The water close to the deep ocean floor is called bottom water. It may be located in deep valleys or trenches. I understand that water flowing to the Arctic will sink there, because it's saltier and therefore heavier than the surrounding water. But once it sinks to the bottom, for example, inside trenches, what causes it to ever leave? As far as I know, there aren't any significant sources of heat down at the ocean floor, like there are at the bottom of the atmosphere, but we still seem to have a highly dynamic system. What causes this bottom water to mix again?

Thermohaline circulation
Figure obtained from University of Washington page


4 Answers 4


The density of a water parcel evolves as it moves in the ocean. A parcel of bottom water (e.g., Antarctic Bottom Water (AABW) the densest water mass) is much denser than the surrounding water when it is formed near the surface (in the case of the AABW in polynyas and below the ice shelf in the Weddell and Ross Seas).

As the parcel of AABW sinks, it flows down the Antarctic continental shelf and slope and moves north along the bottom at speeds of 2-8 cm/s. The issue is that the water parcel of AABW will be interacting and mixing with the surrounding water.

Thus, a parcel of AABW that when it was formed had a temperature of around -0.8°C and a salinity of 34.7, will get mixed with warmer water masses (e.g., Antarctic Intermediate Water, North Atlantic Deep Water) and as it becomes less dense it will be more easily mixed with the surroundings.

If you have surrounding waters that are more saline and with comparable temperatures, then the less saline water parcel (modified AABW for instance compared to NADW) will rise. The interior circulation of the ocean is predominantly driven by thermohaline processes (differences in density) and the thermal wind equation can be used as a simplification of the dynamics, especially away from the edges (surface, bottom).


I wonder why Eckmann transport is not mentioned. You can have wind flowing parallely to the coastlines, for example, along a NS direction. Water driven by this wind will be subject to coriolis force and deflect sideways, and that will produce a net normal component of water movement to the wind direction. This is Ekmann transport, very simply stated.

This way, surface water is moved away, and that creates a pressure gradient, wo balance which, water sinks where it is driven to, and rises where it is driven from. This is the Ekmann transport driven upwelling / downwelling.

The AABW is subject to antarctic eastward flowing polar winds and the resulting northward Ekmann transport. So the surface water is driven north, and that is creating a pressure gradient, which causes increases pressure in just the north of AABW in Indian / south pacific ocean, and that pressure is as well acting on the - albeit from the side - to the AABW, which causes it to rise.

  • 1
    $\begingroup$ Ekman transport is in fact important for AABW evolution. In fact another factor to consider is bottom Ekman transport. As the flow at the bottom is zero and the AABW is moving at a certain speed, a bottom boundary layer is formed and enhanced mixing occurs in it. This will result on additional changes in the water characteristics of the water mass near the bottom. $\endgroup$
    – arkaia
    May 5, 2014 at 19:07

Some areas may well be relatively isolated.

However water will come up when it is pushed from behind! E.g. there are strong sinks in the Norwegian Sea and Weddell Sea. This water must go somewhere and drives the deep water currents. These currents are pushed from behind and will come up when they hit a steep continental shelf. Such areas of upwelling water are often fertile ocean areas where the nutrient-rich deep waters reach the surface. E.g. offshore Namibia ( NASA photo ).

There are sources of heat underwater, but they tend to be on the flanks of the mid ocean ridges (Black & White Smokers). These are actually the outlets of large convection systems which feed large quantities of ocean water through the young hot upper crust. However, they are relatively shallow and not in the ocean depths (and definitely not in the trenches).

  • 4
    $\begingroup$ Doesn't wind also drive upwelling... I am unsure if the Thermal wind equation is relevant $\endgroup$
    – Neo
    Apr 16, 2014 at 19:01
  • $\begingroup$ The water isn't just pushed up as it's pushed from behind, if the water coming down is less dense than the bottom water it will actually just float on top. I think you underestimate the heat sources at the ocean floor, but that alone wouldn't create such concentrated upwelling sites. As is often the case these are sites at which many factors act together, and an important factor you fail to mention is wind. Large atmospheric circulation cells cause oceanic gyres that also cause upwelling. $\endgroup$
    – hugovdberg
    Apr 16, 2014 at 19:02
  • $\begingroup$ The water coming down is coming down because it is dense. Both of those sea sinks there because the water is cold and saline (ice formation will also increase the salinity). $\endgroup$
    – winwaed
    Apr 16, 2014 at 19:08
  • $\begingroup$ Is the thermohaline circulation hitting a shelf really responsible for up-welling in offshore Namibia? My understanding is that water moving through the oceans due to thermohaline circulation has already risen before meeting namibia, in the indian ocean and thus wouldn't dramatically rise there to create upwelling. $\endgroup$
    – G. Gip
    Jul 19, 2016 at 9:02
  • $\begingroup$ Namibia is on the Atlantic cost of Africa. $\endgroup$
    – winwaed
    Jul 19, 2016 at 12:46

Upwelling of deep ocean water can be driven by the wind.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.