[en] Some methods and codes have been developed and used for covariance data evaluation of experimental data, they are parameter analysis, physical analysis, spline fitting etc... These methods and codes can be used in different cases

[en] Double differential nuclear cross-sections measurements (neutron energy-emission angle) usually rely on biparametric counting histograms. Alternatively the detector output may be processed on-line to produce the angular distribution as an expansion in Legendre polynomials of cosθ. The Legendre moments are accumulated in a set of single parameter histograms covering the neutron input energy range so they may be visualized during acquisition. Their covariances are accumulated simultaneously in another set of histograms. The multiple weighted counting method producing these data is presented and discussed in comparison with the classical off-line least-squares analysis of biparametric counting spectra

[en] These proceedings of a specialists'meeting on evaluation and processing of covariance data is divided into 4 parts bearing on: part 1- Needs for evaluated covariance data (2 Papers), part 2- generation of covariance data (15 Papers), part 3- Processing of covariance files (2 Papers), part 4-Experience in the use of evaluated covariance data (2 Papers)

[en] A neutron spectra adjustment procedure requires among others on its input the cross-sections for the reactions of interest with their covariance matrices in the user's energy group structure. The most recent data set, a test version of the International Reactor Dosimetry File (IRDF-90) of the Nuclear Data Section of the International Atomic Energy Agency was compiled in the ENDF-6 format. Computer codes have been developed in order to convert the uncertainty of the IRDF-90 library data into the user's energy grid. Experiences with two conversion codes and some data for this test version of the IRDF-90 library are given

[en] Modern nuclear data evaluation methods demand detailed uncertainty information for all input results to be considered. It can be shown from basic statistical principles that provision of a covariance matrix for a set of data provides the necessary information for its proper consideration in the context of other included experimental data and/or a priori representations of the physical parameters in question. This paper examines how an experimenter should go about preparing the covariance matrix for any single experimental data set he intends to report. The process involves detailed examination of the experimental procedures, identification of all error sources (both random and systematic), and consideration of any internal discrepancies. Some specific examples are given to illustrate the methods and principles involved

[en] Accurate data analysis requires understanding of the roles played by both data and parameter covariance matrices. In this paper the entire data reduction/analysis process is examined, for neutron-induced reactions in the resonance region. Interrelationships between data and parameter covariance matrices are examined and alternative reduction/analysis methods discussed

[en] A simplified method to estimate covariance matrix of the model parameters is presented in this work and applied to calculate covariance matrix of calculated cross sections on ¹⁹F

[en] The 'Evaluated Nuclear Data Files'(ENDF) contain - due to existing physical relations - large portions of data which are not independent. There exist either connections caused by common features of experimental arrangements or built in 'redundancies'caused by physical constraints due to relations between the cross sections contained in the files. Both facts can lead to the loss of consistency each time when new experimental information is added. Some important aspects concerning covariance information in this context, as well as general precautions in the 'design' of covariance matrices are summarized and discussed

[en] This paper presents a brief review of computer codes concerned with checking, plotting, processing and using of covariances of neutron cross-section data. It concentrates on those available from the computer code information centers of the United States and the OECD/Nuclear Energy Agency. Emphasis will be placed also on codes using covariances for specific applications such as uncertainty analysis, data adjustment and data consistency analysis. Recent evaluations contain neutron cross section covariance information for all isotopes of major importance for technological applications of nuclear energy. It is therefore important that the available software tools needed for taking advantage of this information are widely known as they permit the determination of better safety margins and allow the optimization of more economical designs of nuclear energy systems

[en] The ENDF File-30 format provides a means of recording completely, yet reasonably compactly, and type of nuclear data covariances that can be formally separated into (a) a faily small set of parameters with known covariances and (b) tabulated derivatives of the nuclear data with respect to the parameters. In this paper, we discuss the potential usefulness of File 30 for describing the covariances of emission spectra. In particular, it will be shown that, relative to older approaches such as 'SED/SAD', the greater fundamental clarity of the File-30 approach will facilitate the work of both the data evaluator and the data user. To place the topic of such factored covariances into perspective, we also briefly review relevant aspects of the generalized method of least squares and the new technology of computer calculus