Seismo-volcanic Events Research Articles

Grid-search method for short-term over long-term average parameter tuning: an application to Stromboli explosion quakes

The collection of a significant catalogue of seismo-volcanic data involves the selection of relevant parts of raw signals, which can be automatised by using the short-term over long-term average (STA/LTA) method. The STA/LTA method employs the “Characteristic Function” to describe a section of a seismic record in terms of trace amplitude and first-time difference. This function is calculated in a short-term and long-term window; the ratio between the two windows defines a quantity that is controlled through threshold values, i.e., trigger on and trigger off. These threshold values indicate whether there is an increase in the energy in the seismic signal compared to the background noise. The common approach to the selection of the STA/LTA values is the adoption of literature-suggested ones. This could be a limitation as there may be cases in which a choice adapted to a specific raw signal may significantly help in the extraction of the relevant parts. To overcome the possible drawbacks of a non-adaptive choice imposed by such standard literature values, in this study, we propose a methodology for the automatic selection of STA/LTA values that can optimise the extraction of explosion quakes (EQs) from a seismo-volcanic raw signal. The values are obtained through a grid search over an index named quality–numerosity index (QNI) that measures the accordance in the automatic cuts and the consequent number of triggered seismo-volcanic events with the ones suggested by a human expert. The method was applied in the volcano domain for the specific application of the explosion quake signal extraction at Stromboli volcano. The experiments were conducted by selecting a subset of the dataset as training where to search for the best values, which were subsequently adopted in a test set. The results prove that the values suggested by our approach significantly improve the quality of the relevant part compared to the one extracted by adopting the values indicated in the literature. The methodology presented in this study can be applied to a wider typology of signals of volcanic, seismic, and other origin, potentially becoming a widely used approach in parameter optimisation processes.

Amplitude and inter-event time statistics for the island volcanoes Stromboli, Mount Etna, Yasur, and Whakaari

Detailed analyses of past major and minor seismo-volcanic events can help to understand the eruptive behavior of volcanoes and the underlying physical and chemical processes. Catalogs of these eruptions and, specifically, seismo-volcanic events may be generated using continuous seismic recordings at stations in the proximity of volcanoes. Here, we apply a recently-developed automated approach Adaptive-Window Volcanic Event Selection Analysis Module (AWESAM) to seismic data from Stromboli (Italy), Mount Etna (Italy), Yasur (Vanuatu) and Whakaari (New Zealand). We perform an inter-event time analysis to identify characteristic patterns in the events’ recurrence time and the volcanic activity. Using this identical approach for all volcanoes, we were able to discover that despite their differing types and activity, they exhibit similar statistical behaviors. For Whakaari, we noticed a bimodal inter-event time distribution for large events. Since this observation is based on single station data, further in-depth investigations are needed once more data is available in future. We also derive a new amplitude-frequency relationship from seismo-volcanic events. With this relation, we can confirm a change in slope for large events at Stromboli, which is based on 10 years of data. Additionally, we apply a classification model to events from Stromboli to differentiate between low-period (LP) events and high-frequency (HF) events and found an alternating behavior in the frequency of these events before and after the two paroxysms in 2019.

Successive variational mode decomposition-based enhanced Wigner-Ville distribution for seismo-volcanic events identification

Seismicity offers valuable information on the internal activity of volcanoes and the interpretation of seismic signals is crucial for volcano monitoring. As the seismic signal is multi-component & non-stationary, the time-frequency analysis like Wigner-Ville distribution (WVD) has played a significant role, but the presence of cross-term limits the applicability of WVD and other bilinear time-frequency representations (TFRs). This paper introduces a successive variational mode decomposition (SVMD)-based cross-term reduction method in WVD. The proposed method is robust and accurate due to reduced sensitivity to initialization parameters, lesser computational burden, and automatic selection of the number of decomposed modes. SVMD-based WVD outperforms other related signal decomposition-based methods for removing cross-terms and reconstructing auto-terms, as demonstrated by similarity and concentration-based performance measures for synthetic multi-component and seismic signals. SVMD-based WVD is also applied for the seismo-volcanic events identification system development and evaluated on the Llaima volcano dataset to find the applicability of the proposed method in real-world field data. Novel TFR-based features are extracted based on 3-D segmentation of TFR and 2-D normalized correlation between the segments. Feature ranking and SVM classifier optimization result in an overall classification accuracy of 97.10%, surpassing previous approaches on the same dataset.

Distributed dynamic strain sensing of very long period and long period events on telecom fiber-optic cables at Vulcano, Italy

Volcano-seismic signals can help for volcanic hazard estimation and eruption forecasting. However, the underlying mechanism for their low frequency components is still a matter of debate. Here, we show signatures of dynamic strain records from Distributed Acoustic Sensing in the low frequencies of volcanic signals at Vulcano Island, Italy. Signs of unrest have been observed since September 2021, with CO2 degassing and occurrence of long period and very long period events. We interrogated a fiber-optic telecommunication cable on-shore and off-shore linking Vulcano Island to Sicily. We explore various approaches to automatically detect seismo-volcanic events both adapting conventional algorithms and using machine learning techniques. During one month of acquisition, we found 1488 events with a great variety of waveforms composed of two main frequency bands (from 0.1 to 0.2 Hz and from 3 to 5 Hz) with various relative amplitudes. On the basis of spectral signature and family classification, we propose a model in which gas accumulates in the hydrothermal system and is released through a series of resonating fractures until the surface. Our findings demonstrate that fiber optic telecom cables in association with cutting-edge machine learning algorithms contribute to a better understanding and monitoring of volcanic hydrothermal systems.

Spatiotemporal Changes in Fault Displacements Associated with Seismovolcanic Events in and Around Miyakejima and Kozushima in 2000 Inferred from GNSS Data

Spatiotemporal Changes in Fault Displacements Associated with Seismovolcanic Events in and Around Miyakejima and Kozushima in 2000 Inferred from GNSS Data

Automated Seismo-Volcanic Event Detection Applied to Stromboli (Italy)

Many active volcanoes exhibit Strombolian activity, which is typically characterized by relatively frequent mild volcanic explosions and also by rare and much more destructive major explosions and paroxysms. Detailed analyses of past major and minor events can help to understand the eruptive behavior of volcanoes and the underlying physical and chemical processes. Catalogs of these eruptions and, specifically, seismo-volcanic events may be generated using continuous seismic recordings at stations in the proximity of volcanoes. However, in many cases, the analysis of the recordings relies heavily on the manual picking of events by human experts. Recently developed Machine Learning-based approaches require large training data sets which may not be available a priori. Here, we propose an alternative user-friendly, time-saving, automated approach labelled as: the Adaptive-Window Volcanic Event Selection Analysis Module (AWESAM). This strategy of creating seismo-volcanic event catalogs consists of three main steps: 1) identification of potential volcanic events based on squared ground-velocity amplitudes, an adaptive MaxFilter, and a prominence threshold. 2) catalog consolidation by comparing and verifying the initial detections based on recordings from two different seismic stations. 3) identification and exclusion of signals from regional tectonic earthquakes. The strength of the python package is the reliable detection of very small and frequent events as well as major explosions and paroxysms. Here, it is applied to publicly accessible continuous seismic recordings from two almost equidistant stations at Stromboli volcano in Italy. We tested AWESAM by comparison with a hand-picked catalog and found that around 95% of the seismo-volcanic events with a signal-to-noise ratio above three are detected. In a first application, we derive a new amplitude-frequency relationship from over 290.000 seismo-volcanic events at Stromboli during 2019–2020 which were detected by AWESAM. The module allows for a straightforward generalization and application to other volcanoes with frequent Strombolian activity worldwide. Furthermore, this module can be implemented for volcanoes with rarer explosions.

The unexpected Mayotte 2018–2020 seismic sequence: a reappraisal of the regional seismicity of the Comoros

The Mayotte seismic sequence that started on 10 May 2018, with a main shock of magnitude Mw 5.9 on May 15, followed by a major offshore volcanic activity, raises several questions of seismovolcanic hazards in the Comoros region. The unexpected size and duration of the crisis is an opportunity to reassess the distribution and magnitude of the seismicity near Mayotte Island, but also regionally. We present a comprehensive seismicity catalogue of the region including the Mozambique Channel, the Mozambique coast and Madagascar, based partly on previously published data in the region and partly on unpublished data from local catalogues for the Comoros and Madagascar. Our catalogue extends from 1900 onward with a completeness of magnitude 5.5 until 1980s and decreasing only recently to 4.5 after 2010. It comprises the events of magnitude Mlv ≥3.5 for the seismic sequence of Mayotte from May 2018 to October 2020 as the crisis is still ongoing. Present knowledge of the seismicity, largely partial in distribution and magnitude before 1980, makes the seismic sequence of 2018-2020 an exceptional and unprecedented seismo-volcanic event in the region. We discuss the distribution of seismicity in time and space within the context of the south eastward propagation of the East African rift system towards Madagascar.

Advanced signal recognition methods applied to seismo-volcanic events from Planchon Peteroa Volcanic Complex: Deep Neural Network classifier

Advanced techniques in the recognition and classification of seismo-volcanic events are transcendental when studying active volcanoes, not only for their importance as an accurate real time seismic monitoring procedure but also for the use of their results in modeling the dynamics of the volcanic environment. It is well known that real time seismic monitoring deals with such a large amount of data that it would become an overwhelming job for an operator to do manually. Therefore the use of automatic detection and classification techniques based on the Machine Learning approach are suitable in meeting such a challenge.The aim of this work is to test the capability of the Deep Neural Network (DNN) by using different event parametrization as a confident classifier tool that could permit a reliable seismic catalog to be built in a new and un-analyzed volcanic scenario. We tested different configurations in order to build an approach that was as simple as possible to use this classifier with a limited number of events. In this regard, the feature space was explored in order to select the most significant parameters of the seismic signals. The data used for this analysis corresponds to the Planchon Peteroa Volcanic Complex (PPVC) located in the Transitional Southern Volcanic Zone (TSVZ) between Chile and Argentina, South America. The most significant result of this work was not only that it provided an analysis in terms of performance of this algorithm, especially when the training, validation and test dataset is reliable although definitely reduced, but it also gave an insight of into how an optimal event parametrization can significantly improve the automatic detection and classification of seismo-volcanic events.

A deep learning approach for automatic recognition of seismo-volcanic events at the Cotopaxi volcano

The research for developing an automatic recognition system of volcanic microearthquakes have been an important task around the world, based on this, the aim of this paper is to present an automatic recognition system of microearthquakes from the Cotopaxi Volcano based on a deep learning approach. The detection and classification stages were carried out with Convolutional Neural Networks by using spectrograms, which were generated according to the theory of periodograms with different types of windows. In order to enable the training of neural networks with a small database (1187 microearthquakes), the Transfer Learning process was used. This system operates in quasi-realtime, which is able to process records of 20min, accordingly to the requirements of the Instituto Geofísico de la Escuela Politécnica Nacional, with a recognition (detection + classification) time response of one minute, approximately. The system performance presents an accuracy of 99% in the detection stage and an accuracy of 97% in the classification stage.

Genesis and mechanisms controlling tornillo seismo-volcanic events in volcanic areas

Volcanic activity is often preceded or accompanied by different types of seismo-volcanic signals. Among these signals, the so-called tornillo (Spanish for “screw”) events are considered to belong to a unique class of volcano-seismicity characterised by a long-duration coda, amplitude modulation and high-quality factor. These data constitute important evidence for the gas fraction inside magmatic fluids. However, the mechanism behind this unique signal remains not fully understood. Here we report new laboratory evidence showing that two different processes have either scale-invariant or scale-dependent effects in generating tornillo-like events. These processes are respectively the gas pressure gradient, which triggers the event and regulates the slow decaying coda, and the fluid resonance into small scale structures which, in turn, control the frequency content of the signal. Considering that the gas pressure gradient is proportional to the fluid flow, these new findings, as applied to volcanoes, provide new information to better quantify both gas rate and volume, and the dimension of the resonator.

Performance of the ‘material Failure Forecast Method’ in real-time situations: A Bayesian approach applied on effusive and explosive eruptions

Most attempts of deterministic eruption forecasting are based on the material Failure Forecast Method (FFM). This method assumes that a precursory observable, such as the rate of seismic activity, can be described by a simple power law which presents a singularity at a time close to the eruption onset. Until now, this method has been applied only in a small number of cases, generally for forecasts in hindsight. In this paper, a rigorous Bayesian approach of the FFM designed for real-time applications is applied. Using an automatic recognition system, seismo-volcanic events are detected and classified according to their physical mechanism and time series of probability distributions of the rates of events are calculated. At each time of observation, a Bayesian inversion provides estimations of the exponent of the power law and of the time of eruption, together with their probability density functions. Two criteria are defined in order to evaluate the quality and reliability of the forecasts.Our automated procedure has allowed the analysis of long, continuous seismic time series: 13years from Volcán de Colima, Mexico, 10years from Piton de la Fournaise, Reunion Island, France, and several months from Merapi volcano, Java, Indonesia. The new forecasting approach has been applied to 64 pre-eruptive sequences which present various types of dominant seismic activity (volcano-tectonic or long-period events) and patterns of seismicity with different level of complexity. This has allowed us to test the FFM assumptions, to determine in which conditions the method can be applied, and to quantify the success rate of the forecasts. 62% of the precursory sequences analysed are suitable for the application of FFM and half of the total number of eruptions are successfully forecast in hindsight. In real-time, the method allows for the successful forecast of 36% of all the eruptions considered. Nevertheless, real-time forecasts are successful for 83% of the cases that fulfil the reliability criteria. Therefore, good confidence on the method is obtained when the reliability criteria are met.

Evidence of persistent seismo-volcanic activity at Marsili seamount

<p>The Marsili submarine volcano is the largest European volcano, and it can be considered as the key to our understanding of the dynamics of the spreading and back-arc lithosphere formation in the Tyrrhenian sector [Marani et al. 2004, and references therein]. Despite its size, it is very difficult to monitor due to its geographical position [D'Alessandro et al. 2011], and it still remains little known. In 2006, the Centro Nazionale Terremoti (National Earthquake Centre) of the Istituto Nazionale di Geofisica e Vulcanologia (INGV) deployed a broadband ocean-bottom seismometer with hydrophone (OBS/H) [Mangano et al. 2011] on the flat top of Marsili volcano, at a depth of ca. 790 m. In only nine days, the instrument recorded ca. 800 seismo-volcanic events [D'Alessandro et al. 2009]. This revealed the intense seismo-volcanic activity of Marsili volcano for the first time. […]</p><p> </p>

Integration of transient strain events with models of plate coupling and areas of great earthquakes in southwest Japan

We model the crustal deformation caused by two long-term subduction slip transients in southwest Japan, which we refer to as the 2000–2004 Tokai and the 2002–2004 Bungo Channel slow slip events (SSEs). We use re-analysed GEONET position time-series, and a Kalman filter based network inversion method to image the spatiotemporal slip variation of the two events on the plate interface during the period of 1998–2004.67 and 2000–2005. Both events are found to have complex slip histories with multiple subevents. In addition to a newly identified slip subevent in 2002–2003, we find that the major event in the Bungo Channel SSE initiated in early 2003 beneath the northeastern corner of the region and expanded southwestward, in contrast to the slip characteristics suggested by other studies. The re-analysed GPS data in the Tokai region shows a renewed slip activity for the Tokai SSE in early 2003–2004 at a similar location as in the period of 2001–2002. The equivalent M_w for both the Tokai and Bungo Channel SSEs are about 7.0. Our results show that the Tokai SSE appears to start before the Miyaki-Kozu seismovolcanic event. Integrating plate coupling and SSEs shows that the transient slip zones are located in a region between the locked zones and the epicentres of the low frequency earthquakes (LFEs). At least part of the interseismic slip deficit is released by episodic SSEs beneath the Bungo Channel region. We find excellent temporal correspondence between transient slip and adjacent LFEs for both SSE, suggesting that they are closely related and possibly reflect that long-term slow slip may modulate the occurrence of LFEs.

Explosive volcanism during the Holocene in the Upper Limay River Basin: The effects of ashfalls on human societies, Northern Patagonia, Argentina

In order to understand human response to Holocene ashfall events, tephra layers found in archaeological sites along the upper Limay River basin, Northern Patagonia, Argentina were bracketed with radiocarbon dates and correlated with tephra from a lacustrine sediment core and from outcrops and the archaeological evidence was analyzed. A dark tephra associated with seismic activity was identified in the El Trébol rockshelter filling interstices between fallen blocks together with remains of human activity and bones of extinct fauna, marking a seismo-volcanic event occurring between 11 758 and 12 866 cal yr BP. This same seismic event affected the Cuyín Manzano site, where the roof collapse made the site uninhabitable for a time. A white tephra, present in Epullán Grande, Epullán Chica and Traful I caves is correlated with Nahuel Huapi tephra (NHT), equivalent to Laya's Río Blanco and Río Pereyra members, (Río Pireco Formation). NHT is considered to have been derived from the same eruptive event, with dates ranging between ca. 1950–2500 cal yr BP. A dark tephra from Ortega's cave and several tephra from Puerto Tranquilo I rockshelter ranging between 521 and 2069 cal yr BP show how the mobile hunters–gatherers of Northern Patagonia were able to cope with the changing circumstances.

Costa Recta beach, Deception Island, West Antarctica: a retreated scarp of a submarine fault?

Deception Island (South Shetlands, Antarctica) is one of the most active volcanoes in Antarctica, having erupted recently in 1967, 1969 and 1970, damaging scientific stations on the island. It is also seismically very active. The island has attracted the attention of many researchers as it constitutes an undisturbed natural laboratory to study seismo-volcanic events and how they affect landscape modelling and evolution. One of the most remarkable geological and geomorphological features on Deception Island is the linearity of its easternmost coastal landform, the origin of which remains unknown. Some answers, based on presence of strike-slip fault or on the ice cap and beach geomorphological dynamics, have been reported in the literature. Our new work provides several indications of the existence of a dip-slip submarine fault, parallel to the coast (NNW–SSE), which suggests a tectonic origin for this morphological feature. Uplifted marine terraces, incision of a fluvial network over the ice cap, normal faulting parallel to the coast in the north and south rock heads bounding the beach and sharp shelf-break with rather constant slope, constitute some of this evidence. Terrace uplift and fluvial channel incision decreasing southward from Macaroni Point, indicates possible tilt movement across this inferred fault plane.

Slow slip in the focal region of the anticipated Tokai earthquake following the seismo-volcanic event in the northern Izu Islands in 2000

Abstract Transient crustal deformation occurred in the regions of Kanto and Tokai during the seismo-volcanic event in the northern Izu Islands in 2000. In our investigation of the observed deformation, we constructed an optimum-source model of the event between Miyake and Kozu Islands. We then made an inversion analysis of the differences between the observed displacement field and the calculated displacement field from the optimum model, assuming that the differences were caused by the changes in the interplate coupling beneath the Tokai region. From the inversion analysis of data for each of three-month periods, May to August, June to September, and July to October, we found decreased interplate coupling in the early stages of the 2000 event. In the first stage, either a slow slip or a temporary suspension of the plate subduction occurred in the focal region of the anticipated Tokai earthquake. The area then extended to the west and, finally, a slow slip exceeded the secular convergence velocity on the plate interface near Lake Hamana in the fall of 2000. We believe this ongoing slow slip began in August or early September 2000.

Kuwae (≈ 1425 A.D.): the forgotten caldera

In Vanuatu, Tongoa and Epi islands once formed part of a larger landmass, Kuwae, which was partly destroyed during a cataclysmic seismo-volcanic event that is recorded in local folklore. It led to the formation of a 12-km-long and 6-km-wide oval-shaped submarine caldera with two distinct basins and a total area of ∼60 km2 at the level of the rim.The age of this eruption, 1420–1430 A.D., and the structure of the related collapse are discussed and a composite log (143 m) of the pyroclastics surrounding the caldera is presented. They comprise thick hydromagmatic deposits belonging to a terminal hydromagmatic phase of the pre-caldera edifice, which grade upwards into two major sequences of pyroclastic flow deposits, clearly related to the caldera event. Collapse near the caldera edge was at least in the range 650 to 950 m, and may have been as much as 800 to 1100 m. The volume of rocks engulfed during the caldera formation is ∼ 32–39 km3, suggesting the same volume of magma was erupted.Even if two coalescent collapse structures were formed, it is worth noting that the Kuwae caldera is not a reactivated structure, but the result of a single event of short duration which occurred in the first half of the Fifteenth century. This event is one of the seven biggest caldera-forming events during the last 10,000 years, and is comparable with the Santorini Minoan eruption and the Crater Lake eruption.

Contemporary, Holocene, and Quaternary deformation of the Asal Rift, Djibouti: Implications for the mechanics of slow spreading ridges

Because the frequency and character of rifting events along mid‐ocean ridges are largely unknown, how the repetition of such events gives rise to rift structures is unexplored. The Asal rift in the Afar depression of Djibouti, Africa, provides the world's best subaerial analogue for young slow spreading mid‐ocean ridges. Seismic, geodetic, and field observations of a seismovolcanic event in 1978 at Asal yield estimates of the fault and dike locations, geometry, displacement, and volume of basalt extruded in a rifting event. A 6–9 kyr‐old lake shore highstand at Asal has been warped downward by 70 m, providing a Holocene measure of the vertical deformation across the rift. The rift topography furnishes an older datum, which we infer to be 34±6 kyr old using the Holocene deformation rate. We find that faults throughout the rift valley are active; Holocene slip rates diminish beyond 4 km from the rift axis; late Quaternary rates decrease beyond 6–7 km. The Holocene slip rates are used to estimate repeat times by taking the displacement on the faults which slipped in 1978 as characteristic; we find tectonic events on individual faults recur every 200– 300 years. Half the rift faults slipped together in the 1978 event. If this is typical, then groups of faults are activated every 100–150 years. We suggest that half the events take place in the rift axis accompanied by volcanic extrusion; the remainder occur peripheral to the neovolcanic zone and involve fault slip only, both events having a repeat time of 200–300 years. Given the 10 km width of the rift and its 16 mm yr−1 spreading rate, the mean age of the material in the rift should be ∼350 kyr, an order of magnitude older than the inferred age of the formation of the rift topography. The subsidence rate of the rift axis during the past 35 kyr is 8–9 mm yr−1, with the rate of infilling by volcanic extrusion <1 mm yr−1. The resulting net subsidence rate, about equal to the half‐spreading rate of the rift, could not be sustained for 300 kyr without significant infilling by lavas. Thus both observations suggest that the long‐term vertical deformation in the rift has not been steady state. Instead, we suggest that there is a rifting/filling cycle at Asal, with the most recent filling episode ending ∼35 kyr.