This is extremely well written article. I recommend it for publication in Scholarpedia.
I have a few minor remarks.
1. In section “Error growth and the prediction of complex systems”, please include limits in the equation for the largest Lyapunov exponent. 2. In section “Taming the butterfly: the probabilistic approach to prediction”, 3-d paragraph I suggest to drop the word “law”, so that the sentence reads: “One of the novelties brought by the probabilistic description is that the evolution of the underlying probability distributions (described by Liouville, master or Fokker-Planck equations) - which now become the principal quantities of interest - are linear and display strong stability and uniqueness properties.