Friday, 18 September 2026

Bayesian calibration

Today sees the first preprint appearing on the arXiv from my student Sam Sullivan.  It represents the results of work he undertook as part of a BAND fellowship which took him to Michigan State University last year. He used Bayesian methods to fit a new optical potential appropriate to a particular region of the nuclear chart (around zirconium isotopes).  The method intrinsically includes uncertainty quantification, so that any predictions made with the optical potential come with well-formulated error bars.  

Sam has submitted it to Phys Rev C, and we're hoping that's where his first publication will appear.  Fingers crossed 

Here is a plot  from the paper showing predictions of neutron elastic scattering cross sections in Y-89