THE ASSIMILATION OF MARINE SURFACE DATA INTO THE NAVY OPERATIONAL GLOBAL ATMOSPHERIC PREDICTION SYSTEM

As numerical models of the ocean and the atmosphere become more sophisticated, it becomes necessary to account for the interchange of energy at the air/sea interface. Meteorological models require accurate estimates of sea surface temperature and ice fields. Oceanographic models require accurate estimates of the surface winds and heat fluxes. Thus, it is beneficial for scientists in each discipline to have some level of understanding and appreciation for how these various parameters are obtained. With that in mind, we will describe how a meteorological analysis/forecast system uses marine data to estimate the surface wind field, with particular emphasis on the analysis process and the resulting nowcast product. The Navy's data analysis system uses the multivariate optimum interpolation technique to produce updates to background fields provided by the forecast model. The advantage of this technique is that the relative weights given the background and the observations are determined by taking into account the estimated error properties of each. As new data sources become available, the appropriate statistics defining the errors of the observations must be determined. Since we have recently added several new data sources to the assimilation cycle, new values of the observation errors have been determined for all marine surface data. The improvements made by these new data sources, including special sensor microwave/imager wind speeds and synthetic tropical observations, will also be illustrated.