You are here:
MODEL EVALUATION SCIENCE TO MEET TODAY'S QUALITY ASSURANCE REQUIREMENTS FOR REGULATORY USE: ADDRESSING UNCERTAINTY, SENSITIVITY, AND PARAMETERIZATION
Citation:
BABENDREIER, J. E., K. L. WOLFE, R. S. PARMAR, AND G. F. LANIAK. MODEL EVALUATION SCIENCE TO MEET TODAY'S QUALITY ASSURANCE REQUIREMENTS FOR REGULATORY USE: ADDRESSING UNCERTAINTY, SENSITIVITY, AND PARAMETERIZATION. Presented at Board of Scientific Counselors Land Goal Research Program Review, Cincinnati, OH, December 13 - 15, 2005.
Impact/Purpose:
The primary goals are to: (1) Construct a 400-node PC-based supercomputing cluster supporting Windows and Linux computer operating systems (i.e. SuperMUSE: Supercomputer for Model Uncertainty and Sensitivity Evaluation); (2) Develop platform-independent system software for the management of SuperMUSE and parallelization of EPA models and modeling systems for implementation on SuperMUSE (and other PC-based clusters); (3) Conduct uncertainty and sensitivity analyses of the 3MRA modeling system; (4) Develop advanced algorithmic software for advanced statistical sampling methods, and screening, localized, and global sensitivity analyses; and (5) Provide customer-oriented model applications for probabilistic risk assessment supporting quality assurance in multimedia decision-making.
Description:
The EPA/ORD National Exposure Research Lab's (NERL) UA/SA/PE research program addresses both tactical and strategic needs in direct support of ORD's client base. The design represents an integrated approach in achieving the highest levels of quality assurance in environmental decision-making, with technology development focuses on: (a) methods for holistic compositional and performance based model evaluation; (b) tools for uncertainty analysis and parameter estimation; (c) tools for conducting screening, local, and global-based sensitivity analyses; (d) tools for communicating uncertainty in outcomes and how best to reduce it; (e) methods and tools for addressing the computational burdens of quality assurance; (f) model evaluation demonstrations supporting selected research applications; and (g) national and international outreach efforts (leveraging collaborations and tools outside of EPA). This research is enhancing multimedia modeling and performance-based assessments underpinning major Agency initiatives. In the future, model evaluation will be a requisite decision-making component needed to analyze performance in achieving environmental benefits. Improving the overall efficiency of regulatory programs, this research is directly supporting multiple Program Office and ORD missions in allowing EPA to improve models and regulatory programs through better characterization of model sensitivities and uncertainties. Results from complementary approaches being undertaken to enhance model evaluation science and the Agency's capabilities will help EPA form the needed technical basis to: 1) better drive strategic research planning by identifying needs for improved science and data; 2) more accurately identify uncertainty in exposure and risk assessments of solid and hazardous waste management facilities; and 3) directly support the Office of Solid Waste and Emergency Response's (OSWER) Resource Conservation Challenge and One Cleanup Initiatives.