Environmental Engineering Reference
In-Depth Information
Table 1.
Comparison of MBLM and Eta model errors for surface wind predictions
0900 LST
1500 LST
2100 LST
Model
errors
HR
(%)
RMSE
(m s
−
1
)
HR
(%)
RMSE
(m s
−
1
)
HR
(%)
RMSE
(m s
−
1
)
MBLM
32
3.2
50
3.5
41
3.6
Eta
79
5.6
82
6.8
85
6.5
A test was conducted over a 6-month period forcing the MBLM with the Eta
model outputs in order to produce operative daily forecasts (Sraibman and Berri,
2009). Two accuracy measures are used: the hit rate (HR) or percentage of cases with
agreement in the wind direction sector (45° sectors), and the root-mean-squared
error (RMSE) of the horizontal wind components. The BLM surface wind forecasts
are always more accurate, since its averaged hit rate is three times greater and its
averaged RMSE is one half smaller than the Eta forecasts (see
Table 1
). Despite
the large errors in the surface winds displayed by the Eta forecasts, its 850 hPa
winds and surface temperature forecasts are able to drive the BLM model to
obtain surface winds forecasts with smaller errors than the Eta model.
15 km 24-h lagrangian trajectories 5 km
15 km 5 km
Fig. 3.
24-h lagrangian trajectories from two simulated continuous point sources. Left (right)
panels are run at 15 km (5 km) horizontal resolution. The thick
arrow on the upper right corner
indicates the regional scale wind of the experiment
