Notes from literature concerning the relationship between SSHA and density structure of the ocean

• Guinehut et al. (2006):
  • Monthly mean dynamic height anomaly (DHA) and sea level anomaly (SLA) well correlated but there are differences
  • These differences are due to either the barotropic (non-steric) component or due to the baroclinic (steric) component below the level of reference used to compute the DHA (700 m in this paper).
  • They explain that the main difference in the Pacific Ocean is due to the Sverdrup flow that extends below the level of reference.
• Stammer (1997)
  • Decompose the SSH into a component due to atmospheric buoyancy forcing, a component due to baroclinic change in currents below the surface mixed layer and a component due to barotropic advection of density fields (see Gill and Niiler 1973)
  • The dominant component on basin scale and on annual to interannual times scale is the one due to atmospheric buoyancy forcing, especially at mid-latitudes.
• Dhomps et al. (2010)
  • Mesoscale and intraseasonal component defined as signal with a period shorter than 200 days and annual and interannual component with period longer than 200 days
  • The correlation between DHA and SLA decreases (from 0.84 to 0.7) keeping only the mesoscale and intraseasonal component and increases slightly (from 0.84 to 0.87) using the annual and interannual component.
• Ideas drawn from this literature
  • Study the relationship between SSH and DA over several years but at one single point
  • Study the equation of SSH in the OFES model
  • Remove the signal due to atmospheric forcing or see how Chelton does to isolate the mesoscale eddies
  • Can the difference between ARGO DHA and AVISO SSHA be indicative of unresolved mesoscale and submesoscale features by AVISO SSHA? One of the difference is due to the non-zero flow below the level of reference. It is likely that this difference dominates anything else, except in region, like the Tropics, where most of the flow is surface trapped and the assumption of level of no motion is relatively valid.
  • Use the 1/30th-of-a-degree OFES snapshot to see if we can deduce the mesoscale and submesoscale features in SSH
  • Use the ARGO-deduced velocity at 1000 m to correct the DHA estimate.