Title: Information-Thermodynamic Bound on Information Flow in Turbulent Cascade

Abstract: We investigate the nature of information flow in turbulence from an information thermodynamic viewpoint. For the shell model with thermal noise, we show that information of large-scale eddies is transferred to small scales along with the energy cascade. The information transfer rate is characterized by the large-eddy turnover time. Furthermore, we numerically show that the information-thermodynamic efficiency is extremely low compared to other typical information processing systems such as Maxwell's demon. This result implies that transferring information from large to small scales involves enormous thermodynamic costs, indicating the poor performance of turbulence as an information processing system.