Pacific Ocean model results show a consistent error especially in the North Equatorial Counter Current (NECC) region. In the band between 5N and 12N RMS differences between model sea level and observed values from TOPEX/Poseidon altimetry are as large as 8 cm and correlations drop below 0.2. It has been hypothesized that sparse grids do not capture variability brought about by Tuwanapec winds that force ocean eddies in the Eastern tropical Pacific (eg. typical forcing grids for this model are 2x2 degrees resolution). Earlier work (Hackert et al., 2001) has shown that the quality of wind forcing plays a key role in reproducing sea level variability along the equator for the 1997-98 El Nino. In particular, the high accuracy afforded by satellite data in resolving the meridional gradient of the zonal winds is key to getting the sea level variability correct. In the current study, the role of the spatial structure of the wind forcing field is presented. The model uses realistic forcing and was spun up over the period 1985-1999 using ECMWF winds. Then, for the period September 1999 until present the model was forced by QuikScat winds using various resolutions (ie. 2x2, 1x1 0.5x0.5, 0.25x0.25) for the wind forcing. Results of these experiments show that QuikScat winds are able to reproduce the high spatial resolution in the real wind field and thus provide the best forcing for the simulation of the observed 4-dimensional circulation of the ocean.