Entropy Maximization in Nuclear Physics

Abstract

One of the first applications of the maximum entropy principle in nuclear physics—after the thermodynamic treatment of nuclear level densities by Bethe—was the derivation of the Gaussian Orthogonal Ensemble (GOE) by Porter about three decades ago. The GOE furnishes an excellent statistical description of the resonances observed in nuclear reactions (and has acquired a key role in chaos theory more recently). Since then many other applications have been found such as a simple parametrization of fission neutron spectra, establishment of the distribution of R- and S-matrix elements in compound-nuclear theory, and formal studies of the equilibration process leading from doorway states to compound states in nuclear reactions. On a more mundane level, Bayesian and maximum entropy techniques are becoming accepted tools of data reduction in nuclear physics, e.g., unfolding of resolution- and temperature-broadened scattering and other reaction data, and estimation of nuclear model parameters from empirical data.