impossible, thus satellite imagery remains the only option. However, some satellite
systems do not systematically acquire data over oceanic regions, thus specific tasking is
required (e.g. Quickbird).
Effective field validation methods for image classifications remain a challenge in
reef environments due to scales of heterogeneity from individual coral patches to entire
reef ecosystems. Neither field campaigns nor image data can capture all scales, and
integrating the two is difficult. Although global standards for reef substrate monitoring
have been developed (e.g. Reef Check), using this classification scheme or field method
for calibrating and validating image data presents problems with scaling and accuracy
(Joyce et al., 2004) . The Reef Check methods are simple and can be implemented as a
rapid assessment scheme by volunteers with little training (Mumby et al., 1995), thus
are ideal for assisting with image data classifications. However the scaling challenges
related to the differences between image and field data resolutions need to be
overcome.
According to one survey, (Joyce et al., 2002), the main limitations to the use of
remotely sensed data in coral reef environments were perceived to be cost of image
acquisition and inadequate spectral resolution. Increased utilization of these data
requires better integration with GIS and a greater capacity (human and computer) to
effectively process and extract the information. Improved remote sensing technologies
(e.g. data set development, higher spatial and spectral resolution) will be welcomed by
the majority of coral reef monitoring and management agencies, however cost was
noted as a potential constraint, with only 14% believing they would have both technical
and financial capabilities to fully utilize new remotely sensed data sets. The majority
believed they would have neither the technical nor financial capacity, while other
organizations believed that finance would prove to be the only constraint. Although the
majority of respondents identified the cost of remotely sensed data sets as a major
limitation, most were unsure of the cost of their data, due to government provisions,
special research allowances, or infrequency of purchase. This would suggest a limited
knowledge of the full costs of acquiring and processing images for their purpose, or a
limited knowledge of actual prices of image data sets. In either case this is an indication
of the lack of detailed knowledge of the cost, time, and possible accuracy involved in
applying remote sensing for coral reef monitoring activities. An example of this is
aerial photography, where the cost of acquisition is not indicative of the total cost to
integrate fully into a GIS database. As a large portion of the world’s coral reefs occur in
the waters of developing countries, financial constraints are a significant factor in the
methods employed for coral reef management.
Another commonly identified limitation of remotely sensed data is the high degree
of user expertise required to understand and extract the required information. The
survey also indicated the strong need for further research into, and development of,
techniques to best use remote sensing as a monitoring tool.
Based on the results of this survey, it appears that the combination of: 1) the
greater range of image data sets now available, and 2) an inability to consistently
identify indicators of reef condition that can be used to produce reliable change
detection approaches, have placed coral reef remote sensing in a developmental
stage. Recommendations for furthering the utility of remote sensing in coral reef
environments should focus on: 1) the identification and development of algorithms (and
related spectral resolution) to relate reef bio-optical properties with relevant biophysical
controls; 2) further development of techniques to remove the attenuating effects of the
overlying water column; 3) greater incorporation of biogeochemical cycles (e.g.
climatic and oceanographic data) with remote sensing data to understand the processes
239
Integrated Information Acquisition and Management
systems do not systematically acquire data over oceanic regions, thus specific tasking is
required (e.g. Quickbird).
Effective field validation methods for image classifications remain a challenge in
reef environments due to scales of heterogeneity from individual coral patches to entire
reef ecosystems. Neither field campaigns nor image data can capture all scales, and
integrating the two is difficult. Although global standards for reef substrate monitoring
have been developed (e.g. Reef Check), using this classification scheme or field method
for calibrating and validating image data presents problems with scaling and accuracy
(Joyce et al., 2004) . The Reef Check methods are simple and can be implemented as a
rapid assessment scheme by volunteers with little training (Mumby et al., 1995), thus
are ideal for assisting with image data classifications. However the scaling challenges
related to the differences between image and field data resolutions need to be
overcome.
According to one survey, (Joyce et al., 2002), the main limitations to the use of
remotely sensed data in coral reef environments were perceived to be cost of image
acquisition and inadequate spectral resolution. Increased utilization of these data
requires better integration with GIS and a greater capacity (human and computer) to
effectively process and extract the information. Improved remote sensing technologies
(e.g. data set development, higher spatial and spectral resolution) will be welcomed by
the majority of coral reef monitoring and management agencies, however cost was
noted as a potential constraint, with only 14% believing they would have both technical
and financial capabilities to fully utilize new remotely sensed data sets. The majority
believed they would have neither the technical nor financial capacity, while other
organizations believed that finance would prove to be the only constraint. Although the
majority of respondents identified the cost of remotely sensed data sets as a major
limitation, most were unsure of the cost of their data, due to government provisions,
special research allowances, or infrequency of purchase. This would suggest a limited
knowledge of the full costs of acquiring and processing images for their purpose, or a
limited knowledge of actual prices of image data sets. In either case this is an indication
of the lack of detailed knowledge of the cost, time, and possible accuracy involved in
applying remote sensing for coral reef monitoring activities. An example of this is
aerial photography, where the cost of acquisition is not indicative of the total cost to
integrate fully into a GIS database. As a large portion of the world’s coral reefs occur in
the waters of developing countries, financial constraints are a significant factor in the
methods employed for coral reef management.
Another commonly identified limitation of remotely sensed data is the high degree
of user expertise required to understand and extract the required information. The
survey also indicated the strong need for further research into, and development of,
techniques to best use remote sensing as a monitoring tool.
Based on the results of this survey, it appears that the combination of: 1) the
greater range of image data sets now available, and 2) an inability to consistently
identify indicators of reef condition that can be used to produce reliable change
detection approaches, have placed coral reef remote sensing in a developmental
stage. Recommendations for furthering the utility of remote sensing in coral reef
environments should focus on: 1) the identification and development of algorithms (and
related spectral resolution) to relate reef bio-optical properties with relevant biophysical
controls; 2) further development of techniques to remove the attenuating effects of the
overlying water column; 3) greater incorporation of biogeochemical cycles (e.g.
climatic and oceanographic data) with remote sensing data to understand the processes
239
Integrated Information Acquisition and Management