ARB Research Seminar

This page updated June 19, 2013

In-Vehicle Air Pollution Exposure Measurement and Modeling

Photo of Scott Fruin, Ph.D.

Scott Fruin, Ph.D., Preventive Medicine, Division of Environmental Health, Keck School of Medicine, University of Southern California

December 12, 2012
Cal EPA Headquarters, 1001 "I" Street, Sacramento, CA

Research Project


Time spent in vehicles can contribute disproportionately to overall exposure to traffic-related pollutants because of high on-road concentrations. However, if air exchange rates (AERs) are low, in-vehicle loss rates for particulate pollutants, such as diesel PM or ultrafine particles, can be significant, and in-vehicle concentrations can be much lower than outside the vehicle. AER tends to be lower for newer, tighter vehicles; lower at low speeds; and lower at recirculation ventilation settings. However, not until this study had AER been characterized in a generalizable way. We developed new, simpler methods for measuring AER using vehicle occupant CO₂. This allowed testing a large enough sample of vehicles (n=59) to develop successful predictive models (R2=0.7) that require only easy-to-obtain information: age or mileage of vehicle, make and manufacturer, speed, and preferred ventilation setting, appropriate for population- or cohort-sized exposure estimation. Building on these relationships, we further established that similar variables can be used to predict inside-to-outside (I/O) ratios for particulate pollutants. For ultrafine particle (UFP) measurements in 43 cars, the predictive model R2 was 0.79.  I/O ratios ranged from 0.66 0.10 (SD) for outside air ventilation settings, and 0.17 0.13 for recirculation settings, increasing within each range for higher speeds and greater vehicle age. Ratios were relatively insensitive to particle size or in-cabin filter condition. Later measurements showed I/O ratios for particle-bound PAHs, black carbon, and PM2.5 are similarly predictable. Monte Carlo simulations of US fleet characteristics, LA traffic speeds, and on-road LA concentrations demonstrated ventilation setting (outside air or recirculate), road type (freeway or arterial) and an IQR (25th to 75th percentile) of vehicle age and mileage each had roughly similar effects on in-vehicle UFP exposures, each making a 2 to 3 fold difference.

The other part of the exposure equation, on-road concentration, was also modeled, with about 50% of measurement variability explained for PM2.5 and particle-bound PAH, 64% for UFP, and 73% for NOx. Therefore, there is room for improvement in the on-road models. The results of this study show there is little need for further study of inside-to-outside pollutant ratios inside vehicles until on-road predictive models show better performance.

Speaker Biography

Scott Fruin, Ph.D., is Assistant Professor in Preventive Medicine, Division of Environmental Health, Keck School of Medicine, University of Southern California. Dr. Fruin's research focuses on air pollution exposure assessment and includes field measurements in support of population-based, longitudinal health studies. One interest is improving measurement techniques and/or devices to increase both accuracy and ease of assessing exposure. Of particular interest to Dr. Fruin is better characterization of high exposure environments such as in-vehicle, near-vehicle and near-roadway environments, and the use of mobile approaches to map spatial differences in pollution. Dr. Fruin is also interested in bridging the gaps between specific components of particulate matter and health effects through investigating the links between spatial differences in PM composition, biological activity, and human health endpoints.

Other authors and investigators: Ralph J. Delfino, M.D., Ph.D., Jun Wu, Ph.D, and Lianfa Li, Ph.D., University of California, Irvine; Constantinos Sioutas, Sc.D., and Neelakshi Hudda, Ph.D., Univeristy of Southern California.

For a complete listing of the ARB Research Seminars and the related documentation
for the seminars please view the Main Seminars web page