10th International Aerosol Conference September 2 - September 7, 2018 America's Center Convention Complex St. Louis, Missouri, USA
Abstract View
Early Lessons from New Air Pollution Exposure Science: High-resolution Mapping of Urban Air Quality using Google Street View Cars, Low-cost Samplers, and Aerosol Mass Spectrometry
JOSHUA APTE, Kyle Messier, Sarah Chambliss, Michael Brauer, Julien Caubel, Shahzad Gani, Steven Hamburg, Thomas W. Kirchstetter, Julian Marshall, Brian LaFranchi, Melissa M. Lunden, Chelsea V. Preble, Albert A. Presto, Christopher Portier, Allen Robinson, Ellis Shipley Robinson, Rishabh Shah, Karin Tuxen-Bettman, Roel Vermeulen, Ramon Alvarez, University of Texas at Austin
Abstract Number: 1115 Working Group: Aerosol Exposure
Abstract Human exposures to air pollution can vary sharply in space and time. Recent advances in measurement technology enable a rich understanding of the spatiotemporal patterns of outdoor air pollution that influence population exposures and environmental inequities. For example, carefully designed mobile sampling campaigns are now able to reveal patterns of long-term ambient air pollution concentrations at very fine scales (<< 100 m).
Here, we present an overview of the most extensive mobile air pollution measurement campaign conducted to date in a single urban area. Using two specially equipped Google Street View cars, we mapped spatial patterns of air quality in the San Francisco Bay Area between May 2015 and December 2017. The cars were outfitted with reference-grade instruments to measure NO, NO2, black carbon (BC) and ultrafine particle number count at high time resolution (~ 1 Hz). During a 30-month campaign, measurements took place on most weekdays during daytime hours, resulting in a large dataset: ~10 million samples collected during > 4000 h, encompassing 100,000 km of driving. Additional sampling incorporated a dense network of low-cost samplers and a mobile aerosol mass spectrometer, as described below. The overall campaign consisted of several targeted investigations.
During an initial year of already-published measurements, we sampled every road 20-50× within three neighborhoods (~30 km2) in Oakland, CA. Next, during 1.5 years of follow-up measurements, we mapped pollutant concentrations in rural, suburban, and dense urban neighborhoods throughout the SF Bay Area. These measurements reveal how the within-neighborhood pollution structure is overlaid on top of regional spatial gradients in air quality. Repeated measurements over the full 2.5-year period demonstrate persistent spatial variability over time. Patterns of NO, NO2 and BC in Oakland had high correlation (r2>0.85) between the first and second years of measurements. Shorter-term measurement periods (~1-2 months) were generally sufficient to reproduce overall spatial patterns, albeit with ±30% bias in mean concentrations relative to annual-average conditions.
We conducted an especially intensive multi-method study of the mixed residential/industrial West Oakland neighborhood during Summer 2017. Street View cars drove every street in this 5 km2 domain during days, nights, and weekends to characterize the temporal evolution of spatial patterns in air quality. In parallel, a dense network of 100 BC monitors was deployed at fixed sites to characterize temporal patterns at fine spatial scales. Finally, a van with a high-resolution aerosol mass spectrometer (HR-ToF-AMS) repeatedly sampled all city streets in the same domain to collect data on non-refractory PM1 chemical composition. We used positive matrix factorization of AMS data to identify source signatures that contribute to spatial patterns of air quality. Fixed and mobile measurements indicated that BC, NO, and other primary species were elevated on weekdays relative to weekends, with distinct shifts evident in spatial patterns within neighborhoods. UFP concentrations were less spatially variable, and exhibited a pronounced mid-day peak with only small weekday-weekend differences. Mobile AMS data demonstrate a temporal evolution of source and background contributions over the course of each day and spatial hotspots with elevated markers of primary emissions.
This campaign overview presentation (i) summarizes how routine mobile air pollution monitoring can reveal new information about spatial variability in population exposure to air quality, (ii) explores the methodological advantages and tradeoffs of alternative sampling approaches, and (iii) illustrates how established measurement techniques from the atmospheric sciences can add powerful new insights to studies of human exposures in urban areas.