Chapter 14
Validation
Chris M. Roelfsema and Stuart R. Phinn
Abstract Adequate understanding of the validity of information contained in
coral reef remote sensing products is required to support research and management
decisions. This chapter introduces accuracy measures commonly applied to two
types of coral related maps: discrete (e.g., benthic cover type) and continuous (e.g.,
percent coral cover). A critical review of 80 coral reef remote sensing mapping
publications presents the approaches and metrics commonly used to measure
accuracy. The literature review shows that few studies report accuracy information
at all, and when obtained, ‘overall accuracy’ was the most commonly used
accuracy measure. Variations in accuracy levels were not only a result of actual
differences in map accuracy, but are likely also due to: spatial complexity of
benthic features present in the study area; distribution of the calibration and validation samples relative to each other; and the level of detail measured for each
sample. As a result, accuracy measures from different studies should be compared
with caution and with due attention to how the measures were derived. This
chapter enables scientist and managers to understand, design and interpret validation procedures for image-based maps of coral reef environments.
C. M. Roelfsema (&) Á S. R. Phinn
Biophysical Remote Sensing Group, School of Geography,
Planning and Environmental Management,
The University of Queensland, Queensland 4072, Australia
e-mail: c.roelfsema@uq.edu.au
S. R. Phinn
e-mail: s.phinn@uq.edu.au
J. A. Goodman et al. (eds.), Coral Reef Remote Sensing,
DOI: 10.1007/978-90-481-9292-2_14,
Ó Springer Science+Business Media Dordrecht 2013
375
Validation
Chris M. Roelfsema and Stuart R. Phinn
Abstract Adequate understanding of the validity of information contained in
coral reef remote sensing products is required to support research and management
decisions. This chapter introduces accuracy measures commonly applied to two
types of coral related maps: discrete (e.g., benthic cover type) and continuous (e.g.,
percent coral cover). A critical review of 80 coral reef remote sensing mapping
publications presents the approaches and metrics commonly used to measure
accuracy. The literature review shows that few studies report accuracy information
at all, and when obtained, ‘overall accuracy’ was the most commonly used
accuracy measure. Variations in accuracy levels were not only a result of actual
differences in map accuracy, but are likely also due to: spatial complexity of
benthic features present in the study area; distribution of the calibration and validation samples relative to each other; and the level of detail measured for each
sample. As a result, accuracy measures from different studies should be compared
with caution and with due attention to how the measures were derived. This
chapter enables scientist and managers to understand, design and interpret validation procedures for image-based maps of coral reef environments.
C. M. Roelfsema (&) Á S. R. Phinn
Biophysical Remote Sensing Group, School of Geography,
Planning and Environmental Management,
The University of Queensland, Queensland 4072, Australia
e-mail: c.roelfsema@uq.edu.au
S. R. Phinn
e-mail: s.phinn@uq.edu.au
J. A. Goodman et al. (eds.), Coral Reef Remote Sensing,
DOI: 10.1007/978-90-481-9292-2_14,
Ó Springer Science+Business Media Dordrecht 2013
375