Nancy M. Larson

Nancy Larson earned her PhD in computational physics from Michigan State University in 1972, and was a member the ORNL research staff until her retirement on January 31, 2008. She continues to oversee SAMMY-related research and development as a part-time consultant.

Larson's primary research activity has been the development of the powerful and widely-used computer code SAMMY, a tool for analysis of cross-section data in the resonance region. In addition to R-matrix calculations of cross sections in both the resolved and the unresolved resonance regions, this code includes corrections for experimental conditions (finite temperature, finite sample size, etc.), procedures for determining "best-fit" values of parameters, techniques for incorporating measurement uncertainty into the analyses, options for interfacing with the evaluated nuclear data files (ENDF), and many other features.

Documentation for the SAMMY code is now in its ninth revision. The eighth revision of the SAMMY users' manual, completed and released in pdf format in 2008, contains over 600 pages and covers virtually all aspects of the code. In addition, Larson is author or coauthor on nearly 200 publications and presentations, either describing the development of new techniques for data analysis or reporting on specific analyses for which she has served as consultant.

Education of scientists in techniques for analyzing cross section data has been a high priority in Larson's work. Week-long SAMMY workshops have been held approximately ten times at various sites around the world; these workshops include lectures on all aspects of data analysis in the resonance region, plus computer exercises to help a novice understand how to run the SAMMY code. Some of these lecture notes are available on this website.

Before her retirement, Dr. Larson was an active member of the Cross Section Evaluation Working Group (CSEWG), which is the advisory body for ENDF. She has authored and implemented new formats for the ENDF systems, and provided documentation for those formats for inclusion in the official ENDF formats manual. Larson was also involved in international cooperative research projects and other committees.

In addition to consulting with SAMMY users and supervising maintenance of the code and its documentation, Larson is also working to develop new techniques for data analysis. Current research is primarily focused on the generalized least squares fitting procedure, especially with regard to error propagation. Topics of interest include:

• Understanding how measurement uncertainties propagate through the analysis process;
• Development of methodologies to incorporate all relevant uncertainties into analyses;
• Learning what are the effects of small changes in methodology;
• Finding reliable means of comparing covariance matrices, especially those produced by two distinct analyses of the same data.