An Embedding Method for Improving Interannual Variability Simulations in a Hybrid Coupled Model of the Tropical Pacific Ocean–Atmosphere System

Abstract

An embedding approach is developed and tested to improve El Niño?Southern Oscillation (ENSO) simulations in a hybrid coupled model (HCM), focusing on the ocean thermocline effects on sea surface temperature (SST) in the eastern equatorial Pacific. The NOAA/GFDL Modular Ocean Model (MOM 3) is coupled to a statistical atmospheric model that estimates wind stress anomalies based on a singular value decomposition (SVD) of the covariance between observed wind stress and SST anomalies. Analogous to the Cane?Zebiak (CZ) coupled model, a separate SST anomaly model is explicitly embedded into the z-coordinate ocean general circulation model (OGCM). The three components exchange predicted anomalies within the coupled system: The OGCM provides anomalies of ocean currents in the surface mixed layer and the thermocline depth, which are used to calculate SST anomalies from the embedded SST model; wind anomalies are then determined according to the statistical atmospheric model, which in turn force the OGCM. Results from uncoupled and coupled runs with and without the embedding are compared. With the standard coupling, the system exhibits similar behavior to previous HCMs, including interannual variability with a dominant quasi-biennial oscillation and a westward propagation of SST anomalies on the equator. These characteristics suggest that the horizontal advection is playing a more important role than the vertical advection in determining SST changes over the eastern equatorial Pacific. Incorporating the embedded SST anomaly model, with which the thermocline effects on SST can be enhanced in the eastern equatorial Pacific, has a significant impact on performance of the HCM. The embedded HCM exhibits more realistic SST variability and coupled behavior, characterized by 3?4-yr oscillations and a more standing SST pattern along the equator. The results support the hypothesis that current physical parameterizations in the OGCM provide insufficient thermal linkage between the thermocline and the sea surface in the eastern equatorial Pacific. It is demonstrated that the long-known deficiency of some OGCMs in their depiction of the thermocline and its interactions with SST may contribute to unrealistic coupled variability in HCMs of ENSO. The embedding approach not only provides a diagnosis for parameterization deficiencies in current OGCMs but, pending progress on this difficult problem, provides a straightforward means to bypass it and improve coupled model performance.