Detection of records of weak local earthquakes using neural networks

Manual processing of large volumes of continuous observations produced by local seismic networks takes a lot of time, therefore, to solve this problem, automatic algorithms for detecting seismic events are used. Deterministic methods for solving the problem of detection, which do an excellent job of detecting intensive earthquakes, face critical problems when detecting weak seismic events (earthquakes). They are based on principles based on the calculation of energy, which causes multiple errors in detection: weak seismic events may not be detected, and high-amplitude noise may be mistakenly detected as an event. In our work, we propose a detection method capable of surpassing deterministic methods in detecting events on seismograms, successfully detecting a similar or more events with fewer false detections.

Migration-Based Local Event-Location Workflow for Ocean-Bottom Seismometer (OBS) Records in Subduction Zones: A Practical Approach for Addressing a Large Number of Events

ABSTRACT An efficient event-location workflow is highly desired to analyze large numbers of local earthquakes recorded by ocean-bottom seismometers (OBSs) in subduction zones. The present study proposes a migration-based event-location approach for evaluating OBS records to examine local subduction-zone earthquakes. This approach can significantly reduce the amount of manual time picks compared with conventional methods. The event-location workflow was designed to detect arrival onsets of both P and S phases. Synthetic tests have shown that the proposed migration-based event-location method is robust against different types of noise, such as environmental noise and local spike noise. This workflow was then applied to real OBS data in the off-Ibaraki region at the southern end of the Japan trench. The results show that this approach is applicable to real data from subduction-zone events: It gives reasonable agreement with manual time picks for both P and S waves and reasonable error bars, and it demonstrates a clear down-dip trend of seismicity. The results also show fair agreement with event distributions from previous studies of the off-Ibaraki region. This proposed workflow can be used to examine the seismicity of local earthquakes around the subduction zone using OBSs. This approach is especially effective when the seismicity is high and/or in cases in which long-term OBS monitoring has recorded a large number of events.

The Seismo-Performer: A Novel Machine Learning Approach for General and Efficient Seismic Phase Recognition from Local Earthquakes in Real Time

When recording seismic ground motion in multiple sites using independent recording stations one needs to recognize the presence of the same parts of seismic waves arriving at these stations. This problem is known in seismology as seismic phase picking. It is challenging to automate the accurate picking of seismic phases to the level of human capabilities. By solving this problem, it would be possible to automate routine processing in real time on any local network. A new machine learning approach was developed to classify seismic phases from local earthquakes. The resulting model is based on spectrograms and utilizes the transformer architecture with a self-attention mechanism and without any convolution blocks. The model is general for various local networks and has only 57 k learning parameters. To assess the generalization property, two new datasets were developed, containing local earthquake data collected from two different regions using a wide variety of seismic instruments. The data were not involved in the training process for any model to estimate the generalization property. The new model exhibits the best classification and computation performance results on its pre-trained weights compared with baseline models from related work. The model code is available online and is ready for day-to-day real-time processing on conventional seismic equipment without graphics processing units.

One Year of Seismicity Recorded Through Ocean Bottom Seismometers Illuminates Active Tectonic Structures in the Ionian Sea (Central Mediterranean)

Seismological data recorded in the Ionian Sea by a network of seven Ocean Bottom Seismometers (OBSs) during the 2017–2018 SEISMOFAULTS experiment provides a close-up view of seismogenic structures that are potential sources of medium-high magnitude earthquakes. The high-quality signal-to-noise ratio waveforms are observed for earthquakes at different scales: teleseismic, regional, and local earthquakes as well as single station earthquakes and small crack events. In this work, we focus on two different types of recording: 1) local earthquakes and 2) Short Duration Events (SDE) associated to micro-fracturing processes. During the SEISMOFAULTS experiment, 133 local earthquakes were recorded by both OBSs and land stations (local magnitude ranging between 0.9 and 3.8), while a group of local earthquakes (76), due to their low magnitude, were recorded only by the OBS network. We relocated 133 earthquakes by integrating onshore and offshore travel times and obtaining a significant improvement in accuracy, particularly for the offshore events. Moreover, the higher signal-to-noise ratio of the OBS network revealed a significant seismicity not detected onshore, which shed new light on the location and kinematics of seismogenic structures in the Calabrian Arc accretionary prism and associated to the subduction of the Ionian lithosphere beneath the Apennines. Other signals recorded only by the OBS network include a high number of Short Duration Events (SDE). The different waveforms of SDEs at two groups of OBSs and the close correlation between the occurrence of events recorded at single stations and SDEs suggest an endogenous fluid venting from mud volcanoes and active fault traces. Results from the analysis of seismological data collected during the SEISMOFAULTS experiment confirm the necessity and potential of marine studies with OBSs, particularly in those geologically active areas of the Mediterranean Sea prone to high seismic risk.