Towards an operational satellite-based analysis/prediction system using ROMS: An example for the East Australia Current.

Javier Zavala-Garay, John Wilkin, and H. Arango
Institute of Marine and Coastal Sciences, Rutgers the State University of New Jersey.

Assimilation of satellite observations into ocean models is a subject of intense research because of the obvious advantages for operational forecasting and hindcast analysis. A common procedure for incorporating such observations (mainly SST and altimeter data) into the numerical models is the use of empirical relationships, sometimes referred to as “Synthetic CTD”, that project to the subsurface what is seen by the satellites. The best examples where this procedure has been proved to work is in the areas of intense mesoscale variability associated with western boundary currents, such as the East Australia Current (EAC). However, the obvious limitation of such approach is the need of a large number of insitu observations (e.g., XBT and CTD soundings) from which a robust unbiased empirical relationship can be derived. In this talk we will introduce an alternative approach for vertically projecting the satellite information that is based on basic large scale dynamical balances. Comparison of this alternative approach for the EAC suggest that this technique can provide the same kind of information as that of the Syntetic CTD, and therefore having a great potential for improving mesoscale prediction in unsampled regions of the world ocean.