Impact of observed temperature and salinity assimilation in Global Ocean Data Assimilation System (GODAS) using MOM5 

Abstract

The National Centers for Environmental Prediction (NCEP) produces operational ocean reanalysis products based on the Global Ocean Data Assimilation System (GODAS) with the Modular Ocean Model (MOM3) as a physical model and 3D-Var as the data assimilation method. Subsequently, the Indian National Centre for Ocean Information Services (INCOIS) has also operationalized the same GODAS system with MOM4p0d as a physical model to produce ocean analysis since 2013. In this study, we upgraded GODAS with MOM5 to produce ocean reanalysis with NCEP-R2 atmospheric forcing. The improved reanalysis is compared with respect to the EN4 analysis and Ocean Reanalysis System 5 (ORAS5) for the subsurface temperature and salinity. The microwave-based satellite-derived sea surface temperature (SST) is employed for the independent evaluation of the reanalysis SST products. We have assimilated observed temperature and salinity profiles from all in situ platforms over the global ocean and assimilated along-track sea level anomaly from altimeter (Jason1 and Jason2) to produce the improved reanalysis. There is a significant improvement in the sea surface temperature (SST) with biases of temperatures reduced from 1.5 °C to ~0.2 °C as compared to without assimilation. Moreover, a significant improvement is found in the subsurface temperature and salinity fields over the northwest Atlantic and Pacific Ocean, Nino 3.4, and Indian Ocean thermocline ridge regions as well. The biases in this new, improved reanalysis are even lower than ORAS5 when compared with EN4 analysis. Thermocline depth also shows improvement in terms of better representation and capturing seasonal variability. We also show the improvement in the new reanalysis with respect to the old reanalysis based on MOM4p0d.