There is a wide a variety of sources of scattering in the aquatic environment\u2014biological and physical— all with different scattering properties.\u00a0 With this great diversity of scattering properties, there is a corresponding great challenge in interpreting acoustic scattering data, especially when there is more than one type of scatterer present.\u00a0 This figure was developed by Stanton, Lavery, and colleagues at WHOI and appears in various papers in different forms, including in Lavery et al. (2007).\u00a0 It also shows the broadband and narrowband frequencies used in some of our work.<\/p>\n\t Our laboratory measurements have helped to identify dominant scattering mechanisms of a variety of marine organisms.\u00a0 The measurements spanned the frequencies 24 kHz \u2013 1.2 MHz.\u00a0 The organisms were tethered by various means (including with human hair), and were rotated 0-360 degrees in 1-degree increments.\u00a0 We conducted most measurements on land, but took the tank to sea (on the deck of a ship) on two occasions so that we could study live freshly caught organisms.\u00a0 Illustration from Stanton et al. (2000).<\/p>\n\t Using pulse-compression signal processing on our broadband signals, we have been able to resolve individual features of various organisms.\u00a0 In this illustration, we could resolve the echoes from the skull and swimbladder of a fish.\u00a0 In other studies (not shown), we could resolve the echo from the dorsal and ventral sides of a euphausiid (a shrimp-like organism).\u00a0\u00a0 From Reeder et al. (2004).<\/p>\n\t Development of the acoustic scattering models required high resolution 3D information on the morphology of the organism.\u00a0 This illustration shows the information derived from our medical CT scans of a fish (upper panel: outer boundary of fish body and lower panel: \u00a0swimbladder).\u00a0 We have also done the same for a variety of zooplankton. From Reeder et al. (2004).<\/p>\n\t Through controlled acoustic scattering measurements of a variety of marine life, we have identified dominant scattering mechanisms specific to their shapes and material properties.\u00a0 This information was then used in our development of advanced acoustic scattering models.\u00a0 The acoustic models were generally categorized according to \u201cgross anatomical groups\u201d rather than particular species.\u00a0\u00a0 For example, a decapod shrimp and euphausiid, although different species, are in the same anatomical group, acoustically, as they are both elongated, have weakly scattering tissue, and have no gas or hard elastic shell.\u00a0 Adapted from Stanton et al. (1996).<\/p>\n\t When I entered this field, all zooplankton were treated alike— as spheres. \u00a0\u00a0We have developed advanced models that accurately model the shape of elongated organisms, ranging from the deformed bent cylinder through a full 3D model.\u00a0 Differences in material properties are also taken into account (fluid-like, gas-bearing, hard elastic shell) which had not been done to any significant extent before.\u00a0 From review paper: Stanton (2009).<\/p>\n\t This is an example of how well the models fit laboratory data that we collected.\u00a0 The acoustic frequencies spanned 50 kHz \u2013 1.2 MHz.\u00a0\u00a0 Two models were used\u2014a simple ray-based model and a more sophisticated DWBA model.\u00a0 Both gave surprisingly similar results, which demonstrates that under many important conditions, simpler models can suffice.<\/p>\n The below studies involved development of broadband ocean instrumentation and associated methods to study fish. The work was based, in part, on what we developed in the laboratory. The work was focused on swimbladder-bearing fish and associated resonance classification techniques, but the methods are broadly applicable to fish without swimbladders and zooplankton.<\/p>\n\t <\/p>\n End of successful offshore experiment (2007).\u00a0 Left to right:\u00a0 Cindy Sellers, Josh Eaton, Tim Stanton, Mike Jech, and Dezhang Chu.\u00a0 The large sphere (30-cm-diameter solid aluminum) in the foreground was used to calibrate the acoustics channels.<\/p>\n\t By towing the system deep and using pulse-compression processing on the broadband acoustic signal, individual fish could be resolved (lower panel), as opposed to the corresponding echogram (upper panel) from the ship-mounted narrowband echosounder that could not resolve the fish.\u00a0 Adapted from Stanton et al. (2010).<\/p>\n\t The swimbladder (a gas-filled organ) of these 25-cm-long herring resonated at 3.7 kHz at these deep depths. This resonance classification method reduces or eliminates ambiguities in interpretation of acoustic data.\u00a0 This resonance remained at the same frequency for two different patches of fish— one with a strong echo (red curve) and one with a weak echo (blue curve).\u00a0 The fact that the resonance frequency remained the same across the two patches showed that it was a change in numerical density, not size or orientation of fish, that caused the change in echo level.\u00a0 From Stanton et al. (2010).<\/p>\n\t An assemblage of fish containing two distinct size classes— 25 cm and 3-5 cm\u2014results in resonances at correspondingly two distinct frequencies.\u00a0 Through spectral analysis of the broadband echoes, the different sized fish can be spectral resolved (and, hence, quantified), even though they are not spatially resolved.\u00a0 Note that the resonance frequency of the herring decreased as they migrated upwards in the water column (deep in left panel, shallow in right panel).\u00a0 Adapted from Stanton et al. (2012).<\/p>\n We use a standard target (metallic sphere in this case) with known properties.\u00a0 The challenge is the fact that, since the broadband acoustic signal spans such a large frequency range, it encounters strong resonances (in the form of deep nulls) in the signal that cannot easily be directly removed.<\/p>\n\t The resonances in echoes from a sphere are due to interferences from the various types of waves associated with the scattering process.\u00a0 In this example, two waves are shown, one is the \u201cspecular\u201d echo from the front interface and the other is the circumferential wave that travels around the boundary.\u00a0 These two waves alone will combine in both constructive and destructive interference, depending upon the size and material properties of the sphere, as well as acoustic frequency.\u00a0\u00a0 Since the specular echo arrives first, the interference between the specular echo and all other waves that come in later can be eliminated by selecting (via a sampling time window) the specular echo first and not sampling the waves arriving later.\u00a0 This is facilitated through use of pulse-compression processing that reduces the duration of the echo to the inverse bandwidth of the signal.\u00a0 Illustration from Stanton\u2019s tutorial in 2014 on broadband acoustics, presented at MTS\/IEEE Oceans2014 conference in Taipei.<\/p>\n\t![\"\"](\"https:\/\/www2.whoi.edu\/staff\/tstanton\/wp-content\/uploads\/sites\/36\/2018\/05\/interpreting-acoustic-data.png\")
<\/a><\/p>\n\t<\/a><\/h2>\n\tThe Research Challenge<\/a><\/h2>\n
Experimental Determination of Dominant Scattering Mechanisms<\/h2>\n
<\/a><\/p>\n\t<\/a><\/h2>\n\t<\/a><\/p>\n\t<\/a><\/h2>\n\tModeling of Acoustic Scattering<\/h2>\n
<\/a><\/p>\n\t<\/a><\/h2>\n\t<\/a>
\nThese papers review both laboratory methods and scattering models that we have developed for fish and various types of zooplankton.\n\tMeasurements at Sea – Fish<\/h2>\n
<\/a>
\nDeployment of our broadband acoustic system, built by Edgetech. This is a modified version of a commercially available system that was originally designed by Edgetech to be a sub-bottom profiler. One of the modifications was to include more receiver elements to greatly improve receiver sensitivity for detection of echoes from fish. Adapted from Stanton et al. (2010).\n\t<\/a>In preparation for our sea experiments, we spent many days at the test well in the WHOI pier testing the equipment.\u00a0 This well is about 60\u2019 deep and has a crane and equipment van for our topside electronics.\u00a0 Tim Stanton is shown in this test (ca 2011).<\/p>\n\t<\/a><\/p>\nUsing Broadband Acoustics to Improve Resolution<\/h2>\n
<\/a><\/p>\n\tUsing Broadband Acoustics to Determine Size of Fish<\/h2>\n
<\/a><\/p>\n\t<\/a><\/h2>\n\t<\/a><\/p>\n\tCalibration of Broadband System<\/h2>\n
<\/a><\/p>\n\t<\/a><\/h2>\n\t