TUNL Researchers Publish Evaluation of Thermonuclear Reaction Rates

Thermonuclear reaction rates are essential inputs for models of stellar structure, stellar explosions, and the early Universe. Earlier benchmark evaluations—such as those led by Nobel Laureate W. A. Fowler—laid the groundwork for modern stellar modeling but did not assign reaction rates a rigorous statistical meaning based on probability density functions.

In a TUNL-led collaboration, authors Christian Iliadis, Richard Longland, Kiana Setoodehnia, Caleb Marshall, Peter Mohr from ATOMKI (Hungary), and Athanasios Psaltis present a comprehensive reevaluation of experimental thermonuclear reaction rates using modern statistical techniques. The team applies Bayesian inference and Monte Carlo methods to incorporate experimentally-determined nuclear inputs consistently, quantify systematic uncertainties, and propagate correlations among nuclear parameters. The study also examines how measured resonance strengths and indirectly estimated partial widths influence rate calculations across relevant stellar temperatures.

The resulting dataset, published in The Astrophysical Journal Supplement Series (2026, Vol. 283, p. 17), provides 78 experimentally derived reaction rates tabulated over a defined temperature grid. For each reaction, the authors report statistically meaningful uncertainties, offering a transparent, reproducible foundation for next-generation stellar models. By anchoring rates to probability density functions and explicitly tracking correlations, the work enhances the reliability of nucleosynthesis and stellar evolution simulations.