arXiv1999
We consider the motion of a particle in a force field subjected to adiabatic, fluctuations of external origin. We do not put the restriction on the type of stochastic process that the noise is Gaussian. Based on a method developed earlier by us [ J. Phys. A {\bf 31} (1998) 3937, 7301] we have derived the equation of motion for probability distribution function for the particle on a coarse-grained timescale $Δt$ assuming that it satisfies the separation of timescales; $|μ|^{-1} \ll Δt \ll τ_c$, where $τ_c$ is the correlation time of fluctuations. $|μ|^{-1}$ refers to the inverse of the damping rate (or, the largest of the eigenvalues of the unperturbed system) and sets the shortest timescale in the dynamics in contrast to the conventional theory of fast fluctuations. The equation includes a third order noise term. We solve the equation for a Kramers' type potential and show that although the system is thermodynamically open, appropriate boundary conditions allow the distinct steady states. Based on the exact solution of the third order equation for the linearized potential and the condition for attainment of the steady states we calculate the adiabatic noise-induced rate of escape of a particle confined in a well. A typical variation of the escape rate as a function of dissipation which is reminiscent of Kramers' turn-over problem, has been demonstrated.