skip to main content

Title: Confidence in Numerical Simulations

This PowerPoint presentation offers a high-level discussion of uncertainty, confidence and credibility in scientific Modeling and Simulation (M&S). It begins by briefly evoking M&S trends in computational physics and engineering. The first thrust of the discussion is to emphasize that the role of M&S in decision-making is either to support reasoning by similarity or to “forecast,” that is, make predictions about the future or extrapolate to settings or environments that cannot be tested experimentally. The second thrust is to explain that M&S-aided decision-making is an exercise in uncertainty management. The three broad classes of uncertainty in computational physics and engineering are variability and randomness, numerical uncertainty and model-form uncertainty. The last part of the discussion addresses how scientists “think.” This thought process parallels the scientific method where by a hypothesis is formulated, often accompanied by simplifying assumptions, then, physical experiments and numerical simulations are performed to confirm or reject the hypothesis. “Confidence” derives, not just from the levels of training and experience of analysts, but also from the rigor with which these assessments are performed, documented and peer-reviewed.
Authors:
 [1]
  1. Los Alamos National Lab. (LANL), Los Alamos, NM (United States)
Publication Date:
OSTI Identifier:
1170704
Report Number(s):
LA-UR--15-21345
TRN: US1500375
DOE Contract Number:
AC52-06NA25396
Resource Type:
Technical Report
Research Org:
Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)
Sponsoring Org:
USDOE National Nuclear Security Administration (NNSA)
Country of Publication:
United States
Language:
English
Subject:
97 MATHEMATICS AND COMPUTING; COMPUTERIZED SIMULATION; DECISION MAKING; HYPOTHESIS; VERIFICATION; FORECASTING; RANDOMNESS; ACCURACY