Silent Very Long Period Seismic Events (VLPs) at Stromboli Volcano, Italy.

At Stromboli Volcano, Italy, very long period (VLP) seismic signals and Strombolian eruptions have been attributed to the unsteady flow of gas slugs through the shallow plumbing system followed by explosive slug bursting at a free surface. In data from a 2018 seismo‐acoustic deployment, ∼92% of events in two main VLP multiplets do not coincide in time with impulsive infrasonic signals (the expected signal of explosive slug bursting); we term these "silent VLPs." The lack of infrasonically detected explosions relative to repeating VLPs does not support the commonly invoked "gas slug" model. We propose that VLPs may be generated when gas bubbles move into a weak semi‐solid plug in the uppermost portion of the conduit. The plug then acts as a mechanical filter in which pathways vary and guide or trap ascending gas slugs, allowing for passive (silent) gas release and explosive escape mechanisms decoupled in time from VLPs. Plain Language Summary: Very long period (VLP) seismic signals are thought to be produced by magma and gas flow through shallow volcano conduits. At Stromboli Volcano, Italy (the type volcano for Strombolian explosions) and other basaltic volcanoes worldwide, VLPs have been modeled as the result of flow of a gas slug through the conduit followed by an explosion when the slug bursts at a free surface. However, these interpretations do not account for variable crystal content (i.e., systems with three phases: solid, liquid, and gas). Here we show that the standard "slug" model does not hold during a period of intense study at Stromboli in 2018. We compare the timing and characteristics between two repeating VLP families and explosions at the surface, as characterized by acoustic data. The vast majority of VLPs do not have coincident explosions—we term these "silent VLPs." Our observations instead support a "plug" model. We argue that VLPs are generated when gas slugs interact with and move into a weak plug in the uppermost portion of the conduit, which acts as a mechanical filter for gas transport. This interpretation allows for both passive (silent) and explosive gas escape decoupled in time from VLPs. Key Points: Seismo‐acoustic observations from Stromboli in 2018 show hundreds of repeating very long period (VLP) seismic signals without corresponding acoustic signalsVLPs without explosions suggest the slug model is a poor fit for "normal" activity at StromboliWe suggest VLPs may instead result from large gas bubbles interacting with a semi‐permeable solid plug [ABSTRACT FROM AUTHOR]