A Lagrangian stochastic model of a volcanic eruption column

Abstract

We develop a Lagrangian stochastic model (LSM) of a volcanic plume in which the mean flow is provided by an integral plume model of the eruption column and fluctuations in the vertical velocity are modelled by a suitably constructed stochastic differential equation. The LSM is applied to the two eruptions considered by Costa et al. (2016) for the volcanic-plume intercomparison study. Vertical profiles of the mass concentration computed from the LSM are compared with equivalent results from a large-eddy simulation (LES) for the case of no ambient wind. The LSM captures the order of magnitude of the LES mass concentrations and some aspects of their profiles. In contrast with a standard integral plume model, i.e. without fluctuations, the mass concentration computed from the LSM decays (to zero) towards the top of the plume which is consistent with the LES plumes. In the lower part of the plume, we show that the presence of ash leads to a peak in the mass concentration at the level at which there is a transition from a negatively buoyant jet to a positively buoyant plume. The model can also account for the ambient wind and moisture.