Short-time scales in the Kramers problem: past, present, future (review and roadmap dedicated to the 95th birthday of Emmanuel Rashba)

Abstract

The problem of noise-induced transitions is often associated with Hendrik Kramers due to his seminal paper of 1940, where an archetypal example - one-dimensional potential system subject to linear damping and weak white noise - was considered and the quasi-stationary rate of escape over a potential barrier was estimated for the ranges of extremely small and moderate-to-large damping. The gap between these ranges was covered in the 80th by one of Rashba's favourite disciples Vladimir Ivanovich Mel'nikov. It is natural to pose a question: how does the escape rate achieve the quasi-stationary stage? At least in case of a single potential barrier, the answer seems to be obvious: the escape rate should smoothly and monotonously grow from zero at the initial instant to the quasi-stationary value at time-scales of the order of the time required for the formation of the quasi-stationary distribution within the potential well. Such answer appeared to be confirmed with the analytic work of Vitaly Shneidman in 1997. However our works in the end of the 90th and in the beginning of the 2000th in collaboration with one more Rashba's favorite disciple Valentin Ivanovich Sheka and with Riccardo Mannella showed that, at a shorter time-scale, namely that of the order of the period of natural oscillations in the potential well, the escape rate growth generically occured stepwise or even in an oscillatory manner. Analytic results were confirmed with computer simulations. In the present paper, we review those results and provide a roadmap for the development of the subject, in particular demonstrating that various recently exploited experimental systems are excellent candidates for the observation of the above non-trivial theoretical predictions and, moreover, they promise useful applications.