Operational multivariate ocean data assimilation

A fully three‐dimensional, multivariate, optimum‐interpolation ocean data assimilation system has been developed that produces simultaneous analyses of temperature, salinity, geopotential and vector velocity. The system is run in real‐time, and can be executed as a stand‐alone analysis or cycled with an ocean forecast model in a sequential incremental update cycle. Additional capabilities have been built into the system, including flow‐dependent background‐error correlations and background‐error variances that vary in space and evolve from one analysis cycle to the next. The ocean data types assimilated include: remotely sensed sea surface temperature, sea surface height, and sea‐ice concentration; plus in situ surface and sub‐surface observations of temperature, salinity, and currents from a variety of sources, such as ships, buoys, expendable bathythermographs, conductivity–temperature–depth sensors, and profiling floats. An ocean data quality‐control system is fully integrated with the multivariate analysis, and includes feedback of forecast fields and prediction errors in the quality control of new observations. The system is operational at the US Navy oceanographic production centres both in global and in regional applications. It is being implemented as the data assimilation component of the Hybrid Coordinate Ocean Model as part of the US contribution to the Global Ocean Data Assimilation Experiment, and in a limited‐area ensemble‐based forecasting system that will be used in an adaptive sampling, targeted observation application. Copyright © 2005 Royal Meteorological Society