Original scientific paper
MODELLING OF PRIMARY PM10 CONCENTRATIONS FOR THE CITY OF GRAZ, AUSTRIA
Dietmar Oettl
; Government of Styria, Air Quality Department, Austria
Abstract
Within the frame of the European funded LIFE Project KAPA GS (2004-2008), a methodology for modeling the PM10
concentration with a very high resolution (10m x 10m) at the urban scale has been developed. After a first successful application for
the city of Klagenfurt the method has now being applied to the city of Graz (~250.000 inhabitants) in Austria. Based on existing
emission inventories for traffic including resuspension of road dust, domestic heating, and industry, simulations for the spatial
distribution of PM10 have been performed using the Lagrangian particle model GRAL (Graz Lagrange Model), which is coupled
with the mesoscale prognostic model GRAMM (Graz Mesoscale Model). The latter has been operated with a horizontal resolution
of 300m x 300m, while the whole modeling domain was 27km x 39km.
GRAMM has been initialised using a simple statistics of 5 wind speed classes, 36 sectors of wind direction, and Pasquill-Gifford
stability classes. The latter has been used to obtain the vertical profile of potential temperature. For each, out of several hundreds of
such characterised dispersion conditions, a steady-state wind field and subsequently a steady-state concentration field has been
computed.
16 meteorological stations have been used to compare simulated and observed wind direction frequencies as well as wind speeds.
Apart from 12 permanent air quality monitoring stations (AQM) operated by our department, additional 9 particle counters for
PM10, PM2.5 and PM1.0 (GRIMM) were available for comparison purposes.
It is shown that due to extremly bad dispersion conditions in the basin of Graz, PM10 levels can reach or even exceed that of
megacities. A fairly well agreement between modelled and observed average PM10 concentrations could be achieved, although
significant uncertainties arise for each step of the whole model chain (emission inventory, wind field-, and air quality simulations).
It came out that traffic is the most significant source of PM10 in the centre of Graz and nearby busy roads, while domestic heating
may become the dominant source in suburban areas. Especially non-exhaust emissions are comparatively high, as became evident
from road measurements in Graz and Klagenfurt, and additional chemical analyses of PM10 filters. There is also some observational
evidence of local generated secondary PM (e.g. ammonium nitrate) at least in the city of Graz.
Keywords
Lagrangian dispersion model; GRAL; Urban scale; PM10
Hrčak ID:
64274
URI
Publication date:
12.12.2008.
Visits: 1.396 *