Spatial Modeling of PM 10 and NO 2 in the Continental United States, 1985–2000

• DOI: 10.1289/ehp.0900840
• Publication Date: 2009-06-29T17:52:11Z
• Authors: Jaime E. Hart; Jeff D. Yanosky; Robin C. Puett; Louise Ryan; Douglas W. Dockery; Thomas J. Smith; Eric Garshick; Francine Laden
• Keywords: Air Pollutants; Inhalation Exposure; Models, Statistical; Time Factors; Research; Nitrogen Dioxide; 15. Life on land; 01 natural sciences; United States; 03 medical and health sciences; 0302 clinical medicine; 13. Climate action; Air Pollution; Epidemiological Monitoring; 11. Sustainability; Humans; Regression Analysis; Particulate Matter; Particle Size; Environmental Monitoring; Power Plants; Retrospective Studies; Vehicle Emissions; 0105 earth and related environmental sciences

BackgroundEpidemiologic studies of air pollution have demonstrated a link between long-term exposures and mortality. However, many been limited to city-specific average measures or spatial land-use regression exposure models in small geographic areas.ObjectivesOur objective was develop nationwide annual particulate matter < 10 μm diameter (PM10) nitrogen dioxide during 1985–2000.MethodsWe used generalized additive (GAMs) predict levels the pollutants using smooth surfaces available monitoring data information system–derived covariates. Model performance determined cross-validation (CV) procedure with 10% data. We also compared results these commonly interpolation, inverse distance weighting.ResultsFor PM10, road, elevation, proportion low-intensity residential, high-intensity industrial, commercial, transportation land use within 1 km were all statistically significant predictors measured PM10 (model R2 = 0.49, CV 0.55). Distance population density, use, emissions nearest oxides–emitting power plant NO2 0.88, 0.90). The GAMs performed better overall than models, higher precision.ConclusionsThese provide reasonably accurate unbiased estimates for NO2. This approach provides temporal variability necessary describe assessing health effects chronic pollution.