GOAPP Discussion Group, February 24, 2009

Location: 5th floor lounge, Oceanography Building, Dalhousie

Time: 3:30-5:00 pm

Title: Development and evaluation of ice-ocean reanalyses using the S(T) assimilation system

Discussion Leader: Gregory Smith

Abstract:

Abstract: Our focus is on improving the realism of ocean data assimilation schemes and using assimilation to investigate ocean climate signals. Using observations from the Argo array, we demonstrate the benefit of combining temperature (T) and salinity (S) assimilation methods and of assimilating S(T) as the observable property. This allows us to exploit the larger spatial and temporal decorrelations of this quantity, compared with S(z), allowing flow dependent assimilation and recovery of water mass information.

The S(T) algorithm has been implemented into the NEMO global ice-ocean model and two reanalyses have been made: a 50-year reanalysis at 1° resolution, and a 21-year reanalysis at 1/4° resolution. Overall, the assimilation is able to prevent drifts in many ocean metrics, and brings the model in better agreement with accepted values. An evaluation of water mass properties in various ocean syntheses (e.g. ECCO, MERCATOR, ECMWF, SODA) with Argo observations, performed as part of the CLIVAR-GSOP intercomparison, shows that the S(T) reanalyses provide excellent agreement with in situ observations.