This work combines a large number of ultra violet camera based observations of sulphur dioxide release during and following strombolian explosions (which is called a gas flux trace) and smaller degassing events not necessarily associated with ejectiles (i.e. any magma release). Within the paper these are referred to as Vent 1 and Vent 2 events respectively. The benefit of collecting such a large number of gas flux traces is that any trends present or common characteristics between differing events can be compared. This allowed six flux trace types to be identified based on the patterns in gas release following the initial impulsive event associated with a strombolian explosion or smaller degassing event.
Following on from this, building on the work of Tamburello et al. (2012) who noticed that gas flux traces from strombolian explosions contained a coda (a period of elevated flux before a return to background levels), we were able to approximately separate out the distribution of gas mass into the initial impulsive gas release and the resultant coda. This allowed us to determine that a large proportion of the entire event (i.e. the initial release of gas added to the coda) could be contained within the coda - quite an interesting observation!
Alongside this work we performed a number of computational fluid dynamics simulations (using Ansys Fluent) of gas slug flow (Taylor bubble - see previous post). Over a range of parameters for the magma density, viscosity, and the conduit diameter we discovered that it was possible for the simulated gas slugs to shed gas bubbles, called "daughter bubbles", from the base of the slug at varying rates of efficiency. If these slugs were to rise over extended distances (simulations in the manuscript were over short distances in comparison to a full volcanic conduit) then significant mass could be lost from the slug into a daughter bubble train and contribute to the observed gas coda. There are of course other possibilities for the coda generation, however, daughter bubble could certainly play a role in determining how explosive an event may become and has wider implications for activity at other volcanoes.
This is the first study to combine gas flux measurements with computational fluid dynamics models. For full details you will of course have to read the paper! Associated videos are also available.
Pering, T. D., A. J. S. McGonigle, M. R. James, G. Tamburello, A. Aiuppa, D. Delle Donne, and M. Ripepe (2016), Conduit dynamics and post explosion degassing on Stromboli: A combined UV camera and numerical modeling treatment,Geophys. Res. Lett., 43, doi:10.1002/2016GL069001.
Tamburello, G., A. Aiuppa, E. P. Kantzas, A. J. S. McGonigle, and M. Ripepe (2012), Passive vs. active degassing modes at an open-vent volcano (Stromboli, Italy), Earth Planet. Sci. Lett., 359–360, 106–116.