Enhancing the uncertainty quantification of pyroclastic density current dynamics in the Campi Flegrei caldera.

Pisa

In this study we present a new effort to improve the uncertainty quantification (UQ) of pyroclastic density current dynamics in the Campi Flegrei caldera, thanks to the implementation of a new 2D depth-averaged granular flow model in the Monte Carlo simulation of keycontrolling variables. Campi Flegrei caldera is an active and densely populated volcanic area in the urban neighborhood of Napoli, characterized by the presence of many dispersed cones and craters, and by a caldera wall more than one hundred meters high, towards East. Basic mapping of pyroclastic density currents (PDC) hazard at Campi Flegrei has been already reported in previous studies: some related to field reconstruction and numerical modeling of specific past eruptions or individual scenarios, while others endeavored to produce specific or integrated PDC hazard maps in which the variability of important parameters of the volcanic system was explicitly accounted for. In particular, [4, 2] obtained quantitative estimates of probabilistic PDC hazard, based on the implementation of a simplified kinematic invasion model able to represent main topographic effects. This model, called box model, was extensively run thousands of times in the Monte Carlo simulation varying vent location, eruptive scale, and time frequency of the future activity. In this study we build our effort upon the previous research started in [7, 5], and utilize the physical modeling approach of [6], with the effcient numerical solution of depth-averaged equations for the flow mass and momentum, considering the effects of basal and internal, velocity dependent, friction forces. The model describes the gas-particle mixture as a homogeneous flow, assuming a mechanism of particle deposition consistent with that previously implemented in the box model. UQ is performed by assuming three different components in the input space: (i) rheology parameters, (ii)volume scale, (iii) source location. Our statistical analysis focuses on the first two components, considering a relatively small number of source locations or an uncertain source location inside a subregion of the caldera. This is a first step before the exploration of the full spatial variability of the source location. The statistical inversion of box model equations, varying the vent location (x; y) and the value of inundated area A, can provide us with initial probability estimates for the volume scale of the PDC flow, either in terms of runout distance or volume extent of the multiphase mixture. Our depth averaged model relies on these estimates for setting up the volume scale of past flows. The calibration of rheology parameters is performed according to that. Thus, the rheology and volume components of the input space are conjointly explored by means of Latin Hypercube sampling, attempting a hierarchical conditioning on feasible inputs and plausible outputs [3].

[1] A. Bevilacqua, F. Flandoli, A. Neri, R. Isaia, S. Vitale, “Temporal models for the episodic volcanism of Campi Flegrei caldera (Italy) with uncertainty quantification," Journal of Geophysical Research: Solid Earth 121, 11 (2016).
[2] A. Bevilacqua, A. Neri, M. Bisson, T. Esposti Ongaro, F. Flandoli, R. Isaia, M. Rosi, S. Vitale, “The e ects of vent location, event scale, and time forecasts on pyroclastic density current hazard maps at Campi Flegrei caldera (Italy)," Frontiers in Earth Science 5, 72 (2017).
[3] A. Bevilacqua, A.K. Patra, M.I. Bursik, E.B. Pitman, J.L. Macías, R. Saucedo, D. Hyman, “Probabilistic forecasting of plausible debris flows from Nevado de Colima (Mexico) using data from the Atenquique debris flow, 1955," Natural Hazards Earth System Science 19, 791-820 (2019).
[4] A. Neri, A. Bevilacqua, T. Esposti Ongaro, R. Isaia, W.P. Aspinall, M. Bisson, F. Flandoli et al., “Quantifying volcanic hazard at Campi Flegrei caldera (Italy) with uncertainty assessment: 2. Pyroclastic density current invasion maps," Journal of Geophysical Research: Solid Earth 120, 2330-2349 (2015).
[5] T. Esposti Ongaro, S. Orsucci and F. Cornolti, “A fast, calibrated model for pyroclastic density currents kinematics and hazard," Journal of Volcanology and Geothermal Research 327, 257 - 272 (2016).
[6] M. de’ Michieli Vitturi, T. Esposti Ongaro, G. Lari, and A. Aravena, “IMEXSfloW2D 1.0: a depth-averaged numerical flow model for pyroclastic avalanches," Geoscientific Model Development, 12, 581-595 (2019).
[7] M. Todesco, A. Neri, T. Esposti Ongaro, P. Papale, and M. Rosi, “Pyroclastic flow dynamics and hazard in a caldera setting: Application to Phlegrean Fields (Italy)," Geochemistry Geophysics Geosystems, 7, Q11003 (2006).