Ocean Data Assimilation for Operational Ocean Forecasts
Naval Research Laboratory, MS, Ocean Sciences
Ocean data assimilation for operational ocean forecasts encompasses several planned and ongoing research efforts. Observing systems strategies projects seek to enable more effective observation targeting by matching the mobility of various observing platforms (ocean gliders, other UUVs, AUVs, shipboard, and airborne observing systems) with ensemble or variational-based estimates of observation impact to inform asset allocation and maximize observing system performance. Observation impact is further enhanced through properly designed error covariances that relate an observations information content with the representativeness of the model background and other observations. The Calibration of Ocean Forcing with satellite Flux Estimates project seeks to use in situ and remote sensing to calibrate air-sea fluxes and their uncertainties as a step toward fully coupled ocean-wave-atmosphere-ice-modeling. The adjointless 4DVAR project is developing techniques for variational assimilation that rely on appropriately designed ensembles rather than an explicit adjoint code. Other research within ocean data assimilation places particular emphasis on relating relatively abundant surface information from remote sensing to sparsely observed subsurface conditions. Multisensor analyses of sea surface height, temperature, and salinity are adapted to encompass new sensors and various biases. Covariances control observation assimilation within 3DVAR/4DVAR/ensemble assimilation schemes with weights applied to jointly optimize state variables and their gradients. This assimilation research is applied within global and regional ocean and coupled air/wave/ice/bio/ocean models.