KO-7   Wednesday July 10 - 08:45

MS30 Part 2 of 2 - Stochastic and multi-scale methods in climate science

MS30 Part 1 of 2

Abstract:
The climate system and its main constituents, such as atmosphere and oceans, are highly complex dynamical systems with a wide range of spatio-temporal scales. Because of their complexity and multiscale character, analysis of these systems is challenging and often involves numerical simulation. Modern developments in modeling and simulation of the climate system include the use of multiscale methods and/or stochastic modeling for representing small-scale processes in computational models of large-scale atmosphere/ocean flow. Another newly emerging research direction is the development of methodologies for analysis and efficient simulation of extreme events such as heat waves, often employing the theory of large deviations. This minisymposium will feature presentations that highlight these developments both from a theoretical and from more applied perspectives.

Organizers:
Daan Crommelin
Georg Gottwald


08:45 - 09:15 - Laure Zanna - Deep Learning for Ocean Data Inference and Turbulence Parameterisation [Abstract]

09:15 - 09:45 - CANCELLED Etienne Memin - A consistent framework for stochastic representation of large-scale geophysical flows [Abstract]

09:15 - 09:45 - Edriss Titi - Uniform error estimates for a downscaling data assimilation algorithm [Abstract]

09:45 - 10:15 - Jeroen Wouters - Stochastic model reduction for slow-fast systems with moderate time-scale separation [Abstract]

10:15 - 10:45 - Jason Frank - The stochastic limit of backward error analysis of two-scale systems [Abstract]