NERL Research Abstract EPA's National Exposure Research Laboratory GPRA Goal 5: Better Waste Management, Restoration of Contaminated Waste Sites and Emergency Response Significant Research Findings 2003 International Workshop on Uncertainty, Sensitivity, and Parameter Estimation for Multimedia Environmental Modeling Scientific Problem and While there is a high potential for exposure of humans and ecosystems to Policy Issues chemicals released from a single hazardous waste site, the degree to which this potential is realized is often uncertain. Conceptually divided among parameter, model, and modeler uncertainties imparted during simulation, inaccuracy in model predictions result principally from lack of knowledge and data. In comparison, sensitivity analysis can lead to a better understanding of how models respond to variation in their inputs, which in turn can be used to focus laboratory and field-based data collection efforts on processes and parameters that contribute most to uncertainty in outputs. In modeling both simple and complex systems, initially we describe uncertainty for the current state of science and data, then ascertain a prioritized agenda for its reduction via sensitivity analysis. Uncertainty analysis allows for the critical task of making informed wastestream management decisions in the present, and sensitivity analysis, ideally, drives the research planning process. For environmental regulation, these two elements, formulated as action and continued research investigation, represent encompassing statements describing the daily execution of EPA's primary mission to protect human health and the environment. In the context of using models to assist decision-making, it is a combined process deeply rooted in the fundamental engineering principle of cost-benefit analysis, and is paralleled, to a degree, by model evaluation tasking embodied by uncertainty analysis, sensitivity analysis, and parameter estimation. Multiplicity (An Operative Modeling Concept for the Future) As we rapidly push forward to integrate multi-media, multi-pathway, multi- receptor, multi-contaminant, and multi-scale risk assessments associated with waste disposal, we are invariably led to an increasingly complex problem statement and modeling paradigm. Complexity of the problem statement increases substantially in concurrently addressing risks to both human and ecological populations, and their associated subpopulations (e.g., "high end" sensitive receptors, etc.). Further compounding national management approaches for various hazardous wastestreams, national assessment strategies ------- - derived from multiple, site-based risk assessments, present even greater challenges in evaluating confidence in model-based forecasts of population protection. Due to their inherent abstraction, national management strategies also present increasing difficulty in communicating uncertainty in risk to both decision-makers and stakeholders. Research It is recognized across the Federal community that a better understanding of Approach how to conduct uncertainty analysis, sensitivity analysis, and parameter estimation, especially for complex, high-order systems, is paramount in supporting effective decision-making. This endeavor is indeed generic to modeling efforts spanning all of the various GPRA goals and associated programs. As a key element of advancing our understanding in the area of model evaluation, the NERL Ecosystems Research Division has been actively pursing the creation of, and EPA's participation within, a multi-Agency Uncertainty Workgroup covering technical areas of model evaluation. Inter-Agency Memorandum of Understanding (MOU) on Multimedia Environmental Modeling The Uncertainty Workgroup forms a key element of a larger effort now in place to facilitate overall coordination on multimedia modeling research within the Federal community. EPA research staff have met with counterparts from several Federal organizations, identified areas of common interests and needs, and developed a formal multi-agency Memorandum of Understanding (formally initiated in June, 2001) and an associated Federal Interagency Steering Committee on Multimedia Environmental Modeling (ISCMEM). The purpose of the MOU is to establish a framework for facilitating cooperation and coordination among eight signatory agencies in their research and development of multimedia models. A central objective is to provide a mechanism for these agencies to pursue common technologies in multimedia environmental modeling with a shared scientific basis. International Workshop on UA/SA/PE; Rockville MD, August 2003 A primary activity of the Uncertainty Workgroup is to coordinate ongoing and anticipated new research that focuses on uncertainty analysis strategies (UA), sensitivity analysis (SA) techniques, and parameter estimation (PE) methods that support multimedia models. As one of several activities conducted to date, the Uncertainty Workgroup recently organized and sponsored an international workshop on these three themes. The objective of the workshop was to facilitate communication among U.S. Federal Agencies conducting research on the workshop themes, obtain up-to-date information from invited technical experts, and to actively discuss opportunities and new approaches for parameter estimation, sensitivity and uncertainty analyses related to multimedia environmental modeling. ------- Results and The three-day workshop held in Rockville Maryland August 19-21, 2003 was Implications actively organized by seven Federal Agencies including U.S. EPA, the U.S. Nuclear Regulatory Commission, the U.S. Geological Survey, the National Oceanic & Atmospheric Administration, U.S. Department of Energy, the U.S. Department of Agriculture's Agricultural Research Service, and the U.S. Army Corps of Engineers. The workshop brought together over 36 technical speakers and 70 participants from six different countries (U.S., Britain, Canada, Italy, The Netherlands, Australia, and Denmark). For example, among the many members of the "dream team of experts" assembled for this workshop, key presentations were delivered by Andrea Saltelli, Bruce Beck, Keith Beven, Shlomo Neuman, Jon Helton, Ken Reckhow, John Doherty, Eileen Poeter, Christopher Frey, Terry Andres, and Jeroen van der Sluijs. These experts bring to the table a wide array of capabilities and philosophies on the technical themes of the workshop. Despite their high levels of expertise, all of the technical experts responded in some form after the workshop that the 3-day event was extremely illuminating in drawing out these differences, and resolutely defining the commonalities that various researchers share in advancing our knowledge on these subjects. The workshop had a far reach among participants. While too numerous to list individuals, or even all of the research groups they represented, the workshop was attended by a broadly representative set of members from the Federal research community who could benefit most from the technical nature of the workshop. For example, ORD, in sponsoring this event, used EPA's invitation allocation (19 seats), to invite many representatives from various Program and Regional Offices to maximize benefit for ORD's client-base, while simultaneously helping to support knowledge transfer to the CREM (Council for Regulatory Environmental Modeling) design team. This included representatives from the Regional offices, and representatives from the OW, OSWER, OPPTS, and OAR, the Chesapeake Bay and Potomac River Modeling Programs, ORD's Office of Science Policy, and several members from various Divisions within ORD and NERL. Research Collaboration in addition to broad representation among the seven Federal Agencies who and Publications participated, a proceedings of the workshop will be jointly published in October 2004. There are already several instances of how the workshop has initiated essentially a spawning ground for cross-Office and cross-Agency collaboration regarding research into various aspects of uncertainty analysis, sensitivity analysis, and parameter estimation for multimedia modeling. Examples include further developing or maturing relationships between ORD and Program and Regional Offices, as well as research opportunities that will be initiated between EPA staff and staff from other Federal Agencies. Future Research A second workshop is tentatively planned to be held in 2004 in conjunction with a large scientific meeting sponsored by the Society for Risk Analysis. Questions and inquiries can be directed to: Justin Babendreier, Ph.D., P.E. U.S. EPA, Office of Research and Development National Exposure Research Laboratory Ecosystems Research Division ------- 960 College Station Road Phone: 706 355 8344 E-mail: babendreier.justin@epa.gov ------- |