Particulate matter under 10 microns (PM10) is an established health hazard, but exposure measurements often do not have adequate spatial and temporal resolution for public health and other applied science uses. This is especially true in developing countries such as China, despite that the needs for adequate air pollution characterization may be greater due to higher population impacts due to higher pollution levels and greater numbers of people. This project combines remotely-sensed aerosol optical depth (AOD) values obtained from MODIS satellite sensors to improve existing ground-based PM10 data sources. This project also uses other atmospheric variables, such as temperature and humidity, to adjust the effect these conditions may have on remotely-sensed AOD values. The project area covers Beijing municipality for the years 2008-2010. This project to produce a PM10 grid surface data product which can be used for future environmental research studies, and demonstrates the possibilities of future applications of this NASA algorithm for future incorporation of AOD into estimation of ground level PM. As compared to a surface generated solely by ground-based monitors, we expect the surface generated in this project to have better temporal resolution (once a day without missing values), spatial resolution (10 x 10 km regardless of distance from an urban population center), and lower error (as determined by mean RMSD statistics generated during error assessment). Maps will also be produced which can display and compare the generated grid surfaces. County-based mortality data has also been obtained to examine the utility of this new exposure measurement method. The mortality study will contain descriptive data results, including the covariate distributions by PM10 quartiles and chi-squared test p-values that determine any significant differences. Poisson regression models, with a natural spline to correct for seasonally varying variables, will be used to determine relationships between PM10 estimates and mortality. Estimates will be calculated for each of the cities and mixed models will be used to combine the 4 estimates. The regression model will contain relative risks with 95% confidence intervals which will display the strength of any relationships after accounting for various nuisance variables. The results of this analysis will be compared to the results of an analysis using the same data and statistical methods, but simply uses the nearest PM10 monitor to assess outdoor PM10 exposure. We anticipate the results of this project to be published in a peer-reviewed journal and to be used in future studies that will be publishable in public health and other science journals. This study will be the first publication using this innovative method of air pollution assessment in an applied science context. We will contribute to the growing interest using remotely-sensed data to improve air pollution and outdoor environmental factors.