Applying indoor and outdoor modeling techniques to estimate individual exposure to PM2.5 from personal GPS profiles...

Impacts of individual behavior on personal exposure to particulate matter (PM) and the associated individual health effects are still not well understood. As outdoor PM concentrations exhibit highly temporal and spatial variations, personal PM exposure depends strongly on individual trajectories and activities. Furthermore, indoor environments deserve special attention due to the large fraction of the day people spend indoors. The indoor PM concentration in turn depends on infiltrated outdoor PM and indoor particle sources, partially caused by the activities of people indoor. We present an approach to estimate PM2.5 exposure levels for individuals based upon existing data sources and models. For this pilot study, six persons kept 24-hour diaries and GPS tracks for at least one working day and one weekend day, providing their daily activity profiles and the associated geographical locations. The survey took place in the city of Münster, Germany in the winter period between October 2006 and January 2007. Environmental PM2.5 exposure was estimated by using two different models for outdoor and indoor concentrations, respectively. For the outdoor distribution, a dispersion model was used and extended by actual ambient fixed site measurements. Indoor concentrations were modeled using a simple mass balance model with the estimated outdoor concentration fraction infiltrated and indoor activities estimated from the diaries. A limited number of three 24-hour indoor measurements series for PM were performed to test the model performance. The resulting average daily exposure of the 14 collected profiles ranged from 21 to 198 microg m(-3) and showed a high variability over the day as affected by personal behavior. Due to the large contribution of indoor particle sources, the mean 24-hour exposure was in most cases higher than the daily means of the respective outdoor fixed site monitors. This feasibility study is a first step towards a more comprehensive modeling approach for personal exposure, and therefore restricted to limited data resources. In future, this model framework not only could be of use for epidemiological research, but also of public interest. Any individual operating a GPS capable device may become able to obtain an estimate of its personal exposure along its trajectory in time and space. This could provide individuals a new insight into the influence of personal habits on their exposure to air pollution and may result in the adaptation of personal behavior to minimize risks.