Impact of Argo Profiles on Real-time Global Ocean Analyses for GODAE

James A. Cummings
Naval Research Laboratory
Monterey, California 93943

In the absence of in situ data at depth, subsurface properties in the ocean must be inferred from surface-only measurements. The most important observing system for this purpose is satellite altimetry. Assimilation of altimeter sea surface height anomalies (SSHA) in real-time systems is typically done using one of two alternative methods: (1) assimilation of temperature profiles computed using climatological relationships between SSHA and temperature at depth derived from the historical profile archive (synthetic BT method), or (2) correction of the model forecast density field using differences between the model SSHA field and the SSHA field measured by the altimeter (direct method based on Cooper-Haines).

In this study, Argo profile data are used to quantify the advantages and disadvantages of the two assimilation methods. Synthetic temperature-salinity (TS) profiles computed using Argo profile SSHA and altimeter SSHA at the float location and sampling time are compared with the Argo TS profile to determine skill of the synthetic BT method. Skill of the direct method is investigated by comparing Argo float density corrections computed using differences between SSHA from successive Argo cycles and analyzed altimeter fields with the TS profile from the next cycle of the float. The Argo and analyzed altimeter SSHA differences are used as proxies for model error, and the initial Argo float profile provides a perfect model forecast for the direct method. Regional and global results from the comparisons of the altimeter assimilation methods will be discussed. The impact of the source of the SSHA predictor variable (altimeter or in situ) is used to diagnose some limitations of altimeter SSHA to infer subsurface TS properties. These limitations are directly related to the impact of assimilating Argo profiles in real-time ocean forecast systems being implemented as part of GODAE.