Seeing further: broadband acoustics for improved measurements of demersal fishes

Fishery and scientific echosounders have difficulty in seeing fish that are close to the seabed because of the very strong echo from the seabed. This problem hinders the use of acoustic tools and techniques for detecting, measuring, and observing demersal fishes. New technology echosounders are now available that use a much wider range of frequencies (broadband) and allow for more sophisticated acoustic signals and more complex signal processing techniques. These give a much higher frequency and range resolution and will permit the development of techniques to see closer to the seabed. The increased frequency resolution will also facilitate the enhancement of existing species identification techniques, leading to better acoustic determination of fish species. This project has: 1) Carried out experiments to find better ways to acoustically detect fish close to the seabed using broadband acoustic signals. 2) Collected broadband data from fish schools and other marine organisms 3) Developed improved methods of detecting fish close to the seabed 4) Made comparisons of the performance of broadband and narrowband echosounders close to the seabed

- Increased awareness in the fisheries community of advantages and disadvantages of broadband pulses at boundaries - Bottom-specific algorithms implemented in a commercially-available acoustic processing program - Significantly improved support for broadband pulses in a commercially-available acoustic processing program - Enhanced data output options from one broadband echosounder manufacturer that support better processing of echoes close to the seabed - Improved echo classification of targets close to the seabed in conventional fisheries acoustics surveys

The field of fisheries acoustics has two major long-standing methodological problems: the difficulty in resolving fish that are very close to the seabed from the seabed itself (that is, the deadzone problem), and accurate species characterisation or classification. These problems limit the usefulness of acoustic techniques for demersal ecosystems. These problems can be reduced through the use of broadband acoustic techniques, which give considerably higher range resolution and finer frequency spectra than is possible with conventional multi-frequency echosounder systems. Accordingly, this project will apply the advantages of broadband pulses to allow us to see further towards the seabed. Better detection of single fish close to the seabed will yield high quality fish track data which will be used to develop and enhance fish characterisation algorithms. This will be achieved through three experimental operations: 1) tank and netpen measurements to better understand the mechanisms of detecting small targets adjacent to large targets, 2) dedicated at-sea experiments on demersal fish with associated optical and trawl ground-truthing, and 3) at-sea experiments in the context of a conventional demersal trawl-acoustic survey to integrate our methods in routine survey operations.