AAAR 34th Annual Conference
October 12 - October 16, 2015
Hyatt Regency
Minneapolis, Minnesota, USA
Abstract View
Characterization of Local Particulate Matter Concentration Gradients Using Mobile Platform and Fixed-site Monitors and Comparison with R-LINE and CMAQ Air Quality Models
XINXIN ZHAI, James Mulholland, Armistead G. Russell, Yongtao Hu, Timothy Larson, Elena Austin, Christopher Simpson, Timothy Gould, Kris Hartin, Sasakura Miyoko, Mike Yost, Georgia Institute of Technology
Abstract Number: 511 Working Group: Aerosol Exposure
Abstract Spatial gradients in ambient pollutant concentrations affect human exposure, and the strength of those gradients vary by source type. For example, pollutant concentrations decrease rapidly from heavily trafficked roads. Both measurements and simulations can be used to estimate local scale gradients. As part of a study to better characterize exposure fields in Atlanta, detailed monitoring and modeling were used to characterize spatial gradients in pollutant fields to assess the accuracy of various modeling approaches and the representativeness of monitors for use in health studies. Fixed site and mobile monitors were used to measure a number of pollutants, including PM2.5 species and mass, ozone and NOx, from September 7 to September 17, 2013. We compared observations with simulation results developed using the Research LINE source model (R-Line) at 200-m resolution and simulations from the chemical transport model CMAQ at 4-km resolution. Comparison between the measurements of monitors and the mobile platform surrounding the monitors were in good agreement, indicating relatively low spatial gradients within the area. R-LINE simulations yield larger spatial gradients and temporal variations than measurements, due, in part, to the sharp diurnal trends and spatial changes in the emissions. This led to a low correlation between the fine scale R-LINE results (R<0.2). Spatial gradients derived from CMAQ were lower, reflecting the coarser resolution. Our combined model and measurement analysis found that the routine (e.g., CSN and SEARCH) monitors in Atlanta are generally representative of the areas surrounding the monitoring sites. Comparison of the fine resolution simulations with chemical mass balance model results and the PM2.5 monitors in the region suggest that the R-LINE results need to be re-scaled.