Classification and Monitoring of Salt Marsh Habitats with Multi-polarimetric Airborne and Satellite SAR
van Beijma, Sybrand1; Comber, Alexis2; Lamb, Alistair1; Brown, Sarah1
1Astrium GEO-Information Services, UNITED KINGDOM; 2University of Leicester, UNITED KINGDOM

Within the GMES programme there is much interest in the ability of remote sensing technology to deliver operational solutions to many areas of life including environmental management. Integrated Coastal Zone Management (ICZM) seeks to find a balance between human use and sustainable functioning of coastal zone ecosystems. Coupled with general interest in ecosystem services, this research explores methods for characterising coastal salt marsh zone habitats in terms of the environmental benefits and affordances they confer with the use of under-exploited Synthetic Aperture Radar (SAR) remote sensing techniques, thereby providing additional information in support of the management of coastal zones. The research consists of examining the use of full polarimetric airborne S- and X-Band for habitat mapping of salt marsh areas, following common habitat classification schemes as National Vegetation Classification (NVC) or Habitat Directive Annex 1.
This will be done on the basis of a data set which is acquired over the Llanrhidian salt marsh in Wales in July 2010 with the Astrium Airborne Demonstrator, as a precursor test system to the upcoming NovaSAR-S mission. This SAR data set provides an excellent opportunity to research the potential of high resolution fully polarimetric SAR for classification of salt marsh habitats, by characterising botanical and structural composition of specific salt marsh vegetation zones. SAR-derived information, such as backscatter coefficient, band ratios and polarimetric decomposition parameters are extracted. These are used as data layers to perform object-based image analysis to identify the main vegetation habitats, and pixel-based analysis of structural information about vegetation cover and density. Classification results will be validated with newly acquired and existing ecological field data. This presentation will focus on classification results from the airborne SAR data set, the use of polarimetric information and the added value for vegetation mapping.

This research has shown that SAR data can be used to identify salt marsh habitats and to support ICZM. Ongoing research is improving the applicability of satellite and airborne SAR sensors to a range of coastal zone problems and is exploring how the combination of a multi-sensor and multi-temporal SAR approach gives more insights into long-term dynamics of intertidal land cover and ecosystem functions. These are particularly important in intertidal habitat areas and for the ecosystem services they provide. Future research will examine:
1. Broad scale habitat classification with the use of multi-temporal, dual- and quad-polarimetric satellite ALOS PALSAR L-Band SAR data.
2. Correlation analysis between tidal heights (the research area is subject to an average tidal range of 8 metres) and SAR backscatter signatures, based on multi-temporal ALOS PALSAR data.