Approaching Turbulence through Information Theory

Just as a string of words in a text message or a sequence of pixels in an image can be treated as a stream of information, similarly a series of measured velocities, can also be analyzed using the tools of information theory. A simple message (or laminar flow), contains less information (fewer surprises, less entropy S) than a complex story. But both contain redundancies (or correlations), which diminish S. On the other hand, randomly ordered words or white noise have a maximum entropy (``the'' appears often; ``hte'' almost never). In the experiments to be discussed, information theory is used to predict future observations by extrapolating from the immediately preceding velocity string.