A206-02
Green Heart Louisville: Community-Level Assessment of Exposure to Air Pollution

Tuesday, 15 December 2020: 20:37
Virtual
Pradeep S Prathibha, Washington University in St Louis, St. Louis, MO, United States, Eben S Cross, QuantAQ Inc., Somerville, MA, United States, Richard L Strehl, University of Louisville, Louisville, KY, United States, Ray A Yeager II, University of Louisville, Environmental and Occupational Health Sciences, Louisville, KY, United States, Aruni Bhatnagar, University of Louisville, Environmental Health Institute, Louisville, KY, United States and Jay Robert Turner, Washington University in St Louis, Department of Energy, Environmental and Chemical Engineering, St. Louis, MO, United States
Recorded Presentation
Abstract:
Green Heart Louisville (greenheartlouisville.com) is a prospective cohort study examining the linkages between urban vegetation and cardiovascular health, an association potentially mediated by the effect of vegetation on local air quality. We monitor long-term air quality and greenness in a 12 km2 study area in Louisville, KY, in tandem with a clinical study assessing cardiovascular function and risk factors in area residents. A 2.5 km2 contiguous section is the intervention area where mature vegetation is installed beginning in Fall 2019, followed by intensive planting in Fall 2020; the remaining area serves as the control.

To characterize the spatiotemporal variability in ambient air pollution, we employ (1) two-week integrated passive sampling of nitrogen oxides (NO, NO2) and ozone (O3) repeated every two months since May 2018 in a 60-site fixed network; (2) periodic monitoring (1-min) of PM2.5 using collocated Alphasense OPC-N2 (mounted on QuantAQ nodes) and PurpleAir (PA) sensors, at four fixed locations; and (3) recurring mobile-platform monitoring (1-s) of ultrafine particle (UFP) number concentrations throughout the study domain.

Annual average NO (2019) is 3-5х higher near major roadways and commercial facilities and 2-3х higher near church and school parking lots; corresponding O3 is depressed by NO titration. Log-transformed daily mean PM2.5 mass concentrations are well-correlated (R = 0.72, p ≪ 0.001), but the mean ratios of absolute PM2.5 reveal large bias (OPC-N2/PA = 0.19 ± 0.10). Contemporaneous PM2.5 from QuantAQ nodes are highly correlated (0.77 < r < 0.83) at locations with varying proximities to an interstate highway, an international airport, and commercial/industrial facilities. UFP measurements strongly depend on wind direction: for winds parallel to an interstate highway with a 10-ft sound wall, near-road (10-50 m) concentrations remain at background levels during rush hour traffic; under crosswinds, however, 3-5х elevated concentrations extend 750 m downwind of the sound wall even at night.

Collectively, these measurements indicate this approach can capture spatial variations within our study domain. Upcoming work will integrate hyperlocal land use regression models to estimate residential-level air pollution exposure before and after planting to inform the clinical study.

<< Previous Abstract | Next Abstract >>