## Info

1400 L-

Figure 6.32 Profiles of (a) temperature and (b) salinity, along with (c) the corresponding T-S diagrams, to illustrate the mixing of three homogeneous water masses (water types). Stage 1 (top) represents the situations before any mixing has taken place; stage 2 (middle) shows an early stage of mixing when the core of intermediate water is very prominent; by stage 3 (bottom), the core has been eroded.

salinity

Temperature-salinity diagrams and stability

The density (p J of sea water is a function of temperature, salinity and pressure. Rv assuming constant (atmospheric) pressure, it is therefore possible to draw lines of equal density onto T-S diagrams. Variations in the density of seawater are very small - in the oceans as a whole, density at atmospheric pressure ranges between about 1025 and 1028 kg rrr1. For convenience, oceanographers generally write density in terms of o (sigma). where: a = p - 1000,

Using this system* densities of 1026.51 and 1026.94 kg nr* are written as 26.51 and 26.94. and may more easily be compared with one another. These a values were traditionally written without units, but are increasingly given units of kg m 1 (e.g. 26 94 kg m Y

Until the 1980s the temperature value used to calculate p and a for a sample of water was generally the temperature at depth, i.e. its in situ temperature. The G value calculated on the basis of in situ temperature, in situ salinity and atmospheric pressure ts known as a, fsigma-/) In Figure 6.33, the equal-density lines used are lines of equal <7r.

However, ihe situation is complicated by the fact i hut sea water is compressible, albeit only slightly.

Bgure 6.33 Trie T-S curve for Meteor Station 200. at 9' S in the Atlantic Ocean (for use m Question 610] Numbers on ihe curve give ceplbs in hundteds o( metres. The contours are lines ot equal density, expressed as 0| (see text). Observations were made at intervals from 150 rn down to 5000m The r-Scharacleristics ot Antarctic Bottom Water (AA8W), North Atlantic Deep Water (NADW) and Antarctic intermediate Water (AAiW) are also shown. Note that the AABW characteristics shown here are those for Irue AABW, and do nol include that formed wtthm the Antarctic Circumpolar Current, this is also true of Figure 6 19

Thinking back to Section 2.2.2 (concerning convection in the atmosphere), what implication does compressibility of seawater have for the temperature of seawater at depth in the ocean?

The temperature of deep water will be raised through adiabatic compression. As a result, the in situ temperature of a sample of water is higher than the temperature that would be recorded for the same sample of seawater at the surface, under atmospheric pressure. Adiabatic heating of seawater has two important consequences. The first is that because in situ temperature (T) is changed by pressure, it is not strictly a conservative property: however, potential temperature (0) - i.e. in situ temperature corrected for compression and heating - is a conservative property.

The second consequence of adiabatic heating is that a T-S curve may give a misleading impression of the degree of stability of the water column. For a water column to be stable, its density must increase downwards, so that the T-S curve representing it would cross a contours corresponding to successively increasing density values, in the direction of increasing depth (and the greater the rate of increase of density with depth, the more stable the water column). If you look at Figure 6.33. you will see that the part of the T-S curve corresponding to the very deepest water, between 4500 m and 5000 m. curves upwards, apparently indicating that at these depths density decreases with depth. This effect is spurious, and is partly the result of the in situ temperatures at depth being increased through adiabatic heating. If we were to plot 0 against S. and then compare the trend of the resulting curve with contours of sigma-theta (ce) - the a values corresponding to the potential density, i.e. appropriate to the potential temperature. 0. salinity, and atmospheric pressure - we might well find that this water column appears to be stable all the way to the sea-floor (which is what we would expect, as an unstable water column could not persist for long).

However, this is not quite the whole story, because while using 0 and ae takes account of the effect of pressure on temperatures, and hence on calculated densities, it does not compensate for the direct effect of pressure on seawater density through compression. In other words, because o, and ce correspond to density at atmospheric pressure, the shape of a temperature-salinity curve in relation to o contours does not give a completely accurate impression of the stability of very deep water.

Under certain circumstances - including those represented by Figure 6.33 -the effect of compression can be significant. At atmospheric pressure, water with the temperature and salinity characteristics of Antarctic Bottom Water could well be slightly less dense than water with the characteristics of North Atlantic Deep Water - you can see this by comparing the density ranges represented by the blue NADW and AABW rectangles in Figure 6.33. However, because Antarctic Bottom Water is both colder and fresher than North Atlantic Deep Water, it is also more compressible, so at depth in the ocean it becomes sufficiently dense to flow beneath it.*

*Because of the complications caused by compressibility, in studies of deep and bottom water the ae values quoted now often correspond to densities at a reference level other than the surface. In these circumstances. Ge values corresponding to densities at atmospheric pressure are referred to as o0. while values corresponding to densities at (say) 2 km. 3 km and 4km. are referred to as o:. o3 and a4.

Water temperatures are increasingly recorded in terms of 9, and water masses are generally defined in terms of their 0-5 characteristics, rather than their T-S characteristics. Likewise, ae is often used instead of a,. More accurate methods of temperature measurement, combined with continuous profiling techniques, have meant that it is worth correcting for the effect of compression on temperature (and even, in some circumstances, its direct effect on density) in order to reveal interesting aspects of the temperature-salinity curve and/or deep-water flow paths. Nevertheless, for many purposes, measurements of in situ temperature are adequate, and both Tand 0 are widely used.

To consolidate your understanding of this Section, study Figure 6.34 - a 0-S curve for a station to the east of the Azores - and attempt Question 6.11.