Recalibration of the Local Magnitude (ML) Scale for Earthquakes in the Yellowstone Volcanic Region

ABSTRACT The Yellowstone volcanic region is one of the most seismically active areas in the western United States. Assigning magnitudes (M) to Yellowstone earthquakes is a critical component of monitoring this geologically dynamic zone. The University of Utah Seismograph Stations (UUSS) has assigned M to 46,767 earthquakes in Yellowstone that occurred between 1 January 1984 and 31 December 2020. Here, we recalibrate the local magnitude (ML) distance and station corrections for the Yellowstone volcanic region. This revision takes advantage of the large catalog of earthquakes and an increase in broadband stations installed by the UUSS since the last ML update in 2007. Using a nonparametric method, we invert 7728 high-quality, analyst-reviewed amplitude measurements from 1383 spatially distributed earthquakes for 39 distance corrections and 20 station corrections. The inversion is constrained with four moment magnitude (Mw) values determined from time-domain inversion of regional-distance broadband waveforms by the UUSS. Overall, the new distance corrections indicate relatively high attenuation of amplitudes with distance. The distance corrections decrease with hypocentral distance from 3 km to a local minimum at 80 km, rise to a broad peak at 110 km, and then decrease again out to 180 km. The broad peak may result from superposition of direct arrivals with near-critical Moho reflections. Our ML inversion doubles the number of stations with ML corrections in and near the Yellowstone volcanic region. We estimate that the additional station corrections will nearly triple the number of Yellowstone earthquakes that can be assigned an ML. The new ML distance and station corrections will also reduce uncertainties in the mean MLs for Yellowstone earthquakes. The new MLs are ∼0.07 (±0.18) magnitude units smaller than the previous MLs and have better agreement with 12 Mws (3.15–4.49) determined by the UUSS and Saint Louis University.

Station Correction of P-Alert Network to Improve Magnitude Estimation for Earthquake Early Warning

<p>Magnitude estimation for earthquake early warning has been shown that it can be achieved by utilizing the relationship among the first three seconds P-wave amplitude, hypocentral distance and magnitude. However, the regression models in previous studies about P-Alert didn't include station correction factors, which may cause non-negligible effects. Thus, to improve the precision of magnitude estimation, we take station corrections into consideration when building the regression model. For the reason that station corrections are the unobserved latent variables of the model, we adopt the iteration regression method, which is based on the expectation-maximization algorithm, to determine them. By using this method, we are able to approach the values of both the station corrections and the coefficients of the regression model after several iterations. Our preliminary results show that after utilizing the iteration regression method, the standard deviation reduces from 0.30 to 0.26, and the station corrections we get range from -0.70 to 0.66.</p>

Crustal velocity structure beneath the NW Dinarides from 1-D hypocenter-velocity inversion

<p>The investigated area of the NW Dinarides is bordered by the Adriatic foreland, the Southern Alps, and the Pannonian basin at the NE corner of the Adriatic Sea. Its complex crustal structure is the result of interactions among different tectonic units. Despite numerous seismic studies taking place in this region, there still exists a need for a detailed, smaller scale study focusing mainly on the brittle part of the Earth's crust. Therefore, we decided to investigate the velocity structure of the crust using concepts of local earthquake tomography (LET) and minimum 1-D velocity model. Here, we present the results of the 1-D velocity modeling and the catalogue of the relocated seismicity. A minimum 1-D velocity model is computed by simultaneous inversion for hypocentral and velocity parameters together with seismic station corrections and represents the best fit to the observed arrival times.</p><p>We used 15,579 routinely picked P wave arrival times from 631 well-located earthquakes that occurred in Slovenia and in its immediate surroundings (mainly NW Croatia). Various initial 1-D velocity models, differing in velocity and layering, were used as input for velocity inversion in the VELEST program. We also varied several inversion parameters during the inversion runs. Most of the computed 1-D velocity models converged to a stable solution in the depth range between 0 and 25 km. We evaluated the inversion results using rigorous testing procedures and selected two best performing velocity models. Each of these models will be used independently as the initial model in the simultaneous hypocenter-velocity inversion for a 3-D velocity structure in LET. Based on the results of the 1-D velocity modeling, seismicity distribution, and tectonics, we divided the study area into three parts, redefined the earthquake-station geometry, and performed the inversion for each part separately. This way, we gained a better insight into the shallow velocity structure of each subregion and were able to demonstrate the differences among them.</p><p>Besides general structural implications and a potential to improve the results of LET, the new 1-D velocity models along with station corrections can also be used in fast routine earthquake location and to detect systematic travel time errors in seismological bulletins, as already shown by some studies using similar methods.</p>

Velocity models for routine earthquake monitoring at volcanoes: from deterministic to Bayesian

<p>Earthquake location is a primary function of volcano observatories worldwide and the resulting catalogs of seismicity are integral to interpretations and forecasts of volcanic activity.  Ensuring earthquake location accuracy is therefore of critical importance.  However, accurate earthquake locations require accurate velocity models, which are not always available.  In addition, difficulties involved in applying traditional velocity modeling methods often mean that earthquake locations are computed at volcanoes using velocity models not specific to the local volcano.   </p><p>Traditional linearized methods that jointly invert for earthquake locations, velocity structure, and station corrections depend critically on having reasonable starting values for the unknown parameters, which are then iteratively updated to minimize the data misfit.  However, these deterministic methods are susceptible to local minima and divergence, issues exacerbated by sparse seismic networks and/or poor data quality common at volcanoes.  In cases where independent prior constraints on local velocity structure are not available, these methods may result in systematic errors in velocity models and hypocenters, especially if the full range of possible starting values is not explored.  Furthermore, such solutions depend on subjective choices for model regularization and parameterization.</p><p>In contrast, Bayesian methods promise to avoid all these pitfalls.  Although these methods traditionally have been difficult to implement due to additional computational burdens, the increasing use and availability of High-Performance Computing resources mean widespread application of these methods is no longer prohibitively expensive.  In this presentation, we apply a Bayesian, hierarchical, trans-dimensional Markov chain Monte Carlo method to jointly solve for hypocentral parameters, 1D velocity structure, and station corrections using data from monitoring networks of varying quality at several volcanoes in the U.S. and South America.  We compare the results with those from a more traditional deterministic approach and show that the resulting velocity models produce more accurate earthquake locations.  Finally, we chart a path forward for more widespread adoption of the Bayesian approach, which may improve catalogs of volcanic seismicity at observatories worldwide. </p>

The ML scale in western Eurasian Arctic

The local magnitude scale ML was refined for the western part of the Eurasian Arctic on the basis of data from seismic stations operating on the archipelagos of Svalbard, Franz Josef Land, and Severnaya Zemlya: –logA0(R)=1.5*log(R/100)+1.0*10(–4)*(R–100)+3.0. Refinement was carried out on the basis of a sample of 167 earthquakes and 612 amplitude values at 5 seismic stations. The sample covered earthquakes that occurred in the main seismically active zones of the Eurasian Arctic for the period from January 2016 to April 2019. The refined scale can be applied in wide ranges of epicentral distances and magnitudes. The ML scale with the corresponding station corrections will be introduced into the practice of daily processing of seismological data from the western part of the Eurasian Arctic.

Relocation of earthquakes in the Southern and Eastern Alps (Austria, Italy) recorded by the dense, temporary SWATH–D network using a Markov chain Monte Carlo inversion

Local earthquakes with magnitudes in the range of 1–4.2 (ML) in the Southern and Eastern Alps (2017–2018) registered by the dense, temporary SWATH-D network and the AlpArray network reveal seismicity in the upper crust (0–20 km). The seismicity is characterized by pronounced clusters along the Alpine frontal thrust, e.g., Friuli-Venetia (FV) region, in the Giudicarie-Lessini (GL) and Schio-Vicenza domains, as well as in the Austroalpine Nappes and the Inntal area. Some seismicity also occurs along the Periadriatic Fault. The general pattern of seismicity reflects head-on convergence of the Adriatic Indenter with the Alpine orogenic crust. The deeper seismicity in the FV and GL regions indicate southward propagation of the Southern Alpine deformation front (blind thrusts). The first arrival-times of P- and S-waves of earthquakes are determined by an automatic workflow and then visually/manually checked and corrected. We applied a Markov chain Monte Carlo inversion method to achieve precise hypocenter locations of the 344 local earthquakes. This approach simultaneously calculates hypocenters, 1-D velocity model, and station-corrections without prior assumptions such as initial velocity models and earthquake locations. A further advantage of the method is the derivation of the model parameter uncertainties and noise levels of the data. The accuracy of the localization procedure is checked by inverting a synthetic travel-time dataset from a complex 3-D velocity model and using the real stations and earthquakes geometry. The location accuracy is further investigated by the relocation of quarry blasts. The average uncertainties of the locations of the earthquakes are below 500 m in the epicenter and ∼1.7 km in depth when using the average VP and VP/VS models and the station-corrections from the simultaneous inversion.

Recent microseismicity observed at Hekla volcano and first velocity inversion results

<p>Hekla is one of the most active and dangerous volcanoes in Iceland presenting a high hazard to air travel and a growing tourist population. Until now the pre-eruption warning time at Hekla is only around one hour.  In 2018 we installed the real-time seismic network HERSK directly on Hekla's edifice. If microseismicity on Hekla increases prior to the next eruption the network could possibly provide a means to improve early warning. In addition it is hoped that HERSK will better our understanding of the processes driving the evolution of pre-eruptive seismicity. The configuration and tuning of a dedicated real-time detection and location system requires the determination of a suitable velocity model and station corrections. We present a catalogue of recently detected local events that we use to invert for a 1-D velocity model. We observe significant variations in station corrections and conclude that it is important to account for these in the real-time detection and location system which we are developing based on the SeisComp3 software.</p>

Cálculo de Magnitudes e Relação Freqüência-Magnitudes dos Sismos de João Câmara, RN

RESUMO. As magnitudes dos principais sismos da série de João Câmara de 1986-1987 foram calculadas com estações regionais e telessísmicas. Correções das estações foram determinadas permitindo obter-se valores de magnitudes mais homogêneos e com menores desvios padrão. De agosto de 1986 a fevereiro de 1987, 30 sismos tiveram magnitudes maiores ou iguais a 3,5. A magnitude do maior sismo da série (30/11/86 às 05:19:48) foi m = 5,03 ± 0,05. Uma relação empírica entre magnitude e duração do sinal (m = c1 log D + c3) na estação JC01, em João Câmara, foi estabelecida permitindo um cálculo mais rápido de magnitude de microtremores. Para durações medidas até 1 mm pico-a-pico no sismograma, c1 = 2,05 e c3 = –1,61 para m ? 2. O exame das relações frequência-magnitude (log N = a – b m) indica que o coeficiente c1 deve ser menor para magnitudes abaixo de 2, aproximadamente. Para a atividade geral de João Câmara, foi encontrado um valor típico do parâmetro b de 1,12 ± 0,04. Não foi observada variação significativa no valor de b antes e depois do maior sismo de 30/11/1986.Palavras-chave: terremoto, onda de cauda, correções de estação, Rio Grande do Norte. DETERMINATION OF MAGNITUDES AND MAGNITUDE-FREQUENCY RELATION FOR THE EARTHQUAKES OF JOÃO CÂMARA, RNABSTRACT. Magnitudes of the major events of the 1986-1987 João Câmara earthquake swarm were calculated with regional and teleseismic stations. Station corrections were determined allowing more homogeneous magnitudes with smaller standard deviations. From August 1986 to February 1987, 30 events had magnitudes greater than 3.5. The largest (November 30, 1986 at 05:19:48) had m = 5.03 ± 0.05. An empirical relation between magnitude, m, and signal duration, D, (m = c1 log D + c3) at the local station JC01 was established allowing quick estimates of magnitudes for microearthquakes. For durations measured from the P arrival to coda amplitude of 1 mm peak-to-peak, c1 = 2.05 and c3 = –1.61 for magnitudes greater than about 2. The study of the frequency-magnitude relation (log N = a – b m) shows that the coefficient c1 must be smaller for magnitudes less than about 2. For the whole activity of João Câmara, a typical b-value of 1.12 ± 0.04 was found. No significant variation was observed in the b-value before and after the main event of November 30, 1986.Keywords: earthquake, coda wave, station corrections, Rio Grande do Norte State.