[Back]

H. Leeb, G. Schnabel, T. Srdinko:
"What is the proper evaluation method: some basic considerations";
in: "Proceedings of Workshop on Nuclear Measurements", IAEA/INIS, 2014, 231.

Recent developments and applications demand for an extension of the energy range and the inclusion of reliable uncertainty information in nuclear data libraries. Due to the scarcity of neutron-induced reaction data beyond 20 MeV the extension of the energy range up to at least 150 MeV is not trivial because the corresponding nuclear data evaluations depend heavily on nuclear models and proper evaluation methods are still under discussion. Restricting to evaluation techniques based on Bayesian statistics the influence of the a priori knowledge on the final result of the evaluation is considered. The study clearly indicates the need to account properly for the deficiencies of the nuclear model. Concerning the covariance matrices it is argued that they depend not only on the model, but also on the method of generation and an additional consent is required for the comparison of different evaluations of the same data sets. (authors)

https://publik.tuwien.ac.at/files/publik_232599.pdf