Conservative covariance for general-purpose nuclear data evaluation

Tamagno, Pierre

doi:10.1140/epja/s10050-021-00367-1

Cited by 3 publications

(7 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A third method [7] combines a good mathematical definition and the possibility to let the ⃗ θ parameters evolve if necessary. This method is based on the preservation of conservative uncertainty propagation for any quantity varying (linearly) with parameters ⃗ x.…”

Section: Spectral Decomposition Methodsmentioning

confidence: 99%

“…The simple model (7) allows to easily read the value of x : the model value for E = 0. The value of θ can also be estimated as the slope of the curves.…”

Section: Toy Model Analysismentioning

confidence: 99%

“…Several techniques have been designed to transfer uncertainty from θ on x. The first method presented here that is called matrix marginalization but is also called analytical marginalization [4,6,7] or zero-variance penalty marginalization [5] in other works. This method attempts to find a matrix M m…”

Section: "Matrix Marginalization"mentioning

confidence: 99%

“…The second approach (the Bayesian marginalization method) is shown to be possibly biased, namely when nuisance parameters are inaccurate. These methods have been tested on a previous study on a simple but realistic analysis of Prompt Neutron Fission Spectrum [7]. These approaches are tested here on a specifically designed toy model test-case, where they are shown to fail in producing consistent best-estimates and reliable propagated uncertainties.…”

Section: Introductionmentioning

confidence: 99%

“…The failure highlights that the two presented techniques -focusing on the transfer of nuisance parameter uncertainty -neglect the verification of agreement between posterior theory and experiment. An ad hoc procedure [7] is then presented in the paper, it aims at artificially increasing the estimated uncertainties but in a conservative manner with respect to the Bayesian approach and with consideration of agreement between posterior theory and experimental. This new technique is further exemplified on resonant cross section analysis.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Marginalization methods for the production of conservative covariance on nuclear data

Tamagno

2023

EPJ Web Conf.

Self Cite

View full text Add to dashboard Cite

The production of evaluated nuclear data consists not only in the determination of best estimate values for the quantities of interest but also on the estimation of the related uncertainties and correlations. When nuclear data are evaluated with underlying nuclear reaction models, model parameters are expected to synthesize all the information that is extracted from the experimental data they are adjusted on. When dealing with models with a small number of parameters compared to the number of experimental data points – e.g. in resonant cross section analysis – one sometimes faces excessively small evaluated uncertainty compared for instance with model/experimental data agreement. To solve this issue, an attempt was to propagate the uncertainty coming from experimental parameters involved in the data reduction process on the nuclear physics model parameters. It pushed experimentalists to separately supply random (statistical) and systematic uncertainties. It also pushed evaluators to include or mimic the data reduction process in the evaluation. In this way experimental parameters – also called nuisance parameters – could be used to increase evaluated parameter uncertainty through marginalization techniques. Two of these methods: Matrix and Bayesian marginalizations – respectively called sometimes Analytical and Monte-Carlo Marginalizations – that are currently used for evaluation will be discussed here and some limitations highlighted. A third alternative method, also based on a Bayesian approach but using the spectral decomposition of the correlation matrix, is also presented on a toy model, and on a a simple case of resonant cross section analysis.

show abstract

Section: Spectral Decomposition Methodsmentioning

confidence: 99%

“…The simple model (7) allows to easily read the value of x : the model value for E = 0. The value of θ can also be estimated as the slope of the curves.…”

Section: Toy Model Analysismentioning

confidence: 99%

Section: "Matrix Marginalization"mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Marginalization methods for the production of conservative covariance on nuclear data

Tamagno

2023

EPJ Web Conf.

Self Cite

View full text Add to dashboard Cite

show abstract

Evaluation of the neutron-induced cross sections of actinides using the CONRAD code

et al. 2023

View full text Add to dashboard Cite

The CONRAD computer code is being developed by the nuclear data group of CEA Cadarache since mid-2000. It was originally designed to analyse neutron-induced reactions in the resonance energy range and then, was extended to higher energies (several MeV) treatment with inclusion of charged-particles penetration factor. In the thermal energy range, a procedure was implemented to manage the so-called Thermal Neutron Constants, especially devoted to the 239Pu, 241Pu, 233U and 235U nuclei. In the resonance range, nuclear models implemented in the CONRAD program rely on R-matrix model fits with in particular, improved treatment of the fission penetration factor and fluctuations of the prompt neutron multiplicity via a two-step (n,γf) process. Above the resonance range, the neutron continuum energy region of the observed cross sections can be analysed either with the TALYS code or with an in-house optical model code named CCCP that is followed by Hauser-Feshbach calculations according to the compound nucleus deexcitation channels. Non-model least squares fitting procedures have been also tested for neutron cross sections adjustment in the continuum energy range, and such, applied in the framework of the standard cross section group at the IAEA (Vienna). This paper will only focus on the evaluation works performed in the resonance range of minor and major actinides with special emphasis on the experimental corrections suitable to reproduce time-of-flight experiments.

show abstract

Research on Mining of Government Data Based on Enhanced-Object Exchange Model

Sa¹,

Wang³

2021

2021 IEEE 21st International Conference on Software Quality, Reliability and Security (QRS)

View full text Add to dashboard Cite

Conservative covariance for general-purpose nuclear data evaluation

Cited by 3 publications

References 21 publications

Marginalization methods for the production of conservative covariance on nuclear data

Marginalization methods for the production of conservative covariance on nuclear data

Evaluation of the neutron-induced cross sections of actinides using the CONRAD code

Research on Mining of Government Data Based on Enhanced-Object Exchange Model

Contact Info

Product

Resources

About