Automated Resonance Identification in Nuclear Data Evaluation

TL;DR

This paper introduces an automated method for resonance region evaluation in nuclear data, combining optimization and statistical inference to reduce manual effort, bias, and improve reproducibility.

Contribution

It presents a novel automated approach that integrates non-convex optimization with inferential statistics for resonance model inference, reducing manual effort and bias.

Findings

Automated resonance model inference reduces evaluation time.

Method improves reproducibility and documentation.

Enhances workflow efficiency for nuclear data evaluators.

Abstract

Global and national efforts to deliver high-quality nuclear data to users have a broad impact across applications such as national security, reactor operation, basic science, medical fields, and more. Cross section evaluation is a large part this effort as it combines theory and experiment to produce suggested values and uncertainty for reaction probabilities. In most isotopes, the cross section exhibits resonant behavior in what is called the resonance region of incident neutron energy. Resonance region evaluation is a specialized type of nuclear data evaluation that can require significant, manual effort and months of time from expert scientists. In this article, non-convex, non-linear optimization methods are combined with concepts of inferential statistics to infer a set of optimized resonance models from experimental data in an automated manner that is not dependent on prior evaluation(s). This methodology aims to enhance the workflow of a resonance evaluator by minimizing time, effort, and prior biases while improving reproducibility and document-ability, addressing widely recognized challenges in the field.