Archived page: information on this page is no longer updated and may contain broken links and outdated information.

CSIRO Marine and Atmospheric Research
Past Seminars

Seminar Abstract

What sets the heat flux into the tropical Indian Ocean?

Climatologies all agree that there is a substantial net heat flux into the tropical Indian Ocean. This implies that SST is colder than it would be if there were no heat flux, with consequent effects on evaporation and hence on Asian monsoon rainfall. However, the magnitude of the net heat flux varies widely among climatologies; and all ocean models published so far require heat fluxes which are below most climatologies.

The oceanographic problem is made more complex by the massive size of seasonal cycle of wind stress, and hence of ocean currents. To simplify the problem, we split it into two by comparing two runs of a coarse-grid global OGCM — one ("control") with full seasonal and interannual variations of all relevant atmospheric variables, and a second ("12MRM") in which the wind stress (only) is replaced at each point and time by its 12-month running mean. Currents in the latter run are relatively slow, and in Sverdrup balance.

The difference in Indian Ocean area-mean heat flux between the runs, north of the Indonesian Throughflow at 7°S, is surprisingly small (6/watts/m2 compared to 28 w/m2 in the Control run). This encourages the idea that the tropical Indian Ocean heat flux is primarily driven in the model by the Levitus (1988) mechanism — southward annual mean Ekman transport is replaced by a colder, deeper geostrophic inflow, demanding a surface heat flux to replace the southward heat transport. The longterm mean difference flow between the two runs is only 1.3 Sv, but it enters at depths below about 500m, and it warms by some 16°C before exiting in the surface mixed layer. This is about enough to advect the interrun difference in heat flux southward. Thus deep-reaching differences in diapycnal mixing are crucial to understanding the interrun difference. Qualitatively, this interrun difference in annual mean mixing strength and depth comes about because currents in the seasonal run are stronger, giving rise to more mixing — a kind of "rectification" of the seasonal currents. However, the details are rather obscure, due to limitations of the model used. Another feature of these two runs is that interesting interrun flow differences occur across the equator; we suggest that the observed annual mean eastward jet along the equator in the top 100m is primarily driven by nonlinear effects associated with the seasonal Wyrtki Jets.

To explore such issues without the coarse-grid limitations, we also studied the heat flux problem in a fine-grid, idealised model of the tropical Indian Ocean. This consisted of a cross-equatorial rectangular box with a gap for the Indonesian Throughflow, and a shallow outlet at the southern boundary to let Ekman flow pass out of the region. Winds are idealised to have Ekman and Sverdrup transports be equal; in most runs we impose a steady 10 Sverdrups upwelling at the northern wall. We find that the net heat flux into this tropical "Indian Ocean" is remarkably insensitive to parameterisation of diapycnal mixing, because a simple nonlinear constraint places a lower limit on the possible depth of the geostrophic inflow. For a 10 Sv cross-equatorial Ekman flow, this lower limit is about 200m. It seems that diapycnal diffusivities have to be very high before the model can depart much from this minimum depth. Yet if it is close to that limit, our experience so far suggests that the mixing is probably not physically realistic, but is instead generated by numerical artefacts of one kind or another.

If time permits, we will briefly explore some of the issues raised by these results, and discuss preliminary attempts to circumvent these problems.

[Back to Seminars]

CSIRO = Marine Laboratories Auditorium, Castray Esplanade, Hobart

For further information, or to schedule a seminar, contact:
Peter Oke, (Oceanographic seminars) CSIRO Marine Research (03) 6232 5387
Piers Dunstan, (Biological seminars) CSIRO Marine Research (03) 6232 5382
Katrina Nitschke, Antarctic Climate and Ecosystems CRC (03) 6226 2265 & IASOS, University of Tasmania (03) 6226 2509