With the development of computer technology,the application of high resolution numerical ocean model becomes available.And it's absolutely vital to improve the prediction effectiveness that it uses a better data assimilation system because of the absence of observation and bad model-only results.The four-Dimensional Variational assimilation(4DVar)method can assimilate all the observations in the time window which plays an important roles for the forecasting system can improve the predictability.This paper builds the general process of ocean data assimilation aiming at the key technology problems appear when constructing the region ocean data assimilation system at South China Sea,based on the Region Ocean Model System(ROMS).The modular design takes advantages to extend the system application.After that,a series 4DVar testing implemented with three different methods is taken to analysis the stability and performance at the domain South China Sea,using the same quality control observation and spatial filtering covariance method.Then take the numerical experiment with the increment method of 4DVar(I4DVar)to simulate a warm-eddy appears in 2010 at South China Sea,which finds good expression when compare the analysis field with HYCOM data and XBT data. |