ε entropy for a time series of thermal turbulence

Abstract

We consider a dynamic entropy of a Rayleigh-Bénard flow which varies with the characteristic scales ε of the process. Such an ε entropy is used to distinguish and classify the different dynamic regimes recently observed in this system: oscillation, chaos, and soft, hard, and developed turbulences. Moreover, data analysis of an experimental time series reveals a scaling behavior of the ε entropy in a hard turbulent regime, demonstrating that this notion of entropy is suitable to describe dynamic processes with a heirarchy of characteristic instabilities.