Bradley Paul Lipovsky

Distributed acoustic sensing (DAS) is an increasingly popular tool for marine mammal monitoring. Until recently, most monitoring efforts targeted lower frequency sounds made by large baleen whales and were not optimized for the mid-frequency signals made by many other marine mammals. This is partially due to the directivity of the DAS system and the spatial averaging effect of the gauge length, which acts as a low-pass filter, markedly decreasing signal amplitude above 300–500 Hz, depending on the angle of incidence. Adding loops in the cable can mitigate this by integrating the strain around the circumference of the loop to improve sensitivity at higher frequencies. This study deployed a custom fiber optic cable with incorporated loops in Haro Strait (Salish Sea, WA) and recorded signals up to the Nyquist frequency of 2.5 kHz in looped sections, compared to approximately 500 Hz in straight sections. The study successfully recorded mid-frequency (0.5–2.5 kHz) vocalizations from Southern Resident killer whales, Bigg's killer whales, and humpback whales, and stacking channels within loops improved the signal-to-noise ratio by approximately 8 dB on average. These results show that a looped DAS array can supplement existing hydrophone networks or serve as a standalone tool when a large hydrophone array is not feasible.

Seafloor fiber-optic cables could close major observational gaps in ocean basins. A new multispan distributed acoustic sensing (DAS) technique permits strain-rate measurements at ∼10 to 100 m spacing across the entirety of these cables, beyond repeaters. We present the first publicly available multispan DAS dataset from a seafloor observatory. From November 2025 to January 2026, multispan DAS data were collected offshore central Oregon on both cables of the Ocean Observatories Initiative Regional Cabled Array (OOI-RCA) simultaneously with conventional (single-span) DAS data while the observatory remained online. Nonlinear interactions with observatory signals distorted data on some cable sections, but usable multispan DAS data were obtained past three repeaters (> 300 km optical distance). We report coherent recordings of depth-dependent ocean wavefields out to 2900 m water depth, including surface gravity, infragravity, and Scholte waves, with behavior consistent with theory and prior OOI-RCA DAS observations. Background noise characteristics are largely consistent between multispan and conventional DAS techniques. At least 12 earthquakes are recorded with clear P and S waves on both multispan DAS and nearby cabled ocean-bottom seismometers. Our analysis shows that this dataset is suitable for various oceanographic, seismic, and potentially tsunami monitoring applications.

Distributed Acoustic Sensing (DAS) has become a widely used passive acoustic monitoring method but faces challenges recording above a few hundred hertz. The goal of this study was to improve detection and signal-to-noise ratio (SNR) of higher-frequency signals, which could expand monitoring of fish and marine mammals. This study tested a cable with loops and a cable spiraled around a fiberglass tube deployed vertically from a vessel in Puget Sound, Washington, at a depth of approximately 100 m. The systems were tested with varying acquisition parameters, using impulsive signals from a bubble pulser and a 50 Hz to 1.5 kHz sequence from a Lubell VC2C transducer, with hydrophones as references. SNR was estimated in the time-frequency domain to account for unequally frequency-distributed noise. Both configurations improved SNR compared to straight cable sections and the looped configuration recorded up to the Nyquist frequency. Orientation and distance of the cable relative to the sources likely influenced signal detection, but constraints of this study prevented a direct evaluation of incident-angle-dependent sensitivity. The study shows the capability of vertical looped and spiral DAS systems to detect waterborne acoustic signals up to 3 kHz and highlights the importance of experimental design on data quality.

The Cook Inlet segment of the Alaska-Aleutian subduction zone experiences damaging megathrust earthquakes, recurrent slow-slip events, and localized tectonic tremor. We deployed a multicomponent seismic array in the Cook Inlet vicinity to investigate the structure and geophysical properties at and near the subduction interface and understand the factors that control slip, tremor, and plate locking. This deployment combines 11 onshore broadband seismographs and a distributed acoustic sensing (DAS) campaign on offshore fiber-optic cables. Both datasets record a complementary variety of local seismic signals from earthquakes and environmental sources. The broadband network is sensitive to the teleseismic earthquake waves that can constrain deeper subduction zone structure. Noise levels are high, in part because of the basin environment and to a lesser extent because of our decision to use accessible indoor sites for deployment. The DAS data contain rich information about local noise sources, in particular, oceanographic processes within the Cook Inlet that affect land- and sea-based seismic recordings. In addition, this combined dataset provides novel perspectives of a variety of tectonic, environmental, and oceanographic signals and bridges the gap between offshore environmental forcing and onshore seismic response.

Ice cores from the Allan Hills blue ice area in Antarctica have been dated to several million years of age. However, the stratigraphy of these cores is often disturbed, and age reversals are common, hampering the interpretation of the derived climate records. To better understand the physical processes affecting the ice, we here use a variety of microstructural methods to investigate the fabric, texture and bubble characteristics of four depth regimes in the ALHIC1901 core. We find single maximum and girdle CPOs, the latter often contain two distinct maxima. Further, stripes of differently oriented crystals are common. We further observe small, potentially elongated, grains containing indications of dynamic recrystallization. 2D and 3D bubble data show strongly elongated bubbles and very low bubble number densities compared to similar depths of deep ice cores. We interpret these findings as indications of simple shear, including kink and z‐folds at the centimeter scale and potential alterations to the original air bubbles and, thus, their gas content. Extended studies of Allan Hills ice cores using continuous sections are needed to evaluate the spatial scale of the folding and its impact on the climate record.

The accumulation of decades of continuous seismic observations, combined with the emergence of new sensing technologies (e.g., distributed acoustic sensing-DAS- and nodes) and novel computing infrastructure, presents both outstanding challenges and opportunities for the observational geophysical community to tackle data processing at petabytes scale. Methodologies, open-source software practices, and cyberinfrastructure have advanced to a point where mining petabyte-scale archives can be done within a single day of cloud computation (Ni et al, 2025a,b). This contribution reviews research workflows centered around seismic event monitoring with large-scale seismometers and regional DAS networks. We evaluate strategies for both cloud infrastructure (Ni et al, 2025) and edge DAS units (Shi et al, 2025a,b) for a seismic event monitoring pipeline that leverages deep learning for rapid feature extraction, such as classification of seismic source type and picking of P- and S-wave arrivals. Leveraging and advocating for open-source software, we discuss computational considerations and strategies to improve the performance of pre-trained deep learning models through transfer-learning and model architecture adaptation. We illustrate these findings with the United States NSF-National Geophysics Facility archive operated by the EarthScope Consortium, as well as diverse experiments from the University of Washington FiberLab.

High-sensitivity borehole pressure sensors provide a novel means of monitoring the surface mass balance (SMB) of Earth’s great ice sheets. When properly coupled to the ice, these instruments are expected to detect millimeter-scale changes in ice thickness corresponding to variations in snow accumulation. The Boussinesq solution guarantees that a pressure sensor buried at a depth of 1000 m will represent a spatially weighted average of surface mass load variations with an approximate radius of 1000 m at the surface. The sensing system itself consists of a pressure sensor connected to a pressure coupling element such as a sack filled with antifreeze. Once freeze-in related transients have passed, the internal fluid pressure in the sack reflects the full overburden stress exerted by the overlying ice column. Long term drift of the sensors requires special engineering attention and may be corrected for using some combination of pre-deployment pressurization, in situ calibration, and/or the use of multiple in situ sensors. We discuss deployment considerations using the IceDiver thermal melt probe for a potential deployment at Greenland Summit Station. The borehole SMB observatory described here supports altimetry-based inference of glacier mass change, provides a critical dataset for validating firn densification models, and establishes a fundamentally new approach to measuring glacier mass balance.

Long-term submarine observations are critical for understanding subduction zones because the slip of great earthquakes occurs offshore. Geophysical observations suggest that the Cascadia megathrust is locked from the coastline to the deformation front in many places, but off central Oregon they are consistent with a narrowly locked megathrust near the deformation front and creeping behavior beneath the shelf where there are two clusters of earthquakes near the plate boundary, including repeating and very low frequency earthquakes. In this region, scientific objectives include understanding how megathrust locking transitions between the deformation front and the coastline, determining whether there is transient slip behavior, improving constraints on how shallow offshore earthquake clusters are linked to the megathrust, and characterizing the baseline deformation rate and fault slip behavior of the accretionary prism. This summer, the Cascadia Offshore Subduction Zone Observatory (COSZO), an infrastructure project funded by the US National Science Foundation, will add seismic and geodetic instruments to the Ocean Observatories Initiative (OOI) Regional Cabled Array (RCA) off Newport, Oregon. New seafloor science junction boxes, with updates to the RCA design, will be connected to three primary nodes on the continental slope and shelf that currently do not support seafloor geophysical observations. At each new junction box and a fourth site on the shelf where there is an existing science junction box but no geophysical instruments, COSZO will install a Nanometrics Atlantis Cabled Observatory ocean bottom seismic package comprising a buried broadband seismometer, a strong-motion accelerometer, a low-frequency hydrophone, and a differential pressure gauge. The project incorporates two types of calibrated absolute pressure gauges that utilize Paroscientific resonant quartz crystal sensors. The Geodetic and Seismic Sensor Module combines a triaxial accelerometer with two pressure gauges that are periodically calibrated against the internal pressure of the housing measured by a barometer. The Self-Calibrating Pressure Recorder also includes two pressure gauges but performs calibrations with a reference pressure close to ambient generated by a piston gauge. COSZO will also install uncalibrated absolute pressure gauges and Nortek Vector 3-component ocean current meters. Together with sensors already on the OOI RCA at the Slope Base and Hydrate Ridge sites and autonomous long-term geodetic observations, the COSZO infrastructure will form a critical mass observatory on the Cascadia Subduction Zone to support scientific studies and efforts to prototype offshore earthquake and tsunami early warning. COSZO will stream data into EarthScope Data Services and a workshop is planned for spring 2027 to engage early career scientists. Looking forward, each science junction box includes open ports and any unspent COSZO funds and independent PI-driven proposals can add to the suite of cabled instruments. The OOI RCA has also hosted three short fiber sensing experiments, demonstrating the potential for single- and multi-span distributed acoustic sensing concurrent with observatory operations. Implementing permanent fiber sensing on the OOI RCA would complement COSZO by adding additional observations over an expanded footprint.

A recently developed multispan distributed acoustic sensing (multispan-DAS) technique from Nokia Bell Labs enables strain measurements along submarine fiber-optic cables across multiple repeater-separated spans. By leveraging the high-loss loopback couplers within optical repeaters, this technique overcomes the long-standing limitation of conventional DAS to the first span of a repeated cable, typically < 100 km offshore. Dense, continuous arrays of seafloor strain sensors can now extend to hundreds or thousands of kilometers. This technique has been used to successfully record the 2025 M8.8 Kamchatka earthquake and tsunami at teleseismic range with a spatial resolution of ~100 m across 4400 km of a repeated submarine cable.In November 2025, the multispan-DAS system from Nokia Bell Labs was deployed for three months on both repeated submarine cables of the Ocean Observatories Initiative Regional Cabled Array (OOI RCA) offshore Oregon. The deployment traverses the Cascadia subduction zone forearc and extends approximately 500 km offshore to Axial Seamount. During this period, the first span of the southern cable was simultaneously interrogated using a multiplexed conventional DAS unit, while data continued to stream from co-located cabled seismometers, hydrophones, and other oceanographic instruments on the OOI RCA.The multispan-DAS system recorded a regional earthquake beyond the first repeater of both cables during testing as well as the ambient seafloor seismic wavefield, demonstrating sensitivity to a broad range of seismic, oceanographic, and acoustic signals. These observations provide a unique opportunity to directly compare multispan-DAS measurements with conventional DAS and established seafloor instrumentation across a large spatial extent. The resulting dataset will be publicly released following documentation and quality control. We will present preliminary results characterizing the noise floor, sensitivity, and signal fidelity of multispan-DAS relative to co-located sensors, and examine the consistency of observed seismic and oceanographic signals across measurement modalities. These results will highlight the potential of multispan-DAS for applications including routine earthquake monitoring, earthquake early warning, and broader seafloor observation, and represent an important step toward establishing this technique as a new tool for the seismological and oceanographic communities.

Monitoring glacier dynamics is essential for understanding climate change impacts, safeguarding water resources, and protecting communities from related hazard. Distributed acoustic sensing (DAS) provides a unique opportunity to observe these dynamic environments with high spatial and temporal resolution. However, creating comprehensive seismic event catalogs from DAS data requires the development of efficient, automated tools. In this study, we analyzed 17.6 TB of DAS data collected over 31 d from a 9 km fiber-optic cable deployed on Rhonegletscher, Switzerland. The cable recorded strain-rate data along 2225 channels, spanning both snow-covered and bare-ice regions. We developed a robust preprocessing pipeline to address challenges posed by noise, coupling inconsistencies, and large data volumes. Our feature extraction is based on covariance matrix analysis, which allowed us to characterize wavefield properties using the first eigenvalue, coherency function, and eigenvalue variance. We compared unsupervised and supervised approaches, evaluating their relative effectiveness in detecting cryoseismic events across noisy DAS datasets. Although unsupervised methods provided valuable insights into inherent data patterns, their performance was hindered by dataset imbalance and noise. In contrast, supervised methods trained on manually labeled data demonstrated higher classification accuracy and reliability, with random forest emerging as the top performer, achieving 81% cross-validation accuracy, 93% test accuracy using a random split, and 73% test accuracy when evaluated with a grouped split to avoid data leakage. By combining cloud and parallel computing, we develop a scalable framework that streamlines DAS data analysis and supports future operational monitoring and research on glacier dynamics. The resulting comprehensive catalog, comprising thousands of detected and classified cryoseismic events, provides a valuable resource for scientists, fostering advancements in cryosphere monitoring and hazard assessment within the context of a changing climate.