Application Of 4-Dimension Variation Data Assimilation Technique In The Marine Environment Numerical Simulation Of The South China Sea

Data assimilation is a key method to obtain a good initial field for ocean models.And 4-Dimensional Variation Data Assimilation(4D-VAR)is a popular data assimilation method which can combine observation data from different time,different regions and different types.Due to the severe lack of ocean observation in the South China Sea(SCS)region,it is of vital significance to using 4D-VAR technique to improve the numerical simulation level for SCS region.This article illustrated how to establish a 4D-VAR experimental system for the SCS region.A background error covariance matrix B adapting for the SCS region was also constructed.And it was tested by a single experiment with a temperature observation as-similated.The result shows that the B is reasonable.The correlation(C)and multivariate balance operator(K_b),which combine together as B,can transform observation influ-ence by adjusting self-structure in model space and balancing relations between different model state variables respectively.Finally,the satellite altimeter data assimilation was designed and realised in the SCS region 4D-VAR system.And a new method was proposed to evaluate data assimilation performance by taking a vortex-identification algorithm.The experiment shows that the identifying result derived from posterior field had a better performance compared to prior field.Furthermore,it displays a strength for identifying mesoscale eddies as the model owes a higher spatial resolution than S LA observation,which means that it may reveal more accurate structures for mesoscale eddies.