Analysis of GTSPP data assimilation in BCC_GOADS2. 0

The model and assimilation scheme of the second generation Global Oceanic Data Assimilation System of Beijing Climate Center( BCC_GODAS2. 0) are introduced in this paper. Two experiments are designed to test the effect and stability of the assimilation system. One experiment is assimilated with the GTSPP observation from 1990 to 2008,while the other is performed without any observation in the same period. Comparison of the experiment results with the OISST and SODA( Simple Ocean Data Assimilation) datasets indicates that the assimilation of GTSPP can improve the simulation of sea surface temperature( SST) and sea surface salinity( SSS). In detail,the error of SST and SSS in the ocean,especially in the tropical Pacific,is reduced effectively. Moreover,the experiment with assimilation can better represent the temporal evolution of SST in the Nino3 and Nino4 regions as well. Also the assimilation amends the simulation in vertical. Specifically in the mixed layer near 200 m,the root mean square errors of temperature and salinity are decreased by 1. 5 ℃ and 0. 6 psu,respectively.