Craig Brown 12/8/2004 Predictive mapping of benthic habitats using acoustic remote sensing techniques Supervisors: Dr. C.J. Brown and Dr. Rory Quinn (University of Ulster) Dr. Thomas Noji (NOAA) Dr. Vladimir Kostylev (Geological Survey of Canada) Rationale In recent years the need to map the distribution of habitats and marine species on the seabed has arisen with an increase in the demand for information on the status of the natural environment and the impact of human activities. Recent developments in acoustic survey techniques, such as sidescan sonar and multibeam sonar, have revolutionised the way we are able to image, map and understand seabed environments. It is now cost effective to image large areas of the seabed using these techniques, and the information from such surveys provides base line data from which thematic maps of the seabed environment (e.g. bathymetry, surficial sediments, underlying geology and benthic habitat) can be derived when interpreted in conjunction with other benthic data sets (e.g. sediment grab samples, benthic assemblage data, fisheries data etc.). A good understanding of the spatial distribution of these various seabed attributes is crucial for effective environmental and resource management, and high-resolution seabed maps produced in this way have an almost universal application. Despite the recent advances in acoustic survey systems which now allow us to image the seabed with increasing resolution and accuracy, further research is required to establish the most effective and robust methods of interpretation of these data for the production of seabed thematic maps. Acoustic data has been used for many years for understanding the geology of the seabed (Lurton, 2002), but only recently have such data sets been successfully applied to understanding the spatial distribution of seabed biology/habitat (Brown et al 2001, 2002; 2004a and b, in press; Foster-Smith et al. 2004; Kostylev et al. 2001; Service 1998). Although the outcomes of these studies are very encouraging, the approaches have not yet reached the stage of uncritical, routine application. A major challenge in this field of work is the understanding of the relationship between acoustic data (acoustic back-scatter/reflectance), physical seabed attributes (e.g. surficial sediments and seabed topography), hydrographic regime and benthic assemblage structure. A clear understanding of how the physical factors, which can be determined from acoustic data sets, interact to control which biological assemblage predominates in any particular area would allow large areas of the sea-floor to be mapped cost-effectively using remote acoustic survey techniques with minimal ground-truthing. This project would establish the relationship between these factors from intensively studied areas where extensive ground-truthing has been carried out. In particular, the association between acoustic backscatter data and benthic assemblage structure will be investigated. Establishing the extent that backscatter data can be used to predict habitat is a crucial stage in determining whether this parameter can be used to map the spatial distribution of biological attributes over large areas of seabed. Hypothesis: The overall hypothesis to be tested is that sub-littoral benthic habitats can be predicted and mapped over relatively large geographical areas (several km 2 ) based on information derived from acoustic survey techniques (multibeam sonar, sidescan sonar) using limited ground- truthing. Objectives: The following research areas would be targeted: