AAAR 37th Annual Conference October 14 - October 18, 2019 Oregon Convention Center Portland, Oregon, USA
Abstract View
High Resolution (1 km) Chemical Transport Modeling of Fine Particulate Matter in an Urban Area
PABLO GARCIA, Shayak Sengupta, Spyros Pandis, Peter Adams, Carnegie Mellon University
Abstract Number: 756 Working Group: Urban Aerosols
Abstract We undertake high resolution, 1 km chemical transport modeling of fine (Dp < 2.5 µm) particulate matter (PM2.5) in Pittsburgh during winter of 2017 to quantify intra-urban and urban-rural gradients in PM2.5 concentrations and composition. We use PMCAMx to simulate PM2.5 treating aerosols across 10 particle size bins and organic aerosols with an additional 10 volatility bins. Traffic and restaurant activity significantly vary at the urban scale. Consequently, model inputs include new high resolution emissions inventories for traffic and cooking PM sources combining 2011 U.S. EPA National Emissions Inventory (NEI) estimates with a traffic model and restaurant locations, respectively. We evaluate model predictions with an aerosol mass spectrometer, an intra-urban network of Real-time, Affordable, Multi-Pollutant (RAMP) monitors and regulatory monitors from the EPA-CSN network. Average PM2.5 predictions agree with RAMP observations with a mean fractional bias (MFB) of 13% and a mean fractional error (MFE) of 16% highlighting the ability of the models to resolve urban scale gradients. EPA-CSN measurements show an urban-rural gradient of 3.39 µg m-3 that agrees with the 3.56 µg m-3 predicted by PMCAMx, this highlights the ability of the model at reproducing regional concentration fields. We predict an organic aerosol (OA) fraction of the total PM of 54% compared with an observed fraction of 38%. Of this fraction the model predicts 98% to be primary, with an observed fraction of at least 81%. This suggest that OA emissions, including cooking, are over-estimated in the inventory.