348
A. Dekker, V. Brando, J. Anstee, S. Fyfe, T. Malthus and E. Karpouzli
australis cover remained stable whilst Ruppia sp. and
Halophila sp. cover varied slightly. The most significant change had occurred in the cover of Zostera
capricorni, which had been replaced by cover of
macro-algae species like Chara and Nitella or the
Zostera had been overgrown by dense epiphytes.
B. Recent Advances in Remote Sensing
Satellite remote sensing technology changed dramatically at the end of the 1990s: very high spatial
resolution sensors such as IKONOS and QuickBird became available, offering pixel sizes of 0.6–
1.0 m in panchromatic and 2.4–4 m in the multispectral bands. In addition, a new generation of satellite
systems that build on the Landsat and SPOT series
were launched, providing intermediate ground resolution of between 5 and 25 m, an increased number
of spectral bands and higher radiometric sensitivity. These systems include Landsat 7 TM, SPOT 5,
and ASTER. Sensors with increased spatial resolution will better suit the discrimination of small and
patchy, or narrow, linear seagrass beds that commonly occur in small estuaries but they may not
improve the accuracy of mapping large seagrass
meadows (e.g. Mumby and Edwards, 2002; Malthus
and Karpouzli, 2003). However, because of the wide
range of satellite sensors now available, imagery can
be selected to match the scale and objective of almost
any seagrass mapping project.
Multispectral sensors mounted on aircraft became
available in the early 1980s, followed by hyperspectral sensors in the mid-1980s. At first these
were mainly research-type instruments, but from the
early 1990s onwards, commercial companies provided instruments such as the CASI and the HyMap.
Many other hyperspectral instruments were also developed, but since they were usually custom-built
they were never capable of generating multiple use.
Abbreviations: ASTER – advanced spaceborne thermal emission and reflection radiometer; CASI – compact airborne spectrographic imager; CDOM – colored dissolved organic matter;
HyMap – series of airborne hyperspectral sensors; IKONOS – a
high spatial resolution multispectral satellite sensor; Landsat –
a series of land imaging satellites from 1984 to present; Landsat 7 TM – the most recent Landsat sensor: Landsat Thematic
Mapper 7; QuickBird – a high spatial resolution multispectral
satellite sensor; SPOT – systeme probatoir d’observation de la
Terre: Satellite sensor system from 1984 to present; SPOT 5 –
sensor nr 5 of SPOT; RT – radiative transfer of energy theory.
Note: for symbols on light in the sea, see Chapters 12 and 13.
However, the capabilities of these new airborne instruments and the potential they offered were the
main driving force behind research and development of quantitative hyperspectral measurement and
modeling-based, methods for mapping seagrass and
associated ecosystems.
Increasingly, results detailing the high spectral resolution reflectance properties of submerged
aquatic vegetation and associated substrata in situ
are being published. Reported spectra for a range of
species and growth habits, mainly for assessing the
potential for their spectral discrimination in remote
sensing images are reported (Malthus and George,
1997; Alberotanza et al., 1999; Myers et al., 1999;
Hochberg and Atkinson, 2000; Lubin et al., 2001;
Fyfe, 2003; Kutser et al., 2003; Anstee et al., 2004;
Karpouzli et al., 2004).
C. Imaging Seagrass Beds
A number of researchers have investigated the application of airborne hyperspectral or multispectral digital sensors for discriminating and mapping
benthic plant species (e.g. Zacharias et al., 1992;
Bajjouk et al., 1996; Clark et al., 1997; Malthus
and George, 1997; Mumby et al., 1997a; Thomson
et al., 1998; Alberotanza et al., 1999; Pasqualini
et al., 2001) and for estimating seagrass biomass
(Mumby et al., 1997b). Many recent studies into light
in shallow waters and hyperspectral (modeled, in
situ measured and from airborne systems) mapping
of seagrasses, macro-algae, benthic micro-algae and
coral reef species are documented in Limnology and
Oceanography (2003) Volume 48.
The principle of the use of airborne imaging spectrometry for measuring the subsurface reflectance
R(0−) (see Eqs. (11–13)) over a water body with
seagrass is presented in Fig. 1. A CASI image over a
shallow coastal water body near Adelaide (Australia)
was corrected for atmospheric and water column effects using radiative transfer (RT) models. Simultaneous in situ measurements of benthic reflectance
of Posidonia australis were collected using a field
spectrometer at the time of aircraft overpass. Fig. 1
compares the results of Hydrolight-based modeling
of at surface reflectance spectra of Posidonia P. australis under different depths of a water column with
atmospherically corrected CASI image spectra from
the shallow water site where P. australis was measured. The relevance of this work is that a spectrum measured by a remote sensor can increasingly
A. Dekker, V. Brando, J. Anstee, S. Fyfe, T. Malthus and E. Karpouzli
australis cover remained stable whilst Ruppia sp. and
Halophila sp. cover varied slightly. The most significant change had occurred in the cover of Zostera
capricorni, which had been replaced by cover of
macro-algae species like Chara and Nitella or the
Zostera had been overgrown by dense epiphytes.
B. Recent Advances in Remote Sensing
Satellite remote sensing technology changed dramatically at the end of the 1990s: very high spatial
resolution sensors such as IKONOS and QuickBird became available, offering pixel sizes of 0.6–
1.0 m in panchromatic and 2.4–4 m in the multispectral bands. In addition, a new generation of satellite
systems that build on the Landsat and SPOT series
were launched, providing intermediate ground resolution of between 5 and 25 m, an increased number
of spectral bands and higher radiometric sensitivity. These systems include Landsat 7 TM, SPOT 5,
and ASTER. Sensors with increased spatial resolution will better suit the discrimination of small and
patchy, or narrow, linear seagrass beds that commonly occur in small estuaries but they may not
improve the accuracy of mapping large seagrass
meadows (e.g. Mumby and Edwards, 2002; Malthus
and Karpouzli, 2003). However, because of the wide
range of satellite sensors now available, imagery can
be selected to match the scale and objective of almost
any seagrass mapping project.
Multispectral sensors mounted on aircraft became
available in the early 1980s, followed by hyperspectral sensors in the mid-1980s. At first these
were mainly research-type instruments, but from the
early 1990s onwards, commercial companies provided instruments such as the CASI and the HyMap.
Many other hyperspectral instruments were also developed, but since they were usually custom-built
they were never capable of generating multiple use.
Abbreviations: ASTER – advanced spaceborne thermal emission and reflection radiometer; CASI – compact airborne spectrographic imager; CDOM – colored dissolved organic matter;
HyMap – series of airborne hyperspectral sensors; IKONOS – a
high spatial resolution multispectral satellite sensor; Landsat –
a series of land imaging satellites from 1984 to present; Landsat 7 TM – the most recent Landsat sensor: Landsat Thematic
Mapper 7; QuickBird – a high spatial resolution multispectral
satellite sensor; SPOT – systeme probatoir d’observation de la
Terre: Satellite sensor system from 1984 to present; SPOT 5 –
sensor nr 5 of SPOT; RT – radiative transfer of energy theory.
Note: for symbols on light in the sea, see Chapters 12 and 13.
However, the capabilities of these new airborne instruments and the potential they offered were the
main driving force behind research and development of quantitative hyperspectral measurement and
modeling-based, methods for mapping seagrass and
associated ecosystems.
Increasingly, results detailing the high spectral resolution reflectance properties of submerged
aquatic vegetation and associated substrata in situ
are being published. Reported spectra for a range of
species and growth habits, mainly for assessing the
potential for their spectral discrimination in remote
sensing images are reported (Malthus and George,
1997; Alberotanza et al., 1999; Myers et al., 1999;
Hochberg and Atkinson, 2000; Lubin et al., 2001;
Fyfe, 2003; Kutser et al., 2003; Anstee et al., 2004;
Karpouzli et al., 2004).
C. Imaging Seagrass Beds
A number of researchers have investigated the application of airborne hyperspectral or multispectral digital sensors for discriminating and mapping
benthic plant species (e.g. Zacharias et al., 1992;
Bajjouk et al., 1996; Clark et al., 1997; Malthus
and George, 1997; Mumby et al., 1997a; Thomson
et al., 1998; Alberotanza et al., 1999; Pasqualini
et al., 2001) and for estimating seagrass biomass
(Mumby et al., 1997b). Many recent studies into light
in shallow waters and hyperspectral (modeled, in
situ measured and from airborne systems) mapping
of seagrasses, macro-algae, benthic micro-algae and
coral reef species are documented in Limnology and
Oceanography (2003) Volume 48.
The principle of the use of airborne imaging spectrometry for measuring the subsurface reflectance
R(0−) (see Eqs. (11–13)) over a water body with
seagrass is presented in Fig. 1. A CASI image over a
shallow coastal water body near Adelaide (Australia)
was corrected for atmospheric and water column effects using radiative transfer (RT) models. Simultaneous in situ measurements of benthic reflectance
of Posidonia australis were collected using a field
spectrometer at the time of aircraft overpass. Fig. 1
compares the results of Hydrolight-based modeling
of at surface reflectance spectra of Posidonia P. australis under different depths of a water column with
atmospherically corrected CASI image spectra from
the shallow water site where P. australis was measured. The relevance of this work is that a spectrum measured by a remote sensor can increasingly