The fluid dynamics at the impact zone of an impinging jet dictate the initial conditions for the resulting downstream flow. Current understanding of this impact zone is limited to small scale, single phase or monodisperse impinging jet experiments with a focus on industrial and engineering applications. Here, we present multifield numerical modeling results of mono- and polydisperse impinging jets with length scales and physical parameters relevant to large explosive volcanic eruptions. Our modeling shows that particle behavior is sensitive to whether a jet is monodisperse or polydisperse. For a monodisperse jet, the downstream flow behavior can be predicted by the coupling between the particles and gas which we characterize by a form of the Stokes number (Stimp) where the time scale of changes in fluid motion is defined by the length scale and velocity change associated with vertical deceleration as a mixture approaches an impact surface. We find that the length scale of deceleration is sensitive to the mixture Mach number, as high Mach number flows produce standing shocks upstream of the impact site. For low Stimp, monodisperse cases the particles make the transition from axial to radial flow easily and the flow continues downstream relatively unimpeded and well mixed. In contrast, in situations where the monodisperse particles are sufficiently poorly coupled (high Stimp), the particles rebound and/or concentrate at the impact zone, which results in a radial acceleration of gas as it is expelled from the concentrating mixture. In polydisperse jets, larger particles become better coupled to the gas in the free jet zone when in the presence of smaller, well-coupled particles due to particle-particle drag. However, the larger particles still lose significant momentum in the impact zone, which results in a lag effect where the smaller particles and gas are expelled and advance radially at a greater velocity. Although the simulations are simplified compared to real volcanic eruption scenarios, the results suggest that processes in the impact zone contribute directly to the formation of different types of gas-particle flows (concentrated versus dilute) that move outward across the ground.

We analyze the eruptive products of Holocene-aged Dotsero volcano (Colorado, USA; ~4150 y.b.p.), which record evidence of a progression from effusive magmatic activity to explosive phreatomagmatic maar-forming activity to a final explosive magmatic phase. The nature of the deposits suggests that the irregular and mountainous pre-eruptive topography strongly influenced the formation of directed pyroclastic density currents during the phreatomagmatic phase, where the topographically high northern crater rim acted as a barrier and promoted transport to the south. Furthermore, the crater shape strongly controlled the grain size of the final deposits, providing further evidence that deposit grain size can be a misleading proxy for fragmentation or explosion intensity. We test these hypotheses gleaned from fieldwork using multiphase numerical modeling, extending a framework developed to model subsurface explosions in diatremes to include the surface and crater. Using a bidisperse particle population (d1=0.0001 m and d2=0.01 m), we confirm that a crater shape where one rim is higher than the other promotes multiple pulses of particle transport in the direction of the lower rim. The initial currents contain both coarse and fine particles, where the secondary and tertiary currents contain predominately the finer particles. This mechanism provides an alternative explanation to the formation of lapilli tuff/tuff couplets often found at maar-diatreme tephra rings (including Dotsero). Additional modeling shows how the depth of a crater can control the grain size of the resulting pyroclastic density currents. In the case of a deep crater and relatively weak explosion, it is possible that only the fine particles escape following the collapse of the eruptive column, resulting in a fine-grained deposit. This work adds to the expanding hypotheses and models developed to describe the eruptions and dynamics of maar-diatreme volcanoes, but also highlights the limitations of our knowledge of key parameters, especially those related to the processes of magma-water interaction and syn-eruptive hydrology.

Phreatomagmatic explosions are some of the most dynamic and poorly understood phenomena in volcanology. Questions related to magma-water premixing, known to be a necessary process for producing an explosion, are difficult to address in the field or with experiments due to practical and scale constraints. (Abstract shortened by ProQuest.).