Aerial photographs and acoustic ground discrimination systems (AGDS) are established sources of data for remote sensing and habitat mapping in the marine subtidal environment Often these systems are used independently to map specific areas of the marine subtidal aerial photographs for the shallow subtidal and AGDS for the deeper areas. Both systems have inherent technical limitations, using a single system can result in areas left unsurveyed. A preferable solution would be to use both of these data sources in combination to produce detailed maps of the entire marine subtidal environment. This paper presents the results of a case study at Loch Maddy, Western Isles, Scotland, which is a sheltered sealoch, comprising a complex series of interlinked basins Loch Maddy has been submitted as a candidate Special Area of Conservation (SAC) under the European Commission (EC) Habitats Directive Much of the outer loch was over 4 m deep and suitable for acoustic mapping The inner areas were mostly between 0 m and 4 m depth and were impossible to survey using AGDS, due to restricted manoeuvrability of the vessel Detailed aerial photographs were available and provided a means to map these shallow areas
Both sources of data were analysed using the standard image processing techniques of unsupervised and supervised classification Unsupervised classification provided maps to plan a comprehensive ground truthing programme to collect biological data These data were used in a supervised classifications to produce biological resource maps from both data sources A number of technical difficulties were encountered during the processing of the photographs, due to the variations in image quality with water depth This prompted the development of an alternative processing strategy This strategy, along with the techniques used to incorporate both aerial photographs and acoustic data into a single map are discussed.
Remote sensing is an established technique for biological resource mapping in the marine environment. Two common data sources are electromagnetic spectral (EMS) sensors normally mounted on satellites or aircraft, and acoustic sensors mounted on water borne vessels Both sensors record the strength of reflected electromagnetic or acoustic signals from the seafloor which are displayed in the form of images Variations in the physical structure of the seafloor affects its reflectivity and analyses of these images provide the basis for mapping Reflection of electromagnetic spectral radiation from the seafloor is strongly influenced by depth and water clarity Whilst it is possible to apply a depth correction algorithm to an image (Reichelt, 1991), it is not possible to compensate for poor water clarity. Acoustic sensors in the form of sonar are much less restricted by depth or water clarity but are dependent on the operational limits and manoeuvrability of the survey vessel In general EMS images are limited to shallow water (< 20m in areas of good water clarity (Pagett, 1989), >5m in high turbidity waters such as the UK) and acoustic systems can only provide images of the seafloor for deeper water (>5m small boats, >15m research ships)