Synchronized Chaos In Climate Dynamics

摘要

Two parallel, quasi-two-dimensional fluid dynamical channel models, each of which vacillates chaotically between meteorologically significant dynamical regimes, synchronize when only the medium-scale eddy components of the flow are coupled. That the smallest scales need not be coupled suggests a low-dimensional dynamics, in accordance with the theory of inertial manifolds, and validates the use of low-order models in synchronization studies. Generalized synchronization of two channels with forcing terms localized in different sectors implies a relationship between large-scale circulation patterns over the Atlantic and Pacific oceans. However, the resulting correlations are small because of the large time lag in the coupling. In a different interpretation of the two channels - one as "truth" and the other as "model" -synchronization suggests that a limited band of spatial frequencies in observed data will suffice to update weather prediction models, and provides a framework for the assessment of data assimilation strategies.