Chaotic and nonlinear processes are widespread in the ocean and atmosphere. However, most climate fluctuations can be effectively described using linear models. These models, which generally consist of a memory-bearing system, such as the ocean, integrating the stochastic variability of the atmosphere, successfully reproduce the "red noise" signals observed in both historical records and coupled model simulations. For this reason, the climate system is considered predominantly linear, with a limited influence of nonlinear or chaotic processes on low-frequency variability.

Recent advances in climate simulations using global models have revealed processes that transcend the linear paradigm. These include asymmetries in the evolution of events like El Niño and La Niña, the amplification of signals through teleconnections between the tropics and extratropical regions, and the activation of internal modes of variability induced by anthropogenic forcings. These discoveries offer new insights into the causes and predictability of climate variability, opening the door to their exploration through tools from system dynamics.