Ein internes Vorhersagbarkeitsexperiment im Lorenz-Modell

The predictability of the atmosphere usually is quantified by the error doubling time or the limit of predictability. These measures are times when small errors double on average or reach a certain threshold value respectively. In this article we analyse the error doubling time in the Lorenz Model. We want to demonstrate, that averaged measures of predictability are not necessarily typical measures for the system. Predictability is a variable and it is an important task to identify which initial states lead to enhanced, or reduced, predictability