Wednesday, June 17th, 2020. Time: 12 noon
|Seminar video on Youtube|
The objective of the seminar is to give a quick review of the current state of computational glacier dynamics, some of its limitations and its future.
In the last years, advances in computational mechanics and supercomputing capabilities have allowed researchers to improve the numerical models of glaciers and ice-sheets. One of the main goals behind such improvement is to better predict the evolution of glaciers due to climate change. Providing accurate knowledge of the changes in the cryosphere will help mitigate their impact at regional - e.g. water supply availability - and global scales - sea level rise.
Unfortunately, most data available is gathered at the surface, with comparatively scarce observations of subsurface processes. As a result, subsurface processes are poorly constrained, and represented by oversimplified or unvalidated models. Predictions are sensitive to the choice of one physical model over the other, and much of the research that has been done relies on models built upon the same general and simple assumptions that were taken by the first generation of glaciologists, more than 40 years ago. In this context, new efforts are being carried out to validate, improve and understand the limitations of these models.
Juan Pedro ROLDÁN-BLASCO is a CNRS PhD student at IGE (Institute for Geosciences and Environmental research) in Grenoble, France. He was student of the Erasmus Mundus MSc in Computational Mechanics, studying the first year in CIMNE, and the second year in École Centrale de Nantes. During his PhD he is using FEM to improve glacier friction laws - glacier boundary layer equations.