The iReverb project explores technical methodologies employed in the UPFLOW project, which addresses the challenges associated with tidal-induced noise on Ocean Bottom Seismometers (OBS) deployed in the North Atlantic Ocean. OBS sensors, unlike those on land stations, are exposed to oceanic currents, whose coupling with the instruments induces reverberations in the recordings. This project showcases a method for exploration of reverberations of tidally-modulated current-induced noise across different OBS types around the Azores, Madeira and Canaries region. The overarching objective is to present a comprehensive technical framework to enhance our understanding of Ocean Bottom Circulation (OBC), providing insights for calibrating current models. A central aspect of this project is the proposed utilisation of machine learning (ML) algorithms for automating the mapping of resonances and obtaining a proxy for OBC.

Here, we present the iReverb methodology, involving manual curation of a pixel-level annotated spectrogram dataset, after this, a ML classifier is trained to identify features in each spectrogram (known as supervised semantic segmentation). To begin with, 15-minute spectrograms between 1-20 Hz are manually annotated using an open-source labelling tool. A deep Convolutional Neural Network (CNN), consisting of a ResNet Encoder, a UNet decoder, and a pixel-wise classification head, is then trained to classify features in each spectrogram using supervised learning. We explore the various factors, such as architecture changes to the CNN, that contribute to improving the model performance for annotating key features. We also discuss insights into attempted alternatives to standard supervised learning (semi-supervised and synthetic labels).

Finally, we discuss the results and evaluate the effectiveness of the ML/DL approach in mapping resonances, demonstrating its potential in utilising resonances as signals for tracking ocean currents (and whales). 

This project was funded by the UPFLOW project (ERC grant 101001601). This work was partially funded by FCT I.P./MCTES through PIDDAC (UIDB/50019/2020, UIDP/50019/2020 and LA/P/0068/2020).

How to cite: Corela, C., Saoulis, A., Tsekhmistrenko, M., Loureiro, A., Miranda, M., and Ferreira, A.: Decoding Ocean Depths: Machine Learning Insights into Tidal Dynamics with UPFLOW’s OBS Data, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-19546, https://doi.org/10.5194/egusphere-egu24-19546, 2024.

A fruitful close collaboration between teams of researchers and engineers of Geoazur, a research laboratory specializing in seismic observation, and Osean, a dynamic company combining the talents of engineers in low-noise, low-power electronics specialized in underwater acoustics, gives this instrument a guarantee of robustness, reliability and innovation by capitalizing on all their knowledge in their respective fields. Halios has an autonomy of 18 months, dimensions of 1.1 x 1.1 x 1m and a total weight with ballast of 346 kg. Electrical power is supplied by lithium or alkaline batteries. Halios is made entirely from corrosion-resistant materials. The 4-panel, mortise-and-tenon structure naturally creates a well in which the seismometer is protected. This well is covered by two flaps that open when the OBS sinks into the water, and close when the OBS stops on the bottom. During descent, the seismometer is suspended in the well. Once the OBS is in place, the seismometer is released automatically or by an acoustic command to land on the ground without any mechanical contact with the structure. This type of coupling is ideal, and protects the sensor from underwater currents. Communication with the OBS is possible via Ethernet, Wi-fi or acoustics once in the water by non-specialists. An acoustic module makes it possible to retrieve a health report on demand, which groups together the essential parameters of the station, and to modify some of them once the OBS is in the water. Halios incorporates precise clock or a CSAC-type atomic clock, if time precision is required.The seismometer is a compact Nanometrics Trillium, integrated in a dedicated container which also houses an accelerometer. An hydrophone, a precision temperature sensor and an absolute pressure sensor complete the range of sensors integrated on Halios. An dedicated acoustic modem also enables partial data retrieval from the surface.To retrieve the OBS, an acoustic command activates a mechanical release which frees the ballast. Dynamic assistance from leaf springs accelerates the positive thrust provided by the OBS's high-density syntactic foam. This foam is injected into a PHD shell, which effectively protects the whole unit and, thanks to its compact shape, provides effective anti-shock protection. The ascent of the OBS into the water column can be monitored by the acoustic module.Once at the surface, the time difference between the OBS clock and the GPS datum is automatically calculated. The OBS sends its GPS coordinates to the ship by VHF to facilitate recovery. At night, an LED flashing light can also be used to pinpoint its position. Halios is equipped with an interface that enables it to be connected to a real-time cable network, making it a versatile OBS of the highest performance and innovation.

How to cite: Hello, Y., Rebour, C., Sigloch, K., Bonnieux, S., and Philippe, O.: HALIOS  an innovative and clever Broad Band OBS., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-8185, https://doi.org/10.5194/egusphere-egu24-8185, 2024.

In this study, we focus on developing a low-angle active leveling device of geophone for Ocean Bottom Seismometers (OBS). The deployment of OBS in varied underwater environments poses a significant challenge due to unpredictable terrains, which often affect the positioning and orientation of these instruments. In such scenarios, the role of a leveling device becomes crucial, especially for self-floating OBS. Most of the leveling devices are designed with a 360° range, resulting in high costs and limited versality, our approach is informed by previous OBS positional data, suggesting that a limited corrective angle is adequate for underwater conditions. Our innovative leveling mechanism resembles a joystick in its transmission system, topped with a platform to enhance versatility. The control rod below the platform is controlled by two DC motors combined with worm gears, adjusting the rod along the X and Y axes with an approximate corrective angle of 30 degrees. For seismic sensing, our system utilizes one vertical and two horizontal geophones, housed in a custom sensor box to ensure the three axes are orthogonal to each other. This design simplifies the complexity of parts, combining off-the-shelf and custom components, offering a significant size and cost advantage. We have also tested its performance, with a focus on eliminating signal interference from resonance frequencies and assessing the system's stability. This includes a comparative analysis with older leveling mechanisms.

How to cite: Yang, C.-H., Lin, C.-R., Gung, Y., Lin, F.-S., and Chang, K.-H.: The Design of a Low-Angle Active Leveling Device for Short-Period Ocean Bottom Seismometer, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4888, https://doi.org/10.5194/egusphere-egu24-4888, 2024.

Cabled ocean bottom seismometer (OBS) solutions are financially and logistically challenging and autonomous OBS systems do not provide operators with seismic data until recovery. To address this issue Güralp has developed Aquarius, an ultra-low-power, free-fall OBS system, operational at any angle and with the ability to transmit seismic data in near-real-time from the seafloor without the use of cables.

The Güralp Aquarius incorporates seafloor-to-surface acoustic communication technology that allows state-of-health and noise performance interrogation during installation followed by retrieval of seismic data throughout the deployment period.

Omnidirectional broadband seismometer components allow the Aquarius to land and operate on steep slopes without requiring a gimbal mechanism that inherently introduces noise and failure modes. Raw data is recorded uncorrected for orientation to allow users to correct during post-processing.

These unique features allow the sensor to function on uneven seafloor as well as transmitting seismic data to the surface where the operator can use noise characteristics, location, and orientation data to determine if the landing site is suitable.

Intelligent battery design allows for typical 18-month deployments, with charging being possible alongside data transfer. This allows recharging, download and configuration simultaneously on the ship in between deployments.

Ease of configuration, deployment and recovery followed by simple data processing are all central themes to the Aquarius. The capital investment required to purchase OBS systems often means that the OBS instrument must be adaptable to a range of use-cases.

The Aquarius is in use in 6 different countries and is the unit of choice for the Canadian OBS pool, comprising 120 units for deployment around the globe. Aquarius units have been successfully deployed in the Mediterranean, North Atlantic as well as the North and South Pacific. Here, we focus on the mechanics of deployment and recovery for demonstrating to experienced and prospective principle investigators how this system simplifies and improves confidence in deployment.

How to cite: Clark, A., Lindsey, J., Hill, P., Foster, C., Restelli, F., Watkiss, N., Price, E., Nedimović, M., and Cairns, G.: Pushing Boundaries with Ocean Bottom Seismometers (OBS) with a Pool-Ready System: Güralp Aquarius, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-10807, https://doi.org/10.5194/egusphere-egu24-10807, 2024.

During the TYrrhenian Deep-sea Experiment (TYDE) 14 Ocean Bottom stations were deployed in the period December 2000-May 2001 on the southern Tyrrhenian seafloor around the Aeolian islands All the stations were equipped with Hydrophones (OBH) and six of them with 3-component geophones (OBS). This experiment represented one of the first and most important passive ocean bottom data acquisition ever realized in southern Tyrrhenian and the records were mainly used to perform receiver functions, modeling of the seismic velocity field, and detect and locate seismicity.

In this work we present a reprocessing of these old data using modern techniques of analysis to improve the quality of the signals and extract more information of what has been already published.

We show the workflow used for cleaning the records together with Machine Learning approaches to improve the detection of small events and the application of methods that were not common at that time, such as the Ambient Noise analyses.

Work supported by the “Pianeta Dinamico” call, 2023–2025 CAVEAT

How to cite: Calò, M., Gamez Lindoro, J. A., Di Luccio, F., and Bernal Manzanilla, K.: Reprocessing of the ocean bottom seismic data of the TYDE experiment using modern techniques., EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-6977, https://doi.org/10.5194/egusphere-egu24-6977, 2024.

Human activities in the marine environment coupled with the efficient propagation of acoustic waves, have significantly increased ocean noise levels. This rise poses a threat to marine biodiversity, especially impacting vocal-dependent marine mammals. Therefore, obtaining a more comprehensive understanding of the complex ocean soundscape, shaped by both natural and anthropogenic sources, is essential.

Ocean-bottom seismometer (OBS) datasets present an opportunity to explore the marine acoustic noise field. OBS hydrophones can be used to effectively study regional marine soundscapes. Additionally, the three-component OBS seismometers capture acoustic-to-seismic conversions from the water column, enabling the detection and localization of acoustic events.

Through the analysis of OBS hydrophone recordings from the NE Atlantic offshore Ireland, we develop a frequency-dependent empirical sound propagation model for the area. The model provides information on (1) the spectral amplitude decay of sound sources in the region and (2) the distance between OBSs and acoustic sources.

To construct the model, we consider ship noise from individual vessels recorded by OBSs at known distances from the source, using Automatic Identification System (AIS) ship-tracking data. As a ship moves away from its closest point of approach (CPA) to an OBS, the hydrophone records the evolution of the frequency-dependent ship's acoustic signal spectral amplitude with distance. This enables us to compute amplitude decay curves for narrow frequency bands from different vessel signals recorded at various OBSs. We apply a signal-to-noise ratio (SNR) quality control criterion and consider ships sailing along 'straight' paths at a constant speed (< 2 kn variation). After normalizing and scaling the CPAs of the different curves to a reference, we stack them for each frequency band and compute polynomial fits for the stacked amplitude decay curves. We analyse the number of vessel tracks needed to build a reliable empirical model. Numerical simulations are computed to further understand the acoustic propagation properties from our empirical model.

The goal is to investigate (1) how spectral amplitudes at different frequencies behave with distance from the sound sources, (2) the efficiency of frequency-dependent sound propagation in the ocean in regions with variable seafloor morphologies, and (3) the potential use of the empirical model to obtain distances to other sound sources with unknown locations, such as Baleen whales. To achieve this, we compute ratios of spectral amplitudes for pairs of different frequency bands using the fitted amplitude decay curves. We test the model's ability to locate other vessels with known AIS positions before using it to locate different sound sources with unknown positions. Locating a sound source employing the distance provided by the empirical model is valuable when OBSs are too far apart for 'traditional' multi-sensor methods. In such cases, the incoming direction of the sound source is determined by rotating the horizontal-component OBS seismic data to point to the source back-azimuth. Also, the amplitude decay information provided by the model has multiple applications in other areas where noise pollution may be a concern (e.g., marine infrastructure construction).

How to cite: Gómez-García, C. and Bean, C. J.: Empirical insights into ocean acoustic wave propagation and source localization using ocean-bottom seismometers, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-5502, https://doi.org/10.5194/egusphere-egu24-5502, 2024.

The Bransfield Basin, located between the South Shetland Island and the Antarctic Peninsula in Western Antarctica, is formed by slab rollback of the South Shetland Trench and complementary extensional. The region also has large and small active volcanoes like the Orca volcano. Various earthquakes have occurred due to tectonic movement and volcanic activities. Since the Mw 4.9 earthquake struck at a depth of 10 km in the vicinity of South Shetland Islands on 29 August 2020 (USGS), moderate earthquakes have constantly been reported adjacent to Bransfield Basin like an earthquake swarm.The Extreme Geoscience Team (EGG) at the Korea Polar Research Institute (KOPRI) has been operating a seismic station at the King Sejong station. Other countries have been running their seismic stations in the South Shetland Islands, such as the Antarctic Seismographic Argentinean Italian Network (ASAIN) and Chilean seismic stations, to monitor local earthquakes associated with back-arc spreading and submarine volcanic activities in this area. However, these seismic stations are located on the islands, and the inter-station distance is over 50 km. To make up for the poor coverage of seismic stations, we installed 5 OBSs in the earthquake swarm area between 2020 and 2021.

Using cross-correlation with onland seismic data and OBSs data, we conducted the template matching analysis and detected 67,952 events from August 29, 2020, to December 31, 2021. We relocated 2,313 epicenters of the seismic events by manually picking the first arrival of the P-wave. We found that the earthquake swarm’s epicenters were mainly concentrated in the central basin of the Bransfield Strait.

We categorize the characteristics of the earthquake near Bransfield Basin. We will contribute to understanding the dynamics and evolution of subduction zones in Bransfield Basin.

How to cite: Jung, J., Park, Y., Lee, W. S., and Park, S.: Seismicity in the Bransfield Strait, Antarctica, using temporal Ocean Bottom Seismographs data, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-1332, https://doi.org/10.5194/egusphere-egu24-1332, 2024.

The Zabargad Fracture Zone (ZFZ), located between 23.5^o N – 26^o N is the largest rift-axis offset in the Red Sea. The ZFZ is a fundamental tectonic element of ~100 km rift-axis offset that marks the transition between the northern and central Red Sea. Due to data scarcity, our understanding of the seismic activity and the potential presence of transform faults or non-transform offsets in the ZFZ area is limited. Local seismological datasets so far have been restricted to the onshore recordings, thus hampering the ability to study the ZFZ's role in the Red Sea’s tectonic evolution and to assess its potential seismic hazard associated with transform faults accommodating the rift-axis offset in the Red Sea.

To fill this data gap, we deployed 14 broadband Ocean Bottom Seismometers (OBS) and four land stations for 12 months within the ZFZ to collect continuous waveforms. The newly collected data span the period 11/2021 to 11/2022. First, we corrected recorded OBS-waveforms for clock drift using skew values measured at the time of the OBS recovery and time-shifts measured from ambient noise cross-correlations for stations whose skews could not be measured. Next, we applied a deep-learning-based algorithm to automatically detect earthquakes and pick P and S phases. We verified and, if necessary, manually corrected these phase picks; this approach identified more than 3500 local earthquakes.

To obtain reliable hypocentral locations, we inverted for an optimum crustal 1-D seismic velocity model of the ZFZ and station delays simultaneously. Using as reference a nearby land station installed on Precambrian bedrock, we obtained positive OBS station delays of up to 2 seconds. These delays correspond to late phase arrivals, most likely due to thick sedimentary and salt deposits. Our local magnitude (M_L) calculations show that OBS-station M_L values are up to 1.5 units larger than for the land stations. We find that station delays and M_L deviations are correlated, highlighting the importance to account for the variability of shallow geological and bathymetric properties for accurate earthquake location and magnitude estimation.

Our seismicity analysis reveals two major distinct spatial clusters of earthquakes, in the southern (23.95 ^o N – 24.53 ^o N) and northern (24.68 ^o N – 25.43^o N) parts of the ZFZ. Seismicity in the south higher seismicity rate and spatially concentrated than in the north. We also identified three seismic swarms in the south and one seismic swarm in the north lasted for one to three weeks each. Our preliminary analyses document the potential of this new dataset to address key questions on the seismotectonics and seismic activity of the ZFZ.

How to cite: Shiddiqi, H. A., Parisi, L., Cano, E. V., Mai, P. M., Augustin, N., and Jónsson, S.: Seismic Activity in the Zabargad Fracture Zone (Red Sea): Insights from an Amphibious Seismic Network, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-4563, https://doi.org/10.5194/egusphere-egu24-4563, 2024.

Since the introduction of the ocean bottom seismograph (OBS), it has played a pivotal role in the expansion of our comprehension of the Earth’s interior structure and dynamics. Typically deployed through free-fall from a ship, the position of the OBS on the seafloor and its orientation are initially unknown. When using airgun shots to locate and orient an instrument, a severe downside is the often-limited azimuthal distribution of shots around the OBS, possibly leading to large location and, subsequently, orientation errors. We introduce a novel and efficient method aimed at enhancing the accuracy of station locations in active-source seismic experiments. The approach uses airgun shots recorded as part of the experiment, and integrates both acoustic wave travel time information and waveform polarizations. The inverse problem of station location is formulated in terms of Bayesian inference. At each location of a search grid, the misfit of observed and theoretical water wave travel times and the clustering of polarization data are combined into one probabilistic formulation to map the relative likelihood of a station’s position and orientation. We demonstrate the practical utility of the method via application to the location and orientation of OBS deployed during a recent seismic experiment located across the Hawaiian Ridge. Thereafter, we will use the results to compute radial seismograms, extract the travel times of crustal and mantle S waves, and develop S-wave models of the oceanic lithosphere across the Hawaiian Ridge.

How to cite: Cryder, M., Dunn, R., and Xu, C.: Integrated Bayesian Approach for Station Location and Orientation in Marine Active-Source Seismic Shear Wave Studies, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-11876, https://doi.org/10.5194/egusphere-egu24-11876, 2024.

Upward mantle flow is key to understand global mantle geodynamics yet its imaging remains challenging due to potential associated low velocity contrasts and small lateral dimensions. In this study we use the UPFLOW ocean bottom seismometer (OBS) dataset to build images of radial anisotropy to constrain the patterns of mantle upwellings in the Azores-Madeira-Canary Islands region. We use the partitioned waveform inversion (PWI) method whereby non-linear waveform fitting of surface waves filtered between T~ 16 s and T~ 300 s is performed using a successive series of time-frequency windows in two stages. Firstly, the surface wave fundamental mode is extracted via phase match filtering and is used to obtain path average perturbations in shear radial anisotropy and isotropic shear wave velocity from a smoothed combination of the 1-D mantle model ak135 and the crustal model CRUST1.0. These perturbations establish a new initial model, which is subsequently used to estimate the path averaged radial anisotropy model that leads to the best fit between the observed trace and the synthetic waveform obtained by summing all overtones (up to n = 20). An iterative, regularized least squares inversion is used to invert for 3-D radially anisotropic mantle structure. Uncertainties are automatically quantified and used to interpret resolved seismic structures. Preliminary results are compared to previous tomographic models of the Atlantic region.

How to cite: Harris, K., Witek, M., Ferreira, A. M. G., Chang, S.-J., and Miranda, M.: Surface wave mantle anisotropy tomography of the Azores-Madeira-Canaries region using UPFLOW data: initial results, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-16613, https://doi.org/10.5194/egusphere-egu24-16613, 2024.

The Ontong Java Plateau (OJP), formed around 120 Ma, represents a significant event in the Earth’s geologic history. It is the largest preserved Large Igneous Province (LIP) by volume on the Earth. The voluminous and rapid magmatism placed a thick 30-40 km crust on a normal oceanic lithosphere in submarine conditions. The lithosphere has since experienced low subsidence suggesting only modest thermal perturbation. Several hypotheses have been proposed to explain the unique emplacement history, including a hot mantle plume, meteorite impact, or passive mantle upwelling close to a fast-spreading ridge. In this study, we image the crust and upper mantle structure beneath the OJP using body waves and surface waves. A challenge of body wave imaging in submarine environments is the severe reverberations generated in the water column and seafloor sediments. Additionally, multiple reflections from shallow interfaces (e.g., Moho and intra-crustal layers) may interfere with deeper upper mantle phases, making interpretation ambiguous. In this work, we obtain high-resolution Ps-RFs at 17 OBS (ocean bottom seismometer) stations in the OJP region using a transdimensional hierarchical Bayesian deconvolution approach. We then take advantage of two recently developed techniques: (1) FADER (Fast Automated Detection and Elimination of Echoes and Reverberations), which uses autocorrelation and cepstral analysis to identify and remove water and sediment reverberations, and (2) CRISP-RF (Clean Receiver function Imaging with SParse Radon Filters), which uses sparse Radon transforms to eliminate multiples and incoherent noise. We anticipate applying our novel ideas to traditional seismic imaging will provide robust constraints on the discontinuities detected beneath the OJP, shedding new light on its lithospheric structure and origin.

How to cite: Zhang, Z. and Olugboji, T.: How did the Largest Oceanic Plateau modify the Normal Oceanic Lithosphere?, EGU General Assembly 2024, Vienna, Austria, 14–19 Apr 2024, EGU24-13081, https://doi.org/10.5194/egusphere-egu24-13081, 2024.