Talk by Jeroen Wouters

Talk is postponed - new date TBA

Parametrization of fast subgrid-scale processes using Edgeworth
expansions

Many physical systems, including the Earth's climate, are too complex to numerically model in full detail. Therefore, methods for contructing reduced models are necessary.

A classical approach to model reduction is to assume a large time scale separation between slow “climate” and fast “weather” processes. In the unphysical limit of infinite time scale separation this assumption leads to a reduction to a reduced slow system with Gaussian white noise. Realistic physical systems, however, feature a finite time scale
separation.

We will discuss the use of an Edgeworth expansion in orders of the time scale separation parameter to describe the deviations from the limiting equation at finite time scale separation. We then make use of this expansion to construct a reduced model with correlated noise that more accurately approximates the full system.

This is joint work with Georg Gottwald.