Acoustic Discrimination Research Articles

Detection of Rainfall Events Using Underwater Passive Aquatic Sensors and Air–Sea Temperature Changes in the Tropical Pacific Ocean

Abstract Several years of long-term high temporal resolution ocean ambient noise data from the tropical Pacific Ocean are analyzed to detect oceanic rainfall. Ocean ambient noise generated by rainfall and wind are identified through an acoustic discrimination process. Once the spectra are classified, wind speed and rainfall rates are quantified using the empirical algorithms. Rainfall-rate time series have temporal resolutions of 1 min. These data provide a unique opportunity to study the rainfall events and patterns in two different climate regions, the intertropical convergence zone (ITCZ) of the tropical eastern Pacific (10° and 12°N, 95°W) and the equatorial western Pacific (0°, 165°E). At both locations the rain events have a mean rainfall of 15 mm h−1, but the events are longer in the eastern Pacific. After the rain event is defined, the probability that a rain event can be detected using the change in air–sea temperature often associated with the rainfall is investigated. The result shows that the rain event accompanied by the decrease of air temperature is a general feature, but that using the temperature difference to detect the rainfall has a very high false alarm rate, which makes it unsuitable for rainfall detection.

Does auditory discrimination training modify representations in both primary and secondary auditory cortex?

Several components of evoked auditory potentials and magnetic fields have been reported to be enhanced by musical training and by laboratory discrimination training. The P2 component, believed to originate in secondary auditory cortex (A2), is highly plastic, and enhanced both in trained musicians and by a variety of acoustic discrimination tasks. The 40-Hz auditory steady-state response (SSR) is believed to be generated in primary auditory cortex (A1). This response has been reported to be larger for trained musicians; however, in previous research SSR amplitude was not enhanced by laboratory training at pitch discrimination in non-musician adults. To explore whether competitive interactions among frequency representations in A1 may constrain remodeling, we designed a training procedure that provided subjects with experience of a single carrier frequency for 10 training sessions, alternating training and passive control blocks. P2 amplitude increased across sessions ( p < 0.01) on passive as well as training blocks. SSR amplitude was larger in training compared to passive blocks ( p < 0.01) and increased across sessions in both blocks ( p < 0.01), giving the first evidence of acoustic remodeling in A1.

Thematic accuracy assessment of acoustic seabed data for shallow benthic habitat mapping

A single beam acoustic ground discrimination system (AGDS) was used to survey 1.25 km2 of shallow (< 20 m depth) seabed on the northeast coast of Tasmania, Australia. This paper investigates the uncertainties associated with the qualitative interpretation of real time and post‐processed acoustic signal, and the effect of track spacing on the mapping of rocky reef distribution. The survey was repeated with different track spacing (50 m, 100 m, 200 m and 50 × 200 m shore normal) to investigate the influence of data density and ‘knowledge based interpolation’ validated against direct measurements made with an underwater video camera. Habitat area calculations varied substantially only with the 50 × 200 m transect. These results have important implications for the qualitative assessment and application of AGDS technology in shallow water marine habitat mapping.

Acoustic and phonetic discrimination in speech using event-related potentials: A mismatch negativity paradigm

The interaction of acoustic and phonetic discrimination in speech was studied using event-related potentials (ERPs). Participants consisted of 14 neurologically normal right-handed adults. Data were recorded in response to strong and weak acoustic changes signaling consonant distinctions, using a 64-channel encephalogram and a mismatch negativity (MMN) paradigm. The MMN is an automatic response evoked by stimulus change and is elicited when a train of repeated stimuli (standard) is interrupted by an oddball (deviant) stimulus. Stimuli consisted of one standard and two deviants that differed in their acoustic difference from the standard. All three stimuli were synthetic speech syllables in which F2 was manipulated to create a continuum between the syllables ‘‘ba’’ and ‘‘da.’’ The standard was the endpoint item ‘‘da.’’ The ‘‘strong deviant’’ was the ‘‘ba’’ endpoint item; the ‘‘weak deviant’’ was a more intermediate ‘‘ba’’ item, and therefore acoustically closer to the standard. MMNs were identified for both deviants, but the amplitude was smaller for the weak deviant condition. The data suggest that MMNs are sensitive to both phonetic change and also the magnitude of the acoustic difference within this change. The results are discussed with respect to theories of phonetic and acoustic discrimination in speech perception.

In a songbird, the black redstart, parents use acoustic cues to discriminate between their different fledglings

Several studies on parental investment in territorial songbirds have reported the existence of brood division. This is a type of postfledging care in which each parent has long-term feeding preferences for different young within a brood, creating two family units. Recent theoretical work indicates that conflicts between individuals should select for brood division. However, little is known about the mechanisms involved in the onset and maintenance of this behavioural strategy. Given the high rate of fledglings' begging calls, we hypothesized that acoustic discrimination could explain the stability of feeding preferences at a proximate level. In a 3-year field study, we recorded the responses of parent black redstarts, Phoenicurus ochruros, a territorial songbird, to playback of the begging calls of fledglings fed by the male and by the female. Parents responded more to the calls of the fledglings that they preferentially fed. A principal component analysis of the calls suggested that parents may recognize individual offspring. To our knowledge, this study provides the first evidence that a bird can acoustically discriminate between two categories of its own offspring: those that it preferentially feeds and those fed by the other parent.

Passive Acoustic Detection and Measurement of Rainfall at Sea

Abstract Rainfall over the ocean is one of the most important climatic parameters for both oceanic and atmospheric science. Traditional accumulation-type rain gauges are difficult to operate at sea, and so an alternate technique using underwater sound has been developed. The technique of passive monitoring of the ocean rainfall using ambient sound depends on the accuracy of sound pressure level (SPL) detection. Consequently, absolute calibration of the hydrophone is desirable, but is difficult to achieve because typically the geometry of the laboratory calibration process does not fit the measurement geometry over the ocean. However, if one assumes that the sound signal that is generated by wind is universal then the wind signal can be used to provide an absolute calibration. Over 90 buoy months of ambient sound spectra have been collected on the Tropical Atmosphere Ocean (TAO) project array since 1998. By applying the Vagle et al. wind speed algorithm, the instrument noises and sensitivity bias for the absolute calibration of each acoustic rain gauge (ARG) are obtained. An acoustic discrimination process is developed to retrieve the pure geophysical signals. A new single-frequency rainfall-rate algorithm is proposed after comparing the ARG data with R.M. Young self-siphoning rain gauge data, collocated on the same moorings. The acoustic discrimination process and the rainfall algorithm are further tested at two other locations and are compared with R.M. Young rain gauges and the Tropical Rain Measuring Mission (TRMM) product 3B42. The acoustic rainfall accumulations show the comparable results in both long (year) and short (hours) time scales.

Mapping seabed biotopes at two spatial scales in the eastern English Channel. Part 2. Comparison of two acoustic ground discrimination systems

Spatial surveys of marine benthic habitats and biota based on the interpretation of acoustic data were carried out at two sites in the eastern English Channel each representing different scales of geographic area and intensity of survey. A small area (4×12 km) crossing the Hastings Shingle Bank was surveyed at a relatively high intensity (track spacing 400 m) and was nested within a larger area between Hastings and Dungeness (12×40 km), which was surveyed at a lower intensity (track spacing 2 km). Surveys were conducted with two acoustic ground discrimination systems (AGDS), RoxAnn and QTC-VIEW and the primary purpose of the investigation was to compare the performance of the two AGDS using a common approach to analysis of the different data outputs (E1 and E2 for RoxAnn and the Q eigenvectors from QTC-VIEW). Exploratory data analysis using variography indicated that interpolation between tracks was justified for the smaller site to create a complete coverage, but was limited to the creation of a digital image of the track data for the larger area. Grab and video sample data were available for supervised classification of the AGDS data and interpreted sidescan images for comparison with unsupervised classification. Both AGDS gave similar outputs, although RoxAnn consistently gave slightly better levels of performance than QTC-VIEW as measured using error matrices. Although the investigation was not designed to compare the performance of AGDS and sidescan, the outputs from AGDS were similar to the visual interpretation of the sidescan sonar data. It was concluded that despite the inherent limitations of AGDS, they may be suitable for providing distribution maps at a broad scale that can give a context for the interpretation of finer scale survey of smaller, nested areas.

High-frequency acoustic volume scattering from zooplankton and moving oceanic microstructure

It is well accepted that high-frequency acoustic scattering techniques can be used to perform rapid, synoptic surveys of fish and zooplankton over relevant spatial and temporal scales. However, the use of these remote sensing techniques to probe small scale physical processes, such as oceanic microstructure, has not been fully accepted, or exploited. Yet there is a growing body of evidence in the form of both laboratory and field measurements suggesting that their use is feasible. Discrimination of scattering from microstructure versus zooplankton, which span similar spatial scales, is typically a limiting factor. Currently, acoustic discrimination of turbulence from zooplankton relies on either source of scattering being dominant as well as on the availability and accuracy of scattering models. A model for scattering from oceanic microstructure that includes fluctuations in the density and sound speed is presented. The effects on scattering from a layer of microstructure with a mean fluid velocity are discussed. Backscattering predictions are made based on data collected in Hudson Canyon with a tethered free-falling high-resolution vertical microstructure profiler, and compared to two-frequency acoustic data (120 and 420 kHz). The contribution to scattering from zooplankton is also estimated from nearby depth-resolved net tows together with zooplankton scattering models.

Potential acoustic discrimination within boreal fish assemblages

Abstract Differences in the acoustic characteristics of forage fish species in the Gulf of Alaska and the Bering Sea were examined using Kirchhoff ray-mode (KRM) backscatter models. Our goal was to identify species-specific characteristics and metrics that facilitate the discrimination of species using acoustic techniques. Five fish species were analyzed: capelin (Mallotus villosus), Pacific herring (Clupea pallasii), walleye pollock (Theragra chalcogramma), Atka mackerel (Pleurogrammus monopterygius), and eulachon (Thaleichthys pacificus). Backscatter amplitude differences exist among these species, especially between swimbladdered and non-swimbladdered fish. Echo intensities were variable within and among species. The effect of morphological variability was indexed using the ratio of the Reduced-scattering length (RSL) standard deviation over its mean. Morphological variability was low only at fish length to acoustic wavelength ratios less than eight. Target strength differences between pairs of carrier frequencies (ranging from 12 kHz to 200 kHz) differed among species, and were dependent on fish size and body orientation. Frequency differencing successfully discriminated between fish species but the choice of frequency to maximize target strength differences was not consistent among species pairs. Frequency-dependent, backscatter model predictions facilitate comparison of target strength differences prior to acoustic data collection.

Mapping marine benthic biotopes using acoustic ground discrimination systems

This paper describes the use and appraises the performance of acoustic ground discrimination systems (AGDS) for marine benthic biotope mapping for those who are more familiar with satellite or airborne remote sensing. The operation of AGDS is described and the nature of the data produced compared with more conventional remotely sensed data. Analysis of the AGDS data involves the interpolation of track point data to create areal coverages that can then be interpreted using classification and image processing techniques. AGDS data have been shown to be variable both within and between surveys and systems which affects the performance of interpretation. Additionally, the marine environment imposes limitations on the way that ground sampling can be undertaken and the spatial extent of information that can be obtained. In summary, the main issues that contribute to variability and uncertainty of biotope maps based on AGDS-based methodology are: (1) errors in position fixing of both AGDS and the ground samples; (2) inconsistent performance of the AGDS system; (3) single frequency of AGDS and its effect on discrimination; (4) tracking intensity of AGDS survey; (5) heterogeneity of the seabed at the scale of resolution of the system; (6) lack of vista for designating training sites and the reliance on buffers around ground samples; (7) unsuitable biotope categories. Variability due to AGDS hardware is often beyond the control of surveyors, but quality control procedures can at least alert operators to poor quality data. Performance of AGDS surveys also depends on intensity of both tracking and ground sampling in relation to heterogeneity of the sea floor. This relationship can be investigated through exploratory spatial data analysis, particularly the variogram. The performance of biotope maps can be measured using error and similarity matrices. The accuracy of the biotope maps resulting from AGDS surveys is modest when compared with the performance of satellite or airborne remote sensing and the similarity between maps derived from different datasets can be low. It is misleading to base expectations of the performance of AGDS on satellite or airborne remote sensing systems. AGDS are low resolution systems with limited powers for discrimination between different ground types. However, they afford the opportunity to map benthic biotopes in environments where electromagnetic spectra-based systems cannot operate. Thus, AGDS is a useful tool for benthic ecologists to study the spatial distribution of biotopes on the sea floor although the limitations of AGDS for mapping may preclude their use for precise and accurate survey required for some monitoring applications.

Using an Acoustic Ground Discrimination System to map coral reef benthic classes

An Acoustic Ground Discrimination System (AGDS) can extract information on the nature of the seabed. Compared to satellite or airborne sensors, AGDS is rarely used in tropical environments but is easy to operate and produces a modest amount of digital data. This study aimed to assess acoustic surveys of coral reef benthic classes using a RoxAnn @ AGDS in the Philippines. Benthic classes were categorised into four levels of detail using hierarchical classification of field data. Using independent data, subsequent maps were shown to have overall accuracies of: 85% at coarse resolution (three classes), 61% and 54% at two intermediate levels (five and four classes) and 28% at fine resolution (10 classes). These accuracies are assumed to be conservative because of constraints during this study including semi-quantitative data for discriminating class types, benthic changes between AGDS and accuracy assessment surveys and lack of differential global positioning system (GPS). Despite these limitations, AGDS accuracy levels were comparable to those achieved by satellites and it has advantages including greater water penetration and independence from cloud cover. Acoustic data are, therefore, proposed as useful tools for tropical habitat mapping and complementary to satellite and airborne sensors.

Selective habituation shapes acoustic predator recognition in harbour seals.

Predation is a major force in shaping the behaviour of animals, so that precise identification of predators will confer substantial selective advantages on animals that serve as food to others. Because experience with a predator can be lethal, early researchers studying birds suggested that predator recognition does not require learning. However, a predator image that can be modified by learning and experience will be advantageous in situations where cues associated with the predator are highly variable or change over time. In this study, we investigated the response of harbour seals (Phoca vitulina) to the underwater calls of different populations of killer whales (Orcinus orca). We found that the seals responded strongly to the calls of mammal-eating killer whales and unfamiliar fish-eating killer whales but not to the familiar calls of the local fish-eating population. This demonstrates that wild harbour seals are capable of complex acoustic discrimination and that they modify their predator image by selectively habituating to the calls of harmless killer whales. Fear in these animals is therefore focused on local threats by learning and experience.

Rigor and Representativeness in Marine Protected Area Design

Virtually all marine conservation planning and management models in place or proposed have in common the need for improved scientific rigor in identifying and characterizing the marine habitats and processes encompassed. An emerging central theme in the last few years has been the concept of representativeness, or representative systems of marine protected areas (MPAs). Representativeness and similar terms are used in two distinct senses in current literature, leading to considerable confusion. Definitions of these terms as they are currently used are provided. The habitat classification and mapping needed to incorporate considerations of representativeness into MPA planning must logically be carried out at the same scale at which management occurs. Management of highly protected areas is found to occur almost exclusively at local scales or finer, independent of the reservation model or philosophy employed. Current techniques for cost-effective surveying of marine areas are discussed. The emerging field of remote videography, in some cases linked with acoustic discrimination sensors, can provide quantitative data on real biological distributions. Such data lend themselves well to robust multivariate analysis in deriving habitat maps at scales relevant to managers, as a basis for assessing and planning for representation in MPA design.

Detection of acoustic class boundaries in soft sediment systems using the seafloor acoustic discrimination system QTC VIEW

Transition zones (or ‘ecotones’) are ecologically important spatial elements of subtidal landscapes that represent a mixture of different habitat elements. We provide a method for identifying such areas in broad-scale surveys of the seafloor using the acoustic discrimination system QTC VIEW. QTC VIEW is an acoustic processing system that assigns sonar ping stacks to clusters of like points, to identify different habitats on the seafloor. Paired QTC VIEW transects (∼20m spacing) were run at intervals of 200m, at two separate sampling locations, to assess the consistency of clustering of individual ping stacks into acoustic classes. Very consistent spatial patterns of class change were found between transect pairs, suggesting high stability in the classification process. QTC VIEW assigns confidence values to each individual record; running averages calculated using a moving window along transects showed drops in confidence values associated with areas of transition in habitat class assignment, but this was not always consistent. The Berger–Parker index, a class dominance statistic, provided a more consistent transition indicator. Class transition ranged from abrupt to gradual, along with areas where a mixture of acoustic classes occurred. However, acoustically detected transition zones did not consistently respond to visual observations of the sea floor.

Acoustic discrimination of two zooplankton species (mysid) at 38 and 120 kHz

Using bifrequency sonar (38 kHz, 120 kHz) directed towards the open sea from the inside of a submarine cave, two species of Mysidacea (Crustacea), similar in shape but not in size, and with different patterns of behaviour, were observed simultaneously and discriminated, and their biomass was estimated during their migration. The first species, Hemimysis speluncola (<1 cm length), lives in the submarine cave, forming a distinct community that migrates horizontally. The second species, Siriella jaltensis (>1 cm length), lives in the open sea and migrates vertically. These species are both detected clearly at 120 kHz, but only the larger one, Siriella jaltensis, is detected at 38 kHz. This distinction allows one to determine the size limits of planktonic objects detected at these frequencies, i.e. 3 mm at 120 kHz and 10 mm at 38 kHz, corresponding to one quarter of the wavelength. The target strength (TS) of each species was calculated from the field data and compared to results calculated from three models. As the volume of the cave occupied by the Hemimysis population is about 300 m 3, and the density of the swarm is 36,000 individuals/m 3, the number of individuals in the cave is estimated to be around 10 million (or 24 kg). The density of the open-water population of Siriella is estimated to be 13 individuals/m 3.

A REVIEW OF FREQUENCY AND TIME PARAMETERS OF CETACEAN TONAL CALLS

ABSTRACT Properties of the tonal calls of cetaceans are summarised and compared at the species level. Statistics are presented relating to start, end, minimum, maximum and centre frequencies, duration and number of inflections, together with information about recordings (location, number encounters/groups, length of recordings). Evidence of a linear relation between body-size and wavelength is given for odontocetes (R2 between 0.68 and 0.93, depending on frequency variable and least-squares method) but the relationship does not appear to fit well for mysticetes (R2 = 0.64). Results of preliminary investigations into acoustic discrimination of species by multivariate methods are also presented: a simple classifier for 10 species based on results derived from the literature gave a correct classification rate of 28% when tested with independent data. Some methodological recommendations for future descriptive work on cetacean acoustics are made.

Neurophysiological and psychophysical measures of duration discrimination in normal-hearing adults and adults with cochlear implants

The ability to discriminate duration contrasts (a first approximation of voice onset time differences) was evaluated in adult cochlear implant users and normal-hearing adults using a neurophysiological response (the mismatch negativity—MMN). The MMN is an evoked response generated by a deviant stimulus when embedded in a series of standard homogeneous stimuli. Magnitude of the MMN correlates with psychophysical discrimination thresholds indicating that the MMN provides a measure of perceived acoustic change [Kraus et al. (1996)]. The MMN thus provides a neurophysiological correlate of acoustic discrimination. The present study compared an objective statistical measure of MMN magnitude to psychophysical measures in response to duration differences of brief auditory stimuli. A ten-click train was paired with shorter trains to create an oddball sequence. The selected duration differences were similar to voice onset time differences that distinguish voiced from voiceless speech phonemes. The MMN magnitude was measured at four scalp-electrode locations. The relationship between neurophysiological and psychophysical thresholds was determined and comparisons were made between the normal-hearing adults and adults with cochlear implants. While some similarities were found, the results suggest that the sensitivity to duration differences differs between normal-hearing and cochlear implant subjects.

After-effects of human-computer interaction indicated by P300 of the event-related brain potential

After-effects of human-computer interaction (HCI) were investigated by using the P300 component of the event-related brain potential (ERP). Forty-nine subjects (naive non-users, beginners, experienced users, programmers) completed three paper/pencil tasks (text editing, solving intelligence test items, filling out a questionnaire on sensation seeking) and three HCI tasks (text editing, executing a tutor program or programming, playing Tetris). The sequence of 7-min tasks was randomized between subjects and balanced between groups. After each experimental condition ERPs were recorded during an acoustic discrimination task at F3, F4, Cz, P3 and P4. Data indicate that: (1) mental after-effects of HCI can be detected by P300 of the ERP; (2) HCI showed in general a reduced amplitude; (3) P300 amplitude varied also with type of task, mainly at F4 where it was smaller after cognitive tasks (intelligence test/programming) and larger after emotion-based tasks (sensation seeking/Tetris); (4) cognitive tasks showed shorter latencies; (5) latencies were widely location-independent (within the range of 356 - 358 ms at F3, F4, P3 and P4) after executing the tutor program or programming; and (6) all observed after-effects were independent of the user's experience in operating computers and may therefore reflect short-term aftereffects only and no structural changes of information processing caused by HCI.

Acoustic discrimination of passerine anti-predator signals by Australian raptors

Acoustic discrimination of anti-predator calls was examined in 11 species of Australian raptors, including 5 Falco species and 2 species of Elanus kites, by their responses to the playback of alarm and distress calls of Australian passerines. The present study investigated the ability of raptors to discriminate between alarm and distress calls that have different acoustic properties and are emitted in different behavioural contexts. The raptors were tested with broad-band calls (containing a wide range of frequencies) given as distress calls, mobbing calls and alarm calls to terrestrial predators, and with narrow-band calls (comprising a narrow range of frequencies) typically given as a response to flying predators. Raptor responses were categorised into three classes based on head orientation (or lack thereof) towards the sound source (i.e. one of 2 or 4 speakers positioned in the cage set-up); (1) ‘correct response’ – the raptor looked directly at the speaker; (2) ‘incorrect response’ – the raptor detected the sound, but oriented the head in a direction other than towards the sound source; (3) ‘no response’. All raptor species showed a higher percentage of correct responses (60–100%) for broad-frequency vocalisations and a lower percentage of correct responses (usually 0–40%) and more incorrect responses for narrow-band vocalisations. Further, all raptors showed a greater rate of overall responsiveness to broad-band alarm and distress calls than narrow-band calls, indicating a higher interest level in the former. The behavioural implications of acoustic discrimination by Australian raptors to different types of alarm call are discussed.

Marine Biological Mapping for Environment Management Using Acoustic Ground Discrimination Systems and Geographic Information Systems

Marine Biological Mapping for Environment Management Using Acoustic Ground Discrimination Systems and Geographic Information Systems