Environmental Engineering Reference
In-Depth Information
(Woo et al., 2003). The emissions are provided on a 0.5° by 0.5° grid. The reference
year is 2006 and emissions are given in ton per year (http://www.cgrer.uiowa.edu/
people/carmichael/ACESS/Emission-data_main.html).
Meteorological fields, required as input for AURORA, were simulated using
the Advanced Regional Prediction System (ARPS), a non-hydrostatic mesoscale
atmospheric model developed by the University of Oklahoma (Xue et al., 2000,
2001). The equations of this model are finite-difference on a Arakawa C-grid.
Terrain-following coordinates are employed in the vertical direction. Leapfrog
time stepping in a fourth order central differencing scheme is used to solve the
advection, while the turbulence is represented by a 1.5 order TKE model with
the Sun and Chang (1986) parameterization for the convective boundary layer.
Furthermore, ARPS contains detailed parameterizations for the cloud micro-
physics, cumulus convection and radiation transfer. Thanks to its nesting capa-
cities, large-scale atmospheric features are allowed to enter the domain through
the lateral boundaries. A detailed land surface scheme was also incorporated in
ARPS, with an improved land surface model for urban surfaces (De Ridder and
Schayes (1997); De Ridder (2006)).
The air quality model used in this study is AURORA (Air quality modelling in
urban regions using an optimal resolution approach, Mensink et al. (2001)). In this
model, the vertical diffusion is calculated with the Crank-Nicholson method (De
Ridder and Mensink, 2002), while the horizontal diffusion uses a Walcek (2000)
scheme. The gas phase chemistry is treated by the Carbon-Bond IV scheme (Gery
et al., 1989), which has been enhanced to take into account biogenic isoprene
emissions. For particulate matter (PM
10
and PM
2.5
), a distinction has been made
between primary and secondary particles. These secondary particles are simulated
in a simple fashion using constant gas-to-particle conversion rates for both the
transitions between SO
2
and sulphate aerosols and between HNO
3
and nitrate
aerosols.
Both the amount and distribution of green vegetation cover are taken from the
SPOT-VEGETATION satellite. Terrain height is taken from the Global 30 Arc-
second Elevation Data Set, distributed by the U.S. Geological Survey.
The model effort is centered on the city of Shenyang, one of the ten largest
cities of China and capital of the Liaoning Province, in North-Eastern China.
Together with its neighboring cities it forms one of the most important industrial
centers of China. Within the AMFIC program we obtained measurements of air
pollution data in Shenyang which makes it an interesting area to test the perfor-
mance of the our AURORA model.
The measurement data comes from eight stations in and around the city of
Shenyang and measures three distinct pollutants, TSP (Total Suspended Particles),
NO
2
and SO
2
, which therefore will be in our focus in this paper. The simulations
have been made on resolutions of 30, 10, 3 and 1 km (with in each case 51 × 51
horizontal grid points and 35 vertical levels) centered on the city of Shenyang, in
which the runs on a finer resolution are nested in the runs with a coarser reso-
lution. The 30 km run has been nested for the meteorology in the GFS-FNL (Global
Forecast System-Final) model of the NCEP (National Centers for Environmental
