American Association for Aerosol Research - Abstract Submission

AAAR 38th Annual Conference
October 5 - October 9, 2020

Virtual Conference

Abstract View


Source-Resolved Primary Organic Aerosol Exposure Estimates in the United States

Provat Saha, Ellis Shipley Robinson, Wenwen Zhang, Steven Hankey, Julian Marshall, ALLEN ROBINSON, Albert A. Presto, Carnegie Mellon University

     Abstract Number: 253
     Working Group: Aerosol Exposure

Abstract
Primary organic aerosol (POA) has a significant contribution to fine particulate matter concentration at urban scales. High-resolution aerosol mass spectrometer (HR-AMS) data and Positive Matrix Factorization analysis offer source-resolved speciation of organic aerosol, which has been extensively applied in atmospheric field studies. We bring this technique in exposure research. We develop traffic and cooking POA exposure estimates through highly spatially resolved HR-AMS mobile sampling in three North American cities (Oakland, Pittsburgh, Baltimore) and empirical land-use regression modeling. As expected, we find both cooking and traffic POA concentrations vary substantially within and between cities. Within-city spatial variabilities are (4-6x) greater than between-city spatial variabilities (2x). Surprisingly, cooking POA concentration is higher than traffic POA concentration in each city. In land-use regression model development, traffic-related land-use variables (road density, transportation land use) explain the spatial variably of traffic POA, and cooking related variables (restaurant density, commercial land use) explain the spatial variability of cooking POA. External validation of model estimates against data from different atmospheric field studies across the US shows good agreement. This indicates the extrapolation of models over wider geographic areas is possible. We apply our models to predict national concentration surfaces of cooking and traffic POA in the US at the census block level. We use these estimates to assess the POA exposure inequality by race, income, and region.