Mr. Tom's Blog

The ultimate iso-isopycnal dispersion experiment?!

In a previous blog I wrote that it was good to know that it wasn’t crucial if isopycnal floats didn’t settle on exactly the same density surface since once they were more than O(10) km apart, lateral shear would dominate over vertical shear and control their subsequent rates of separation. This because the lateral shear is controlled by the mesoscale velocity field, which is strongly coherent in the vertical, meaning it doesn’t matter exactly which isopycnal a float is on. But suppose you really wanted to know more about lateral dispersion in an isopycnal layer, i.e. in a layer in which vertical shear is ‘guaranteed’ absent. In that earlier study we could minimize but not guarantee absence of vertical shear by restricting ourselves to floats that were within 0.1°C of each other. But is there such a place? I think so, it’s an absolutely fascinating area.

East of Barbados there is an extensive region where temperature and salinity in the main thermocline don’t decrease more or less gradually across the main pycnocline but instead in distinct steps – what are called sheets and layers – where T and S jump across thin ‘sheets’ or interfaces between thicker nearly uniform layers. The sheets are but a few meters thick and the layers 10-20 m thick. The pycnocline is of course stably stratified, but we must thank temperature for that for it is strongly unstable with respect to salt, this because tropical Atlantic water is far saltier than the fresher Antarctic Intermediate Water at depth. While the salt is held up by the thermal stratification some salt does leak down through the pycnocline. It does so though a curious arrangement that is connected to its much lower diffusivity than that of heat. Consider a layer of warm salty water floating on top of fresh cold water. Now imagine a little downward bump of the interface. The salt in the bump will remain in it, but its temperature will cool slightly as it gets exposed to the cooler water of the lower layer. As a result the bump will get heavier since it is still salty but cooler; so it starts to sink, and it does so in the form of a narrow pipe or ‘finger’. Collectively, these are known as salt fingers. You can demonstrate them in a kitchen sink experiment. In the ocean these fingers flux salt across thin sheets into thicker homogeneous layers. The supply of salt from the sheet at the top of a layer and loss of salt at the bottom of a layer into the sheet below makes it unstable leading to convective motions that mix and maintain its uniformity. These salt fingers are amazing – perhaps a cm or two wide and perhaps a meter long? But what I find equally mind-boggling is that this sheet-layer organization is horizontally coherent for several hundred km! How could this be, given stirring motions by the mesoscale eddy activity – if this is the case? I mention this because these layers are not homogeneous across space: as you go from SE to NW temperature and salinity in each layer both increase with salinity dominating so density increases slightly as well. So, while they are well-defined, coherent layers stretching 100s of km, their properties are also set by the vertical flux of salt suggesting that the mesoscale stirring activity must be limited? I don’t think I was aware of these questions at the time, but we had the tool to study eddy motion in these layers.

The isopycnal float RAFOS float saw its first use in the Gulf Stream in 1984. That they worked well was evident from how little its temperature would change as it shoaled and descended 100s of meters during its lateral motions across the current. (see the blog on the compressee.) Somewhere around this time we became aware of an upcoming study called C-Salt to study the sheet and layer structure of the main pycnocline east of Barbados. Our new float technology seemed like a good match, but we were just getting started in the Gulf Stream. Years later (2003) Dave Hebert and I conducted the LIDEX (see Iso-isopycnal dispersion) in which we could show that the initial rate of separation of isopycnal floats on the same isopycnal is very gradual. But they were deployed in stratified water such that the slight difference in temperature (0.1°C) between floats might subject them to a slight vertical shear that would contribute to their initial rates of separation. But suppose you could deploy them in one of the horizontally extensive C-Salt layers?

It seems to me it might be interesting to conduct another LIDEX type study, this time in the C-Salt area. In LIDEX we deployed clusters of floats at five sites, ten on one isopycnal and nine on another deeper isopycnal. Something like this in the C-Salt area might be quite instructive. The clusters would inform us on the initial rates of separation while the separate clusters would inform us on the mesoscale activity – the scale and intensity of isopycnal stirring. Careful thought would have to go into the float design. As isopycnal devices, the floats will naturally want to lie in the stratified sheets, but we’ll want them to follow action in the layers, not the sheets. Also, since density of a layer increases from north to south, the floats will need some kind of micro-buoyancy adjustment to keep them in the same layer. I have little doubt this can be done. Imagine putting a cluster of floats in a layer where there is – presumably very little vertical shear, how long will it take for the cluster to expand to the point where the ambient mesoscale addy activity takes over? And what is that scale? I don’t know how important these questions are, but iso-isopycnal floats would be a good match to address these questions!


Schmitt, R. W., H. Perkins, J. D. Boyd and M. C. Stalcup, 1987. C-SALT: an investigation of the thermohaline staircase in the western tropical North Atlantic. Deep-Sea Research., 34, 1655-1665.