Challenges for Climate and Weather Prediction in the Era of Exascale Computer Architectures: Oscillatory Stiffness, Time- parallelism, and the Role of Long-Time Dynamics

Beth Wingate

University of Exeter

Tuesday, July 18, 1:30 pm

Abstract:

For weather or climate models to achieve exascale performance on next-generation heterogeneous computer architectures they will be required to exploit on the order of hundred-million-way parallelism. This degree of parallelism far exceeds anything possible in today’s complex models even though they are highly optimized. In this talk I will discuss one of the mathematical issues that leads to the limitations in space- and time-parallelism for climate and weather prediction models – oscillatory stiffness in the PDE that leads to time scale separation. I will discuss the case when the time scale separation is infinite, including a fast-converging HMM-type parareal method and a time-parallel matrix exponential. In addition I will present new convergence results for the case when the time scale separation is finite.

Biography:

Professor Beth Wingate's main research interest is the study of oscillations in fluid mechanics, mathematics, and numerics for high performance computing. Her recent research is focused on physics of the Arctic Ocean, direct numerical simulations, and time-stepping methods for HPC and climate modeling, and the fluid mechanics of the slow/fast manifolds. She did her PhD work at the University of Michigan studying numerics, waves and ocean fluid dynamics. Other interests include spectral element methods, in particular the investigation of near optimal interpolation on triangles. She spent many years at the Los Alamos National Laboratory in New Mexico, USA before moving to the University of Exeter in Devon, UK in 2013.