A first step phenomenology for the statistics of non-equilibrium fluctuations

TL;DR

This paper develops a phenomenological framework for understanding non-equilibrium fluctuations, revealing different classes of stationary distributions and their relation to relaxation processes and intrinsic disorder.

Contribution

It introduces a new phenomenology for non-equilibrium fluctuations, identifying three classes of distributions based on a Hurst exponent and linking them to relaxation dynamics and disorder.

Findings

Identifies three classes of stationary distributions based on the Hurst exponent.

Shows the distribution functions are scale free near the origin and crossover to Maxwell-Boltzmann behavior.

Establishes a generalized fluctuation dissipation relation for non-equilibrium systems.

Abstract

The paper assesses stationary probability distributions in out of equilibrium systems. In the phenomenology proposed, no free energy can be well defined. Fluctuations of Landau free energy couplings arise when the intrinsic chemical potential leads to intrinsic disorder. The relaxation is shown to take the form of a geometrical random process. Systems of this kind show criticality features as well as that of first order transitions, which encapsulate in the form of a generalized static fluctuation dissipation relation. This will help determine three classes of distributions, which are, by defining a Hurst exponent for the relaxation rate: the regular Maxwell-Boltzmann for all H < 1/2; the usual scale free universal distributions with power law tails for H in ]1/2,1]; and a new class. The latter lies in the intermediate case, when H = 1/2. The distribution functions are scale free close to the origin, and cross-over to a Maxwell-Boltzmann asymptotic behavior, with both the scaling exponent and an effective temperature that depend on the magnitude of the intrinsic disorder.