The variability of hazardous air pollutants (HAPs) is an important Fac
tor in determining human exposure to such chemicals, and in designing
HAP measurement programs. This study has investigated the factors whic
h contribute to HAP variability in an urban area. Six measurement site
s separated by up to 12 km collected data with 3 h time resolution to
examine spatial variability within neighborhoods and between neighborh
oods. The measurements were made in Columbus, OH. The 3 h results also
were used to study temporal variability, and duplicate samples collec
ted at each site were used to determine the component of variability a
ttributable to the measurement process. Hourly samples collected over
10 days at one site provided further insight into the temporal resolut
ion needed to capture short-term peak concentrations. Measurements at
the 6 spatial sites focused on 78 chemicals. Twenty-three of these spe
cies were found in at least 95% of the 3 h samples, and 39 chemicals w
ere present at least 60% of the time. The relative standard deviations
for most of these 39 frequently detected chemicals was 1.0 or lower.
Variability was segmented into temporal,spatial, and measurement compo
nents. Temporal variation was the major contributor to HAP variability
for 19 of the 39 Frequently detected compounds, based on the 3 h data
. Measurement imprecision contributed less than 25% for most of the vo
latile organic species, but 30% or more of the variability for carbony
l compounds, trace elements, and particle-bound extractable organic ma
ss. interestingly, the spatial component contributed less than 20% of
the total variability for all the chemicals except sulfur. Based on th
e data with hourly resolution, peak to median ratios (hourly peak to 2
4 h median) averaged between 2 and 4 for most of the volatile organic
compounds, but there were two species with peak to median ratios of ab
out 10. Copyright (C) 1996 Published by Elsevier Science Ltd