Coupling functions in climate

Abstract

We examine how coupling functions in the theory of dynamical systems provide a quantitative window into climate dynamics. Previously we have shown that a one-dimensional periodic non-autonomous stochastic dynamical system can simulate the monthly statistics of surface air temperature data. Here we expand this approach to two-dimensional dynamical systems to include interactions between two sub-systems of the climate. The relevant coupling functions are constructed from the covariance of the data from the two sub-systems. We demonstrate the method on two tropical climate indices; The El-Ni\~{n}o--Southern Oscillation (ENSO) and the Indian Ocean Dipole (IOD), to interpret the mutual interactions between these two air-sea interaction phenomena in the Pacific and Indian Oceans. The coupling function reveals that ENSO mainly controls the seasonal variability of the IOD during its mature phase. This demonstrates the plausibility of constructing a network model for the seasonal variability of climate systems based on such coupling functions.