Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
1834516 | Nuclear Data Sheets | 2014 | 4 Pages |
In order to provide realistic evaluated uncertainties, well-prepared prior information but also adequate quantification of experimental uncertainties and their correlations are needed. In contemporary evaluations, the latter are often estimated roughly or neglected altogether — especially when it comes to correlations of uncertainties between experiments. In this contribution, a statistical analysis of 237Np(n,f) cross section data for 0.5–200 MeV is presented where special care was taken to estimate realistic correlations of uncertainties of the same and between different experiments. It is shown that taking correlations between experiments into account leads to larger evaluated uncertainties in the present example. A program to analyze experimental data and their uncertainties and correlations in a more reproducible and systematic manner is under development at LANL.