Dynamic topography
4.3.3 INSIGHTS FROM MODE
34 ~ N
115
Figure 4.24 (a) A compilation of the tracks of all the Sofar floats launched in the MODE area (in the vicinity of the red dot), as recorded from August 1972 to June 1976. Note how the floats became dispersed from the original area, so that after a number of years they had become scattered over the western part of the North Atlantic subtropical gyre.
The most exciting result to come out of analysis of the MODE records was the ubiquity of mesoscale eddies, which were found to dominate flow with periods greater than the tidal and inertial periods (Section 3.5.2). This was despite that fact that - as has since become c l e a r - the MODE site is relatively 'quiet" eddy-wise, when compared with areas closer to intense frontal currents (Figure 3.34).
Figure 4.24(a) is a compilation of the tracks recorded during MODE. Most of the floats seem to have been carried in eddy currents at some time or another, and at about 30 ~ N, a number of floats were trapped in an eddy (or eddies) for a considerable time. The track of one of these floats is shown in Figure 4.24(b) (overleaf), which shows that before being caught in a fast southward flowing current, the float spent most of the monitoring period circulating within a cyclonic eddy which was moving due west.
Figure 4.24 (b) The path of one of the Sofar floats tracked during the MODE experiment. The float was drifting in the sound channel at 2000 m. Dots show daily positions, so the track shows the path taken by the float over 5! months. The float first circulates at speeds of about 0.5 m s -1 within a westward-moving eddy. It is then caught in a fast southward-flowing current.
Figure 4.25 illustrates geostrophic flow patterns at a depth of 150 m in the MODE area, on the basis of density determined from T and S data.
"o~ ~~ ~- ~Z -= :m- ~ 0 r ,~. ,-~. O~ o = o ~~ o ~_~ ~= ~ ~"o = ,,, ~.~= ~ ~o~ ~ " ~- ~ "-4" o 0 ~ ~,, ~,, ~'~ ~,." I~'~ "-~ ~ r ~ =.o o= ~,= o ~ _~ = 0 c~ ~ -'-- (.t) ~ ~ = ~ o ~ ~ ~ o ~ 5 ~ 5 =_ ~ ~ ~ ~ ~ ~ = = ~ ' o ~ _~ ~ o -. o ~ 9. ~ ~ =-.-5< z =. .~ .. =~ 9 --. ,_. CIQ 9 ~ ~ --* ... ~ 9
~_~
~ ~ o ~~ ...
. ~ .} ~. ~ 9-= ~ ~,~= ~ _. .~-~. ~ ~ ~ .~ .~ "* ,--- '-'l 0 ~.. ~,~ "--4Figure 4.26 (a) Examples of freely drifting surface buoys. Left. simple spar buoy, with radar reflector and parachute drogue. Middle and right.
more sophisticated buoys with satellite-tracking devices, for use in shallow seas (with cruciform drogue) and the deep ocean (with tubular drogue at - 700 m depth).
Apart from these obvious disadvantages, it is difficult to interpret results obtained using simple drifters. Drifters moving near the surface may be directly affected by the wind; others, designed to be just negatively buoyant at the sea-bed so that they trail along the bottom, are affected by friction with the sea-bed (as is the layer of water in which they move) and so their movement cannot be taken to represent that of the main body of water.
Also, it is hard to assess the effects on the floats' trajectories of tidal flows and other fluctuating currents.
In tracking or deducing the paths of drifting objects, we are effectively following the path taken by a parcel of water as it moves relative to the Earth;
the velocity of the parcel of water may be calculated from its change of position with time. Such methods of current measurement are described as Lagrangian*. Much of our knowledge about oceanic circulation patterns and current velocities has been obtained by Lagrangian methods, including use of ship's drift (e.g. Figure 5.12). An enormous amount of information has been obtained from the tracks of Sofar floats and, later, from tracks of more sophisticated neutrally buoyant floats, including ALACE (Autonomous
119
Figure 4.26 (b) Typical operational cycle of a PALACE float. Some more sophisticated versions actively 'seek' the required depth, rather than simply drifting at the depth for which they have been ballasted.
Lagrangian Circulation Explorer) floats. These were developed during the 1990s to fulfil the needs of the World Ocean Circulation Experiment (WOCE, Section 6.6.1 ), which required about a thousand subsurface floats to be deployed worldwide, for which acoustic tracking was not feasible.
Like Sofar floats, ALACE floats are designed to become neutrally buoyant at a specified depth, remain submerged for a fixed period (typically a few weeks) and then return to the surface where they transmit their position (accurate to 1-3 kin) via the
Argos
satellite. They remain at the surface, transmitting, for 24 hours and then return to their operating depth. Those deployed in the 1990s could undertake 70 or more such cycles over a five- year period. (For examples of tracks from ALACE floats, see Figures 5.34 and 6.45.) There are now versions of ALACE and other autonomous floats that carry sensors for temperature, conductivity (for salinity) and pressure (for depth). These collect data on water properties as they rise or sink and are known as profiling floats. Figure 4.26(b) shows the operational cycle of an upward-profiling PALACE (Profiling Autonomous Lagrangian Circulation Explorer) float.Alternatively, currents may be measured by Eulerian* methods, in which the measuring instrument is held in a fixed position, and current flow past that point is measured. The greatest challenge with Eulerian methods of current measurement is keeping the measuring instrument fixed. Nowadays, this is commonly done by anchoring a subsurface float to the sea-bed and securing
*Both these terms are named alter mathematicians: Joseph Louis Lagrange ( 1736-1813) and Leonard Euler (1707-83). respectively.
Figure 4.27 Some examples of current-meter moorings (vertical scale greatly compressed).
(a) Surface buoys carrying meteorological instruments (anemometer etc.) may either support current-meter arrays directly, or act as marker buoys for a bottom mooring.
(b) A deep-ocean bottom mooring incorporating an ADCP (see Section 4.3.7).
The 'logger' internally records data on current velocity, salinity (as conductivity, C) and temperature, T.
one or more current meters to the taut mooring cable (Figure 4.27(a), left).
Current-meter arrays can also be suspended from a securely anchored surface buoy (Figure 4.27(a), right), but such moorings are affected by surface wave motions which are transmitted to the meters. They are also particularly vulnerable to damage by shipping or fishing trawls. Even bottom-mounted current-meter arrays (Figure 4.27(b)) may be damaged by trawling activity, which now extends to depths of well over 1000 m. Occasionally, current meters are deployed directly from anchored ships.
In the most common type of current meter, current flow causes a propeller to rotate at a rate that is proportional to the current speed. Some current meters are designed to align themselves in the direction of flow so that the propeller rotates about an axis parallel to the current direction (Figure 4.28(a)). By contrast, in the Savonius rotor, two hollow S-shaped rotors are mounted about
a v e r t i c a l axis (Figure 4.28(b)). This type of meter is very sensitive and responds even to very weak currents, but it only rotates in one direction, so that rapidly reversing flows caused, for example, by surface waves, all contribute to the total number of rotations recorded. In the other type of meter, the propeller rotates in both directions so that the effects of rapidly reversing flows cancel out.
121
Figure 4.28 (a) A typical propeller-type current meter. (b) A Savonius rotor.
The meters are oriented with respect to the current by their large vanes.
Figure 4.29 (a) The paths taken by a number of Sofar floats at a depth of 700 m in the western North Atlantic. The arrows are 100 days apart; note the high velocity of the floats caught in the Gulf Stream.
(b) The same region as in (a), showing the mean current flow as it would have been measured by fixed current meters.
In most modem current meters, current speed and direction (measured by means of a magnetic compass) are internally recorded electronically. The record is retrieved at the end of the experiment, when the moorings are released from the anchor (generally by means of an acoustic signal) and the current meter bobs up to the surface. Sometimes, current data are transmitted to shore stations via satellite, so that they can be analysed immediately and used for forecasting.
When moored current meters were first used, their recovery rate was often as low as 50%. Technological advances have permitted the development of robust mooring systems and extremely strong cables that can withstand the effects of corrosion, extreme tension, and even fish, which have been known to bite through nylon mooring lines. Modern moorings can be deployed in almost any part of the ocean for a considerable period (over a year), although eventually they may become affected by 'biofouling' (colonization by marine organisms),
\vhich allcr~ their dra~ characleristics.
Figure 4.29 shows the paths taken by a number of Sofar floats in the region of the Gulf Stream and, for comparison, the mean current flow as it would have been measured by means of moored current meters (although, in fact, such measurements were not made).
For completeness, we should mention the use of coloured d y e - in a sense, the ultimate Lagrangian method. This approach is generally used to study relatively small-scale turbulence, and to investigate rates of dispersion, i.e. the extent to which a patch of initially coherent water becomes widely spread, or dispersed, as a result of mixing and turbulence (cf. Figure 4.24(a)).
All the methods mentioned above provide information about horizontal velocity but nothing at all about the vertical component of current velocity. Vertical flow velocities are generally very much smaller than horizontal velocities but in certain unusual circumstances have been measured using floats with fins so arranged that the float rotates at an angular velocity proportional to the vertical current velocity.
An instrument which measures both vertical and horizontal velocities is the Acoustic Doppler Current Profiler (ADCP). ADCPs are now widely used, and can be deployed in various ways (e.g. on moorings, cf. Figure 4.27(b)).
The way in which they exploit sound waves is explained in Section 4.3.7.