Dipartimento di Scienze della Terra, Università di Cagliari, Cagliari, Italia

Disasters from explosive volcanic eruptions are infrequent and experience in emergency planning and mitigation for such events remains limited. The need for urgently developing more robust methods for risk assessment and decision making in volcanic crises has become increasingly apparent as world populations continue to expand in areas of active explosive volcanism. Nowhere is this more challenging than at Vesuvius, Italy, with hundreds of thousands of people living on the flanks of one of the most dangerous volcanoes in the world. We describe how a new paradigm, evidence-based volcanology, has been applied in EXPLORIS to contribute to crisis planning and management for when the volcano enters its next state of unrest, as well as in long-term land-use planning. The analytical approach we adopted enumerates and quantifies all the processes and effects of the eruptive hazards of the volcano known to influence risk, a scientific challenge that combines field data on the vulnerability of the built environment and humans in past volcanic disasters with theoretical research on the state of the volcano, and including evidence from the field on previous eruptions as well as numerical simulation modelling of eruptive processes. Formal probabilistic reasoning under uncertainty and a decision analysis approach have provided the basis for the development of an event tree for a future range of eruption types with probability paths and hypothetical casualty outcomes for risk assessment. The most likely future eruption scenarios for emergency planning were derived from the event tree and elaborated upon from the geological and historical record. Modelling the impacts in these scenarios and quantifying the consequences for the circumvesuvian area provide realistic assessments for disaster planning and for showing the potential risk–benefit of mitigation measures, the main one being timely evacuation, but include for consideration protecting buildings against dilute, low dynamic pressure surges, and temporary roof supports in the most vulnerable buildings, as well as hardening infrastructure and lifelines. This innovative work suggests that risk-based methods could have an important role in crisis management at cities on volcanoes and small volcanic islands.

Correlation of distal ash deposits with their proximal counterparts mainly relies on chemical and mineralogical characterization of bulk rock and matrix glasses. However, the study of juvenile fragments often reveals the heterogeneity in terms of clast shape, external surface, groundmass texture and composition. This is particularly evident in small scale eruptions, characterized by a strong variability in texture and relative abundance of juvenile fragments. This heterogeneity introduces an inherent uncertainty, that makes the compositional data alone inadequate to unequivocally characterize the tephra bed. Pyroclast characteristics, if described and quantified, can represent an additional clue for the correct identification of the tephra. The paper presents morphological, textural and compositional data on the products of an ash eruption from Middle Age activity of Vesuvius, to demonstrate the information that can be extracted from the proposed type of analysis. Juvenile fragments from five ash layers throughout the studied products were randomly hand-picked and fully characterized in terms of external morphology, particle outline parameterization, groundmass texture and glass composition. Statistical analysis of shape parameters characterized groups of fragments that can be compared with the other textural and physical parameters. The main result is that the data do not show important cross-correlation so suggesting that all of these parameters, together with accurate field data are needed for the complete fingerprinting of a tephra bed. We suggest that this approach is especially important for characterizing the products of small scale, compositionally undistinguishable, eruptions and represents the necessary step to deal with before going into more detailed compositional analyses.

The concentration of thirty nine geochemically relevant trace elements, from 7Li to 238U, was determined in standard silicate glasses (NIST610, NIST612, BCR-2) using the Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS) instrumentation at the Department of Earth Sciences of the University of Cagliari (Italy). The device is a Quadrupole ICP-MS (Perkin Elmer Elan DRC-e) coupled with a 213 nm Nd:YAG laser probe (New Wave Research). This configuration allows rapid, high quality, in-situ trace elements analysis in glasses and minerals. The calibration strategy, achieved using synthetic multi-element glasses (NIST612), with 44Ca as internal standard, gives an analytical accuracy within 5% error level, providing a precision between 1% and 9%, at 40 µm of crater size for all elements. At a laser spot size of 40 m, the lower limit of detection (LLD) ranges between 0.001 and 1 ppm for all the elements; it increases by about one order of magnitude, without any significant fractionation among the different elements, for a laser spot size of 15 µm. Quality control of LA-ICP-MS analyses is routinely performed analysing a natural standard glass, the BCR-2, certified by the USGS, considered as unknown sample. Results indicate that the instrumentation capabilities are suitable for the geochemical characterisation of various materials of mineralogical, petrological, geological and environmental interest.