Developing a spatial vent opening probability map of Somma-Vesuvius caldera

Abstract

The Somma-Vesuvius (SV) volcanic complex has shown in his history a moderate variability of eruptive styles associated with a significant spatial variability of the associated vent locations. This is proved by the presence of numerous eccentric vents which fed effusive eruptions and also by the variability of the vent area of the large explosive eruptions that showed a shift within the present multistage SV caldera. Numerical simulations of explosive eruptions with variable vent locations inside the caldera have shown that this variability, although restricted to an area a few square kilometers, can have a major effect on the associated hazard, particularly for the threat associated to the occurrence of pyroclastic density currents produced by column collapse. This work summarizes some of the activities that have been put forward with the aim of producing a first long-term vent opening probability map for the area of the Somma-Vesuvius caldera. These activities mainly consist in the recognition and collection of key volcano features that can be linked to the spatial distribution of volcanic activity as well as used for their probabilistic treatment. Key variables that have been considered so far include: a) location of Plinian and sub-Plinian volcanic vents; b) location of moderately explosive eruptions (Violent Strombolian to ash emission); c) location of parasitic vents and eruptive fissures; d) distribution of subsurface faults from DTM analysis; e) distribution of the main deep lineaments, as deduced from published geophysical inversions. Locations of Plinian and sub-Plinian volcanic vents have been represented considering their uncertainties based on the available reconstruction of deposits and expert judgment. Parasitic cone and eruptive fissure locations have been also compiled after a comparative analysis of different bibliographic sources, including geological, geomorphological and topographic maps. Distribution of faults and fractures have been finally derived by integrating data from literature studies and new analysis of different digital terrain models (DTM). All the data have been imported into a GIS-based workspace that allowed to organize, analyze and elaborate different datasets. By assuming that each dataset can contribute to the probability distribution of vent opening through the assignment of appropriate weights (e.g. based on expert elicitation), preliminary vent opening (susceptibility) maps will be produced. Results will be used in the production of more accurate hazard maps of the range of expected explosive phenomena in case of a future reactivation of Somma-Vesuvius.