Microphysical processes of volcanic ash aggregation and their implications for volcanic eruption dynamics

Although numerous hazard models exist to assess possible ash fallout from explosive volcanic eruptions around the world, these models frequently neglect to consider ash aggregation or use a simple percent proxy to represent aggregation, without considering the varying processes at work throughout the volcanic flow. Eruption dynamics are sensitive to ash aggregation, and ash aggregates are commonly found in eruptive deposits, yet few experiments have been conducted on aggregation phenomena using natural materials. In this work, experiments were developed to produce both probabilistic and process-based relationships for the efficiency of ash aggregation with respect particle size, collision kinetic energy, atmospheric water vapor and residence time. A synthetic ash proxy, ballotini, and ash from the 2006 eruption of Tungurahua, Ecuador, and the 1980 eruption of Mount St. Helens, WA, were examined for their aggregation potential. Two aggregation regimes, wet and dry, were identified based on their potential for aggregation. The wet flow regime occurs when particles are circulated in high relative humidity environments long enough to develop a water layer with a thickness that exceeds the particle roughness scale. Hydrodynamic forces control aggregation in the wet flow regime. The dry flow regime includes particles in low relative humidity environments as well as those that circulate too briefly in high humidity environments to fully develop a water layer. Electrostatic forces control aggregation in the dry flow regime. Aggregation efficiency in both regimes was dominantly controlled by collision kinetic energy; however, this effect is significantly dampened in the wet flow regime. Equations governing the relationships between aggregation efficiency, collision kinetic energy and the related forcings in the wet or dry flow regimes have been developed for implementation into large-scale numerical volcanic models. The results of this experimental work have been developed into a probability distribution that has been integrated and incorporated into a multifluid numerical model. The numerical simulation was tested on a range of explosive depths and overpressure estimates from the 1790 eruption of Kilauea volcano, HI. The model output was compared to field data collected on the deposit thickness moving away from the source and the distribution, including both size and density, of aggregates. The mass fraction of ash removed from the eruption column in the form of aggregates was also calculated to examine how efficiently aggregation processes remove ash throughout the eruption. Cumulatively, the work presented here furthers our understanding of aggregation processes and the role they play in volcanic eruptions.Ph.D