Abstract

The roles of zonal and meridional wind stress and of surface heat flux in the seasonal cycle of sea surface temperature (SST) are examined with a primitive equation (PE) model of the tropical Pacific Ocean. While a variety of previous numerical and observational studies have examined the seasonal cycle of SST in the eastern tropical Pacific, it is noteworthy that different mechanisms have been invoked as primary in each case and different conclusions have been reached regarding the relative importance of the various components of surface forcing. Here, we perform a series of numerical experiments in which different components of the surface forcing are eliminated and the resulting upper ocean variability is compared with that of the climatological experiment. The model used for these experiments reproduces a realistic climatological seasonal cycle, in which SST emerges as an independent quantity.

We find that the different cases all produce qualitatively reasonable seasonal cycles of SST, though only the most complete model is also able to reproduce the seasonal cycle of near surface currents, tropical instability waves (TIWs), and net surface heat fluxes consistent with historical observations. These results indicate that simply reproducing a qualitatively accurate seasonal cycle of SST does not necessarily allow meaningful conclusions to be made about the relative importance of the different components of surface forcing. The results described here also suggest that a model simulation must at least reproduce all the documented near surface kinematic features of the equatorial Pacific cold tongue region reasonably well, before accurate inferences can be made from model experiments. This provides useful guidelines to current efforts to develop and evaluate more complex fully coupled air-sea models and shows that results for simple or intermediate ocean models that do not have this level of fidelity to the observations will be difficult to interpret.