“…Here the initial, large pressure front results from the liberation of pressurized gas at the beginning of the simulation. The same process has been shown to potentially generate supersonic blast waves, with initial Mach number > 6, at the beginning of explosions at several volcanoes [Marchetti et al, 2013;Medici et al, 2014]: the largely supersonic apparent velocity we found for our initial waves (Table 1) suggests that this could be the case also at Stromboli volcano. The different families of sustained waves can be tentatively attributed, based on their geometrical similitude with the modeled ones, to well-established types of jet noise, including (1) vortex ring noise, (2) predominantly forward directed noise from large shear layer structures, like toroidal or helical modes in the jet, (4) turbulent mixing noise, usually propagating sideward from the shear layer, (5) broadband shock-associated noise from supersonic jets, which is predominantly directed backward from a position about 5-7 jet diameters above the vent, and (6) screech noise, a tonal component due to a feedback of the shock-associated noise with the ground [Tam, 1995;Schulze and Sesterhenn, 2008].…”