I think, the first step is that the system is aperiodic at all scales, and (perhaps?) it is non deterministic at all scales (that is, given any model M that tried to model this system, no matter how narrow the time window or how small the region of space, the system always evolve in some way outside the prediction by the model)
(I am actually not sure if that's even a scientific thing...)
(A rough analogy is something that is random at all scale, and does not get any more predictable no matter how much you zoom in, unlike a wave packet where if you take a shot enough time interval, you ca…