Remote and local forcing of tropical linear Rossby waves

Rossby waves are the large scale mechanism of adjustment to perturbations in geophysical fluids. In the ocean, they can be generated locally by the wind stress curl or by perturbations of the thermocline level in the eastern boundary. The latter will be referenced as remote forcing and can be due to wind stress variations or integrated basin-wide stochastic wind forcing.

Satellite altimeters provided sturdy evidence of the existence of these waves, through measurements of sea level anomaly (SLA). However, the original interpretation of westward propagating features as Rossby waves has been challenged by recent studies, which suggest the observed variability in the altimetric database is mostly due to mesoscale nonlinear eddies.

This work aims to test whether classic linear dynamic is a reasonable explanation for observed SLA and to verify which kind of forcing, local or remote, is more relevant in the genesis of theses waves.

A linear, reduced gravity, non-dispersive, annual Rossby wave model was used to estimate the SLA forced by satellite-derived Ekman pumping. A zero-pumping eastern boundary condition has been tested as well as a data based one. The latter is a proxy for the remote forcing. Comparisons were made with both unfiltered and filtered SLA from AVISO altimetric records.

The model results agree better with observations in the tropical region of all ocean basins. Correlations between them are up to 0.88. In mid-latitudes, the correlations become insignificant. Nevertheless, a significant part of SLA annual variability is explained by linear Rosby wave dynamics in the tropical oceans.

The relative contributions of eastern boundary forcing and local wind forcing in the generation of Rossby waves is estimated. Forcing at eastern boundary seems to be more relevant for the formation of linear long baroclinic Rossby in most ocean basins. The exception is the western Pacific ocean, were local forcing yields better results.