Megan S. Ballard

Abstract Warm, saline Atlantic waters and fresher Arctic‐origin waters converge in the central Nordic Seas, creating strong mesoscale and submesoscale variability that influences both hydrography and sound propagation. During the Northern Ocean Rapid Surface Evolution 2022 experiment, high‐resolution temperature and salinity measurements were collected using shipboard profiling and drifting Wirewalker systems, alongside acoustic transmissions in the 500–1,500 Hz band. Water‐mass structure was characterized using two‐dimensional temperature–salinity histograms, which revealed distinct surface and intermediate water types and their spatial and temporal evolution. Two processes dominated variability: topographically trapped internal tides over the East Jan Mayen Ridge and a propagating anticyclonic eddy composed of modified Atlantic water. Both features produced measurable shifts in the depth and thickness of the intermediate Atlantic water layer, which forms a regional sound channel. These hydrographic changes led to predictable modulation of waterborne acoustic arrival patterns and, in the case of the eddy, downward refraction strong enough to eliminate the waterborne path through bathymetric blocking. The results demonstrate how evolving water‐mass structure in high‐latitude frontal systems directly governs mid‐frequency acoustic propagation, with implications for acoustic observing, environmental prediction, and interpretation of variability in Arctic‐influenced basins.

Climate change is altering environmental conditions that govern acoustic propagation across oceanic, atmospheric, and terrestrial systems, while also creating new opportunities for acoustic sensing of these changes. This joint Special Issue of the Journal of the Acoustical Society of America and JASA Express Letters highlights recent advances at the intersection of acoustics and climate science, including studies of ocean acoustic tomography, marine and terrestrial soundscapes, coastal ecosystems such as coral reefs and seagrass meadows, and the impacts of anthropogenic noise. Together, these contributions demonstrate how acoustic methods can provide unique insights into environmental variability, ecosystem response, and long-term climate processes across a range of spatial and temporal scales.

Abstract Noise from calving icebergs, cracking ice, and melting ice dominates the underwater soundscape of glacierized fjords creating one of the loudest recorded ambient ocean environments. While progress has been made toward identifying and describing individual sound sources—including the automatic detection of calving and quantification of ice‐mass loss—the relative contributions of multiple, simultaneous processes, and how these contributions evolve over time, remain underexplored, limiting robust interpretation of ice‐ocean interactions. Here, we show that unsupervised machine learning separates a series of recordings captured over 8 months into five dominant sound profiles related to glacier activity. We deployed an array of hydrophones approximately 400 m from the terminus of Xeitl Sít’ (LeConte Glacier) in Southeast Alaska and recorded sound regularly between October 2016 and May 2017. Using the k‐means clustering algorithm, we cluster spectral shapes of 10,440 background acoustic spectra, defined as the 25th‐percentile spectral level of each recording. We identify five distinct acoustic clusters and relate their temporal occurrence to environmental time series including ice movement, meteorology, and oceanographic data. We further link spectral shapes to known glacier sources such as calving and ice melt. Our analysis reveals that these clusters correspond more closely with glacier and ice‐mélange activity than with other environmental variables, confirming the dominance of glacier behavior on fjord soundscapes. This research demonstrates the effectiveness of clustering passive acoustic data and provides a framework for analyzing large, complex acoustic data sets of undersampled environments—such as glacierized fjords—to guide interpretation and track changes in dominant environmental processes.

Studies of acoustic propagation in seagrass meadow environments have demonstrated a high degree of sensitivity of passive and active acoustics to diurnal photosynthetic cycles and seasonal growth patterns. While empirical metrics based on these measurements in combination with traditional ecological monitoring methods provide insight into trends in primary productivity, estimates of seagrass biomass and other ecological condition indicators have not yet been calculated directly from acoustic measurements. Towards this goal, an acoustic inference was developed utilizing a Bayesian framework to infer estimates of seagrass biomass. This work was based on active acoustic measurements collected over the course of a 25-month remote monitoring experiment conducted in a seagrass meadow in a shallow-water sub-tropical estuary. A ray-based acoustic propagation model was developed to incorporate effects of the seagrass leaf canopy and model parameter estimates related to the height and leaf density of the canopy are used to predict above-ground seagrass biomass over the course of the experiment. The study found biomass to be consistent with seasonal trends found in previous measurements. Comparison of acoustic estimates with an 11-month record of direct measurements of biomass shows good agreement, demonstrating the potential of acoustic inversion to facilitate acoustic-based ecosystem monitoring in seagrass meadow environments.

INTRODUCTION Nasal chondromesenchymal hamartoma (NCMH) is a rare benign tumor of the nasal cavity and paranasal sinuses. Its clinical relevance has increased following the discovery of its association with DICER1 syndrome-a genetic condition predisposing individuals to multiple neoplasms. These two cases highlight important clinical extremes of this disease entity and offer practical evaluation and management recommendations for practicing otolaryngologists. METHODS A retrospective review was conducted from 1999 to 2025 at a single institution. The report includes two female teenagers presented at ages 13 and 17 years, respectively, diagnosed with NCMH-one with and one without DICER1 syndrome. Key outcomes included recurrence rates, anatomical spread, genetic testing results (including evaluation for DICER1 variants), and disease-status at follow-up. A thorough literature review on NCMH and DICER1 syndrome was conducted to provide the reader with management recommendations. RESULTS CONCLUSIONS: Genetic testing for DICER1 variants should be standard in all newly diagnosed NCMH cases. Annual nasal endoscopic and imaging follow up is recommended in all individuals with a constitutional or mosaic pathogenic DICER1 variant. There is no published guidance for recurrent NCMH at the olfactory groove/cribriform plate region. Conservative local excisions should be pursued for tumor recurrences.

This paper reports on measurements of out-of-plane arrivals from an airgun transect over the north slope of the Atlantis II Seamount Complex in the North Atlantic. The data were collected by a set of three near-bottom single-channel acoustic recorders. Time delay analysis is applied to assess the directionality of anomalous arrivals in the received time series that are interwoven between the in-plane bottom-surface and multiple bottom-surface reflected paths. Back propagation using estimated arrival angles in a three-dimensional ray-tracing model facilitates identification of bathymetric features associated with the arrivals. The responsible features were most commonly identified as local maxima that appear as steep-sided spires rising several hundred meters above the seafloor on the slope and near the base of Atlantis II. Arrival times for the predicted propagation paths are recreated by combining two-dimensional ray traces from the source to the bathymetric feature and from the bathymetric feature to the receiver to obtain modeled arrival times for comparison with observations. Multiple examples are analyzed, providing insight into likely contributors to the complex propagation in seamount environments. Forward modeling of the identified propagation paths using three-dimensional ray traces highlights challenges for ray-based acoustic modeling of propagation in the presence of highly variable bathymetry.

In wind-wave driven ambient sound environments, the cross-correlation of acoustic data recorded on vertically separated hydrophones can provide estimates of the average sound speed between hydrophones. Deployment of a 52-element vertical line array of hydrophones located on a 425 m deep ridge in the Nordic waters near Jan Mayen Island enabled estimates of the water column sound speed profile. Sound speed profiles were estimated for each 24-min recording, which were collected every four hours over the course of the year-long experiment, supplying a wide range of environmental conditions. Over the full experiment, estimated sound speeds had a root-median-square error of 0.79 m/s when compared to direct measurements.

The oceanographic environment surrounding the New England Seamounts frequently involves the complex interaction of the colder Slope Sea waters with the warmer Sargasso Sea waters, where a meandering frontal boundary is caused by the Gulf Stream. The sound speed profiles and associated available acoustic propagation paths differ significantly on either side of the Gulf Stream. Regional historical temperature profile data gathered from the Argo float program are used to investigate common acoustic trends associated with the region. A clustering algorithm organizes the dataset by the three dominant water temperature profile types, allowing for the direct investigation of Slope Sea, Gulf Stream, and Sargasso Sea environments. Cycle distance and propagation path type are employed as acoustic metrics to assess common depth-dependent features. Comparison of the metrics calculated for Argo data with those for the closest Hybrid Coordinate Ocean Model predictions indicate the existence of spatiotemporal features that are present in the observational data but not in the numerical ocean model predictions. Small biases in modeled temperature fields can result in the absence of common acoustic propagation features found in the data.

In the ice-free Arctic waters near the island of Jan Mayen (Nordic Seas), wind and wave-generated sound is used to estimate the sound speed between vertically separated hydrophones. The experimental data were recorded on a 52-element evenly spaced vertical line array moored on Jan Mayen Ridge, an undersea volcanic ridge with a depth of approximately 420 m at the array location. The coherence between vertically separated hydrophone pairs, sampled at a frequency of 8096 Hz and averaged over the 24-min recording period, provides sufficient resolution in the cross-correlation shape to determine features related to the acoustic travel time between hydrophones. The impact of hydrophone-pair spacing on sound speed estimates is explored, from the smallest separation distance of 7.25 m, determined by array design, to distances exceeding 100 m in high-wind conditions. The sound speed profile was estimated every 4 h over the yearlong experiment, and the fluctuations compared well with measurements made from thermistors mounted along the array. The method is robust, enabling accurate estimates (RMSE < 1 m/s) in a wide variety of environmental conditions over the course of the experiment, including low wind conditions, flow noise, and array tilt.

This paper presents measurements collected on a vertical line array (VLA) of hydrophones moored near Jan Mayen Island, an area where warm, salty Atlantic waters mix with fresher polar waters from the Greenland Basin. The dataset compares signals simultaneously collected along two propagation paths: a moored source located 42 km north of the array and a drifting source that originated near the array and then moved northeast. Water-column properties were measured extensively using sensors on the moorings, on the buoy of the drifting source, and on ship deployed instruments. The water-column sound-speed profile included a warm surface mixed layer and a subsurface acoustic duct with an axis near 200 m. Both the moored and drifting sources were located at a depth of 100 m, and signals from both sources presented a ducted arrival measured on the VLA. While the sound-speed conditions between the moored source and VLA were more quiescent, the drifting source became trapped in an eddy with a deep mixed layer, intensifying the downward refraction of sound, and lengthening the ray cycle distance. As the source opened to a range of 43 km, the ducted path became bathymetrically blocked, and the signal could no longer be detected.

Seagrasses provide a multitude of ecosystem services and act as important carbon sinks. However, seagrass habitats are declining globally, and they are among the most threatened ecosystems on earth. For these reasons, long-term and continuous measurements of seagrass parameters are of primary importance for ecosystem health assessment and sustainable management. This talk will present results from both active and passive acoustical methods for ecosystem monitoring in seagrass meadows. Examples of both techniques will be presented based on data collected as part of a two-year continuous deployment of an acoustical measurement system operating in a seagrass bed dominated by Thalassia testudinum (turtle grass) in Corpus Christi Bay, Texas. From a propagation perspective using a broadband acoustic source, gas bodies contained within the seagrass tissue as well as photosynthetic-driven bubble production results in attenuation and scattering of sound that produces increased transmission loss. For the passive approach, the detachment of gas bubbles from the plants is an important component of the ambient soundscape. The data show annual trends related to the seasonal growth pattern of Thalassia as well as diurnal trends correlated with photosynthetically active radiation. [Work supported by ONR, NSF, ARPA-E.]

Out-of-plane propagation resulting from acoustic interactions with abruptly varying bathymetry is often difficult to predict and can make the disentanglement of noise and reverberation from multipath propagation challenging. As a case study of these effects, an airgun survey conducted during the New England Seamounts Acoustics Experiment is analyzed for the presence of out-of-plane arrivals associated with prominent bathymetric features of the Atlantis II Seamount Complex. Time delay analysis is applied to a set of three near-bottom acoustic recorders to assess the directionality of distinct arrivals in the received time series. Back propagation using three-dimensional (3-D) ray tracing facilitates the identification of bathymetric features associated with the arrivals. Predicted propagation paths are recreated by combining two-dimensional ray traces from the source to the bathymetric feature and from the bathymetric feature to the receiver to obtain modeled arrival times for comparison with observations. Multiple examples of out-of-plane reflections from specific bathymetric features on the slope and near the base of the seamount are identified, providing insight into likely contributors to the complex propagation in seamount environments. Forward modeling of the identified propagation paths using 3-D ray traces highlights challenges for ray-based acoustic modeling of propagation in the presence of highly variable bathymetry. [Work supported by ONR.]

Noise from calving icebergs, cracking ice, and melting ice dominate the underwater soundscape of glacierized fjords. The combination of these sources results in one of the loudest recorded ambient ocean environments. Until now, the combined symphony of sounds has made identifying specific sources difficult, limiting its value to provide insights into the physical processes influencing ice-ocean interactions. Here, we show that unsupervised machine learning can separate the signal into five dominant sound profiles related to glacier activity. For this work, we deployed an array of hydrophones approximately 400 m from the terminus of Xeitl Sít’ (LeConte Glacier) in Southeast Alaska and recorded sound at regular intervals between October 2016 and May 2017. Using the k-means clustering algorithm, we cluster spectral shapes of 10,440 background acoustic spectra, defined as the 25th-percentile spectral level of each recording. We identify five distinct acoustic clusters and associate their temporal occurrence with environmental time series including ice movement, local meteorological conditions, and oceanographic data. We further associate the spectral shape and audio signals to known glacier sources such as calving and ice melt. Our analysis reveals that these acoustic clusters correspond more closely with glacier and ice-mélange activity than other environmental variables, confirming the dominance of glacier behavior on fjord soundscapes. This research demonstrates the straightforward application and effectiveness of clustering passive acoustic data and sets a foundation for using soundscapes to monitor and detect environmental changes in glacierized fjords.

Benthic biological processes influence seabed heterogeneity and contribute to variability in geoacoustic properties. To investigate these relationships, measurements were conducted to quantify spatial variability in the upper few decimeters of sediment near the water-seabed interface within a fine-grained sediment deposit on the New England continental shelf. At each measurement location, an acoustic multicorer was deployed to sample the seabed. Acoustic probes were inserted into the sediment to collect direct in situ measurements of sediment compressional wave speed and attenuation (30-100 kHz) under near-ambient conditions, after which cores were collected from the inter-probe propagation paths. Sediment physical properties, organic carbon, infaunal community composition, and ex situ compressional wave speed and attenuation spanning two frequency decades (104-106 Hz) were subsequently measured in the laboratory. The frequency dependence of sound speed ratio and attenuation was analyzed in the context of sediment acoustics models for mud based on the viscous grain shearing and extended Biot models. Sites with greater abundance of larger-bodied infauna (>1 mm) displayed higher variability in sound speed and attenuation. Correlation was found between sediment compressional wave modulus and total organic carbon, suggesting that organic matter in the sediment matrix also affects bulk acoustic properties.

Oxygen ebullition is typically ignored in long-term measurements of net ecosystem productivity (NEP), which reflects whether systems are net heterotrophic or autotrophic based on diel changes in oxygen. The solubility of oxygen in seawater is a function of temperature, salinity, and pressure. Warm, high-salinity seawater has low oxygen solubility, and when combined with the photosynthetic productivity of macrophytes in shallow, clear waters, oxygen ebullition frequently occurs. The presence of mixed-phase oxygen in supersaturated seawater creates difficulties for oxygen measurements because sensors cannot measure dissolved and gas phases simultaneously. Therefore, dissolved oxygen measurements must be taken in conjunction with separate ebullition measurements to develop an accurate oxygen budget needed to characterize NEP. Here, we seek to understand the drivers of NEP for a Thalassia testudinum meadow in the Gulf of Mexico and calculate ebullition rates using acoustic sensing. We hypothesize that oxygen ebullition will substantially increase NEP, particularly during the summer months because of photosynthetic dependence on temperature and irradiance. We found that the seagrass meadow is typically oxygen saturated during the hours 0700 to 2200 with maximum saturation occurring at 1400. Acoustic-based methods and general additive modeling both found highest ebullition rates in July/August. High respiration led to an apparent net heterotrophic system (NEP = −2.1 mmol O 2 m −2 d −1 ), but accounting for air–sea exchange and ebullition causes the system to be autotrophic (NEP = 54 mmol O 2 m −2 d −1 ). This study demonstrates the importance of including ebullition into NEP calculations and the viability of acoustics as a tool for monitoring aquatic productivity.

Abstract Wind, wave, and acoustic observations are used to test a scaling for ambient sound levels in the ocean that is based on wind speed and the degree of surface wave development (at a given wind speed). The focus of this study is acoustic frequencies in the range 1–20 kHz, for which sound is generated by the bubbles injected during surface wave breaking. Traditionally, ambient sound spectra in this frequency range are scaled by wind speed alone. In this study, we investigate a secondary dependence on surface wave development. For any given wind‐speed, ambient sound levels are separated into conditions in which waves are 1) actively developing or 2) fully developed. Wave development is quantified using the non‐dimensional wave height, a metric commonly used to analyze fetch or duration limitations in wave growth. This simple metric is applicable in both coastal and open ocean environments. Use of the wave development metric to scale sound spectra is first motivated with observations from a brief case study near the island of Jan Mayen (Norwegian Sea), then robustly tested with long time‐series observations of winds and waves at Ocean Station Papa (North Pacific Ocean). When waves are actively developing, ambient sound levels are elevated 2–3 dB across the 1–20 kHz frequency range. This result is discussed in the context of sound generation during wave breaking and sound attenuation by persistent bubble layers.

Seagrasses provide a multitude of ecosystem services and act as important carbon sinks. However, seagrass habitats are declining globally, and they are among the most threatened ecosystems on earth. For these reasons, long-term and continuous measurements of seagrass parameters are of primary importance for ecosystem health assessment and sustainable management. This paper presents results from both active and passive acoustical methods for ecosystem monitoring in seagrass meadows. From a propagation perspective, gas bodies contained within the seagrass tissue as well as photosynthetic-driven bubble production result in attenuation, dispersion, and scattering of sound that produce increased transmission loss. For the passive approach, the detachment of gas bubbles from the plants is an important component of the ambient soundscape. Examples of both techniques will be presented based on data collected as part of a two-year continuous deployment of an acoustical measurement system operating in a moderately dense seagrass bed dominated by Thalassia testudinum (turtle grass) in Corpus Christi Bay, Texas. The data show annual trends related to the seasonal growth pattern of Thalassia as well as diurnal trends correlated with photosynthetically active radiation.

This paper reports on an original set of direct sound speed measurements collected with the acoustic coring system in the New England Mud Patch (NEMP) and shelf break area to the south. Cores collected within the NEMP show range-dependence of the mud with slower sound speed and lower attenuation on the west side. In the shelf break region, the highest sound speeds are observed between the 200- and 350-m isobaths. The depth-dependence of the mud layer in the NEMP includes a surficial layer with a negative sound speed gradient of 28 s-1. The remainder of the mud column has a weak positive sound speed gradient of 6.2 s-1 over an isovelocity layer. Comparison between in situ and ex situ sound speed measurements provides an assessment of the effects of sediment disturbance from gravity coring operations. Small differences in the upper 2.5 m were attributed to the changes in the geoacoustic properties caused by disturbance from the coring process. Below 2.5 m, the average difference is close to zero, suggesting that these sediments were minimally disturbed. Finally, an in situ measurement of shear speed was obtained near the depth of maximum penetration. The shear speed was well correlated with sound speed from approximately the same depth interval.

The acoustical oceanography (AO) curriculum at The University of Texas at Austin (UT) consists a number of courses and thesis research, which often includes field work. The core course is EE/ME 384N-5, Underwater Acoustics, which covers acoustic properties of the ocean, propagation, reflection, reverberation, scattering and target strength, ocean noise, array and signal processing and basic sonar design. The course is offered in alternate years and is cross-listed as both an electrical and mechanical engineering course. Prior to this, students usually take two semesters of physical acoustics: EE/ME 384N-1 and 2, Acoustics I and II, which covers plane waves in fluids, transient and steady-state reflection and transmission, lumped elements, refraction, ray acoustics, absorption and dispersion, spherical and cylindrical waves, radiation and scattering, multipole expansions, Green’s functions, waveguides, Fourier acoustics, and Kirchhoff theory of diffraction. Both are offered every year. Another course commonly taken by AO students is EE/ME 384N-3: Electromechanical Transducers, which covers basic modeling, analysis and design of acoustics and vibration transducers, including calibration. Recent student thesis topics have included marine acoustic ecology, the investigation of methane seeps, acoustic seagrass monitoring, and the assessment of glacial processes. Appropriate courses in UT’s natural and earth sciences departments supplement the acoustics courses.

This paper presents inversion results for three datasets collected on three spatially separated mud depocenters (hereafter called mud ponds) during the 2022 Seabed Characterization Experiment (SBCEX). The data considered here represent modal time-frequency (TF) dispersion as estimated from a single hydrophone. Inversion is performed using a trans-dimensional (trans-D) Bayesian inference method that jointly estimates water-column and seabed properties along with associated uncertainties. This enables successful estimation of the seafloor properties, consistent with in situ acoustic core measurements, even when the water column is dynamical and mostly unknown. A quantitative analysis is performed to (1) compare results with previous modal TF trans-D studies for one mud pond but under different oceanographic condition, and (2) inter-compare the new SBCEX22 results for the three mud ponds. Overall, the estimated mud geoacoustic properties show no significant temporal variability. Further, no significant spatial variability is found between two of the mud ponds while the estimated geoacoustic properties of the third are different. Two hypotheses, considered to be equally likely, are explored to explain this apparent spatial variability: it may be the result of actual differences in the mud properties, or the mud properties may be similar but the inversion results are driven by difference in data information content.

Commercially available underwater acoustic recorders have become commonplace tools in ocean-acoustics research, due to their ease of deployment, compact size, and relatively low cost. The size of these systems typically results in a configuration with the hydrophone in close proximity to the electronics housing, flotation devices, and other equipment that can degrade the generally assumed omnidirectional response of the hydrophone. The mid-frequency acoustic regime (0.5–10 kHz) is particularly affected due to the similarity between the length-scales of these structures and the corresponding acoustic wavelengths. Calibration measurements made at an open-water test facility with a calibrated source/receiver pair characterized the frequency-dependent receiver directivity for three underwater recording devices with different housings and hydrophones: the PVC air-filled Loggerhead Snap, PVC oil-filled SoundTrap ST300, and titanium air-filled SoundTrap ST600. Furthermore, directivity measurements of a TOSSIT mooring [Zitterbart et al., HardwareX (2022)] were taken with the SoundTrap ST300 and SoundTrap ST600. Results suggest the frequency-dependent acoustic directivity of the recorders should not be neglected. In particular, the Loggerhead Snap had variations in receive level of over 20 dB as a function of receiver orientation angle and frequency, introducing a bias that obscures spectral levels of the in situ environment.

Both physical and biological benthic processes can influence seabed heterogeneity and contribute to spatiotemporal variability in geoacoustic properties. In particular, how biological processes affect both sediment acoustic properties and their variability is poorly understood. To address this deficiency, recent measurements investigated spatial variability in the upper few decimeters of sediment near the water-seabed interface within a fine-grained sediment deposit on the New England shelf. At each measurement location, acoustic probes were inserted into the sediment to collect direct in situ measurements of sediment sound speed and attenuation at near-ambient conditions, after which cores were collected from the inter-probe propagation paths for ex situ analysis of sediment physical, biological, and acoustic properties. Relationships among sediment properties, such as bulk density, porosity, grain size distribution, organic matter composition, infaunal community composition, and acoustic measurements spanning several frequency decades (10–1000 kHz) will be explored in this paper. Frequency dependence of sediment acoustic properties will also be discussed in the context of sediment acoustics models for mud based on the viscous grain shearing and extended Biot theories. [Sponsored by ONR.]

In sandy sediments, the salinity of the pore water only affects the sound speed of the mixture by way of affecting the sound speed of the pore water. With clay mineral and saltwater mixtures, the intergranular forces that depend on the salinity of the pore water affect the sediment acoustic properties beyond simply affecting the sound speed of the pore water. These dependencies were investigated via image analyses of prepared samples of two kaolinite clay types (Flat DS and RSA) via analyses of their settling dynamics, porosity, and rigidity as a function of pore water salinity. There exists an observable dependence on salinity to mixture porosity and rate of settling. In addition, three regimes of slip-stick dynamics are shown to exist at different salinity ranges: (a) an immediate transition to liquid-like behavior; (b) a viscoelastic transition to liquid-like behavior with long term creep; and (c) a delayed transition to viscoelastic behavior with long term creep. [Work supported by ONR.]

Wind, wave, and acoustic observations are used to test a scaling for ambient sound levels in the ocean that is based on the relative penetration depth of active bubbles during surface wave breaking. The focus is on acoustic frequencies in the range 1-10 kHz, which are typically scaled by wind speed alone. Wind and wave information are combined in a parametric form to describe the depth of the active bubble layer (which produces sound) relative to the depth of the passive bubble layer (which attenuates sound). The relative depth scaling has a primary dependence on wind speed and a secondary dependence on any departure of significant wave height from fully-developed, open-ocean conditions. The scaling is tested with long time-series observations of winds and waves at Ocean Station Papa (North Pacific Ocean), as well as with a case study with fetch limitation near the island of Jan Mayen (Norwegian Sea). When waves are less developed (e.g., limited by fetch) at a given wind speed, the attenuating layer is relatively thin and the sound levels are higher. The scaling is a plausible explanation for the observed reduction in sound levels during high wind events (winds greater than 15 m/s).

This paper reports on an observation of three-dimensional (3D) arrivals for which the change in the direction of horizontally refracted sound is nearly 180°. The experimental site is Jan Mayen Channel (JMCh), which connects the Greenland and Norwegian Seas. During the experiment, signals from a moored source transmitting a 500-1500 Hz sweep every 4 h were recorded by three surface drifters equipped with hydrophone arrays. Over a 3-day period, the drifters moved north across JMCh toward the moored source. In each recording, an in-plane arrival is identified. In a subset of these recordings, a second arrival is observed, having travel time consistent with propagation from the moored source, turning at the ridge on the south side of the channel, and arriving at the drifters. In a smaller subset of recordings, a third arrival is also observed having travel time consistent with a turning point on the face of the bathymetric rise on the west end of the channel that forms the Jan Mayen volcano. A 3D ray trace is employed to show the change in direction results from repeated reflections from the seafloor such that it is classified as horizontal refraction and not a single-bounce reflection.

This paper explores the challenging polyphonic sound event detection problem using machine learning architectures applied to data recorded in the Beaufort Sea during the Canada Basin Acoustic Propagation Experiment. Four candidate architectures were investigated and evaluated on nine classes of signals broadcast from moored sources that were recorded on a vertical line array of hydrophones over the course of the yearlong experiment. These signals represent a high degree of variability with respect to time-frequency characteristics, changes in signal-to-noise ratio (SNR) associated with varying signal levels as well as fluctuating ambient sound levels, and variable distributions, which resulted in class imbalances. Within this context, binary relevance, which decomposes the multi-label learning task into a number of independent binary learning tasks, was examined as an alternative to the conventional multi-label classification (MLC) approach. Binary relevance has several advantages, including flexible, lightweight model configurations that support faster model inference. In the experiments presented, binary relevance outperformed conventional MLC approach on classes with the most imbalance and lowest SNR. A deeper investigation of model performance as a function of SNR showed that binary relevance significantly improved recall within the low SNR range for all classes studied.

The acoustic coring system (ACS) is a probe-equipped gravity corer that provides in situ measurements of compressional and shear wave speed and attenuation. During an April 2022 coring survey, the ACS was deployed at 36 locations within the New England Mud Patch and New England Shelf Break areas. Data from these measurements will be presented to characterize the depth-dependent structure of the geoacoustic seabed properties as well as their spatial variability. The in situ measurements will be interpreted in the context of stratigraphic layering measured by a seismic survey. Historically, in situ shear speed measurements of the seabed have proved difficult to obtain. From the 2022 survey, depth-dependent (up to 4 meters) shear speed profiles at 800 and 1200 Hz were produced from a subset of the 36 deployments. These results, along with methods for analysis, challenges in the measurements, and supplemental laboratory experiments will be discussed. The results will also be compared to measurements from previous experiments. [Work sponsored by ONR.]

The 2023 New England Sea Mounts Acoustics (NESMA) experiment took place from April to June at the Atlantis II Seamount Complex. One goal NESMA seeks to address is to further the understanding of acoustic propagation around prominent bathymetric features and associated diffraction and refraction effects. The bathymetry of the experimental site exhibits extreme variability with a difference in elevation of three kilometers from the abyssal plane to the peaks of the seamounts and slopes upwards of 15°, which has a significant influence on bottom-interacting acoustic propagation. As part of the experiment, multiple broadband impulsive sound sources were deployed around the seamounts, including explosive charges, seismic air-gun, and rupture induced cavitation events. Three-dimensional ray trace models of the experiment are presented for comparison with the measured acoustic receptions. Model results are primarily concerned with signals received on three single-channel acoustic recorders deployed on the abyssal plane near the seamounts. Comparisons provide insight into variation in the structure of the measured arrivals associated with varying ray paths refracted around and between the seamounts for different source positions. The influence of geotechnical properties on the measured signals is also assessed through modeling. [Work supported by ONR.]

In situ measurements of geoacoustic properties provide direct characterization of the seabed at near ambient conditions. The Acoustic Coring System (ACS) is a gravity corer equipped with acoustic probes that obtain in-situ compressional wave (30–200 kHz) and shear wave (400–1200 Hz) measurements as the corer penetrates the seabed. During the April 2022 R/V Endeavor coring survey, the ACS was deployed at 36 locations within the New England Mud Patch and New England Shelf Break areas. Data from these measurements will be presented to characterize the depth-dependent structure of the geoacoustic seabed properties as well as their spatial variability. The in-situ measurements will be interpreted in the context of stratigraphic layering measured by a seismic survey. Depth-dependentprofiles of compressional speed from a subset of these deployments in the NEMP will be compared to profiles previously collected at nearby locations in 2016. In situ compressional wave records from both areas will be compared with ex-situ sediment core measurements, including data collected from core loggers and laboratory analyses. Finally, preliminary shear speed results will be discussed. [Work sponsored by ONR.]

In recent years, detailed geological characterization of the seafloor and shallow sub-surface on the northern U.S. Atlantic margin have been an important component of multi-disciplinary ocean acoustics and environmental sensing projects to support experiment planning and inform geoacoustic models of sound propagation. Comprehensive surficial sediment and core sampling across shallow continental shelf, shelf-edge, and upper continental slope sites offshore southern New England provide the opportunity to evaluate the importance of many sediment characteristics that impact sound propagation through the seabed. Beginning in the mid-shelf New England Mud Patch and continuing into deeper water, we have begun to develop a detailed understanding of the composition and physical properties of the shallow sediment column record using a combination of surficial and sub-surface sediment sampling tools, coupled with an extensive suite of discrete laboratory analyses. Quantitative data on sediment density, porosity, mineral and biogenic composition, and grain size distribution provides both valuable input parameters for geoacoustic models and helps establish a framework for interpreting independently derived measurements and modeling results. The physical, biological, and other environmental sediment characteristics can be integrated with existing geological assessments of the region and used to investigate ocean and sub-surface acoustic propagation. [Work supported by the Office of Naval Research.]

Measurements of surficial sediment properties at five locations in the New England Mud Patch (NEMP) were conducted during a coring survey in April 2022. At each location the Acoustic-Multi-Corer (AMC) sampled the upper few decimeters of sediment near the water-seabed interface. The AMC inserted acoustic probes into the sediment to collect in situ measurements of compressional waves (30–140 kHz) and shear waves (0.8–1.2 kHz) at near ambient conditions, after which cores were collected from the inter-probe propagation paths for ex situ analysis of the sediment properties. Onboard the ship, the cores were scanned with a Core and Resonance Logger (CARL), which provided high-frequency sound speed and attenuation measurements (100–1000 kHz) along the lengths of the cores, as well as resonance-based measurements of sound speed at 12–16 kHz. A subset of cores was sieved for infauna onboard the vessel, and the remaining cores were later analyzed in an onshore laboratory for porosity, bulk density, mineral and organic matter composition, and grain size. The frequency dependence of the surficial sediment acoustic properties will be discussed in the context of the mud’s physical and biological properties as well as sediment acoustics models. [Work sponsored by ONR.]

Seagrasses serve major ecological roles in biodiversity promotion, coastal protection, and nutrient cycling. Furthermore, seagrasses have been proposed as a nature-based solution to mitigate effects of climate change due to their capacity for sequestering marine carbon. Current global estimates of seagrass coverage are uncertain; therefore, developing improved methods to assess seagrass coverage and rates of decline are critical to promote sustainable seagrass conservation efforts. Acoustic propagation in seagrass meadows is highly sensitive to oxygen bubble production via photosynthesis and gas volumes encapsulated within seagrass tissue, both acting as biophysical markers. This paper discusses an acoustic method to monitor seagrass oxygen production and biomass with high temporal resolution and over long time-scales. An 18-month acoustic propagation experiment was conducted in a seagrass meadow located in a shallow bay on the Texas Gulf of Mexico Coast. A piezoelectric sound source transmitted broadband frequency-modulated chirps (0.5–100 kHz) every ten minutes, and the signal was measured on horizontal hydrophone array. Dissolved oxygen, photosynthetically active radiation, water temperature, salinity, and depth were concurrently measured with oceanographic probes. Additionally, cores were collected for point-estimates of seagrass biomass. Our work demonstrates that acoustic propagation offers a valuable alternative to experimental measurements of photosynthesis. [Work sponsored by NSF.]

Seagrasses are foundation species in many coastal ecosystems, but these environments are declining globally due to climate change and other anthropogenic impacts. Ballard et al. [J. Acoust. Soc. Am. 147, 2020] established the efficacy of acoustic remote sensing techniques for seagrass monitoring by exploiting acoustic sensitivity to gas bubbles produced by photosynthesis and gas channels within the seagrass leaves. However, the effects of seagrass on acoustic propagation are not understood with sufficient quantitative detail, and an improved model describing propagation through a mixture of seagrass leaves, free gas bubbles, and seawater is needed to aid in integrating acoustic methods into conservation efforts. This talk provides an overview of developments in the modeling of acoustic propagation through a Thalassia testudinum meadow using a homogeneous effective medium approach to represent the seagrass leaves and seawater. The model accounts for the complex microstructure of seagrass leaves including the encapsulated gas channels and the elastic properties of the seagrass tissue. The model is intended for use in geoacoustic inference algorithms for the overall goal of providing estimates of seagrass productivity and biomass. Candidate effective medium models include 2D cylindrical seagrass leaves and a micromechanical model of a seagrass leaf cross-section. [Work supported by NSF.]

The top meter of marine sediments is estimated to store a total of 2322 Pg of carbon [Atwood et al., 2020. Front. Mar. Sci., vol. 7 p.165], which is twice that of terrestrial soils. Physical disturbance and remineralization of these carbon stocks could further accelerate climate change. Rapid and accurate quantification of sediment carbon stocks can (1) better inform carbon budgets and management of human activities in the ocean to minimize carbon remineralization and (2) monitor changes in sediment carbon stocks due to anthropogenic and natural disturbances. Sediment acoustics, which have been linked with sediment total organic carbon/content in mud banks, seagrass beds, and estuarine environments, can be measured in situ and offer a scalable solution toward rapid estimation of carbon stocks. However, before this modality can be applied broadly, a fundamental understanding between sediment acoustic properties and sediment organic/inorganic constituents must be realized. Here, direct ex situ measurements of sediment acoustic properties, sediment organic carbon, and grain size distribution will be compiled and compared across a variety of sediment types and locations from our own measurements and from published datasets. The development of a non-site-specific constitutive-based relationship will be discussed. [Work supported in part by ONR.]

Seagrasses provide a multitude of ecosystem services and act as important carbon sinks. However, seagrass habitats are declining globally, and they are among the most threatened ecosystems on earth. For these reasons, long-term and continuous measurements of seagrass parameters are of primary importance for ecosystem health assessment and sustainable management. After a brief historical overview, this talk will present results from both active and passive acoustical methods for ecosystem monitoring in seagrass meadows. From a propagation perspective, gas bodies contained within the seagrass tissue as well as photosynthetic-driven bubble production results in attenuation and scattering of sound that produces increased transmission loss. For the passive approach, the detachment of gas bubbles from the plants is an important component of the ambient soundscape. Examples of both techniques will be presented based on data collected as part of an 18-month continuous deployment of an acoustical measurement system operating in a moderately dense seagrass bed dominated by Thalassia testudinum (turtle grass) in Corpus Christi Bay, Texas. The data show annual trends related to the seasonal growth pattern of Thalassia as well as diurnal trends correlated with photosynthetically active radiation. [Work supported by NSF.]