Acoustical and Optical Sensor Integration

Chairman: Peter Ortner
Rapporteur: Lewis Incze

Participants: Charles Barans, Mark Berman, Cabell Davis, Brad Doyle, Charles Greenlaw, Alex Herman, Van Holliday (briefly), Mark Huntley (briefly), Jules Jaffe, Gus Paffenhöfer and Rudi Strickler

Scientific Context

GLOBEC intends to study the population dynamics of key species and the processes controlling their abundance in a variety of marine ecosystems. The first field program will study Georges Bank and will focus, at least initially, on the copepod Calanus finmarchicus and on the early life history of cod, Gadus morhua. Virtually all biological rates are assumed to be modulated by physical conditions and motions. The influence of physics on biology occurs over a wide range of spatial and temporal scales. The population-level response of Calanus on Georges Bank depends upon, among other factors, the spatial extent, seasonal timing, frequency and amplitude of external forcing from atmospheric weather, oceanic anomalies (e.g., Gulf Stream rings), cross-bank advection, and the migrations of predator populations (e.g., fish). The goal of GLOBEC is to understand, and ultimately to predict, population changes by determining (from first principles) the processes affecting variability in population abundance. That is, GLOBEC seeks to understand the combination of physical and biological interactions affecting reproduction, distribution, and mortality of selected taxa. This requires identification and documentation of small-scale processes and distributions and then quantitative linkage between these processes and larger-scale forcing functions.

Working Group Deliberations

Members of the working group agreed that the following observation problems would have to be resolved to meet GLOBEC's stated goals.

Distribution

Scale Linkage

We must discover and document whatever linkages exist among the various temporal and spatial scales in order to relate small-scale processes (those affecting the individual) to population level responses. We must understand how these various time and space scales are mechanistically connected and must design appropriate observational techniques to obtain the necessary data to validate models of population response to climatic conditions.

Processes

We must be able to measure relevant biological processes and understand their variability among individuals under different conditions at spatial and temporal scales relevant to the particular process. Such processes include grazing, predator/prey interactions, and mating and reproduction. We felt our principal task was to consider whether, and to what degree, an integration of biooptical and bioacoustic sensor technology could address these problems and how they might be addressed realistically given the state-of-the-art and the prospects for advances in sampling technology.

As an initial exercise we attempted to specify, a priori, what reciprocal benefits could be obtained by merging technologies or obtaining complementary data. From the viewpoint of biooptical system users, bioacoustics could fill the following needs:

  1. Provide a spatial map of the broad scale distribution of selected taxa and pinpoint features of interest for fine-scale process studies;

  2. Provide data on vertical migrations;

  3. Provide data on larger-scale three dimensional structure;

  4. Make it possible to track a group of organisms over periods enabling a series of process studies to be made;

  5. Provide doppler measurements of swimming speeds and statistics on population behavior (e.g., number swimming up/down/not at all); and

  6. Provide information on larger, rarer organisms that might constitute predators upon organisms whose interactions are being studied with current bio-optical systems.
From the viewpoint of bio-acousticians, bio-optics could fill the following needs:

  1. Provide information on target orientation;

  2. Provide independent size distribution estimates perhaps as transect sample series data (e.g., utilizing current towed platforms like UOR, SeaSoar, or Batfish) within a bioacoustic map; and

  3. Provide taxonomic identification of bioacoustic targets.
All participants agreed simultaneous sampling with traditional sampling devices (nets, pumps) was still essential since no single method was sufficient, and confidence in accuracy could only be obtained by using independent methods and ascertaining the degree to which their estimates converged. All participants agreed that GLOBEC's objectives implied concomitant physical (temperature, salinity, and current) data collection. Quite likely physical, bio-optical, and bioacoustical methods would have to be integrated to obtain estimates of micro- and fine-scale turbulence and its effects on behavior and organism-organism interactions.

Current bio-optical methods sample in close proximity to the sensor. Bioacoustic methods sample on nearly the same scale but can also sample larger water volumes much further away from the sensor. At these large scales bioacoustic methods sacrifice resolution. This loss, however, may not be the greatest problem to be faced in linking various scales. The group felt the more difficult problem will be confidently linking 1 mm to 1 cm individual organism behavioral scales to 100 m to km subpopulation scales. Last, although we felt the scale problem was likely to be a primary issue in the Fisheries Acoustic Working Group, the participants recognized that while GLOBEC is focusing on key species in the plankton the bioacoustic survey systems used will have to encompass the lower frequencies necessary to sample larger organisms such as euphausiids and fish.

Available Technology

Prior to detailing specific recommendations the group enumerated promising bio-optical or bioacoustic technologies and instruments and tried to characterize their relative costs and availability. We described the following general groups.

Some instruments holding promise for GLOBEC are currently available off-the-shelf for comparatively moderate cost. These include the following devices:

  1. Optical Plankton Counter (laboratory or towed) - A towed or profiling sensor used to count and size zooplankton in the 0.25 mm - 3.0 cm range. It was developed by A. Herman of the Bedford Institute and is commercially available from Focal Technology, Inc.;

  2. CritterCam [R] (laboratory, lowered on a cable, mounted on a submersible or a ROV) This camera system using an IR diode laser was developed by J.R. Strickler. It is commercially available from LNG Technical Services;

  3. ADCP Backscatter (vessel mounted or moored) - Software and hardware are now commercially available for this purpose from RDI, Inc. The method was described in the literature by C. Flagg and S. Smith;

  4. Commercial Echo Integration (dual- or split-beam on towed bodies or vessel mounted) - Dual-beam systems at various frequencies currently are available from various manufacturers including BioSonics, Inc. Results using these systems to sample micronekton have been published by C. Greene and P. Wiebe. Split-beam systems are available from Simrad.
Other instrument systems require modification or adaptation to be applicable to GLOBEC problems. Others have only been developed as prototypes in certain laboratories. As a result these will likely cost more money to bring on-line. These include the following instruments:

  1. Moored: OPC, CritterCam [R] (video systems or serial plankton recorders) - The OPC system that might be used in this application is under development and is substantially different from that commercially available today. A plankton recorder of this type has been developed and used at WHOI (C. Butman). Field trials with a moored CritterCam[R] are planned for the fall of 1991;

  2. Video Plankton Recorder (VPR) - A towed video camera system under development at WHOI. It is intended to sample on centimeter scales over many kilometer transects;

  3. Plankton Image Analyzer- A device developed at URI/NMFS-Narrragansett to enumerate samples of zooplankton or their recorded images and classify individuals into taxonomic groups;

  4. Simple Multiple-frequency Systems- Systems of this type are under development at a number of institutions. A prototype of a fully modular quantitative echo-integration system employing up to six frequencies, and capable of real-time data analysis and display, has been deployed on MOCNESS and in situ plankton cameras by NOAA/AOML and Tracor investigators. Development is also underway on such systems in Norway (Dalen), at BIO in Canada (Sameoto), and at the University of Wisconsin (Clay);

  5. Commercial ROVS with Bioacoustic Imaging/Location Systems - (Greene et al., 1991).

  6. High Resolution "Shadowgraph" Side-scan Sonar- A system of this type was used by C. Barans and V. Holliday to sample large demersal fish species.
The final category includes systems or methods that are likely to be costly either initially because of large research and development commitments (although the individual units eventually may be produced at moderate cost), or because they are inherently complex and likely to remain expensive. For the latter group a "facility" model of operation and maintenance may be required.

  1. Major Research and Development Projects

  2. Potential Facilities

Specific Recommendations

Integrate bioacoustic and bio-optical sampling technology so as to reduce the ambiguity inherent to purely bioacoustical measurements.

Without target identification bioacoustic measurements of the biota will not provide the information required by GLOBEC. Optical data can efficiently provide much of the requisite calibration data (including e.g., target orientation information). In addition to these bio-optical data traditional sampling will be required to ground-truth indirect methodologies. Such integration currently is being pursued and needs to be more explicitly emphasized as especially fundamental to GLOBEC.

Utilize bioacoustic sampling to extrapolate bio-optically based process information to larger time and space scales.

Video and camera information on organism behavior and feeding are typically obtained on small sample volumes over comparatively short time scales. By nesting such experiments inside larger-scale bioacoustic maps the results can be generalized to the regional or population level.

Develop bioacoustic and bio-optical techniques that provide information within a 1 m3 volume to characterize processes operating on scales <1 cm.

At the present time no such systems are readily available but are considered to be essential to GLOBEC's fundamental process orientation given the size of the target organisms selected. Both technologies can resolve targets on these scales and would be employed most fruitfully in conjunction with one another.

Develop processing and analysis technology to the point where population distributions of target organisms can be visualized in near real-time.

Without such advances it will be impossible to accomplish certain GLOBEC requirements including sampling a coherent population over time and conducting a series of process or behavioral experiments at the scales implicit in the GLOBEC program.

Develop integrated bio-optical and bioacoustic systems that can be deployed at various depths on fixed moorings instrumented with physical (temperature, salinity, and current) and chemical (fluorescence) sensors.

Such systems are likely to be required by GLOBEC in its initial field program to characterize the advection of biological populations on and off shallow banks. They should be designed with the capability of 2-way telemetry so that sampling rates can be modified if unusual events are detected, and to monitor system performance. They also could be equipped with sample collection systems of various kinds.

Explicitly recognize the significant problems of data assimilation, archiving, and retrieval inherent in utilization of such bioacoustic and bio-optical sampling systems.

These systems generate volumes of data orders of magnitude larger than we are currently equipped to process and store. Simple storage of raw data is likely precluded. Whatever systems are eventually adopted will doubtless be dependent upon the availability of sufficient computer capability at sea to analyze raw data and transform it into manageable processed units like images or maps. An advanced computer system needs to be placed on at least one ship to be used in GLOBEC. Moreover, the accumulation of data may well exceed even enhanced shipboard storage capacities and a high-speed data link to shore via satellite communication will be essential. The same facility may be critical to coordinate the activities of multiple platforms during a GLOBEC field experiment.

Develop bio-optically and bioacoustically instrumented Lagrangian platforms that can be deployed at various depths (or are capable of changing depth) so as to follow a targeted population.

This application presents special technical challenges beyond deploying similar systems on fixed moorings. It may not be essential to GLOBEC in its early phase but may become essential when the behavioral responses of target populations and the most significant regulatory processes are better understood.

Modify commercial echo sounder technology to the higher frequencies suitable for initial GLOBEC target organisms (copepods) to permit routine (although perhaps not entirely quantitative) mapping.

This approach is felt to present few technical challenges and to be especially cost-efficient.