and the reader is referred to (Phinn et al., 2000a; Phinn et al., 2001b; Phinn, 1998b;
Phinn et al., 2000b).
Part 2. Evaluation of Remotely Sensed Data and Processing Approaches for Indicator
Monitoring Each indicator (e.g. seagrass extent) is directly compared to relevant
remotely sensed data sets and processing approaches listed in Part 1 to determine the
suitability of remotely sensed solutions for monitoring an indicator (i.e. Operational,
Feasible, Likely/Possible or Unlikely/Impossible).
To arrive at a direct link between the specified indicator(s) and suitable remote
sensing data and processing approaches, a three-stage procedure is implemented. At the
completion of this procedure, a clear link is established between each indicator and the
remotely sensed data set that could be used for its measurement (e.g. Table 1). This
linkage includes specifications of the most appropriate remotely sensed data, image
processing techniques, required personnel, hardware, and software necessary to
complete the task. An estimated cost of mapping, verification, and monitoring for the
indicator can be provided for each potentially suitable data type. A final assessment can
then be made for each data type and processing operation in terms of its “feasibility”
for operational monitoring of select indicators.
The first stage of this process involves determining a direct link between
environmental variables that could be mapped, measured, and monitored from remotely
sensed data, and relevant environmental indicators (Tables 2 and 3). If an indicator can
not be matched with a remotely sensed variable or surrogate it is removed from the
evaluation process and considered to be in the “Impossible” category. An extensive
review of past and current remote sensing applications in coastal environments should
be used as a basis for this evaluation (e.g Edwards, 1999a; Dekker et al., 2001b;
Dadouh-Guebas, 2002; Malthus, 2003). This information is then condensed into Table
3, where the level of match between indicators and remotely sensed variables is
identified. For example, processing of airborne or satellite image data sets to produce
benthic cover maps provides the information required to assess several indicators.
The next stage is to link “appropriate” remotely sensed data sets to each remotely
sensed variable. This is achieved in Table 4 by taking all of the commercially available
remotely sensed data types and identifying the remotely sensed variable(s) they have
been used to derive. Next, the most “appropriate” remotely sensed data set(s) for
deriving remotely sensed variables linked to an associated indicator are identified
(Table 5).
The final stage specifies the resources required to map and monitor indicators from
the most appropriate form of remotely sensed data and image-derived variables. A
direct assessment of the feasibility and costs of selected indicators derived through
remote sensing is provided. Table 5 contains an example of the results of the
assessment. The format of each table first specifies the relevant remotely sensed
variable and its spatial and temporal dimensions. The most appropriate data sets
selected for each remotely sensed variable are then added, along with their dimensions
and a listing of:
• Processing technique(s) required to convert remotely sensed data to the
relevant environmental variable and indicator
• Specifications (and costs estimates) for the necessary data, hardware and
software systems required to complete the processing of remotely sensed data
to map or monitor the relevant environmental variable and indicator
226
Phinn, Joyce, Scarth and Roelfsema
Phinn et al., 2000b).
Part 2. Evaluation of Remotely Sensed Data and Processing Approaches for Indicator
Monitoring Each indicator (e.g. seagrass extent) is directly compared to relevant
remotely sensed data sets and processing approaches listed in Part 1 to determine the
suitability of remotely sensed solutions for monitoring an indicator (i.e. Operational,
Feasible, Likely/Possible or Unlikely/Impossible).
To arrive at a direct link between the specified indicator(s) and suitable remote
sensing data and processing approaches, a three-stage procedure is implemented. At the
completion of this procedure, a clear link is established between each indicator and the
remotely sensed data set that could be used for its measurement (e.g. Table 1). This
linkage includes specifications of the most appropriate remotely sensed data, image
processing techniques, required personnel, hardware, and software necessary to
complete the task. An estimated cost of mapping, verification, and monitoring for the
indicator can be provided for each potentially suitable data type. A final assessment can
then be made for each data type and processing operation in terms of its “feasibility”
for operational monitoring of select indicators.
The first stage of this process involves determining a direct link between
environmental variables that could be mapped, measured, and monitored from remotely
sensed data, and relevant environmental indicators (Tables 2 and 3). If an indicator can
not be matched with a remotely sensed variable or surrogate it is removed from the
evaluation process and considered to be in the “Impossible” category. An extensive
review of past and current remote sensing applications in coastal environments should
be used as a basis for this evaluation (e.g Edwards, 1999a; Dekker et al., 2001b;
Dadouh-Guebas, 2002; Malthus, 2003). This information is then condensed into Table
3, where the level of match between indicators and remotely sensed variables is
identified. For example, processing of airborne or satellite image data sets to produce
benthic cover maps provides the information required to assess several indicators.
The next stage is to link “appropriate” remotely sensed data sets to each remotely
sensed variable. This is achieved in Table 4 by taking all of the commercially available
remotely sensed data types and identifying the remotely sensed variable(s) they have
been used to derive. Next, the most “appropriate” remotely sensed data set(s) for
deriving remotely sensed variables linked to an associated indicator are identified
(Table 5).
The final stage specifies the resources required to map and monitor indicators from
the most appropriate form of remotely sensed data and image-derived variables. A
direct assessment of the feasibility and costs of selected indicators derived through
remote sensing is provided. Table 5 contains an example of the results of the
assessment. The format of each table first specifies the relevant remotely sensed
variable and its spatial and temporal dimensions. The most appropriate data sets
selected for each remotely sensed variable are then added, along with their dimensions
and a listing of:
• Processing technique(s) required to convert remotely sensed data to the
relevant environmental variable and indicator
• Specifications (and costs estimates) for the necessary data, hardware and
software systems required to complete the processing of remotely sensed data
to map or monitor the relevant environmental variable and indicator
226
Phinn, Joyce, Scarth and Roelfsema