Search

Detecting phase arrivals and pinpointing the arrival times of seismic phases in seismograms is crucial for many seismological analysis workflows. For land station data, machine learning methods have already found widespread adoption. However, deep learning approaches are not yet commonly applied to ocean bottom data due to a lack of appropriate training data and models. Here, we compiled an extensive and labeled ocean bottom seismometer (OBS) data set from 15 deployments in different tectonic settings, comprising ∼90,000 P and ∼63,000 S manual picks from 13,190 events and 355 stations. We propose PickBlue, an adaptation of the two popular deep learning networks EQTransformer and PhaseNet. PickBlue joint processes three seismometer recordings in conjunction with a hydrophone component and is trained with the waveforms in the new database. The performance is enhanced by employing transfer learning, where initial weights are derived from models trained with land earthquake data. PickBlue significantly outperforms neural networks trained with land stations and models trained without hydrophone data. The model achieves a mean absolute deviation of 0.05 s for P-waves and 0.12 s for S-waves, and we apply the picker on the Hikurangi Ocean Bottom Tremor and Slow Slip OBS deployment offshore New Zealand. We integrate our data set and trained models into SeisBench to enable an easy and direct application in future deployments. Key Points We assembled a database of ocean Bottom Seismometer (OBS) waveforms and manual P and S picks, on which we train PickBlue, a deep learning picker Our picker significantly outperforms pickers trained with land-based data with confidence values reflecting the likelihood of outlier picks The picker and database are available in the SeisBench platform, allowing easy and direct application to OBS traces and hydrophone records

Observations of spatio-temporal variability of the deep ocean are rare and little is known about occurrence of deep ocean mesoscale dynamics. Here, we make use of 2.5 years of time series data from three distributed sensor arrays, which acquired high-resolution temperature, pressure and sound speed data of the bottom layer offshore northern Chile. Estimating salinity and density from the direct observations enable access to the full spectrum of hydrographic variability from a multi-hourly to annual time scale and with average inter-station distances of less than 1 km. Analyses revealed interannual warming over the continental slope of 0.002 °C yr−1–0.003 °C yr−1, and could trace periodic hydrographic anomalies, likely related to coastal-trapped waves, as far as to the lower continental slope. A concurrent change in the shape of the warm anomalies and the rate of deep-sea warming that occurs with the crossing of the deep-sea trench suggests that the abyssal part of the eastern boundary current system off Chile does not extend past the deep sea trench. Furthermore, the comparison of anomaly timing and shape in between stations implies southwards flow over the mid to lower continental slope, centred closer to the trench.

24.02.2023 - 11.04.2023, Talcahuano (Chile) - Guayaquil (Ecuador)

We deployed a dense geodetic and seismological network in the Atacama seismic gap in Chile. We derive a microseismicity catalog of >30,000 events, time series from 70 GNSS stations, and apply a transdimensional Bayesian inversion to estimate interplate locking degree. We identify two highly locked regions of different sizes whose geometries appear to control seismicity patterns. Interface seismicity concentrates beneath the coastline just downdip of the highest locking. A region of lower interplate locking around 27.5ºS coincides with higher seismicity levels, a high number of repeating earthquakes and events extending further towards the trench. Having shown numerous signs of aseismic deformation (slow-slip events and earthquake swarms), this area is situated where the Copiapó Ridge is subducted. While these findings suggest that the structure of the downgoing oceanic plate prescribes patterns of interplate locking and seismicity, we note that the Taltal Ridge further north lacks a similar signature.

24.02.2023 - 11.04.2023, Talcahuano (Chile) - Guayaquil (Ecuador)

24.02.2023 - 11.04.2023, Talcahuano (Chile) - Guayaquil (Ecuador)

24.02.2023 - 11.04.2023, Talcahuano (Chile) - Guayaquil (Ecuador)

24.02.2023 - 11.04.2023, Talcahuano (Chile) - Guayaquil (Ecuador)

At the Blanco transform fault system (BTFS) off Oregon, 138 local earthquakes and 84 double-couple focal mechanisms from ocean-bottom-seismometer recordings jointly discussed with bathymetric features reveal a highly segmented transform system without prominent fracture zone traces longer than 100 km. In the west, seismicity is focused at deep troughs (i.e., the West and East Blanco, and Surveyor Depressions). In the east, the BTFS lacks a characteristic transform valley and instead developed the Blanco Ridge, which is the most seismically active feature, showing strike-slip and dip-slip faulting. Sandwiched between the two main segments of the BTFS is the Cascadia depression, representing a short intra-transform spreading segment. Seismic slip vectors reveal that stresses at the eastern BTFS are roughly in line with plate motion. In contrast, stresses to the west are clockwise skewed, indicating ongoing reorganization of the OTF system. As we observed no prominent fracture zones at the BTFS, plate tectonic reconstructions suggest that the BTFS developed from non-transform offsets rather than pre-existing transform faults during a series of ridge propagation events. Our observations suggest that the BTFS can be divided into two oceanic transform systems. The eastern BTFS is suggested to be a mature transform plate boundary since ∼0.6 Ma. In contrast, the western BTFS is an immature transform system, which is still evolving to accommodate far-field stress change. The BTFS acts as a natural laboratory to yield processes governing the development of oceanic transform faults. Key Points Local seismicity of the Blanco transform fault system (BTFS) reveals along-strike variations dominated by strike-slip and oblique dip-slip The BTFS developed from non-transform offsets rather than discrete transform faults in response to plate rotation and ridge propagation The BTFS consists of a mature plate boundary in the east and an immature system in the west, separated by a central

24.02.2023 - 11.04.2023, Talcahuano (Chile) - Guayaquil (Ecuador)

24.02.2023 - 11.04.2023, Talcahuano (Chile) - Guayaquil (Ecuador)

Machine‐learning (ML) methods have seen widespread adoption in seismology in recent years. The ability of these techniques to efficiently infer the statistical properties of large datasets often provides significant improvements over traditional techniques when the number of data are large (millions of examples). With the entire spectrum of seismological tasks, for example, seismic picking and detection, magnitude and source property estimation, ground‐motion prediction, hypocenter determination, among others, now incorporating ML approaches, numerous models are emerging as these techniques are further adopted within seismology. To evaluate these algorithms, quality‐controlled benchmark datasets that contain representative class distributions are vital. In addition to this, models require implementation through a common framework to facilitate comparison. Accessing these various benchmark datasets for training and implementing the standardization of models is currently a time‐consuming process, hindering further advancement of ML techniques within seismology. These development bottlenecks also affect “practitioners” seeking to deploy the latest models on seismic data, without having to necessarily learn entirely new ML frameworks to perform this task. We present SeisBench as a software package to tackle these issues. SeisBench is an open‐source framework for deploying ML in seismology—available via GitHub. SeisBench standardizes access to both models and datasets, while also providing a range of common processing and data augmentation operations through the API. Through SeisBench, users can access several seismological ML models and benchmark datasets available in the literature via a single interface. SeisBench is built to be extensible, with community involvement encouraged to expand the package. Having such frameworks available for accessing leading ML models forms an essential tool for seismologists seeking to iterate and apply the next generation of ML techniqu

Volcano seismology is an essential tool for monitoring volcanic processes in the advent and during eruptions. A variety of seismic signals can be recorded at volcanoes, of which some are thought to be related to the migration of fluids which is of primary importance for the anticipation of imminent eruptions. We investigate the volcanic crises at Villarrica volcano in 2015 and report on a newly discovered very-long-period (VLP) signal that accompanies phases of periodic long period (LP) signal burst. Despite their low amplitude emergent character, we can locate the source region of the 1 Hz LP signals to the close vicinity of the volcano using a network-based correlation method. The source of the VLP signal with a period of about 30–100 s appears to locate in the vicinity of two stations a few kilometres from the summit. Both stations record very similar VLP waveforms that are correlated with the envelope of the LP bursts. A shallow magma reservoir was inferred by Contreras from surface deformation as the source of inflation following the eruption in 2015. Cyclic volume changes of 6 m3 in this reservoir at 3 km depth can explain the observed amplitudes of the vertical VLP signal. We propose that the LP signal is generated by the migration of gas or gas-rich magma that is periodically released from the inflating reservoir through a non-linear valve structure which modulates the flux, and thereby causes bursts of flow-related LP signals and pressure changes observed as VLP deformation. Our model predicts that the correlated occurrence of LP bursts and VLP surface motion depends on the intensity of the fluid flux. A weaker flux of fluids may not exceed the opening pressure of valve structure, and higher rates might maintain pressure above the closing pressure. In both cases, the VLP signal vanishes. Our observation provides constrains for models of fluid transport inside volcanoes. At Villarrica the VLP signal, and its relation to the LP activity, reveal additional inf

Seismic event detection and phase picking are the base of many seismological workflows. In recent years, several publications demonstrated that deep learning approaches significantly outperform classical approaches and even achieve human-like performance under certain circumstances. However, as most studies differ in the datasets and exact evaluation tasks studied, it is yet unclear how the different approaches compare to each other. Furthermore, there are no systematic studies how the models perform in a cross-domain scenario, i.e., when applied to data with different characteristics. Here, we address these questions by conducting a large-scale benchmark study. We compare six previously published deep learning models on eight datasets covering local to teleseismic distances and on three tasks: event detection, phase identification and onset time picking. Furthermore, we compare the results to a classical Baer-Kradolfer picker. Overall, we observe the best performance for EQTransformer, GPD and PhaseNet, with EQTransformer having a small advantage for teleseismic data. Furthermore, we conduct a cross-domain study, in which we analyze model performance on datasets they were not trained on. We show that trained models can be transferred between regions with only mild performance degradation, but not from regional to teleseismic data or vice versa. As deep learning for detection and picking is a rapidly evolving field, we ensured extensibility of our benchmark by building our code on standardized frameworks and making it openly accessible. This allows model developers to easily compare new models or evaluate performance on new datasets, beyond those presented here. Furthermore, we make all trained models available through the SeisBench framework, giving end-users an easy way to apply these models in seismological analysis.

The Ligurian Sea lies in the heart of Europe. With its densely settled coastline, large cities, and thriving tourism, it is one of the most populous regions in Europe, and yet there is still little known about the tectonic structures forming this Sea. The Ligurian Sea is bordered to the north and east by the Western Alpine and Apennine orogenies respectively. In fact, the Sea formed as a back-arc basin related to Apennine-Calabria-Maghrebides subduction zone retreat. Opening began in the late Oligocene and continued into the early Miocene. During opening the Corsica-Sardinia block rifted eastwards and rotated counter-clockwise to its current position. Geodetic and seismic data show that the region has ceased to open and is now under compression, with shortening of 0.3-1.5cm/year. Compression in the basin has the potential to reactivate local rifting-related faults. The Ligurian margin has hosted large earthquakes in the past, most notably the 1887 Mw 6.3-7.5 earthquake near the town of Imperia, Liguria. A similar magnitude earthquake could cause substantial damage to the densely populated coastline, especially if it were to trigger a tsunami. To better understand the risks posed by seismicity in the area it is important to improve our understanding of the tectonic setting, including the nature of the Ligurian Basin’s crust and how the Western Alpine zone continues offshore. In the frame of the DFG Priority Program “Mountain Building Processes in Four Dimensions (MB-4D)”, we use seismic data from airgun shots and local seismicity recorded on Ocean Bottom Seismometers (OBS) from the AlpArray OBS network and land stations as input for a 3D P-wave tomographic model of the Ligurian Basin and surrounding coastline into the South-Western Alps. We use Pg and Pn phases to investigate the P-wave velocity structure of the crust and uppermost mantle using the inversion program FMTOMO. This poster presents preliminary results of this 3D P-wave velocity model for the Ligurian Sea, Plain-language Summary: The Ligurian Sea lies in the heart of Europe, and yet there is little known about the tectonic setting of this densely populated area. The Ligurian Sea is bordered to the north by the Western Alps and to the east by the Apennine mountains. It began opening in the Late Oligocene and continued opening until the early Miocene. During this time the islands of Corsica and Sardinia rotated counter clockwise from near the French coast to their current locations. GPS and seismic data show that the region is now under compression. Some historical large earthquakes have been recorded in the region, and future earthquakes could be very damaging to the densely populated coastline. This is why it is important to learn about the local structure of the crust and upper mantle, so we can better understand the risk posed by seismic activity to the area. This project is part of the DFG Priority Program “Mountain Building Processes in Four Dimensions (MB-4D)” and the AlpArray initiative. We present the first results of 3D P-wave velocity tomography of the Ligurian Sea region. This model uses data from airgun shots and local earthquakes which were recorded on land and sea stations