A Bayesian method to rank different model forecasts of the same volcanic ash cloud: Chapter 24

Geophysical Monograph
By: , and 



Volcanic eruptions often spew fine ash high into the atmosphere, where it is carried downwind, forming long ash clouds that disrupt air traffic and pose a hazard to air travel. To mitigate such hazards, the community studying ash hazards must assess risk of ash ingestion for any flight path and provide robust and accurate forecasts of volcanic ash dispersal. We provide a quantitative and objective method to evaluate the efficacy of ash dispersal estimates from different models, using Bayes theorem to assess the predictions that each model makes about ash dispersal. We incorporate model and measurement uncertainty and produce a posterior probability for model input parameters. The integral of the posterior over all possible combinations of model inputs determines the evidence for each model and is used to compare models. We compare two different types of transport models, an Eulerian model (Ash3d) and a Langrangian model (PUFF), as applied to the 2010 eruptions of Eyjafjallajökull volcano in Iceland. The evidence for each model benefits from common physical characteristics of ash dispersal from an eruption column and provides a measure of how well each model forecasts cloud transport. Given the complexity of the wind fields, we find that the differences between these models depend upon the differences in the way the models disperse ash into the wind from the source plume. With continued observation, the accuracy of the estimates made by each model increases, increasing the efficacy of each model’s ability to simulate ash dispersal.

Additional publication details

Publication type Book chapter
Publication Subtype Book Chapter
Title A Bayesian method to rank different model forecasts of the same volcanic ash cloud: Chapter 24
Series title Geophysical Monograph
DOI 10.1029/2012GM001249
Year Published 2012
Language English
Publisher American Geophysical Union
Contributing office(s) Volcano Science Center
Description 12 p.
Larger Work Type Book
Larger Work Title Lagrangian Modeling of the Atmosphere
First page 299
Last page 310
Google Analytic Metrics Metrics page
Additional metadata about this publication, not found in other parts of the page is in this table