HR: 1340h
AN: A33C-0924    [Abstracts]
TI: Study of the Spatial Variability of PM2.5 Concentrations Using Satellite Remote Sensing Data
AU: * Liu, Y
EM: yangliu@hsph.harvard.edu
AF: Harvard School of Public Health, 401 Park Drive, 4th floor west PO Box 15677, Boston, MA 02215
AU: Franklin, M
EM: mfrankli@hsph.harvard.edu
AF: Harvard School of Public Health, 401 Park Drive, 4th floor west PO Box 15677, Boston, MA 02215
AU: Koutrakis, P
EM: petros@hsph.harvard.edu
AF: Harvard School of Public Health, 401 Park Drive, 4th floor west PO Box 15677, Boston, MA 02215
AB: The capability of MISR and MODIS aerosol data in characterizing the spatial variability of ground level PM2.5 concentrations is studied in two steps in a 300 km by 300 km region surrounding St. Louis, Missouri. Two regression models are developed using NOAA meteorological fields and MISR and MODIS AOT, respectively, to predict ground level daily PM2.5 concentrations collected from 23 EPA monitoring stations in the region. Both MISR and MODIS AOT are highly significant predictors of PM2.5 concentrations. Overall, the regression model involving MISR AOT is able to explain approximately 60 - 63 percent of the variability in PM2.5 while the regression model involving MODIS AOT is able to explain approximately 52 - 55 percent of the variability. In addition to predicting PM2.5, we examine both the annual and seasonal spatial variability of EPA monitoring site concentrations and compare the spatial pattern seen on the ground to the spatial pattern of AOT measured by MISR and MODIS. The comparison is made by fitting separate variograms to the data and assessing their similarity. Spatial patterns in MISR and MODIS AOT that are comparable to the spatial pattern of ground level measurements indicate that satellite measurements are able to capture the true variability in pollution over the study region. Our study of the relationship of AOT to ground level concentrations of PM2.5 coupled with an understanding of the spatial distribution of AOT shows that MISR and MODIS measurements of AOT can be used as an important surrogate to extend measurements made at sparse ground PM2.5 monitoring stations to cover a large geographic region.
DE: 0305 Aerosols and particles (0345, 4801, 4906)
DE: 0345 Pollution: urban and regional (0305, 0478, 4251)
DE: 0480 Remote sensing
SC: Atmospheric Sciences [A]
MN: Fall Meeting 2005