Improving the accuracy of entrainment coefficients in one-dimensional volcanic plume models

Ash-cloud transport models can be used to help to mitigate the downstream impact of volcanic ash. The accuracy of these models relies on using correct values for input parameters for plume height H and mass eruption rate MER. One-dimensional plume models can be used to infer MER from H, but the accuracy of the inversion is limited by uncertainties in the air entrainment rate, especially in the bending region. The radial entrainment coefficient α is better constrained than the crossflow entrainment coefficient β. Wind tunnel experiments used stereo particle image velocimetry SPIV to measure the centerline trajectory of a jet orthogonally injected at speed U into crossflow at speed W under varying ratios U/W. Regression methods compared the experimental data to trajectories predicted by a one-dimensional plume model. Values of a and b were incremented in the model, and best-fit a-β combinations were found for each ratio U/W. The overall best-fit coefficient values (α = 0.07, β = 0.46) were applied to model estimates of historical eruptions to gauge model improvement. The results constrain the effects of data uncertainties on the accuracy of 1-D plume models, which are used to provide input to ash cloud models.