10.4 Range binning 315
10.5 Doppler binning 319
10.6 Oceanic backscatter 324
11 Scatterometers 331
11.2 Background 333
11.3 How scatterometers derive the wind velocity 336
11.4 NSCAT scatterometer 342
11.5 AMI and ASCAT scatterometer 343
11.6 The rotating beam scatterometers 346
11.7 Advantages and disadvantages of the different scatterometers 354
11.8 The ISS-RapidScat 355
11.9 Cross-calibrated multi-platform winds (CCMP) 356
11.10 Applications and examples 356
12 The altimeter 362
12.2 Shape of the Earth 363
12.3 Past, present and future altimetric satellites 368
12.4 TOPEX/POSEIDON 368
12.5 JASON-1/JASON-2 378
12.6 Altimeter interaction with a specular sea surface 380
12.7 Effect of surface waves on the altimeter return 385
12.8 Errors and biases in retrieval of sea surface height 389
13 Imaging radars 401
13.2 Background 402

x Contents
13.3 Resolution of side-looking radars (SLRs) 409
13.4 How the SAR achieves its resolution 409
13.5 RADARSAT-2 SAR 415
13.6 Other operational SARs 422
14 Other instruments: the gravity missions, ICESat-1 and -2, CryoSat-2, SMOS
and Aquarius/SAC-D 436
14.2 Gravity missions 436
14.3 The ICESat-1, ICESat-2 and CryoSat-2 missions 441
14.4 SMOS and Aquarius/SAC-D 449
Appendix 455
References 458
Index 489
The color plates will be found between pages 000 and 000

Preface
Since the publication of the first edition a decade ago, the variety and use of ocean observing
satellites has continued to grow. Combined with a similar expansion in computer resources
and in surface receiving and distribution networks, this growth has greatly increased our
knowledge of the properties of the upper ocean and the overlying atmosphere.
Ten years ago, many satellites were large, managed by single countries and carried
multiple sensors. Now, by international agreement, different countries collaborate on con-
stellations of smaller satellites that fly in complementary orbits and focus on a single
oceanic or atmospheric feature such as biology, winds or sea surface temperature (SST).
Many of these data sets such as SST from the constellations are available in a common
format from public archives that also provide software tools for working with the data.
These constellations and their archives greatly improve research opportunities for students
and professionals.
For remote sensing, the use of the electromagnetic spectrum combined with our under-
standing of the oceanic surface and atmospheric properties has stimulated innovations
in instrumentation. Satellite remote sensing also uses gravity measurements that have
improved our knowledge of the Earth’s geoid, measured the ice loss from the major ice
caps, and monitored changes in the ocean circulation. Many of the experimental sensors
of the 1980s are now the operational tools of oceanography. These include narrow-band
optical sensors to estimate biological productivity, infrared sensors to measure sea sur-
face temperature that approach an accuracy necessary to observe climate change, passive
microwave sensors that provide global cloud-independent observations of winds and sea
surface temperature and salinity, and altimeters capable of measuring sea surface height to
within 2 cm.
Because remote sensing involves many disciplines, the book provides under one cover
the necessary background in electromagnetic theory, atmospheric and seawater properties,
physical and biological oceanography, physical properties of the sea surface and the prop-
erties of satellite orbits. The contents range from the reflective and emissive properties of
clouds and foam to the radar-scattering properties of ocean waves, to the optical properties
of plankton-associated pigments. It also includes many examples. The book describes
the development of satellite oceanography from 1975 to 2013, and outlines pending
xi

xii Preface
missions. The book requires only an introductory knowledge of electromagnetic theory
and differential equations.
The text divides into five parts. Chapters 1–3 introduce satellite systems, ocean surface
properties and electromagnetic theory. Chapters 4–7 discuss remote sensing in the visible
and infrared spectrum, including atmospheric properties, the ocean/atmosphere interface,
the visible retrieval of ocean color and the infrared retrieval of sea surface temperature.
Chapters 8 and 9 discuss the passive microwave, including antennas, instruments, atmo-
spheric properties and the retrieval of ocean surface and atmospheric variables. Chapters
10–13 discuss the active microwave, including a variety of radars to retrieve wind speed and
direction, sea surface height and images of the ocean surface. Finally, Chapter 14 describes
a variety of gravity and sea surface salinity missions, and sea ice and ice sheet laser and
radar altimeter satellites.
I began this book during 1993–94, when I was a visiting scientist at the National
Institute of Polar Research in Tokyo. I wrote the second draft following my retirement
from the University of Washington in 2011. The book benefited from my work with the
National Aeronautics and Space Administration (NASA); from my service on committees
in 1980s and 1990s, from 2006–2008 when I worked at NASA Headquarters as program
manager for the cryosphere, and from 2009–2012, when I performed a variety of services
for the Airborne Operation IceBridge (OIB) program. I am grateful to NASA for these
opportunities. I particularly thank Dixon Butler, who was head of the Earth Observing
System (EOS) program, and Waleed Abdalati and Jack Kaye for their support during my
time at headquarters.
At the University of Washington, I taught remote sensing both singly and jointly with
Miles Logsdon. I thank Miles and all of our students, who always managed to focus on
those points that I did not understand. In my teaching and writing, I benefited from the class
notes of Dudley Chelton, James Mueller and Carlyle Wash, and the textbooks of Charles
Elachi, George Maul, Ian Robinson and Robert Stewart.
At NASA Goddard Space Flight Center (GSFC), I thank Ziauddin Ahmad, Gene Eplee,
Don Cavalieri, Josephino Comiso, Charles McClain, Claire Parkinson, Jeremy Werdell
and Meng-Hua Wang; at the Jet Propulsion Laboratory (JPL), Ron Kwok, Lee-Lueng Fu,
Ben Holt and Simon Yueh. At MacDonald, Dettwiler and Associates (MDA), I thank
Jeff Hurley and Wendy Keyser. At the National Oceanic and Atmospheric Administration
(NOAA), I thank Alexander Ignatov, Boris Petrenko and Mayra Pazo; at Oregon State
University, Dudley Chelton; at Earth and Space Research, Gary Lagerloef and Hsun-Ying
Kao; at Remote Sensing Systems, Chelle Gentemann, Tom Meissner and Frank Wentz; at
NASA Headquarters, Paula Bontempi. I also thank Peter Wadhams from the University of
Cambridge and Peter Minnett from the University of Miami for their encouragement and
support. At Cambridge University Press, I thank Kirsten Bot, Laura Clark, Susan Francis
and David Mackenzie for their help and support. For his careful line-by-line reading of the
manuscript, I thank my freelance editor, Steven Holt.
At the University of Washington, I thank Jamie Morison, Cecilia Peralta-Ferriz as well
as the staff of the UW Libraries for their support and for their extensive online collection

Preface xiii
of journals. For their critical readings of draft chapters I thank Peter Cornillon for Chapter
1 and Boris Petrenko for Chapter 7. I also thank Alexander Ignatov for his help with
understanding the NOAA SST processing. Any errors are my own.
I thank my son and daughter, Carl William Coryell-Martin and Maria Elizabeth Coryell-
Martin, for putting up with all this even after they have left home and my wife, Julie Esther
Coryell, for her optimism that I might ﬁnish the book, for reading all of the chapters in
draft and for her support. Finally, I ask the reader to remember that each of the satellites,
instruments and algorithms described in this book began as an idea generated by a single
individual or a small committee.

Chemical symbols
Ar Argon
CH4 Methane
CO Carbon monoxide
CO2 Carbon dioxide
Fe Iron
H2O Water
N2 Nitrogen
N2O Nitrous oxide
O2 Oxygen
O3 Ozone
Hα, Hβ, Hγ Hydrogen lines in the Fraunhofer spectrum
Mg–I Magnesium–iodine line
O2-A Oxygen-A line
xiv

Mathematical symbols
Symbol Unit Definition
A m2
Area, or instrument aperture area
Ae m2
Effective antenna aperture area
AFOV area Antenna half-power field-of-view
Ai(400) m−1
Reference absorption at 400 nm; i refers to particulates or
CDOM
a(λ) m−1
Volume absorption coefficient
â(λ; θ, φ) – Ratio of gray-body to blackbody absorption; in VIR, the
absorptance, in microwave, the absorptivity
aCDOM m−1
CDOM absorption coefficient
aw m Amplitude of ocean surface waves
aw(λ), ap, aφ, aT m−1
Absorption coefficients for seawater, particulate,
phytoplankton and total absorption
B W m−2
sr−1
Brightness, used for radiance in the passive microwave
B tesla m−1
Magnetic field vector
Bf J m−2
sr−1
Frequency form of spectral brightness
b(λ) m−1
Volume scattering coefficient of seawater
bb(λ), bbw(λ) m−1
Backscatter coefficient of pure seawater
bbT(λ) m−1
Total backscatter coefficient of seawater
°C Degrees Celsius
Ca mg Chl-a m−3
Chlorophyll concentration
Cw, C1 – Concentrations of open water and sea ice
c m s−1
Speed of light in vacuum
c(λ) m−1
Volume attenuation coefficient of seawater
D cm, m Aperture diameter of a lens or length of an antenna
ˆd (λ) – Normalized absorption depth
da(λ) m Absorption depth of radiation in seawater
E W m−2
Irradiance, the incident flux density per unit area
E V m−1
Electric field vector
Ê J Energy of a photon
E0 V m−1
Reference amplitude of an electric field vector
Ed(λ, 0+) W m−2
Downwelled solar irradiance measured just above the ocean
surface
xv

xvi Mathematical symbols
ER(χ, ψ) km Height of reference ellipsoid above Earth’s center of mass
Eu(0−) W m−2
Upwelled solar irradiance just below the water surface
EV, EH V m−1
Vertically and horizontally polarized components of the
electric ﬁeld vector
e(λ; θ, φ) – Emissivity, which is the ratio of gray-body to blackbody
radiance
e0 – Temperature- and salinity-dependent emissivity of a specular
ocean surface
F(λ, z) W m−2
nm−1
Solar irradiance at a height z in the atmosphere
Fn – Normalized power or radiation pattern
FS(λ) W m−2
nm−1
solar irradiance at the top of the atmosphere
FS(λ) W m−2
nm−1
FS(λ) attenuated by two passes through the ozone layer
f s−1
Coriolis parameter
f Hz Frequency
f(x) V m−1
Antenna illumination pattern
f L m Focal length
f N s−1
Nyquist sampling frequency
fp(T, λ) W m−3
sr−1
Planck blackbody radiance
G – Antenna gain
G0 – Maximum antenna gain
GR – Gradient ratio used in the derivation of sea ice concentration
g m s−2
Acceleration of gravity
H km Radial distance of a satellite from Earth’s center of mass
H1/3 m Signiﬁcant wave height
Hz s−1
Cycles per second
h length Height of satellite above ocean surface
hS length Height of sea surface above Earth’s center of mass
hs length Temporal mean of sea surface height
h J s Planck constant, 6.626 × 10−34
J s
I deg Inclination, the angle between the Earth’s rotation axis and
the normal to the orbit plane
I(r, θ, φ) W sr−1
Radiant intensity
I0 W sr−1
Maximum radiant intensity
i Imaginary part of complex number
J Joules
K Degrees Kelvin
k, kim m−1
Real and imaginary part of the wavenumber
k m−1
Vector wavenumber
kB J K−1
Boltzmann constant, 1.38 × 10−23
J K−1
kw m−1
Wave number of ocean waves
L mm Columnar equivalent of non-raining cloud liquid water
L(λ) µW cm−2
nm−1
sr−1
Radiance
W m−3
sr−1
(Alternative units of L)
LA(λ) µW cm−2
nm−1
sr−1
Path radiance generated by aerosol atmospheric scattering
LE km Equatorial separation between successive orbits

Mathematical symbols xvii
Lf (λ) J m−2
sr−1
Frequency form of spectral radiance
LR (λ) µW cm−2
nm−1
sr−1
Path radiance generated by Rayleigh scattering
Ls(λ) µW cm−2
nm−1
sr−1
Solar radiance at the top of the atmosphere
LT(λ) µW cm−2
nm−1
sr−1
Total radiance received at the satellite
Lw(λ) µW cm−2
nm−1
sr−1
Water-leaving radiance
[Lw(λ)]N µW cm−2
nm−1
sr−1
Normalized water-leaving radiance
Lλ(λ) µW cm−2
nm−1
sr−1
Wavelength form of spectral radiance
l m Length of an imaging radar
M W m−2
Exitance, or emitted flux or power density
N(χ, ψ) m Geoid undulation, or height of geoid relative to the
reference ellipsoid ER
Np, nepers – Units of atmospheric absorption used in microwave
NE T K Noise-equivalent delta-temperature
NE L µW cm−2
nm−1
sr−1
Noise-equivalent delta-radiance
NE σ0 – Noise-equivalent delta-sigma-zero
n – Real part of the index of refraction
P – For radiometers, subscript indicates V or H polarization.
For radars, subscript indicates VV or HH polarization
P(θ) sr−1
Atmospheric scattering phase function
PR – Polarization ratio used in the derivation of sea ice
concentration
PR(θ) sr−1
Rayleigh atmospheric scattering phase function
p kg m−1
s−2
Atmospheric pressure
Q – Coefficient used in description of the water-leaving
radiance
R(λ) – Plane irradiance reflectance
R(λ, 0−) – Irradiance reflectance evaluated just below the surface
R0 km Distance from radar to target
Rc mm, µm Radius of curvature of the sea surface
RF(λ) – Irradiance reflectance of foam
RR mm h−1
Rain rate
Rrs(λ) – Remote sensing reflectance
r length Radius
r length Vector radius (r, θ, φ)
r(θ) – Unpolarized radiance reflectance
S psu Salinity
SN – Signal-to-noise ratio
SS psu Surface salinity
T °C, K Temperature
¯T °C, K Mean temperature of the lower troposphere
T(θ) – Interface transmittance
T3, T4, T5 K AVHRR brightness temperatures for bands 3, 4, 5
T22, T23, T31, T32 K MODIS brightness temperatures for bands 22, 23, 31,
32
TA K Antenna temperature

xviii Mathematical symbols
Ta K Air temperature
Tb K Brightness temperature
Tb °C Buoy or bulk temperature
TBV, TBH K Vertically and horizontally polarized components of
brightness temperatures
Text K Extraterrestrial brightness temperature exclusive of the
Sun
Tgal K Brightness temperature of the Milky Way galaxy
TS °C, K Ocean surface skin temperature
Tsfc °C, K Externally supplied surface temperature to algorithms
Tsol K Solar contribution to the antenna brightness temperature
Tsun K Solar brightness temperature
Tuniv K The 2.7-K Universe background temperature
Tw s Period of ocean surface waves
t Time
t – In the visible/infrared, the atmospheric transmittance; in
the microwave, the atmospheric transmissivity
tD(λ) – Diffuse transmittance
U m s−1
The scalar wind speed at a 10-m height
U0 m s−1
Spacecraft velocity
ULOS m s−1
Line-of-sight wind speed, the wind speed in the
azimuthal look direction of a passive microwave
radiometer
u, v m s−1
x- and y-components of an ocean current
V mm Equivalent height in liquid water of the columnar water
vapor
v m s−1
Local phase speed of light
w m Width of an imaging radar
x length Vector position (x, y, z)
X, Y – Coefficients used in discussion of particulate scattering
properties
XS length Imaging radar cross-track swath width
YS length Imaging radar along-track swath width
ZH km Reference height for the top of the atmosphere
α deg Scattering angle relative to the forward direction
α – ˚Angström exponent used to describe aerosols
αS sr Solid angle resolution of an ideal optical instrument
β(α, λ) km−1
sr−1
, m−1
sr−1
Atmospheric and oceanic volume scattering function
˜β(α, λ) sr−1
Oceanic scattering phase function
β0 km−1
sr−1
, m−1
sr−1
Isotropic scattering phase function
βT, βw, βp, βφ m−1
sr−1
Total, pure seawater, particulate and phytoplankton
volume scattering function
Ê J Energy difference associated with a change in the
internal state of a molecule or atom
f Hz, MHz Instrument bandwidth, also used to describe Doppler
shift

Mathematical symbols xix
hion m Range delay caused by atmospheric free electrons
T45 K Temperature difference between AVHRR channels 4 and 5,
T45 = T4 – T5
T53 K Temperature difference between AVHRR channels 5 and 3,
T53 = T5 – T3
x, y m Radar resolution in the cross-track and along-track direction
θ1/2 deg Half-power beamwidth; for imaging radars, the half-power
beamwidth in the cross-track direction
ø1/2 deg Half-power beamwidth in the along-track direction
ε farad m−1
Electrical permittivity
ε(λ, λ0) – Single-scattering color ratio for aerosols
ε0 farad m−1
Permittivity in vacuum
εr – Complex dielectric constant, εr = ε + iε
ζ m Sea surface height relative to the geoid
ζD m Dynamic height, or the oceanographic height calculated from
the vertical density structure
η – Complex index of refraction, η = n + iχ
η m Vertical displacement of ocean surface waves
ηM – Main beam efficiency of a microwave antenna
θ deg Incidence, look or zenith angle
θ S deg Solar zenith angle
θv deg View or scan
κA, κE, κS km−1
Atmospheric absorption, extinction and scattering coefficients
κR km−1
Rayleigh scattering attenuation coefficient
κoxy km−1
Oxygen absorption coefficient
κvap km−1
Water vapor absorption coefficient
λ nm, µm Radiation wavelength
λw mm, m Wavelength of ocean surface waves
µ henry m−1
Magnetic permeability
µ0 henry m−1
Vacuum permeability
W m−4
sr−1
The atmospheric radiative source term
ρ kg m−3
Density of seawater
ρa kg m−3
Density of air
ρH, ρV – Horizontal, vertical reflection coefficients
ρion TECU Free-electron columnar density
ρw(λ) – Extraterrestrial reflectance generated by the water-leaving
radiance
[ρw(λ)]N – Normalized extraterrestrial reflectance
σ siemens m−1
Electrical conductivity
σ m2
Radar scattering cross section
σ2
– Mean-square sea surface slope
σ0 – Normalized radar scattering cross section (pronounced
sigma-zero)
σN – Standard deviation of noise

xx Mathematical symbols
σVV, σHH, σHV, σVH – Normalized radar scattering cross section for VV, HH, HV
and VH transmitting and receiving
ση m Root-mean-square sea surface height
τ s Pulse duration or length
τ(λ) – Optical depth
τA – Optical depth associated with aerosol scattering
τOZ – Optical thickness of the ozone layer
τR(λ) km Rayleigh optical thickness
W Radiant flux or power
N W Noise generated internally to an instrument
T W Total radiant flux or power transmitted by an antenna
(V, H) W V-pol or H-pol radiant flux received by an antenna
λ W µm−1
Spectral form of the radiant flux
σ W Received power corrected for atmospheric attenuation
ø deg Azimuth angle
øR deg Azimuthal angle relative to the wind direction
øW deg Azimuthal wind direction
χ – Imaginary part of the index of refraction
χ, ψ deg Latitude, longitude
sr Solid angle
E s−1
Angular rotation of the Earth
M sr Main beam solid angle of a microwave antenna
P sr Pattern solid angle of a microwave antenna
ω s−1
Radian frequency of an electromagnetic wave
ω 0 (λ) – Single-scattering atmospheric albedo
ωA(λ) – Aerosol single-scattering albedo
ωR(λ) – Rayleigh single-scattering albedo

Abbreviations and acronyms
A-Train The A- or afternoon train is a constellation of satellites in the
same orbit with a 1:30 pm equator crossing time.
AATSR Advanced ATSR (ESA)
ABI Advanced Baseline Imager (instrument on GOES-R)
ACSPO Advanced Clear-Sky Processor for Ocean (NOAA)
ADEOS-1, -2 Advanced Earth Observing Satellite (Japan)
AGC Automatic Gain Control (altimeter function)
AHRPT Advanced High Resolution Picture Transmission (METOP)
ALOS Advanced Land Observing Satellite (Japan)
ALT Altimeter on TOPEX/POSEIDON
AMSR Advanced Microwave Scanning Radiometer (Japan) on
ADEOS-2
AMSR-E AMSR-EOS (Japan) on AQUA
AOML Atlantic Oceanographic and Meteorological Laboratory (NOAA)
AOP Apparent Optical Properties
APC Antenna Pattern Correction
APT Automatic Picture Transmission (data transfer mode for AVHRR)
AQUA Second major EOS satellite (not an abbreviation)
ASAR Advanced SAR (ENVISAT)
ASCAT Advanced Scatterometer (METOP)
ATSR Along-Track Scanning Radiometer (ESA)
AVHRR Advanced Very High Resolution Radiometer (United States)
AVISO Archiving, Validation and Interpretation of Satellite
Oceanographic data (France)
CalTech California Institute of Technology
C-band Frequencies of about 5 GHz
CCMP Cross-Calibrated Multi-Platform (mind dataset)
CDOM Colored Dissolved Organic Material
CHAMP CHAllenging Minisatellite Payload (German gravity mission)
Chl-a Chlorophyll-a
xxi

xxii Abbreviations and acronyms
CDR Climate Data Record
CEOS Committee on Earth Observation Satellites
CONAE Comisi´on Nacional de Actividades Espaciales (Argentinian
Space Agency)
CNES Centre National d’Etudes Spatiales (National Center for Space
Studies, France)
CryoSat-2 ESA radar satellite for Sea ice and ice sheet studies
CRTM Community Radiative Transfer Model
CSA Canadian Space Agency
CZCS Coastal Zone Color Scanner
dB Decibels
DMSP Defense Meteorological Satellite Program (United States), also
name of a satellite
DOD Department of Defense (United States)
DORIS Doppler Orbitography and Radiopositioning Integrated by
Satellite (France)
ECMWF European Centre for Medium-range Weather Forecasts
EDR Environmental Data Record
EFOV Effective Field-Of-View; shape of the FOV after time-averaging
EM ElectroMagnetic
EMR ElectroMagnetic Radiation
ENVISAT Environmental Satellite (ESA)
EOS Earth Observing System (United States, with international
components)
ERS-1, -2 European Remote-sensing Satellite
ESA European Space Agency
ESMR Electrically Scanned Microwave Radiometer (United States)
EUMETSAT European Organization for the Exploitation of Meteorological
Satellites
FLH Fluorescence Line Height
FM Frequency Modulation
FOV Field-Of-View, see also EFOV, IFOV
FRAC Full Resolution Area Coverage (AVHRR, MODIS, VIIRS)
FY Feng Yun (Wind and Cloud) as in FY-1C and FY-1D; name of
satellite (China)
FY First Year, as in ﬁrst-year sea ice
GAC Global Area Coverage (AVHRR data mode)
Gbit Gigabit or 109
bits
GCOM Global Change Observation Missions (Japan)
GDAS Global Data Assimilation System (NCEP)
GEO Group on Earth Observations

Abbreviations and acronyms xxiii
GES DISC Goddard Earth Sciences, Data and Information Services Center
(NASA)
GEOSS Global Earth Observation System of Systems
GLAS Geoscience Laser Altimeter System (United States)
GLI Global Imager, ocean color instrument on ADEOS-2 (Japan)
GMES Global Monitoring for Environment and Security (European
satellite program)
GOCE Gravity Field and Steady-State Ocean Circulation Explorer (ESA)
GODAE Global Ocean Data Assimilation Experiment
GOES Geostationary Operational Environmental Satellite (United
States)
GHz Gigahertz
GHRSST GODAE High Resolution STT
GIOVANNI Geospatial Interactive Online Visualization ANd aNalysis
Infrastructure; often written as Giovanni
GMPE GHRSST Multi-product Ensemble (UK Met Ofﬁce)
GRACE Gravity Recovery and Climate Experiment
GSM Garver–Siegel–Maritorena algorithm (ocean biology)
HH Antenna that transmits and receives with an H-polarization
H-pol Horizontally polarized
HRD Hurricane Research Division (NOAA)
HRPT High Resolution Picture Transmission (AVHRR data transfer
mode)
HV Antenna that transmits with an H-polarization and receives with a
V-polarization
HY Haiyang (Ocean) satellite as in HY-1 (China)
IAPSO International Association for Physical Sciences of the Ocean
ICESat Ice, Cloud and land Elevation Satellite (United States)
IEEE Institute of Electrical and Electronics Engineers
IFOV Instantaneous Field-Of-View, or Instrument Field-Of-View
IJPS Initial Joint Polar-orbiting operational satellite System (United
States, EUMETSAT)
IOP Inherent Optical Properties
IPO Integrated Project Ofﬁce (NPOESS)
IR Infrared
ITCZ Inter-Tropical Convergence Zone
JASON-1, -2, -3 United States/Frame altimeter satellites (Not an abbreviation)
JAXA Japan Aerospace Exploration Agency (replaced NASDA)
JERS-1 Japanese Earth Resources Satellite
JMA Japan Meteorological Agency
JMR Jason Microwave Radiometer
JPL Jet Propulsion Laboratory (NASA), operated by CalTech

xxiv Abbreviations and acronyms
JPSS Joint Polar Satellite System
K-band Frequencies between 11 and 36 GHz
Ku-band Frequencies of about 14 GHz
KOSMOS USSR satellite series
LAC Local Area Coverage (data mode for AVHRR)
L-band Frequencies of about 1 GHz
LRA Laser Retroreflector Array
M-AERI Marine-Atmosphere Emitted Radiance Interferometer (United
States)
Mbps Megabits-per-second
MCSST Multi-Channel Sea Surface Temperature (algorithm)
MEDS Maritime Environmental Data Service (Canada)
MERIS Medium Resolution Imaging Spectrometer (ENVISAT)
METEOSAT Geosynchronous Meteorology Satellite (EUMETSAT)
METOP-A, -B, -C MÉTéorologie OPérationnelle (Operational Meteorology)
(EUMETSAT satellite)
MHz Megahertz
MOBY Marine Optical BuoY (ocean color calibration buoy near Hawaii)
MODI Moderate Resolution Visible/Infrared Imager (China)
MODIS Moderate Resolution Imaging Spectroradiometer on TERRA,
AQUA
MODTRAN Program for calculation of atmospheric transmissivity
MOS Modular Optical Scanner (Germany)
MSL Mean Sea Level
MVIRSR Multispectral Visible–Infrared Scanning Radiometer (China)
MY Multiyear, as in multiyear sea ice
NASA National Aeronautics and Space Administration (United States)
NASDA National Space Development Agency (Japan), see JAXA
NCEP National Centers for Environmental Prediction (NOAA)
NDBC National Data Buoy Center (United States)
NDT Nitrate-Depletion Temperature
NESDIS National Environmental Satellite Data and Information Service
(United States)
NIR Near-infrared
NLSST NonLinear SST (algorithm)
NOAA National Oceanic and Atmospheric Administration (United
States)
NOAA-18, -19, . . . Names of NOAA operational polar orbiting satellites
NOMAD NASA bio-Optical Marine Algorithm Dataset
NPOESS National Polar-orbiting Operational Environmental Satellite
System (United States)
NPP NPOESS Preparatory Project (United States)

Abbreviations and acronyms xxv
NRCS Normalized Radar Cross Section
NSCAT NASA Scatterometer (ADEOS-1)
NWP Numerical Weather Prediction
OC3M Ocean Chlorophyll Version 3 MODIS bio-optical algorithm
OC4 Ocean Chlorophyll Version 4 SeaWiFS bio-optical algorithm
OBPG Ocean Biology Processing Group (NASA)
OCTS Ocean Color and Temperature Sensor (Japan)
OISST Optimally Interpolated SST
OKEAN Series of satellites (Russia/Ukraine)
OLS Optical Line Scanner (visible/infrared instrument on DMSP)
OVWM Ocean Vector Wind Mission
OW Open Water (sea ice algorithms)
PALSAR Phased Array L-Band SAR (Japan)
Pixel Abbreviation for picture element
PMEL Pacific Marine Environmental Laboratory (NOAA)
POD Precision Orbit Determination
PO.DAAC Physical Oceanography Distributed Active Archive (NASA JPL)
POES Polar Operational Environmental Satellite (United States)
POLDER Polarization and Directionality of the Earth’s Reflectances
(France), ocean color instrument on ENVISAT
POSEIDON Premier Observatoire Spatial Étude Intensive Dynamique Océan
et Nivosphère, French contribution, TOPEX/POSEIDON
satellite.
PRF Pulse repetition frequency
psu Precision salinity units (units of oceanic salinity)
RA-2 Radar Altimeter-2 (ENVISAT altimeter)
RADARSAT-1, -2 SAR satellites (Canada)
RGB Red–Green-Blue color mixing
RGPS RADARSAT Geophysical Processing System (United States)
rms Root-mean-square
rss Root-sum-of-the-squares
RTE Radiative Transfer Equation
SAC-D Satelite de Aplicaciones Cient´ıficas-D
SAR Synthetic Aperture Radar
SASS SEASAT-A Satellite Scatterometer (United States)
ScanSAR Wide-swath SAR imaging mode (partial abbreviation)
SDR Sensor Data Record
SeaBAM SeaWiFS Bio-optical Algorithm Mini-Workshop
SEASAT First ocean observing satellite (1979, United States)
SeaWiFS Sea-viewing Wide Field-of-view Sensor (United States)
SeaWinds Radar vector wind instrument (not an abbreviation)
SEVIRI Spinning Enhanced Visible and Infrared Imager (EUMETSAT)

xxvi Abbreviations and acronyms
SGLI Second-generation GLobal Imager (Japan)
SIRAL SAR Interferometric Radar Altimeter (ESA)
SLAR Side-Looking Airborne Radar
SLR Side-Looking Radar
SLR Satellite Laser Ranging
SMMR Scanning Multichannel Microwave Radiometer (United States)
SMOS Soil Moisture and Ocean Salinity instrument (ESA)
SSALT Solid State Altimeter on TOPEX (France)
SSH Sea Surface Height
SSM/I Special Sensor Microwave/Imager (United States)
SSMI/S Special Sensor Microwave Imager/Sounder (SSM/I upgrade)
SSS Sea Surface Salinity
SST Sea Surface Temperature
SWH Significant Wave Height (H1/3)
TECU Total Electron Content Unit (1 TECU = 1016
electrons m−2
),
columnar concentration of free electrons
TERRA First major EOS satellite (not an abbreviation)
TIR Thermal-Infrared
TIROS-N Television Infrared Observation Satellite-N (early version of
POES satellite)
TIW Tropical Instability Waves
TMI TRMM Microwave Imager (Japan)
TMR TOPEX Microwave Radiometer
TOA Top Of the Atmosphere
TOGA-TAO Tropical Ocean Global Atmosphere–Tropical Atmosphere Ocean
TOMS Total Ozone Mapping Spectrometer
TOPEX TOPography EXperiment (United States/France)
TRMM Tropical Rainfall Measuring Mission (United States/Japan)
TRSR Turbo Rogue Space Receiver BlackJack GPS receivers (Satellite
GPS receivers used on JASON-1)
UK Met Office United Kingdom Meteorological Office
UTC Universal Time Coordinated
UV Ultraviolet
VAM Variational Analysis Method
VH Antenna that transmits with a V-polarization and receives with an
H-polarization
VIRR Visible and Infrared Radiometer (China)
VIIRS Visible/Infrared Imager/Radiometer Suite (NPP instrument)
VIR Visible/Infrared
VNIR Visible/Near-Infrared
V-pol Vertically polarized
VV Antenna that transmits and receives with a V-polarization

Abbreviations and acronyms xxvii
WindSat Polarimetric radiometer for vector wind measurements (not an
abbreviation)
WVSST Water Vapor Sea Surface Temperature (algorithm)
X-band Frequencies of about 10 GHz

CUUK2533-01 CUUK2533/Martin ISBN: 978 1 107 01938 6 November 25, 2013 13:47
1
Background
1.1 Introduction
During the past forty years, rapid technological growth has advanced the ability of satellites
to observe and monitor the global ocean and its overlying atmosphere. Because of similar
advances in computer hardware and software, it is now possible to acquire and analyze, at
short time delays, large satellite data sets such as the global distribution of ocean waves, the
variations in sea surface height associated with large-scale current systems and planetary
waves, surface vector winds and regional and global variations in ocean biology. The
immediate availability of these data allows their assimilation into numerical models, where
they contribute to the prediction of future oceanic weather and climate.
The ocean covers approximately 70% of the Earth’s surface, is dynamic on a variety
of scales, and contains most of the Earth’s water as well as important marine ecosystems.
The ocean also contains about 25% of the total planetary vegetation, with much of this
restricted to a few coastal regions (Jeffrey and Mantoura, 1997). Regions of high biological
productivity include the Grand Banks off Newfoundland, the Bering Sea and Gulf of
Alaska, the North Sea and the Peruvian coast. Between 80% and 90% of the world’s ﬁsh
catch occurs in these and similar regions. For its role in climate, determination of the
changes in ocean heat storage and measurement of the vertical ﬂuxes of heat, moisture and
CO2 between the atmosphere and ocean are critical to understanding global warming and
climate change.
Large-scale ocean currents carry about half of the heat transported between the equator
and the poles; the atmosphere transports the remainder. Away from the polar regions,
the combination of these transports with the large oceanic heat capacity relative to the
atmosphere means that the ocean moderates the global climate and improves the habitability
of the continents (Stewart, 1981; Chelton, 2001). For the polar regions, the recent increase
in the melting of the Greenland and Antarctic icecaps and the dramatic decrease in the
arctic summer sea ice cover (Comiso, 2010) show that the ability to monitor the extent and
thickness of the Arctic and Antarctic ice covers is important both for short-term navigation
needs and for long-term climate studies. All these examples illustrate the need to monitor
and observe the ocean on a range of local to global scales.
1

2 Background
The growth in satellite systems has been driven in part by technology and in part by
societal concerns. Societal concerns include the importance of the ocean to national security
and naval operations, global commerce, the prediction of severe storms and hurricanes,
fisheries management, the extraction of offshore gas, oil and minerals, and public health
and recreation. Regarding commerce, in 2012, there were about 100 000 ships engaged
in commerce, oil, gas and mineral exploration, fisheries and recreation (Allianz, 2012).
Increasingly, these concerns also include global sea level rise and the change in the areal
extent of the Arctic and Antarctic sea ice. In addition, about half of the global population
lives within 200 km of the coast, where fourteen of the seventeen largest cities are coastal. Of
these, eleven are Asian, including Bangkok, Jakarta, Shanghai, Tokyo, Ho-Chi-Minh City,
Calcutta and Manila (Creel, 2003). These populations are vulnerable to natural hazards
such as the storm surge and flooding associated with the combination of sea level rise
and hurricanes or typhoons. There are also public heath considerations associated with
the oceanic disposal of urban runoff, sewage and garbage, and with the monitoring and
prediction of the growth of pathogenic organisms such as red tides. Satellite observing
systems and the interpretation of the resultant data play a central role in addressing these
concerns.
In the 1970s, the United States launched the first ocean remote sensing satellites. Since
that time, many countries have launched satellites that carry oceanographic instrumenta-
tion, and, as Section 1.8 describes, beginning in about 2002 there has been an international
effort to organize satellites from different countries into what are called observing “con-
stellations”. These constellations are made up of satellites that carry similar instruments,
observe the same oceanic variables and fly in complementary orbits, so that the coverage
by a single satellite is enhanced by observations from the other constellation members. The
data from the constellation are then placed in a common format and distributed among the
participants and other interested parties.
With these observations, there is an emphasis on the rapid dissemination of the data to
the various government and private-sector users, and the use of this near-real-time data in
numerical models and in other areas such as search-and-rescue, oceanographic research
cruise support and the routing of cargo ships to avoid storms. Examples of the oceanic
variables observed by these satellites include sea surface temperature (SST), the height and
directional distribution of ocean swell, wind speed and direction, atmospheric water content
and rain rate, the changes in sea surface height associated with ocean tides, currents and
planetary waves, concentrations of phytoplankton, sediments and suspended and dissolved
material, and the areal extent and types of polar sea ice.
Prior to the 1980s, such properties were determined from dedicated and expensive ship
expeditions, or in the polar regions from surveys made from aircraft, drifting ships and
ice islands. This meant that the ocean could be surveyed only slowly and incrementally.
At present, satellite imagers can make simultaneous observations of the desired variables
with scales of 1–1,000 km that are difficult to observe even from multiple ships. For
variables such as the near surface air temperature that are not retrievable by remote sensing,
some satellites are designed to relay data from moored and drifting buoys that make direct

1.2 Definition of remote sensing 3
measurements of such quantities to national data centers. Even for those ocean depths that
are inaccessible to satellite observations, instruments called Argos floats are deployed in
large numbers that profile the ocean interior and periodically come to the surface, where
they report their observations by satellite.
Because satellites survey a variety of oceanic properties with near global coverage
and at intervals of 1–10 days, then rapidly transmit these observations to national and
international forecast centers, these data are of great operational importance. In addition,
the observations contribute to long-term studies and numerical modeling of global climate
change, sea level rise, and the decadal-scale atmospheric and oceanographic oscillations,
including the Pacific Decadal Oscillation (PDO), North Atlantic Oscillation (NAO), El
Niño/Southern Ocean Oscillation (ENSO), and Arctic Oscillation (AO).
In the following, Section 1.2 defines remote sensing and describes its oceanographic
applications. Section 1.3 describes the satellite orbits used in remote sensing and summa-
rizes the hazards faced by satellites. Sections 1.4 and 1.5 describe the geosynchronous and
Sun-synchronous satellites. Section 1.6 discusses the imaging techniques used by satellites
in Sun-synchronous and other low Earth orbits. Section 1.7 describes the different process-
ing levels of satellite image data and the NASA data archives. Section 1.8 gives a brief
history of the changes in satellite remote sensing over the past forty years, describes the
international context of these observations, and presents a table of past, present and pending
satellite missions through 2015.
1.2 Definition of remote sensing
Earth remote sensing is primarily defined as the use of electromagnetic radiation to acquire
information about the ocean, land and atmosphere without being in physical contact with
the object, surface or phenomenon under investigation. Remote sensing is not unique
to electromagnetic radiation, as this book shows, there are also techniques for studying
changes in ocean circulation and ice sheet properties through observations of gravity
anomalies. Unlike shipboard measurements of quantities such as SST or wind speed, which
are direct measurements made at a point by a thermometer or anemometer, remote sensing
measurements of such quantities cover broad areas and are indirect, in that the geophysical
quantity of interest is inferred from the properties of the reflected or emitted radiation.
The sensors can range from a radiometer mounted on a ship, oil platform or aircraft to a
multispectral satellite imager. The following briefly describes the concepts behind remote
sensing and the various observing bands.
Because the satellite instrument is not in physical contact with the phenomena under
investigation, its properties must be inferred from the intensity and frequency distribution of
the received radiation. This distribution depends on how the received radiation is generated
and altered by its propagation through the atmosphere. This radiation has three principal
sources: blackbody radiation emitted from the surface, reflected solar radiation, and, for
the directed energy pulses transmitted by satellite radars, the backscattered energy received

4 Background
at the sensor. The properties of the received radiation also depend on the sensor, which
must be designed so that its observing wavelengths are appropriate for the phenomenon in
question. Finally, the received data must be organized into images or data sets so that the
spatial distributions of the quantity under investigation can be viewed. This is the generally
accepted definition of remote sensing; in the past decade, it has been expanded to include
the use of satellite measurements of gravity to infer changes in land, ice sheet and ocean
properties.
Because of the atmospheric contributions to the reflected and received radiation
described in Chapters 4 and 9, there are three electromagnetic wavelength bands or win-
dows, called the visible, infrared and microwave, through which the ocean is viewed. In the
visible and extending into the near infrared, the observations depend on reflected sunlight
and are restricted to daytime cloud-free periods. Because the visible spectrum contains the
only wavelengths at which light penetrates to oceanic depths of order 10–100 m, visible
observations yield the only information on the depth-averaged color changes associated
with phytoplankton and sediment concentrations. In the infrared, the observations measure
the blackbody radiation emitted from the top few micrometers of the sea surface. Although
these observations are independent of daylight, infrared satellite observations are restricted
to cloud-free conditions.
In the microwave and especially at the longer microwave wavelengths, the surface
can be viewed through clouds and is obscured only by heavy rain. Microwave observations
divide into passive and active. Passive microwave instruments observe the naturally emitted
blackbody radiation, which can be used to retrieve such atmosphere and ocean surface
properties as the areal extent of ice cover, the atmospheric water vapor and liquid water
content, sea surface temperature (SST), salinity, and, through the directional dependence
of the sea surface roughness, the vector wind speed.
In contrast, different kinds of radars make active measurements; these instruments
transmit pulses of energy toward the ocean, then receive the backscatter, so that they
provide their own illumination. The active microwave instruments include imaging radars
(the Synthetic Aperture Radar or SAR), directed, pulsed vertical beams (altimeter), several
pulsed fan beams at oblique angles to the satellite orbit (scatterometer), and an oblique
rotating pulsed beam (also scatterometer). The scatterometers are highly directional radars
that receive the backscatter from relatively small surface areas. Together, these instruments
provide information on the roughness and topography of the sea surface, wind speed and
direction, wave heights, directional spectra of ocean surface waves and the distribution and
types of sea ice.
1.3 Satellite orbits
The orbit of an Earth-observing satellite divides into two parts, the satellite motion in
its orbit plane relative to the Earth’s center of mass, and the satellite position relative to
the rotating Earth. The time-dependent position of the satellite in its orbit is called the

satellite ephemeris. For the rotating Earth, the orbit is frequently described in terms of its
ground track, which is the time-dependent location of the surface intersection of the line
between the satellite and the Earth’s center of mass. The point directly beneath the satellite
is called the satellite nadir. The first of the following sections considers the theoretical case
of satellite motion in its orbit plane, and describes how the addition of the Earth’s rotation
determines the satellite ground track; the second considers the actual space environment
of these satellites, and the constraints imposed on the satellites and their instruments by
space debris and uncontrolled satellites, gravity-induced orbit perturbations, solar storms
and radiation, and radio-frequency interference (RFI).
1.3.1 Satellite orbits and their applications
Rees (2001, Chapter 10), Elachi (1987, Appendix B) and Duck and King (1983) survey
the commonly used, near circular orbits used in remote sensing. These orbits are described
in a rectangular coordinate system with its origin at the Earth’s center of mass. The z-axis
is in the northerly direction and co-located with the Earth’s rotation axis, the x-axis is
in the equatorial plane and points in the direction γ of a star in the constellation Aries, and
the y-axis is in the direction appropriate for a right-handed coordinate system. Relative to
these axes, the six Keplerian orbital elements describe the satellite location. Because two
of these are specific to elliptical orbits, for circular orbits, the six elements are reduced to
four.
As Figure 1.1 shows, these four elements are as follows. First, the right ascension of
the ascending node, or simply the ascending node , is the angle between the x-axis and
the point at which the orbit crosses the equator. Second, the radial distance H is the height
of the satellite above the Earth’s center of mass. Third, the orbit true anomaly θ is the
angular position of the satellite in its orbit relative to . Fourth, the inclination I is the
angle between the Earth’s axis and the normal to the orbit plane with the convention that
I is always positive. Of these variables, I and specify the orientation and position of
the orbit plane relative to the fixed stars; H and θ specify the satellite position within the
orbit plane. The advantage of this description is that I, and H are either fixed or slowly
varying, so that, over short periods, θ describes the instantaneous satellite position. Based
on the magnitude of I, there are three kinds of orbits. If I = 90°, the orbit is polar; if I <
90°, the orbit is prograde and precesses in the same direction as the Earth’s rotation as in
Figure 1.2; if I > 90°, the orbit is retrograde and precesses in the opposite direction.
In remote sensing, interest is generally not in the satellite position in its orbit, but rather
in its location on its surface ground track. For a non-rotating spherical Earth, the orbit
track is a great circle, or, on the Mercator map shown in Figure 1.2(a), a simple sine wave
(Elachi, 1987, Section B-1–4). Because of the Earth’s rotation, the orbit track is steadily
displaced to the west, yielding the succession of tracks shown in Figure 1.2(b). On the
tracks, the numbers i, ii, iii mark the beginning and end of each orbit, where, for example,
the points marked ii are at the same time and geographic location. Another orbit property

6 Background
N
I
Earth’s rotation
Prograde
orbit
Ω
Ω
z
x
y
Equator
Normal
to orbit plane
Equator
plane
H
x
Ascending node
θ
γ
γ
y
View from North Pole
N
Fig. 1.1. For a circular orbit, the Keplerian parameters used to describe the orientation of the orbit
plane and the satellite position along the orbit.
Equator
N
Equator
N
EastWest
EastWest
Earth’s rotation
LE
iii
0o
360o
360o
iii iii
iii
Orbit
displacement
(a)
(b)
Fig. 1.2. Mercator map of the satellite ground track for the orbit shown in Figure 1.1 and for
(a) non-rotating Earth and (b) rotating Earth. See the text for further description. (Adapted from
Elachi (1987, Figure B-6).

N
S
Geosynchronous orbit
Sun-synchronous orbit
Low-inclination orbit
Eq
Fig. 1.3. Examples of the Sun-synchronous, geosynchronous and low-inclination orbits, where “Eq”
is the equator. (Adapted from Asrar and Dozier (1994), Figure 3).
concerns the equatorial separation LE between successive orbits. If division of a multiple
of the equatorial circumference by LE is an integer, the orbit is an exact repeat orbit, so
that, after a given period of time, the satellite repeats the same track lines. This property
is particularly valuable for instruments such as the altimeter, since it allows successive
measurements of sea surface height along the same ground track.
The three common Earth observation orbits are called geosynchronous, Sun-
synchronous, and near equatorial low inclination (Figure 1.3). There is also a fourth
altimeter orbit used for observations of sea surface topography that is at a slightly higher
altitude than the Sun-synchronous orbits, and there are also various low-altitude non-Sun-
synchronous orbits used for observations of phenomena such as winds and rainfall. The
following summary shows that each particular orbit has advantages and disadvantages.
Because no single orbit allows coverage of all space and time scales, there is no such
thing as a “perfect” satellite orbit or system. Instead, the choice of orbit depends on the
phenomenon under investigation.
The geosynchronous orbits are located at an altitude of 35 800 km above the equator.
The geostationary orbit is a special case; it lies in the Earth’s equatorial plane (I = 0°). In
this orbit, although the satellite is orbiting the Earth such that it moves in and out of the
Earth’s shadow, its position remains over a fixed equatorial location so that it continuously
observes the same surface area. The plane of the more general geosynchronous orbit is
tilted relative to the equator (I = 0°), so that, although the mean surface position of this
satellite is stationary, its ground path is described by a figure eight centered on the equator
(Elachi, 1987). The period of a geosynchronous satellite is 23.93 hours, which is the time
in which the Earth rotates around its axis relative to the fixed stars. In contrast, the 24-hour
day is the time between successive noons, defined as when the sun is directly overhead, so
that the length of day is determined from a combination of the Earth’s rotation about its
axis and the Earth’s rotation in its orbit.

8 Background
Satellite
orbit plane
Earth
90 days
later
Orbitplane
Fig. 1.4. Rotation of the plane of a Sun-synchronous orbit in the Earth–Sun orbit plane.
Operators and managers of geosynchronous satellites work in terms of a “geosyn-
chronous belt”, deﬁned as the region extending 200 km above and below the geosyn-
chronous altitude and ±15° in latitude (IADC, 2007; Weeden, 2010). Within this belt, the
satellites occupy slots that measure about 2° in longitude, where their operators try to main-
tain the satellite within a 0.1° box (Weeden, 2010). In Earth observations, geosynchronous
satellites provide observations of weather, SST and ocean color, and provide data relay
services.
The Sun-synchronous orbit is retrograde with I > 90°, and has an altitude of about 800
km, or a much lower altitude than the geosynchronous orbits. The Sun-synchronous period
is about 90 minutes, corresponding to about sixteen orbits per day. The reason why this
orbit is called Sun-synchronous is that throughout the year each orbit crosses the equator
at the same local time of day. Consequently, is not constant, but changes slowly with
time. The drift occurs because of the Earth’s equatorial bulge, which causes the plane of
a near polar orbit to rotate slowly around the pole (Rees, 2001). For a retrograde orbit,
the inclination and orbit height can be set so that the orbit rotates about 1° per day in the
ecliptic or Earth–Sun plane, and in an equal but opposite direction to the orbital motion
of the Earth around the Sun. Relative to the ﬁxed stars, the Sun-synchronous orbit plane
rotates once per year, so that its orbit plane remains at a constant angle to the line between
the Sun and Earth. Figure 1.4 shows the change in the angular position of the orbit in the
Earth–Sun plane as the Earth moves an angular distance of 90° in its orbit, during a period
of approximately 90 days.
Sun-synchronous satellites are the most common of the ocean-observing satellites and
are often referred to as polar orbiters. Their orbits are described in terms of their daytime
equatorial crossing times, as in a 0730 descending or a 1330 ascending orbit, where
descending refers to a southward satellite velocity, ascending refers to a northward velocity,
and the crossing time is local. The orbits are also described in terms of their crossing times,

as “early morning”, “mid-morning” and “early afternoon”. Because the Sun-synchronous
equator crossings always occur at the same local time of day, satellites in this orbit can
make daily observations of SST or ocean chlorophyll at the same time in their diurnal cycle.
Since cloudiness over the ocean generally increases throughout the day, the crossing time
can be chosen to minimize cloudiness under the satellite.
One difficulty with this orbit is that, because of the tilted orbit plane, the satellite does not
pass directly over the poles. This means that the regions around the poles may be excluded
from instrument coverage; this lack of coverage is called the “hole at the pole”. Figures
4.2 and 9.18 give examples of the swath coverage for this orbit, and show that, depending
on the instrument, a single Sun-synchronous satellite can provide near global coverage at
1–2-day intervals.
The near-equatorial low-inclination orbit used for missions such as the Tropical Rainfall
Measuring Mission (TRMM) is circular with an altitude of 350 km and an inclination angle
of 35°. This orbit covers approximately half the globe, and, in a one-month period, observes
any specific area at every hour of the day with a sampling rate that is roughly twice that
of a polar orbiter. The advantage of this orbit is that it allows TRMM to determine the
variability of tropical rainfall throughout its diurnal cycle. The successor to this mission
is the joint US/Japanese Global Precipitation Measurement (GPM) Core mission, with a
greater inclination angle of 65° that is scheduled for launch in 2014. Another member of the
GPM constellation in a similar orbit is the Indian/French Megha-Tropiques rainfall mission
with an inclination angle of 22° that was launched in 2010.
Finally, the altimeter occupies an orbit designed to measure sea surface height. Because
the tidal bulge associated with the 12- and 24-hour tides always lies directly beneath a
satellite in a Sun-synchronous orbit, some altimeters operate at a higher non-synchronous
altitude of 1200–1400 km. Consequently, the orbit is not in phase with the tides and the
satellite experiences a smaller atmospheric drag. Altimeter satellites in this orbit include
the US/French TOPEX/POSEIDON JASON-1, JASON-2 and the forthcoming JASON-3
mission discussed in Chapter 12.
1.3.2 The satellite environment: Solar storms, radiation pressure, the
South Atlantic Anomaly, gravitational perturbations, space debris, graveyard
orbits and radio frequency interference (RFI)
In space, various factors perturb the satellites, their orbits and their instruments. First, the
lunar and solar gravity fields and radiation pressure from the solar wind exert forces on
the satellites and perturb their orbits. Second, there are two bulges in the Earth’s gravity
field called libration points, one over India (105° W) and the other at the longitude of the
US Rocky Mountains (75° W), that also affect the orbits (Weeden, 2010). For this reason,
all satellites have engines and carry fuel so that they can maintain their desired orbits.
Third, the satellite can be damaged or destroyed by collisions with space debris or other,
sometimes decommissioned, satellites.

10 Background
The NASA Orbital Debris Program Office (NASA, 2012a) monitors space debris; ESA
(2012a) describes the ESA monitoring of debris. As of 2009, ESA (2012a) states that there
were 14 000 catalogued pieces of space debris, and approximately 600 000 uncatalogued
pieces of debris with dimensions greater than 1 cm. Depending on their relative velocity,
even a small object can damage or destroy a satellite. In the low Earth orbits (LEO), the
maximum amount of debris occurs at two altitudes: the polar orbit altitudes at 800–1000
km and the altimeter satellite altitude of 1400 km. For the geosynchronous belt, the amount
of debris is about two orders of magnitude less than in LEO.
ESA (2012a) describes the growth in the amount of debris and its sources. For example,
in January 2007, the Chinese use of an anti-satellite missile to destroy the Sun-synchronous
Feng-Yun 1C satellite led to a 25% increase in catalogued debris. In February 2009, the
first accidental collision of two satellites occurred in LEO when the American commercial
satellite, Iridium-33, collided with a Russian military satellite, Kosmos-2251, destroying
both satellites and generating a large amount of debris. For the rest of 2009, five satellites,
namely the remote sensing satellites AQUA and Landsat-7 at altitudes of about 700 km,
the Space Station and Space Shuttle at an altitude of 400 km, and a NASA Tracking and
Data Relay Satellite (TDRS-3) in geosynchronous orbit, maneuvered to avoid collisions
with debris (David, 2010). Based on the current growth in satellite debris, Donald Kessler
has forecast the occurrence of what is called a “Kessler” syndrome or cascade, where the
frequency of collisions will increase at such a rate and generate so much debris that all of
the satellites in LEO would be destroyed (Kessler interview in David, 2010).
For geosynchronous satellites, Weeden (2010) states that, in 2010, there were 1238
catalogued objects in the geosynchronous belt, of which 391 were under control, 594 were
drifting, 169 had been captured by the libration points, and the remainder were lost or
undocumented. He also describes the fate of the Intelsat Galaxy-15 satellite that, during a
solar storm in April 2010 when the satellite was positioned at 130° W, lost contact with
its ground controllers. Because of this, it drifted east toward the North American libration
point, and received the nickname “Zombiesat”. As it drifted east, its transponders continued
to receive and transmit data broadcast from the ground, causing both radio interference and
hazards to other satellites. This situation continued until January 2011, by which time the
satellite had passed through the orbital slots of about fifteen communication satellites, when
Intelsat restored communications with Galaxy-15, and returned it to a safe position (Space
News, 2011).
Given these problems with space debris, 11 nations with space programs and ESA formed
the 12-member Inter-Agency Space Debris Coordination Committee (IADC, 2012). The
IADC recommends that, to avoid further generation of debris, two protected regions be
established. The first contains the LEO, which IADC defines as the global region extending
in altitude from the surface to 2000 km, and covering the Sun-synchronous and altimeter
orbits; the second contains the geosynchronous orbits (GEO). For LEO, IADC (2007)
recommends that, when the satellite approaches the end of its lifetime, it be deorbited into
the atmosphere. For GEO, IADC recommends that a satellite approaching its end of service
should be placed into a graveyard orbit located at an altitude of about 100–200 km above

1.000
0.500
0.200
0.100
0.050
0.020
0.010
0.005
0.002
–135 –90 –90 0 45 90 135 180
60
45
30
15
0
–60
–45
–30
–15
Fig. 1.5. Graphic of the South Atlantic Anomaly (SAA) showing the contours of the relative prob-
ability for space systems to suffer single anomalous events caused by high-energy protons at an
altitude of 1000 km. See the text for further description. (Reprinted from Brautigam (2002, Figure
8), copyright 2002, with permission from Elsevier.)
the geosynchronous belt. For both sets of orbits, to minimize the generation of debris by
break-up of the satellites, all fuel tanks should be depressurized and any energy contained
in momentum wheels should be depleted.
Another satellite hazard is that solar storms and flares generate highly charged particles
that can cause temporary or permanent damage to satellite electronics. Such storms are
monitored by the NOAA Space Weather Prediction Center (SWPC), which issues warnings
to satellite operators (SWPC, 2012). These particles are primarily a problem at GEO
altitudes, but for LEO, and as Brautigam (2002) describes, they occur in a location over
South America called the South Atlantic Anomaly (SAA). The SAA is a permanent anomaly
in the Earth’s magnetic field, generated by the misalignment between the axis of the Earth’s
rotation and the axis of the magnetic field. This misalignment means that the charged
particles in the Van Allen belt dip down toward the Earth’s surface in an area over Brazil
and the South Atlantic Ocean (Figure 1.5). Within this region, high-energy protons can
cause temporary or permanent damage to the spacecraft electronics. Dodd et al. (2010)
describe the effect of the SAA on the Moderate Resolution Imaging Spectroradiometer
(MODIS) instrument on the AQUA and TERRA spacecraft. For these satellites, the high-
energy particles can reduce the efficiency of instrument detectors and can cause bits to flip
spontaneously in computer circuitry, which led to a decision that, when the spacecraft is in
the SAA, no critical commands are to be sent to it.
Finally, as Chapter 9 discusses in more detail, in the microwave, the limited spectrum
available for remote sensing observations and the presence of many other broadcast sources

12 Background
strongly affect the satellite observations by causing radio-frequency interference (RFI). As
Chapter 9 discusses, the growth in the number of direct broadcast satellites, including
satellite radio, television and telephone, the existence of powerful space observation radars
and the pressures to open up new radio spectra for these purposes and for cellular commu-
nications at the surface have increased the presence of RFI, led to a reduction in the width
of bands used for Earth observations, and, in some cases, reduced the global coverage of
the remote sensing observations.
1.4 Geosynchronous satellites
The geosynchronous satellites important to oceanography include observation, weather
and data relay satellites. The website GOES (2012) summarizes the different kinds of
geosynchronous satellites, which are classified according to their scanning methods, called
spin-scan and fixed orientation. The spin-scan satellites consist of a cylindrically symmetric
spinning part, mounted on a non-spinning section that contains the antennas for broadcasting
the data to ground stations. The spinning section is oriented such that its long axis is parallel
to the Earth’s rotation axis, where its rotation rate is about 100 revolutions per minute. On
each spin, a visible/infrared sensor sweeps across the Earth’s disk where the resultant data
are stored or broadcast. On the next revolution, the north–south sensor view angle changes
slightly, and the scan is repeated. From such multiple scans, it takes about 20 minutes
to create an image of the Earth’s disk. The spinning helps keep the satellite in thermal
equilibrium and stabilizes the satellite in its orbit.
Satellites that use this technique are the European Meteosat series and the out-of-service
Japanese Geostationary Meteorological Satellite (GMS) series (GOES, 2012).
Newer satellites such as the US Geostationary Operational Environmental Satellites
(GOES) series have a fixed orientation and use a different scanning technique. For this
case, the images are acquired by a scanner that employs two mirrors, one sweeping across
the Earth’s disk, the other stepping north-to-south. The future EUMETSAT and Japanese
satellites will employ similar systems.
The two European agencies involved with ocean remote sensing are the European Space
Agency (ESA), founded in 1973, and the European Organization for the Exploitation
of Meteorological Satellites (EUMETSAT), founded within ESA in 1986. ESA has the
overall responsibility for space programs; EUMETSAT manages the geosynchronous and
Sun-synchronous weather satellites (EUMETSAT, 2012). In 2012, the ESA governing
council included members from nineteen countries: Austria, Belgium, the Czech Republic,
Denmark, Finland, France, Germany, Greece, Ireland, Italy, Luxembourg, the Netherlands,
Norway, Portugal, Romania, Spain, Sweden, Switzerland and the United Kingdom. Under
a special agreement, Canada is also a member of the council (ESA, 2012b).
A network of geosynchronous weather satellites provides global coverage between
±60° latitude. As of February 2012, NOAA maintains two GOES satellites. These satel-
lites, called GOES East and GOES West are located over the equator at approximately

1.5 Sun-synchronous satellites 13
60
0
30
–60
–30
0 30–30 60–60 90–90 120–120 150–150 180–180
GOES-East
75 W
GOES-West
135 W
MTSAT
145 E
Meteosat
0 E
Meteosat
57 E
Fig. 1.6. Field-of-view of the five geosynchronous meteorological satellites that provide near-global
coverage. The boxes give the names of the satellites and their center longitudes; the ovals show their
respective coverage. See the text for further description. (Reprinted from Vignola et al. (2012, Figure
6), copyright 2012, with permission from Elsevier.)
75° W and 135° W, or at the longitudes of the east and west coasts of the United States.
EUMETSAT maintains two spin-scan geosynchronous weather satellites called Meteosat,
one over the Atlantic at approximately 0° and the other over the Indian Ocean at about
60° E. Russia and India also maintain satellites at 75° E, although India generally reserves
its data for domestic use. Japan maintains its geosynchronous weather satellite, called the
Multi-functional Transport SATellite-2 (MTSAT-2) at 145° E. Consequently, the globe is
covered by five overlapping fields-of-view (Figure 1.6), placed at approximately equal
intervals around the globe, with a sixth from China at 105° E.
These five satellites produce publically available imagery at about 3-hour intervals. Even
though these imagers cannot view the polar regions, they provide sequential visible and
infrared imagery of clouds and SST patterns at 20–30-minute intervals for the equatorial
and temperate latitudes. The second class of geosynchronous satellites is constituted by the
data relay satellites, which transfer data from the polar orbiters to the ground. The United
States maintains the Tracking and Data Relay Satellite System (TDRSS) that consists of
about four active satellites and three on standby. TDRSS is the primary communication
link between the TERRA and AQUA spacecraft and the surface. ESA, China and Japan
also maintain data relay satellites.
1.5 Sun-synchronous satellites
Several countries maintain operational Sun-synchronous satellites with oceanographic
instrumentation, where the term operational means that the data from these satellites are

14 Background
regularly used in oceanographic or atmospheric forecasting. In the United States, three
government agencies operate satellites with ocean applications. The National Aeronautics
and Space Administration (NASA) maintains a series of research satellites, the National
Oceanic and Atmospheric Administration (NOAA) maintains the operational meteorologi-
cal and oceanographic satellites, and the Department of Defense (DOD) maintains the two
Defense Meteorological Satellite Program (DMSP) meteorological satellites with oceano-
graphic applications that are administered by NOAA. Other operational Sun-synchronous
satellite programs include the Russian Meteor series and the Chinese Feng Yun (Wind and
Cloud) FY-1C and FY-1D series.
In the United States, the NOAA satellites are launched by NASA, administered by
NOAA, and carry instruments from France and the United Kingdom. Previous to 1994, the
DOD and NOAA maintained parallel sets of operational satellites. For NOAA, the Polar
Operational Environmental Satellite (POES) program administered these satellites, which
were called POES or NOAA satellites. The DMSP satellites carry the visible–infrared
Optical Line Scanner (OLS) and the passive microwave Special Sensor Microwave/Imager
(SSM/I). As Chapters 9 and 10 discuss, the SSM/I and the post-2003 Special Sensor
Microwave Imager/Sounder (SSMI/S) modification of the SSM/I provide time series of sea
ice extent.
The POES satellites were built by NASA and operated by NOAA. During construction
and before launch, these satellites are described by letters, as in NOAA-K; after launch
they are described by numbers, so that, for example, NOAA-K became NOAA-15. In
addition to a variety of instruments used to gather atmospheric data as input to numerical
weather forecasts, the principal oceanographic instrument on the NOAA satellites is the vis-
ible/infrared Advanced Very High Resolution Radiometer (AVHRR) used for SST retrieval.
AVHRR observations began in 1978 with the launch of the Television Infrared Observation
Satellite-N (TIROS-N); the first AVHRR specifically designed for SST retrieval was the
AVHRR/2 launched in 1981 on NOAA-7. The AVHRR data are continuously broadcast in
an open format, so that with the use of a relatively simple ground station these data can be
downloaded over most of the globe. As Chapter 7 discusses, AVHRR observations provide
a three-decade time series of global SST.
Like their current replacements, the NOAA satellites operated at altitudes between 830
km and 870 km, where the orbit of the morning satellite was such that the satellite descended
or moved south across the equator with local crossing time of 0730, while the orbit of the
afternoon satellite had an ascending equator-crossing time of 1330. For POES, because
the crossing times of the two satellites are approximately 6 hours apart, with nighttime
equator crossings of approximately 1930 ascending and 0130 descending, the satellites
acquired imagery from almost every point on the Earth’s surface at 6-hour intervals. For
comparison, the DMSP satellites operate at a nominal altitude of 830 km with dawn–dusk
crossing times.
In 1994, a presidential decision transferred the management of all these satellites to the
new National Polar-orbiting Operational Environmental Satellite System (NPOESS). The
purpose of NPOESS was to reduce the number of operational satellites from four to three,
of which the United States would provide two satellites; the Europeans, one. NPOESS

also transferred operation of the DMSP satellites to NOAA. As part of this transition,
the European MÉTéorologie OPérationnelle-A (Operational Meteorology or METOP-A)
satellite launched in October 2006 joined the observing constellation.
NPOESS also carried out the planning and construction of the NPOESS Preparatory
Project (NPP) satellite, designed to be the transition between POES and NPOESS. Although
NPP was completed and launched in October 2011, then renamed the Suomi-NPP after the
inventor of the spin-scan satellite, the construction costs of the other NPOESS satellites so
greatly exceeded their budget that in February 2010 the NPOESS program was terminated.
Its replacement is the Joint Polar Satellite System (JPSS), which is a collaboration between
NOAA and NASA, where NOAA operates the satellites and NASA acquires them (JPSS,
2013a). In 2013, the JPSS space segment consists of the Suomi-NPP in an early afternoon
orbit, a DMSP satellite in a dawn–dusk orbit and METOP-B in a mid-morning orbit. In
about 2017, the satellite JPSS-1 will replace Suomi-NPP, where JPSS-1 has a 7-year lifetime
and will carry the same instruments as Suomi-NPP (JPSS, 2013b).
The coverage of these satellites is as follows. The DMSP satellite is in early morning
orbit with a descending equator-crossing time of 0530 local. The next in the series is the
mid-morning METOP-B satellite with a descending crossing time of 0930 local, where
METOP-B also carries an AVHRR. Finally, Suomi-NPP has an early afternoon ascending
crossing time of 1330 (CGMS, 2012). These three satellites provide coverage of most
of the Earth’s surface at 4-hour intervals. Suomi-NPP carries the replacement for the
AVHRR, called the Visible/Infrared Imager/Radiometer Suite (VIIRS). Chapter 7 describes
the AVHRR; the following and Chapters 6 and 7 describe VIIRS.
1.6 Imaging techniques
Satellites use several scanning methods to generate images. As Section 1.4 describes,
the geosynchronous satellites use spin-scan or fixed-orientation step-scanners to acquire
images. For the Sun-synchronous and other low Earth orbits, in the visible/infrared satellites
use different but related scanning techniques to generate images. As Chapters 8, 10 and 14
show, different scanning methods are used by passive and active microwave instruments.
Section 1.6.1 describes the geometry used for a sensor viewing the Earth’s surface, then
show for a simple telescope how the surface field-of-view changes with view angle. Sections
1.6.2–1.6.4 discuss three scanning techniques used with low Earth orbits called cross-track
or whiskbroom, along-track or pushbroom, and what this book calls hybrid whiskbroom,
where each of these depends on the satellite motion along its trajectory. Section 1.6.5
concludes with a discussion of resolution.
1.6.1 Viewing the Earth’s surface
Figure 1.7 shows the terminology and geometry for a satellite sensor viewing the Earth’s
surface. On this figure, the point on the surface beneath the satellite is its nadir point; the
point observed by the instrument is its scan point. Zenith means directly overhead. The

16 Background
θ
θV
Sensor
S
Satellite
nadir point
Satellite
scan point
Sun
θ
Fig. 1.7. The angles used to describe the sensor view direction and the solar angle relative to a
spherical Earth. θV is the view or scan angle that is associated with the satellite sensor and defined
relative to satellite nadir. θ is the viewing zenith angle and θS is the solar zenith angle, both defined
relative to the local vertical at the satellite scan point.
angle between the nadir line and the instrument look direction is the scan angle θV and, at
the scan point, the angle between the view direction and the local vertical is the viewing
zenith or look angle θ. At off-nadir view angles, θ and θV differ because of the Earth’s
curvature. The figure shows that the solar zenith angle θS is also measured relative to the
local vertical. Given that oceanic surface properties are functions of the viewing zenith
angle θ, the following chapters primarily use θ to describe the operation of the satellite
instruments (View angles, 2013).
Many optical instruments employ telescopes with circular lenses and apertures to view
the Earth at a variety of view angles (Figure 1.8). For this case, the instrument solid angle
= A/r2
is a constant, where A is the surface area observed by the telescope at
nadir and r is the distance from the instrument to the surface. The surface area is also
called the instrument field-of-view or equivalently the instantaneous field-of-view (IFOV),
or often simply the field-of-view (FOV). For a nadir view, the FOV is a circle; because of
the Earth’s curvature at off-nadir view angles, the FOV is an ellipse.
1.6.2 Cross-track or whiskbroom scanners
The next sub-sections describe three scanning techniques that are primarily used in the vis-
ible/infrared and in low Earth orbits, while Chapter 8 describes the analogous microwave
scanners. First, whiskbroom scanners construct images from the combination of the satel-
lite motion along its trajectory and the rotation of a telescope–mirror combination relative
to the spacecraft. For these instruments, three directions describe the scan: along-track is
in the direction of the satellite trajectory, cross-track is at right angles to the trajectory

r
ΔA
ΔΩ
ΔΩ
Sensor
Earth’s surface
θ V
Ω
Fig.. 1.8. The surface area observed by an optical instrument with a constant-solid-angle ﬁeld-of-
view, for nadir and off-nadir view angles.
Rotating mirror
Calibrator
λ
Satellite nadir track
Scan
direction
Field-of-view
Swath width
Cross-track
Along-track
λ λ λ λ λ
(a) (b)
Along-scan
Detector
1
1 2 3 4
5
Fig. 1.9. Schematic drawing of a cross-track or whiskbroom scanner. The circles show the ﬁelds-
of-view. The gray ellipse shows the instrument FOV. The radiation from the FOV is focused on the
detector, also shown in gray. (a) Single-Wavelength scanner. (b) Multi-wavelength scanner. The λ1
are the center wavelengths of the detectors.
and along-scan is in the scan direction of the sensor on the surface. Examples of whiskb-
room instruments include the AVHRR and the Sea-viewing Wide Field-of-View Sensor
(SeaWiFS).
For this scanner, Figure 1.9 shows a schematic drawing of the surface scanning pattern
and operation of idealized single and multichannel instruments. The single-channel scanner
in Figure 1.9(a) collects radiation from the FOV at a single wavelength band; the multichan-
nel scanner in Figure 1.9(b) collects radiation from the same FOV at several wavelength

18 Background
bands. The instrument operates as follows. For each wavelength band, the detectors are
focused on a mirror mounted at a 45° angle to its axis of rotation that rotates uniformly
around 360°. At the same time as the rotating mirror sweeps the FOV across the surface, the
satellite motion moves it along the satellite trajectory, so that an image is constructed from
the successive parallel scans. Because the mirror rotates as the satellite advances, the scan
lines lie at an oblique angle to the satellite trajectory. The figure also shows a calibration
source that is held at a constant radiance. The source is located such that, after completion
of a surface scan, each channel views and stores a calibration value. A great advantage of
the cross-track scanners is that the sensors are calibrated once per rotation.
A property of the whiskbroom scanners is that, as the off-nadir angle increases, the FOV
increases and its shape changes from a circle to an ellipse. The growth in FOV can be large.
For a Sun-synchronous satellite at an altitude of 800 km, the FOV area at θV = 45° exceeds
its nadir value by a factor of 1.5 in the along-track direction and by a factor of 3.5 in the
along-scan direction; at 55°, the area exceeds its nadir value by factors of respectively 2
and 6. For these scanners, the mirror rotation rate is set so that on successive scans the
nadir FOVs are adjacent to one another. Consequently, as the off-nadir FOVs increase in
area they overlap. Because of this growth in the FOV with angle, the overall shape of a
scan resembles a bowtie, so that this growth in FOV with increasing off-nadir scan angle is
called the bowtie effect.
The received data are also averaged over short periods of time into a series of successive
time blocks. This further increases the FOV, where the time-averaged FOV is called the
effective field-of-view (EFOV). As Section 1.7 describes in more detail, on the ground
the data are resampled to a uniform grid, where each cell in the grid has the area of the
nadir FOV. Given the increase in both atmospheric interference and EFOV with increasing
zenith angle, data taken at θV greater than 45–55° are noisier than data taken near nadir.
Finally, some sensors such as the Optical Line Scanner (OLS) on the DMSP satellite and
the Day–Night Band (DNB) on VIIRS use a variety of techniques such as a variable-focus
telescope to adjust the instrument solid angle so that the FOV area is independent of look
angle.
1.6.3 Along-track or pushbroom scanners
In contrast to the whiskbroom scanner, the pushbroom scanner uses long linear arrays
of sensors to observe the surface in the cross-track direction, where each sensor, or, for
multiple bands, each set of sensors, is focused on a specific track line beneath the satellite
(Figure 1.10). For this instrument, the nadir FOV is a circle; the off-nadir FOVs are ellipses.
The advantage of this technique is that the dwell time, or time interval for which the sensor
is focused on a specific surface area, is greater than for the whiskbroom. Because it allows
one to obtain a greater signal-to-noise ratio and a higher spatial resolution than is possible
for whiskbroom sensors, this increased dwell time is one of the most useful properties
of the pushbroom instruments. Examples include the 30-m resolution Enhanced Thematic

Detectors
IFOVs
λ
λ λ
λ
(b)(a)
λ
Lens system
Lens system
1
λ1
2
3
4
5
Fig. 1.10. A schematic representation of the along-track or pushbroom scanner. (a) Single-wavelength
scanner. (b) Multi-wavelength scanner. The ellipses show the FOVs; the gray ellipses are simultane-
ously viewed by the strip of detectors. Part (b) shows how the dark gray ellipse is viewed at multiple
bands by the strip of dark gray detectors. See the text for further description.
Mapper Plus (ETM+) on the LANDSAT-7 satellite, the German Modular Optical Scan-
ner (MOS) on the Indian IRS-P3 and the ESA Medium Resolution Imaging Spectrometer
(MERIS) on ENVISAT with its 1200-km swath width. The advantages of the pushbroom
scanner are longer dwell time and better spatial resolution; the disadvantages are that the
individual sensors can lose their calibrations relative to one another, making the instru-
ment less accurate. Also, given that the pushbroom scanner requires one sensor for each
surface pixel, the pushbroom instruments generally have a narrower swath width than the
whiskbrooms, because otherwise the large number of required sensors would generate an
unwieldy instrument.
1.6.4 Hybrid cross-track scanner
Third, the need for wide-swath, high-spatial-resolution scanners led to the development
of hybrid cross-track scanners that combine the properties of the whisk and pushbroom
scanners. The hybrid scanner uses linear arrays of sensors with their long axis oriented in
the along-track direction. These arrays receive radiation from within a large-aspect-ratio
elliptical FOV with its along-track length much longer than its cross-scan length. The
advantage of this scanner is that it provides a way to increase dwell time and obtain high
resolution from a wide-swath instrument while still permitting calibration of the sensors at
each rotation.

20 Background
Examples include MODIS on TERRA and AQUA with its 2300-km swath width, and
VIIRS on Suomi-NPP with its 3000-km swath width. At nadir, the overall MODIS FOV
dimensions are 10 km in the along-track direction and 1 km in the cross-scan direction
(Barnes et al., 1998; Wolfe et al., 2002). In the along-track direction, and depending on
the observational wavelength, the number of detectors is 10, 20 or 40, corresponding to the
nadir resolution of 1.0, 0.5 and 0.25 km. As listed in Table A.2 in the Appendix, MODIS has
36 spectral bands, where, at nadir, 29 of the bands have a 1-km resolution, five have a 0.5-km
resolution, and two have a 0.25-km resolution. The advantage of this scanning technique
is that, if this multiple-detector system were replaced by a single-sensor whiskbroom, the
mirror would have to spin ten times as fast to obtain the same spatial resolution, reducing
the dwell time and increasing the noise, both by a factor of ten. A problem that occurs with
the MODIS sensor is the bowtie effect, where, at the swath edge, the 1-km nadir resolution
increases to 2 km in the along-track direction and 5.6 km in the cross-track direction (Wolfe
et al., 2002).
VIIRS on Suomi-NPP is the replacement for AVHRR and MODIS, and has a similar
set of along-track sensors to MODIS. As Table A.3 in the Appendix shows, although
VIIRS has a better spatial resolution than MODIS, it has only 22 bands compared with the
36 MODIS bands (Welsch et al., 2001). Of these bands, one is the Day–Night Band (DNB)
discussed in Section 1.6.2; the others are discussed below. Compared with MODIS, the
smaller number of VIIRS bands reduces the VIIRS complexity, cost and weight relative to
MODIS (VIIRS, 2012a). VIIRS gathers data using a rotating telescope and linear arrays
of along-track sensors. VIIRS has a cross-track view angle of ±56° and a 3000-km swath
width, which is 30% greater than the MODIS swath width.
At nadir and similar to MODIS, the VIIRS FOV extends about 12 km in the along-track
direction and 750 m in the along-scan direction. Within the instrument, the FOV radiances
are focused onto two linear detector arrays, one for the sixteen 750-m resolution bands,
called “Moderate” or “M” bands, and one for the five 375-m resolution bands, called
“Imaging” or “I” bands, where these resolutions are at nadir. The Moderate bands have
sixteen detectors in the along-track direction; the Imaging bands have 32 (VIIRS, 2012b).
A unique feature of VIIRS is that, in the along-scan direction, each detector is made up of
three sub-detectors. VIIRS uses these sub-detectors to partially correct for the bowtie effect
by constraining the increase in the field-of-view with scan angle. As the following shows,
VIIRS compensates for this increase by having the number of along-scan sensors decrease
as the view angle increases.
Figure 1.11 shows the configuration of the VIIRS along-scan sensors, and, for specific
values of the scan or view angle θV, the approximate IFOV dimensions for the M-bands.
For 0° < θV < 32°, three sensors determine the IFOV, where, as Figure 1.11(a) shows for
the nadir case, the FOV generated by the sensors is nearly square and measures 0.75 km ×
0.75 km. For 32° < θV < 45°, the number of sensors that determine the IFOV decreases from
three to two, yielding at 32° an IFOV of 1.1 km × 1.3 km, so that it remains approximately
square. For angles greater than 45°, the number of sensors decreases from two to one,
yielding at 45° an IFOV measuring 1.6 km × 1.6 km.

762m
786 m
262 m
Along-scan direction
1100m 1260 m
630 m
1600m
1600 m
Along-trackdirection
32o < < 45o< 32oθV θV θV45o < < 56o
Fig. 1.11. The along-scan configuration of the number of detectors used to determine the FOV as a
function of view angle for the VIIRS Moderate resolution bands. The gray rectangles represent the
sensors used in the retrieval of the surface radiance, while the ranges of angles above the rectangles
show the range of applicability of the sensor configuration in terms of the view angle; the adjacent
dimensions give the size of the surface FOV for (a) nadir view, (b) θV = 32◦
and (c) θV = 45◦
. See
the text for further description. (Adapted from Guenther et al. (2011)).
For comparison of the MODIS and VIIRS IFOVs, Figure 1.12 shows the dependence of
their along-scan dimension on scan angle, and, for VIIRS, shows how the reduction in the
number of sensors reduces the along-scan IFOV dimension. Because of the reduction in
the number of sensors with view angle, the along-scan dimensions of the IFOV increase by
a factor of two, instead of by the factor of six that occurs for MODIS. Finally, for different
locations on the swath, Figure 1.13 compares the IFOV of the AVHRR, MODIS and VIIRS
bands.
1.6.5 Resolution
As the next section describes in detail, the data from these instruments are resampled into a
uniform grid, where the grid spacing approximately corresponds to the nadir FOV diameter.
Each element in the grid is called a pixel, which is the abbreviation for picture element.
Typically, for AVHRR and SeaWiFS, the pixel measures 1 km by 1 km, referred to as a
1-km pixel, where the pixel area equals that of the nadir FOV. For this case, the instrument is
also described as having a 1-km resolution, meaning that objects smaller than 1 km cannot
be distinguished by the imager. In the visible, infrared and passive microwave, resolution is
defined as equal to the nadir FOV. For radars and as Section 13.2.2 describes, the definition
of resolution is different, in that the smallest pixel size equals half the resolution.

An introduction to ocean remote sensing (2nd ed.) [s. martin, 2014]

An introduction to ocean remote sensing (2nd ed.) [s. martin, 2014]

Empfohlen

Empfohlen

Weitere ähnliche Inhalte

Was ist angesagt?

Was ist angesagt? (20)

Andere mochten auch

Andere mochten auch (10)

Ähnlich wie An introduction to ocean remote sensing (2nd ed.) [s. martin, 2014]

Ähnlich wie An introduction to ocean remote sensing (2nd ed.) [s. martin, 2014] (20)

Mehr von Marchel monoarfa

Mehr von Marchel monoarfa (11)

Kürzlich hochgeladen

Kürzlich hochgeladen (20)

An introduction to ocean remote sensing (2nd ed.) [s. martin, 2014]