** Deadline for applications for this workshop is February 5, 2019 **
This workshop will be held at the
The forecasting of natural hazards, whether earthquakes, volcanic eruptions, floods, or landslides, poses a difficult challenge for earth scientists. Often a cascade of events accompanies a natural disaster — rain weakens the soil on a fire-scorched hillside, which gives way causing a large landslide, killing people and damaging property. Models of the principal disaster, the landslide, have many inputs, most of which are uncertain and many of which are affected by rain and fire. The trigger for the landslide is a problem of soil stability, which is then coupled with a mass flow model, in order to predict the regions at risk.
Scientists are beginning to understand the propagation of uncertainty through mathematical models, to enable predictions of the likely outputs given the uncertainty of inputs. More challenging is the propagation of uncertainty through coupled models. Outputs of the first model are usually not directly related to the inputs of the second model — so how does one effect the coupling? How do the assumptions and uncertainties of the first model propagate through to outputs of the second? Although coupling has been recognized as an important issue in quantifying uncertainty, the analysis of coupled models has only very recently been the subject of mathematical and statistical study.
Earth scientists, applied and computational mathematicians, and statisticians are invited to meet to examine modeling of coupled hazards, to identify the primary challenges in making predictions, and to devise effective strategies to advance methodologies for tackling these complex problems. There will be an examination of the physics of the dominate geophysical process such as mudslides, landslides, avalanches, volcanic mass flow events and airborne ash clouds, and their attendant uncertainties. Researchers will examine and apply new statistical methodologies that show promise for the efficient coupling of hazards.
Questions: email email@example.com