Activity, Unrest, and Hazard Evaluation at Stratovolcanoes and Calderas II
Presiding: G De Natale, INGV-Osservatorio Vesuviano; C Troise, INGV-Osservatorio Vesuviano
V23B-01 INVITED 13:30h
Structure and formation of ring faults in composite volcanoes
Most collapse calderas form by vertical displacement (subsidence) along ring faults. One principal problem regarding the formation of collapse calderas is thus to explain the stress conditions that favor the formation of, or slip on existing, ring faults. The ring faults are shear fractures, more specifically dip-slip faults, so that they cannot form unless the local driving shear stress in the hosting composite volcano satisfies the criteria of shear failure. Many ring faults, however, are perhaps more accurately described as mixed-mode fractures: partly shear and partly extension fractures. This follows because the ring faults are commonly occupied by ring dikes which, like dikes in general, are extension fractures. Also, ring-fault development at the surface commonly involves tension-fracture formation. For a ring fault to form, or an existing one to slip, suitable stresses must concentrate at a certain radial distance from the center of the associated magma chamber. Similarly, at the surface, suitable stresses must peak in the circular region that eventually develops the ring fault. Most unrest periods do not generate stress fields suitable for formation of, or slip on, ring faults. By contrast, many unrest periods are favorable to dike injection, some of which reach the surface to feed eruptions. Thus, even in existing calderas, most unrest periods do not generate the special stress fields needed for ring-fault slip. Here we present new numerical models on ring-fault formation in composite volcanoes. Two common geometries of magma chambers are considered: spherical and sill-like (oblate ellipsoidal). In some models, the chamber itself is located in a single thick layer, whereas in others it is hosted by several layers of different mechanical properties. In all the models, the crustal segment above the chamber, up to the surface of the volcano, is composed of 30 adjacent layers with stiffnesses (Young's moduli) ranging from 1 GPa to 100 GPa. The thickness of the crustal segment hosting the chamber is 20 km, whereas its width is either 20 km or 40 km. The loading conditions include (1) crustal segments subject to tensile stress of 5 MPa, (2) crustal segments subject to magmatic pressure of 10 MPa at the bottom (that is, doming of the volcanic field containing the chamber), (3) a combination of tension and doming, and (4) internal excess pressure or underpressure in the chamber. In all models, the top of the chamber is at 3 km depth; the diameter of a sill-like chamber is 8 km (height 2 km), that of a spherical chamber 4 km. Some of the main results can be summarized as follows. (1) Internal excess-or underpressure in a spherical or sill-like chamber is not very likely to give rise to a ring fault; this loading rather results in dike injections. (2) For doming or tension, a spherical magma chamber is normally unlikely to give rise to a ring fault but favors dike injection. However, a spherical chamber in a very soft (10 GPa) layer, or one with recent dike injections, may generate a ring-fault. (3) A sill-like chamber under tension, doming, or tension and doming may generate a ring fault (and often a ring dike). When the volcanic field is 20 km wide, tension or tension and doming may result in ring-fault formation. When the volcanic field is 40 km wide, doming alone can also result in ring-fault formation. In conclusion, the numerical results suggest that composite volcanoes with sill-like magma chambers located within volcanic fields subject to tension, doming, or both are most likely to generate stress fields suitable for ring-fault formation.
V23B-02 14:00h
The Campi Flegrei Caldera (Southern Italy): Modeling, Interpretation and Hazard Estimation
This work reviews the main geophysical observations, modelling interpretations and hazard estimation at Campi Flegrei caldera, both in normal periods and during unrests involving very high uplift and seismicity rates. Campi Flegrei caldera is a volcanic area partially including the city of Naples (Southern Italy). Explosive activity, mainly hydromagmatic, characterises this area, in which huge ground deformations, well known since ancient Roman times, occur. In the last 30 years, two episodes at least of spectacular ground deformation and seismicity have occurred. This work reviews the main issues and problems in the interpretation of Campi Flegrei activity, presenting a coherent model which involves both elastic and thermal fluid-dynamical effects. The basic concepts, which are thought to be distinctive features of activity at all the calderas, involve the close interplay among magmatic stress, bordering caldera faults and shallow geothermal system, with an important contribution of background regional stress. The work presents new insight for interpreting pre-eruptive sequences at calderas, and new methodologies for the estimation of hazard in such a densely populated area.
V23B-03 14:15h
Controls on Recent Unrest at Campi Flegrei Caldera, Southern Italy
Campi Flegrei, in Southern Italy, is an active caldera that has shown signs of unrest since 1969. Because the caldera has a population of 400,000 people, it is especially important to understand the mechanisms driving the unrest and their implication for the probability of a future eruption. Since its last ignimbrite eruption 12,000 years ago (which produced the Neapolitan Yellow Tuff), volcanic activity in Campi Flegrei has consisted of numerous eruptions (volumes ~0.1 km3 or less) surrounding the inferred caldera rim. For at least the last 3,700 years, the caldera has been subsiding at mean rates of 14-17 mm per year, punctuated by two known periods of mean uplift (1430-1538 and 1969-Present). The first period produced a net uplift of about 30 m at the port of Pozzuoli and was followed in 1538 by the eruption of Monte Nuovo (20 million m3) some 4 km to the west. The second period has to date consisted of two episodes of uplift (in 1969-72 and 1982-84), each raising Pozzuoli by about 2 m. Studies of the second period have attributed uplift either to magmatic intrusion or to the expansion of water in heated aquifers. These interpretations assumed a stationary reference condition. It is here proposed that the reference condition in fact corresponds to subsidence at about 17 mm per year. Slower subsidence then reflects the difference between background subsidence and actual intrusion of magma. The revised interpretation suggests a two-component source for the recent episodes of uplift: (1) intrusion of two batches of magma of ~0.1 km3 that have produced a permanent uplift of about 2.8 m, and (2) the expansion and later dissipation of heated water, which produced a temporary uplift of about 0.7 m that has since disappeared. The similar volumes of recent intrusions and post-NYT eruptions further suggest that Campi Flegrei is fed by discrete batches of magma. The caldera today may thus be underlain by a collection of modest magma bodies rather than a single, large reservoir.
V23B-04 14:30h
First Volcanological-Probabilistic Pyroclastic Density Current and Fallout Hazard Map for Campi Flegrei and Somma Vesuvius Volcanoes.
Integrated volcanological-probabilistic approaches has been used in order to simulate pyroclastic density currents and fallout and produce hazard maps for Campi Flegrei and Somma Vesuvius areas. On the basis of the analyses of all types of pyroclastic flows, surges, secondary pyroclastic density currents and fallout events occurred in the volcanological history of the two volcanic areas and the evaluation of probability for each type of events, matrixs of input parameters for a numerical simulation have been performed. The multi-dimensional input matrixs include the main controlling parameters of the pyroclasts transport and deposition dispersion, as well as the set of possible eruptive vents used in the simulation program. Probabilistic hazard maps provide of each points of campanian area, the yearly probability to be interested by a given event with a given intensity and resulting demage. Probability of a few events in one thousand years are typical of most areas around the volcanoes whitin a range of ca 10 km, including Neaples. Results provide constrains for the emergency plans in Neapolitan area.
V23B-05 14:45h
The Global Water Cycle Drives Volcanism on Seasonal to Millennial Timescales
Global rates of occurrence of volcanic eruptions show periodic behaviour on timescales ranging from <1 yr (seasonal) to >106 years. At long timescales (>106 to 107 years), rates of eruption are controlled by plate tectonics. At shorter timescales, the periodic nature of volcanism is forced by the global water cycle. Historical records of the rates of onset of eruption for the past 300 years are dominated by small-scale activity at a number of persistently, or repeatedly, active volcanoes around the world. This record shows statistically significant evidence for `seasonality': globally, rates of eruption are about 18% higher during northern hemisphere winter than northern hemisphere summer. This pattern of seasonality is strong for volcanoes at high northern latitudes; but also exists for volcanic regions in the southern hemisphere (e.g. Chile) and at specific volcanoes (e.g. Sakurajima, Japan). Seasonality is weak at certain ocean-island volcanoes (e.g. Hawaii), and certain volcanic regions (e.g. Mediterranean). The only external parameters that account for the periodic nature of small-scale volcanism (i.e. the observation that eruption rates peak between November and March in both hemispheres) are those related to the global water cycle. Movement of water (including atmospheric vapour; soil moisture; snow and ice) between the northern-hemisphere continents and the world's oceans is responsible for an annual deformation of Earth's surface that is weakly defined in equatorial regions, and stronger at higher latitudes. This external modulation of the Earth's surface has an amplitude of the order of centimetres, and an associated (vertical) strain rate of ~ 10-16 s-1. This deformation is slow enough to be felt by the Earth's interior, and is of the same order of magnitude as the (horizontal) strain rates experienced in tectonically active continental regions. This modulation effectively applies a time-dependence to the `threshold' point at which a volcano will begin to erupt. In this way, the subtle, small-magnitude but long-wavelength changes associated with the annual hydrological cycle leads to clustering of volcanic eruptions. Peaks of eruption onset are associated with periods of changing sea-level or atmospheric pressure, rather than with the maxima or minima. Longer timescale (> 103 year) variability in rates of volcanism during the Pleistocene, associated with large scale climatic changes, is evident in long-term terrestrial and ice-core records of volcanism. Linkage between the global hydrological cycle and volcanism over annual to millennial timescales plays an important role in land-atmosphere coupling by modulating volcanic emissions and, thereby, the volcanic component of climate forcing.