Chapter 15
Remote Sensing of Seagrass Ecosystems:
Use of Spaceborne and Airborne Sensors
Arnold Dekker
∗ , Vittorio Brando, and Janet Anstee
CSIRO Land and Water, GPO Box 1666,Canberra, ACT 2601, Australia
Suzanne Fyfe
University of Wollongong, Wollongong, Australia
Timothy Malthus and Evanthia Karpouzli
University of Edinburgh, Edinburgh, UK
I. Introduction
A. Remote Sensing and Aerial
Photographic Approaches
The focus of this chapter lies in describing digital multispectral and hyperspectral remote sensing
developments and applications in the mapping and
monitoring of seagrass ecosystems. Multispectral
refers to a sensor that registers light in a limited
number of relatively broad spectral bands (bandwidths of 20–60 nm); hyperspectral (also referred to
as imaging spectrometry) is defined for sensors that
measure the entire spectrum under consideration in
contiguous narrow spectral bands (bandwidths between 2 and 20 nm).
Currently, seagrass maps are still predominantly
being produced from the interpretation of aerial photography although it is likely that airborne and spaceborne remote sensing methods will rapidly take over
this role given the advantages they present in terms of
accuracy, repeatability, versatility, and information
content. Nevertheless, retrospective studies of seagrass change using the more modern methodologies
will still need to make use of results generated by the
more traditional methods since aerial photographs
are the dominant archival source of historical spatial
∗ Author for correspondence, email: arnold.dekker@csiro.au
information on seagrass meadows. The subject of
aerial photography and the assessment of seagrass
vegetation has been extensively dealt with in Global
Seagrass Research Methods (Short and Coles, 2001)
and the reader is referred to reviews in that publication, as well as to relevant sections of Green and
Short (2003).
Traditional satellite sensors offer a cost-effective
approach for mapping seagrass over large areas
and in remote locations (Ferguson and Korfmacher,
1997; Mumby et al., 1999) particularly where the
meadows are large, monospecific, and continuous.
The value of the repeated temporal cover provided
by satellite sensors has been shown by Jensen and coworkers who investigated the application of Landsat
MSS and SPOT multitemporal data to successfully
evaluate change in cattail and seagrass species in the
Everglades (Jensen et al., 1995). Multi-date satellite remote sensing is geometrically highly repeatable and a cost-effective method for detecting large
changes in seagrass distribution or extent over time
(Robblee et al., 1991; Zainal et al., 1993; Ward et al.,
1996; Macleod and Congalton, 1998). Anstee et al.
(2004) detected seagrass and macro-algae change
in a shallow coastal tidal lake in Australia using
archival Landsat satellite image data from 1988 to
2002, by applying hyperspectral measurement and
modeling techniques to the multispectral Landsat
images. They were able to determine that Posidonia
347–359.
A. W. D. Larkum et al. (eds.), Seagrasses: Biology, Ecology and Conservation, pp.
c
2006 Springer. Printed in the Netherlands.
Remote Sensing of Seagrass Ecosystems:
Use of Spaceborne and Airborne Sensors
Arnold Dekker
∗ , Vittorio Brando, and Janet Anstee
CSIRO Land and Water, GPO Box 1666,Canberra, ACT 2601, Australia
Suzanne Fyfe
University of Wollongong, Wollongong, Australia
Timothy Malthus and Evanthia Karpouzli
University of Edinburgh, Edinburgh, UK
I. Introduction
A. Remote Sensing and Aerial
Photographic Approaches
The focus of this chapter lies in describing digital multispectral and hyperspectral remote sensing
developments and applications in the mapping and
monitoring of seagrass ecosystems. Multispectral
refers to a sensor that registers light in a limited
number of relatively broad spectral bands (bandwidths of 20–60 nm); hyperspectral (also referred to
as imaging spectrometry) is defined for sensors that
measure the entire spectrum under consideration in
contiguous narrow spectral bands (bandwidths between 2 and 20 nm).
Currently, seagrass maps are still predominantly
being produced from the interpretation of aerial photography although it is likely that airborne and spaceborne remote sensing methods will rapidly take over
this role given the advantages they present in terms of
accuracy, repeatability, versatility, and information
content. Nevertheless, retrospective studies of seagrass change using the more modern methodologies
will still need to make use of results generated by the
more traditional methods since aerial photographs
are the dominant archival source of historical spatial
∗ Author for correspondence, email: arnold.dekker@csiro.au
information on seagrass meadows. The subject of
aerial photography and the assessment of seagrass
vegetation has been extensively dealt with in Global
Seagrass Research Methods (Short and Coles, 2001)
and the reader is referred to reviews in that publication, as well as to relevant sections of Green and
Short (2003).
Traditional satellite sensors offer a cost-effective
approach for mapping seagrass over large areas
and in remote locations (Ferguson and Korfmacher,
1997; Mumby et al., 1999) particularly where the
meadows are large, monospecific, and continuous.
The value of the repeated temporal cover provided
by satellite sensors has been shown by Jensen and coworkers who investigated the application of Landsat
MSS and SPOT multitemporal data to successfully
evaluate change in cattail and seagrass species in the
Everglades (Jensen et al., 1995). Multi-date satellite remote sensing is geometrically highly repeatable and a cost-effective method for detecting large
changes in seagrass distribution or extent over time
(Robblee et al., 1991; Zainal et al., 1993; Ward et al.,
1996; Macleod and Congalton, 1998). Anstee et al.
(2004) detected seagrass and macro-algae change
in a shallow coastal tidal lake in Australia using
archival Landsat satellite image data from 1988 to
2002, by applying hyperspectral measurement and
modeling techniques to the multispectral Landsat
images. They were able to determine that Posidonia
347–359.
A. W. D. Larkum et al. (eds.), Seagrasses: Biology, Ecology and Conservation, pp.
c
2006 Springer. Printed in the Netherlands.