GFS .032 behind ECMWF

1
GFS .032 behind ECMWF .005 behind Met
Office
2
6 dropouts of 20 points worse than ECMWF half
of difference from ECMWF
GFS .083 behind ECMWF .027 behind Met
Office
3
April 2008
GFS, EC Both improved Over other centers
April 2007
4
ECMWF Close to medley
Medleys best
APRIL 2008
April 2007
5
Init 4/24/0Z 12-36 h
Parallel better
6
Init. 4/23/00z 36-60
Parallel worse
7
Both have heavy precip too far north, Parallel
better over Arkansas
8
Parallel underdoes precip over DC, overdoes
precip over se Ohio
9
Parallel too strong maxima over se US
10
Parallel better pattern
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
500 hPa geopotential height
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
Parallel slightly worse precip scores over CONUS
Jan-May 08 midlat height scores not
significantly different tropical wind errors
reduced (?) GFS hurricane tracks improved
29
(No Transcript)
30
(No Transcript)
31
ECMWF explains 65 of variance of 500 hPa
height Met Office 57.2 GFS 57.0 CDAS 38
32
ECMWF explains 62 of variance of 500 hPa
height Met Office 56 GFS 52 CDAS 30
33
GFS 38 less error than CDAS
34
GFS 36 less than CDAS
35
GFS 16 less than CDAS
36
GFS 13 less than CDAS
37
GFS 36 less than CDAS
38
GFS 30 less than CDAS
39
(No Transcript)