Dissertations / Theses on the topic 'Volcano seismology'

Electrical activity at volcanoes has been recently recognized as a potential new remote sensing technique for plume-forming eruptions. Volcanic electrical activity takes place in the conduit and plume and therefore has the benefit of being a direct indicator of surface activity. This is unlike seismic signals, which indicate magma/gas movement underground, and infrasound signals, which indicate a surface explosion but not necessarily the formation of an ash plume. There are two distinct types of volcanic electrical discharges: volcanic lightning and continual radio frequency (CRF) impulses. This dissertation explores the relationships between these two electrical signals and other commonly monitored volcanic parameters. For volcanic electrical activity to be widely adopted into monitoring platforms it is important to understand how electrical discharges at volcanoes are related to other monitored signals. I present a case study of the electrical activity at Sakurajima Volcano, Japan. The lightning mapping array (LMA) is used to record both lightning and CRF. I relate CRF to ash properties and show that CRF corresponds to eruptions containing more juvenile magma that has undergone milling as it is transported out of the conduit. Seismic, infrasound, and video data are used in conjunction with multivariable statistical methods on a suite of electrical parameters to show that high levels of volcanic electrical activity are related to eruptions with large infrasound signals (> 107 J), high initial velocities (> 55 m/s), and relatively tall plume heights (> 1 km). Finally, an examination of globally detected lightning at Bogoslof Volcano, AK shows the potential for volcanic lightning in plume tracking (0-100 km), even after the end of the explosive phase of the eruption.

Dans l'histoire des éruptions volcaniques, le Papandayan à l'Ouest de Java est considéré comme l'un des plus meurtriers après avoir causé la mort de 2957 personnes et des dégâts sérieux en 1772. L'éruption la plus récente de ce volcan a eu lieu en 2002 et était de type phréatique. Cette éruption a été précédée d'une augmentation soudaine de l'activité sismique moins de deux jours avant l'éruption. Aucune victime n'a été déplorée. La nature de cette éruption est indéfinie. Cette thèse regroupe plusieurs études utilisant différentes techniques en vue d'améliorer la prédictibilité des éruptions du volcan Papandayan, principalement via l'interprétation des signatures sismiques.

Le monitoring sismique passif a débuté en décembre 2009 par l'installation d'une station sismique permanente à large bande dans le cratère du Papandayan. L'année suivante, une station météorologique a été installée pour compléter les mesures. La troisième année, 8 stations sismiques temporaires ont été déployées autour du volcan en réponse à une augmentation de l'activité sismique en 2011.

Nous avons conduit différentes études; (1) Nous avons examiné l'évolution de l'activité volcanique par réalisation d'une revue complète de l'histoire éruptive du volcan, autant pour la période préhistorique qu'historique. (2) Nous avons réalisé une analyse temps-fréquence des événements sismiques, étudié leurs caractéristiques et proposé une nouvelle classification avec une description des processus physiques supposés les générer. (3) Nous avons étudié les signatures sismiques précurseur de l'éruption de 2002 et pendant la crise volcanique de 2011 en implémentant différentes méthodologies, dont: la détection automatique d'événements sismiques à l'aide de filtres récursifs STA/LTA, l'analyse spectrale des formes d'onde, la mesure continue de l'amplitude spectrale du signal (SSAM), la polarisation des ondes et l'analyse de la distribution fréquence/magnitude (b-value). Nous avons alors réalisé un modèle chronologique des séquences sismiques du Papandayan. (4) Pour améliorer la compréhension de la dynamique des fluides sous le volcan Papandayan, nous avons réalisé une analyse des fréquences complexes des événements longue période (LP) et leurs variations temporelles peuvent être utilisées pour estimer (a) la composition des fluides présents dans les fractures sous le volcan et/ou (b) l'évolution des dimensions de ces fractures. Ces variations des fréquences complexes des événements LP peuvent être interprétées comme les réponses dynamiques du système hydrothermal à des changements d'impulsions de chaleur transférées par les flux de gaz volcaniques du magma sous le volcan. (5) nous avons calculé l'évolution temporelle du rapport spectral horizontal-sur-vertical (HVSR) en utilisant le bruit sismique ambiant enregistré par une station unique pour estimer les variations de vitesse de propagation des ondes de cisaillement en lien avec l'activité dynamique du volcan. Nous avons trouvé une corrélation claire entre les variations de fréquence de résonnance HVSR et l'augmentation de la sismicité.

Enfin, nous proposons des hypothèses sur les processus physiques qui se produisent sous le Papandayan. Cette étude est une première tentative d'utilisation de cette méthode pour surveiller l'activité volcanique en continu.

Doctorat en Sciences
info:eu-repo/semantics/nonPublished

Concepción is the most active composite volcano in Nicaragua, and is located on Ometepe Island, within Lake Nicaragua. Moderate to small volcanic explosions with a volcanic explosivity index (VEI) of 1-2 have been characteristic of this volcano during the last four decades. Although its current activity is not violent, its volcanic deposits reveal stages of violent activity involving Plinian and sub-Plinian eruptions that deposited vast amounts of volcanic tephra in the Atlantic Ocean. These observations, together with the 31,000 people living on the island, make Concepción volcano an important target for volcanological research. My research focuses on the investigation of the stability of the volcano edifice of Concepción, using geophysical data such as gravity, geodetic global positioning system (GPS), sulphur dioxide (SO2) flux, real-time seismic amplitude (RSAM), and satellite remotely-sensed data. The integration of these data sets provides information about the short-term behavior of Concepción, and some insights into the volcano's long-term behavior. This study has provided, for the first time, information about the shallow dynamics of Concepción on time scales of days to weeks. I furnish evidence that this volcano is not gravitationally spreading in a continuous fashion as previously thought, that its bulk average density is comparable to that of a pile of gravel, that the volcano edifice is composed of two major distinctive lithologies, that the deformation field around the volcano is recoverable in a matter of days, and that the deformation source is located in the shallow crust. This source is also degassing through the relatively open magmatic conduit. There are, however, several remaining questions. Although the volcano is not spreading continuously there is the possibility that gravitational spreading may be taking place in a stick-slip fashion. This has important implications for slope stability of the volcano, and the associated hazards. The factors influencing the long term slope stability of the volcano are still not fully resolved, but internal volcanic processes and anthropogenic disturbances appear to be the major factors.

Using joint P-wave seismic tomography, receiver functions, and ambient noise we image the magmatic structure beneath Newberry Volcano, located near Bend, Oregon. Use of active source and teleseismic events in a joint tomographic inversion provides the ray crossings necessary to resolve a low velocity body around 4 km depth. Receiver functions show large lateral heterogeneity and are consistent with the location of a low velocity body derived from the tomography but require a larger low velocity anomaly. Ambient noise autocorrelations are used to image a low velocity reflector, located at ~3 km depth, shallower than the imaged low velocity body recovered using tomography and receiver functions. Ultimately, our results reveal a magma chamber at 3-4 km depth beneath Newberry caldera, with an overlying partially molten sill at ~3 km depth. These results show the usefulness of dense seismometer deployments over volcanoes.

A multi-prong approach was taken in this dissertation to understand volcanic processes from both a long-term and more immediate hazard perspective. In the long-term, magma sources within the crust may produce measurable surficial response and long-wavelength gravity anomalies that provide information about the extent and depth of this magma. Long-term volcanic hazard forecasting is also improved by developing as complete a record as possible of past events. In the short-term, a long-standing question has been on the casting of precursory volcanic activity in terms of future volcanic hazards. Three studies are presented in this dissertation to address these issues. Inversion of high-resolution ground magnetic data in Amargosa Valley, NV indicates that anomaly B could be generated by a buried shield volcano. This new information changes the event count in this region which in turn affects the overall volcanic hazard estimation. Through the use of Finite Element Models (FEM) an in-depth characterization of the surficial response to magma underplating is provided for the Tohoku Volcanic Arc, Japan. These models indicate that surficial uplift was dominantly driven by mid-crustal intrusions and the magnitude and wavelength of this uplift was mainly controlled by the elastic layer thickness. In Dominica, seismic data were used as weights in spatial intensity maps to generate dynamic volcanic hazard maps influenced by changes in seismicity. These maps show an increasing trend in the north that may be indicative of an increase in earthquake and volcanic hazards.

L'éruption de 2010 du Merapi est la première grande éruption explosive du volcan qui a été observée instrumentalement. Dans ce travail, nous étudions les précurseurs de l'éruption et le comportement du volcan avant l'éruption en reliant les caractéristiques sismiques avec d'autres observations disponibles. Nous présentons les principaux aspects de l'activité sismique au cours de la crise de 2010, tels que la chronologie de la sismicité, l'évolution spatio-temporelle des positions de source de séisme et les changements de vitesse sismique. En effectuant des localisations absolues et relatives, nous obtenons des preuves de l'existence de zones asismiques, concordant avec des études antérieures, que nous interprétons comme des zones plus ductiles. La migration du magma de la partie profonde à la partie superficielle du conduit à travers la zone asismique supérieure est mise en évidence par un déplacement vers le haut des hypocentres. Nous analysons l'énergie sismique quantifiée par le RSAM calculé pour plusieurs bandes de fréquences. Ces fonctions affichent des accélérations claires dans les dernières semaines avant l'éruption. Ce comportement est utilisé pour effectuer des prévisions d'éruption volcanique rétrospective avec la méthode « Material Failure Forecast » ou FFM. Le début de la première éruption est estimé avec une bonne précision. Nous proposons une méthode originale de détection d'événement basée sur un rapport d'énergie. En utilisant cette méthode et la corrélation de la forme d'onde, nous identifions 10 familles de séismes similaires. Ces multiplets sismiques sont situés en dessous ou au -dessus de la zone asismique supérieure et sont composés soit d'événements volcano-tectoniques soit d'événements basse fréquence. Certains de ces groupes ont été actifs pendant plusieurs mois avant la crise éruptive alors qu'une famille qui comprend 119 événements répétitifs est apparue 20 heures avant le début de l'éruption. Nous estimons des variations de vitesse sismique, liées principalement à l'activité magmatique, en utilisant la coda des multiplets et les fonctions d'intercorrélation du bruit sismique. Ces variations montrent une forte variabilité spatiale et temporelle de leur amplitude et de leur signe. Bien qu'elles ne puissent pas être décrites par une simple tendance unique, ces variations de vitesse peuvent être considérées comme un précurseur de l'éruption. En utilisant les résultats précédents ainsi que d'autres observations, nous déterminons les particularités associées à la grande éruption explosive de 2010. En outre, nous proposons un scénario chronologique de l'activité pré- éruptive du Merapi
The 2010 eruption of Merapi is the first large explosive eruption of the volcano that has been instrumentally observed. In this work, we study the eruption precursors and the pre-eruptive volcano behaviour by linking seismic features with other available observations. The main characteristics of the seismic activity during the 2010 crisis, including the chronology of seismicity, the spatio-temporal evolution of earthquake source positions and the seismic velocity changes, are presented. By performing absolute and relative locations, we obtain evidences of aseismic zones which are consistent with earlier studies and are interpreted as more ductile zones. Magma migration from the deep to the shallow part of the conduit through the upper aseismic zone is revealed by an upward shift of the hypocenters. We analyse the seismic energy quantified by RSAM calculated for several frequency bands. These functions display clear accelerations in the last few weeks before the eruption. This behaviour is used to perform hindsight eruption forecasting with the Material Failure Forecast method (FFM). The onset of the first eruption is estimated with a good precision. We propose an original method of event detection based on energy ratio. Using this method and waveform correlation, we identify 10 families of similar earthquakes. The seismic multiplets are located either below or above the upper aseismic zone and are composed of either volcano-tectonic or low-frequency events. Some of the clusters were active during several months before the eruptive crisis while a family that includes 119 repeating events appeared 20 hours before the eruption onset. Seismic velocity variations associated mainly with magmatic activity are estimated using the coda of both multiplets and noise cross correlation functions. These variations display strong temporal and spatial variability of their amplitude and sign. Although they cannot be described by a unique simple trend, these velocity variations can be considered as an eruption precursor. Using the preceding results together with other observations, we determine the specific features associated with the large explosive eruption of 2010. Furthermore, we propose a chronological scenario of the pre-eruptive activity of Merapi 2010 unrest

An abnormally high number of explosion quakes were noted during the monitoring effort for the 2007 eruption of Pavlof Volcano on the Alaskan Peninsula. In this study we manually counted the explosion quakes from their characteristic ground-coupled air waves. This study makes an effort at better quantifying the number of explosion quakes and how the characteristic ground-coupled air waves are affected by wind direction and wind speed. Additionally this study investigates how the ground coupled air waves might be used in a monitoring or analysis effort by calculating energy release and gas mass release. Over 3.2x104 quakes were recorded. It was found that wind direction affects the travel time of the air wave by up to 0.7 seconds depending on station location and wind direction. Wind direction and speed, however, are demonstrated not to cause an appreciable difference in ground-coupled air wave frequencies or amplitude ratios. The energy release from the explosions is calculated to be 3.04x1011 J. and the total gas mass (assuming 100% water) released was 729 metric tons. These values are compared to other volcanoes in the literature and found to be somewhat lower. Nevertheless, the tracking of explosion quakes has the potential to become a valuable member of the seismic monitoring arsenal.

UNRAVELLING VOLCANIC TREMOR SOURCE AT MOUNT ETNA FROM QUANTITATIVE MULTIPARAMETRIC ANALYSIS AND MOMENT TENSOR INVERSION Developments of multiparametric monitoring networks, real time analysis techniques and the acquisition of high-resolution data, have allowed to improve the knowledge of the structures and dynamics that characterize the active volcanoes. To understand the dynamics of an active volcano, it is possible to proceed with two different types of studies: i) multiparametric studies based on the joint analysis of different data (such as geophysical, geochemical, petrological data); ii) detailed studies concerning specific data. In order to understand source volcanic tremor at Mt. Etna volcano, volcanic tremor recorded during the vigorous summit of 2011-2012 was analyzed. In particular, this thesis shows the results obtained from the investigation of the volcanic tremor source through a duplex study: i) a multiparametric study based on the quantitative comparison between different time series such as volcanic tremor amplitudes and geochemical data such as soil CO2 flux and the SO2 flux at summit craters; ii) a moment tensor inversion analysis of volcanic tremor recorded during a lava fountain episode of the considered period. For these studies, data acquired from the multiparametric monitoring network managed by Istituto Nazionale di Geofisica e Vulcanologia were used.The quantitative comparison between seismic and geochemical data was performed in order to identify and quantitatively estimate similar trends variations between the compared time series, both synchronous and characterized by time lags. To this end, four different analysis techniques have been implemented, such as Cross-Correlation, Randomized Cross-Correlation, wavelet analysis and linear regression. The obtained results allowed to infer pressurization and depressurization dynamics of the plumbing system during the different eruptive phases observed during the considered period. The moment tensor inversion analysis of the volcanic tremor recorded at Mt. Etna represents a novelty. Specifically, this analysis was applied to volcanic tremor recorded during the lava fountain episode that occurred on 5 August 2011, one of the most vigorous of the 25 episodes that characterized the 2011-2012 period. The results allowed to highlight a source mechanism generated by a sub-horizontal crack near the summit crater area, which can be interpreted as a valve through which gas-rich magma is ejected during the lava fountain activities.

Telica Volcano, Nicaragua, is a persistently restless volcano with high rates of seismicity that can vary from less than ten events to over a thousand events per day. Low-frequency (LF) events dominate the seismic catalogue and seismicity rates at Telica show little clear correlation with periods of eruption. As such, traditional methods of forecasting of volcanic activity based on increases in seismicity and recognition of LF activity are not applicable. A single seismic station has been operating at Telica since 1993, and in 2010 we installed a broadband seismic and continuous GPS network (TESAND network) at Telica. In this study we investigate the seismic characteristics surrounding a nine-month period of phreatic to phreatomagmatic explosions in 1999, and also from the initial three-and-a-half year deployment of the TESAND network, including a three-month phreatic vulcanian eruptive period in 2011. We demonstrate that pertinent information can be obtained from analysis of single-station data, and while large seismic networks are preferable when possible, we note that for many volcanoes this is not possible. We find unusual patterns of seismicity before both eruptive periods; rather than a precursory increase in seismicity as is observed prior to many volcanic eruptions, we observe a decrease in seismicity many months prior to eruption. We developed a new program for cross-correlation of large seismic data catalogues and analysed multiplet activity surrounding both eruptive periods. We observed that the formation of new multiplets corresponds to periods of high event rates (during inter-eruptive periods) and high percentages of daily events that belong to a multiplet. We propose a model for the seismicity patterns observed at Telica, where changes in seismicity are related to a cyclic transition between open-system degassing and closed-system degassing. Periods of open-system degassing occur during non-eruptive episodes and are characterised by high event rates, a broad range of frequency content of events and high degrees of waveform correlation. A transition to closed-system degassing could be due to sealing of fluid pathways in the magmatic and/or hydrothermal system, or due to magma withdrawal. Periods of closed-system degassing are characterised by low event rates, higher frequency contents and low degrees of waveform correlation. Eruptive periods may then represent a transition from closed-system degassing to open-system degassing, however the system must also be capable of transitioning to open-system degassing without eruption. These observations have important implications for volcano monitoring and eruption forecasting at persistently restless volcanoes. Rather than a precursory increase in seismicity as is often observed prior to eruption at other volcanoes, our observations indicate that phreatic eruptions at Telica occur after a decrease in seismicity, a corresponding change in the frequency content of events, and a decrease in waveform correlation. These changes may represent a period of closed-system degassing that could culminate in phreatic eruptions. The inclusion of real-time analysis of variations in frequency content and multiplet activity provides critical information for volcano monitoring institutions.

La multiplication du nombre de réseaux sismiques fait exploser le nombre de données sismologiques. Manuellement, leur traitement est long et fastidieux, d'où la nécessité d'automatiser la détection, la classification et la localisation des événements pour aider les observatoires qui surveillent continuellement la sismicité, mais aussi, dans un intérêt plus scientifique, rechercher et caractériser les phénomènes. La thèse se décompose en 2 axes majeurs : (1) la détection / localisation des séismes, avec le logiciel Waveloc. On a amélioré les outils pré-existants, ajouté de nouvelles fonctionnalités pour une analyse plus détaillée de la sismicité et validé le code avec les données du Piton de la Fournaise ; (2) la classification des séismes. Après calcul des attributs décrivant au mieux les signaux, on a démontré l'efficacité de 2 méthodes d'apprentissage supervisé (régression logistique et SVM) pour le Piton de la Fournaise et soulevé les difficultés pour un cas plus complexe (le Kawah Ijen)
For some time now the quantity of available seismological data has kept increasing. Manually, their processing is long and tedious. Then, the automation of the detection, location and classification of seismic events has become necessary and aims to help the local observatories and to search and characterize some rarer or not well-known phenomena. The work is divided into 2 main directions : (1) the detection and location of seismic events with the Waveloc software (we improved the pre-existing tools, added some new functions for a more detailed analysis of the seimicity and applied the code to data from the Piton de la Fournaise volcano) ; (2) their classification (after computing the seismic attributes, we proved the efficiency and reliability of 2 supervised learning methods - logistic regression and SVM - for the Piton de la Fournaise volcano, underlined the difficulties for a more complex case - the Kawah Ijen volcano - and tried to apply new strategies)

Kawah Ijen (2386 m) is a stratovolcano located within Ijen Caldera, at the easternmost

part of Java island in Indonesia. Since 2010, the volcano has been equipped with seismometers

and several sensors (temperature and level) have been immersed in its acidic lake waters and in the acidic river seeping on the volcano flanks. While finding instruments capable of resisting to such extreme conditions (pH~0) has been challenging, the coupling of lake monitoring techniques with seismic data improves the knowledge of the volcanic-hydrothermal dynamics. Moreover, the monitoring capabilities have been considerably

enhanced supporting the decision-making of the authorities in case of emergency.

Several methods and processing techniques were used to analyze the seismic data. Much effort has been given to implement the seismic velocities (Moving Window Cross Spectral Analysis (MWCSA)) calculations. At Kawah Ijen, the frequency band that is less affected by the volcanic tremor and the seasonal fluctuations at the source ranges between 0.5-1.0 Hz. Moreover, a stack of 5 days for the current CCF gives reliable results with low errors and allows to detect fluctuations which are missed using a 10-day stack.

The background seismic activity mostly consists in low frequency events and a continuous tremor of low amplitude. Fluctuations of the lake temperature and level result from the recharge of the hydrothermal system during the rainy season. Kawah Ijen lake waters are not perfectly mixed and a shallow stratification occurs during the rainy season, because meteoric waters are less dense than the lake fluids.

Different unrest occurred during our study. Some of them strongly affected the volcanic lake, while others did only weakly. In the first category, a strong unrest commenced in October 2011 with heightened VT (Volcano Tectonic) earthquakes and low frequency events activity, which culminated mid-December 2011. This unrest was correlated with an enhanced heat and hydrothermal fluids discharge to the crater and significant variations of the relative velocities (~1%). This suggests an important build-up of stress into the system. VT earthquakes opened pathways for the fluids to ascend, by increasing the permeability of the system, which latter allowed the initiation of monochromatic tremor (MT) when the steam/gases interacted with the shallow portions of the aquifer. Our calculations evidence a higher contribution of steam in March 2012 that might explain the increase of the MT frequency when bubbles were observed at the lake surface. This period was also characterized by short-lived but strong velocity variations, related to water level

rises containing important amount of bubbles, and important heat and mass discharges

into the lake. On the contrary, the second category of unrest did only slightly affect the

lake system. This could be explained by a dryer hydrothermal system and/or locations of

the seismic sources, which were not directly linked to the lake.

While a magmatic eruption will likely be preceded by a strong seismic activity, the major challenges remain to understand why the unrest we studied did not lead to an eruption and to identify precursory signs of a phreatic eruption. Even a small phreatic eruption would be devastating for the people working everyday in the crater and the ones

who live nearby the voluminous acidic lake.
Doctorat en Sciences
info:eu-repo/semantics/nonPublished

Forecasting eruptions using volcano seismology is a subject that affects the lives and property of millions of people around the world. However, there is still much to learn about the inner workings of volcanoes and how this relates to the chance of eruption. This dissertation attempts to increase the breadth of knowledge aimed at helping to understand when a volcano is likely to erupt and how large that eruption might be. Chapters 2 and 3 focus on a technique that uses changes in the local stress field beneath a volcano to determine the source of these changes and help forecast eruptions, while Chapter 4 focuses on a technique that shows great potential to be used to image magma chambers beneath volcanoes by using receiver functions. In Chapters 2 and 3 the source mechanisms of shallow volcano-tectonic earthquakes recorded at Mount St. Helens are investigated by calculating hypocenter locations and fault plane solutions (FPS) for shallow earthquakes recorded during two eruptive periods (1981-1986 and 2004-2008) and two non-eruptive periods (1987-2004 and 2008-2011). FPS show a mixture of normal, reverse, and strike-slip faulting during all periods, with a sharp increase in strike-slip faulting observed in 1987-1997 and an increase in normal faulting between 1998 and 2004 and again on September 25-29, 2004. FPS P-axis orientations (a proxy for ó1) show a ~90° rotation with respect to regional ó1 (N23°E) during 1981-1986 and 2004-2008, bimodal orientations (~N-S and ~E-W) during 1987-2004, and bimodal orientations at ~N-E and ~S-W from 2008-2011. These orientations are believed to be due to pressurization accompanying the shallow intrusion and subsequent eruption of magma as domes during 1981-1986 and 2004-2008, and the buildup of pore pressure beneath a shallow seismogenic volume during 1987-2004 and 2008-2011. Chapter 4 presents a study using receiver functions, which show the relative response of the Earth beneath a seismometer. Receiver functions are produced by deconvolving the vertical component of a seismogram from the horizontal components. The structure of the ground beneath the seismometer can then be inferred from the arrivals of P-to-S converted phases. Receiver functions were computed for the Katmai Volcanic Group, Alaska, at two seismic stations (KABU and KAKN) between January 2005 and July 2011. Receiver functions from station KABU clearly showed the arrival of the direct P-wave and the arrival from the Moho; however, receiver functions from station KAKN did not show the arrival from the Moho. In addition, changes in the amplitude and polarity of arrivals on receiver functions suggested that the structure beneath both KABU and KAKN was complex. Station KABU is likely underlain by dipping layers and/or anisotropy, while station KAKN may lie over a basin structure, an attenuating body, or some other highly complex structure. However, it is impossible to say for certain what the structure is under either station as the azimuthal coverage is poor and thus the structure is unable to be modeled. This dissertation also includes a section (Chapter 6) on the assessment of spreadsheet-based modules used in two Introductory Physical Geology courses at the University of South Florida (USF). When faculty at USF began using spreadsheet-based modules to help teach students math and geology concepts the students complained that they spent more time learning how to use Excel than they did learning the concepts presented in the modules. To determine whether the sharp learning curve for Excel was hindering learning we divided the students in two Introductory Physical Geology courses into two groups: one group was given a set of modules which instructed them to use Excel for all calculations; the other group was simply told to complete the calculations but was not instructed what method to use. The results of the study show that whether or not the students used Excel had very little to do with the level of learning they achieved. Despite complaints that Excel was hindering their learning, students in the study attained high gains for both the math and geology concepts presented in the modules whether they used Excel or not.

We investigated the relationship between seismo-volcanic events, recorded at La Fossa crater of Vulcano (Aeolian Islands, Italy) during 2004-2009, and the dynamics of the hydrothermal system. During the period of study, six episodes of increasing numbers of seismo-volcanic events took place at the same time as geothermal and geochemical anomalies were observed. These geothermal and geochemical anomalies have been interpreted as resulting from an increasing deep magmatic component of the hydrothermal fluids. Four classes of seismic events (long period, high frequency, monochromatic and tornillos events), characterised by different spectral content and various similarity of the waveforms, have been recognised. These events, clustered mainly below La Fossa crater area at depths of 0.5 1.1 km b.s.l., were space-distributed according to the classes. Based on their features, we can infer that such events at Vulcano are related to two different source mechanisms: (1) fracturing processes of rocks and (2) resonance of cracks (or conduits) filled with hydrothermal fluid. In the light of these source mechanisms, the increase in the number of events, at the same time as geochemical and geothermal anomalies were observed, was interpreted as the result of an increasing magmatic component of the hydrothermal fluids, implying an increase of their flux. Indeed, such variation caused an increase of both the pore pressure within the rocks of the volcanic system and the amount of ascending fluids. Increased pore pressures gave rise to fracturing processes, while the increased fluid flux favoured resonance and vibration processes in cracks and conduits. Finally, a gradual temporal variation of the waveform of the hybrid events (one of the subclasses of long period events) was observed, likely caused by heating and drying of the hydrothermal system. After careful analysis of the seismo-volcanic events of the Aeolian Islands area, the attention was paid to the tectonic events, in order to find possible relationships with the volcanic activity in the area. The aim of this part of the thesis was to identify spatial clusters of earthquakes, locate active seismogenic zone and their relationships with the volcanic activity in the Aeolian Islands. High precision locations were performed in the present thesis, by applying the concept of the velocity model-hypocentres joint inversion and earthquake relocations, along with an analysis of the fault plane solutions. In order to improve our knowledge on the active seismo-tectonics areas we exploited a dataset encompassing 351 events recorded during a 17 year period (1993-2010). Overall, our results show that part of the seismicity is clustered along active seismogenic structures that concur with the main regional tectonic trends whose activity furnishes new elements to better understand the dynamics of the area. A cluster of 24 events in the northern part of Vulcano, NE-SW oriented, marks the presence of a structure that seems to play a key role in magma uprising at Vulcano. These earthquakes suggest the existence of a seismogenic structure (passing just below Vulcanello), which could be interpreted as a discontinuity linking the two magma accumulation zones, thereby representing a possible preferential pathway along which magma may intrude as well as being responsible for fluid migration toward the surface. The results presented in this thesis suggest that the comparison of seismic, ground deformation and temperature data can be useful for better understanding the dynamics of a complex volcano-hydrothermal system, including a better definition of the origin of a volcano unrest, and hence for improving the estimation of the level of the local volcanic hazard.

Deep non-volcanic tremor is a newly discovered seismic phenomenon with low amplitude, long duration, and no clear P- and S-waves as compared with regular earthquake. Tremor has been observed at many major plate-boundary faults, providing new information about fault slip behaviors below the seismogenic zone. While tremor mostly occurs spontaneously (ambient tremor) or during episodic slow- slip events (SSEs), sometimes tremor can also be triggered during teleseismic waves of distance earthquakes, which is known as "triggered tremor". The primary focus of my Ph.D. work is to understand the physical mechanisms and necessary conditions of triggered tremor by systematic investigations in different tectonic regions. These include Taiwan, California, southwest Japan, Alaska and the Aleutian Arc, Cascadia, and New Zealand. In the first chapter of my dissertation, I conduct a systematic survey of triggered tremor beneath the Central Range (CR) in Taiwan for 45 teleseismic earthquakes from 1998 to 2009 with Mw ≥ 7.5. Triggered tremors are visually identified as bursts of high-frequency (2-8 Hz), non-impulsive, and long-duration seismic energy that are coherent among many seismic stations and modulated by the teleseismic surface waves. A total of 9 teleseismic earthquakes has triggered clear tremor in Taiwan. The peak ground velocity (PGV) of teleseismic surface waves is the most important factor in determining tremor triggering potential, with an apparent threshold of ~0.1 cm/s, or 7-8 kPa. However, such threshold is partially controlled by the background noise level, preventing triggered tremor with weaker amplitude from being observed. In addition, I find a positive correlation between the PGV and the triggered tremor amplitude, which is consistent with the prediction of the 'clock-advance' model. This suggests that triggered tremor can be considered as a sped-up occurrence of ambient tremor under fast loading from the passing surface waves. Finally, the incident angles of surface waves also play an important rule in controlling the tremor triggering potential. The next chapter focuses on a systematic comparison of triggered tremor around the Calaveras Fault (CF) in northern California (NC), the Parkfield-Cholame section of the San Andreas Fault (SAF) in central California (CC), and the San Jacinto Fault (SJF) in southern California (SC). Out of 42 large (Mw ≥7.5) earthquakes between 2001 and 2010, only the 2002 Mw 7.9 Denali fault earthquake triggered clear tremor in NC and SC. In comparison, abundant triggered and ambient tremor has been observed in CC. Further analysis reveal that the lack of triggered tremor observations in SC and NC is not simply a consequence of their different background noise levels as compared to CC, but rather reflects different background tremor rates in these regions. In the final chapter, I systematically search for triggered tremor following the 2011 Mw9.0 Tohoku-Oki earthquake in the regions where ambient or triggered tremor has been found before. The main purpose is to check whether triggered tremor is observed in regions when certain conditions (e.g., surface wave amplitudes) are met. Triggered tremor is observed in southwest Japan, Taiwan, the Aleutian Arc, south-central Alaska, northern Vancouver Island, the Parkfield-Cholame section of the SAF in CC and the SJF in SC, and the North Island of New Zealand. Such a widespread triggering of tremor is not too surprising because of the large amplitude surface waves (minimum peak value of ~0.1 cm/s) and the associated dynamic stresses (at least ~7-8 kPa), which is one of the most important factors in controlling the triggering threshold. The triggered tremor in different region is located close to or nearby the ambient tremor active area. In addition, the amplitudes of triggered tremor have positive correlations with the amplitudes of teleseismic surface waves among many regions. Moreover, both Love and Rayleigh waves participate in triggering tremor in different regions, and their triggering potential is somewhat controlled by the incident angles. In summary, systematically surveys of triggered tremor in different tectonic regions reveal that triggered tremor shares similar physical mechanism (shear failure on the fault interface) as ambient tremor but with different loading conditions. The amplitude of the teleseismic surface wave is one of the most important factors in controlling the tremor triggering threshold. In addition, the frequency contents and incident angles of the triggering waves, and local fault geometry and ambient conditions also play certain roles in determining the triggering potential. On the other hand, the background noise level and seismic network coverage and station quality also could affect the apparent triggering threshold. Because triggered tremor occurs almost instantaneously during the teleseismic surface waves, and the tremor amplitude is somewhat controlled by the amplitude of the triggering waves, the occurrence time and the size of the triggered tremor could be somewhat predictable, so long as we know the amplitude and period of surface waves and associated time-varying dynamic stresses. Hence, further analysis of triggered tremor may provide important new clues on the nucleation and predictability of seismic events.

Regional-scale complete Bouguer gravity anomalies underlying the Lassen and Shasta -Medicine Lake regions in northern California and southern Oregon are associated with subduction of the Gorda plate beneath North America. These generally negative anomalies reflect where underplating has deepened to form the mantle wedge, and where subduction has given rise to a series of Quaternary volcanoes comprising the southernmost end of the Cascade range. Multiple conductive bodies were identified by Park and Ostos (2013) in their magnetotelluric (MT) study of the broader Lassen volcanic region. Their broadband and long period measurements were conducted along a 250 km profile spanning from the California-Nevada border, to just west of the Great Valley in California. Utilizing their MT conductor geometries as a starting point, a forward gravity model was generated along the same profile, and agrees well with what they interpret to be the locations and depths of mid-crustal magma bodies in the Lassen and surrounding regions. The excess mass and volume of modeled anomaly (a) - most closely attributed to underlying Lassen Peak - were estimated at -2 x 1014 kg and 7 x 1011 m3, respectively.

In this thesis I combine data from 29 volcanic earthquake swarms that follow the pattern predicted by the Generic Volcanic Earthquake Swarm Model (GVESM; Benoit and McNutt, 1996) to investigate whether the relative timing of various parameters of pre-eruptive volcanic earthquake swarms could be used to forecast the time of an impending eruption. Based on the analysis of seismic unrest preceding many eruptions, the GVESM suggests that it is common to see an increase first in high-frequency earthquakes, then low-frequency earthquakes, then the onset of volcanic tremor. While this pattern is useful to volcano-seismologists, the relative timing and durations of these three different types of volcanic seismicity, is explored here for the first time. The parameters investigated are the onset times of (i) low-frequency (LF) events and of (ii) tremor, and the time at which (iii) the peak rate (PR) of volcano-tectonic (VT) events and (iv) the maximum magnitude (MM) earthquake occur in relation to normalized time defined by swarm onset and end (i.e., eruption). The normalized time starts at the swarm onset (0%) and ends with the eruption (100%) allowing a comparison and joint consideration of parameter occurrences across swarms of different actual duration. We identify the normalized onset time of for each parameter (LF, tremor, PR, MM) with respect to the duration of each swarm. Each swarm has onset time uncertainties of the swarm itself and of its parameters. A swarm with large onset uncertainty could bias the normalized onset time of each parameter and we use weighted means to decrease the influence of swarms with large uncertainties on overall results. The weighted means of LF onset, tremor onset, MM and PR occurrence are 79% ± 23%, 96% ± 10%, 78% ± 29% and 75% ± 34%, respectively. Errors are the standard deviation of each parameter. The uncertainties for LF, MM and PR are large because their normalized onset times have the characteristics of a uniform distribution and therefore seem to have no predictive value. In contrast, tremor onset has a narrow distribution towards the end of swarms. A possible tremor mechanism consistent with this observation could be boiling of groundwater as magma nears the surface. LF onset always seems to precede tremor onset. LF and tremor start early (at less than 80% of normalized time) at five volcanoes with high SiO2 content possibly related to lower density and higher gas content of the resulting magma.

The observational data recorded by the monitoring networks during El Hierro 2011-2012 volcanic unrest and eruption, allowed us to conduct a study of the volcanic eruptive phenomena with that wide view. The lack of previous instrumental data describing such a process in the Canaries gave us the invaluable chance of analyzing for the first time, novel geophysical signals registered before, during and after an eruptive event, allowing the study of evidence of the energy involved in the emplacement and migration of magma through the lithosphere. The aim of this PhD Thesis is to make a comprehensive revision of the physics of this eruptive process and its causes, through the analysis and processing of geophysical data which has been recorded since the time of the first instrumental record to exist in the area, to the present. Three main aspects are considered, focusing on the eruptive phenomenon of El Hierro at different temporal and spatial scales, going from the general to the specific. The first aspect is the study of the potential relation between the regional geodynamics and the eruptive event of El Hierro 2011–2012. This potential relation, through the analysis of long-time series of geophysical observational data, has not been studied before. Highly reliable seismic and geodetic data acquired from 1996 to 2014 is analyzed, covering: from the North Atlantic Ridge to the West, to the Azores-Gibraltar boundary to the North (including the NW African margin) and the Canary Islands. A joint regional- and local scale analysis based on this data enabled the identification of early signs of anomalous tectonic activity from 2003 onwards, the intensity of which increased in2007,accelerating three months before the onset of the volcanic eruption on El Hierro in October 2011. The second aspect is the study of the precursory signals recorded during the El Hierro volcanic episode (19 July to 10 October, 2011), and their correct interpretation when only limited data is available. The seismic and deformation data corresponding to the pre-eruptive unrest is reanalyzed using novel methods, taking into account new information about the internal structure of the island. Results indicate that important changes in the medium properties and in the magmatic mechanism occurred throughout the process, identifying different phases with distinct types of fracturing. A triggered shear seismicity is observed on pre-existent faults caused by the magma pressure on the structure underneath the island, and the crossing of the Moho discontinuity. The third aspect is the detailed study of the type and geometry of the irreversible mechanism, acting during the final phase of the unrest episode (6-10 October, 2011) and during the first days of tremor (10-15 October, 2011) recording on the seismic monitoring stations. Application of time varying fractal analysis to the seismic data and the characterization of the seismicity pattern and the strain and the stress rates, allows the identification of different stages in the source mechanism, and to infer the geometry of the path used by the magma and associated fluids to reach the Earth’s surface.
El análisis de datos geofísicos y geodésicos en un amplia área que comprende al oeste la Dorsal Oceánica Atlántica, al norte la falla Azores-Gibraltar y el Sur de Iberia y al sur el NW de Marruecos y las Islas Canarias, nos ha permitido realizar un estudio entre la posible relación entre magmatismo y actividad tectónica. En concreto el estudiar su posible influencia en el desencadenamiento de la última erupción ocurrida en la isla de El Hierro en 2011. La ausencia de datos instrumentales previos que describen tal proceso en Canarias, nos dio la oportunidad de analizar por primera vez las señales geofísicas registradas antes, durante y después de un evento eruptivo, permitiendo el estudio de la actividad al emplazamiento y migración del magma a través de la litosfera. El objetivo de esta tesis doctoral es hacer una revisión integral de la física de este proceso eruptivo y sus causas, a través del análisis y procesamiento de datos geofísicos que se han registrado desde la época del primer registro instrumental existente en el área, el presente. Se consideran tres aspectos principales, centrados en el fenómeno eruptivo de El Hierro en diferentes escalas temporales y espaciales, pasando de lo general a lo específico. El primer aspecto es el estudio de la posible relación entre la geodinámica regional y el evento eruptivo de El Hierro 2011-2012. Se analizan los datos sísmicos y geodésicos adquiridos de 1996 a 2014, que abarcan desde el Atlántico Norte hasta el Oeste, hasta la frontera Azores-Gibraltar al Norte (incluido el margen NW-África) y Canarias. Un análisis conjunto permitió identificar signos tempranos de actividad tectónica anómala a partir de 2003, cuya intensidad aumentó en 2007, acelerándose tres meses antes del inicio de la erupción volcánica en El Hierro en octubre de 2011. El segundo aspecto es el estudio del periodo de unrest (10 Julio-10 Octubre). Los resultados muestran en El Hierro evidencias de: inyección de magma bajo la corteza, sobrepresurización, fracturación hidráulica (inicialmente de fluídos ricos en gas), migración hacia el Sur, superación del Moho desencadenando sismicidad inducida en fallas preexistentes y migración hacia la superficie, probablemente en condiciones de irreversibilidad. El tercer aspecto estudiado los las condiciones finales de ascenso de magma a la superficie y de establecimiento de la erupción (6 Octubre-15 Octubre), mostrando cambios en la geometría asociados al establecimiento del edificio volcánico, así como cambios en la composición del magma en los primeros días de la erupción.

Le Yasur est un petit stratovolcan situé au coeur d'une grande caldeira, dans la partie sud-est de l'Ile de Tanna (Sud Vanuatu). Le cratère sommital est occupé par trois évents : les évents A et B dans le sous-cratère sud et l'évent C dans le sous-cratère nord. L'activité du Yasur est caractérisée par des explosions de bulles de gaz et l'émission de petits panaches de cendres. Afin de mieux comprendre les processus volcaniques qui se produisent au sein de l'édifice, un large réseau de capteurs, comprenant 12 antennes sismiques et 10 stations large-bande, a été déployé autour du Yasur en 2008, enregistrant le signal sismique en continu. Le modèle de vitesse de la structure superficielle du volcan ainsi que de la caldeira de Siwi a été estimé jusqu'à 200 m de profondeur en utilisant les méthodes SPAC et f-k - deux techniques d'analyse du bruit ambiant inclues dans le logiciel Geospy - sur les signaux enregistrés par sept antennes. Quatre zones principales ont été distinguées et interprétées en terme de couches géologiques et de répartition des fluides (aquifère, système hydrothermal) dans la structure volcanique. Le modèle de vitesse a ensuite été intégré dans un algorithme de tracé de rai et combiné à une méthode de triangulation sismique pour localiser la source des explosions et des événements longue période. Les localisations ont permis de suivre l'évolution de l'activité en 2008 dans les différents évents, et de proposer deux modèles de dynamique éruptive pour expliquer la périodicité des explosions stromboliennes et l'intense activité observée au Yasur. Les localisations ont également permis l'estimation de la vitesse d'ascension des poches de gaz ainsi qu'une estimation de la géométrie du conduit volcanique.

Development of homes, roads, and commercial buildings in northern Utah has grown significantly during the last several decades. Construction has expanded from the valley floor to higher elevations of benches, foothills, and other elevated regions of the Wasatch Mountain Front. Construction in the higher elevation areas are a concern due to potential for landslides, both new and reactivated. Landslides have been identified in this region and are dated as Pleistocene to historical in age. A possible landslide of about 0.5 km2 on the south slope of Traverse Mountain has been mapped by the Utah Geological Survey in 2005. Its surface exhibits hummocky topography and is comprised of Oligocene-age volcanic ash, block and ash flow tuffs, and andesite lava. Landslides along the Wasatch Mountain Front are complex features usually characterized by dense vegetation and poor outcrop and require a combination geological and geophysical methods to study their thickness, slope, lateral extent, and style of emplacement. Our study incorporates trenching, boreholes, and LiDAR aerial imagery. Unique to the study of landslides is our use of seismic reflection with a vibroseis source over the mapped landslide deposit. The seismic parameters of source, station spacing, and processing method provide a coherent, albeit low-resolution, image of the upper 500 m of the subsurface beneath the landslide. A major reflector boundary in our seismic profiles has an apparent dip of 4° to the south, approximately parallel with the surface topography. Its elevation and seismic character are indicative of a contact between the Oligocene-age volcanic rocks on top of a portion of the Pennsylvanian-age Bingham Mine Formation, a mixed carbonate and siliciclastic sequence. The reflector defines an asymmetric graben-like structure bounded by a north-northwest-trending normal fault system. Analysis of trenches, boreholes and local geology reveals a faulted, chaotic body of block and ash flow tuffs, surrounded by andesite lavas. Using LiDAR and surface geological reconnaissance, a possible toe or margin of a landslide has been interpreted in the north-west portion of the study area. The combination weakened block and ash flow tuffs and abundant clay production from this unit contribute to the likelihood of a coalescence of landslides in this mapped landslide area. The integration of LiDAR, trenching, boreholes and reflection seismology provides the range and resolution of data needed to assess the complex geology of landslides.

Depuis maintenant une dizaine d'années, la vision du cycle sismique en zone de subduction a beaucoup évolué. Des découvertes récentes ont mis en évidence une grande diversité des régimes de glissement dans ces zones, avec notamment des glissements asismiques transitoires appelés « séismes lents » (SSE) et des vibrations de faibles amplitudes, persistantes dans le temps, appelées « trémors non volcaniques » (NVT). Ce travail a pour objectif l'étude des trémors non volcaniques afin de caractériser ces nouvelles manifestations des zones de faille. Nous avons abordé ce problème avec deux approches distinctes :1. Observer les trémors dans le milieu naturel afin de déterminer leurs caractéristiques. La zone étudiée correspond à la lacune sismique de Guerrero le long de la subduction mexicaine. Nous avons développé une méthode de détection et de localisation des NVT au Mexique grâce à des analyses d'antennes par formation de voie sur les corrélations. Cette méthode permet de mettre en évidence cer taines caractéristiques des NVT : une complexité des sources pour un épisode de trémors, une corrélation entre les activités de NVT et les pics de vitesse des glissements lents à plus long terme. Par ailleurs, l'étude de l'impact du séisme de Maule (2010, Chili, Mw 8.8) au Mexique montre qu'il a déclenché le second sous évènement du séisme lent de 2009-2010. Ce déclenchement d'un SSE s'ac- compagne de fortes activités de trémors, modulées par les ondes du séisme de Maule dans un premier temps, puis simplement associées au SSE.2. Modéliser les trémors expérimentalement et numériquement pour mieux com- prendre leur origine physique et leurs évolutions sur le long terme. Nous avons en particulier utilisé une expérience de frottement à faible vitesse qui indique une corrélation systématique entre les accélérations d'un glissement et l'émission de signaux qui ressemblent à des NVT. Une modélisation numérique de la zone de subduction mexicaine est également présentée et montre la possibilité de reproduire des trémors en considérant une transition d'affaiblissement critique associée à un processus de décrochage
The vision of the seismic cycle in subduction zones has considerably evolved over the last 10 years. New discoveries has pointed the diversity of slip behaviors in these zones with aseismic slow slip called « slow slip events » (SSE) and persistent low amplitudes vibrations called « non-volcanic tremors » (NVT). The goal of this thesis is to study the non-volcanic tremors in order to characterize these new manifestations of fault zones. We used two different approaches: 1. We first observed the non-volcanic tremors in the nature in order to characterize this phenomenon. The area of interest is the Guerrero seismic gap along the Mexican subduction zone. We develop a new detection and location method based on beamforming of correlations of seismic signals. This new method exhibits some characteristics of NVT: a complex source for a single tremor episode and a correlation between the NVT episodes and the long-term peak of movement velocity in southwards direction. Moreover, the study of the consequences of the Maule earthquake on the Mexican subduction zone showed that this earthquake triggered the 2009-2010 SSE in Guerrero. This triggering of slow slip is accompanied by strong seismic tremor actvity that are first modulated by the passing waves and then associated to the SSE. 2. We model numerically and experimentally the tremors in order to better understand their physical origin and their long-term evolution. We used a very slow friction experiment that indicates a systematic correlation between slip acceleration of a slider and emission of acoustic signals that are similar to NVT. A numerical modeling of the Mexican subduction zone is also presented and shows the possibility to reproduce NVT with a critical depinning transition

A detailed comparative geochemical characterization of three different types of Iceland glacial systems was conducted during June, August, and October, 2016. The study was carried out at a total of 11 outlet glacier rivers flowing from the icecaps Vatnajökull, Eyjafjallajökull, and Mýrdalsjökull. A total of 75 grab samples were collected (25 for each sampling period). The hydrogeochemical variations of Icelandic glacial meltwater are influenced by volcanic activity, temporal changes, and geographical location, which differed between the sampling sites within the glaciers and icecaps. Lower pH range, and comparatively higher and variable specific conductivity, SO4, S and F is linked to higher volcanic influences, including residuals from the 2010 eruption at Eyjafjallajökull (located above a tectonic plate boundary zone). High concentrations of Al and Fe were found at Kötlujökull and Kvíárjökull, both of which are close to active volcanic zones. Changes in hydrogeochemistry of the meltwater caused by volcanic activity may be used to forecast eruptions and jökulhlaups; however, given the variability of Icelandic meltwater chemistry, high-resolution monitoring should be done in order to determine a precursor threshold for an volcanic event, as the chemical composition of one jökulhlaup could be within normal range for a different glacier. TSS concentrations depicted high spatial and temporal variation as the highest and lowest values of TSS drained from the same glacier. Hydrogeochemical weathering is driven by Na-HCO3 and Ca-HCO3 dissolution. Concentrations of ions varied with respect to their geographic location, as specific conductivity increase distance downstream from glaciers, proglacial lagoons, and river reaches. Ca, Mg, K, Na, and HCO3 increased from 1984 to 2016 for Fjallsjökull, which may be from an increased weathering rate, due to temperature, CO2 increase, and increased erosion beneath glaciers under a changing climate. This study of hydrogeochemical variation in Icelandic glaciers complements the database of physical and chemical compositions of understudied glaciers. The hydrogeochemical variations of Icelandic glacial meltwater throughout a diverse sample of glaciers and their respective icecaps are related to internal and external factors, and their diversity indicates a much more complex set of processes underway at the different icecaps and their respective glaciers.

Cette étude traite de la dynamique des éruptions volcaniques explosives, depuis les mécanismes de sub-surface jusqu'aux processus d'émission et de dispersion des pyroclastes. A cet effet un radar Doppler sol est utilisé (VOLDORAD), lequel renseigne sur la charge / vitesse des ejectas. Les données sont intégrées avec d'autres techniques géophysiques, et des modèles numériques sont développés afin de simuler les émissions pyroclastiques, générer des signaux radar synthétiques, pour finalement améliorer notre compréhension des processus qui leurs sont sous-jacents. L'Arenal (Costa Rica) est utilisé comme volcan cible, où de fréquentes éruptions de faible magnitude émettent des panaches de cendres et des projections balistiques jusqu'à quelques centaines de mètres au-dessus de l'évent. Dans un premier temps, nous combinons des données sismiques et radar afin d'explorer la relation entre les processus de conduit et les émissions pyroclastiques. Leurs interactions complexes sont interprétées via un modèle conceptuel, lequel décrit les fractures parsemant le bouchon de lave comme responsables du dégazage du système, et en retour des signaux sismiques et radar collectés (ces derniers dépendants de la charge en cendres des émissions de gaz). Par ailleurs, nous investiguons la dynamique des émissions pyroclastiques à travers l'étude de radargrammes Doppler. La distribution spatio-temporelle de la vitesse des ejectas indique l'existence de deux phénomènes aux dynamiques distinctes. Des modélisations numériques permettant la reconstruction de signaux synthétiques indiquent qu'il s'agit de l'émission simultanée de blocs balistiques et de panaches de cendres. Une procédure d'inversion de type Monte Carlo couplée d'un algorithme d'optimisation permet de retrouver les radargrammes synthétiques qui reproduisent au mieux ceux observés. Les résultats apportent des contraintes sur divers paramètres éruptifs, tels que les tailles, trajectoires, vitesses des ejectas et des gaz, ainsi que la vitesse / direction de dispersion des panaches de cendres par le vent. Enfin, nous discutons du potentiel des radars Doppler appliqués à la surveillance opérationnelle des émissions volcaniques. En particulier, la possibilité de quantifier les masses éjectées dans l'atmosphère ou retombant sur les flancs du volcan, fournit des paramètres éruptifs à la source pouvant alimenter les modèles de dispersion de panaches de cendres.

Les séismes de type longue période (LP) sont aujourd’hui enregistrés sur la plupart des volcans dans le monde entier. Malgré cela, le mécanisme à leur source n’est encore que très peu compris. A l’heure actuelle les modèles proposés pour expliquer leur origine sont : 1) la résonance d’une fracture remplie de gaz ou de fluides excités par des instabilités dans l’écoulement des fluides ou par la rupture fragile du magma ; 2) la fracturation lente des sédiments peu consolidés à la surface des volcans, dans des conditions de transition entre le ductile et le fragile. L’outil le plus utilisé pour comprendre leur nature est aujourd’hui l’inversion du tenseur des moments. Au cours des dernières années, les inversions du tenseur des moments se concentraient principalement sur la compréhension du mécanisme physique à l’origine des séismes LP qui souvent supposaient des milieux géologiques très simples, voire homogènes. Des études récentes ont montré l’influence des sédiments peu consolidés à la surface des volcans sur la propagation des ondes à basse fréquence et en conséquence, sur l’inversion du tenseur des moments quand ils ne sont pas pris en compte dans le processus d’inversion. Le but de cette thèse est de mieux comprendre les processus physiques qui génèrent les séismes LP et de quantifier les incertitudes liées à leur interprétation
Long-period (LP) seismic events are abundantly recorded during rest and unrest periods at many volcanoes worldwide. However, their source mechanism is still poorly understood. Models which have been proposed so far to explain their origin are: 1) the resonance of a fluid-filled cavity triggered by fluid instabilities or the brittle failure of magma; 2) slow-rupture earthquakes occurring in the low consolidated materials composing the shallow portion of the volcanic edifice. Nowadays the main tool used to get insights into their nature is moment tensor (MT) inversion. MT inversions carried out in the past years focused mainly on the understanding of the physical origin of LP events and often supposed a relative simple geological structure of the medium. Recent studies highlighted the strong influence of shallow unconsolidated materials on the retrieved MT solutions and the importance of considering geological inhomogeneity in the inversion process. The principal aim of this thesis is to gain a better understanding of the source processes that generate LP events and to quantify the uncertainties related to the MT inversion process

This dissertation investigates the link between volcanic unrest and the occurrence of moderate-to-large earthquakes with a specific type of focal mechanism. Vertical compensated-linear-vector-dipole (vertical-CLVD) earthquakes have vertical pressure or tension axes and seismic radiation patterns that are inconsistent with the double-couple model of slip on a planar fault. Prior to this work, moderate-to-large vertical-CLVD earthquakes were known to be geographically associated with volcanic centers and vertical-CLVD earthquakes were linked to a tsunami in the Izu-Bonin volcanic arc and a subglacial fissure eruption in Iceland. Vertical-CLVD earthquakes are some of the largest and most anomalous earthquakes to occur in volcanic systems, yet their physical mechanisms remain controversial largely due to the small number of observations. Five vertical-CLVD earthquakes with vertical pressure axes are identified near Nyiragongo volcano in the Democratic Republic of the Congo. Three earthquakes occur within days of a fissure eruption at Nyiragongo, and two occur several years later in association with the refilling of the lava lake in the summit crater of the volcano. Detailed study of these events shows that the earthquakes have slower source processes than tectonic earthquakes with similar magnitudes and locations. All five earthquakes are interpreted as resulting from slip on inward-dipping ring-fault structures located above deflating shallow magma chambers. The Nyiragongo study supports the interpretation that vertical-CLVD earthquakes may be causally related to dynamic physical processes occurring inside the edifices or magmatic plumbing systems of active volcanoes. Two seismicity catalogs from the Global Centroid Moment Tensor (CMT) Project are used to search for further examples of shallow earthquakes with robust vertical-CLVD focal mechanisms. CMT solutions for approximately 400 target earthquakes are calculated and 86 vertical-CLVD earthquakes are identified near active volcanoes. Together with the Nyiragongo study, this work increases the number of well-studied vertical-CLVD earthquakes from 14 to 101. Vertical-CLVD earthquakes have focal depths in the upper ~10 km of the Earth's crust, and ~80% have centroid locations within 30 km of an active volcanic center. Vertical-CLVD earthquakes are observed near several different types of volcanoes in a variety of geographic and tectonic settings, but most vertical-CLVD earthquakes are observed near basaltic-to-andesitic stratovolcanoes and submarine volcanoes in subduction zones. Vertical-CLVD earthquakes are linked to tsunamis, volcanic earthquake swarms, effusive and explosive eruptions, and caldera collapse, and approximately 70% are associated with documented volcanic eruptions or episodes of volcanic unrest. Those events with vertical pressure axes typically occur after volcanic eruptions initiate, whereas events with vertical tension axes commonly occur before the start of volcanic unrest. Both types of vertical-CLVD earthquakes have longer source durations than tectonic earthquakes of the same magnitude. The isotropic and pure vertical-CLVD components of the moment tensor cannot be independently resolved using our long-period seismic dataset. As a result, several physical mechanisms can explain the retrieved deviatoric vertical-CLVD moment tensors, including dip-slip motion on ring faults, volume exchange between two reservoirs, the opening and closing of tensile cracks, and volumetric sources. An evaluation of these mechanisms is performed using constraints obtained from detailed studies of individual vertical-CLVD earthquakes. Although no single physical mechanism can explain all of the characteristics of vertical-CLVD earthquakes, a ring-faulting model consisting of slip on inward- or outward-dipping ring faults triggered by the inflation or deflation of a shallow magma chamber can account for their seismic radiation patterns and source durations, as well as their temporal relationships with volcanic unrest. The observation that most vertical-CLVD earthquakes are associated with volcanoes with caldera structures supports this interpretation.

Hauptaufgabe der Vulkanseismologie ist die qualitative and quantitative Beschreibung einer oder mehrerer unbekannter seismischer Quellen, die sich in einer unbekannten Tiefe unter dem Vulkan befinden. Auch wenn viele Vulkane der Erde ähnliche Signalcharakteristiken aufweisen, war es bis heute nicht möglich, für Vulkane ein seismisches Standard-Quellmodell zu finden, analog dem Double- Couple in der Erdbebenseismologie. Kontinuierlich tätige Vulkane, wie z.B. Stromboli (Italien), stellen für den Vulkanseismologen ein ideales natürliches Feldlabor dar, diese Fragestellung zu untersuchen. Die vorliegende Arbeit untersucht auf Stromboli registrierte Explosionsbeben und vulkanischen Tremor in einem breiten Frequenzband und behandelt die Frage nach der Lage und dem Mechanismus der seismischen Quelle(n). Seismische und Infraschallmessungen von strombolischen Explosionsbeben zeigen, dass sich eine Hochfrequenz-Phase mit einer Geschwindigkeit von etwa 330 m/s fortbewegt. Die seismische Quelle kann durch eine Explosion am oberen Ende der Magmasäule erklärt werden, die durch aufsteigende Gasblasen verursacht wird. Sowohl die seismische P-Welle, als auch die Luftwelle werden zum gleichen Zeitpunkt an ein und demselben Ort generiert. Die verschiedenen Laufwege und Geschwindigkeiten der seismischen und der Luftwelle resultieren in einem Laufzeitunterschied dt, der zur Bestimmung des Magmenstandes und der Schallgeschwindigkeit in der Eruptionss¨aule im Schlotinnern genutzt werden kann. In Kraternähe installierte Stationen zeigen, dass Infraschall- und seismische Messungen des kurzperiodischen Tremors (> 1 Hz) den gleichen Frequenzgehalt und ähnliche Fluktuationen der seismischen Energie aufweisen. Daher wird der kurzperiodische vulkanische Tremor auf Stromboli durch das kontinuierliche Aufsteigen und Platzen kleiner Gasblasen im oberen Teil der Magmasäule verursacht. Das Spektrum des auf Stromboli registrierten langperiodischen Tremors besteht hauptsächlich aus drei Maxima bei 4.8 s, 6 s und 10 s, deren Spektralamplitude mit der jeweiligen Wettersituation variieren. Sie werden daher nicht von einer lokalen vulkanischen Quelle erzeugt, sondern durch Meeresmikroseismik (MMS). Der Durchzug eines lokalen Tiefdruckgebietes scheint die Ursache für Spektralenergie bei 4.8 s and 10 s, die jeweils die Doppelte bzw. die Primäre Frequenz der MMS darstellen. Als Ursache des spektralen Maximums bei 6 s könnte ein Tief nahe der Britischen Inseln in Frage kommen. Seismische Daten, die von dem ersten auf Stromboli installierten Breitband- Array registriert wurden, zeigten überraschend einfache Wellenformen, die einen anfänglich kontraktierenden Quellmechanismus anzeigen. Die Analyse der Partikelbewegung und die Anwendung seismischer Arraytechniken ermöglichten eine Lokalisierung der seismischen Quelle in Oberflächennähe. Die Anwendung verschiedener Inversionsmethoden gestattete es, Eruptionsparameter und Charakteristiken der seismischen Quelle während der Strombolieruption am 5. April 2003 abzuschätzen. Als Ergebnis kann festgehalten werden, dass der paroxystische Ausbruch durch eine langsame Überschiebungsdislokation mit einer Momentenmagnitude von Mw = 3.0 verursacht wurde, ausgelöst durch einen vorher durch Dike-Intrusion verursachten Bruch. Während des Paroxysmus konnte in den seismischen Signalen mindestens eine Blow-out Phase mit einer Momentenmagnitude von Mw = 3.7 identifiziert werden. Diese kann durch einen vertikalen linearen Vektordipol, zwei schwächere horizontale lineare Dipole in entgegengesetzter Richtung, zuzüglich einer Vertikalkraft repräsentiert werden. Seismische Messungen, die während kontrollierter und reproduzierbarer Blowout Experimente unter Verwendung von einem in einer Basaltschmelze eingeschlossenen Gasvolumen durchgeführt wurden, ergaben folgende Ergebnisse: Monochromatische Signale sind Anzeiger für einen Blow-out in einem duktilen Regime, wohingegen ein breitbandigerer Frequenzgehalt auf einen Sprödbruch hinweist. Je grösser die Länge des Schmelztiegels ist, desto schwächer sind die seismischen Signale. Ein grösser Gasdruck bewirkt eine stärkere Fragmentation des Magmas, aber keine höhere Austrittsgeschwindigkeit des Magmapropfens und auch keine grössere seismische Amplitude. Auch wenn die langperiodischen Signale, wie beispielsweise Tilt, im Labor nicht simuliert werden konnten, sind die Blow-out Experimente überraschend gut in der Lage, die am Vulkan Stromboli registrierten kurzperiodischen seismischen Signale zu reproduzieren
The main purpose of volcano-seismology concerns the qualitative and quantitative description of one or more unknown seismic source(s) located at some unknown depth beneath a volcano. Even if many different volcanoes show similar seismic signal characteristics, up to now it was not possible to find a standard seismic source model for volcanoes, as the double-couple in earthquake seismology. Volcanoes with a continuous activity, like Stromboli (Italy), represent for the volcano seismologist a perfect natural laboratory to address this question. This thesis treats the study of explosion-quakes and volcanic tremor recorded on Stromboli in a broadband frequency range, and discusses the location and the possible mechanisms of the seismic source(s). Seismic and infrasonic recordings of explosion-quake from Stromboli showed that the high-frequency phase propagates with a velocity of approximately 330 m/s. The seismic source can be explained as an explosion at the top of the magma column generated by rising gas bubbles. The seismic P-wave and the air-wave are both generated in the same point at the same time. The different path lengths and velocities for the seismic wave and the air-wave result in a difference in arrival times dt, that could be used to deduce the magma level and sound speed in the eruption column inside the conduit. Stations installed near the active crater reveal that infrasonic and seismic recordings of the short-period tremor (> 1 Hz) share the same spectral content and show similar energy fluctuations. Therefore, the short-period volcanic tremor at Stromboli originates from the continuous out-bursting of small gas bubbles in the upper part of the magmatic column. The spectrum of the long-period tremor recorded at Stromboli consists of three main peaks with periods at 4.8 s, 6 s and 10 s, and amplitudes varying with the regional meteorological situation. Hence, they are not generated by a close volcanic source but rather by ocean microseisms (OMS). The passage of a local cyclone seems to be the seismic source for spectral energy at 4.8 s and 10 s, which represent the Double Frequency and the Primary Frequency of the OMS, respectively. Concerning the 6 s peak, a cyclone near the British Isles could act as a seismic source. Seismic data from the first broadband array deployed on Stromboli showed surprisingly simple waveforms, indicating an initially contracting source mechanism. The analysis of particle motion and the application of seismic array techniques allowed the location of a seismic source in the shallow part of the volcano. Eruption parameters and seismic source characteristics of the April 5, 2003 Stromboli eruption have been estimated using different inversion approaches. The paroxysm was triggered by a shallow slow thrust-faulting dislocation event with a moment magnitude of Mw = 3.0 and possibly associated with a crack that formed previously by dike extrusion. At least one blow-out phase during the paroxysmal explosion could be identified from seismic signals with an equivalent moment magnitude of Mw = 3.7. It can be represented by a vertical linear vector dipole and two weaker horizontal linear dipoles in opposite direction, plus a vertical force. Seismic measurements performed during controlled and reproducible blow-out experiments with a gas volume entrapped in basaltic melt revealed the following: Monochromatic seismic signals suggest a blow-out in a more ductile regime, whereas broader frequency content indicates rupture in a more brittle environment. The longer the crucible, the weaker the seismic signals. An increase in pressure results in a stronger fragmentation, but not in a higher ejection velocity of the plug neither in a higher seismic amplitude. Even if the very long period observations like the tilt signal could not be simulated in the laboratory, the blow-out experiments simulate very well the short-period seismic signals recorded at Stromboli volcano