Random transitions in stochastic turbulent flows

Many turbulent systems exhibit random switches between qualitatively different attractors. The transition between two different attractors is often an extremely rare event that can not be computed through direct numerical simulations due to complexity limitations. In this talk, we present results for the calculation of the most probable transition trajectories or instantons between non-equilibrium stationary states (attractors) in the 2D stochastic Navier-Stokes equations. By representing the transition probability between two given states as a path integral, we can determine the instanton by the minimization of an appropriate action functional.