Reduced Modeling with Optimal Bases

Daan Crommelin
Courant Institute of Mathematical Sciences

Abstract
It is known that atmospheric models based on spectral expansions are not very efficient. By using a different model basis, the number of degrees of freedom needed to produce a certain degree of realism with a model can be significantly reduced - a gain that can be valuable for extended simulations of atmosphere/ocean models. I will discuss several types of optimal bases, and show results from a comparison study.

Back to agenda