Uncertainty Assessment of Pyroclastic Density Currents Uncertainty Assessment of Pyroclastic Density Currents Cone Model

Abstract

Pyroclastic density currents (PDCs) are extremely dangerous phenomena so their modeling is essential for haz­ ard and risk purposes. However, PDCs are governed by very complex processes, making their deterministic prediction impossible. Probabilistic approaches are in a pioneering phase and feature large (and still unknown) uncertainties, from the natural variability of PDCs (aleatory uncertainty) to the main sources of epistemic uncertainty (input, parametric, theoretical and structural). In this chapter, we quantify these uncertainties by using the Energy Cone Model (ECM) in a Monte Carlo scheme applied to Mount Vesuvius. According to our results, theoretical uncertainty has the largest impact, 5 to 100 times bigger than input uncertainty, which seems to play a minor role. We find that conditional probabilities of PDC arrival (given an eruption of a specific size) show spatial distributions related to the surrounding topography. In particular, for medium and large eruptions, the conditional probability of PDCs traveling beyond Mount Somma is 1%–15% and 50%–60%, while they reach the Napoli airport in about 0%–1% and 0%–15% of the simulations, respectively. Small‐eruption PDCs remain restricted to the south flank and summit area. These results may guide future research devoted to reduce epistemic uncertainties and improve volcanic hazard analyses associated with PDCs.