Monitor-to-monitor temporal correlation of air pollution and weather variables in the North-Central U.S.

Abstract

Numerous time series studies have reported associations between daily ambient concentrations of air pollution and morbidity or mortality. Recent personal exposure studies have also reported relatively high longitudinal correlation between personal exposures to particulate matter ( PM ) and home outdoor PM concentrations, lending support to the health effects reported in time series studies. However, the question remains as to how well the temporal fluctuations in the air pollution levels observed at an outdoor monitor represent the temporal fluctuations in the population exposures to pollution of outdoor origins in a city, and how such representativeness affects the size and significance of risk estimates. Also, such spatio -temporal correlations would vary from pollutant to pollutant, likely influencing their relative significance of statistical associations with health outcomes. In this study, we characterized the extent of monitor -tomonitor correlation over time among multiple monitoring sites for PM less than 10 m (PM 10 ), gaseous criteria pollutants, and several weather variables in seven central and eastern contiguous states ( IL, IN, MI, OH, PA, WI, and WV ) during the study period of 1988 ± 1990. After removing seasonal trends, the monitor -to -monitor temporal correlation among the air pollution / weather variables within 100 -mile separation distance in these areas could be generally ranked into three groups: ( 1 ) temperature, dew point, relative humidity ( r > 0.9 ); ( 2 ) O 3 , PM 10 , NO 2 ( r: 0.8 ± 0.6 ); and ( 3 ) CO, SO 2 ( r < 0.5 ). Using the subsets for separation distance less than 100 miles, regression analyses of these monitor -to -monitor correlation coefficients were also conducted with explanatory variables including separation distance, qualitative ( land use, location setting, and monitoring objectives ) and quantitative ( large and small variance ) site characteristics, and region indicators for Air Quality Control Region ( AQCR ). The separation distance was a significant predictor of monitorto -monitor correlation decline especially for PM 10 and NO 2 ( $0.2 drop over 30 miles ). Site characteristic variables were, in some cases, significant predictors of monitor -to -monitor correlation, but the magnitude of their impacts was not substantial. Regional differences, as examined by AQCR, were in some cases ( e.g., in Metropolitan Philadelphia ) substantial. In these areas, the pollutants that had generally poor monitor -to -monitor correlation in the overall seven states data ( i.e., for SO 2 and CO ) showed higher monitor -to -monitor correlations, comparable with PM 10 and O 3 , within the AQCR. These results are useful in interpreting some of the past time series epidemiological results. The differences in monitor -to -monitor correlations found across pollutants in this work ( i.e., r$0.8 vs. r$0.4 ) are sufficiently large that they could be a factor in the different pollutant significance levels reported in the epidemiologic literature. It is recommended that future epidemiol...

Comparing exposure metrics in the relationship between PM2.5 and birth weight in California

“…In New York City, for example, measurements of sulfur dioxide at one aerometric monitoring station was not found to be representative of overall exposure in the city in studies of acute effects (Goldstein, 1979;Goldstein et al, 1979). Another study examining monitor-to-monitor correlations in the North-Central US reported that correlations varied by location for PM 10 , gaseous criteria pollutants, and several weather variables (Ito et al, 2001).…”

Epidemiological studies of acute ozone exposures and mortality

“…This may, in turn, reduce such timeseries methods' ability to detect the effects of some pollutants. The use of multiple -site averages for a city or region in such aggregate epidemiological studies should reduce this problem, and the problem is less severe for more spatially correlated pollutants ( e.g., see Ito et al, 1995Ito et al, , 2001 ). In the case of O 3 air pollution, which is highly correlated spatially across a metropolitan area, the chief concern is actually that the pollutant may not permeate equally in all buildings, for example, in air-conditioned versus non-air-conditioned buildings (Wechsler et al, 1989 ).…”

Estimation of long-term average exposure to outdoor air pollution for a cohort study on mortality

“…Other investigators have used kriging to interpolate concentrations (Wartenberg et al, 1991;Liu and Rossini, 1996 ). Kriging assumes that data from geographically close sites are more similar than more distant sites (see Ito et al, 2001 ). The procedure uses a function describing the correlation versus separation distance.…”