relevant indicators for local, national, and global scale coastal monitoring and management
programs (Rice, 2003) and successful processes for developing and implementing indicatorbased monitoring programs (Kabuta, 2003). Recent reviews have identified several
hundred types of environmental indicators with relevance to coastal environments.
Selecting an appropriate set of indicators for the environmental monitoring/
management issue and decision maker in question is considered a key task (Bromberg,
1990; McKenzie, 1992; Rice, 2003). The selection of indicators to use for monitoring
and management applications should be based on scientific validity, clear links to
management goals, incorporation into the management process, understand-ability,
time required for implementation and maintenance, and cost efficiency. The majority
of indicators presented in the literature are based on data collected through field
surveys, with limited use of remote sensing techniques, except in the case of mapping
the extent of surface cover features (e.g. vegetation communities, weeds or algal
blooms).
3.3 LINKS BETWEEN ENVIRONMENTAL INDICATORS AND REMOTE
SENSING
Environmental indicators provide one of most useful links from environmental
monitoring and management programs to remotely sensed data. The utility of the
indicators is that they define set environmental parameters to be mapped, at a specific
spatial scale, and usually over a set time period. This information provides a basis for
selecting suitable remotely sensed data sets and processing techniques to deliver map(s)
of the requested environmental indicator (Phinn, 1998a; Foody, 2003). The framework
proposed (Phinn, 1998a) and refined (Phinn et al., 200b, 2003) by Phinn provides a
basis for making the link between environmental indicators and suitable remotely
sensed data, and techniques for processing remotely sensed data. The framework is
outlined in detail in the next section and defines the process used in other studies
linking environmental indicators to remotely sensed data (Wallace and Campbell, 1998;
Green et al., 2000; Foody, 2003; Trinder, 2003).
4. Linking remotely sensed data sets to environmental indicators, the community,
and policy-makers
4.1 A FRAMEWORK FOR LINKING ENVIRONMENTAL INDICATORS TO
REMOTELY SENSED DATA
The framework outlined in Figure 1 and described in Parts 1 – 2 below
provides a guide to evaluating commercially available remotely sensed data sets for
mapping or monitoring selected environmental indicators. Originally this approach was
developed for use in coastal wetlands (Phinn, 1998a), then modified for tropical
wetlands and tropical rainforest environments (Phinn et al., 2000b, 2002a, b). The key
to this approach is linking the spatial and temporal scale(s) of data and information
required to remotely sensed data with corresponding dimensions. Explicit consideration
is also given to the full costs of processing image data to a map product in terms of
necessary hardware, software, ancillary data, and skilled personnel. A worked example
is provided in Tables 2-5 for monitoring one selected indicator of coastal ecosystem
health, the extent of seagrass beds in Moreton Bay, Queensland, Australia (Phinn et al.,
2001b). Implementation of the framework requires initial specifications of: 1) the
indicator to map/monitor, the area to cover, and the timeframe; 2) available financial
221
Integrated Information Acquisition and Management
programs (Rice, 2003) and successful processes for developing and implementing indicatorbased monitoring programs (Kabuta, 2003). Recent reviews have identified several
hundred types of environmental indicators with relevance to coastal environments.
Selecting an appropriate set of indicators for the environmental monitoring/
management issue and decision maker in question is considered a key task (Bromberg,
1990; McKenzie, 1992; Rice, 2003). The selection of indicators to use for monitoring
and management applications should be based on scientific validity, clear links to
management goals, incorporation into the management process, understand-ability,
time required for implementation and maintenance, and cost efficiency. The majority
of indicators presented in the literature are based on data collected through field
surveys, with limited use of remote sensing techniques, except in the case of mapping
the extent of surface cover features (e.g. vegetation communities, weeds or algal
blooms).
3.3 LINKS BETWEEN ENVIRONMENTAL INDICATORS AND REMOTE
SENSING
Environmental indicators provide one of most useful links from environmental
monitoring and management programs to remotely sensed data. The utility of the
indicators is that they define set environmental parameters to be mapped, at a specific
spatial scale, and usually over a set time period. This information provides a basis for
selecting suitable remotely sensed data sets and processing techniques to deliver map(s)
of the requested environmental indicator (Phinn, 1998a; Foody, 2003). The framework
proposed (Phinn, 1998a) and refined (Phinn et al., 200b, 2003) by Phinn provides a
basis for making the link between environmental indicators and suitable remotely
sensed data, and techniques for processing remotely sensed data. The framework is
outlined in detail in the next section and defines the process used in other studies
linking environmental indicators to remotely sensed data (Wallace and Campbell, 1998;
Green et al., 2000; Foody, 2003; Trinder, 2003).
4. Linking remotely sensed data sets to environmental indicators, the community,
and policy-makers
4.1 A FRAMEWORK FOR LINKING ENVIRONMENTAL INDICATORS TO
REMOTELY SENSED DATA
The framework outlined in Figure 1 and described in Parts 1 – 2 below
provides a guide to evaluating commercially available remotely sensed data sets for
mapping or monitoring selected environmental indicators. Originally this approach was
developed for use in coastal wetlands (Phinn, 1998a), then modified for tropical
wetlands and tropical rainforest environments (Phinn et al., 2000b, 2002a, b). The key
to this approach is linking the spatial and temporal scale(s) of data and information
required to remotely sensed data with corresponding dimensions. Explicit consideration
is also given to the full costs of processing image data to a map product in terms of
necessary hardware, software, ancillary data, and skilled personnel. A worked example
is provided in Tables 2-5 for monitoring one selected indicator of coastal ecosystem
health, the extent of seagrass beds in Moreton Bay, Queensland, Australia (Phinn et al.,
2001b). Implementation of the framework requires initial specifications of: 1) the
indicator to map/monitor, the area to cover, and the timeframe; 2) available financial
221
Integrated Information Acquisition and Management