The analysis is performed with the Ensemble Filter as described in Compo et al. (2006).Observations of surface pressure and sea level pressure from the International Surface Pressure Databank version1.1 and ICOADS version 2.4 were assimilated every six hours. The short-term forecast ensemble is generated in parallel from 56 9-hour integrations of a state-of-the-art atmospheric general circulation model, the atmospheric component of NCEP's operational Climate Forecast System model (Saha et al. 2006). Briefly, the model has a spatial resolution of nearly 200-km on an irregular Gaussian grid in the horizontal (corresponding to a spherical harmonic representation of model fields truncated at total wavenumber 62, T62). In the vertical, for efficiency, we have reduced the resolution from a finite differencing of 64 levels to 28 levels with no detrimental effects to the reanalysis quality. The model top is at 0.2 hPa. The model has a complete suite of physical parameterizations as described in Kanamitsu et al. 1991) with recent updates detailed in Moorthi et al. (2001). Additional updates to these parameterizations, specific to this version of the model, are described in Saha et al. and include revised solar radiation transfer, boundary layer vertical diffusion, cumulus convection,and gravity wave drag parameterizations. In addition, the cloud liquid water is a prognostic quantity with a simple cloud microphysics parameterization. The radiation interacts with a fractional cloud cover that is diagnostically determined by the predicted cloud liquid water. The specified boundary conditions needed to run the model in atmosphere-only mode are taken from the time-evolving sea surface temperature and sea ice fields of the HadISST1.1 dataset obtained courtesy of the United Kingdom Met Office Hadley Centre Rayner et al. 2003).