Abstract:
Habitat maps are vital to resource management efforts as well as for many ecological studies. Automated classification of satellite imagery can be a valuable method of generating habitat maps for shallow subtidal habitats, including reefs, in temperate regions. However, current methods require expensive proprietary software and/or specialised knowledge and skills to implement and are difficult to integrate with field based reference data for the production of quantitative accuracy assessment. A set of free open source software tools will be described that will make these methods more accessible and efficient and integrate lowcost field data for classifier training and accuracy assessment. Two of these tools, Benthic Photo Survey (BPS) and the Accuracy Assessment QGIS plugin, are currently available for use. BPS has been used with an inexpensive camera, GPS, and depth logger to map the extent of an Ecklonia radiata die-back. In conjunction with the Accuracy Assessment plugin, BPS has been used to gather reference data from the field for classifier training and accuracy assessment in the production of preliminary habitat maps from 8-band WorldView- 2 satellite imagery. Increased accessibility and availability of cost-effective tools for mapping subtidal habitats, such as those presented here, will be of value to marine spatial planning initiatives and ecological studies in coastal regions worldwide.