1. Yang Liu +, Ralph Kahn #, Solene Turquety *, Robert M Yantosca ++, and Petros Koutrakis + + Harvard School of Public Health, Boston, MA; # Jet Propulsion Laboratory, Pasadena, CA * Universite Pierre et Marie Curie, Paris, France; ++ Harvard Division of Engineering and Applied Sciences, Cambridge, MA The health effects studies of chronic exposure to PM2.5, which have played an essential role in shaping EPA’s air quality standards, are difficult to conduct partially due to the lack of reliable PM2.5 mass exposure estimates. Studying the health effects of specific PM2.5 constituents is more difficult because of data scarcity. We develop a novel method to use the fractional aerosol optical depth (AOD) values derived from the Multi-angle Imaging SpectroRadiometer (MISR) aerosol data, along with aerosol transport model constraints, to estimate PM2.5 mass concentrations and its major constituents. Using 2005 data in the United States, we conducted a regression study with fractional AOD as major predictors. When assessed by the available EPA ground truth, these models show substantially improved predictive power as compared to similar models using total-column AOD as a single predictor. These encouraging results suggest that this method may provide a new data source with much more comprehensive spatial coverage to support the health effects studies of PM2.5 constituents. A Novel Method to Estimate PM2.5 Constituent Concentrations and Size Distributions Using Satellite Retrieved Fractional Aerosol Optical Depth Harvard School of Public Health Methods Data Discussion Abstract 1.MISR Aerosol Properties [Kahn et al., 2005] Description: eight lognormally distributed particle components ranging from 0.06 to 2.0 μm in characteristic diameter and a climatology of 74 aerosol mixtures each containing up to three of the eight components. Parameters: AOD for each of the 74 mixtures at 17.6 x 17.6 km spatial resolution. Complete coverage in about a week. 2.GEOS-CHEM Aerosol Simulation [Park et al., 2004] Description: global 3D CTM driven by NASA assimilated meteorology. Parameters: vertical profiles of SO4, NO3, OC, EC and dust particles at 2°x 2.5° horizontal resolution. Global coverage with 3-hr time intervals. 3.EPA STN PM2.5 Speciation Data Description: ~ 130 sites in the Speciation and Trend Network (STN) established to study the spatial characteristics of PM2.5 constituents. Parameters: 24-hr SO4, NO3, OC and EC concentrations at ground level. Spatial distribution of EPA STN sites in the US. The coarse resolution of GEOS-CHEM simulation result is less likely to introduce substantial data redundancy given the relatively sparse distribution of STN sites. The two AERONET sites are used to compare with the estimated particle size distribution. 1. Calculation of MISR fractional AODs 2.Calculation of lower-air MISR fractional AODs using GEOS-CHEM data 3.Statistical Model linking PM2.5 with MISR fractional AODs Results Scatterplots of fitted PM2.5 mass and SO4 concentrations vs. EPA observations. Solid lines represent simple linear regression results with intercepts excluded. Final data used to derive the regression coefficients are displayed as small black squares and outliers are displayed as large hollow squares. The adjusted R2 is 0.56 for PM2.5 in both the east and the west, 0.62 for SO4 in the east, and 0.40 for SO4 in the west. The PM2.5 models underpredict PM2.5 by approximately 7- 8% in both the east and the west. The SO4 model underestimates by approximately 10% in the east and approximately 17% in the west. The mean percentage contribution of each significant component to the concentrations of a PM2.5 constituent can be used to estimate particle size distributions. Plots to the left are estimated annual average size distribution of PM2.5 and SO4 in the eastern US (upper panel), and PM2.5, SO4, and NO3 in the western US (lower panel). [Liu et al, 2006a] [Liu et al, 2006b] Estimated size distributions capture the overall features of PM2.5, SO4 and NO3 reported in ground measurements. However, estimated peak sizes are smaller [Koutrakis et al., 1993]. 1.The fractional AOD method shows much higher predicting powers than previous models using AOD as a single predictor because it is able to adjust for the different compositions of fine particles in the east from the west. In addition, they are flexible enough to allow better relative humidity correction based on the hygroscopicity of individual aerosol components. 2.In this case study, we are able to predict PM2.5 and SO4 concentrations in both the east and west, and NO3 concentrations in the west reasonably well, compared with available EPA ground-truth. 3.The discrepancy between estimated and observed size distributions is likely due to the satellite sampling bias caused by the inability to retrieve aerosols through cloud cover, and the impact of particle hygroscopicity on measured particle size distributions at ground level. 4.As more mature MISR particle property products and longer data time series are now becoming available, this method has the potential to provide valuable information on the spatial patterns of OC and EC 1.Kahn, R. A.; Gaitley, B. J.; Martonchik, J. V.; Diner, D. J.; Crean, K. A.Holben, B.; Multiangle Imaging Spectroradiometer (MISR) global aerosol optical depth validation based on 2 years of coincident Aerosol Robotic Network (AERONET) observations; J. Geophys. Res.-Atmos. 2005, 110, Article No. D10S04. 2.Park, R. J.; Jacob, D. J.; Field, B. D.; Yantosca, R. M.Chin, M.; Natural and transboundary pollution influences on sulfate-nitrate- ammonium aerosols in the United States: Implications for policy; J. Geophys. Res.-Atmos. 2004, 109(D15):Art. No. D15204. 3.Y. Liu, R. Kahn, and P. Koutrakis. Estimating PM2.5 Component Concentrations and Size Distributions Using Satellite Retrieved Fractional Aerosol Optical Depth: Part I - Method Development. Submitted to Journal of the Air & Waste Management Association. December 2006. 4.Y. Liu, R. Kahn, S. Turquety, R.M. Yantosca, and P. Koutrakis. Estimating PM2.5 Component Concentrations and Size Distributions Using Satellite Retrieved Fractional Aerosol Optical Depth: Part II – A Case Study. Submitted to Journal of the Air & Waste Management Association. December 2006. 5.Koutrakis, P.Kelly, B. P.; Equilibrium Size of Atmospheric Aerosol Sulfates as a Function of Particle Acidity and Ambient Relative Humidity; J. Geophys. Res.-Atmos. 1993, 98, 7141-7147. References This study is supported by Harvard-EPA Center on Particle Health Effects (R827353-01-0). The work of R. Kahn is supported by NASA's Climate and Radiation Research and Analysis Program and the EOS-MISR instrument project. We thank Dr. Lyatt Jeagle of University of Washington and Rynda Hudman of Harvard University for providing access to the GEOS-CHEM simulation results.