Ancient and modern subaqueous calderas form in deep- to shallow-marine oceanic settings and are primary sites for volcanogenic massive sulphide (VMS) deposits. Caldera structures hosting VMS deposits are concentrated in the Archean Abitibi greenstone belt and in the Wabigoon subprovince. The Hunter Mine and Normetal calderas, two little-known effusive-dominated edifices, are emphasised and Sturgeon Lake is the best known explosive-dominated caldera. Although VMS deposits are linked to calderas, their 5-30 km diameter size, poses problems in pin-pointing highly favourable sites for VMS exploration. The overall caldera geometry is readily recognised based on regional-scale mapping but detailed volcanic facies mapping is required to distinguish caldera subsettings in the Archean. Caldera subenvironments include: (1) the caldera wall featuring (a) chaotic breccias, (b) dyke intrusions, (c) synvolcanic faults and (d) pyroclastic debris; (2) the intracaldera moat or floor with (a) horst and graben structures (synvolcanic faults), (b) superposed dome-flow-hyaloclastite complexes, (c) extensive central dyke swarm and (d) small explosive volcanic edifices formed by magmatic fountaining eruptions and (3) caldera apron deposits showing (a) extensive volcaniclastic debris of pyroclastic and autoclastic origin and (b) local dome-flow-hyaloclastite complexes. All loci are possible sites for VMS formation, but the caldera wall favours large VMS deposits.Based on the studied calderas, a new hydrothermal alteration model is proposed that helps explain chert-Fe carbonate assemblages in Archean volcanic sequences. The early hydrothermal silica seals the volcaniclastic-dominated rocks (cap rock) at or near the edifice-seawater interface. These silicified volcaniclastic turbidite deposits have generally been referred to as chert or exhalites but this interpretation necessitates reconsideration. Overprinting the silicification phase is a pervasive semi-conformable carbonate hydrothermal alteration zone with a discordant focussed root zone along faults. Three distinct carbonate pairings are observed: (1) proximal siderite (side-roplesite) - Fe-ankerite next to the VMS-deposit, (2) an intermediate ankerite-Fe-dolomite zone and (3) a distal calcite-dolomite zone. Transitions between zones are subtle and changes are indicated mineral assemblage overlaps. Our results shed new light on hydrothermal alteration patterns, but also resolve some of the problems associated with chert-iron carbonate formations. The chert and hydrothermal carbonates as well as the VMS deposits are of the replacement type, rather than chemical precipitants and black smoker deposits.
Collapse calderas are common on Earth and some other solid planetary bodies, particularly on Io (a satellite of Jupiter), Mars and Venus. Caldera structures are generally similar on all these bodies but the sizes vary considerably. Here I present numerical models of caldera-fault formation in volcanoes with shallow, spherical or sill-like magma chambers. In all the anisotropic models, the crustal segment (including the volcano) above the shallow chamber is composed of 30 comparatively thin layers with stiffnesses (Young's moduli) alternating between 1 and 100 GPa. The chamber itself is located in a single, thick layer with stiffness different from that of the other layers. The crustal segment hosting the chamber is either 20 or 40 km wide but has a constant thickness of 20 km. The loading conditions considered are crustal segments subject to: (a) underpressure (lack of magmatic support) of 5 MPa; (b) tensile stress of 5 MPa; (c) excess magmatic pressure of 10 MPa at the bottom of the crustal segment (doming of the volcanic field containing the chamber) and (d) combination of tension and doming. In all the models, the magma-chamber top is at 3 km depth; the diameter of the sill-like chamber is 8 km (its thickness is 2 km) while that of the spherical chamber is 4 km.The main results are as follows: (1) Underpressure and excess pressure in a shallow, crustal chamber normally result in dyke injection rather than caldera formation. (2) For doming or tension, a spherical magma chamber normally favours dyke injection rather than ring-fault initiation. However, when the spherical chamber is located in a very soft (10 GPa) layer, the local stress field may be suitable for caldera-fault formation. (3) For a sill-like chamber in a 20-km-wide volcanic field, a ring fault may be initiated either during horizontal tension or a combination of tension and doming. (4) For a sill-like chamber in a 40-km-wide volcanic field, doming alone is sufficient to initiate a caldera fault. The results indicate that the local stresses in composite volcanoes most likely to initiate caldera faults are associated with sill-like chambers subject to tension, doming or both.
Most caldera volcanoes are associated with circular dike intrusions. Ring-dikes form during complete or partial subsidence of the caldera floor and may be responsible for eruption locations that surround a structural basin. Through a systematic set of numerical models, this paper summarizes a variety of types, mechanisms, and patterns of caldera ring-dikes that can be observed in nature. Caldera subsidence is simulated by magma chamber depressurization; three main sets of models are distinguished. First, local linear and circular faults are included in order to understand their effect on caldera-related displacements. Second, passive opening at a ring-fault is studied in order to understand where ring-dike intrusions may occur. Third, models are designed to exemplify how processes external to the caldera, such as a tectonic earthquake or an eccentric intrusion, may affect the location of a ring-dike intrusion. These models suggest that ring-dikes commonly form “incompletely,” i.e. only part of a ring can be intruded because of the nonuniform displacement field around the ring-fault. As described in the discussion, these models help explain the locations of ring-dikes in various volcanic regions.
The Sierra Madre Occidental (SMO) is the largest continuous ignimbrite province in the world. It covers the NW portion of Mexico and has a minimum estimated volcanic rock volume of about 400,000 km(3). The southern part of the North American Basin and Range extensional province is in Mexico and was formed by NW- to NE-trending normal faults that bound many large grabens, which are particularly long and deep in the southern SMO. Graben formation and ignimbrite-forming pyroclastic flow eruptions coincided in space and time, particularly for the 38-23 Ma period, which has been referred to as the Ignimbrite Flare-up event. Geologic observations in the southern SMO indicate that the vents of several large-volume silicic ignimbrites were related to the graben master faults. We propose the new name "graben caldera" for this type of explosive-volcano-tectonic-collapse structure. The evidences for vent location of graben calderas include large pyroclastic dikes, co-ignimbrite lithic-lag breccias, and post-ignimbrite aligned rhyolitic domes and rhyolitic lava dikes. All these features are found along the graben caldera walls or on the graben's margins and thus were controlled by the main graben faults. Large-volume ignimbrite-forming eruptions occurred during graben collapse, usually through several vents along the graben caldera walls. Often, only part of the graben acted as a collapse caldera. In most cases, the downdropped blocks inside the collapsed segment of the graben have several distinct tilting directions, opposite to those caused by regular domino faulting (with tilting outwards from the graben's axis), indicating a chaotic collapse of blocks. This is interpreted as collapse of blocks during magma evacuation from an elongated and possibly batholith-sized magma chamber controlled by the regional extensional tectonics that existed under the collapsed segment of the graben. This collapse produced a piece-meal-like caldera confined within a graben. Generally, the pyroclastic dikes occur as discontinuous elongate lenses rather than regular tabular bodies. Usually, an elongated lava dome filled the vent and this totally covers the pre-existent pyroclastic dike, but in some cases the pyroclastic dike is exposed. Faulting and subsidence continued for several millions of years after the collapse and ignimbrite emplacement, displacing the intra-graben caldera products downward into the tectonic depression, but preserving or even intensifying the chaotic arrangement of the collapsed blocks. Fluvio-lacustrine continental deposits are generally found beneath and over the main and most voluminous ignimbrite, indicating that subsidence started before ignimbrite emplacement and graben collapse, and that continued after these events. In many cases the graben caldera vents are related to gold and silver hydrothermal mineralisation; thus understanding the relationship between ignimbrites, graben caldera vents, and fluvio-lacustrine deposits will be important for economic purposes.
This paper focuses on the role that hydrothermal systems may play in caldera unrest. Changes in the fluid chemistry, temperature and discharge rate of hydrothermal systems are commonly detected at the surface during volcanic unrest, as hydrothermal fluids adjust to changing subsurface conditions. Geochemical monitoring is carried out to observe the evolving system conditions. Circulating fluids can also generate signals that affect geophysical parameters monitored at the surface. Effective hazard evaluation requires a proper understanding of unrest phenomena and correct interpretation of their causes. Physical modelling of fluid circulation allows quantification of the evolution of a hydrothermal system, and hence evaluation of the potential role of hydrothermal fluids during caldera unrest. Modelling results can be compared with monitoring data, and then contribute to the interpretation of the recent caldera evolution. This paper: (1) describes the main features of hydrothermal systems; (2) briefly reviews numerical modelling of heat and fluid flow through porous media; (3) highlights the effects of hydrothermal fluids on unrest processes and (4) describes some model applications to the Phlegrean Fields caldera. Simultaneous modelling of different independent parameters has proved to be a powerful tool for understanding caldera unrest. The results highlight the importance of comprehensive conceptual models that incorporate all the available geochemical and geophysical information, and they also stress the need for high-quality, multi-parameter monitoring and modelling of volcanic activity.
In the problem of forecasting a volcanic eruptive scenario, empirical or observational time series must be examined in order to look for precursory behaviour. A key concept in this respect is that of memory or persistence of a given time series. A time series that does not keep some memory of its past cannot, in fact, provide information about the future of the evolving volcanic process, i.e. it cannot help to forecast an eruption. The geostatistical approach aims to identify this memory, if it exists, quantify its duration and exploit its potential in forecasting, and can be applied not only in the time but also in the space domain. In this paper, a review is presented of the state of the art, and an application to the case of a possible reawakening of a caldera is presented.
A complete study of collapse caldera formation should ideally involve multiple aspects such as regional tectonics, system geometry, magma and host rock properties, fluid-structure interaction, pre-existing structural discontinuities, and deformation history. Due to the complexity of such a comprehensive analysis, studies so far have centred on relevant but atomised topics. From a methodological point of view, and in addition to essential field and petrological studies, collapse calderas have also been investigated through analogue and theoretical models and geophysical imaging. Each approach presents advantages and disadvantages. We review the most significant contributions, summarise the relevant outcomes, and highlight the strong points and weaknesses of each approach. Analogue models enable a qualitative study of the structural evolution of a collapse process and allow us to infer which geometric factors play a relevant role. Differences among employed models lie in the applied experimental devices, the host rock analogue material (dry quartz sand, flour, etc.), and the magma chamber analogue (water or air-filled balloons, silicone reservoirs, etc.). However, the results obtained from different experimental setups are not substantially different if basic input parameters are kept similar in the experiments. Discrepancies in results mainly stem from restrictions of experimental designs. Theoretical (mathematical) models have grown in importance during the past decades, in combination with the development of computational resources. Nowadays, these models constitute a significant source of information on caldera-forming processes and can predict semi-quantitatively general conditions for fault formation and propagation. Theoretical studies can be classified in two groups according to their objectives. One group focuses on the evolution of pressure within the magmatic reservoir during a caldera-forming eruption. The second looks more into the structural conditions for caldera collapse and hence relate to analogue models. Both analogue and theoretical models are employed to gain a fundamental understanding of caldera processes and their resulting structures. Additionally, geophysical imaging helps to construct a regional image of the subsurface at active calderas, thus imposing constraints on the structural investigations based on analogue and mathematical modelling. A revision of each of these three complementary approaches to the study of collapse calderas is given in this paper, together with a combined analysis of their main findings and restrictions.
This paper reviews the times that silicic magmas related to major caldera systems spend in the crust prior to eruption. The significance of the time information is evaluated and combined with magma volumes and temperatures to quantify the mass and thermal fluxes associated to calderas. The data discussed includes the largest explosive eruptions on Earth: Taupo Volcanic Zone (New Zealand), the Youngest Toba Tuff (Indonesia), Yellowstone system (USA), Long Valley (USA), Carter Lake (USA), Valles-Toledo complex (USA), La Garita caldera (USA), La Pacana (Chile) and Kos (Greece). Magma residence times are calculated from the difference between the eruption age and the age obtained by radioactive clocks and minerals that are a closed system at high magmatic temperatures (e. g., U-Pb system in zircon).Large ranges of residence times between different systems are found. The shortest residences (4-19 ky) are those of some magmas from the Taupo Volcanic Zone (Oruanui and Rotoiti) and Yellowstone (Dry Creek and Lava Creek). There is not a good correlation between magma volume and residence time, although most eruptions <10 km(3) have residence times <100 ky, and those >100 km(3) have longer residences, some up to 300-500 ky (Fish Canyon, La Pacana). The residence times of some small (<10 km(3)) pre-and post-caldera magmas indicate that they were extracted from the same reservoir as the caldera-forming magma (e. g., Long Valley, Taupo). However, the time information from most small-volume magmas seems to reflect the recycling of crystals from previous cycles of caldera-forming magmas (Yellowstone), from plutonic rocks of the same caldera cycle with or without erupted equivalents on the surface (Crater Lake, Taupo, Long Valley), or from a partially solidified magma reservoir (Taupo). These interpretations are in agreement with cooling rates and solidification times obtained from simple thermal models of magma reservoirs.Magma production rates were calculated from the ratio of erupted volume and residence time, and they vary between <0.001 km(3) y(-1) for small deposits (<10 km(3)) and ca. 0.1 km(3) y(-1) for the Oruanui eruption (530 km(3)). Estimates for most eruptions >500 km(3) are within 2 +/- 2 x 10(-2) km(3) y(-1). These high magma production rates are probably transient and comparable to global eruptive fluxes of basalts (e. g., Hawaii). Magma cooling rates for deposits >100 km(3) were calculated from the difference between the liquidus and pre-eruptive temperatures over their residence times, and they vary between 2 x 10(-4) and 3 x 10(-3) Ky(-1). Integration of the calculated residence times and magma fluxes with a simple rheological model of the crust is not possible and should be a main topic of research if we are to understand the mechanisms and rates which permit large amounts of silicic magma to be stored below calderas.
This work is an exploratory effort to address some key concepts about physical processes that influence the mobility, sedimentology and depositional mechanisms of pyroclastic density currents. The main mechanisms of particle support are critically reviewed and their influence on depositional mechanisms is discussed. The depositional behaviour of pyroclastic density currents is discussed in the light of new models that consider the different typologies within a continuum spectrum that spans from very dilute (fluid-dominated) to very concentrated (solid-dominated) flows. The combination of progressive aggradation and en masse freezing models is proposed on the basis of a model that takes into account the stepwise aggradation of discrete pulses that develop within a single current and that stop en masse when resisting bulk forces exceed driving ones. The influence of various types of morphological settings and obstacles on depositional mechanisms and mobility of pyroclastic density currents is examined and discussed in detail. Some examples of stepwise aggradation within pulsating PDCs are discussed using analyses of lithofacies and lithofacies associations of deposits emplaced under different flow-boundary conditions.
Unrest at explosive collapse calderas is the manifestation of complex subsurface processes. Geophysical signals recorded during unrest can be caused by the migration and emplacement of magma, or by tectonic or hydrothermal activity. Geodetic techniques represent a crucial part of a monitoring programme as they provide means to quantify volume changes in the feeder system of restless calderas. However, deformation data alone cannot discriminate between magma and aqueous fluid intrusions. Time-lapse gravity measurements can constrain the mass of the intrusion, and consequently the combination of geodesy and gravity measurements can be used to infer the density of the intrusive fluids and can better constrain the deformation source. Here, we consider the application of gravimetric and geodetic techniques to study caldera unrest with examples from Long Valley, Campi Flegrei, Las Canadas and Nisyros. We identify problems with current time-lapse gravimetric techniques, discuss several approaches to model the source of unrest from deformation and gravity data and provide an outlook into future challenges for integrated geodetic studies.
Campi Flegrei caldera, which includes the highly urbanized city of Naples, is the most dangerous volcanic area in the world as it represents a serious threat to around two million people. During the last four decades it has experienced a huge uplift phase, which reached about 3.5m in 1985 when a subsidence phase started. Recent geodetic data demonstrates that such a subsidence phase has terminated, and a new uplift episode started in November 2004 has so far resulted in about 0.04m of uplift (October 2006). Here we show that the present episode, which appears slower but longer than previous small uplifts, sheds light both on the origin of small and large unrests, and on the conditions to evolve toward large unrests. Our simple but robust method is based on continuous GPS measurements and compares the ratio between maximum horizontal to vertical displacements. Results show that such unrests are due to overpressure in a deeper source of fluids of magmatic origin. When the resulting stress increase causes fracturing of the overlying rocks, magmatic fluids are injected in shallower aquifers and generate large uplifts. This means that, given the common magmatic origin, both small and large unrests contribute to the increase in pressure of the magma chamber which, given sufficient magnitude, can cause fracturing of the overburden and produce an eruption. In this paper, we show that the maximum horizontal to vertical displacement ratio can be a powerful indicator of source changes, and can give important information for volcanic eruption forecast.
A number of analogue models studying caldera architecture and development have been recently performed under different conditions (apparatus, materials, scaling parameters, stress conditions). An overview of the experiments reveals a consistent scenario for caldera structure and development, regardless of imposed boundary condition. In fact, a complete collapse can be summarised through four main stages, proportional to the amount of subsidence, progressively characterised by a: (1) downsag; (2) reverse ring fault; (3) peripheral downsag and (4) peripheral normal ring fault. A brief comparison to natural cases shows that all these experimental structures, as well as their development, are commonly observed, even at various scales. Such a consistency between models and nature suggests a general applicability of experimental results. The four evolutionary stages adequately explain the architecture and development of the established caldera end-members (downsag, piston, funnel, piecemeal, trapdoor) along a continuum, where one or more end-members may correspond to a specific stage. While such a continuum is controlled by progressive subsidence, specific collapse geometries result from secondary contributory factors (roof aspect ratio, collapse symmetry, pre-existing faults). The proposed evolutionary scheme incorporates not only the geometric features of calderas, but more importantly, also their genetic features.
This paper sets out to assess the failure forecast method (FFM) and its application to volcanology. After first reviewing the history of its development and the published literature, a special focus will be given to the possible use of FFM in the analysis of reawakening volcanoes and caldera unrest. We present results from its application to the recent episode of unrest at the Las Canadas caldera, Tenerife (Spain). Here the FFM procedure was automated, solving the problem of subjectivity in the application of FFM.
Lithic-rich pyroclastic units and facies are often associated with caldera-forming eruptions. Petrographic and quantitative studies on the variety of lithic types, and the spatial and vertical variations in their proportions, provide a powerful tool for understanding (a) the subsurface and pre-caldera geology, and (b) conduit-vent processes during caldera eruptions. In particular, lithic assemblages may include unique samples of deep plutonic-basement features, hydrothermal systems and ancient volcanic landforms destroyed by caldera fragmentation. When interpreting caldera eruptions, studies of lithic clasts can constrain vent configurations, and the depth and style of conduit wall rock fragmentation.The 186 ka Abrigo ignimbrite, representing the last major caldera-forming eruption of the Las Canadas volcanic edifice, Tenerife, contains a diverse lithic population including (a) syenite, and rare syenogabbroid and gabbroid fragments from a deep plutonic-contact metamorphic core, (b) abundant altered fragments, representing a relatively deep extensive zone of hydrothermal alteration, and (c) shallow-and surface-derived mafic to felsic, crystalline and glassy volcanic clasts, and welded to non-welded pyroclastic and epiclastic breccia clasts, all of which are consistent with being derived from a pre-Abrigo constructive phase of the Las Canadas edifice. Significant lateral variations in the proportions of lithic clast types, within depositional units, is consistent with an eruption involving multiple vents around a caldera that underwent piecemeal collapse, and this is further supported by lateral geochemical variations in juvenile clast populations. Vertical variations in lithic clast proportions between depositional units suggest an increasing depth of conduit wall rock fragmentation during the eruption. This study highlights vertical caldera collapse as a major process in the evolution of the Las Canadas caldera complex.
This chapter presents an introduction to the book, Vesuvius Education, Security, and Prosperity, in Italian language and provides a global perspective of the territory as seen by a group of intermediate students of the Vesuvius area.
The Vesuvius area communities at high volcanic risk comprised in 2001 about 550 000 people. One part of this population left the territory during the past decade. This contributed to a decrease of about 30 000 people and shows an inversion of the population growth of previous decades. The emigrating population leaves along the seacoast which is comprised of poorly governable communities with high human loads and being subjected to economic impoverishment and risk from the volcano. This migration flux is associated principally with the people who do not work in the Vesuvius area and can thus be relocated into less populated areas at the north of Naples that are connected with this city with various transportation systems. The population is also subject to a lowering of the fertility rate, but this is decreasing at a slower rate than that of the central and northern parts of the country.The educational level of Vesuvians is low, with an increase in the illiteracy level and decrease in secondary school and college graduates as the distance from Naples increases. The economic reality of the area is precarious and officially only one-fifth of the population works. In spite of the invading urbanization and strong parceling of land, the agriculture has always produced good results, due both to the high fertility of the volcanic soil and favorable climatologic and morphologic conditions of the area. A considerable number of people work in the public sector (schools, sanitation, public administration, transports) and thus cause damage to the already little-efficient public service sector.
This chapter summarizes a comparative study of shear-wave velocity models and seismic sources in the Campanian volcanic areas of Vesuvius and Phlegraean Fields. These velocity models were obtained through the nonlinear inversion of surface-wave tomography data, using as a priori constraints the relevant information available in the literature. Local group velocity data were obtained by means of the frequency-time analysis for the time period between 0.3 and 2 s and were combined with the group velocity data for the time period between 10 and 35 s from the regional events located in the Italian peninsula and bordering areas and two station phase velocity data corresponding to the time period between 25 and 100 s. In order to invert Rayleigh wave dispersion curves, we applied the nonlinear inversion method called hedgehog and retrieved average models for the first 30-35 km of the lithosphere, with the lower part of the upper mantle being kept fixed on the basis of existing regional models.A feature that is common to the two volcanic areas is a low shear velocity layer which is centered at the depth of about 10 km, while on the outside of the cone and along a path in the northeastern part of the Vesuvius area this layer is absent. This low velocity can be associated with the presence of partial melting and, therefore, may represent a quite diffused crustal magma reservoir which is fed by a deeper one that is regional in character and located in the uppermost mantle. The study of seismic source in terms of the moment tensor is suitable for an investigation of physical processes within a volcano; indeed, its components, double couple, compensated linear vector dipole, and volumetric, can be related to the movements of magma and fluids within the volcanic system. Although for many recent earthquake events the percentage of double couple component is high, our results also show the presence of significant non-double couple components in both volcanic areas.
The distribution of pyroclasts from 79 A. D. eruption of Vesuvius is analysed to assess the ejection velocities of ballistic particles pertaining to the white and gray eruption phases. This distribution is related to the energy of the eruptive mixture and conditions of the atmosphere during the eruption. Ballistic debris is common in the deposits, and within the sampled area (3-14 km S-SE from the vent) the ejected blocks are scattered throughout the fine-grained pumice fall. We measured about 300 ballistic blocks with diameters between 0.07 and 1 m. Some fragments as large as 0.3 m are located at 9 km from the vent, which probably represents the ballistic limit of such fragments. By using a ballistic model for large blocks permitted an assessment of their initial velocities which range from 170 to 2300 m/s, and since some of these velocities exceed the maximum observed velocities of plinian eruptions we conclude that the ballistic model is deficient.The trajectories of smaller blocks (0.1<d<0.3 m) are not truly ballistic, because these can be sustained in the eruptive column and dispersed by means of the finger-like projections from the jet thrust region of the column from where they fall or produce gravity currents on the slopes of the volcano. The gas expansion in the column reduces the drag force on particulates and aids in their vertical and lateral transport. In modeling an explosive scenario at Vesuvius it is thus necessary to account for a wide variety of particulate sizes in the presence of local and stratospheric wind conditions and changing characteristics of magma as it is being evacuated from the volcanic system.
