This article observes fish acoustic telemetry RMS acceleration data as stochastic process system. The system is tested for stationarity and differentiability. In order to access continuation of the stationary stochastic system, it is considered Lagrange mean theorem and Wiener-Levy process. The continuation is tested on transition matrix of probability distributions between cause and product state levels of acceleration. The Wiener-Levy process estimates the diffusion system as the transition distributions with finer resolution, while preserving the mean value and standard deviation of the original distribution. The approach may help to simulate and understate the behavior of fish in captive.