To what extent can dynamical models describe statistical features of turbulent flows?

Statistical features of ``bursty'' behaviour in charged and neutral fluid turbulence are compared to statistics of intermittent events in a GOY shell model, and avalanches in different models of Self-Organized Criticality (SOC). It is found that inter-burst times show a power law distribution for turbulent samples and for the shell model, a property which is shared only in a particular case of the running sandpile model. The breakdown of self-similarity generated by isolated events observed in the turbulent samples is well reproduced by the shell model, while it is absent in all SOC models considered. On this base, we conclude that SOC models are not adequate to mimic fluid turbulence, while the GOY shell model constitutes a better candidate to describe the gross features of turbulence.