13.4: Observations of the Deep Circulation

The abyssal circulation is less well known than the upper-ocean circulation. Direct observations from moored current meters or deep-drifting floats were difficult to make until recently, and there are few long-term direct measurements of current. In addition, the measurements do not produce a stable mean value for the deep currents. For example, if the deep circulation takes roughly 1,000 years to transport water from the north Atlantic to the Antarctic Circumpolar Current and then to the North Pacific, the mean flow is about 1 mm/s. Observing this small mean flow in the presence of typical deep currents having variable velocities of up to 10 cm/s or greater, is very difficult.

Most of our knowledge of the deep circulation is inferred from measured distribution of water masses with their distinctive temperature and salinity and their concentrations of oxygen, silicate, tritium, fluorocarbons and other tracers. These measurements are much more stable than direct current measurements, and observations made decades apart can be used to trace the circulation. Tomczak (1999) carefully describes how the techniques can be made quantitative and how they can be applied in practice.

Water Masses

The concept of water masses originates in meteorology. Vilhelm Bjerknes, a Norwegian meteorologist, first described the cold air masses that form in the polar regions. He showed how they move southward, where they collide with warm air masses at places he called fronts, just as masses of troops collide at fronts in war (Friedman, 1989). In a similar way, water masses are formed in different regions of the ocean, and the water masses are often separated by fronts. Note, however, that strong winds are associated with fronts in the atmosphere because of the large difference in density and temperature on either side of the front. Fronts in the ocean sometimes have little contrast in density, and these fronts have only weak currents.

Tomczak (1999) defines a water mass as

a body of water with a common formation history, having its origin in a physical region of the ocean. Just as air masses in the atmosphere, water masses are physical entities with a measurable volume and therefore occupy a finite volume in the ocean. In their formation region they have exclusive occupation of a particular part of the ocean. Elsewhere they share the ocean with other water masses with which they mix. The total volume of a water mass is given by the sum of all its elements regardless of their location.

Plots of salinity as a function of temperature, called \(T\text{-}S\) plots, are used to delineate water masses and their geographical distribution, to describe mixing among water masses, and to infer motion of water in the deep ocean. Here’s why the plots are so useful: water properties, such as temperature and salinity, are formed only when the water is at the surface or in the mixed layer. Heating, cooling, rain, and evaporation all contribute. Once the water sinks below the mixed layer temperature and salinity can change only by mixing with adjacent water masses. Thus water from a particular region has a particular temperature associated with a particular salinity, and the relationship changes little as the water moves through the deep ocean.

Figure \(\PageIndex{1}\): Temperature and salinity measured at hydrographic stations on either side of the Gulf Stream. Data are from tables \(10.5.1\) and \(10.5.3\). Left: Temperature and salinity plotted as a function of depth. Right: The same data, but salinity is plotted as a function of temperature in a \(T\text{-}S\) plot. Notice that temperature and salinity are uniquely related below the mixed layer. A few depths are noted next to data points.

Thus temperature and salinity are not independent variables. For example, the temperature and salinity of the water at different depths below the Gulf Stream are uniquely related (figure \(\PageIndex{1}\), right), indicating they came from the same source region, even though they do not appear related if temperature and salinity are plotted independently as a function of depth (figure \(\PageIndex{1}\), left).

Temperature and salinity are conservative properties because there are no sources or sinks of heat and salt in the interior of the ocean. Other properties, such as oxygen, are non-conservative. For example, oxygen content may change slowly due to oxidation of organic material and respiration by animals.

Each point in the \(T\text{-}S\) plot is a water type. This is a mathematical ideal. Some water masses may be very homogeneous and they are almost points on the plot. Other water masses are less homogeneous, and they occupy regions on the plot.

Mixing two water types leads to a straight line on a \(T\text{-}S\) diagram (figure \(\PageIndex{2}\)). Because the lines of constant density on a \(T\text{-}S\) plot are curved, mixing increases the density of the water. This is called densification (figure \(\PageIndex{3}\)).

Figure \(\PageIndex{2}\): Upper: Mixing of two water masses produces a line on a \(T\text{-}S\) plot. Lower: Mixing among three water masses produces intersecting lines on a \(T\text{-}S\) plot, and the apex at the intersection is rounded by further mixing. After Defant (1961: 205).

Figure \(\PageIndex{3}\): Mixing of two water types of the same density (\(L\) and \(G\)) produces water that is denser (\(M\)) than either water type. After Tolmazin (1985: 137).

Water Masses and the Deep Circulation

Let’s use these ideas of water masses and mixing to study the deep circulation. We start in the South Atlantic because it has very clearly defined water masses. A \(T\text{-}S\) plot calculated from hydrographic data collected in the South Atlantic (figure \(\PageIndex{4}\)) shows three important water masses listed in order of decreasing depth (Table \(\PageIndex{1}\)): Antarctic Bottom Water (AAB), North Atlantic Deep Water (NADW), and Antarctic Intermediate Water (AIW). All are deeper than one kilometer. The mixing among three water masses shows the characteristic rounded apexes shown in the idealized case shown in figure \(\PageIndex{2}\).

The plot indicates that the same water masses can be found throughout the western basins in the South Atlantic. Now let’s use a cross-section of salinity to trace the movement of the water masses using the core method.

Table \(\PageIndex{1}\). Water Masses of the South Atlantic between 33\(^{\circ}\)S and 11\(^{\circ}\)N.

			Temperature \((^{\circ}\text{C})\)	Salinity
Antarctic water	Antarctic Intermediate Water	AIW	3.3	34.15
	Antarctic Bottom Water	ABW	0.4	34.67
North Atlantic water	North Atlantic Deep Water	NADW	4.0	35.00
	North Atlantic Bottom Water	NABW	2.5	34.90
Thermocline water	Subtropical Lower Water	U	18.0	35.94
From Defant (1961: table 82)

Core Method

The slow variation from place to place in the ocean of a tracer such as salinity can be used to determine the source of the water masses such as those in figure \(\PageIndex{4}\). This is called the core method. The method may also be used to track the slow movement of the water mass. Note, however, that a slow drift of the water and horizontal mixing both produce the same observed properties in the plot, and they cannot be separated by the core method.

Figure \(\PageIndex{4}\): \(T\text{-}S\) plot of data collected at various latitudes in the western basins of the south Atlantic. Lines drawn through data from 5\(^{\circ}\)N, showing possible mixing between water masses: NADW–North Atlantic Deep Water, AIW–Antarctic Intermediate Water, AAB Antarctic Bottom Water, U–Subtropical Lower Water.

A core is a layer of water with extreme value (in the mathematical sense) of salinity or other property as a function of depth. An extreme value is a local maximum or minimum of the quantity as a function of depth. The method assumes that the flow is along the core. Water in the core mixes with the water masses above and below the core and it gradually loses its identity. Furthermore, the flow tends to be along surfaces of constant potential density.

Let’s apply the method to the data from the South Atlantic to find the source of the water masses. As you might expect, this will explain their names.

Figure \(\PageIndex{5}\): Contour plot of salinity as a function of depth in the western basins of the Atlantic from the Arctic Ocean to Antarctica. The plot clearly shows extensive cores, one at depths near 1000 m extending from 50◦S to 20◦N, the other at is at depths near 2000 m extending from 20◦N to 50◦S. The upper is the Antarctic Intermediate Water, the lower is the North Atlantic Deep Water. The arrows mark the assumed direction of the flow in the cores. The Antarctic Bottom Water fills the deepest levels from 50◦S to 30◦N. pf is the polar front, saf is the subantarctic front. See also figures 10.15 and 6.10. After Lynn and Reid (1968).

We start with a north-south cross section of salinity in the western basins of the Atlantic (figure \(\PageIndex{5}\)). If we locate the maxima and minima of salinity as a function of depth at different latitudes, we can see two clearly defined cores. The upper low-salinity core starts near 55\(^{\circ}\)S and it extends northward at depths near 1000 m. This water originates at the Antarctic Polar Front zone. This is the Antarctic Intermediate Water. Below this water mass is a core of salty water originating in the far north Atlantic. This is the North Atlantic Deep Water. Below this is the most dense water, the Antarctic Bottom Water. It originates in winter when cold, dense, saline water forms in the Weddell Sea and other shallow seas around Antarctica. The water sinks along the continental slope and mixes with Circumpolar Deep Water. It then fills the deep basins of the south Pacific, Atlantic, and Indian ocean.

The Circumpolar Deep Water is mostly North Atlantic Deep Water that has been carried around Antarctica. As it is carried along, it mixes with deep waters of the Indian and Pacific Ocean to form the circumpolar water.

The flow is probably not along the arrows shown in figure \(\PageIndex{5}\). The distribution of properties in the abyss can be explained by a combination of slow flow in the direction of the arrows plus horizontal mixing along surfaces of constant potential density with some weak vertical mixing. The vertical mixing probably occurs at the places where the density surface reaches the sea bottom at a lateral boundary, such as seamounts, mid-ocean ridges, and along the western boundary. Flow in a plane perpendicular to that of the figure may be at least as strong as the flow in the plane of the figure shown by the arrows.

The core method can be applied only to a tracer that does not influence density. Hence temperature is usually a poor choice. If the tracer controls density, then flow will be around the core according to ideas of geostrophy, not along core as assumed by the core method. The core method works especially well in the South Atlantic with its clearly defined water masses. In other ocean basins, the \(T\text{-}S\) relationship is more complicated. The abyssal waters in the other basins are a complex mixture of waters coming from different areas in the ocean (figure \(\PageIndex{6}\)). For example, warm, salty water from the Mediterranean Sea enters the North Atlantic and spreads out at intermediate depths, displacing intermediate water from Antarctica in the North Atlantic, and adding additional complexity to the flow as seen in the lower right part of the figure.

Figure \(\PageIndex{6}\): \(T\text{-}S\) plots of water in the various ocean basins. After Tolmazin (1985: 138).

Other Tracers

I have illustrated the core method using salinity as a tracer, but many other tracers are used. An ideal tracer is easy to measure even when its concentration is very small; it is conserved, which means that only mixing changes its concentration; it does not influence the density of the water; it exists in the water mass we wish to trace, but not in other adjacent water masses; and it does not influence marine organisms (we don’t want to release toxic tracers).

Various tracers meet these criteria to a greater or lesser extent, and they are used to follow the deep and intermediate water in the ocean. Here are some of the most widely used tracers.

1. Salinity is conserved, and it influences density much less than temperature.
2. Oxygen is only partly conserved. Its concentration is reduced by the respiration by marine plants and animals and by oxidation of organic carbon.
3. Silicates are used by some marine organisms. They are conserved at depths below the sunlit zone.
4. Phosphates are used by all organisms, but they can provide additional information.
5. \(^{3}\text{He}\) is conserved, but there are few sources, mostly at deep-sea volcanic areas and hot springs.
6. \(^{3}\text{H}\) (tritium) was produced by atomic bomb tests in the atmosphere in the 1950s. It enters the ocean through the mixed layer, and it is useful for tracing the formation of deep water. It decays with a half life of 12.3 years and it is slowly disappearing from the ocean. Figure \(10.8.4\) shows the slow advection or perhaps mixing of the tracer into the deep north Atlantic. Note that after 25 years little tritium is found south of 30\(^{\circ}\)N. This implies a mean velocity of less than 1 mm/s.
7. Fluorocarbons (freon used in air conditioning) have been recently injected into the atmosphere. They can be measured with very great sensitivity, and they are being used for tracing the sources of deep water.
8. Sulphur hexafluoride, \(\text{SF}_{6}\), can be injected into sea water, and the concentration can be measured with great sensitivity for many months.

Each tracer has its usefulness, and each provides additional information about the flow.

North Atlantic Meridional Overturning Circulation

The great importance of the meridional overturning circulation for European climate has led to programs to monitor the circulation. The Rapid Climate Change/Meridional Overturning Circulation and Heat Flux Array (RAPID/MOCHA) deployed an array of instruments that measured bottom pressure plus temperature and salinity throughout the water column at 15 locations along 24\(^{\circ}\)N near the western and eastern boundaries and on either side of the mid-Atlantic ridge beginning in 2004 (Church, 2007). At the same time, flow of the Gulf Stream was measured through the Strait of Florida, and wind stress, which gives the Ekman transports, was measured along 24\(^{\circ}\)N by satellite instruments. The measurements show that transport across 24\(^{\circ}\)N was zero, within the accuracy of the measurements, as expected. The one-year average of the Meridional Overturning Circulation was \(18.7 \pm 5.6 \ \text{Sv}\), with variability ranging from \(4.4\) to \(35.3 \ \text{Sv}\). Accuracy of the measurement was \(\pm 1.5 \ \text{Sv}\).