GSI/WRF Regional Data Assimilation System and Its Application in theWeather Forecasts over Southwest Asia - Data Assimilation for Atmospheric, Oceanic and Hydrologic Applications - page 624

a

b

Skill Score Wind Speed at 10-m

30

20

10

0

-10

ABCDEF GH I

c

Skill Score Precip

100

80

60

40

20

0

-20

-40

-60

-80

-100

ABCDEFG HI

Fig. 25.16 Mean square error skill scores ( SS ) for 2-m temperature ( a ), 10-m wind speed ( b )and
24 h accumulated precipitation ( c ). Results are plotted for averaged 24- and 48-h forecasts as a
function of defined locations (Updated from Xu et al. ( 2009 ))

error. Satellite data assimilation, at least for the AMSU-A and AMSU-B radiances,
seems not to make a significant contribution to the accuracy of surface temperature
forecasts in the higher mountain areas.
In contrast, the 10-m wind speed in Fig. 25.16 b shows a reverse SS value from
the surface temperature. Six of nine locations including all high mountain areas
(B, D, E) show a negative skill score, which means the satellite data assimilation
produced a negative impact, but the SS in the Arabian Sea increases by 25 % and
20 % for 24- and 48-h forecasts, respectively. For the precipitation forecasts, the
results suggest (Fig. 25.16 c) that the satellite data assimilation only has a positive
impacts on improvement of forecast biases over Iraq (A), North of Iran (B) and
Saudi Arabia desert (F, G). The other five sub-regions become worse.