Chemistry Reference
In-Depth Information
Fig. 4 Average source contributions to road dust loadings measured at different locations in
Barcelona and Zurich
Receptor models are widely used tools for apportioning concentrations of
pollutants to different sources. They can be factor analytical methods (PMF,
PCA, UNMIX, etc.) or chemical mass balance (CMB). On the one hand, these
methods revealed to be very valuable to identify the main sources/categories of PM
pollution (road traffic, secondary particles, fuel oil combustion, sea salt, etc.) but on
the other hand they experienced difficulties in separating the contributions of
collinear sources such as mineral dust (natural resuspension) and road dust (anthro-
pogenic) or co-variant sources such as vehicle exhaust and road dust [
34
,
44
,
45
,
49
,
55
,
58
,
110
-
113
]). Significant improvements were made with the use
of combination of models or constrained models such as the Multilinear Engine
(ME-2).
In spite of the low number of studies and the wide range of methods used, it is
already possible to observe geographical variability on the road dust emission
strength and air quality impact. Central Europe experiences the lowest road dust
contributions due to the wet climate and the ban of studded tyres. This low “signal”
hampers the task of quantifying road dust contribution. In Germany, Beuck et al.
[
42
] estimated the contribution to PM
10
from road dust in 2.4
g/m
3
(8%) and 0.3
(2%) at the urban and regional background sites, respectively. In Stuttgart non-
exhaust emissions from road traffic are estimated to be about twice as high as
exhaust emissions [
114
]. Bukowiecki et al. [
33
] have analysed the traffic-related
emissions (through trace elements, BC and nitrogen oxides) at a heavily congested
street canyon in Z¨rich and assigned 21% of the traffic related PM
10
emissions to
brake wear, 38% to road dust and 41% to exhaust emissions. Astel [
115
] could
distinguish road dust contributions from those of soil and vehicular exhaust in
Cracow and Vienna by combining several models (CMB, PCA-APCS, PMF and
UNMIX).
Thorpe et al. [
116
] proposed the roadside incremental concentration of coarse
particles above the urban background as a first estimate of the sum of source
strength road dust resuspension and the coarse fraction of wear emissions. Other
studies succeeded in separating different traffic emissions by means of multivariate
receptor models applied to PM size distribution data ([
84
,
117
].
m
