Indistinguishable states. II: The imperfect model scenario (Q1881719): Difference between revisions

[For part I of this paper see ibid. 151, 125--141 (2001; Zbl 1052.62092).] Given a perfect model of chaotic dynamical systems and an arbitrary long series of noisy observations, it is not possible to determine the state of this system precisely, or to accurately forecast a unique future state. In the paper the case where no perfect model is at hand, indeed where the class of available models does not contain a perfect model, is considered. The concept of indistinguishable states for perfect models and methods for calculating them is extended to the case of imperfect models. It is shown that in order to maintain consistency between observations of a system and imperfect models, it is necessary to study pseudo-orbits instead of trajectories. Using pseudo-orbits one finds that they are consistend with observations and these can be used to estimate the projection of the system state. Main conclusions: 1) There is no state of the model that can be identified with the state of the system. 2) Great care must be taken when using an imperfect model to forecast the system. The forecast may not shadow the future behavior for very long, even if one were able to obtain a noisy-free projection of the system state.