11:45 AM - 12:00 PM
[SVC33-10] A volcanic ashfall simulation accounting for ash fingers
Keywords:Ash finger, Volcanic ash transport and dispersion model, Sakurajima Volcano, vulcanian eruption
Atmospheric advection and diffusion transport volcanic ash particles emitted from the vent during explosive eruptions. Ashfall distribution can be predicted using volcanic ash transport and dispersion models (VATDMs), which commonly assume that particles fall at their terminal velocity. However, the concentrated ash currents that can fall faster than the terminal velocity are observed at the base of volcanic plumes. These currents, known as ‘ash fingers,’ arise from gravitational instabilities driven by density differences between ash plumes and the surrounding atmosphere, and they can enhance sedimentation through their higher fall velocity (e.g., Carazzo and Jellinek, 2013). This study investigates how ash fingers modify volcanic ash transport using a VATDM incorporating their effects.
We developed a VATDM incorporating ash-finger physics based on the FALL3D framework (Costa et al., 2006; Folch et al., 2020), which numerically solves advection–diffusion equations to track the evolution of particle concentration and size distribution. The particle size distribution was discretized into bins from −2φ to 4.5φ in 0.5φ increments. For each size bin, the terminal velocity vt was calculated from the particle size using the formula by Suzuki (1983). The finger falling velocity vf was calculated from the particle concentration and a representative terminal velocity—defined as the mass-weighted average across size bins—using the formula of Hoyal et al. (1999). Following the experiments (Fries et al., 2024), we assumed that a particle settles at its terminal velocity vt when vt > vf but descends at the ash-finger velocity vf when vt < vf. We followed the numerical treatments in FALL3D.
We tested the model for Vulcanian eruptions under the topographic and atmospheric conditions of the Sakurajima volcano. We used a 5 km mesh dataset from the mesoscale model (MSM) provided by the Japan Meteorological Agency and downscaled it to 0.5 km resolution using the WRF model (ver. 4; Skamarock et al., 2019). For an eruption with a 2.5 km plume and 1.7 x 107 kg of erupted mass, particles coarser than 3φ settled at vt, while finer particles descended as ash fingers, resulting in heavier proximal deposits. For an eruption with a 4.7 km plume and 4.4 x 107 kg of erupted mass, we observed notable differences in fine-particle distribution around 5 km from the vent when the effects of ash fingers were included. In this talk, we compare the simulation results with the field data and discuss the impact of incorporating ash fingers on estimating volcanic ash deposits' spatial distribution and arrival time.
We developed a VATDM incorporating ash-finger physics based on the FALL3D framework (Costa et al., 2006; Folch et al., 2020), which numerically solves advection–diffusion equations to track the evolution of particle concentration and size distribution. The particle size distribution was discretized into bins from −2φ to 4.5φ in 0.5φ increments. For each size bin, the terminal velocity vt was calculated from the particle size using the formula by Suzuki (1983). The finger falling velocity vf was calculated from the particle concentration and a representative terminal velocity—defined as the mass-weighted average across size bins—using the formula of Hoyal et al. (1999). Following the experiments (Fries et al., 2024), we assumed that a particle settles at its terminal velocity vt when vt > vf but descends at the ash-finger velocity vf when vt < vf. We followed the numerical treatments in FALL3D.
We tested the model for Vulcanian eruptions under the topographic and atmospheric conditions of the Sakurajima volcano. We used a 5 km mesh dataset from the mesoscale model (MSM) provided by the Japan Meteorological Agency and downscaled it to 0.5 km resolution using the WRF model (ver. 4; Skamarock et al., 2019). For an eruption with a 2.5 km plume and 1.7 x 107 kg of erupted mass, particles coarser than 3φ settled at vt, while finer particles descended as ash fingers, resulting in heavier proximal deposits. For an eruption with a 4.7 km plume and 4.4 x 107 kg of erupted mass, we observed notable differences in fine-particle distribution around 5 km from the vent when the effects of ash fingers were included. In this talk, we compare the simulation results with the field data and discuss the impact of incorporating ash fingers on estimating volcanic ash deposits' spatial distribution and arrival time.