The following HTML text is provided to enhance online
readability. Many aspects of typography translate only awkwardly to HTML.
Please use the page image
as the authoritative form to ensure accuracy.
Models in Environmental Regulatory Decision Making
This report looks specifically at the use of computational models in environmental regulatory activities, particularly at EPA. The use of computational models is central to the regulatory decision-making process because the agency must do prospective analyses of its policies, including estimating possible future effects on the environment, human health, and the economy. Obtaining a comprehensive set of measurement data is not feasible in many cases because of time and resource constraints. The agency uses models to generate estimates (or predictions) when data are not available. EPA also uses models to analyze measurement data for trends and effects. The results of models can become the basis for such decisions as initiating environmental cleanup or regulation. In sum, models are critical tools that help to inform and set priorities in environmental policy development, implementation, and evaluation at EPA.
Because of the critical role played by models, EPA has developed a variety of policies and programs to improve models and their use at the agency. One laudable step has been the establishment of the Council for Regulatory Environmental Modeling (CREM) in 2000 to support modeling activities across the agency and to provide an important resource for interested parties outside of EPA.
The National Research Council (NRC) convened the Committee on Models in the Regulatory Decision Process in response to a request from CREM to independently assess evolving scientific and technical issues related to the selection and use of computational and statistical models in decision-making processes at EPA. The full charge is provided in Box S-1 at the end of the Summary.
MODEL USE IN THE REGULATORY PROCESS AT EPA
Models will always be constrained by computational limitations, assumptions, and knowledge gaps. They can best be viewed as tools to help inform decisions rather than as machines to generate truth or make decisions. Scientific advances will never make it possible to build a perfect model that accounts for every aspect of reality or to prove that a given model is correct in all respects for a particular regulatory application. These characteristics make evaluation of a regulatory model more complex than solely a comparison of measurement data with model results. They suggest that model evaluation be viewed as an integral and ongoing part of the life cycle of a model, from problem formulation and model conceptualization to the development and application of a compu-