Monthly average surface wind fields, and related quantities, over the tropical Pacific Ocean (29°N-29°S, 122°E-290°E) are compared for three products: the objective FSU winds (FSU2), the NCEP/NCAR Reanalysis (NCEPR1), and the NCEP/DOE AMIP-II Reanalysis (NCEPR2). Accurate surface wind fields are essential for ocean modeling studies and comparisons are needed to understand biases in the available products. Previous comparisons to ship and satellite measured winds have shown the NCEPR1 winds to be underestimated. The analysis to be presented shows the NCEPR2 be an improvement over NCEPR1 for the tropical Pacific Ocean.

A wide range of comparisons of the winds, divergence, and pseudo-wind stress curl will be presented. Long-term monthly wind averages from 11ºN-11ºS, 122°E-290ºE reveal large differences in the three products. The FSU2 winds are stronger (5.6 ms-1) than either the NCEPR1 (4.1 ms-1) or the NCEPR2 (4.6 ms-1). These results show that NCEPR2 improvements in forecast model and data assimilation system have produced better estimations of the surface winds in this region. The ITCZ and the SPCZ are found to be sharper in the mean wind divergence fields in the FSU2 and NCEPR2 as compared to the NCEPR1. In addition, interannual and seasonal signals are stronger in the FSU2 product than in the reanalyses, in particular the ENSO anomalies.