An extension of the Lyapunov analysis for the predictability problem

Abstract

The predictability problem for systems with different characteristic time scales is investigated. It is shown that even in simple chaotic dynamical systems, the leading Lyapunov exponent is not sufficient to estimate the predictability time. This fact is due the saturation of the error on the fast components of the system which therefore do not contribute to the exponential growth of the error at large errors. It is proposed to adopt a generalization of the Lyapunov exponent which is based on the natural concept of error growing time at finite error size. The method is first illustrated on a simple numerical model obtained by coupling two Lorenz systems with different time scales. As a more realistic example, this analysis is then applied to a toy model of Atmospheric circulation recently introduced by Lorenz.