Preparing NCEP's Global Ocean Data Assimilation System to make full use of available Satellite Data

The Global Ocean Data Assimilation System (GODAS) used at the National Centers for Environmental Prediction (NCEP) has recently been upgraded to a hybrid data assimilation scheme using the Local Ensemble Transform Kalman Filter (LETKF) in conjunction with the 3D-Var approach currently used in the operational GODAS. The ensemble component of this scheme allows new observational data sources to be included with greater ease than has been possible with previous systems. We are now preparing GODAS for the assimilation of all available satellite-based observations of the ocean surface, including sea surface temperature, height anomalies (e.g. Jason-1; Envisat, Jason-2 and Cryosat-2), and salinity data (Aquarius and SMOS). We allow these data sources to impact the atmospheric surface forcing fields that are applied to the ocean forecast model via a recursive filter. In addition, the GRACE satellite data provide constraints on the vertically integrated temperature and salinity fields by measuring bottom pressure. While such global constraints present difficulty within the localization approach used by LETKF, they can be implemented directly through the 3D-Var component of the hybrid assimilation scheme. A particularly important element of data assimilation when assimilating numerous data sources is the estimation of observation errors. We address an approach for estimating these errors.