
 The spectral ratio between horizontal and vertical components (H/V ratio) of microtremors measured at the ground surface has been used to estimate fundamental periods and amplification factors of a site, although this technique lacks theoretical background. The aim of this article is to formulate the H/V technique in terms of the characteristics of Rayleigh and Love waves, and to contribute to improve the technique. The improvement includes use of not only peaks but also troughs in the H/V ratio for reliable estimation of the period and use of a newly proposed smoothing function for better estimation of the amplification factor. The formulation leads to a simple formula for the amplification factor expressed with the H/V ratio. With microtremor data measured at 546 junior high schools in 23 wards of Tokyo, the improved technique is applied to mapping site periods and amplification factors in the area.

Ground Motion Characteristics Estimated from Spectral Ratio between Horizontal and Verticcl Components of Mietremors.

Seismology is a data rich and data-driven science. Application of machine learning for gaining new insights from seismic data is a rapidly evolving sub-field of seismology. The availability of a large amount of seismic data and computational resources, together with the development of advanced techniques can foster more robust models and algorithms to process and analyze seismic signals. Known examples or labeled data sets, are the essential requisite for building supervised models. Seismology has labeled data, but the reliability of those labels is highly variable, and the lack of high-quality labeled data sets to serve as ground truth as well as the lack of standard benchmarks are obstacles to more rapid progress. In this paper we present a high-quality, large-scale, and global data set of local earthquake and non-earthquake signals recorded by seismic instruments. The data set in its current state contains two categories: (1) local earthquake waveforms (recorded at “local” distances within 350 km of earthquakes) and (2) seismic noise waveforms that are free of earthquake signals. Together these data comprise ~1.2 million time series or more than 19,000 hours of seismic signal recordings. Constructing such a large-scale database with reliable labels is a challenging task. Here, we present the properties of the data set, describe the data collection, quality control procedures, and processing steps we undertook to insure accurate labeling, and discuss potential applications. We hope that the scale and accuracy of STEAD presents new and unparalleled opportunities to researchers in the seismological community and beyond.

/pdf/stanford-earthquake-dataset-stead-a-global-data-set-of-2ulnt20z63.pdf

STanford EArthquake Dataset (STEAD): A Global Data Set of Seismic Signals for AI

The dynamics of magma deep in the Earth’s crust are difficult to capture by geophysical monitoring. Since May 2018, a seismically quiet area offshore of Mayotte in the western Indian Ocean has been affected by complex seismic activity, including long-duration, very-long-period signals detected globally. Global Navigation Satellite System stations on Mayotte have also recorded a large surface deflation offshore. Here we analyse regional and global seismic and deformation data to provide a one-year-long detailed picture of a deep, rare magmatic process. We identify about 7,000 volcano-tectonic earthquakes and 407 very-long-period seismic signals. Early earthquakes migrated upward in response to a magmatic dyke propagating from Moho depth to the surface, whereas later events marked the progressive failure of the roof of a magma reservoir, triggering its resonance. An analysis of the very-long-period seismicity and deformation suggests that at least 1.3 km3 of magma drained from a reservoir of 10 to 15 km diameter at 25 to 35 km depth. We demonstrate that such deep offshore magmatic activity can be captured without any on-site monitoring. Recent seismicity near Mayotte in the Indian Ocean is due to dyke propagation from and drainage of a 25–35 km deep magma reservoir, according to an analysis of earthquake and deformation data.

/pdf/drainage-of-a-deep-magma-reservoir-near-mayotte-inferred-52xnxyqigi.pdf

Drainage of a deep magma reservoir near Mayotte inferred from seismicity and deformation

On May 10th, 2018, an unprecedented long and intense seismic crisis started offshore, east of Mayotte, the easternmost of the Comoros volcanic islands. The population felt hundreds of events. Over the course of one year, 32 earthquakes with magnitude greater than 5 occurred, including the largest event ever recorded in the Comoros (Mw = 5.9 on May 15th, 2018). Earthquakes are clustered in space and time. Unusual intense long lasting monochromatic very long period events were also registered. From early July 2018, Global Navigation Satellite System stations and Interferometric Synthetic Aperture Radar registered a large drift, testimony of a large offshore deflation. We describe the onset and the evolution of a large magmatic event thanks to the analysis of the seismicity from the initiation of the crisis through its first year, compared to the ground deformation observation (GNSS and InSAR) and modelling. We discriminate and characterise the initial fracturing phase, the phase of magma intrusion and dike propagation from depth to the sub-surface, and the eruptive phase that starts on July 3rd, 2018, around fifty days after the first seismic events. The eruption is not terminated two years after its initiation, with the persistence of an unusual seismicity, whose pattern has been similar since summer 2018, including episodic very low frequency events presenting a harmonic oscillation with a period of ~16 s. From July 2018, the whole Mayotte Island drifted eastward and downward at a slightly increasing rate until reaching a peak in late 2018. At the apex, the mean deformation rate was 224 mm yr-1 eastward and 186 mm yr-1 downward. During 2019, the deformation smoothly decreased and in January 2020, it was less than 20% of its peak value. A deflation model of a magma reservoir buried in a homogenous half space fits well the data. The modelled reservoir is located 45 ± 5 km east of Mayotte, at a depth of 28 ± 3 km and the inferred magma extraction at the apex was ~94 m3 s-1. The introduction of a small secondary source located beneath Mayotte Island at the same depth as the main one improves the fit by 20%. While the rate of the main source drops by a factor of 5 during 2019, the rate of the secondary source remains stable. This might be a clue of the occurrence of relaxation at depth that may continue for some time after the end of the eruption. According to our model, the total volume extracted from the deep reservoir was ~2.65 km3 in January 2020. This is the largest offshore volcanic event ever quantitatively documented. This seismo-volcanic crisis is consistent with the trans-tensional regime along Comoros archipelago.

https://eartharxiv.org/repository/object/1013/download/2250/

The 2018–2019 seismo-volcanic crisis east of Mayotte, Comoros islands: seismicity and ground deformation markers of an exceptional submarine eruption

Most seismological analysis methods require knowledge of the geographic location of the stations comprising a seismic network. However, common machine learning tools used in seismology do not account for this spatial information, and so there is an underutilized potential for improving the performance of machine learning models. In this work, we propose a graph neural network (GNN) approach that explicitly incorporates and leverages spatial information for the task of seismic source characterization (specifically, location and magnitude estimation), based on multistation waveform recordings. Even using a modestly‐sized GNN, we achieve model prediction accuracy that outperforms methods that are agnostic to station locations. Moreover, the proposed method is flexible to the number of seismic stations included in the analysis and is invariant to the order in which the stations are arranged, which opens up new applications in the automation of seismological tasks and in earthquake early warning systems.

/pdf/automated-seismic-source-characterization-using-deep-graph-1a39cgob4q.pdf

Automated Seismic Source Characterization Using Deep Graph Neural Networks

A Deep Convolutional Neural Network for Localization of Clustered Earthquakes Based on Multistation Full Waveforms

/pdf/review-can-animals-predict-earthquakes-o0cuohifgz.pdf

Review: Can Animals Predict Earthquakes?

. The Alpine mountains in central Europe are characterized by a heterogeneous crust accumulating different tectonic units and blocks in close proximity to sedimentary foreland basins. Centroid moment tensor inversion provides insight into the faulting mechanisms of earthquakes and related tectonic processes but is significantly aggravated in such an environment.
Thanks to the dense AlpArray seismic network and our flexible bootstrap-based inversion tool Grond, we are able to test different setups with respect to the uncertainties of the obtained moment tensors and centroid locations. We evaluate the influence of frequency bands, azimuthal gaps, input data types, and distance ranges and study the occurrence and reliability of non-double-couple (DC) components. We infer that for most earthquakes ( Mw≥3.3 ) a combination of time domain full waveforms and frequency domain amplitude spectra in a frequency band of 0.02–0.07 Hz is suitable. Relying on the results of our methodological tests, we perform deviatoric moment tensor (MT) inversions for events with Mw>3.0 . Here, we present 75 solutions for earthquakes between January 2016 and December 2019 and analyze our results in the seismotectonic context of historical earthquakes, seismic activity of the last 3 decades, and GNSS deformation data. We study regions of comparably high seismic activity during the last decades, namely the Western Alps, the region around Lake Garda, and the eastern Southern Alps, as well as clusters further from the study region, i.e., in the northern Dinarides and the Apennines. Seismicity is particularly low in the Eastern Alps and in parts of the Central Alps. We apply a clustering algorithm to focal mechanisms, considering additional mechanisms from existing catalogs. Related to the N–S compressional regime, E–W-to-ENE–WSW-striking thrust faulting is mainly observed in the Friuli area in the eastern Southern Alps. Strike-slip faulting with a similarly oriented pressure axis is observed along the northern margin of the Central Alps and in the northern Dinarides. NW–SE-striking normal faulting is observed in the NW Alps, showing a similar strike direction to normal faulting earthquakes in the Apennines. Both our centroid depths and hypocentral depths in existing catalogs indicate that Alpine seismicity is predominantly very shallow; about 80 % of the studied events have depths shallower than 10 km.

/pdf/regional-centroid-moment-tensor-inversion-of-small-to-4iuai04chz.pdf

Regional centroid moment tensor inversion of small to moderate earthquakes in the Alps using the dense AlpArray seismic network: challenges and seismotectonic insights

Abstract An earthquake swarm affected the Bransfield Strait, Antarctica, a unique rift basin in transition from intra-arc rifting to ocean spreading. The swarm, counting ~85,000 volcano-tectonic earthquakes since August 2020, is located close to the Orca submarine volcano, previously considered inactive. Simultaneously, geodetic data reported up to ~11 cm northwestward displacement over King George Island. We use a broad variety of geophysical data and methods to reveal the complex migration of seismicity, accompanying the intrusion of 0.26–0.56 km 3 of magma. Strike-slip earthquakes mark the intrusion at depth, while shallower normal faulting the ~20 km long lateral growth of a dike. Seismicity abruptly decreased after a Mw 6.0 earthquake, suggesting the magmatic dike lost pressure with the slipping of a large fault. A seafloor eruption is likely, but not confirmed by sea surface temperature anomalies. The unrest documents episodic magmatic intrusion in the Bransfield Strait, providing unique insights into active continental rifting. 

/pdf/massive-earthquake-swarm-driven-by-magmatic-intrusion-at-the-1mhzl6bl.pdf

Gesa Petersen

Papers

Drainage of a deep magma reservoir near Mayotte inferred from seismicity and deformation

A Deep Convolutional Neural Network for Localization of Clustered Earthquakes Based on Multistation Full Waveforms

Review: Can Animals Predict Earthquakes?

Regional centroid moment tensor inversion of small to moderate earthquakes in the Alps using the dense AlpArray seismic network: challenges and seismotectonic insights

Massive earthquake swarm driven by magmatic intrusion at the Bransfield Strait, Antarctica