You are here:
APPLICATION OF THE RANDOM COMPONENT SUPERPOSITION (RCS) MODEL TO PM2.5 PERSONAL EXPOSURE AND INDOOR AIR QUALITY MEASUREMENTS IN DIFFERENT CITIES
Citation:
Wallace, L A. AND W R. Ott. APPLICATION OF THE RANDOM COMPONENT SUPERPOSITION (RCS) MODEL TO PM2.5 PERSONAL EXPOSURE AND INDOOR AIR QUALITY MEASUREMENTS IN DIFFERENT CITIES. Presented at International Society of Exposure Analysis 2002 Conference, Vancouver, Canada, August 11-15, 2002.
Impact/Purpose:
The main objective is to investigate human exposure to fine and coarse particles (and PAHs) from several important sources such as cooking, woodsmoke, and household cleaning. A second objective is to investigate the observed increased personal exposure (compared to indoor air concentrations measured by a fixed monitor) to particles: the so-called "personal cloud," that has been observed in many occupational and some environmental studies. A third objective is to incorporate the findings into a mass-balance indoor air quality model.
Description:
The RCS model allows us to estimate the distribution of population exposure to air pollutants in any city given only the outdoor measurements in that city. Since outdoor measurements are made in many cities, but personal exposures are measured in few, the model could conceivably be very useful. The fundamental assumption of the model is that the distribution of exposures to pollutants emitted by personal and indoor sources is invariant across cities. Although of course cities have somewhat different levels of smoking, frequency of open windows, etc., it is at least testable whether these differences significantly affect the observed distribution of exposures. To test the model, it is necessary to have probability-based studies of personal exposure, including outdoor measurements, in two cities. From these measurements, one can separate the outdoor from the non-outdoor (indoor and personal) components of personal exposure. One can then use the outdoor measurements in one city to estimate the total personal exposure measurements in the other, by simply adding the invariant distribution of non-outdoor exposures to the measured outdoor concentrations in the second city. This predicted distribution of total exposures can then be compared to the observed distribution to test the adequacy of the model. Besides estimating personal exposures, the model can also estimate indoor concentrations given the proper probability-weighted fixed indoor measurements.
The RCS model has been applied with success to PM10 indoor and personal exposure measurements from field studies conducted in 3 cities - Toronto, Canada; Phillipsburg, NJ, and Riverside, CA. Until recently, only one probability-based personal exposure study of PM2.5 was available (Toronto, Canada), but the completion of a second probability-based personal exposure study in Indianapolis has made it possible to test the model for PM2.5 as well as PM10. Indoor PM2.5 measurements made in Riverside, CA; Toronto, Canada; and Indianapolis will also be explored in this paper.
Regression analysis is useful for applying the RCS model to personal and indoor air measurements, because the physically based superposition model of indoor concentrations and personal exposures yields the same basic equation used in the standard linear regression model. Thus, regression approaches, when applied in RCS modeling, yield parameter values that have an important physical interpretation, such as the particle exposure attenuation factor a = pa/(a + k), where p is the penetration factor, a is the air exchange rate, and k is the particle deposition rate.
This abstract has been subjected to Agency review and approved for publication.