Regulatory models and the environment: practice, pitfalls, and prospects.

Area of Science:

• Environmental Science
• Computational Modeling
• Regulatory Science

Background:

• Computational models are crucial for environmental regulation, aiding knowledge evaluation, regulation assessment, and compliance. However, inherent uncertainties exist due to system complexity.
• The regulatory environment adds demands for accountability, transparency, public accessibility, and technical rigor to modeling challenges.

Purpose of the Study:

• To summarize findings and recommendations from the National Research Council's Committee on Regulatory Environmental Models.
• To provide guidance on the effective use of computational models in the U.S. Environmental Protection Agency's regulatory processes.

Main Methods:

• The committee's activities involved reviewing the challenges and best practices for regulatory environmental models.
• Focus on evaluating models not for absolute validation but for their suitability as tools for specific regulatory questions.
• Emphasis on a life-cycle approach to model evaluation, incorporating diverse peer review and uncertainty analysis methods.

Main Results:

• Models cannot be validated as 'true' but must be evaluated for fitness-for-purpose in regulatory contexts.
• Evaluating regulatory models is more complex than comparing data to results; it requires balancing accuracy with reproducibility, transparency, and utility.
• Standard validation methods are insufficient; regulatory models need specialized evaluation over their entire lifecycle.

Conclusions:

• Model evaluation must balance accuracy with reproducibility, transparency, and usefulness for specific regulatory decisions.
• A comprehensive life-cycle approach, including varied peer review and uncertainty analysis, is essential for regulatory models.
• The suitability of environmental models for regulation hinges on a holistic evaluation tailored to their intended use and the regulatory context.