Environmental Engineering Reference
In-Depth Information
model performance evaluation statistics (bias and error) for these comparisons.
For PM
2.5
, CAMx tended to show less bias and error than CMAQ and for O
3
,
CMAQ tended to show better performance. However, for both O
3
(8 h max) and
total PM
2.5
, there were only slight differences in model performance between the
three horizontal resolutions for each model.
Table 1.
Model Performance Evaluation Statistics for CMAQ and CAMx for O
3
and total PM2.5
Modeled at Various Horizontal Grid Resolutions
N
12 km
4 km
1 km
Bias
Error
Bias
Error
Bias
Error
CMAQ
O
3
8 h Max (ppb)
93
1.001
8.527
−0.685
8.917
−1.526
8.949
CAMx O
3
8 h Max (ppb)
93
−
0.884
9.260
−
2.956
10.230
−
3.200
10.313
CMAQ PM
2.5
(µg/m
3
)
259
3.990
8.217
4.512
8.182
4.932
8.760
CAMx PM
2.5
(µg/m
3
)
259
−
0.053
6.315
0.344
5.247
0.483
6.583
In this study, we also compared the MAHM to the CMAQ and CAMx modeled
concentrations for 1 km horizontal resolutions.
Table 2
shows the model perfor-
mance evaluation statistics (bias and error) for the MAHM, CMAQ and CAMx.
The CMAQ and CAMx concentrations were only extracted for the AERMOD
domain for this evaluation. For the toxics species, the EPA Initial Screening Risk
Table 2.
Model Performance Evaluation Statistics for the MAHM and CMAQ and CAMx for 1
km horizontal resolution
N
ISRL
MAHM
CMAQ 1 km
CAMx 1 km
Bias
Error
Bias
Error
Bias
Error
PM
2.5
(µg/m
3
)
185
6.083
6.213
9.318
0.896
6.375
−2.222
EC+OC (µg/m
3
)
25
−1.173
3.125 −1.098
2.340 −3.760
3.765
Benzene (µg/m
3
)
36
0.1
−1.049
3.203 −0.797
2.739 −1.974
3.075
Formaldehyde
(µg/m
3
)
32
181.8
−3.112
4.520
−3.150
4.491
−3.323
4.600
Acetaldehyde
(µg/m
3
)
32
0.5
0.353
1.117
0.321
1.164
−0.259
1.097
1,3-Butadiene
(µg/m
3
)
36
0.03
0.044
0.144
0.045
0.137
−0.017
0.115
Dichlorobenzene
(µg/m
3
)
46
0.14
0.107
−0.072
0.107
−0.099
0.113
−0.081
