The "dB differencing" or "delta Sv" technique for classifying echo traces
Contents: Background information | Literature and other resources | Echoview templates and documentation | Hydroacoustics User Forum topic
Related links: About procedures | About techniques | Sharing techniques | Disclaimer
Keywords: Classification; single-beam echosounder; echo trace; multi-frequency acoustics; remote species identification; frequency response; dB differencing; delta Sv
Background information
Frequency response
The same object can provide a different echo return at different frequencies; this is known as its frequency response. For example, an individual fish might have a target strength (TS) of -68 dB when insonified at 38 kHz and a TS of -72 dB at 120 kHz.
A frequency response arises because the TS of a given object is related not only to its size, shape, orientation and material properties, but also to the frequency of sound with which it is insonified (e.g. Demer & Martin 1995).
A common acoustic data-processing technique that exploits the frequency response in order to classify species is known as "dB differencing" or "delta Sv".
Using the frequency response to aid classification
While the TS of an individual object relates to its size, shape, orientation and material properties, on its own this is typically not enough to define what that object is.
For example, at 38 kHz a 30 cm-long non-swimbladdered fish might have the same TS as a 10 cm-long swimbladdered fish; on its own, TS would not enable to you discern the difference.
However, at 120 kHz the non-swimbladdered fish might have a higher TS than at 38 kHz, while the swimbladdered fish might exhibit no difference in TS across these two frequencies.
By looking at the frequency response, a point of difference is found between the acoustic measurements of these two different objects.
Relating the frequency response to the identity of an object
Once the frequency response of a given object or aggregation has been measured, clues to its identity can be derived from theory.
For example, Korneliussen & Ona (2003) present a series of theoretical sound-scattering models for different types of marine organism (see their Fig. 1 below) that show how differences in their frequency responses can be used to hint at their identity.
A general schematic description of the relative frequency response, r(f). Bands indicate typical positions of selected acoustic categories when measured at frequencies 18-200 kHz (ref. Figure 1 of Korneliussen & Ona 2003)
Taking swimbladdered fish and fluid-like zooplankton as an example, their frequency-response curves are very different; the fish show similar backscatter returns across multiple frequencies, while the zooplankton returns increase with increasing frequency. So for a given echo trace, if you were to observe a higher backscattering value at 120 kHz than at 38 kHz you might classify this trace as zooplankton rather than fish.
An example of the frequency response in practice: studies of Antarctic krill
Madureira et al. (1993a, 1993b) found that the mean volume backscattering strength (Sv) of Antarctic krill (Euphausia superba) was between 2 and 16 dB higher at 120 kHz than at 38 kHz.
Watkins & Brierley (2002) used this algorithm to confirm that the identification of krill in this way matched the visual scrutiny of echograms by experienced operators, thereby heralding a move from subjective to objective classification techniques that can be performed by any analyst regardless of experience.
While the 2-16 dB "window" for krill has and will continue to undergo refinement on the basis of increasingly comprehensive, empirically-validated scattering models (e.g. Demer & Conti 2003), the dB-differencing technique has been used effectively in a range of acoustic studies of this species (e.g. Pauly et al. 2000; Hewitt et al. 2004; Jarvis et al. 2010).
Literature and other resources
- Demer DA and Conti SG (2003) Reconciling theoretical versus empirical target strengths of krill: effects of phase variability on the distorted-wave Born approximation. ICES Journal of Marine Science, 60: 429-434
- Demer DA and Martin LV (1995) Zooplankton target strength: volumetric or areal dependence? Journal of the Acoustical Society of America, 98: 1111-1118
- Hewitt RP and 16 others (2004) Biomass of Antarctic krill in the Scotia Sea in January/February 2000 and its use in revising an estimate of precautionary yield. Deep-Sea Research II, 51: 1215-1236
- Jarvis T, Kelly N, Kawaguchi S, van Wijk E and Nicol S (2010) Acoustic characterisation of the broad-scale distribution and abundance of Antarctic krill (Euphausia superba) off East Antarctica (30-80E) in January-March 2006. Deep-Sea Research II, 57: 916-933
- Kang M, Furusawa M and Miyashita K (2002) Effective and accurate use of difference in mean volume backscattering strength to identify fish and plankton. ICES Journal of Marine Science, 59: 794-804
- Korneliussen RJ and Ona E (2003) Synthetic echograms generated from the relative frequency response. ICES Journal of Marine Science, 60: 636-640
- Madureira LSP, Everson I and Murphy EJ (1993a) Interpretation of acoustic data at two frequencies to discriminate between Antarctic krill (Euphausia superba Dana) and other scatterers. Journal of Plankton Research, 15: 787-802
- Madureira LSP, Ward P and Atkinson A (1993b) Differences in backscattering strength determined at 120 and 38 kHz for three species of Antarctic macroplankton. Marine Ecology Progress Series, 93: 17-24
- Pauly T, Nicol S, Higginbottom I, Hosie G and Kitchener J (2000) Distribution and abundance of Antarctic krill (Euphausia superba) off East Antarctica (80-150E) during the Austral summer of 1995/1996. Deep-Sea Research II, 47: 2465-2488
- Watkins JL and Brierley AS (2002) Verification of the acoustic techniques used to identify Antarctic krill. ICES Journal of Marine Science, 59: 1326-1336
Echoview templates and documentation
Please read about the Echoview-template naming convention here.
Further information about each template can be found in the various notes sections of each Echoview file under View > EV File Properties: Notes and (for each variable) Object > Variable Properties: Notes.
Template 1 - Myriax Software 
Description
Two aggregations are separated into distinct classes on the basis of their frequency response at 38 and 120 kHz. This template uses the data presented in Chapter 5 of the Echoview Technical Manual.
View the data flow for this template here.
Downloads
Echoview template version 4.90.48 (25 KB, 31st May 2010)
Data (9.8 MB, 31st May 2010)
