TracMyAir smartphone application for modeling exposures to PM and ozone - Integration with air quality networks and location-activity sensors.

Epidemiologic studies of ambient fine particulate matter (PM) and ozone (O) often use outdoor concentrations from central-site monitors or air quality model estimates as exposure surrogates, which can result in exposure errors. We previously developed an exposure model called TracMyAir, which is an iPhone application that determines seven tiers of individual-level exposure metrics for ambient PM and O using outdoor concentrations, home building characteristics, weather, time-activities. The exposure metrics with increasing information needs and complexity include: outdoor concentration (C, Tier 1), building infiltration factor (F, Tier 2), indoor concentration (C, Tier 3), time spent in microenvironments (ME) (T, Tier 4), personal exposure factor (F, Tier 5), exposure (E, Tier 6), and inhaled dose (D, Tier 7). In this study, we extended TracMyAir with two sets of additional features: (1) time-resolved exposures using smartphone geolocations with a ME classification model (MicroTrac) and official PM and O monitoring network, and (2) exposures based on low-cost outdoor PurpleAir (PA) PM monitoring network, non-ambient indoor PM using indoor-outdoor PA monitors, and inhaled dose based on physical activity data from smartphone and smartwatch. To demonstrate the two sets of extended features, we applied TracMyAir to estimate hourly PM and O exposure metrics for two corresponding panel studies with participants living in central North Carolina, USA. For Tier 4, the MicroTrac estimates were compared with 24-h diary data, and correctly classified the ME for 97 % of the daily time spent by the participants. Overall, the TracMyAir estimates showed considerable temporal and building-to-building variability of F, and C (Tiers 2-3), and person-to-person variability of C, T, F, E, and D (Tiers 1, 4-7). Our study demonstrates the capability of extending TracMyAir with air quality monitors, location-activity sensors, and models to determine fine-scale exposures, in support of epidemiologic studies and public health strategies to help reduce exposures to air pollutants.