Environmental Engineering Reference
In-Depth Information
European domain (see http://www.bsc.es/caliope/). This contribution describes the
annual dynamics and the statistical performances of the model system for a
reference year (2004). The analysis is centered on summer months, as the most
severe ozone and other pollutant episodes mainly occur during warm and sunny
periods. The evaluation comprises the inter-comparison of O
3
, NO
2
, and SO
2
in
the gas phase and particulate matter PM
10
and PM
2.5
, together with observations
performed at EMEP measurement sites.
2. Methodology
The WRF-ARW meteorological model contains 38 vertical layers reaching up to
50 hPa. Initial and boundary conditions are obtained from the NCEP reanalysis
data. The vertical resolution of the CMAQ chemistry-transport model (version 4.5;
Byun and Schere, 2006) was increased from 8 to 15 layers in order to simulate
vertical exchanges more accurately. The vertical sigma-p coordinates cover up to
the lower stratosphere (5 hPa) with an increased resolution within the planetary
boundary layer. Gas phase boundary conditions are provided by the LMDz-
INCA2 global climate-chemistry model (see Hauglustaine et al., 2004). The
aerosol module utilized is AERO4, in comparison to the AERO3 module used in
previous simulations. The Dust REgional Atmospheric Model (DREAM, Nickovic
et al., 2001) model simulates long-range transport of mineral dust from the African
continent over the domain under study. For this European simulation (resolution:
12 km, 1 h), emissions are disaggregated from the EMEP expert emission inven-
tory for 2004 to the utilized resolution using the criteria implemented in the
HERMES emission model (Baldasano et al., 2008). In order to evaluate the
performances of the CALIOPE system, model simulations were compared with
ground-based measurements from the EMEP network which applied quality
criteria and controls on the data (
Fig. 1)
. Note that the months of March, April and
May are referred to as “spring”, and June, July and August as “summer”, res-
pectively. Note that only hourly data are discussed for gas phase species, while
aerosol measurements are available on daily basis only. According to previous
studies of single and inter-comparison model evaluations and to the US-EPA
(1991, 2007) suggestions, a suite of discrete statistics has been used to characterize
the model behaviour, such as the Mean Normalized Bias Error (MNBE) (%),
Mean Normalized Gross Error (MNGE) (%), Root Mean Square Error (RMSE)
(µg m
−3
) and correlation (cor). Apart from the correlation, statistics are calculated
using recommended thresholds (O
3
: 80 µg.m
−3
; NO
2
: 5 µg.m
−3
; SO
2
: 3 µg.m
−3
;
PM
10
/
2.5
: 10 µg.m
−3
).
