AbstractA high‐resolution global daily analysis of ocean surface vector winds (1987 onward) was developed by the Objectively Analyzed air‐sea Fluxes (OAFlux) project. This study addressed the issues related to the development of the time series through objective synthesis of 12 satellite sensors (two scatterometers and 10 passive microwave radiometers) using a least‐variance linear statistical estimation. The issues include the rationale that supports the multisensor synthesis, the methodology and strategy that were developed, the challenges that were encountered, and the comparison of the synthesized daily mean fields with reference to scatterometers and atmospheric reanalyses. The synthesis was established on the bases that the low and moderate winds (<15 m s−1) constitute 98% of global daily wind fields, and they are the range of winds that are retrieved with best quality and consistency by both scatterometers and radiometers. Yet, challenges are presented in situations of synoptic weather systems due mainly to three factors: (i) the lack of radiometer retrievals in rain conditions, (ii) the inability to fill in the data voids caused by eliminating rain‐flagged QuikSCAT wind vector cells, and (iii) the persistent differences between QuikSCAT and ASCAT high winds. The study showed that the daily mean surface winds can be confidently constructed from merging scatterometers with radiometers over the global oceans, except for the regions influenced by synoptic weather storms. The uncertainties in present scatterometer and radiometer observations under high winds and rain conditions lead to uncertainties in the synthesized synoptic structures.