Continuous hourly measurements of isoprene and 30 other hydrocarbons were p
erformed at an urban centre site in Lille, France, from May 1997 to April 1
999. Parallel mass emissions of the same hydrocarbons from in-service passe
nger vehicles were determined from measurements made on a chassis dynamomet
er using the European MVEG driving cycle. On the one hand, descriptive stat
istics and principal component analysis revealed the strong traffic origin
of isoprene in winter months and its double biogenic and anthropogenic orig
in during the summer. On the other hand, the emission measurements of indiv
idual hydrocarbons in exhaust gases confirmed the presence of isoprene in p
etrol fuelled (with or without catalytic converters) and diesel car exhaust
s. Finally, the isoprene/acetylene ratios, both of them derived from ambien
t concentrations and emission factors, were compared. No statistically sign
ificant difference was found in winter, indicating the strict traffic origi
n of isoprene during that period. For the winter period, a simple regressio
n analysis was performed on daily isoprene concentrations vs. those of acet
ylene and three other exhaust gases tracers-propene, ethylene and 1,3-butad
iene. The established regression equations, together with the four tracer c
oncentrations, were used to estimate the vehicle exhaust fractions of isopr
ene. From November to March, vehicle exhaust explained the totality of isop
rene levels. While traffic remained the major source of isoprene with a con
tribution greater than 50% during the growing season, it still constituted
a non-negligible source of isoprene in summer, anti-correlated to temperatu
re and fluctuating between 10% and 50%. The application with 1,3-butadiene
gives the greatest estimation of the anthropogenic fraction of isoprene. Ot
her sources of 1,3-butadiene, acetylene, ethylene and propene were suspecte
d in addition to their known traffic origin. (C) 2001 Elsevier Science Ltd.
All rights reserved.