prEN 15910

SLOVENSKI STANDARD
01-junij-2009
.DNRYRVWYRGH1DYRGLOR]DRFHQMHYDQMHãWHYLOþQRVWLULE]PRELOQRKLGURDNXVWLþQR
PHWRGR
Water quality - Guidance on the estimation of fish abundance with mobile hydroacoustic
methods
Wasserbeschaffenheit - Aufnahme von Daten zur Fischpopulation mittels
hydroakustischer Verahren
Qualité de l'eau - Guide sur l'estimation de l'abondance des poissons par des méthodes
hydroacoustiques mobiles
Ta slovenski standard je istoveten z: prEN 15910
ICS:
13.060.70 Preiskava bioloških lastnosti Examination of biological
vode properties of water
2003-01.Slovenski inštitut za standardizacijo. Razmnoževanje celote ali delov tega standarda ni dovoljeno.

EUROPEAN STANDARD
DRAFT
NORME EUROPÉENNE
EUROPÄISCHE NORM
March 2009
ICS
English Version
Water quality - Guidance on the estimation of fish abundance
with mobile hydroacoustic methods
Qualité de l'eau -Echantillonnage de données de Wasserbeschaffenheit - Aufnahme von Daten zur
populations de poissons par hydroacoustique Fischpopulation mittels hydroakustischer Verahren
This draft European Standard is submitted to CEN members for enquiry. It has been drawn up by the Technical Committee CEN/TC 230.
If this draft becomes a European Standard, CEN members are bound to comply with the CEN/CENELEC Internal Regulations which
stipulate the conditions for giving this European Standard the status of a national standard without any alteration.
This draft European Standard was established by CEN in three official versions (English, French, German). A version in any other language
made by translation under the responsibility of a CEN member into its own language and notified to the CEN Management Centre has the
same status as the official versions.
CEN members are the national standards bodies of Austria, Belgium, Bulgaria, Cyprus, Czech Republic, Denmark, Estonia, Finland,
France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Malta, Netherlands, Norway, Poland, Portugal,
Romania, Slovakia, Slovenia, Spain, Sweden, Switzerland and United Kingdom.
Recipients of this draft are invited to submit, with their comments, notification of any relevant patent rights of which they are aware and to
provide supporting documentation.
Warning : This document is not a European Standard. It is distributed for review and comments. It is subject to change without notice and
shall not be referred to as a European Standard.
EUROPEAN COMMITTEE FOR STANDARDIZATION
COMITÉ EUROPÉEN DE NORMALISATION
EUROPÄISCHES KOMITEE FÜR NORMUNG
Management Centre: Avenue Marnix 17, B-1000 Brussels
© 2009 CEN All rights of exploitation in any form and by any means reserved Ref. No. prEN 15910:2009: E
worldwide for CEN national Members.

Contents Page
Foreword .3
Introduction .3
1 Scope .3
2 Normative references .4
3 Terms and definitions .4
4 Principle and field of application .4
5 Equipment .6
6 Survey design .8
7 Survey data acquisition . 12
8 Post-processing of acoustic data . 13
9 Results and reporting . 18
Annex A (informative) Common abbreviations used in this document . 27
Annex B (informative) Supplementary data . 28
Annex C (informative) Methods for estimates of fish abundance . 29
Annex D (informative) Interpretation of TS into fish length and weight . 30
Annex E (informative) Deconvolution procedure . 35
Annex F (informative) Determination of the Elementary Distance Sampling Unit (EDSU) . 37
Annex G (informative) EIFAC/CEN Acoustic Workshop . 38
Bibliography . 39

Foreword
This document (prEN 15910:2009) has been prepared by Technical Committee CEN/TC 230 “Water analysis”, the
secretariat of which is held by DIN.
This document is currently submitted to the CEN Enquiry.
WARNING — Persons using this European Standard should be familiar with normal laboratory and fieldwork
practice. This standard does not purport to address all of the safety problems, if any, associated with its
use. It is the responsibility of the user to establish appropriate safety and health practices and to ensure
compliance with any national regulatory conditions.
IMPORTANT – It is absolutely essential that tests conducted according to this European Standard be carried
out by suitably trained staff.
Introduction
This is one of several European Standards developed for the evaluation of species composition, abundance and age
structure of fish in rivers, lakes and transitional waters. The following standards have already been published:
EN 14011, Water quality — Sampling of fish with electricity
EN 14757, Water quality — Sampling of fish with multi-mesh gill nets
EN 14962, Water quality — Guidance on the scope and selection of fish sampling methods
Common abbreviations that are used in this document are compiled and explained in Annex A.
The initial draft of this document was constructed by an international group of experts during an ad hoc joint
EIFAC/CEN workshop (see Annex G).
1 Scope
This European Standard describes a standardised method for data sampling and procedures for data evaluation of
fish populations in large rivers, lakes and reservoirs, using hydroacoustic equipment deployed on mobile platforms
(boats and vessels).
This standard covers fish population abundance estimates of pelagic and profundal waters > 15 m mean depth with
the acoustic beam oriented vertically, and the inshore and surface waters of water bodies > 2 m depth with the beam
oriented horizontally. The size structure of fish populations can only be determined to a relatively low degree of
precision and accuracy, particularly from horizontally-deployed echosounders. As acoustic techniques are presently
unable to identify species directly, other direct fish catching methods should always be used in combination.
This standard provides recommendations and requirements on equipment, survey design, data acquisition, post-
processing of data and results and reporting. A selected literature with references in support of this standard is given
in the Bibliography.
2 Normative references
The following referenced documents are indispensable for the application of this document. For dated references,
only the edition cited applies. For undated references, the latest edition of the referenced document (including any
amendments) applies.
EN 14757, Water quality — Sampling of fish with multi-mesh gillnets
EN 14962, Water quality — Guidance on the scope and selection of fish sampling methods
3 Terms and definitions
For the purposes of this document, the terms and definitions given in EN 14962 apply.
4 Principle and field of application
Hydroacoustic (or echosounding) technologies are effective and efficient methods for sampling fish in the water
column [37]. Fisheries acoustics methods are analogous to remote sensing techniques and advantageous to other
sampling methods as nearly the entire water column can be sampled quickly and non-destructively, areal coverage
is continuous, data resolution is on the order of tenths of metres, and data can be post-processed in a variety of
ways. However, other methods and procedures are required for determination of species identity and age structure.
Acoustics is used to gather information remotely by transmitting a pulsed beam of sound energy into a water body
and subsequently detecting and analysing the returning echoes. Systems are available with single-, dual-, split- and
multi-beams, although the latter two types have now superseded the other two systems. Acoustic systems are
usually deployed from a moving boat in large water bodies. A computer is required for control of the echo sounder in
the field and for the data processing.
This standard covers acoustic sampling of deep lakes, reservoirs, shallow lakes and wide lowland rivers. The pelagic
and profundal waters of lakes > 15 m depth are surveyed with the acoustic beam oriented in the vertical axis, whilst
inshore and surface waters of lakes and lowland rivers > 2 m depth are surveyed with the beam oriented horizontally
[20], [24]. Water bodies of all trophic levels can be sampled acoustically and a wide range of fish communities and
targets, ranging from young of the year to large mature fish can be detected and quantified (Table 1).
Mobile acoustic surveys provide several layers of information; from relatively simple presence / absence studies of
target species, to spatial (or temporal) distributions of individuals or groups, to fully quantitative density and (when
combined with other sampling techniques) system-wide biomass estimates.
Correctly obtained acoustic sampling data are directly related to population density. The strategy shall be to sample
a defined area or volume of lake or river using appropriate equipment (Clause 5), data collection (Clause 7) and data
processing procedures (Clause 8), presenting the results in a standard reporting format (Clause 9) to provide
estimates of fish abundance. Abundance in this context can be either a relative or an absolute measure of
assessment based on a single survey of a known area or volume of water.

Table 1 — Suitability of hydroacoustic sampling techniques for inland water bodies and fish communities
Application Objectives Water Types Target Species and Life Stages Limitations
a
Vertical Beaming Fish population abundance estimates Fish in pelagic and profundal waters Poor coverage of surface and littoral waters
Lake Categorie 1
Fish population size structure YOY to adult Must be used in conjunction with direct capture methods for species
b
Lake Categorie 3
composition and age structure
a
Horizontal Beaming Fish population abundance estimates Lake Categorie 1 Fish in littoral and surface waters Poor coverage of pelagic and profundal waters
b
Fish population size structure Lake Categorie 3 YOY to adult Vulnerable to interference from macrophytes and entrained air
c
Low confidence in size-structure from lakes and slow-flowing rivers
River Categorie 3
Must be used in conjunction with direct capture methods for species
d
River Categorie 4
composition
e
River Categorie 5
a
Combined Vertical Fish population abundance estimates Lake Categorie 1 Fish in pelagic, profundal, littoral and Horizontal beaming vulnerable to interference from macrophytes and
and Horizontal surface waters entrained air
b
Fish population size structure Lake Categorie 3
Beaming
YOY to adult Low confidence in size-structure from horizontal beaming
Must be used in conjunction with direct capture methods for species
composition
Categories of lakes and rivers see EN 14962:
a
With a pelagic or profundal zone, area < 0,5 km
b
With a pelagic and profundal zone, Area > 0,5 km
c
Width < 30 m, maximum depth > 2 m
d
Width 30 m to 100 m, maximum depth > 2 m
e
Width > 100 m maximum depth > 2 m
5 Equipment
Although current acoustic equipment is accurate and reliable, it must be used correctly with a fundamental
understanding of factors that can affect its performance. Sources of systematic error or bias in acoustic survey
results include calibration errors, hydrographic conditions, diel fish behaviour and migration [37]. Other practical
limitations are sources of unwanted echoes (reverberation), such as plankton, debris, submerged macrophytes and
entrained air bubbles.
5.1 System performance
Recommended equipment specifications are given below as minimum and optimum requirements:
Minimum:
Whilst it is accepted that useful information may be obtained from a wide variety of echosounder types, the minimum
requirement for a scientific survey is that a “Scientific” sounder with the following characteristics be used:
 quantitative fisheries echosounder (calibrated) and operating at an appropriate frequency for the waterbody and
target fish species, probably between 38 kHz and 1,8 MHz [34];
 enables data storage of calibrated data for reprocessing;
 enables data processing in order to generate abundance and size distribution outputs.
Optimum:
Because of their inherent and obvious advantages, it is recommended that scientific split or multi-beam sounders be
used if possible.
5.2 Calibration
5.2.1 General
Calibrations are conducted to ensure that the echosounder and transducer are measuring fish abundance and fish
size correctly. Secondly, they verify that the complete acoustic system is operating properly and remaining stable
over time, permitting comparisons among survey periods and allowing inter-echosounder comparisons. All
calibrations should be based on and follow the manufacturer’s manual and recommendations.
5.2.2 Types of calibration
5.2.2.1 “Full” instrument and equipment calibration
This is usually conducted by the manufacturer, once in a lifetime for most transducers, but it should also be done
whenever the transducer has been subjected to physical damage.
5.2.2.2 “Beam pattern” calibration
This should be conducted at least once per year or whenever the transducer or cable is suspected of being
subjected to physical damage.
5.2.2.3 “Standard Target” tests
These should be conducted at each survey site in order to verify that the system is operating properly and to correct
for environmental factors.
The specific requirements for each calibration / standard target test are summarised below.
5.2.3 Full calibration
Full calibrations shall be conducted by the manufacturer, or at a facility approved by the manufacturer.
Shall be done separately for each transmitted pulse duration, transmit source level and receiver gain settings being
used.
Should also be done if the transducer, transducer cable or echosounder have experienced any physical damage.
Records shall be kept of each calibration (if possible, raw data should be stored) in order to assess substantial
changes in power parameters during the lifetime of the transducer.
5.2.4 Beam pattern calibration
For both vertical and horizontal applications (i.e. vertical deep or shallow lake surveys and horizontal lake and river
surveys), beam pattern calibrations shall involve:
 Vertical calibration in a free field (i.e. one with no lateral boundaries) under high signal to noise ratio (SNR)
conditions.
 Confirmation of temperature and salinity in order to accurately determine the speed of sound and absorption
coefficient. Mean water temperature should be measured as a depth profile in 1 m intervals over the whole
water column.
1)
 A minimum target distance of 2 x the theoretical near-field
 Avoidance in the beam of scattering layers such as thermal stratification, fish, air bubbles or zooplankton.
 A minimum distance of 2 x the transmitted pulse length between the calibration sphere and the bottom.
 Parameters to be measured should include beam-width and angle-offset measurements.
 After physical trauma to the cable and transducer housing, damage shall be repaired and a new beam pattern
calibration shall be conducted.
 If the calibration parameters do not deviate too much from previous calibrations, the transducer and cable can
be considered fully functional. The manufacturer should be able to provide information about acceptable
deviation.
5.2.5 Standard target test
For both vertical and horizontal applications (i.e. vertical deep or shallow lake surveys and horizontal lake and river
surveys) the standard target test should ideally be carried out at the start of every new survey or day (irrespective of
the survey location or strategy). It shall include:
 The passage of a standard target through the beam to check that results are, within tolerances, as expected
(e.g. Table 2). Tolerances will vary depending on beam orientation (vertical or horizontal) and the signal to noise
ratio (SNR). A minimum of 250 echoes is recommended on the acoustic axis and within each quadrant.
 The transducer shall be acclimated to water temperature and air bubbles removed from the transducer face and
standard target.
1) The transducer may need to be lowered well below the surface of a deep water body to avoid, for example, wave action and
bubbles at the surface, whilst still having the necessary range available.
 The standard target test shall be conducted in the same environmental conditions (water temperature and
salinity) as are experienced during the survey.
 Standard target tests shall be conducted with the same pulse durations, transmit powers, and bandwidths used
during the survey.
 For mobile horizontal surveys, a horizontal standard target test, ideally with the standard target positioned at
different ranges from the transducer, shall be performed. This is in order to determine if the Time Varied Gain
(TVG) function follows normal spherical spreading. Otherwise, potential bias may be introduced when
interpreting acoustic fish-size data.
 For mobile horizontal surveys, periodic fixed location temperature profile measurements shall be taken in order
to verify normal spherical spreading of the acoustic beam.
No adjustments shall be made to the equipment settings as a result of this test, but a beam pattern or full calibration
is required if the result is unsatisfactory. For shallow lakes or horizontal surveys this may require relocating to a
suitable test site.
Table 2 — Target strengths (TS) of tungsten carbide spheres with different diameters for sound speed of
s-1
1 450 m in fresh water [37]
Frequency Diameter Fresh WaterTS
kHz mm dB
38 38,1 -42,1
70 36,4 -40,9
70 38,1 -40,6
120 33,2 -40,8
120 38,1 -39,8
200 36,4 -39,5
200 38,1 -39,5
420 8,9 -52,3
420 21,2 -43,5
6 Survey design
6.1 General
Acoustic surveys are conducted to investigate large volumes of water. In practice, owing to the limited time available
to perform the survey, only a small proportion of this volume can be observed acoustically. Transect-based surveys
are, therefore, based on the assumption that the measurements, which are made along the survey tracks, are
representative samples of the wider distribution of the target species in the water volume under study [37]. Since
only a portion of the overall area of concern is actually sampled, any survey design consists of choices that need to
address specific objectives, which can vary from an overall estimate of abundance for an entire population to simply
the identification of locations of fish concentrations.
6.2 Design for appropriate resolution and detection
When planning a vertical acoustic survey, sampling should be planned in order to produce a three-dimensional
picture of fish density using depth strata at least to the resolution of EN 14757 gillnet layers (0 m to 3 m, 3 m to 6 m,
etc.). For both vertical and horizontal surveys, the signal to noise ratio should be maximised.
6.3 Pre-planning
Prior to conducting an acoustic survey, the following information should be assembled for the water body under
study:
 Sufficient bathymetric data. If necessary, pre-surveys specifically for the collection of depth data should be
conducted. For surveys of reservoirs, it is important to make a record of water depth at the time of the survey.
 Resident fish species data and limnological information.
 Potential temperature and oxygen stratification.
 Access permissions.
 Weather forecast (particularly wind speeds and direction).
 Identification of the cruise track:
 Define the area to be covered by the survey. Ideally, this would be the entire lake or river, however some
areas may not be feasible for hydroacoustics (e.g. too shallow or obstructed by stands of macrophytes).
 Within the area under consideration, the choice of spacing and track layout (e.g., systematic parallel,
random parallel, systematic zig-zag, etc.) should reflect an understanding of the serially correlated nature of
the acoustic sampling technique and a consideration of the expected patchiness of the population of
interest. For lakes:
 The first preferred cruise track is a systematic parallel design, with allowance for inshore bathymetry
and weather conditions.
 The second preferred option is a zig-zag design.
 Other options should only be considered if conditions preclude the above.
 For rivers, the first preferred option is moving up one bank beaming horizontally to the far bank, returning
along the other bank. Ideally, the surveyed stretch should be between impounding structures such as locks
and weirs.
 When designing the cruise track, it is important to understand how the precision of the results depends upon the
transect spacing. The coefficient of variation (CV) of the abundance estimate depends upon the degree of
coverage [1], defined as:
Λ = D / A (1)
where
D is cruise track length;
A is the area being surveyed.
Then:
−0,5
CV = a(Λ) (2)
where
a is a variable between 0,4 and 0,8, depending on fish distribution. Higher values of a are appropriate when
fish are concentrated in a few large schools, low values when the fish are more uniformly distributed [37].
 The working time available for the collection of acoustic data should be calculated. For both water bodies, time
may have to be factored in for stationary measurements for specific purposes, e.g. aspect identification for
sizing Target Strength (TS) in horizontal surveys, hydrographic sampling etc.
 The survey plan (i.e. waypoints and transects) should be in a format suitable for transfer to GPS.
The selection of an appropriate vessel is important. This should be stable and low noise (with a preference for
-1
4-stroke over 2-stroke motors, or electric if feasible). Cruise speed should be a maximum of 10 km hour Actual
speed selection should be appropriate for the ping rate and water depth, aiming for a minimum of 3 hits on a fish of
interest when target-tracking / trace-counting.
6.4 Timing of surveys
The timing of acoustic surveys should consider the following factors:
 Surveys should be conducted during the period when the target fish are in open water and most dispersed.
 The following seasonal factors should be considered when planning an acoustic survey [39]:
 Recruitment patterns. Depending on the objective of the survey, under-yearlings may be deliberately
included or excluded.
 Beware of spawning time generally.
 Beware of winter aggregations.
 Beware of migrations (e.g. diadromous species).
 Beware of sources of acoustic interference (e.g. Chaoborus larvae, other invertebrates, fish larvae,
macrophytes, bubbles as a result of decreased hydrostatic pressure associated with draw-down, leaves,
increased noise during high flows, boat traffic etc.).
Note that optimal sampling periods may differ between countries and regions.
 Diel timing of acoustic surveys is also important:
 If no pre-existing information on fish distribution patterns is available, then carry out both day and night
surveys.
 For night surveys, avoid the full moon.
 Avoid transitional times (usually dawn and dusk). Restrict survey time from 1 h after sunset to 1 h before
sunrise.
 Night is usually best for surveys of both lakes and rivers.
 Surveys shall be conducted under homogeneous environmental conditions. If conditions change
significantly during the course of a survey, it should be abandoned.
6.5 Specific factors with respect to transducer orientation and position
In 6.3 and 6.4 factors, that are common to both vertical and horizontal surveys, are considered. There are also
factors that are specific to the survey mode, requiring different approaches to equipment deployment and operation.
In general, the preferred method for acoustic sampling is vertical beaming. This is due to a number of factors:
 The uncertainties and potential errors are much greater for horizontal data. For example, horizontal surveys
usually have lower signal to noise ratios, the aspect of fish to the transducer is often unknown, etc.
 Horizontal surveys are more susceptible to adverse weather conditions, particularly wind and heavy rain.
 Horizontal surveys are more susceptible to vessel instability.
 Horizontal surveys are more susceptible to acoustic interference from boat traffic.
Factors that are specific to vertical surveys include:
 The maximum pulse repetition rate shall be calculated according to the maximum depth sampled.
 The transducer depth should be as shallow as possible, but greater than depths that generate micro-bubbles.
 The transducer should be oriented as near as possible to the vertical.
Factors that are specific to horizontal surveys include:
 Transducers with a short nearfield and small side-lobes should be used in small, shallow rivers (width = 15 m,
depth = 2 m).
 The transducer should be on an adjustable mount allowing small changes in both vertical (tilt) and horizontal
(pan) planes.
 The transducer shall be at least one transducer face dimension below the surface of the water.
 The transducer shall be tilted in order to approach the Maximum Useable Range (MUR). This is a function of the
space between the surface and bottom boundaries and the beam shape [21].
 The transducer tilt and pan angles shall be optimised, recorded and maintained during the course of an acoustic
survey.
 The acoustic beam should be approximately perpendicular (or slightly forwards) relative to the cruise track.
 Care should be exercised selecting a ping-rate when beaming across substantial water widths towards a
distant, non-smooth shore. One approach is to measure reverberation levels with increasing ping-rate, the
optimal rate being just before reverberation levels sharply increase.
Combined vertical and horizontal surveys may be required on deeper lakes when a large proportion of the fish
population are distributed close to the water surface [20].
2)
6.6 Specific factors with respect to acoustic inter-comparisons
When inter-calibrating or comparing the outputs from different acoustic systems or survey teams, care must be taken
to ensure there is no acoustic interference between the test echosounders.
Trials shall be conducted prior to the investigation to test for significant cross-talk between the systems on the
survey vessel. In the absence of cross-talk, the echosounders can be operated simultaneously on the same boat.

2) EC Mandate M/424, (“Mandate for standardisation addressed to CEN for the development or improvement of standards in
support of the water framework directive” WFD, 2000/60/EC) received from the European Commission, DG Environment,
requires the validation of all standard methods according to ISO 5725. Collaborative studies based on reference materials or field
trials must be performed before the draft texts are transmitted to CEN in the form of prENs. In the case of the Hydroacoustics
standard, data from the inter-agency comparisons are awaited.
Where significant interference is detected, the second system shall be mounted on a separate boat. The second
vessel shall follow the first vessel along the same route, approximately 300 m apart [38]. The two vessels should
interchange the lead on alternate transects.
It is important to ensure there is no acoustic interference from other devices on the vessel (e.g. depth sounders,
invertors).
7 Survey data acquisition
7.1 Acoustic data
For both vertical and horizontal surveys, the following acoustic data are required in order to measure abundance:
 Intensity samples with as high range resolution and ping-rate as the equipment, data storage capacity and
environment will allow.
 The time of each ping.
 The start and end ranges of sample collection.
 Where possible, unthresholded raw data should be collected. Otherwise, the lowest possible threshold should
be used for data collection appropriate for normal survey conditions.
NOTE Echo Integration and Single Echo Detection (SED) / Single Target (ST) are regarded as post-processing techniques
and are described in Clause 8.
7.2 Echosounder settings
The following echosounder settings shall be adjusted and recorded in order to optimise the acoustic data collected:
 Ping-rate, based on maximum depth or range to be surveyed.
 The transmit power, pulse length and bandwidth should be set to give the best signal to noise ratio for the
appropriate range resolution. These parameters are directly linked.
 Speed of sound (c) and sound attenuation (α). Setting c and α require a priori knowledge of environmental
conditions (water temperature and salinity).
7.3 Data acquisition from additional equipment
Fish sampling is affected by physical and geographical factors. Therefore supplementary data should be collected
(Annex B, Table B.1). As a minimum, acoustic data should be supported by the following geographic and
environmental information as part of a survey log:
 Geographical position.
 Temperature (required for sound attenuation and speed of sound values in echosounder settings). For vertical
surveys, determining average water temperature from a temperature profile is recommended. For mobile
horizontal surveys, periodic fixed location temperature profile measurements should be taken in order to verify
normal spherical spreading of the acoustic beam.
 Salinity / conductivity (required for sound attenuation and speed of sound values in echosounder settings).
 Tilt angle of transducer. The use of an attitude or orientation sensor is recommended.
 Weather data, including prevailing wind strength and direction.
 Moon phase.
 Prevailing water currents.
 Lake surface area, mean and maximum depths.
A comprehensive survey log is required, recording all anomalies during the survey and any changes to system
settings.
All data shall be recorded in a consistent format. The recommended format is the HAC format of the ICES FAST
technical committee, however .RAW (HTI and Simrad), .SPL (HTI), .DG (Simrad) and .DT4 formats (BioSonics) are
also acceptable.
8 Post-processing of acoustic data
8.1 General
Post-processing of acoustic data consists of a pre-analysis phase and an analysis phase, and requires the use of
3)
specialist acoustic post-processing software (e.g. Echoview, Echoscape, Visual Analyzer and Sonar5-Pro
8.2 Pre-analysis
Pre-analysis data processing includes bottom detection, noise-cleaning by setting thresholds, and Single Echo /
Target Detection (SED or ST) selection.
8.2.1 Bottom detection
Bottom detection is an essential component of accurate acoustic abundance estimates:
In horizontal surveys, the ‘bottom’ (i.e. the opposite bank) is often of variable strength and range. The maximum
range should be defined as where the background noise exceeds a level 6 dB below the minimum target size.
Manually setting the bottom is often the most practical option.
In vertical surveys, the collection of samples from the bottom should be avoided as they usually represent a very
large acoustic biomass. Bottom tracking algorithms should be used for setting the bottom position, beyond which no
usable data are accepted. A bottom window or margin can also be used to decrease the probability of echo
integrating the bottom signal, however this will increase the ‘blind zone’ and fish close to the bottom will not be
recorded [37]. The margin should not exceed 0,5 m for most applications. When automatically applying the bottom
line within the software, the line should be visually inspected and corrected for any errors.
8.2.2 Discrimination
Discrimination is the separation of unwanted echoes such as air bubbles, debris, surface and bottom reverberation
and plankton from target echoes. In order to set target threshold levels, the noise-level at range shall be defined. As
modern digital echosounders enable all sample data to be collected, there is no need to filter the data during data
collection, and all filtering procedures can be conducted during data processing.

3) These are available from; Sonar5-Pro (Lindem Data Acquisition, Oslo, Norway), EchoView (Myriax, Tasmania, Australia),
EchoScape (HTI, Seattle, USA), Visual Analyser (BioSonics, Seattle, USA).
8.2.3 Separating noise reverberation and setting TS thresholds
Weak unwanted signals (noise echoes) are usually extremely numerous and can seriously bias abundance
estimates. Setting a noise threshold (i.e. a low threshold) is the most common way of drawing a clear line between
wanted and unwanted signals. In less frequent cases, a high threshold can be used where targets bigger than a
certain size are excluded. This can be useful, for example, when larvae are targets of interest and larger fish are
considered interference [7], [16].
The theory behind setting both low and high thresholds is similar. The investigator should know; (1) what size targets
require analysis, and (2) what is possible:
1) The investigator should decide on the size range of investigated fish in terms of length and Target Strength.
Minimum and maximum lengths of fish should be converted into Target Strength using an appropriate
regression based on fish species present and aspect to the transducer (examples of which are given in
Annex D). The resulting TS and TS provide the first guidance for setting an appropriate threshold.
min1 max1
2) The population of targets should be processed using a TS threshold 20 dB lower than TS in order to
min1
obtain the TS frequency distribution of targets (Figure 1) around the suggested TS . At this stage, SED /
min1
ST criteria can be kept on default values (see Single Echo / Target Detection Scenarios below). The
analysis threshold should then be set at a trough in the TS frequency distribution, thereby accepting or
rejecting a whole peak of small targets (PST; Figure 1, Thr. 1 or Thr. 2).
Setting a threshold that cuts through a peak in the TS frequency distribution is poor practice (Figure 1,
Thr. 3), as this would result in an undefined population of targets being analysed. For example, a small
change in the threshold or a small increase in fish length would result in a false increase in abundance.
The decision about accepting or rejecting a peak of small targets should comply with the aims of the survey
and/or the nature of PST assessed by direct biological sampling.
The threshold established by this protocol (TS ) should be displayed on a 40 log R TVG amplitude
min2
echogram in order to check whether the targets are thresholded beyond background noise levels (noise
signals often do not satisfy SED / ST criteria but can be recorded on amplitude echograms). If the targets
are not safe from the noise, the threshold setting procedure above and hence TS should be
min2
reconsidered.
Key
1 noise
2 larger fish
X TS (dB)
Y frequency
NOTE PST represents the peak of small targets, often young-of-the-year fish, which are usually very numerous in natural
populations
Figure 1 — Hypothetical example of unsuitable (Thr.3) and more suitable (Thr. 1 and 2) ways of setting the
noise threshold for fisheries surveys
NOTE 1 In poor signal-to-noise conditions, finding the natural minima of a TS distribution can be difficult. In such conditions,
the abundance estimates of small targets will not be reliable.
NOTE 2 Noise levels usually increase with the range, so it is possible to estimate small and larger fish at close ranges but only
larger fish at further ranges. If estimates of small fish are required, it can be necessary to lower the range. Larger fish can be
estimated in all ranges using the same (higher) threshold.
NOTE 3 If echo-integration is being used, TS has to be transferred to volume backscattering strength (S ) values. If
min2 v
processing 20 log R echogram is a priority, the S threshold should be adjusted until approximately the same population of targets
v
can then be seen on both 20 log R and 40 log R echograms. At longer ranges, the 20 log R threshold can be range dependent.
8.2.4 Separating noise echoes by manual classification
Some sources of noise cannot be removed with a threshold and other methods must be used to remove them from
the analysis. For example, reverberation from bubbles (e.g. boat wake or surface entrainment), transducer ring-down
and interference from other echosounders all fall into this category. These sources of noise must be manually or
automatically excluded from the analysis by setting exclusion or ‘bad data’ areas on the echogram.
8.2.5 Single Echo / Target Detection (SED / ST) scenarios
SED / ST criteria shall be selected according to the chosen abundance estimation method (see 9.2.1) and the signal
to noise ratio (SNR) defined below.
8.2.5.1 Accurate size distribution when SNR is high (> 10 dB)
Under high SNR conditions, high precision TS measurements are obtained from the central part of the beam, when
the phase is stable, there is good suppression of multiple echoes and a low degree of unwanted detections from
noise and background. This is normally obtained with SED / ST criteria set below:
 Echo Length detector = 0,7 times to 1,4 times the transmitted pulse length.
 Pulse length determination level = 6 dB.
4)
 If available, Maximum Phase Standard Deviation = 0,15
 Max Gain Compensation = 3 dB (one-way).
 Multi-peak suppression applied.
8.2.5.2 Track counting when SNR is medium (< 10 dB)
When track-counting (trace-tracking) in a medium SNR environment, wider echo selection criteria can be applied:
 Echo Length detector = 0,6 times to 1,6 times the transmitted pulse length.
 Pulse length determination level = 6 dB.
 If available, Maximum Phase Standard Deviation = 0,5.
 Max Gain Compensation = 6 dB (one-way).
 Multi-peak suppression should not be applied.
8.2.5.3 Track counting when SNR is low
If the SED or ST criteria for medium SNR fail (i.e. it generates insufficient detections in each size-class in question to
be statistically valid) and sample data are available, methods designed for low SNR conditions such as a cross filter
detector [4] or the track-before-detection method [13] can be used.
8.3 Analysis
8.3.1 Abundance estimate methods
Three different methods can be used to provide acoustic abundance and biomass estimates; S /TS scaling, track-
v
counting (trace-tracking) and echo-counting. Selection of the appropriate method for any given survey depends on
target density, quality of SEDs / STs, the signal to noise ratio and the quality of GPS data. A description of each
method and conditions under which each may be used is summarised in Annex C (Table C.1).
These methods can take four different sources for the involved size distribution; single echo detections, tracked fish
in-situ, tracked fish ex-situ and catch data. The following subclauses give advice on which source to select.
In-situ means that the size distribution will be built from echoes within the analysed region of the echogram. Ex-situ
means that the size distribution will be built from targets outside the analysed region.

4) Maximum Phase Standard Deviation is the standard deviation in mechanical samples measured in degrees. Different
echosounders operate with different definitions. In general, the manufacturer’s recommendations for a strict setting should be
applied.
8.3.2 Vertical surveys
The following analytical methods are recommended for vertical surveys:
8.3.2.1 Method 1 - Echo integration
In high SNR environments with high densities of targets, echo-integration (S /TS scaling) is recommended. The
v
backscattered energy should be scaled using the size distribution of tracked fish, as they provide more accurate TS
estimates. If densities are too high for target-tracking, use SEDs / STs, ensuring the TS data originates from the
layers of interest. If these densities are too high, SEDs / STs from other layers or other locations and periods should
be used assuming they apply to a given case (see 8.2.2 and Annex C, Table C.1).
8.3.2.2 Method 2 - Track-counting
Track-counting should be used in all SNR conditions with low target densities.
8.3.2.3 Method 3 - Echo-counting
In high SNR environments with a low density of targets and in the absence of GPS information, echo-counting can
be employed.
8.3.3 Horizontal Surveys
Horizontal surveys of rivers and lakes are generally conducted in low SNR environments. Track-counting is therefore
recommended under these conditions.
8.3.4 Biomass estimates
Conversion from numerical density estimates to biomass estimates requires information on the size distribution of
the targets of interest.
These can be established for acoustic data from:
1) Target Strength of tracked fish.
2) Target Strength of Single Echo Detections.
3) Catch (converted via appropriate TS - length and length - weight regressions).
8.3.5 Vertical surveys
For vertical surveys the recommended hierarchy of size distribution data sources are tracks, then SED / ST, then
catch.
8.3.6 Horizontal surve
...