Ambient Measurements\nof Hazardous Air Pollutants in\nthe United States Routinely Exceed Predictions from Screening-Level\nExposure Models

Hazardous air pollutant\n(HAP) emission regulations in\nthe US often\nrely on modeled estimates of ambient exposures. Model accuracy compared\nto real-world measurements of HAPs is crucial for understanding and\nmitigating exposure and associated health harms. While previous work\nshows ambient measurements are higher than regulatory model estimates,\nthe implications for health risk assessments are rarely discussed.\nWe provide a comprehensive comparison of the modeled and measured\nconcentrations at 489 US monitoring sites for 79 HAPs. We quantify\nhow model-measurement discrepancies affect the estimation of the exposure\nand risk of adverse health effects. Measurements were higher than\nmodeled concentrations in 74% of comparisons over all monitors, chemicals,\nand years assessed, with measurements a median 2 (IQR 1–9)\ntimes higher than model estimates. Measurements exceeded noncancer\nadverse health effect thresholds, while the model did not (model false\nnegatives) for nine pollutants. Adjusting for model bias in two industrial\ncenters, we found the number of people with multipollutant exposure\nabove the US EPA’s acceptable excess lifetime cancer risk increased\nby a factor of 30 times in Houston, Texas, and 13 times in Baton Rouge,\nLouisiana. Our results imply that assessments relying exclusively\non models like those we evaluated likely underestimate the spatial\nextent and magnitude of health hazards and risk.