Does the cold tongue advect from the Peru coast?

Does the cold tongue advect west from the Peru coast?

Observing the westward expansion of the annual cold tongue in the eastern Pacific, it is tempting to infer that this is due to cold water advected northward along the Peru coast and then westward along the equator in the South Equatorial Current. However, in the Gent/Cane GCM forced with annual cycle winds, this is not the case. The SEC is weak in the far east and zonal gradients of SST are small. The model cold tongue arises primarily from local upwelling along the equator. A test in the model confirms this impression.

The model was run with the addition of a thin barrier running west from the Peru coast along 4.5°S out 700km to 89°W. This weir blocks advection from the main coastal upwelling region onto the equator. Note that the barrier is approximately perpendicular to the local winds, thus generating no additional Ekman up- or downwelling. The following figures compare the run with realistic coastlines to that with the barrier.

Read the abstract and download the paper on the annual evolution of the cold tongue: Kessler, Rothstein and Chen (1997).

Mean differences are small (0.2-0.5°C) except right in the coastal upwelling region.
(This figure shows the location of the barrier as a thick line).
Mean SST with and without the barrier
Although the barrier blocks flow of cool water into the cold tongue, the phase of the annual cycle is almost unchanged. Differences have relatively small scale and high-frequency.
SST one the equator with and without the barrier
The time series of SST in the central cold tongue is very similar with and without the wall.
Cold tongue SST with and without the barrier

Return to main figure list