LANS Informal Seminar: Noemi Petra

Seminar Title: From data-to-predictions under uncertainty for Antarctic ice sheet flow
Speaker: Noemi Petra, Assistant Professor, School of Natural Sciences, University of California, Merced

Date/Time: 2014-12-03 15:00
Location: Building 240 Room 1404-1405

Description:
We present efficient and scalable algorithms for an end-to-end data-to-prediction process under the Gaussian approximation and in the context of modeling the flow of the Antarctic ice sheet and its effect on sea level. The ice is modeled as a viscous, incompressible, creeping, shear-thinning fluid. The observational data come from InSAR satellite measurements of surface ice flow velocity, and the uncertain parameter field to be inferred is the basal sliding parameter, which is represented by a heterogeneous coefficient in the Robin boundary condition at the base of the ice sheet. For illustration we consider a simplified prediction, namely we take the quantity of interest as the present-day ice mass flux from the Antarctic continent to the ocean. We show that the work required for executing this data-to-prediction process – measured in number of forward (and adjoint) ice sheet model solves – is independent of the state dimension, parameter dimension, data dimension, and number of processor cores. The key to achieving this dimension independence is to exploit the fact that, despite their large size, the observational data typically provide only sparse information on model parameters. This property can be exploited to construct a low rank approximation of the linearized parameter-to-observable map via randomized SVD methods and adjoint-based actions of Hessians of the data misfit functional.

Bio: Noemi Petra is an assistant professor in the Applied Mathematics department in the School of Natural Sciences at the University of California, Merced. She earned a bachelor degree in Mathematics and Computer Science and a master degree in Computer Science from Babes-Bolyai University (Romania) in 2003 and 2004, respectively, and master and PhD degrees in Applied Mathematics from the University of Maryland, Baltimore County in 2007 and 2010, respectively. Prior to joining UC Merced, Noemi was the recipient of the ICES Postdoctoral Fellowship during 2010-2011 in the Institute for Computational Engineering and Sciences (ICES) at The University of Texas at Austin, after which she joined the Center for Computational Geosciences and Optimization group in ICES as research associate. Her research interests include PDE-constrained optimization, inverse problems, uncertainty quantification and optimal experimental design.