Aarhus University Seal / Aarhus Universitets segl

No. 144: Remote sensing based classification of structural elements of coastal habitats

Groom, G., Juel, A. & Ejrnæs, R. 2015. Remote sensing based classification of structural elements of coastal habitats. Aarhus University, DCE – Danish Centre for Environment and Energy, 44 pp. Scientific Report from DCE – Danish Centre for Environment and Energy No. 144. http://dce2.au.dk/pub/SR144.pdf

Summary

Inclusion of additional habitat types in the NOVANA nature monitoring places new demands on the work. For several of these habitat types it is not so much the vegetation species composition and chemical parameters that are central to assessment of the conservation status and trend, but more structural factors, such as the land distribution, geomorphology, coverage of woody plants, dwarf shrubs, etc., which are essential. The project that this report describes the work of covers 14 coastal habitat types, for which remote sensing data in the form of aerial orthophotos has been used as a supplement to field-based monitoring.

The six-year EU-reporting cycle of habitat conservation status includes reporting of habitat extents and the structure and function of habitats. With remote sensing (aerial photographs and/or satellite), it will be possible to monitor the entire Danish coastal zone, while traditional field-based methods only cover Natura 2000 sites or a random sample of the coastal zone.

The project has shown that the method is useful to identify and map structural elements. A final implementation of the method in NOVANA monitoring requires further development based on the data already collected, but even now there is good reason to be optimistic for the future applications.

This scientific report describes the possibilities for automated classification of structural element types in Danish coastal habitat types using aerial orthophotos and elevation models, towards development of a remote sensing based method for monitoring and mapping of these habitats that is better suited than the currently used field based methods. The assessment focuses on the use of remote sensing based data available around 2010 and a classification of structures present in this year. The report includes assessment of classification options, classification accuracy, and the use of the classification parameters based on field reference data from 2012.