The northern Indian Ocean is unique in a number of respects. First, Ekman fluxes cause coastal upwelling, but western boundary current dynamic processes cause temperatures of upwelled water to be much colder than they might otherwise be; the resultant strong cooling of the western Indian Ocean probably has a strong influence on the Asian monsoon. Secondly, observed cross-equatorial WBC flow is essentially always northward in 100-500m; this flow into the (closed) northern Indian Ocean via the WBC must change its density before it can permanently escape, because it is trapped in a vorticity trap. The only escape routes are via the southward surface Ekman transport, and via the WBC (at some depth where southward flow occurs). Finally, seasonal flows are so vigorous that water at 100m depth crosses the equator and upwells to the surface at 10N, all during the summer monsoon; the winter monsoon can carry (warmed) water back across the equator, to create a seasonal ″eddy flux″. These facts have strong implications for the heat budget of the northern Indian Ocean.
We performed two runs of a global, coarse-grid ocean model: one with full annual and interannual wind stresses and heat fluxes, and one in which wind stresses (decoupled from wind speeds) are replaced by their 12-month running means (12MRM). The two runs have very similar 12-month running mean horizontal and overturning streamfunctions; yet the 12MRM run is much simpler, being virtually in Sverdrup balance at all times. Seasonal SSTs warm by up to 4C near Somalia in the 12MRM run relative to the control, implying less heat absorption in the 12MRM run. The dynamics of heat transport in the simple 12MRM run are analysed, and contrasted with those in the strongly seasonal control run. In particular we examine the size and mechanisms of seasonal eddy fluxes of heat, within our (admittedly coarse) model. |
|
|