Impact of an upgraded model in the NCEP Global Ocean Data Assimilation System: The tropical Indian Ocean

Abstract

The National Centers for Environmental Prediction (NCEP) and the Indian National Centre for Ocean Information Services (INCOIS) produces global ocean analysis based on the Global Ocean Data Assimilation System (GODAS). This study shows how upgrades to the forward model simulations from MOM4p0d to MOM4p1 impact ocean analyses over the tropical Indian Ocean in GODAS. Three experiments were performed with same atmospheric forcing fields: (i) using MOM4p0d (GODAS_p0), (ii) using MOM4p1 (GODAS_p1), both using observed temperature and synthetic salinity, and (iii) using MOM4p1 (GODAS_p1S) assimilating both observed temperature and observed salinity. Validation with independent observations shows significant improvement of subsurface temperature and salinity in the new analysis using MOM4p1 versus MOM4p0d. There is also improvement in the upper ocean current of the equatorial Indian Ocean. The impact of observed salinity on the upper ocean surface current is marginal, but there is significant improvement in the subsurface current. The seasonal and inter-annual variability of the Wyrtki jet and the equatorial undercurrent is improved in GODAS_p1 versus GODAS_p0. All analyses reproduced the Indian Ocean dipole, with the GODAS_p1S simulated sea surface temperature (SST) the most accurate. The temperature inversion over the north Bay of Bengal (BoB) is reproduced only in GODAS_p1S. The mean sea level over BoB and equatorial Indian Ocean improved in GODAS_p1S as compared with AVISO observation. The combined model upgrade and assimilation of observed salinity led to reduced root-mean-square deviation and higher correlation coefficient values in the sea level anomaly (SLA) when compared with satellite observations.