Role of assimilation of ARGO data in the improvement of tropical Pacific Ocean simulations

Eric Hackert(1) , Joaquim Ballabrera-Poy(1) , Antonio Busalacchi(1) , and Raghu Murtugudde(1)

(1) Earth System Science Interdisiplinary Center, 2227 CSS BLD. (BLD 224), College Park, MD 20842-2465, United States

Abstract
A series of data assimilation experiments is used to investigate the impact of ARGO data to constrain tropical Pacific dynamics and thermodynamics of the reduced-gravity, sigma-coordinate, ocean model of Gent and Cane (1989). In these experiments, subsurface data (Tz, Sz) are assimilated with satellite-derived fields of sea level (SL), obtained from TOPEX/Poseidon/Jason1, and sea surface temperature (SST) from Reynolds and Smith (1994). Observations are assimilated using a Reduced-Order Kalman Filter (ROKF) algorithm method built on the multivariate EOFs of the model. These MEOFs statistically project the information of the data correction to all the variables of the model including salinity. This study focuses on the added value of ARGO data with respect to conventional subsurface data (XBT, CTD, BATHY, and other instruments) from the Global Temperature-Salinity Profile Program (GTSPP) database. In this manner, the relative contribution of ARGO versus other subsurface observations (both Tz, Sz) can be isolated.