Impact of Sea Level Assimilation on Salinity Variability in the Western Equatorial Pacific

Abstract

In the primitive equation model for the tropical Pacific at the National Centers for Environmental Prediction, subsurface temperature observations are assimilated. The addition of TOPEX/Poseidon sea level observations to the NCEP assimilation scheme has resulted in large differences in sea level during 1996. These differences are suggested to be related to salinity variability. A bivariate assimilation scheme is presented that corrects both temperature and salinity. The method is tested with synthetic data in an identical triplets experiment, in which a westerly wind burst is simulated. In this experiment, the correction of salinity improves the density simulation and attenuates errors in salinity. A four-year assimilation experiment with real data is performed to test the system?s performance for 1993?96. In this experiment, the assimilation of TOPEX/Poseidon observations improves dynamic height simulation without degrading the temperature field. This application of altimetry improves the mean salinity in the western equatorial Pacific and leads to a more pronounced salinity variability in the ocean model.