Volcanic ash generation: Effects of componentry, particle size and conduit geometry on size-reduction processes

Grain size distributions (GSD) of pyroclastic materials are the product of processes ranging from primary fragmentation efficiency to tephra transport. As such, a detailed description of their physical and chemical state can provide pivotal information regarding such processes. By constraining the GSD of volcanic deposits one can thereby deliver powerful constraints on the energetics and dynamics of eruptions. In order to do so we must distinguish between two primary controls: 1) fragmentation of magma to tephra, and 2) secondary transport-related processes (such as abrasion and comminution), within the conduit and until deposition of the particles. Variations in particle interactions in the conduit together with conduit geometry, may be major factors in modifying the GSD. As in all physicochemical processes, for eruptive dynamics an experimental basis is an essential element of calibration and quantification. Here, we have conducted the first experimental investigation linking fragmentation and ash production via the influence of 1) particle-componentry and particle-size, and 2) the conduit geometry (constricted versus unconstricted) on the GSD. Rapid decompression experiments with loose tephra material from the fall deposit of Pomici Principali eruption (10.3 ka, Campi Flegrei) have been conducted in an optically transparent setup that enables the optical monitoring of particle dynamics with a high-speed camera. The samples employed can be classified into two main groups; 1) pumices, and 2) dense clasts (including crystals, lava clasts and wall rock fragments). Our results indicate that 1) a conduit diameter constriction (simulating obstacles in the conduit walls) is likely to reduce the average diameter of individual clasts and increase the generation of ash (<63 μm, fine ash); 2) the presence of larger pumice particles in the starting sample results in more efficient ash production; 3) the presence of dense particles is linked to a higher efficiency of size-reduction processes (whereby the presence of crystal-rich pumice appears to have a counterproductive effect). These findings correlate well with existing studies, enabling a general assessment of possible secondary size-reduction processes controlling the GSD of a volcanic deposit. Further, our results also suggest that secondary-generated ash should be considered for hazard assessment and modeling, since increased amounts of ash may affect, e.g., the plume dispersion dynamics, and in turn the area affected by ash deposition.