Observing-System Research and Ensemble Data Assimilation at JAMSTEC - Data Assimilation for Atmospheric, Oceanic and Hydrologic Applications - page 513

V − spread and Vorticity ( σ =0.85, 12N − 18N)

2010
16SEP

21SEP

26SEP

1OCT

6OCT

11OCT

16OCT

21OCT

26OCT

100E 110E 130E 140E 160E

contour: vorticity (x1.0e05 s − 1 )

120E 150E

− 2, − 1, 1, 2, 4, 6, 8

1 1.2 1.4 1.6 1.8

Fig. 21.8 A longitude-time section of the analysis ensemble spread of meridional winds ( shading )
and vorticity ( contours )at 0.85 averaged between 12 and 18 N. Blue dots indicate the locations

of the center of a tropical cyclone

D

during field campaigns such as PALAU 2008, the Mirai Arctic Ocean Cruise 2010
and VPREX 2010. These experiments show the importance of additional observa-
tions in data-sparse regions over the ocean. Details will be reported elsewhere.
Our next step is to incorporate the variability of the ocean. With the coupled
atmosphere-ocean model for the Earth Simulator (CFES) ( Komori et al. 2008 ),
we began the development of data assimilation system called CLEDAS. In this
system, the ocean boundary conditions are replaced by the ocean general cir-
culation model for the Earth Simulator (OFES) ( Pacanowski and Griffies 2000 ;
Masumoto et al. 2004 ) with the sea-ice process ( Komori et al. 2005 ). Currently,
only atmospheric observations can be assimilated. Preliminary tests indicate that the
replacement of AFES with CFES contributes to an increased ensemble spread in the
lower troposphere. In the future, CLEDAS will allow ocean and land observations to
be assimilated. We plan to use CLEDAS as a test bed for the a study of the problem
of assimilating data with different spatiotemporal scales and degrees of nonlinearity.
CLEDAS may also be used to evaluate and design the observation systems and field
campaigns conducted by JAMSTEC.