Improvements of AMSR sea surface wind speed retrieval algorithm

1

• Improvements of AMSR sea surface wind speed
  retrieval algorithm
• Akira Shibata
• JAXA
• AMSR-E meeting, at
  Washington, June 25-26, 2009

2
Study Purposes
Improve wind speed retrieval algorithm for
AMSR2 / GCOM-W Tune coefficients in AMSR
algorithm comparing the scatterometer data of
ADEOS-II
3
SSW retrieval algorithm
s36 (36H - (s (36V a) b ) ) / fac
parameters a ,b, s, fac depend on SST
36V
fac
calm surface
roughened surface
s
a
s36
36H
b
4
Removal of relative wind direction dependency -
6H / s36 plane -
cross
(K)
16m/s
up
down
6H
7m/s
(K)
s36
6H AMSR_6H atmos_effect_6H calm_ocean_6H
5
Conversion of s36 to wind speed
K
s36
m/s
wind speed
6
Cross talk with SST
AMSR - Seawinds
bias (m/s)
1.0
5
20
10
15
25
SST 0C
Seawinds wind speed 10m/s
-1.0
30
7
Cross talk with water vapor
bias (m/s)
1.0
vapor 5mm
15
55
Seawinds wind speed 10m/s
-1.0
35
45
25
8
Cross talk with cloud liquid water
bias (m/s)
1.0
0.3
cloud liquid water less than 0.05mm
0.4
0.5
0.1
0.2
Seawinds wind speed 10m/s
-1.0
9
Dependency on relative wind direction
downwind
upwind
downwind
Seawinds wind speed 20m/s
10m/s
0m/s
m/s
10
Density distribution
1.0
0.0
20m/s
10m/s
AMSR wind speed
11
Conclusion
The framework of SSW algorithm seems to work
well, and tuning parameters has been almost done.
But, several points of further works seem to
be necessary, in particular for tuning the cloud
liquid water effect.