Begin Main Content

Argo salinity data are downloaded from the USGODAE Argo GDAC and processed through several quality control parameters. The data are passed through a density inversion check where profile levels with a density inversion greater than 0.05 kg/m3 are rejected. The salinity profiles are then compared to the Levitus mean climatology and profiles differing from the mean by more than 2 in the upper 50 meters and by more than 1 below 50 meters are rejected. A 'superob' technique is performed when two Argo profiles occure on the same day and are less than 0.5 degrees apart spatially. The RMS difference between the two profiles is calculated and if the RMS is less than 0.1, the two profiles and their positions are averaged. If the RMS is greater than 0.1, then the profile with the most number of levels is accepted. In the depth direction, levels of data that are less than 10 meters apart are averaged and data is interpolated to ensure a data level exists every 50 meters.

Argo salinity data are assimilated with synthetic profiles developed for temperature profiles from XBTs and moorings using the T-S from Levitus-Boyer (1994) climatology. Synthetic salinity "observations" are assimilated with an observation error that is 200% of the background standard deviation while the Argo salinity observations are assimilated with an observation error that is 120% of the background standard deviation. The covariance scales used are the same as those for temperature in the OI and with the derived spatially varying scales in the MvOI or EnKF.

Argo Salinity Distribution Plots

Date	Global Equatorial	Global
2004
2003
2002
2001
2000

Evaluation of Salinity Analyses Against Argo Data

Mean and RMS Plot by Basin for 2002
Argo Salinity Positions for 2002

XBT03_TS Assimilation Plots

Daily Plots and Profile Plots for the Pacific and Indian Ocean