We review the development and implementation of a Bayesian Hierarchical Model (BHM) for surface wind fields in the tropical Indian and western Pacific Oceans. Fifty realizations of the sur- face wind process are stored, at 6-hourly intervals, on a 0.5 degree grid, for the period July 1999 through 25 December 2003. Two research vignettes demonstrate the utility of ensembles of surface winds from the BHM with respect to atmosphere and ocean signals of a single Madden-Julian Oscilla- tion (MJO) active phase from November 2000. In one, the ensemble daily averages of surface convergence field realizations reduce the noise in these derivative fields such that a correlation with out-going longwave radiation becomes evident on daily timescales. In the second, an Ocean General Circulation Model (OGCM) ensemble response to forcing by BHM winds can be used to map regions of model uncertainty during the MJO and associated westerly wind burst. The emphasis of the talk is on methodology and observations with respect to organized processes in the tropics. This leads to the introduction of a new satellite mission concept that is germane to a wide-variety of research and operational applications.