Predicting Long-Range Transport:  A Systematic Evaluation of Two Multimedia Transport Models

• Published in: Environmental science & technology Vol. 35; no. 6; pp. 1181 - 1189
• Main Authors: Bennett, Deborah H, Scheringer, Martin, McKone, Thomas E, Hungerbühler, Konrad
• Format: Journal Article
• Language: English
• Published: Washington, DC American Chemical Society 15-03-2001
• Subjects: Applied sciences; Atmospheric pollution; Chemicals; Decision Making; Environmental Pollutants; Exact sciences and technology; Humans; International Cooperation; Models, Theoretical; Multimedia; Organic Chemicals; Particle Size; Pollution; Predictions; Public Policy; Reproducibility of Results; Uncertainty; Pollutant behavior; Decision making; Pollution prevention; Forecast model; Transport process; Persistent organic pollutant; Compartmental model; Long range pollutant transport; Air pollution; Phase partition; Decision criterion; Reliability; Organic compounds
• ISSN: 0013-936X; 1520-5851
• DOI: 10.1021/es001278u

The United Nations Environment Program has recently developed criteria to identify and restrict chemicals with a potential for persistence and long-range transport (persistent organic pollutants or POPs). There are many stakeholders involved, and the issues are not only scientific but also include social, economic, and political factors. This work focuses on one aspect of the POPs debate, the criteria for determining the potential for long-range transport (LRT). Our goal is to determine if current models are reliable enough to support decisions that classify a chemical based on the LRT potential. We examine the robustness of two multimedia fate models for determining the relative ranking and absolute spatial range of various chemicals in the environment. We also consider the effect of parameter uncertainties and the model uncertainty associated with the selection of an algorithm for gas-particle partitioning on the model results. Given the same chemical properties, both models give virtually the same ranking. However, when chemical parameter uncertainties and model uncertainties such as particle partitioning are considered, the spatial range distributions obtained for the individual chemicals overlap, preventing a distinct rank order. The absolute values obtained for the predicted spatial range or travel distance differ significantly between the two models for the uncertainties evaluated. We find that to evaluate a chemical when large and unresolved uncertainties exist, it is more informative to use two or more models and include multiple types of uncertainty. Model differences and uncertainties must be explicitly confronted to determine how the limitations of scientific knowledge impact predictions in the decision-making process.