A hierarchical Bayesian framework to probe the Big-Bang nucleosynthesis

25/04/2018 - 14:00 - 15:00
Auditório IAG, Bloco G

A hierarchical Bayesian framework to probe the Big-Bang nucleosynthesis


Rafael S. de Souza

Dept. of Physics & Astronomy

University of North Carolina - USA


We developed a principled statistical framework  for uncertainty quantification of nuclear reaction rates  at energies of interest for the Big-Bang nucleosynthesis (BBN).  The Big-Bang theory rests mainly on three observations: BBN, the cosmic expansion, and the cosmic microwave background radiation. Primordial nucleosynthesis occurs during the first 20 minutes after the big bang, at temperatures near 1 GK, and is responsible for the production of the light nuclides.

Although a large body of directly measured data exist, the estimation of thermonuclear reaction rates from the measured cross section (or S-factor) data remains challenging. Rates obtained using traditional statistics are plagued by a number of problems, including the treatment of systematic and statistical uncertainties, and the effects of electron clouds in the nucleus potential. Our approach simultaneously account for intrinsic, statistical, and systematic uncertainties, and the effects of electron screening.

Twelve nuclear processes of interest take place during BBN,  whose reactions have been measured directly in the laboratory at the energies of astrophysical interest. As a case in point, this work reports astrophysical S-factors and thermonuclear rates for the helium-3 deuterium reaction, which impacts the primordial abundances of helium-3 and Lithium-7 and subsequent history of stellar and galaxy chemical evolutions.  

For the nuclear reaction model we implemented a full Bayesian single-level, two-channel approximation of the R-matrix theory. Our analysis of the helium-3 deuterium S-factor data results in improved estimates for the thermonuclear rates and the associated uncertainties of interest for BBN. Finally,  the incorporation of R-matrix expressions into the Bayesian model represents a testbed for future studies of more complex nuclear reactions.