Welcome to Adobe GoLive 6

The Global Ocean Data Assimilation System (GODAS) provides near real-time operational analyses of the ocean state in at the U.S. National Centers for Environmental Prediction (NCEP) where it is used to initialize the ocean component of coupled seasonal to interannual climate forecasts. The GODAS is built on the GFDL Modular Ocean Model, Version 3, (MOMv3) and a three-dimensional variational assimilation scheme. In its operational configuration, the GODAS assimilates temperature profiles from sources that provide data in near real-time (XBTs, TAO/TRITON/PIRATA moorings, Argo floats).

The GODAS also assimilates synthetic salinity profiles which are computed from each temperature profile and the local T-S correlation. Here we report on the impact of the Argo float network on the GODAS analyses. The Argo network has grown steadily over the last 5 years and in 2003 the number of Argo profiles used in each GODAS analysis exceeded the number of XBT profiles. With its relative freedom from the constraint of commercial ship tracks, salinity profiling, and deep profiling capability, the Argo network, however, is more than a simple replacement for the XBT network.

Using the standard GODAS, some reconfigured versions of GODAS, and a CONTROL that assimilates no data, we will look at the impact of Argo on the GODAS in 3 ways.

1) Given the current availability of XBT profiles, what is the impact of withholding Argo profiles? Likewise, what is the impact of withholding XBT profiles?
2) What is the impact of replacing synthetic salinity profiles as used in the standard GODAS with observed Argo salinity profiles?
3) In the standard GODAS data are assimilated only into the upper 750 meters, a depth corresponding to the maximum depth of many XBT profiles. What is the impact of taking advantage of the greater depth of Argo profiles by extending the assimilation depth of GODAS to 1500 meters?

All of the model results will be evaluated by cross-validation and comparisons with independent datasets.