Characterizing Uncertainty in Air Pollution Damage Estimates

This study uses Monte Carlo methods to characterize the uncertainty associated with per-ton damage estimates for 100 power plants in the contiguous United States (U.S.) This analysis focuses on damage estimates produced by an Integrated Assessment Model (IAM) for emissions of two local air pollutants: sulfur dioxide (SO2) and .ne particulate matter (PM2:5). For each power plant, the Monte Carlo procedure yields an empirical distribution for the damage per ton of SO2 and PM2:5:For a power plant in New York, one ton of SO2 produces $5,160 in damages with a 90% percentile interval between $1,000 and $14,090. A ton of PM2:5 emitted from the same facility causes $17,790 worth of damages with a 90% percentile interval of $3,780 and $47,930. Results for the sample of 100 fossil-fuel .red power plants shows a strong spatial pattern in the marginal damage distributions. The degree of variability increases by plant location from east to west. This result highlights the importance of capturing uncertainty in air quality modeling in the empirical marginal damage distributions. Further, by isolating uncertainty at each module in the IAM we .nd that uncertainty associated with the dose-response parameter, which captures the in.uence of exposure to PM2:5 on adult mortality rates, the mortality valuation parameter, and the air quality model exert the greatest in.uence on cumulative uncertainty. The paper also demonstrates how the marginal damage distributions may be used to guide regulators in the design of more efficient market-based air pollution policy in the U.S.