A Bayesian Hierarchical Model (BHM) is developed to estimate surface vector wind fields (SVW),and associated uncertainties, over the Mediterranean Sea. The BHM-SVW incorporates data-stageinputs from analyses and forecasts of the European Centre for Medium-Range Weather Forecasts(ECMWF) and from the QuikSCAT data record. The process model stage of the BHM-SVWis based on a Rayleigh Friction Equation model for surface winds. Dynamical interpretationsof posterior distributions of the BHM-SVW parameters are discussed. Ten realizations from theposterior distribution the BHM-SVW are used to force the data assimilation step of an experimentalensemble ocean forecast system for the Mediterranean Sea in order to create a set of ensembleinitial conditions. Ensemble initial condition spread is quantified by computing standard deviationsof ocean state variable fields over the 10 ensemble members, driven by 10 realizations from theBHM-SVW posterior distribution over a 14-day sequential data assimilation period. Ensemblespread occurs on mesoscale time and space scales, in close association with strong synoptic scalewind forcing events. A companion paper compares the performance of the MFS ensemble forecastsgiven initial condition generation and forecast forcing from the BHM-SVW, with forecasts basedon more traditional methods of ensemble generation
展开▼
