Displays

We have developed a 3D model of the Mount St Helens (MSH) magmatic plumbing system extending from the upper magma storage zone (> 3.5 km bsl) to Moho depths (40-45 km) by combining results from 2D and 3D active source seismic tomography and reflection imaging, and autocorrelation reflectivity imaging. The data are from the ~6000 high frequency seismographs used in the 2014 iMUSH active seismic experiment.

We developed a 3D Vp tomography model of melt distribution in the upper-middle crust (Kiser et al, 2018). The model suggests the plumbing system is a complex sill structure consisting of several interconnected bodies that lie beneath MSH at 3.5-14 km depth and that extend ~25 km laterally. Bright reflections in 3D autocorrelation reflectivity depth migrations are strongly correlated with the melt model, illuminating its interior as well as a system of more geographically extensive thin sills that are invisible to the tomography. High amplitude reflectivity occurs near the top of the sill complex, suggesting the system grows by successive emplacement at the top of the complex. Inversion of the autocorrelation reflection volume for melt content suggests melt concentrations exceed 30% locally in the sill complex.  The highly reflective center of the sill complex is likely the magma storage zone that feeds dacitic composition MSH eruptions. We speculate that some of the more geographically widespread dikes feed the Indian Heaven basalt fields.

Deeper reflectivity trends to the northeast of MSH and intersects the Lower Crustal Conductor in Bedrosian et al’s (2018) MT interpretation. They interpret high conductivity values as indicative of 3-10% interconnected melt in the crust at depths > 20 km, which is consistent with our reflectivity images. We also observe asymmetric crustal thickening toward and thinning away from MSH along the strike of the Cascades. Moho reflectivity is weak directly beneath MSH, agreeing with previous studies (Kiser et al, 2016; Hansen et al, 2016). Zones of strong autocorrelation and wide-angle reflectivity cross the refraction Moho and extend some distance into the upper mantle. 

How to cite: Levander, A. and Kiser, E.: iMUSH Autocorrelation Reflectivity and Active Seismic Imaging of the Magma Plumbing System under Mount St Helens, Washington, USA, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11763, https://doi.org/10.5194/egusphere-egu2020-11763, 2020.

The Hengill area experienced an intensive and long-lived series of earthquakes in the 1990s. This coincided with a period of inflation near the Hengill volcano, which was interpreted as new influx of magma at ~7 km depth. Feigl et al. (2000) postulated that the observed seismicity was triggered by the strain accumulation associated with the magma-influx. In a similar area ~3 km to the NW, subsidence has been occurring since 2006. The timing of this subsidence coincides with the onset of geothermal production at Hellisheidi in the west and enlargement of the Nesjavellir powerplant in the North. The source of the subsidence near Hengill volcano is however estimated between 5.6 and 7 km depth and at significant lateral distances from these production sites (Juncu et al. 2016). In this study we apply newly developed methods in time-dependent seismic tomography (Hobé et al. 2020) in the Hengill area, to study if significant velocity changes can be attributed to these inflation/deflation episodes. The dataset employed for the tomography covers the inflation period, the subsidence period, and the time in-between, with varying station coverage and geometry. In this study, the artificial velocity variations due to variations in source and receiver geometries are first separated from “true” velocity variations. In the approximate source region of the 2006-onwards deflation the preliminary results show a low Vp/Vs ratio anomaly between ~4-7 km depth, with an EW extent of ~8-10 km and an NS extent of ~4 km. This anomaly coincides with a significant amount of seismicity. This may indicate an increase in the amount of compressible fluids, accompanied with hydro-fracturing. The seismicity terminates below this low Vp/Vs anomaly, underneath which there is an area of increased Vp/Vs ratios (associated with melt) in the approximate center of the inflation episode in the 1990s. Thus, this investigation provides new information about the nature of the deformation sources, and the surrounding hydrothermal system. We will further investigate the apparent connection between the current subsidence and geothermal production.

References:

Feigl et al. (2000): Crustal deformation near Hengill volcano, Iceland 1993-1998: Coupling between magmatic activity and faulting inferred from elastic modeling of satellite radar interferograms, J. Geophys. Res.

Hobé et al. (2020): Imaging the 2010-2011 inflationary source at Krysuvik, SW Iceland, using time-dependent Vp/Vs tomography, WGC 2020, forthcoming

Juncu et al. (2016): Anthropogenic and natural ground deformation in the Hengill geothermal area, Iceland, J. Geophys. Res.

How to cite: Tryggvason, A., Hobé, A., Gudmundsson, O., and Geirsson, H. and the SIL Seismological Group: Relation of Time-Varying Vp/Vs ratio to Inflation and Deflation Episodes near Hengill Volcano, Iceland, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18971, https://doi.org/10.5194/egusphere-egu2020-18971, 2020.

Volcanic eruptions progress by co-evolving the fluid and solid systems. The fluid mechanics can be observed through the evolution of plumes and ejecta. How does the solid evolve? When does the conduit open? When does it close? Seismology can potentially tell us about these processes by measuring the failure of the solid rock. However, such inferences require detection of earthquakes during an explosive eruption. Standard earthquake detection methods often fail during this time as the eruption itself produces seismic noise that obscures the earthquakes. In this work, we address this problem by applying both a supervised and unsupervised search techniques to the existing catalog of the 2008 Okmok Caldera eruption to find brittle failure signals during the continuous eruptive sequence. We were able to detect >4500 new earthquakes using the 419 events previously located by the Alaska Volcano Observatory (AVO). A spatiotemporal analysis of the occurrence of earthquakes during the eruption reveal interesting observations: Seismic bursts during the eruption are not synchronized with the exhalation of large ash and steam plumes, suggesting that the dynamics of the eruption are controlled by a clog-and-crack mechanism; most of the Caldera co-eruptive seismicity that is not located at the focus of the eruption occurs under the intra-Caldera cones, showing the activation of their hydrological system due to a system-wide pressurization; the end of the eruption is marked by a large burst of small, deep earthquakes trending SW-NE, possibly related to a propagating lateral dike similar to those observed in other basaltic calderas; the magnitude distribution of seismicity through time shows that the largest earthquakes in the bursts do not happen at the beginning of the sequence like in typical mainshock-aftershock sequences. Furthermore, high precision earthquake relocations highlight a ring-fault structure inside of Okmok Caldera which is thought to be acting as the pathway for fluids to the surface.

How to cite: Garza-Giron, R., Brodsky, E., Spica, Z., and Haney, M.: Hidden earthquakes unveil the dynamic evolution of a large-scale explosive eruption, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-14124, https://doi.org/10.5194/egusphere-egu2020-14124, 2020.

Rotational seismology is an emerging field of seismology with rotational sensors such as blueSeis-3A as portable devices. We deployed one of these rotational sensors on Etna volcano from August to September 2019 in the middle of a 26 stations broadband seismic array and a fibre-optic cable deployed for Distributed Acoustic Sensing (DAS). We, therefore, recorded continuously the full seismic wavefield using a 6C station (rotational sensor co-located with a broadband seismometer) for 30 days.

We will present an overview of our work on the rotational data in combination with a broadband seismometer. We will (i) compare the translational with rotational data and show how they complement each other, (ii) calculate back azimuths using only a 6C station or using merely the horizontal components of the rotational sensor, (iii) determine Love and Rayleigh wave velocities from the rotation rate and (iv) perform a simple inversion for the shallow velocity structure below the station, and finally (v) discuss the usefulness of such a sensor in a volcanic environment and (vi) highlight what new it would bring to volcano-related research.

How to cite: Eibl, E. P. S., Currenti, G., Wassermann, J., Jousset, P., Vollmer, D., Larocca, G., Pellegrino, D., Pulviventi, M., Contrafatto, D., and Yuan, S.: Rotational sensor on a volcano: New insights from Etna, Italy, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18862, https://doi.org/10.5194/egusphere-egu2020-18862, 2020.

In volcanoes, topography, shallow heterogeneity, and even shallow morphology can substantially modify seismic coda signals. Coda waves are an essential tool to monitor eruption dynamics and model volcanic structures jointly and independently from velocity anomalies: it is thus fundamental to test their spatial sensitivity to seismic path effects. Here, we apply the Multiple Lapse Time Window Analysis (MLTWA) to measure the relative importance of scattering attenuation vs absorption at Mount St. Helens volcano (MSH) before its 2004 eruption. The results show the typical dominance of scattering attenuation in volcanoes at lower frequencies (3 - 6 Hz), while absorption is the primary attenuation mechanism at 12 Hz and 18 Hz. Still, the seismic albedo (measuring the ratio between seismic energy emitted and received from the area) is anomalously-high (0.95) at 3 Hz.

A radiative-transfer forward model of far- and near-field envelopes confirms this is due to strong near-receiver scattering enhancing anomalous phases in the intermediate and late coda across the 1980 debris avalanche and central crater. Only above this frequency and in the far-field, diffusion onsets at late lapse times.  We also implemented a layered model with a shallower layer with increased scattering properties to model late coda envelopes. While the broadening of late coda phases improves, this model cannot explain the phases of the intermediate coda with higher amplitude than the direct waves.

The scattering and absorption parameters derived from MLTWA are used as inputs to construct 2D frequency-dependent bulk sensitivity kernels for the S-wave coda in the multiple-scattering (using the Energy Transport Equations - ETE) and diffusive (AD, independent of MLTWA results) regimes. At 12 Hz, high coda-attenuation anomalies characterise the eastern side of the volcano using both kernels, in spatial correlation with low-velocity anomalies from literature. At 3 Hz, the anomalous albedo, the forward modelling, and the results of the tomographic imaging confirm that shallow heterogeneity beneath the extended 1980 debris-avalanche and crater enhance anomalous intermediate and late coda phases, mapping shallow geological contrasts.

The geomorphological map of MSH highlights extremely rough landforms (hummocky structures) of the already complex morphology of the debris avalanche. The comparison with the attenuation tomography reveals several matches, not only with the debris avalanche itself but also with other areas in the south flank of MSH, like the volcanoclastic plane, affected by intense eruptions in the past (e.g. Cougar stage, 28-18 ka).

We remark the effect those may have on coda-dependent source inversion and tomography, currently used across the world to image and monitor volcanoes.

How to cite: Gabrielli, S., De Siena, L., and Spagnolo, M.: Geomorphological controls on seismic recordings in volcanic areas, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-511, https://doi.org/10.5194/egusphere-egu2020-511, 2020.

Deep Long Period (DLP) earthquakes have been observed in many volcanic regions and are often considered as one of the important precursors to volcanic eruptions. At the same time, the physics of the source of these earthquakes remains unclear. We focus our study on Klyuchevskoy group of volcanoes in Kamchatka, Russia, one of the World’s most active volcanic system. The DLP earthquakes in this region occur at the limit between the lower crust and the upper mantle at depths of 30-35 km where ductile flow is expected to dominate rock deformation. Their occurrence also appears to correlate with the eruptive activity. Therefore, this is natural to consider that their generating mechanism is not related to brittle mechanism but rather to pressure fluctuations in the magmatic system as often suggest for the LP seismicity in general. We suggest a possible generating mechanism related to the rapid pressure changes caused by nucleation and growth of gas bubbles in response to the slow decompression of over-saturated magma. The pressure variation is simulated using the mathematical model of bubble nucleation and growth accounting for multiple dissolved volatiles (H₂O-CO₂) and diffusive gas transfer from magma into growing bubbles. Results of simulations show that fast pressure increase followed by its relaxation almost to its initial level is not very sensitive to the assumptions on the values of governing parameters. Typical pressure changes of a few tens of MPa in a volume of 3500 m³ occurring on time scales of fractions of a second to a second following bubble nucleation and growth can generate seismic waves with amplitudes similar to those recorded by seismographs in the vicinity of the Klyuchevskoy volcano.

How to cite: Shapiro, N. M., Melnik, O., Lyakhovsky, V., Galina, N., and Bergal-Kuvikas, O.: Bubble nucleation and growth in basaltic magmas as a possible source of Deep Long Period Volcanic earthquakes, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8251, https://doi.org/10.5194/egusphere-egu2020-8251, 2020.

Deep volcanic processes and magma intrusion episodes through the crust are typically accompanied by a variety of seismic signals, including volcano-tectonic (VT) seismicity, very long period (VLP) signals and deep low-frequency (DLF) events. These signals can reveal the migration of magma batches and the resonance of magma reservoirs and dikes. The recent 2018-2019 unrest offshore the island of Mayotte, Comoros archipelago, represents the first case of a geophysically monitored magmatic intrusion from a deep sub-Moho reservoir through the whole crust reaching the surface. At Mayotte, a huge magma movement and the following drainage of a deep reservoir were accompanied by a complex seismic sequence, including a massive VT swarm and energetic long-duration very long period (VLP) signals recorded globally. The identification and characterization of ~7000 VTs and ~400 VLPs by applying waveforms-based seismological methods allowed us to reconst the unrest phases: early VTs, migrating upward, were driven by the ascent of a magmatic dike, and tracked its propagating from Moho depth to the seafloor, while later VTs marked the progressive failure of the reservoir’s roof, triggering its resonance and the generation of long-duration VLPs. At the Eifel, Germany, weak DLFs earthquakes have been recorded over the last decades and located along a deep channel-like structure, extending from sub-Moho depth (~40-45 km) to the upper crust (~5-10 km). While not showing any clear migration, they reveal a different way of fluid transfer from depth towards the surface, possibly marking intermediate small reservoirs along a feeding channel. Here, brittle failure occurring in the vicinity of the reservoirs may cause their resonance. The Mayotte and Eifel observations are example of end member models for deep fluid transfer processes through the crust. These examples show that, by listening to seismic signals at different distances and by analysing them with modern waveform based methods, we can provide a detailed picture of deep magmatic processes and enable future eruption early warning.

How to cite: Cesca, S., Dahm, T., Heimann, S., Hensch, M., Letort, J., Razafindrakoto, H. N. T., Isken, M., and Rivalta, E.: The seismic sound of deep volcanic processes, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-6813, https://doi.org/10.5194/egusphere-egu2020-6813, 2020.

Seismic antennas are often used on volcanoes to analyse emergent signals as LP events or tremor.  In fact, they can be used for any kind of seismicity whether the signal is impulsive or emergent. In this work we are using a seismic antenna as an instrument for monitoring the continuous seismic signal, with the objective of a real-time application.

A seismic antenna composed of 5 broadband stations equipped with Guralp CMG-6TD stations was installed in November 2013 close to the summit of Merapi, on the site called Pasar Bubar. Sensors have a flat response characteristic from 30 s to the Nyquist frequency (50 Hz). This network has an aperture of 280 m. The shortest distance between sensors is 100 m.

In the perspective of a real-time application, the main analysis, which consists of estimating the slowness vector, requires a shorter computation time than the data acquisition time. We thus focused on a signal processing technique based on the calculation of time delays on the vertical component only and in a single frequency band. Given a set of time delays and associated errors calculated between each couple of sensors in the frequency domain, the corresponding slowness vectors can be recovered by inversion. Slowness vectors are estimated for successive time-windows in the frequency band 0.5-3 Hz. Temporal series of back-azimuth and apparent slowness are deduced with respect to time.

The analysis strategy for monitoring is then the following: A weight function expressed as a function of the derivatives of the time delays is calculated for successive moving time-windows. This function was designed in order to identify areas of stability of the back-azimuth values as function of time. A PDF of the back-azimuth and apparent slowness is then estimated for time series of 1 hour. This gives information on the dominant activity by time unit.

We will show the results obtained with the analysis of several months of continuous signal which are including different types of events generated by the on-going eruptive activity of Merapi: 1) volcano-tectonic events, 2) Multi-Phase (MP) events related with magma ascent in the conduit, 3) low-frequency events, 4) Rock-falls and 5) Pyroclastic density currents.

How to cite: Metaxian, J.-P., Santoso, A. B., Beauducel, F., Dahamna, N., Monteiller, V., and Fahmi, A.: Towards real-time monitoring with a seismic antenna at Merapi volcano, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-19068, https://doi.org/10.5194/egusphere-egu2020-19068, 2020.

Seismic activity at La Soufrière volcano of Guadeloupe is composed of various transient signals, which are classified manually by the Observatoire Volcanologique et Sismologique de Guadeloupe (OVSG-IPGP) considering waveforms recorded at several stations. Although five main types of signals are recognized in the data analysis by the observatory (Moretti et al., 2020), only three main classes readily distinguishable on seismic traces during the daily analytical protocol have been catalogued: Volcano-Tectonic events, Long-Period events and Nested events, each related to a distinct physical process.

Automatic classification of seismo-volcanic signals of La Soufrière was performed by using an architecture based on supervised learning, available at github.com/malfante/AAA. Seismic waveforms are transformed into a large set of features (34 features for each representation domain) computed from three representation domain of the signal (time, frequency, quefrency). The resulting vectors of features are then used for the modeling. We are using the Random Forest Classifier algorithm from the scikit-learn library.

At first, we trained the model with the dataset given by the OVSG consisting of 845 available labeled events (542 VT, 217 nested and 86 LP) recorded in the period 2013-2018. We obtained an average classification rate of 72 %. We determined that the VT class includes a variety of signals covering the LP, Nested and VT classes. Reviewing in details the waveforms and the spectral characteristics of the signals belonging to the 3 classes we then introduced Hybrid events and also defined a monochromatic class (so-called Tornillo) of LP signals, thus matching the full description of signals provided in Moretti et al. (2020).

Then, using the new information, a new model was trained with 5 classes and tested. We obtained a much better classification average rate of 84 %. The classification is excellent for Nested events (93 % of accuracy and precision) and Tornillo events (93% of accuracy and precision). The classification of VT events (90% accuracy, 89% precision) and LP events (86% accuracy, 82% precision) were also very good. The most difficult class to recognize is the Hybrid class (64 % accuracy, 69 % precision). Hybrid events are often mixed with VT and LP events. This may be explained by the nature of this class and the physical process that includes both a fracturing and a resonating component with different modal frequencies.

Machine learning is a powerful tool to handle large datasets. From a dataset built manually, the processing we applied allowed to obtain a reliable automatic classification by refining class definitions. This has important implications for observatory data processing during unrest and eruptive activity.

How to cite: Falcin, A., Metaxian, J.-P., Mars, J., Stutzmann, E., Moretti, R., and Komorowski, J.-C.: Automatic classification of seismo-volcanic signals at La Soufrière of Guadeloupe, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-10234, https://doi.org/10.5194/egusphere-egu2020-10234, 2020.

Cotopaxi volcano (5,897 m) is located in Central Ecuador, 50 km south of Quito. It has a long eruptive history including more than 70 eruptions with an estimated VEI between 2 and 4 since 1534. Its last low magnitude eruption occurred in 2015. The summit of the volcano is covered by a glacier down to about 5000 m elevation. The volcano is monitored by the Instituto Geofísico (IG) whose monitoring network includes permanent seismic stations. The closest station to the summit (BREF) is located 1 km below the summit (2.2 km distance), about 400 m from the base of the glacier. It is used as a reference station by the IG to characterize the seismicity. The station records transient events related to volcanic activity such as Long Period (LP) and Volcano Tectonic (VT) events, as well as icequakes (IQ) issued from the neighboring glacier. IQs may have various origins including fracture propagation or opening, collapse of ice blocks, basal friction or forced water flow within the glacier. These signals may be difficult to distinguish from VTs or LPs.

We examined data from station BREF recorded between January 2013 and October 2018, with the aim of identifying families of characteristic similar events. We applied a 3-step procedure including: (1) an automatic detection of transient events, (2) a classification of the detected events into families of similar events and (3) a re-composition of the temporal evolution of the largest families using matched-filtering. This procedure outlines the presence of numerous families and points out 4 characteristic temporal evolutions with respect of the 2015 eruption. These evolutions allow to distinguish precursory LP events from background seismicity and outline the presence of long lasting families which may persist for years. We use amplitude ratios calculated between BREF and a station more distant from the summit to distinguish shallow families from deeper ones. We also locate sources of long-lasting families with a seismic antenna installed at the foot of the glacier from April to September 2018. Locations indicate shallow sources below the glacier corresponding to IQs. These results confirm that background seismicity close to the summit of Cotopaxi is dominated by IQs. Temporal evolutions of these families also suggest that the large (Mw=7.8) subduction earthquake which occurred near Pedernales on April 16, 2016, 250 km from the volcano, had a stronger influence on the glacier or its shallow substratum than the 2015 eruption.

How to cite: Battaglia, J., Hidalgo, S., Helmstetter, A., Espín, C., Velez, L., Cordova, M., and Proaño, A.: Discriminating icequakes from volcanic seismicity at Cotopaxi volcano (Ecuador), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11749, https://doi.org/10.5194/egusphere-egu2020-11749, 2020.

Deception Island is the most active and documented volcano in the South Shetland Islands (Antarctica). Since its last eruption (1970) several experiments have targeted the reconstruction of its magmatic systems. Geophysical imaging has provided new insight into Deception's interior, particularly when using space-weighted seismic attenuation tomography for coda waves. Here, sensitivity kernels have been used to invert coda wave attenuation (Q_c⁻¹). We obtain a multifrequency-dependent model of the magmatic systems at Deception Island using active data, paying particularly attention to data selection and model optimisation. The results have been framed in the extensive knowledge of the tectonics and the geomorphology of the volcano with a GIS, underlining a spatial correlation between high-attenuation anomalies and high thermal activity regions. This inter- and multi-disciplinary analysis improves the interpretation of the dynamics of Deception Island and its related hazards.

How to cite: Guardo, R., De Siena, L., Caselli, A., Prudencio, J., and Ventura, G.: Space-weighted seismic attenuation multi-frequency tomography at Deception Island volcano (Antartica), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9986, https://doi.org/10.5194/egusphere-egu2020-9986, 2020.

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.

How to cite: Möllhoff, M., Rezaeifar, M., Bean, C. J., Vogfjörd, K. S., Bergsson, B. H., and Buxel, H.: Recent microseismicity observed at Hekla volcano and first velocity inversion results, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18954, https://doi.org/10.5194/egusphere-egu2020-18954, 2020.

Western South Iceland Seismic Zone (SISZ) plate boundary lies adjacent to the Hengill central volcano. The sinistral SISZ connects the two arms of the divergent Mid-Atlantic Ridge (MAR) plate boundaries (Western and Eastern Volcanic Zones; WVZ, EVZ), while Hengill is a part of the WVZ. Seismicity in western SISZ, also known as the Hjalli-Ölfus region, closely interacts with the seismicity and magmatism in Hengill. For instance, the  4 June 1998 Mw 5.4 Hengill earthquake witnessed aftershocks that extended south to meet the Hjalli-Ölfus segment. This segment then hosted the Mw 5.1 Hjalli-Ölfus earthquake that occurred on 13 November 1998; elucidating the Hengill-Ölfus interaction. Relative relocations of earthquakes from July 1991 to December 1999 in Hjalli-Ölfus indicate that the seismogenic zone is predominant at 4-8 km depth, with 80% of the events occuring along an ~ENE-WSW trending seismic zone with lateral extension of ~12 km. The remaining occur along N-S faults, much like the observed norm of dextral faulting along the rest of the SISZ (e.g., 17 June 2000, 29 May 2008 earthquakes; Árnadottir et al., 2001; Brandsdottir et al., 2010). These relocated earthquake sequences were used to perform stress inversions within specified spatio-temporal grids. The results show that from 1994 to 1997, the western part of the Hjalli-Ölfus region exhibits an oblique normal stress regime, while the eastern part remains consistently strike-slip in nature. From mid-1997 to June 1998 western Hjalli-Ölfus shifts from an oblique normal to a strike-slip stress regime, while the eastern part maintains the strike-slip character of the SISZ. However, two months after the 4 June 1998 Hengill earthquake, the western part shifts back to an oblique normal regime, which loses a part of its normal-faulting tendency after the 13 November 1998 Hjalli-Ölfus earthquake. This variation in stress fields between two significant events on conjugately oriented prodominantly strike-slip faults is a clear example of these features influencing one another between seismic episodes. 

How to cite: Bjarnason, I. Th., Parameswaran, R. M., and Thorbjarnardóttir, B. S.: Interseismic stress field variations in Hjalli-Ölfus, SW Iceland, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-8521, https://doi.org/10.5194/egusphere-egu2020-8521, 2020.

Geysers are hot springs whose frequency of water eruptions remain poorly understood. We setup a local broadband seismic network for one year at Strokkur geyser, Iceland, and developed an unprecedented catalog of 73,466 eruptions. We detected 50,135 single eruptions, but find that the geyser is also characterized by sets of up to six eruptions in quick succession. The number of single to sextuple eruptions exponentially decreased, while the mean waiting time after an eruption linearly increased (3.7 to 16.4 min). While secondary eruptions within double to sextuple eruptions have smaller mean seismic amplitudes, the amplitude of the first eruption is comparable for all eruption types. We statistically assess and model the eruption frequency assuming discharges proportional to the eruption multiplicity and a constant probability for subsequent events within a multi‐tuple eruption. We conclude that the waiting time after an eruption is predictable, but not the type or amplitude of the next one.

How to cite: Eibl, E. P. S., Hainzl, S., Vesely, N. I. K., Walter, T. R., Jousset, P., Hersir, G. P., and Dahm, T.: Seismic Eruption Catalog of Strokkur Geyser, Iceland, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-16535, https://doi.org/10.5194/egusphere-egu2020-16535, 2020.

The Great Geysir is within a tectonically active region bordering the eastern flank of the Western Volcanic Zone (WVZ), south of Langjökull glacier. The geothermal area has been active at least throughout the Holocene (Torfason, 1985). It is a high-temperature system, which is not common in Iceland for geothermal areas located outside the neovolcanic zones. Its longevity suggests continuously active tectonics in the region. Indeed, half a century of seismic monitoring shows relatively high activity of minor earthquakes (magnitude<4.0). The general pattern of seismicity is rather constant through time, but comes in bursts of activity. We attempt to elucidate the driving forces in this unusual and poorly tectonically understood area, by analyzing the most modern seismic data collected in the years 1995-2016 within a study area ~25x25 km² enclosing the Geysir area. It is, for instance, observed how the large (Mw~6.5) earthquakes in June 2000, located ~45 km south and southwest of Geysir in the South Iceland Seismic Zone (SISZ), induced seismicity kilometers away within hours after their occurrence. The heightened level of activity, an order of magnitude in terms of number of earthquakes, lasted half a year after the 2000 events in large parts of the study area and finally tapered down in 2001. Within the first two weeks of the 2000 events, the main activated faults are within 5 km of the Great Geysir. The activation is mostly at shallow depth (< 4 km). However, none pass directly through the Geysir geothermal area. That may explain the only minor change observed in the dormant state of the Great Geysir, which has now lasted approximately a century. There are historic accounts on how several large South Iceland earthquakes in the SISZ activated the Great Geysir, lasting for years or decades. The last such activation was in 1896. In its full might, it erupts up to a height of 70-80 m (Torfason, 1985).  Its currently active neighbor, Strokkur geysir, usually erupts to heights of 15-20 m.  Cross-sections of the seismicity near Geysir suggest several near vertical  right-lateral ~NNE trending faults.  Focal mechanisms indicate strike-slip movements, but also oblique-normal and thrust events in between. This may suggest fault jogs and high horizontal stresses. Approximately 6 km north of Geysir, in the Sandfell and Sandvatn area, there is a persistent ~ENE trending ~5 km long seismic pattern with main activity between 4-8 km depth. This seismicity has occurred, on and off, through the history of seismic observations. Here the faulting is also complicated (strike-slip and thrust), but focal mechanisms suggest the main component to be normal to oblique normal. Cross-sections, although unclear, suggest possible dip to the ~SSE. We intend to calculate stress inversions in the study area prior to the conference.

How to cite: Thorbjarnardóttir, B. S., Bjarnason, I. Th., and Parameswaran, R. M.: The Great Geysir and tectonic interactions in South Iceland, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-16388, https://doi.org/10.5194/egusphere-egu2020-16388, 2020.

The volcanic long-period seismicity, composed of long-period events and volcanic tremors, constitutes an important attribute of volcanic unrest. Its detection and characterization is therefore a key aspect of volcano monitoring. In the present work, a method based on the seismic network covariance matrix, the equivalent in the frequency domain of the cross-correlation matrix, is used to automatically detect and locate long-period events of the Teide volcano on the island of Tenerife (Canary Islands, Spain). The method is based on the analysis of eigenvalues and eigenvectors of the network covariance matrix.

Long-period events are detected through the time evolution of the width of the network covariance matrix eigenvalues distribution, which is a proxy of the number of sources acting in the wavefield. Each detected long-period event is then located using the moveout information of the corresponding first eigenvector. Three years of seismic data (from 2017 to 2019) continuously recorded by the Red Sísmica Canaria (C7), a permanent monitoring network composed of 17 broadband stations operated by the Instituto Volcanológico de Canarias (INVOLCAN), are analysed. The obtained locations are compared with potential locations from INVOLCAN’s catalog, obtained by a standard approach based on manual phases picking.

How to cite: Soubestre, J., D'Auria, L., Barrancos, J., Padilla, G. D., Seydoux, L., and Peréz, N. M.: Characterizing the long-period seismicity of Teide volcano in Tenerife (Canary Islands, Spain), EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-4714, https://doi.org/10.5194/egusphere-egu2020-4714, 2020.

The 11 November 2018 Mayotte event was first introduced in the media by Maya Wei-Haas (2018) on National Geographic Magazine as a strange earthquake of which seismic waves were recorded by instruments around the world, but unusually nobody felt them. The Mayotte event in the absence of body waves caused long-duration, long-period surface waves traveling around the globe. Cesca et al. (2020) by analyzing regional and global seismic and deformation data suggested drainage of a deep magma reservoir. Tono Research Institute of Earthquake Science recorded the data with the broadband seismometer (STS-1) and gravimeter (gPhone) installed in Mizunami, Japan (Murakami et al., 2019). The records by Iranian broadband stations clearly showed the long-period seismic signals around 10 (UTC) on November 11, 2018. We studied records by 26 stations distributed throughout the country. The stations are operated by National Center of Broadband Seismic Network of Iran, International Institute of Earthquake Engineering and Seismology (IIEES). Since the frequency content of Fourier amplitude spectra appeared the signal of the surface waves as a peak around 0.06 Hz, we applied a bandpass filter of 0.05-0.07 Hz to the waveform data. To separate Rayleigh from love in surface waves, the filtered horizontal components were rotated to the radial and transverse components based on an assumed epicenter location at the latitude of 12.7S and longitude of 45.4E degrees. The stations considered as an array and the investigation was carried out in two ways. First, the position of each station was taken as the reference point of the array coordinate, and arrival delay times at the other stations relative to the reference were calculated. The phase velocity and the back-azimuth of each station were estimated through the least-square regression method. The estimated back azimuths were within 13 degrees from the back azimuths from the assumed epicenter. The average phase velocity for Rayleigh and Love phases are calculated as 2.97 and 3.31 km/sec, respectively. Second, we applied semblance analysis to six stations with the shortest spacing distances. However, the distance between the adjacent stations relative to the signal wavelength was not enough short to prevent spatial aliasing. Nevertheless, the interesting was that the semblance results were different for radial and transverse components. We calculated surface-wave magnitude (Ms) for the event and a number of recorded earthquakes occurring in the Mayotte area from May 13 to June 1, 2018. Linear regression was used to define relationships between the calculated Ms and the USGS body-wave magnitude (mb) and the local magnitude by BRGM catalog (Bertil et al. , 2019), and the moment magnitude (Mw) from the CMT solutions of HRVD and USGS.

How to cite: Sadeghi, H. and Suzuki, S.: The 11 November 2018 Mayotte event was observed at the Iranian Broadband seismic stations, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-9767, https://doi.org/10.5194/egusphere-egu2020-9767, 2020.

Deep Low-Frequency earthquakes (DLFs) beneath volcanoes are possible evidence for deep-seated magmatic activities in the crust and uppermost mantle. After the 2011 Tohoku Earthquake (Mw 9.0), the number of DLFs beneath Zao volcano in the northeast Japan started increasing. The hypocenters of these DLFs form two clusters at shallow (20~28 km) and deep (28~38 km) depths. The shallow and deep clusters are located central and lower part of a high Vp/Vs zone, respectively (e.g., Okada et al., 2015), and the fact suggests different fluid involvement and source processes of DLFs at two clusters. In addition, after the activation of DLFs in 2012, increase in shallow ( < 2 km depth) seismicity has been observed since 2013, which implies the interaction of shallow and deep volcanic fluids. However, the small magnitude of DLFs makes it rather difficult to discuss detailed spatio-temporal characteristics of DLFs and focal mechanism of individual DLF. Therefore, in this study, we first detected DLFs using waveform correlation and determine their hypocenter, and then classified DLFs using waveform correlation to reveal the spatio-temporal characteristics and focal mechanism of each event type.

To detect DLFs, we applied the matched filter method to the continuous three-components waveform data recorded at stations operated by Tohoku Univ., NIED, and JMA. 146 DLFs listed in the JMA unified earthquake catalog between Jan. 2012 and Sept. 2016 were selected as templates. For each newly detected DLF, we estimated the differential arrival times using cross-correlation between the detected DLF and the template having maximum correlation, and determined the relative hypocenter using the master event method. As a result, we determined hypocenters of 1202 DLFs between Jan. 2012 to May 2018, which is about 4 times the number of DLFs listed in the JMA catalog.

We then classified newly detected DLFs using the hierarchical clustering method based on the waveform correlation, and classified 939 events into seven types (Type A: 241 events, B: 222, C: 295, D: 79, E: 42, F: 37, G: 23). The characteristics of individual waveform types are summarized as follows: Type C shows high frequency components (4-8 Hz) superimposed on the P wave, while the other types only have low frequency components (1-4 Hz); S-wave/P-wave spectral ratio of type C observed at each station shows larger azimuthal variation than that of the other types, and shows maximum peaks in northeast and southwest direction; Type C occurs mainly in the deep cluster while the other types occur in the shallow cluster; The activity of type C started in 2012 and showed rapid increase in 2015, while the other types show similar temporal changes in 2013 and 2016.

These results of this study suggest fluid transportation in the crust and different dynamic processes at each depth beneath the volcano.

How to cite: Ikegaya, T. and Yamamoto, M.: Spatio-temporal characteristics and focal mechanisms of deep low-frequency earthquakes beneath Zao volcano, Japan, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-12533, https://doi.org/10.5194/egusphere-egu2020-12533, 2020.

Sierra Negra is a basaltic shield volcano in the Galapagos Archipelago (Ecuador) and is the largest of the Galapagos volcanoes. The 2018 eruption was a complex event that included eruptive fissures opening on the northern rim and north-western flank. In this study, we report observations of seismic signals recorded on a temporary dense local network consisting of 14 seismometers and nearby permanent seismic stations, and analyze this data set to retrieve the source mechanisms of moderate pre- and co-eruptive seismic events (body-wave magnitude range of M3.5-5.3). Because of the shallow depths of the seismic events (<2 km) and short source-receiver distances (~1.5-10 km), that are comparable to or shorter than the wavelengths of radiated waves, the effect of near- and intermediate-field terms on dynamic displacements can be significant and hence the far-field assumption may not be well-suited for determining fault-plane solutions. Therefore, we pay special attention on the polarization properties of seismic waves excited at the near-field and intermediate-field ranges, and model and analyze complete displacement wave-fields to determine seismic sources. The source mechanism solutions are also interpreted in light of the volcanic unrest leading to the 2018 eruption, GPS observations, and reported regional centroid moment tensors.

How to cite: Nooshiri, N., Lokmer, I., Bean, C., Bell, A., Möllhoff, M., and Ruiz, M.: Source mechanisms of seismic events during the 2018 eruption of Sierra Negra Volcano (Galapagos) determined by using polarization properties of complete (near-field and far-field) body waves, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11297, https://doi.org/10.5194/egusphere-egu2020-11297, 2020.

Changes in external stress state and fluid content alter the mechanical properties of an geological media. Variations in seismic wave velocity can be used as proxies for changes in stress the onset of mechanical demage and/or possible fluid ingression. Temporal variations in seismic wave velocity have previously been monitored and observed prior to volcanic eruptions. In the absence of additional constraints related to stress or fluid changes on the volcano, these pre-eruptive changes are difficult to interpret and hence the causes of them are often not well understood. In this study, Coda Wave Interferometry (CWI) is used to measure time-lapse changes in seismic velocity on seismic multiplets (repeating similar earthquakes). In particular, we focus our analysis on using this technique to calculate the velocity changes on the data recorded prior to the 2018 eruption of Sierra Negra volcano, Galapagos Island. On 26th June 2018 at 09:15 UTC, a magnitude 5.3 earthquake occurred near the south-west caldera rim and an intense seismic swarm started around 17:15 UTC. Seismic tremor dominated at about 19:45 UTC, which marked the onset of the eruption. A very large seismicity sequence preceded the eruption. The pricise relationship between the magnitude 5.3 event and the eruption is not fully constraind. Here we search for multiplets in order to achieve high time resolution velocity change information in the hours between the large earthquake and the eruption. Our aim is to understand whether changes in seismic velocity measured with CWI on multiplets method provide new insight into the physical processes related to the eruption.

How to cite: Longobardi, M., Grannel, J., Bean, C., Bell, A., and Ruiz, M.: Multiplet Based Time Lapse Velocity Changes Prior to the 2018 Eruption of Sierra Negra Volcano, Galapagos Island Observed with Coda Wave Interferometry, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18213, https://doi.org/10.5194/egusphere-egu2020-18213, 2020.

Volcanic tremor are persistent seismic signals observed near active volcanoes. They are often associated with eruptions, although the exact relationships are not well constrained. To gain a better insight into the generation mechanisms of volcanic tremor, we study tremor that occurred during the 2018 eruption at Sierra Negra volcano, Galapagos. Located 1000 km west of continental Ecuador, Sierra Negra is a shield volcano with a large summit caldera and is one of the most active volcanoes in the Galapagos archipelago. The 2018 eruption started at about 19:55 UTC on 26th June and lasted about two months. Two tremor phases with very different frequency characteristics are identified before and after the eruption onset. The pre-eruptive phase is characterized by a narrow frequency band (2.5 – 4 Hz) and the co-eruptive phase has a broad frequency band (1 – 15 Hz). Location of the two phases by a seismic amplitude ratio method suggests that they are likely to be generated by different physical processes. The pre-eruptive phase is likely generated by dike opening while the co-eruptive phase is associated with lava flow. This interpretation is consistent with a time-lapse P-wave velocity structure of the volcano imaged by local-earthquake travel-time tomography.

How to cite: Li, K. L., Rezaeifar, M., Bean, C. J., Grannell, J., Bell, A., Ruiz, M., Hernandez, S., and Möllhoff, M.: Different mechanisms of the pre- and co-eruptive tremor during the 2018 eruption at Sierra Negra volcano, Galapagos, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-18975, https://doi.org/10.5194/egusphere-egu2020-18975, 2020.

Volcanic tremor is a feature of most explosive eruptions. Pre-eruptive tremors can be characterized by monotonic increases in the maximum frequency, frequency bandwidth and amplitude that are correlated with increases in gas flux from a volcanic vent. An enigmatic feature of this behavior is that is observed at volcanoes with widely ranging conduit geometries and structures. Accordingly, the ``magma wagging'' model introduced by [1] and extended by [2] hypothesizes an underlying mechanism that is only weakly-sensitive to volcano architecture: Within active volcanic conduits, the flow of gas through a permeable foamy annulus of gas bubbles excites and maintains an oscillation of a central magma column through a well-known Bernoulli effect. Furthermore, this oscillation has spectral properties that evolve depending on annulus thickness and permeability and the total flow of gas. 

In this thesis, we carry out a critical experimental test of the underlying mechanism for excitation. We explore the response of columns with prescribed elastic and linear damping properties to forced air annular airflows. From high-speed video measurements of linear and orbital displacements and time series of accelerometer measurements we characterize and understand the excitation, evolution, and steady-state oscillating behaviors of analog magma columns over a broad range of conditions. Where the time scale for damping is much longer than the natural period of free oscillation, column oscillation is continuously excited by relatively short period Bernoulli modes through a reverse energy cascade. We also identify three distinct classes of wagging: i. rotational modes that confirm predictions for whirling modes by [3]; as well as ii. mixed-mode; and iii. chaotic modes that are extensions to previous studies[1,2]. Our results show that rotational modes are favored for symmetric, and high-intensity forcing. Mixed-mode responses are favored for a symmetric and intermediate intensity forcing. Chaotic modes occur in asymmetric or low intensity forcing. To confirm and better understand our laboratory results and also extend them to conditions beyond what is possible in the laboratory we carry out two-dimensional numerical simulations of our analog experiments.

Taken together, results from our experimental and numerical studies can be extended to a natural system to make qualitative predictions testable in future studies of pre- and syn-eruptive volcano seismicity. Far before an eruption, the total gas flux is low and magma wags in a chaotic mode no matter what is the spatial distribution of the gas flux. At a pre-eruptive state, as gas flux increases, if the distribution of gas flux is approximately symmetric, we expect a transition to mixed and possibly rotational modes. During an eruption, fragmentation and explosions within the foamy annulus can cause spatial heterogeneity in permeability resulting in non-uniform gas flux that favors chaotic wagging behavior. 

[1] A. M. Jellinek and D. Bercovici. Seismic tremors and magma wagging during explosive volcanism. Nature, 470(7335):522-525, 2011

[2] D. Bercovici, A. M.  Jellinek, C. Michaut, and D. C. Roman. Volcanic tremors and magma wagging: gas flux interactions and forcing mechanism. Geophys. J.Int., 195(2):1001-1022, 2013

[3] Y. Liao and D. Bercovici. Magma wagging and whirling: excitation by gas flux. Geophys. J.Int., 215(1):713-735, 2018

How to cite: dehghanniri, V. and Jellinek, A. M.: An Experimental Study of Volcanic Tremor Driven by Magma Wagging, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-11365, https://doi.org/10.5194/egusphere-egu2020-11365, 2020.

Volcanoes and glaciers are both productive sources of seismic activity which may be easily confused for each other, leading to potential missed warnings or false alarms. This presents a challenge for organizations monitoring active volcanoes with glaciers on or near the edifice. Cryogenic earthquakes (i.e. icequakes) have been studied at only a few volcanoes around the world and there is a ready need to develop robust methods for efficiently differentiating them from volcanic events. Here we present results from an ongoing study of icequakes at active ice-covered volcanoes in the Southern Chilean Volcanic Zone. The primary focus of the project so far has been on seismo-acoustic data collected at Llaima volcano, one of the largest and most active volcanoes in the region. The data, recorded in 2015 and 2019, was analysed using a combination of automatic multi-station event detection and waveform cross-correlation to find candidate repeating icequakes. We identified 11 persistent families of repeating events in 2015, and over 50 families in 2019; the largest family containing over 1000 events from January to April 2019. The persistent, repetitive nature of these events combined with their waveform characteristics and source locations suggest they originated from multiple sub-glacial sources on the upper flanks of the volcano. Low levels of volcanic activity at Llaima volcano since 2009 have allowed this clear discrimination of icequake events. We are also targeting Villarrica volcano in early 2020 with a network of seismo-acoustic sensors and to record icequake activity in concurrence with the ongoing eruptive activity at the summit. Altogether, the results from this project so far suggest icequakes may be more common than previously thought and has strong implications for how seismic data at ice-covered volcanoes may be interpreted.

How to cite: Lamb, O., Lees, J., Franco Marin, L., Lazo, J., Rivera, A., Shore, M., and Lee, S.: Identifying icequakes at ice-covered volcanoes in Southern Chile, EGU General Assembly 2020, Online, 4–8 May 2020, EGU2020-851, https://doi.org/10.5194/egusphere-egu2020-851, 2020.