Search

Data are collected from a variety of acoustic systems in many countries to address a range of ecosystem monitoring and stock management objectives. A key step in the analysis of fisheries acoustics data is target classification, i.e. categorizing the backscatter data, ultimately by target species, so that it can be converted into estimates of abundance or biomass. The information needed to classify acoustic targets may be contained within the acoustic measurements, particularly if they are made over a range of frequencies. The SIMFAMI project, financed by the European Union, presented some multifrequency methods for species identification (Fernandes et al., 2006). Readers should also note that there are two other ICES reports on related topics: CRR No. 238 Report on Echo Trace Classification (Reid, 2000) and Acoustic seabed classification of marine physical and biological landscapes (ICES, 2007). However, as these reports were written when multifrequency and wideband methods were less mature, they mostly focus on single-frequency methods. Acoustic classification of biological targets is a fast-moving field. While most of the theoretical principles in the earlier reports are still relevant, there is a need to evaluate recent developments, expand their applications to contemporary technologies, and recommend target-classification protocols for use in fisheries research and ecosystem surveys. Several ICES Member Countries and observer countries have identified these needs and conveyed them to ICES Working Group on Fisheries Acoustics, Science, and Technology (WGFAST) and Science Committee (SCICOM). This is the first ICES CRR to detail the latest multifrequency and wideband methods for acoustic target classification.

Acoustic methods used in fish abundance estimation constitute a key part of the analytic assessment that makes the basis for fisheries research and international regulations of marine resources. The methods rely on power budget equations and calibrated systems. Different instrument specific formulations of power budget equations and calibration factors are used in modern scientific echosounder and sonar systems, such as the Simrad EK500 and the more recent Simrad EK60, ES60, ME70 and MS70 systems. These are instruments extensively used internationally for such applications, over the last decade or more. However, the lack of sufficient documentation in prior literature, on the actual power budget equations and calibration factors employed in these instruments, and their relationships to the traditional theory of fish abundance measurement, has caused some uncertainty and confusion among users. Control with systematic errors and drift demands the functional relationship of the abundance measurement to be fully known in terms of the electrical and acoustical parameters of the system used. The paper presents a unifying theory which seems to explain the different (instrument specific) power budget equation formulations and calibration factors employed in the mentioned echosounder and sonar systems, how they are related, and their relationship to the traditional and generic (instrument independent) theory of fish abundance measurement. Inconsistencies in prior literature are explained and corrected. Prior literature on the traditional theory is extended to provide more complete power budget equations for fish abundance estimation and species identification, by accounting for electrical termination, representation of echo integration, and the full range of electrical and acoustical echosounder parameters. These extensions apply both to the instrument specific and the generic instrument independent formulations. The expressions derived here provide a consistent and relatively complete theoretical basis for improved understanding and control in use of conventional methods and instruments for fish abundance measurement and species identification employed today. Akustiske metoder for mengdemåling av fisk representerer et nøkkelelement innen den teknologiske basis for moderne fiskeriforskning og internasjonal regulering av marine ressurser. Akustiske metoder er basert på bruk av effektbudsjett-ligninger og kalibrerte ekkolodd-system. Ulike instrument-spesifikke formuleringer av effektbudsjett-ligninger og kalibreringsfaktorer benyttes i moderne vitenskapelige ekkolodd og sonarsystemer, så som Simrad EK500 og de nyere Simrad EK60, ES60, ME70, og MS70 - systemene. Dette er instrumenter som over de senere tiår i utstrakt grad er benyttet internasjonalt for slike anvendelser. Mangel på tilstrekkelig dokumentasjon i tidligere litteratur, vedrørende de aktuelle effektbudsjett-ligningene og kalibreringsfaktorene som benyttes i disse instrumentene, og deres sammenheng med den tradisjonelle teorien for mengdemåling av fisk, har forårsaket noe usikkerhet og forvirring blant brukere av slike systemer. Kontroll med systematiske feil og drift i slike instrumenter, forutsetter at mest mulig fullstendige matematiske funksjonsuttrykk for bestandsestimering er kjent, i form av de elektriske og akustiske parametre som benyttes i aktuelle systemer. Rapporten presenterer en enhetlig teori som ser ut til å kunne forklare de ulike (instrument-spesifikke) effektbudsjett-ligningene og kalibreringsfaktorene som benyttes i ekkolodd- og sonarsystemer. Dette inkluderer hvordan funksjonsuttrykkene som er brukt i disse instrumentene er relatert, samt deres sammenheng med den tradisjonelle og generiske (instrument-uavhengige) teorien for bestandsestimering av fisk. Inkonsistens i tidligere litteratur er forklart og korrigert. Tradisjonell teori for mengemåling av fisk er videreutviklet og generalisert til å dekke mer fullstendige effektbudsjett-likninger for bestandsestimering og artsgjenkjennelse, ved å ivareta elektrisk terminering, representasjon av ekko-integrasjon, og de aktuelle elektriske og akustiske ekkolodd-parametre. Slik videreutvikling innbefatter både de instrument-spesifikke og de instrument-uavhengige (generiske) formuleringene. De funksjonsuttrykk som blir utledet her, utgjør en konsistent og relativt komplett teoretisk basis for forbedret forståelse og kontroll, i bruk av de konvensjonelle metoder og instrumenter for mengdemåling og artsgjenkjennelse av fisk som benyttes i dag.

For abundance estimation and species identification on fisheries acoustic surveys, control with temperature and finite amplitude effects in the sea demands the calibration factor C of the conventional echo-integrator equation to be fully known in terms of the sonar system’s electrical and acoustical parameters. As no such expression is available from earlier literature, the theoretical basis for the power budget and echo-integrator equations is revisited. The objective is to provide (a) a derivation of these equations for integration in a more complete functional relationship for abundance measurement, including an expression for the calibration factor C; (b) to account for electrical termination; (c) to formulate these expressions in terms of voltage signal echo integration processing; and (d) thereby generalize the Clay-Medwin formulations based on echo integration of “in-water” sound pressure signals, to account for the transducer and electronics components of the echosounder system. Under conditions of small-amplitude (linear) sound propagation, electroacoustic power budget equations are derived for the backscattering cross section in single-target backscattering, used in echosounder calibration; and the volume backscattering coefficient for multiple-target backscattering, used in oceanic surveys. On this basis a more complete echo-integrator equation is derived for two operational cases, “short-ping-and-long-gate”, and “long-ping-and-short-gate”. The results are consistent with and extend previous work in this area.

Rapporten er tospråklig: engelsk og norsk. This report summarizes technology, operations, experience, and scientific results from the Hermes Lander project. The Project established an autonomous, multi-sensor, sub-sea sensor platform powered by batteries. The platform was launched at a coral reef location in Hola off Vesterålen and collected data during seven months in 2010 (March – September). Oceanographic and echo-sounder data were analyzed and compared with similar data from sporadically collected data from research vessels as well as model-produced data. The results demonstrate a partly chaotic variation. The project neither confirmed nor excluded a connection between coral reef and biomass density. The data and experience are crucial for efficient planning, development, and operation of the planned Statoil-financed cable observatory in the same area.

only.

Traditionally, collection of acoustic data has been optimised for a single frequency, while requirements for the optimal combination of multifrequency data has been given less attention. Even though data manipulation can enable single-frequency data to be combined to some extent, optimal multifrequency comparisons cannot be made from a system if the input data are not collected properly. Multiple single-frequency [s(f)] data may be collected in various ways onboard research and fishing vessels. For detailed analysis, the physical and spatial characteristics of acoustic data should be as similar as possible. While direct comparability is impossible in all aspects, ideal data is defined as a reference point for the collection and analysis of multiple s(f) data. Acoustic data from several single frequencies are defined as ideal in this context if they can be used to generate combined frequency [c(f)] data at the same resolution as the original. This requires comparable physical measurements, carried out simultaneously from identical volumes, limited only by the effective range of the higher frequencies. Requirements necessary for recording ideal multi-frequency acoustic data are presented. Consequences of combining acoustic data originating from beams of different widths and separation distances are also analysed and illustrated.

The relative frequency response is an important acoustic feature used to characterise acoustic targets. This response has been defined as the sv, volume backscattering coefficient, for a specific frequency relative to that of a reference frequency (38 kHz). The acoustic data commonly used in these calculations are derived from integrated measurements in a region containing multiple targets. In this study the relative frequency responses at 18, 38, 70, 120, and 200 kHz have additionally been measured using filtered target strength data on all frequencies. The spatial comparability of the sv-data is thus avoided, while the single-target detection becomes a new challenge. Target strength was extracted from in situ measurements, using calibrated and digitised data from a Simrad EK60 with split-beam transducers transmitting simultaneously at all five frequencies. Selected series with nearly pure catches of Atlantic cod (Gadus Morhua L.), saithe (Pollachius virens L.) and Norway pout (Trisopterus esmarkii L.) were analysed. The frequency response derived by the new method is compared with standard integration method.

There is a current desire to harvest marine resources by managing total marine ecosystems rather than single species of the ecosystems. By means of algorithms applied on high-quality multi-frequency acoustic data, species, or rather acoustic categories, of the ecosystem can be identified. This information may significantly increase the accuracy of acoustic survey estimates of fish and to some extent also for zooplankton. Multi-frequency split beam echo sounders with nearly identical and overlapping acoustic beams have been regularly used in acoustic surveys for fish stock abundance estimation at Institute of Marine Research for the last five years. Calibrated raw data from up to six simultaneously working echo sounders at 18, 38, 70, 120, 200 and 364 kHz was used as input to a stepwise, modular sequence of analysis, like bottom detection, noise quantification and removal, target categorisation and school detection in near real-time. Direct generation of new, synthetic echograms, based upon the measured or modelled relative frequency response of the targets is one of the most useful features of the systems. The result of the categorisation process can be used to show the spatial distribution of different acoustic categories in a single synthetic echogram, or to keep some and remove other acoustic categories in echograms at a single frequency.

The accuracy of the multi-frequency species identification and separation algorithms may be significantly improved when the detailed acoustic scattering characteristics of the fish are known, including its expected variability and uncertainty. The modeling approach is effective for interpreting measured data collected for different fish species and may give a better understanding of the acoustic scattering characteristics. The paper is motivated by development of algorithms of multi-frequency species identification and separation of three economically valuable North European fish species: Atlantic mackerel, Norwegian spring spawn herring and Norway pout. The backscattering by these fish is studied using an advanced modeling approach, and the results of the analysis are compared with selected multi-frequency recordings. This allows for an explanation of the main backscattering mechanisms over the selected frequency range and often a better understanding of the dominating acoustic scattering parameters. The impact of fish orientation behavior, pressure (for herring) and the morphological state of the fish is also discussed. Typical backscattering frequency responses for herring, mackerel and Norway pout are explained theoretically. The obtained results may be a base for improving the automatic identification algorithms now introduced in scientific echo sounder post-processing systems.

The purpose of this work has been to test complementary methods in order to classify marine organisms, with particularly attention to zooplankton and fish. Algorithms to separate fish and zooplankton have been developed and implemented at IMR and at IRD. A novel optimised model framework based on known scattering models are used to classify zooplankton and to separate these from fish. Acoustic data from up to 6 frequencies were collected to test the scattering model framework, while concurrent biological samples from multi-net oblique or horizontal MOCNESS tows, WP2 vertical net hauls and pelagic trawl were also obtained and analysed. Great attention are given on one side to the inter calibration and the comparability of all the frequencies, and to the space and time coherence between the samples collected and the acoustical data which are processed. All algorithms involve zooplankton scattering models, the high-pass ones from Stanton, or the more complex ones like the truncated fluid sphere from Holliday or the DWBA set of models from Chu and Stanton. A set of reliable acoustical and biological data has been chosen in order to proceed to comparisons between the results of the acoustic data processing through the classification algorithms and the results of the biological processing.

Calibrated, digitised data from simultaneously working multi-frequency echo sounders with nearly identical and overlapping acoustic beams have been used to generate new, synthetic echograms where only targets identified as Atlantic mackerel are retained. Echo sounder raw data are processed stepwise in a modular sequence of analysis to improve the ability to categorise acoustic targets. The relative frequency response measured over up to six acoustic frequencies, 18, 38, 70, 120, 200 and 364 kHz, is the main acoustic feature used to characterise acoustic backscatter. Mackerel seems to have a frequency-independent backscatter below approximately 100 kHz, and above approximately 200kHz, but at 4 times higher level of the backscatter. Results from numeric modelling explaining the measured relative frequency response of mackerel are shown. Synthetic echograms containing targets identified acoustically as mackerel are presented and evaluated against trawl catches. Even though catching of the fast-swimming mackerel is difficult, the trawl catches from two Norwegian research vessels confirm that the targets identified acoustically as mackerel is really mackerel.

Calibrated and digitised data from two or more discrete echo sounder frequencies can be combined for the purpose of separating and extracting the acoustic scattering from zooplankton and fish in mixed recordings. Methods are described for echogram manipulation and for the construction of new synthetic combined-frequency [m] echograms. Examples of extracted scattering information from mixed layers of fish and small scattering organisms, such as copepods and euphausiids, are shown, and the potential of each of the different cm methods is discussed. Key words: Acoustic, interpretation, echo integration, multi-frequency, synthetic echogram. NORSK SAMMENDRAG: Kalibrerte digitaliserte data fra to eller flere akustiske frekvenser på ekkolodd kan kombineres med formål å skille fra hverandre og trekke ut ekko fra zooplankton og fisk i blandede forekomster. Metoder for manipulering av ekkogram og for bmbinering av akustiske data inn i nye syntetiske ekkogram blir beskrevet. Eksempler på hvordan ekko informasjon fra blandede registreringer av fisk og smil organismer som kopepoder og krill vises, og potensialet til hver enkelt av kombinert frekvens metodene diskuteres.

Survey data from six herring assesment cruises in the wintering areas for spring spawning Norwegian herring were analysed with regard to variation in abundance and depth with time of day. The data were quite consistant between cruises, showing a downwards vertical migration at in situ dawn and an upwards migration at dusk, with a subsequent descent of parts of the herring through the dark period. This behaviour was attributed to schooling and predator avoidance during the day, and energy conservation and swimbladder refilling during the night. There were significantly high average SA values during the short winter day compared to very low values at dawn and dusk, and also quite low values at night. This variation was mainly attributed to articulate angular distribution of the herring during dusk and dawn, and also during the night time swim/sink/glide energy conservation exercise. The implications on assesment surveying is discussed.

The Bergen Echo Integrator (BEl) is a software system designed for convenient postprocessing of echo sounder data. BEl is realized in international standards and is essentially machine independent. Development of BEl for analysis of such diverse scatterers as plankton, pelagic fish, bottom fish and bottom itself is reviewed. The system design is described and the latest advances are discussed. Among these: a more flexible database, an improved man-machine interface, an easier procedure for accessing data and a system for mapping the data. General operating procedures are described for extraction of information while scrutinizing acoustical data.

In 1992 an integrated model for marine research data was developed at the Institute of Marine Research. Traditionally, the marine research disciplines collect data from their own fields, normally used in a limited context. In many research programs there is a strong need to crossrelate data from various sources. To accomplish this it is essential to establish an integrated database, with a uniform data representation and a structure that is flexible enough to represent all kinds of marine data. A centralized database increases the availability, avoids redundancy and enforces access control and protection against loss of data. The model is based on a generic data structure where "platform", "operation", "object" and "measurement" are central entities. In an operation (e.g. trawling) on a platform (e.g. ship), series of measurements (e.g. lengths) of different objects (e.g. fish) are gathered. Data can be classified according to different systems (e.g. taxonomy, regions and methods). The database is implemented in the Ingres relational database system running on HP9000/755 computers with Unix operating system and connected to the institute's local network and Internet. Tests have been carried out to evaluate functionality and performance. While the test results are acceptable, minor immediate modifications are needed. The database will be subject to further development in the future. Currently, work is in progress on data quality assurance, inserting data into the database and adapting software applications. The database is planned to be in regular use at the end of 1993.

Data are collected from a variety of acoustic systems in many countries to address a range of ecosystem monitoring and stock management objectives. A key step in the analysis of fisheries acoustics data is target classification, i.e. categorizing the backscatter data, ultimately by target species, so that it can be converted into estimates of abundance or biomass. The information needed to classify acoustic targets may be contained within the acoustic measurements, particularly if they are made over a range of frequencies. The SIMFAMI project, financed by the European Union, presented some multifrequency methods for species identification (Fernandes et al., 2006). Readers should also note that there are two other ICES reports on related topics: CRR No. 238 Report on Echo Trace Classification (Reid, 2000) and Acoustic seabed classification of marine physical and biological landscapes (ICES, 2007). However, as these reports were written when multifrequency and wideband methods were less mature, they mostly focus on single-frequency methods. Acoustic classification of biological targets is a fast-moving field. While most of the theoretical principles in the earlier reports are still relevant, there is a need to evaluate recent developments, expand their applications to contemporary technologies, and recommend target-classification protocols for use in fisheries research and ecos ystem surveys. Several ICES Member Countries and observer countries have identified these needs and conveyed them to ICES Working Group on Fisheries Acoustics, Science, and Technology (WGFAST) and Science Committee (SCICOM). This is the first ICES CRR to detail the latest multifrequency and wideband methods for acoustic target classification. Refereed 14.A 14.4 Fish abundance and distribution Manual Guide