This discussion paper applies bounding analysis to environmental risk analysis as suggested in Morgan (2001) "The neglected art of bounding analysis" Environ Sci Technol. 35(7):162A-164A. The goal is to bound the fraction of lung cancer occurrences not attributed to specific well-studied causes, in order to keep estimates of the less well delimited risks consistent with known risks. Available data and expert judgment are used to attribute a portion of the observed lung cancer cases to known causes such as smoking, residential radon and asbestos exposure, to describe the uncertainty surrounding these estimates, and quantify the interaction between pollutants. Then an upper bound on the risk related to diesel particulates is inferred using a coherence constraint on the total number of deaths and a principle of maximum plausibility, a concept from the field of imprecise probabilities.
Keywords. lung cancer, risk, transferable belief model, environmental pollutant
Authors addresses:
Center for Integrated Study of Human Dimensions of Global Change
Department of Engineering and Public Policy.
Carnegie Mellon University
Pittsburgh, Pennsylvania 15213-3890
USA
E-mail addresses:
| Minh Ha Duong | minh.ha.duong@cmu.edu |
Related Web Sites