Trends of Present-Day On-Road Motor Vehicle Particulate Matter Emissions to Total PM Using a Low-Cost Sensor Network in Baltimore City

COLBY BUEHLER, Misti Zamora, Daniela Flores, Marley Macarewich, Lei Hao, Abhi Datta, Kirsten Koehler, Drew Gentner, Yale University

     Abstract Number: 492
     Working Group: Aerosols Spanning Spatial Scales: Measurement Networks to Models and Satellites

Abstract
Historically, urban particulate matter concentrations (PM_2.5) have been driven by combustion-related sources, including mobile sources. In recent decades, emissions from motor vehicles have decreased substantially due to a number of factors, including improvements in emissions control technology, fuel composition, and the retirement of older high-emitting vehicles. This decrease in emissions inherently leads to larger relative contributions from other local urban sources (e.g., restaurants) or from transported pollution, which can differ spatially and over time. These changes have implications for human exposure to pollutants and associated health impacts.

Here, we evaluate the relative role of near-roadway PM_2.5 emissions (and other combustion-related pollutants) on concentration enhancements using high spatiotemporal data from the Solutions for Energy, AiR, Climate, and Health (SEARCH) low-cost air quality network (N=45) in Baltimore City, MD. Each monitoring site in the SEARCH network and nearby regulatory reference monitors are characterized using various land use metrics, traffic data, and roadway datasets. This is used to estimate the influence of different types of roadways (e.g., interstates, freeways, local roads, etc.) on spatiotemporal variations in concentrations over the course of multiple years. In the urban core of Baltimore City, annual average PM_2.5 concentrations have decreased from over 15 μg m^-3 in the year 2000 to less than 10 μg m^-3 in recent years while maintaining similar average annual daily traffic volumes. While overall long-term concentrations are low, significant site-to-site differences (e.g., >5 μg m^-3) relative to the network average were observed on time scales up to a month, with even greater differences observed at hourly time scales. Using a combination of meteorological data, reference monitoring stations, and SEARCH network sites that are upwind of the urban core, we estimate regional background PM_2.5 concentrations to determine PM_2.5 enhancements attributable to locally-emitted sources.