Mobile monitoring of multi-pollutant mixtures on roadways: Characterizing temporal variation in response to pollutant plumes across multiple instruments.! Adriana Falcon1, Erin Riley2, Lanae Schaal2, Kris Hartin2, Timothy Gould2, Miyoko Sasakura2, Timothy Larson2, Michael Yost2, Christopher Simpson2! 1University of Puerto Rico, 2University of Washington!

Introduction! Results! Discussion!

The UW Center for Clean Air Research (CCAR) Synchronism required particular attention as Peak Type #1! focuses on the cardiovascular health effects of near- BC FWHM (s) ! Peak Type #2! offsets of more than a minute were observed. These roadway pollution. It uses a multi-pollutant BC Ptrack BC Ptrack time-lags could be further analyzed according to 16000 12000 20 8000 8000 mixture model, that takes into account the emission 18 16 atmospheric chemistry and factors not necessarily 12000 14 source, the atmospheric chemical processes and 8000 12 4000 4000 related to the instruments’ performance. In profiling physical aging of pollutants. This comprehensive 8000 10 ng/cm3 8 BC ng/cm3 4000 events and characterizing them, synchronism is an approach serves as a stepping stone for further 6 0 0 4000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

4 parcle count/cm3

parcle count/cm3 important factor that cannot be assumed. research regarding multi-pollutant relationships 2 0 0 0 and their health effects. 1 2 3 4 5 6 7 8 9 10 11 12 0 10 20 30 40 50 60 70 80 90 Peak Type #3! According to the FWHM, the instruments’

Synchronicity (s)! 25 FWHM (s) ! BC Ptrack responses were fairly similar on average, except for CCAR Project 1 is aimed at exposure mapping 2.5 24000 16000 and characterization of pollutant mixtures. It 2 20 Grimm. Variation was also observed within employs mobile monitoring that permits a realistic, 1.5 15 16000 instruments. This contrasts with the peak CO 8000 CO n=10 10 1 8000 thorough approach to exposure mapping. The ng/cm3 classification for which 23 peaks (28%) were found 0.5 5 mobile measurement platform houses multiple 0 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 to change types in different instruments, which parcle count/cm3 instruments, which are expected to react differently -50 -40 -30 -20 -10 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 4.5 16 Type #2 peaks in BC are two consecutive could mean that a certain instrument was sensitive 4 14 when challenged with an exhaust plume. This sharp peaks that occurred within 30 3.5 12 enough to differentiate between two distinct events poster captures the instrument responses, allowing 3 10 seconds of each other. 2.5 CO2 n=18 CO2 8 Type #3 peaks in BC are elevated from 2 that happened consecutively while another would for a more accurate characterization of pollution 6 1.5 baseline and contain several local 4 1 minimums and maximum values. present both events as a single, wide peak. Further events. 0.5 2 0 0 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 research related to the instruments’ particular

14 20 18 mechanisms would be encouraged. Methods! 12 16 10 14 12 Synchronicity shows a 8 10 G1 6 G1 n=30 8 broad distribution of events, 6 Conclusion! Data from mobile monitoring of Baltimore 4 4 as compared with BC. Ptrack during the dates (15-28 June, 2012) was analyzed. 7 2 2 0 Conclusion 0 0 10 20 30 40 50 60 70 80 90 was the only instrument with Characterizing temporal variation across different instruments were compared, using black -50 -40 -30 -20 -10 0 10 20 30 40 50 60 8 16 fairly similar timing to BC. instruments in a data set is of particular importance carbon (BC) as a reference. 81 peaks were selected 7 14 6 12 This broad range of responses in multi-pollutant research, as these factors from the BC data and classified into three 5 10 4 8 G2 n=27 G1 could be due to the variable

! Frequency ! Frequency complicate the interpretation of roadway sources. categories. The events from BC type 1 peaks were 3 6 2 4 nature of pollutant sources, then used to assess the synchronicity between 1 2 This research is the first step in better understanding 0 0 detector response times and instruments and event durations as captured by the -50 -40 -30 -20 -10 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 how pollutant plumes vary in an urban setting. the physical arrangement of Full Width at Half Max (FWHM). Some BC peaks 8 45 7 40 35 6 the instruments in the did not correspond to an event on another 30 5 25 platform. 4 Acknowledgements! instrument, these have FWHM of zero. NOx n=23 20 NOx 3 15 2 10 1 5 FWHM show that the This poster was supported by Award Number 5R25ES021646-02 Instruments with the measured pollutant: 0 0 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 from the National Institute of Environmental Health Sciences. distribution of responses are The content is solely the responsibility of the authors and does Black 7 16 Magee Scientific microAeth model AE52 not necessarily represent the official views of the National Carbon 6 14 different, but the modes are 12 5 Institutes of Health. CO 10 similar to the BC channel. This Langan T15v 4 8 3 PAH n=21 PAH Special thanks to Erin Riley, Christopher Simpson, Gabriela Falcón, and the CO2 Senseair CO2 Engine k-30-FS Sensor 6 indicates that the instrument 2 4 Environmental Health Research Experience Program (EHREP) for their help in 1 responses are similar for the Particles Grimm laser aerosol spectrometer 1.109 2 the preparation of this poster and for this research opportunity. 0 0 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 instruments across the Grimm 1 (.025-.3 m) For questions contact Adriana Falcon at [email protected] or Erin Riley 20 20 platform, with variability at [email protected].! 15 15 Grimm 2 (.35-1 m) likely due to characteristics of 10 10 August 2013! NOx 2B Technologies Model 410 Nitric Oxide Monitor 5 Ptrack n=28 5 Ptrack the plume. For FWHM all 33 PAH EcoChem PAS2000 0 0 -50 -40 -30 -20 -10 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 peaks were considered; 0 was Ptrack P-TRACK UPC w/diffusion screen Time (s)! Time (s)! assigned to absent peaks.