a)
1
=
(1)
0.8
0
0
a.: a... 0.6
0
0
>- 0.4
cu
E
0
c: 0.2
«
0
0
b) 1
~
0.8
0
0
a.: a... 0.6
0
0
>- 0.4
' m
E
0
c: 0.2
«
0
0
113
IJm'''~ ..... .
IlJlJ.'il.tIj::!:::;;;;;~l~" u ....... .
12
7
" 16
~ ...... : .. \ . . .. : .' ',._,' .. .. (
~;\~\~:\? WV'~ , :
500 hPa height
2
Day
""~\P~,\ ' i)'
10
i
11
5
1
15
l'
500 hPa height
2
4
Day
1V
20
21
it
:aa
32
27
23
27
22
~A
23
~
32
. .
~g
Figure 11 : Ten-day ensemble forecast dispersion as measured by anomaly correlation
between 500 hPa geopotential height of control and perturbed forecasts, over Europe
(light lines). Skill of lO-day control forecast (heavy line) , a) forecasts from 30 October
1993, b) fore casts from 13 November 1993. From Palmer et at (1994).
1
=
(1)
0.8
0
0
a.: a... 0.6
0
0
>- 0.4
cu
E
0
c: 0.2
«
0
0
b) 1
~
0.8
0
0
a.: a... 0.6
0
0
>- 0.4
' m
E
0
c: 0.2
«
0
0
113
IJm'''~ ..... .
IlJlJ.'il.tIj::!:::;;;;;~l~" u ....... .
12
7
" 16
~ ...... : .. \ . . .. : .' ',._,' .. .. (
~;\~\~:\? WV'~ , :
500 hPa height
2
Day
""~\P~,\ ' i)'
10
i
11
5
1
15
l'
500 hPa height
2
4
Day
1V
20
21
it
:aa
32
27
23
27
22
~A
23
~
32
. .
~g
Figure 11 : Ten-day ensemble forecast dispersion as measured by anomaly correlation
between 500 hPa geopotential height of control and perturbed forecasts, over Europe
(light lines). Skill of lO-day control forecast (heavy line) , a) forecasts from 30 October
1993, b) fore casts from 13 November 1993. From Palmer et at (1994).