Search

International audience; Recent rifting episodes highlight the role of magmatic systems with propagating dikes on crustal spreading. However, our knowledge of magmatic systems is usually limited to surface observations and geophysical data. Eastern Iceland allows direct access to extinct and eroded deeper magmatic systems. Here we collected field structural and AMS (anisotropy of magnetic susceptibility) data on 187 and 19 dikes, respectively, in the 10-12 Ma old Alftafjordur magmatic system. At a paleodepth of~1.5 km, the extension due to diking is at least 1-2 orders of magnitude larger than that induced by regional tectonics, confirming magmatism as the key mechanism for crustal spreading. This magma-induced extension, inferred from the aspect ratio of the magmatic system, was of~8 mm/yr, lower than the present one. AMS data suggest that most of dikes have geometrically normal fabric, at least at the margins, consistent with prevalent subvertical magma flow and propagation.

Understanding shallow magma transfer and the related vent distribution is crucial for volcanic hazard. Here we investigate how the stress induced by topographic scarps linked to normal faults affects the distribution of monogenic volcanoes at divergent plate boundaries. Our numerical models of dyke propagation below a fault scarp show that the dykes tend to propagate towards and erupt on the footwall side. This effect, increasing with the scarp height, is stronger for dykes propagating underneath the hanging wall side, and decreases with the distance from the scarp. A comparison to the East African Rift System, Afar and Iceland shows that: 1) the inner rift structure, which shapes the topography, controls shallow dyke propagation; 2) differential loading due to mass redistribution affects magma propagation over a broad scale range (100–105 m). Our results find application to any volcanic field with tectonics- or erosion-induced topographic variations, and should be considered in any volcanic hazard assessment.

The axis of divergent plate boundaries shows extension fractures and normal faults at the surface. Here we present evidence of contraction along the axis of the oceanic ridge of Iceland and the continental Main Ethiopian Rift. Contraction is found at the base of the tilted hanging wall of dilational normal faults, balancing part of their extension. Our experiments suggest that these structures result from dike emplacement. Multiple dike injection induces subsidence above and uplift to the sides of the dikes; the transition in between is accommodated by reverse faults and subsequent peripheral inward dipping normal faults. Our results suggest that contraction is a direct product of magma emplacement along divergent plate boundaries, at various scales, marking a precise evolutionary stage and initiating part of the extensional structures (extension fractures and normal faults).

Plate boundaries are the most active, unstable and hazardous areas on Earth. The aim of this study is twofold. 1) Provide an overview of the main structural features along divergent and convergent plate boundaries and their relationships to volcanism. The considered divergent plate boundaries include the continental East African Rift System (EARS), the transitional Afar Rifts and slow (Iceland) and fast (East Pacific Rise) oceanic ridges. The analysis of the convergent plate boundaries refers to the extensional (Taupo Volcanic Zone, New Zealand), strike-slip (Sumatra), contractional (NE Japan) and more complex (Central Andes) volcanic arcs. 2) Propose an original and innovative frame to understand tectono-magmatic processes along plate boundaries, based on two major points. a) Magmatism may effectively control the development of plate boundaries. At immature continental divergent plate boundaries (as the non-magmatic portions of the EARS), regional extension plays a major role in extending the upper crust; however, along mature continental and oceanic rifts magmatism is most effective in spreading plates apart through dikes. At convergent plate boundaries, the possibility to develop extensional, strike-slip, contractional and oblique systems along the volcanic arc provides highly variable structural conditions controlling magmatism. At least three magmatic processes (diking, sill emplacement, heat-induced strain localization) may control the development of volcanic arcs. In addition, megathrust earthquakes may induce transient extension, enhancing volcanism. Therefore, the evolution of both divergent and convergent plate boundaries is significantly magma-controlled and magmatic processes seem largely self-sustained, requiring limited direct tectonic contribution; and b) magmatism along divergent and convergent plate boundaries usually focuses in magmatic systems, associated with a dominant volcano. Magmatic systems provide a convenient framework to relate the geometric and compositional features of a magmatic plate boundary to its tectonic setting. In particular, the larger the opening rate, the lower is the aspect ratio A (A = width W/length L) of the magmatic system. Therefore, the linear magmatic accretion of divergent plate boundaries and extensional arcs is replaced by a punctiform accretion in contractional arcs. Several aspects (magmatic rates, lateral vs. vertical dike propagation along divergent plate boundaries, structural control on volcanism along strike-slip arcs) are still poorly defined and should be the target of future research.

[1] Magma transfer, i.e., dike propagation, is partly controlled by Young's modulus (elasticity) contrasts (ratio upper layer to lower layer modulus) in the host rock. Here we try to better constrain the elasticity contrasts controlling the propagation velocity of dikes and their arrest. We simulate dike propagation in layered elastic media with different elasticity contrasts. Salted gelatin and water represent host rock and magma, respectively. For common density ratios between magma and host rock (~1.1), velocity variations are observed and a critical threshold in the elasticity contrast between layers results in the Young's modulus ratio of 2.1 ± 0.6. Naturally occurring elasticity contrasts can be much higher than this experimental threshold, suggesting that dike arrest due to heterogeneous elastic host rock properties is more frequent than expected. Examples of recently deflected or stalled dikes inside volcanoes and the common presence of high-velocity bodies below volcanoes suggest that better defining elasticity contrasts below volcanoes helps in forecasting eruptions.

The evaluation of the kinematic setting and the structural control of volcanic arcs are important in defining the tectono-magmatic processes along convergent plate boundaries. However, our knowledge is fragmented and the available data highlight different behaviours. This study analyzes the kinematic setting, the structural control and the volcanic productivity of 16 arcs. These arcs are characterized by predominant extensional, compressional, strike–slip or oblique motions. There is an overall coupling between the normal vs. parallel motions along the arc and those of the underlying slab. Therefore, the higher the trench-normal (or parallel) component of the subduction rate, the higher is the amount of arc-normal (or parallel) motion. This relation confirms that strain partitioning is, in general, feasible at many convergent settings, involving also the volcanic arc portion. The arc-normal motion may be characterized by extension or compression, as a function of the increase in the trench-normal convergence (or subduction) velocity. The lack of an evident relation between the subduction rate (or convergence rate) and the volcanic productivity of the arc is partly inconsistent with previous studies, which highlight a relation between the subduction rate and melt production below arcs. This discrepancy mainly suggests that different processes control the generation, rise and eruption of magma at different depths, varying the intrusive/extrusive ratio along arcs. The structure of the arc does not control the distribution of the volcanoes; however, it does control the volcanic output through different processes, even though regional or local extension (associated with strike–slip or compressive structures) is the ultimate requisite, in any setting. In general, the higher the amount of extension, the higher is the output rate along the arc.

Calderas are topographic depressions formed by the collapse of a partly drained magma reservoir. At volcanic edifices with calderas, eruptive fissures can circumscribe the outer caldera rim, be oriented radially and/or align with the regional tectonic stress field. Constraining the mechanisms that govern this spatial arrangement is fundamental to understand the dynamics of shallow magma storage and transport and evaluate volcanic hazard. Here we show with numerical models that the previously unappreciated unloading effect of caldera formation may contribute significantly to the stress budget of a volcano. We first test this hypothesis against the ideal case of Fernandina, Galápagos, where previous models only partly explained the peculiar pattern of circumferential and radial eruptive fissures and the geometry of the intrusions determined by inverting the deformation data. We show that by taking into account the decompression due to the caldera formation, the modeled edifice stress field is consistent with all the observations. We then develop a general model for the stress state at volcanic edifices with calderas based on the competition of caldera decompression, magma buoyancy forces and tectonic stresses. These factors control: 1) the shallow accumulation of magma in stacked sills, consistently with observations; 2) the conditions for the development of circumferential and/or radial eruptive fissures, as observed on active volcanoes. This top-down control exerted by changes in the distribution of mass at the surface allows better understanding of how shallow magma is transferred at active calderas, contributing to forecasting the location and type of opening fissures.

The Taupo Volcanic Zone (TVZ), New Zealand, comprises several segments with variable components of extension, and is characterised by longitudinal segmentation into a central zone dominated by rhyolitic calderas and extremities with only andesitic composite volcanoes. Modern (< 300ka) TVZ was investigated to define (a) the overall relationship between regional structures and volcanism and (b) the structural control on the evolution of calderas. New remote sensing and field structural data have been combined with previously published data for TVZ. The results show a general correlation between the amount of extension and the volume and style of eruption in each segment. The segments with the greatest extension coincide with the highly active Taupo and Okataina caldera complexes; conversely, the segments with the greatest dextral transtension correspond to volumetrically less active andesitic stratovolcanoes. Within Modern TVZ, two types of caldera are distinguished based on their structure and development. Calderas within the main zone of rifting (Taupo, Okataina) are multiple (two or more major eruptions) collapse structures with rectilinear margins, overprinted by younger volcanism and faulting. Their complexity is related to the proximity (and influence) of active faulting within TVZ. In contrast, calderas peripheral to the main rift (Rotorua, Reporoa) are sub-circular monogenetic collapse structures, with minor post-collapse volcanism. Their simpler evolution and structure is a attributable to their location away from the main rift zone. Thus, TVZ exemplifies the situation where firstly, magma is erupted as a function of the extension rate along the axial rift zone, and secondly, volcanism is manifest as distinctive caldera structures according to their proximity to the axial rift zone.

The volcanic island of Ischia is located on the Tyrrhenian margin of Central Italy, characterized by Plio-Quaternary NW–SE- and NE–SW-trending extensional fractures. Ischia displays a resurgent dome uplifted by at least 800 m in the last 33 ka. Remote sensing and field data have been collected to study the structural setting of the island, the deformation pattern associated with resurgence and the superimposition of the regional and the resurgence-induced stress fields. NW–SE and NE–SW extensional fracture systems predominate throughout the island and around the resurgent block, suggesting a relationship with the regional extensional structures. These systems were formed before resurgence and were partly reactivated during resurgence. The reactivation of pre-existing regional systems during resurgence confined the extent of the uplifted area. N–S- and E–W-trending systems have been found exclusively at the borders of the dome and are interpreted as being induced by resurgence. The topmost resurgent block shows an octagonal shape in map view and is tilted at an angle of 15° around a NE–SW-trending horizontal axis; the block is partly bordered by high-angle, inward-dipping regional faults. More than 90% in volume of the volcanic products coeval with resurgence on Ischia have been erupted outside the resurgent block area, suggesting that the resurgence process locally replaced volcanic activity in the last 33 ka.

Mount Etna produces frequent eruptions from its summit craters and from fissures on its flanks. The flank fissures trend approximately radially to the summit, and are mainly concentrated in three rift zones that are located on the NE, S and W flanks. Many flank eruptions result from lateral magma transfer from the central conduit into fractures intersecting the flanks, although some eruptions are fed through newly formed conduits that are not directly linked to the central conduit. We analyzed the structural features of eruptions from 1900 to the present, one of the most active periods in the documented eruptive history of Etna, which comprised 35 summit and 33 flank events. Except for a small eruption on the W flank in 1974, all of the flank eruptions in this interval occurred on or near the NE and S rifts. Eruptions in the NE sector were generally shorter, but their fissure systems developed more rapidly and were longer than those in the S sector. In contrast, summit eruptions had longer mean durations, but generally lower effusion rates (excluding paroxysmal events characterized by very high effusion rates that lasted only a few hours). This database was examined considering the main parameters (frequency and strike) of the eruptive fissures that were active over the last ~2 ka. The distribution in time and space of summit and flank eruptions appears to be closely linked to the dynamics of the unstable E to S flank sector of Etna, which is undergoing periodic displacements induced by subvolcanic magma accumulation and gravitational pull. In this framework, magma accumulation below Etna exerts pressure against the unbuttressed E and S flanks, which have moved away from the rest of the volcano. This has caused an extension to the detachment zones, and has facilitated magma transfer from the central conduit into the flanks.

Mount Etna is characterized by significant flank instability, whose triggering factors are still a matter of debate. We use analog models to investigate the role of different factors. In the models, a cone and base of granular material simulate the volcanic edifice and its basement. The asymmetric geometry of the basement simulates the topographic gradient around the volcano. Injections of silicone (sets A and B) and low-viscosity vegetable oil (set C) simulate the pressurization of plutonic complex, deep reservoirs, and the emplacement of dikes, respectively. Other experiments (set D) reproduce regional extensional tectonics in the last 10(5) years, within layers with different cohesion simulating strength differences in the basement. Laser scanner and control points allow the tracking of surface deformation with submillimeter precision. The asymmetric topography enhances flank instability on the side with the weakest confinement (i.e., seaside), providing the preparing factor for instability. In sets A and B, any type of pressurized reservoir enhances, up to 10 times, the amount of flank instability toward the seaside with respect to the other flanks of the volcano. In set C, dike emplacement enhances seaside flank instability up to 8 times. Regional tectonics and crustal layering in set D enhance flank instability up to 2 and 1.3 times, respectively. Considering the duration and frequency of the simulated processes in nature, we propose a semiquantitative evaluation and hierarchy of the factors controlling flank instability at Etna. Magmatic activity (point-like source inflation and dike emplacement) provides the most important triggering factor. Extensional tectonics, in the last 10(5) years, and crustal layering are more than 10 times less effective. This study shows the importance of differential buttressing at the volcano base for flank instability. This condition is an important indication to expect asymmetric activity in any volcano (flank eruptions and deformation).

This work was funded by Istituto Nazionale di Geofisica e Vulcanologia (INGV) and the Italian Dipartimento per la Protezione Civile (DPC) (DPC-INGV project V4 “Flank”).

[1] Fracture reactivation is a widespread process in nature even though evidence of magma-induced reactivation is less documented. Here we provide evidence of the reactivation of a fracture system on the upper flank of the Mt. Etna volcano and consider its possible implications in understanding the recent volcanic and tectonic activity. A NNW–SSE trending fracture, partly accompanied by magma emplacement in the form of a laterally propagating dike, formed in 1989 on the upper SE flank of Etna. Lava effusions in 1991–1993, 2001, and 2006 were associated with volcano-tectonic (VT) seismicity and ground deformations on the upper part of the volcano, which document the seismogenetic involvement of the 1989 fractures, although without eruptive phenomena along the discontinuity. In addition to the aforementioned episodes of VT seismicity, differences in the characteristics of the background seismic radiation (volcanic tremor) were measured at stations close to these fractures during the eruptive activity on 24 November 2006, for which more detailed volcanological and seismological time histories are available. Moving on from these findings, we analyze volcanic tremor data recorded close to the summit and along the S flank of the volcano to highlight the interactions between seismic radiation and the 1989 fracture system. Centroid location of volcanic tremor and wave field characteristics at stations of the permanent local seismic network of Etna highlight the guidance role played by the 1989 fractures during the eruptive activity on 24 November 2006. In addition, the collected data shed light on hitherto unknown structural features, which appear to connect the volcano summit to the lower SE slope and also play an important role in controlling the instability of the E flank. More generally, this study shows how (1) using an integrated approach, it is possible to link apparently different features to a common structure, showing uniform and distinct dynamics relevant at the volcano scale, and (2) fracture reactivation can also occur by means of magma intrusion, playing an important role in the transfer of magma within a volcanic edifice.

[1] Volcano deformation may occur under different conditions. To understand how a volcano deforms, as well as relations with magmatic activity, we studied Mt. Etna in detail using interferometric synthetic aperture radar (InSAR) data from 1994 to 2008. From 1994 to 2000, the volcano inflated with a linear behavior. The inflation was accompanied by eastward and westward slip on the eastern and western flanks, respectively. The portions proximal to the summit showed higher inflation rates, whereas the distal portions showed several sectors bounded by faults, in some cases behaving as rigid blocks. From 2000 to 2003, the deformation became nonlinear, especially on the proximal eastern and western flanks, showing marked eastward and westward displacements, respectively. This behavior resulted from the deformation induced by the emplacement of feeder dikes during the 2001 and 2002–2003 eruptions. From 2003 to 2008, the deformation approached linearity again, even though the overall pattern continued to be influenced by the emplacement of the dikes from 2001 to 2002. The eastward velocity on the eastern flank showed a marked asymmetry between the faster sectors to the north and those (largely inactive) to the south. In addition, from 1994 to 2008 part of the volcano base (south, west, and north lower slopes) experienced a consistent trend of uplift on the order of ∼0.5 cm/yr. This study reveals that the flanks of Etna have undergone a complex instability resulting from three main processes. In the long term (103–104 years), the load of the volcano is responsible for the development of a peripheral bulge. In the intermediate term (≤101 years, observed from 1994 to 2000), inflation due to the accumulation of magma induces a moderate and linear uplift and outward slip of the flanks. In the short term (≤1 year, observed from 2001 to 2002), the emplacement of feeder dikes along the NE and south rifts results in a nonlinear, focused, and asymmetric deformation on the eastern and western flanks. Deformation due to flank instability is widespread at Mt. Etna, regardless of volcanic activity, and remains by far the predominant type of deformation on the volcano.

[1] Oceanic ridge segments interact forming different types of structures, of which transform faults and overlapping spreading centers represent the most common end-members. This study uses analogue models for testing the role of the initial configuration of the ridges on the geometry and kinematics of the resulting interaction zones. The models, run in a centrifuge, consist of a lighter silicone (LDS), rising within a denser silicone (HDS) below a brittle layer simulating the oceanic lithosphere. Precut fractures in the brittle layer simulate the initial configuration of oceanic ridges. The rise and lateral spreading of the LDS induce the propagation, widening, interaction, and linkage of the fractures. We varied the offset and overlap between the fractures, obtaining 10 distinctive types of interaction. Variations in fracture overlap and offset define several interaction types. Increase of fracture overlap leads to interaction zones of lower aspect ratio, with fractures propagating at lower angles to the mean extension direction. Increase of offset leads to the elongation of the interaction zones with the new fractures propagating subparallel to the extension direction. The comparison with several examples of natural ridges shows close geometric similarities, confirming the existence of predominant types of ridge interaction. Among these, ridges with smaller offsets develop interactions similar to overlapping spreading centers, whereas ridges with larger offsets have geometries reminiscent of transform zones. The comparison between experimental and natural examples suggests that the observed wide spectrum of ridge interaction types in nature results from the initial configuration of the divergent plate boundary.

Panarea, characterized by gas unrest in 2002–2003, is the volcanic island with the least constrained structure in the eastern-central Aeolian Arc (Italy). Based on structural measurements, we define here its deformation pattern relative to the Arc. The main deformations are subvertical extension fractures (63% of data), normal faults (25%) and dikes (12%). The mean orientation of the extension fractures and faults is ∼N38°E, with a mean opening direction of N135° ± 8°, implying extension with a moderate component of dextral shear. These data, matched with those available for Stromboli volcano (pure opening) and Vulcano, Lipari and Salina volcanoes (predominant dextral motions) along the eastern-central Arc, suggest a progressive westward rotation of the extension direction and an increase in the dextral shear. The dextral shear turns into compression in the western arc. The recent unrest at Panarea, coeval to that of nearby Stromboli, may also be explained by the structural context, as both volcanoes lie along the portion of the Arc subject to extension.

No eruption, no caldera collapse, and no large landslide can takeplaceinavolcanounlessitsstateofstressissuitablefortheassociatedtype of rock failure. The state of stress, in turn, results in deformation,andbothstressanddeformationdependonthemechanicalpropertiesof the rocks that constitute the volcano. Understanding stress anddeformation in volcanoes is thus of fundamental importance forunderstanding unrest periods and for accurate forecasting volcanofailure, such as may result in large-scale lateral and vertical collapsesand eruptions.Deformation studies in volcanic areas have significantly improved,both in quality and quantity, during the past decade. While theinterest of the scientific community, and the volcanological commu-nity in particular, has primarily been in the description andinterpretation of the deformation, especially at active volcanoes,new and powerful techniques have recently become available toimprove our understanding of the local stresses and mechanicalproperties of volcanoes and how they relate to surface deformation.Volcano deformation studies in the last years have of courseincluded standard field and analytical techniques, but in additionthere has been increasing use of advanced analogue and numericalmodels, as well as modern geophysical monitoring techniques such asGPS and InSAR. All these techniques have contributed to an improvedunderstanding of various aspects of both regional and local volcanodeformation. At the regional scale, the relationship between tectonicsand magmatism at divergent plate boundaries is now better under-stood (e.g. Kendall et al., 2005; Wright et al., 2006) and so is thegeneral relationship between arc volcanism and seismicity (e.g.Walter and Amelung, 2007). At a more local scale, significantimprovements have been made in understanding the mechanics ofmagma-chamber rupture (e.g. Gudmundsson, 2006), dyke propaga-tion(e.g.Polandetal.,2004;Gudmundssonetal.,2008),thestructureof calderas (e.g. Acocella, 2007; Gottsmann and Marti, 2008; Geshi,2009), the collapse of volcanoes (e.g. Cervelli et al., 2002; Amelunget al., 2007), the relationships between volcanic activity and edificedeformation (e.g. Neri et al., 2009), the seismicity associated withvolcanism (Benson et al., 2008) and the interactions betweenvolcanoes (e.g. Miklius and Cervelli, 2003; Andrew and Gudmunds-son, 2008). These and other studies confirm the crucial importance ofunderstandingdeformationandstressinvolcanicareas,asisgenerallyrecognised by the volcanological and tectonic communities.Oneimportanttopicrelatedtovolcanodeformationandstressthatis still not well understood is how active deformation can be used toforecast eruptions and their behaviour. This applies particularly to themechanics of dike emplacement. The vast majority of volcaniceruptions are fed by dikes, and understanding dike propagation andarrest is necessary so as to mitigate various hazards associated withvolcanic eruptions, caldera collapses, and landslides. In particular,major eruptions, seismicity, and sector collapses pose threats out togreat distances from volcanoes. Understanding these processes,however, depends on knowing the more general relationshipsbetween deformation, stress, and volcanic activity.The structure and contents of this special issue are motivatedprimarily by our belief that understanding deformation and stress inactive volcanoes is a necessary condition for reliable forecast ofvolcanic eruptions, caldera collapses, and large landslides. Thesetopics all belong to the field of volcanotectonics. Also, we think thatcorrect interpretation of volcano deformation, in terms of localstresses and mechanical properties, is fundamental for understandingvolcano behaviour and, thereby, assessing volcanic hazards.This special issue derives from a session on volcanotectonics at the2007EGUAssemblyinVienna,Austria.Thefocusisonfivemaintopicswithin the field of volcanotectonics. The first topic concerns generalaspects of volcano deformation, stress, and failure (Bonafede andFerrari, 2009-this volume; Eggert and Walter, 2009-this volume;Gudmundsson, 2009-this volume). The second topic is collapsecalderas (Battaglia and Hill, 2009-this volume; Jonsson, 2009-thisvolume; Kusumoto and Gudmundsson, 2009-this volume), followedby dike propagation and arrest (Acocella and Neri, 2009-this volume;Contents lists available at ScienceDirect

Eruptions are fed by dikes; therefore, better knowledge of dike propagation is necessary to improve our understanding of how magma is transferred and extruded at volcanoes. This study presents an overview of dike patterns and the factors controlling dike propagation within volcanic edifices. Largely based on published data, three main types of dikes (regional, circumferential and radial) are illustrated and discussed. Dike pattern data from 25 volcanic edifices in different settings are compared to derive semi-quantitative relationships between the topography (relief, shape, height, and presence of sector collapses) of the volcano, tectonic setting (presence of a regional stress field), and mean composition (SiO2 content). The overview demonstrates how dike propagation in a volcano is not a random process; rather, it depends from the following factors (listed in order of importance): the presence of relief, the shape of the edifice and regional tectonic control. We find that taller volcanoes develop longer radial dikes, whose (mainly lateral) propagation is independent of the composition of magma or the aspect ratio of the edifice. Future research, starting from these preliminary evaluations, should be devoted to identifying dike propagation paths and likely locations of vent formation at specific volcanoes, to better aid hazards assessment.

[1] A structural model for Campi Flegrei Caldera (CFC; Italy), one of Earth's most hazardous calderas, is provided. Data from the north (outer) rim show that this formed by downsag soon after the Neapolitan Yellow Tuff emplacement (∼15 ka). Campanian Ignimbrite's (CFC's other major eruption; ∼39 ka) displacement at depth also suggests the activity of a buried caldera fault. Therefore, during the Campanian Ignimbrite eruption the north rim was activated as caldera-bounding fault and during the Neapolitan Yellow Tuff eruption as downsag. Previously published data reveal an inner collapsed structure to the south, formed during the Campanian Ignimbrite eruption and reactivated during the Neapolitan Yellow Tuff eruption. So, CFC consists of 2 nested depressions, both active during the Campanian Ignimbrite and Neapolitan Yellow Tuff eruptions, but with different amounts of collapse. CFC is a major example of how 2 nested calderas may: (1) form during the same collapse episode; (2) be reactivated together during caldera-forming eruptions.

Several factors may control the development of a silicic caldera during an eruption: here we match previous studies of caldera forming eruptions with analogue experiments to consider the role of the aspect ratio (thickness/width) of the magma chamber roof. These data suggest that large silicic caldera formation is controlled by: (1) the availability of magma; (2) initial explosive eruption through a conduit connecting the reservoir and the surface; (3) fast depressurization of the reservoir; and (4) chamber roof collapse, with newly-formed fractures connecting the reservoir with the surface and creation of a continuous ring conduit feeding annular vents, capable of sustaining the eruption. The occurrence of the latter condition may depend upon the aspect ratio of the chamber roof: lower ratios (< 1.6) form coherent collapse, ensuring connection from the chamber to the surface; higher ratios (> 1.6) form incoherent collapse, which may hinder the continuation of the eruption after an initial partial emptying of the chamber. The aspect ratio of the chamber roof should therefore be considered as a further factor, capable of controlling the formation of large silicic calderas.

The Eastern Cordillera (Central Andes, ∼ 24°S) consists of a basement-involved thrust system, resulting from Miocene-Quaternary eastward migrating compression, separating the Puna plateau from the Santa Barbara System foreland. The inferred Tertiary strains arising from shortening in the Eastern Cordillera and Santa Barbara System are similar, higher than in the Puna. Slip data collected on the major ∼ N-S trending faults of Eastern Cordillera show a westward progression from dip-slip (contraction) to dextral and sinistral motions. This, consistently with established tectonic models, may result from partitioning due to the oblique Mio-Quaternary underthrusting of the Brazilian Shield north of 24°S. This strain partitioning has three main implications. (1) As the dextral and sinistral shear in the Eastern Cordillera are ∼ 62% and 29% of the compressive strain respectively, the Eastern Cordillera results more strained than Santa Barbara System foreland, contrary to previous estimates. (2) The partitioning in the Eastern Cordillera may find its counterpart in that to the west of the Central Andes, giving a possible structural symmetry to the Central Andes. (3) The easternmost N-S strike-slip structures in the Eastern Cordillera coincide with the easternmost Mio-Pliocene magmatic centres in the Central Andes, at ∼ 24°S. Provided that, further to the east, the crust is partially molten, the absence of magmatic centres may be explained by the presence of pure compressive structures in this portion of the Eastern Cordillera. Fil: Acocella, V.. Università Roma Tre; Italia Fil: Vezzoli, L.. Università degli studi dell'Insubria; Italia Fil: Omarini, Ricardo Hector. Universidad Nacional de Salta. Facultad de Ciencias Naturales; Argentina. Consejo Nacional de Investigaciones Científicas y Técnicas; Argentina Fil: Matteini, M.. Universidade do Brasília; Brasil Fil: Mazzuoli, R.. Università degli Studi di Pisa; Italia

[1] Understanding how shallow magma is emplaced within volcanoes is crucial for hazard assessment. The 2002–2003 Stromboli eruption provides the opportunity to investigate shallow magma emplacement resulting from orthogonal feeder dikes and its possible effects. Stromboli erupted in 2002–2003, when effusive activity replaced Strombolian activity. On December 28, a NE-SW fissure propagated from the lava-filled northernmost summit crater. On December 29, a NW-SE fissure propagated north of the craters, feeding NW-SE aligned vents. On December 30, this area collapsed, reaching the sea and generating a tsunami. In mid February 2003, the NW-SE fissure became inactive, while the NE-SW effusive fissure continued until July. A model for shallow magma emplacement is proposed. The lateral propagation of a NE-SW dike from the northernmost crater was triggered. Below, a NW-SE dike, propagating from the magma-filled NE tip of the NE-SW elongated conduit, fed the NW-SE aligned vents. In February, the conduit periphery became solidified, freezing the NW-SE dike, and the transport of magma was limited to the central part, focusing its rise below the craters. This fed the NE-SW fissure until the supply decreased further (July), returning to the ordinary level sustaining Strombolian activity. Orthogonal dike emplacement followed the trajectories of the maximum (gravitational) stress σ1, partly controlled by the irregular topography of the uppermost edifice. The emplacement of orthogonal dikes in a limited area is feasible at non-perfectly conical active volcanoes, where the maximum gravitational stress may show variations from a purely radial path.

[1] Analogue experiments investigate how flank collapse affects dike propagation within volcanoes. Water (magma analogue) is injected within a gelatin cone (volcano analogue) with a lateral collapse. The injections form dikes that, away from the collapse, become radial. The dikes propagating nearby the collapse focus towards the collapse sides, becoming subparallel to them, because of the stress reorientation due to the unbuttressing. Only dikes formed along the collapse axis (passing through the mean points of the collapse, map view) propagate radially within the collapse. This general model is applied at Stromboli. The Stromboli dikes cluster along and parallel to the collapse sides, similar to the experiments. Nevertheless, the current location of the conduit, along the collapse axis, leads to dike propagation within the collapse infill, rather than at its sides. This may have occurred during the 2002–03 eruption, when diking in the collapse triggered a landslide and tsunami within the collapse.

Mount Etna produced two significant eruptions in 2001 and 2002–2003, which we have analysed using geological, seismic and deformation data. These eruptions showed some similarities, such as the activating of two magmatic plumbing systems (central–lateral and eccentric), but they differed in their triggering mechanisms. While the 2001 eruption was largely the result of the emplacement of a N–S eccentric dike (independent from the central conduits) consistent with E–W regional extension, the 2002–2003 eruption occurred in response to a major flank slip on the eastern and southeastern sides of the volcano. This is demonstrated by the spatial and temporal distribution of seismicity and deformation preceding and accompanying the two eruptions. During the months prior to the 2001 eruption, most epicenters were concentrated on the southern flank, at depths of 5–15 km below sea level. During the 4 days before the eruption, earthquake hypocenters migrated to shallower levels (from 5 km bsl. upward) indicating the emplacement of the eccentric dike. This is confirmed by the patterns of ground fracturing observed in the field and deformation documented by electronic distance measurements (EDM). In contrast, the months before the 2002–2003 eruption were characterised by shallower seismicity, mainly concentrated along the active faults bordering the slipping flank sector. Flank slip accelerated in September 2002 and a second, more vigorous acceleration of flank slip occurred on 26–27 October 2002, accompanying the opening of eruptive vents. The very short (2 h) seismic crisis preceding the onset of eruptive activity stands in neat contrast with the 4 days of intense seismicity before the 2001 eruption. Subsequently, flank slip-deformation extended all over the eastern and southeastern flanks of the volcano, causing serious damage in this sector. The events of 2001–2003 can be seen as a continuous chain of intimately interacting processes including regional tectonics, magma accumulation and eruption, and flank instability. In this scenario the 2001 eruption led to increased flank instability that subsequently accelerated and culminated with the massive flank slip, which in turn facilitated the 2002–2003 eruption. This sequence of events points to a long-term feedback mechanism between magmatism and flank instability at Etna.

The structure and shape of collapses and resurgences is often controlled by pre-existing discontinuities, such as normal faults in rift zones. In order to study the role of extensional structures on collapse and resurgence, we used analogue models. Dry sand simulated the brittle crust; silicone, located at the base of the sand-pack, simulated magma. In the experiments, regional extension pre-dated collapse or resurgence, forming normal faults in a graben-like structure; the graben was filled with additional sand, simulating post-rift deposits. A piston then moved the silicone downward or upward, inducing collapse or resurgence within the previously deformed sand. The collapses showed an ellipticity (length of minor axis/length of major axis) between 0.8 and 0.9, with the major axis parallel to the extension direction. The partial reactivation of the pre-existing normal faults was observed during the development of the caldera reverse faults, which, conversely to what was expected (from experiments without pre-existing extension), became partly inward dipping. Resurgence showed an elongation of the uplifted part, with the main axis perpendicular to the extension direction. At depth, pre-existing normal faults were partly reactivated by the reverse faults formed during resurgence; these locally became outward dipping normal faults. A total reactivation of pre-existing faults was also observed during resurgence. The experiments suggest that the observed elongation of calderas and resurgences is the result of the reactivation of pre-existing structures during differential uplift. Such a reactivation is mainly related to the loss in the coefficient of friction of the sand. The results suggest that elliptic calderas and resurgences in nature may develop even from circular magma chambers.

The Ethiopian Rift is characterized by several Quaternary calderas. Remote sensing and field analyses were used to investigate the regional structural control on three calderas (Fantale, Gariboldi, Gedemsa) in the axial part of the rift. These calderas are located along the Wonji Fault Belt (WFB), a zone of Quaternary NNE–SSW normal faults and extensional fractures. The three calderas show E–W elongation and major E–W vent alignments, oblique with regard to the mean NW–SE extension direction. No significant evidence of E–W tectonic structures has been found near the calderas, the only relevant systems being those of the WFB. Conversely, left-lateral E–W-trending faults are present at the rift borders and on the Nubia and Somalia plateaus, implying a predominant pre-rift activity. The E–W fractures were partly reactivated during rifting, possibly controlling the development of the magma chambers. Thus, the E–W elongation of the calderas would be the surface expression of such a control, rather than the result of regional extension. An evolutionary model on the role of different structures on magmatism at different crustal levels within the rift is proposed.

We present centrifuge experiments to study the surface deformation induced by shallow pluton emplacement in a rheologically stratified crust. Sand simulates the topmost brittle crust; plastilina and denser silicone represent more and less competent crustal portions, respectively; lighter silicone simulates a buoyant intrusion. In the models, density differences affect the rate of intrusion but not their evolution or shape, whereas viscosity and strength stratifications control both the shape and rate of the intrusions. With a higher viscosity contrast (10 2–4 ) between the intrusion and the embedding media, the rise of the lighter silicone induces a laccolith-like intrusion, responsible for doming and thinning of the overburden; an apical depression may form, inducing silicone extrusion. Conversely, with a lower (10 1 ) viscosity contrast, the overburden and the intrusion exhibit a lens-shaped form, with a broad central depression bordered by an upward flexure towards the periphery. A sag in the floor of the intrusion is commonly observed; no silicone extrusion occurred. The intrusion is a hybrid between a laccolith and a lopolith. The comparison with nature (1) confirms roof uplift as an important means of accommodating space during pluton emplacement and (2) suggests that, where roof uplift plays a major role, pluton emplacement can induce a well-correlated sequence of events at surface: doming, the development of a depression and extrusion.

Resurgent doming consists of the uplift, usually accompanied by volcanic activity, of part of a collapse caldera. Analogue models were used to investigate the architecture of resurgent domes. Dry sand simulates the brittle crust; uprising silicone, located at the base of the sand-pack, simulates magma. The deformation pattern depends mainly upon: (1) the ratio (aspect ratio) between the thickness of the sand overburden and the width of the silicone intrusion; (2) the duration of experiment. For aspect ratios ≈1, two concentric domes develop; the first-formed outer dome is bordered by inward-dipping reverse ring faults, while the inner dome by outward-dipping normal ring faults. The layers inside the dome are uniformly dipping. For aspect ratios ≈0.4, the dome shows a crestal depression, surrounded by radial fractures, followed by an apical extrusion of silicone. The internal structure of the dome is made up of domed layers. Independently from the aspect ratio, the duration of the experiment enhances silicone extrusion. A consistent structure is observed in most resurgent domes in nature. The comparison between experiments and nature suggests that two distinct resurgence modes occur, mainly depending on the aspect ratio (thickness/width) of the crust overlying the magma chamber. Aspect ratios ≈1 develop a resurgent block with uniformly-dipping layers and peripheral volcanic activity (Ischia and Pantelleria type). Aspect ratios ≈0.4 develop a resurgent dome with a crestal depression, domed layers within and peripheral and internal volcanic activity (Valles and Long Valley type).

The Tyrrhenian margin of Central Italy underwent extension during Pliocene and Quaternary. Extension occurred mainly through NW–SE normal faults, bordering a sequence of Plio-Quaternary basins. These basins are offset by coeval NE–SW faults, which show strike–slip and normal motions and have been interpreted as transfer faults. Plio-Quaternary volcanic activity along the margin occurred along a NW–SE belt, systematically in correspondence with NE–SW transverse systems. The Campi Flegrei Volcanic District (CFVD), on the Southern Tyrrhenian margin, consists of an active NE–SW volcanic ridge developed along NE–SW fractures. We performed a structural field analysis with analogue and mechanical models to investigate the role of transverse structures upon volcanism at Campi Flegrei. Field analysis at Campi Flegrei recognized NE–SW and, to a lesser extent, coeval NW–SE active fractures. Analogue experiments have simulated the development of transfer fault systems in brittle extensional domains. The experiments show that subvertical transfer faults connect offset adjacent normal faults dipping 60°. The mechanical model is based on the stress equations in uniaxial lithostatic conditions and absence of regional stresses. It shows how pre-existing subvertical fractures require the smallest magmatic pressures to be penetrated. For a given magmatic pressure, subvertical fractures might be penetrated more deeply, tapping more easily primitive magmas. These results suggest that the CFVD is located along a NE–SW transfer zone connecting NW–SE regional normal faults. Volcanic activity along such NE–SW trend would be induced by the subvertical dip of the transfer faults. The subvertical dip of transfer faults also suggests an explanation for the emission of the more primitive products along NE–SW systems at Campi Flegrei. These considerations find a wider application on the remaining volcanic districts of the margin, located in the same overall structural setting.