Breadcrumb
Resolving Antarctic Ice Mass Trends
Fri 15 March 2013, 14:15
Andrew Zammit-Mangion
Bristol
Organisers: Nick Whiteley, Feng Yu
ABSTRACT
Sea level rise is one of the most serious consequences of climate warming. The largest potential contributors, and the ones with the greatest uncertainty, are the ice sheets covering Antarctic and Greenland. We propose a statistical synthesis of model output and observations, focusing on recently-collected data from altimetry, gravimetry, and GPS. In this talk I will describe the spatial framework currently being set up to use this data for resolving mass trends, with particular emphasis on the physical aspects of the problem. I will discuss the use of frequency analysis on forward-models to establish suitable Matern fields for the latent spatial fields and the use of finite-elements to induce the sparsity required to maintain computational tractability for inference at the glacier scale (Lindgren et al). This is joint work with Jonathan Rougier (Dept. of Maths), Jonathan Bamber and Nana Schoen (Dept. of Geographical Sciences).