[Paper Review] Superstatistics Based on the Microcanonical Ensemble

Christophe Vignat, A. Plastino+1 more|arXiv (Cornell University)|May 24, 2005

Statistical Mechanics and Entropy11 citations

TL;DR

This paper introduces a novel form of superstatistics based on the microcanonical ensemble, where the fluctuating parameter is the nonextensivity index $ q $ of Tsallis statistics rather than temperature. By superposing Tsallis distributions with $ q $-fluctuations following a quasi-Poisson distribution, the authors derive a Gaussian distribution as the effective equilibrium state, demonstrating that Gibbsian statistics emerges from a superposition of finite, $ q $-statistical baths.

ABSTRACT

Superstatistics is a "statistics" of "canonical-ensemble statistics". In analogy, we consider here a similar theoretical construct, but based upon the microcanonical ensemble. The mixing parameter is not the temperature but the index q associated with the non-extensive, power law entropy Sq.

Motivation & Objective

• To develop a new superstatistics framework based on the microcanonical ensemble, replacing temperature fluctuations with fluctuations in the nonextensivity index $ q $.
• To explore how the canonical (Gibbs-Boltzmann) distribution can arise as an effective equilibrium state from a superposition of nonextensive (Tsallis) statistics.
• To provide a physical justification for the emergence of Gaussian statistics in systems interacting with finite heat baths of varying size.
• To connect the observed $ q $-fluctuations to the statistical mechanics of finite reservoirs, particularly through Poisson-like distributions of bath particle numbers.
• To establish a reciprocal mechanism to Beck and Cohen’s temperature-fluctuation superstatistics, where finite baths give rise to $ q $-statistics, and their superposition yields Gibbs statistics.

Proposed method

• Proposes a superstatistics framework where the mixing parameter is the nonextensivity index $ q $, rather than inverse temperature $ \beta $, using a multiplicative convolution over $ q $-distributions.
• Defines the Tsallis distribution $ S_{q,\sigma^2}(x) $ with variance $ \sigma^2 \frac{q-1}{3q-1} $, normalized via a gamma function expression.
• Introduces a discrete distribution $ p_k(a) $ with quasi-Poisson weights $ \propto \frac{a^{k+1/2}}{\Gamma(k+3/2)} $ to superpose Tsallis distributions.
• Derives a key identity showing that a weighted sum of Tsallis distributions with $ q = 1 + \frac{1}{k} $ and $ \sigma^2 = 2a \sigma^2 $ yields a Gaussian distribution over $ [-\sigma\sqrt{2a}, \sigma\sqrt{2a}] $.
• Establishes a link between the number of particles $ N $ in a finite heat bath and the nonextensivity index $ q $, finding $ q = \frac{N}{N-1} $.
• Uses the ergodic hypothesis to interpret time-averages over short intervals $ \Delta T $ as $ q $-averages with $ q $-values distributed via a Poisson-like law, while long-time averages recover Gibbs statistics.

Experimental results

Research questions

• RQ1 Can a superstatistics framework be constructed based on the microcanonical ensemble, with $ q $ as the fluctuating parameter instead of $ \beta $?
• RQ2 How does the superposition of Tsallis distributions with fluctuating $ q $ lead to a Gaussian equilibrium distribution?
• RQ3 What physical mechanism underlies the emergence of Gibbs-Boltzmann statistics from a collection of finite, nonextensive baths?
• RQ4 Why do the weights in the superposition follow a quasi-Poisson distribution, and what does this imply about the effective size of the heat bath?
• RQ5 How does this $ q $-fluctuation mechanism differ from Tsallis’ own scenario of relaxation to $ q $-statistics followed by relaxation to Gibbs statistics?

Key findings

• A superposition of Tsallis distributions with $ q = 1 + \frac{1}{k} $ and $ \sigma^2 = 2a\sigma^2 $, weighted by a quasi-Poisson distribution $ p_k(a) \propto \frac{a^{k+1/2}}{\Gamma(k+3/2)} $, yields a Gaussian distribution over $ [-\sigma\sqrt{2a}, \sigma\sqrt{2a}] $.
• The normalization factor in the superposition ensures the total weight integrates to unity, confirming the validity of the $ p_k(a) $ distribution.
• The nonextensivity index $ q $ is related to the number of particles $ N $ in a finite heat bath via $ q = \frac{N}{N-1} $, with $ q \to 1 $ as $ N \to \infty $.
• The effective Gaussian distribution emerges as the long-time average, while short-time averages correspond to $ q $-statistics, suggesting a dual-time-scale ergodicity.
• The mechanism provides a physical basis for the emergence of Gibbs statistics not from a single infinite bath, but from a superposition of finite baths with fluctuating sizes.
• The quasi-Poisson weights arise naturally from the statistical mechanics of a system interacting with a finite number of particles in a volume $ \Delta V $, consistent with Poisson statistics for particle counts.