Predictability in Dynamical Systems Relevant to Climate and Weather

Richard Kleeman
Courant Institute of Mathematical Sciences

Abstract
Variability of predictability from one forecast to another is an important utilitarian consideration in practical climate and weather prediction. Here we review new theoretical tools for examining this question derived from information theory. We also examine the question in a selection of simple and complex/realistic models appropriate for both ENSO and weather predictions. Our conclusions for the weather case differ from what is often conventionally assumed in practical statistical prediction.

Back to agenda