Pr

1
SWOT viewed with some OSSEs C. Ubelmann, J.
Verron, LEGI, Grenoble
30

• Model OPA1.12 on a 1/4 tropical Atlantic
  configuration
• Assimilation SEEK filter (Brasseur Verron,
  2006)
• OSSEs method Twin experiments with perturbed
  winds (not pure twin), CORE vs. ERA40
• Virtual observations Altimetric scenarios
  including some for SWOT Temperature and
  Salinity PIRATA profiles
• Errors Instrumental error simulated (roll error
  for SWOT white noise), model error
  parameterization

-30
Sea Surface Height (cm) from Opa configuration on
day 230. 1whole field, 2 Jason extraction, 3
SWOT extraction
2
Results on V in depth
Results on SSH
No correction
Full SH
Sea Surface Height RMS error (cm) for various
altimetric scenarios assimilation.
Meridional velocity RMS error (m/s) averaged on
TIW zone for various altimetric scenarios
assimilation. WHM1 on a 10 days orbit. WHM2 on
a 21 days orbit.

• 10 days WHM always better than 20 days WHM.
  Error reduction
• Down to 30 on SSH and Velocity (max control 23)
• Down to 75 on heat content (max control 72 !)
• Limits
• Focus on Tropical Atlantic
• Usual limits of OSSEs
• In particular of DA technique
• but here we account for model errors (wind)
  contrary to most twin exp.)
• Roll error relatively easy to be accounted for
  with data assimilation

Results on Heat Content
Heat content error () for various altimetric
scenarios assimilation.