Limits to Fluctuation Dynamics

The fluctuation of an experimentally measured observable, along with its mean, constitutes the fundamental ingredient of a non-equilibrium system involving randomness. Despite previous efforts, a comprehensive framework for characterizing the temporal dynamics of fluctuations of observables remains elusive. In this manuscript, we develop a ubiquitous theory concerning rigorous limits to the rate of fluctuation growth. We discover a simple principle that the time derivative of the standard deviation of an observable is upper bound by the standard deviation of an appropriate observable describing velocity. This indicates a hitherto unknown tradeoff relation between the changes for the mean and standard deviation, i.e., the sum of the squares for these quantities cannot exceed certain cost determined by dynamical processes. The cost can be kinetic energy for hydrodynamics, irreversible entropy production rate for thermodynamic processes, energy fluctuations for unitary quantum dynamics, and quantum Fisher information for dissipative quantum dynamics. Our results open an avenue toward a quantitative theory of fluctuation dynamics in various non-equilibrium systems, encompassing quantum many-body systems and nonlinear population dynamics, as well as toward our understanding of how to control them.Comment: Main text: 7 pages, 1 figure; Supplemental Materials: 7 pages, 1 figur