![]()
1. The general circulation
of water masses in the southwestern Mediterranean.
2. The origin, structure
and trajectories of Algerian Eddies.
ELISA rationaleThe general circulation of the water masses in the Mediterranean Sea has been studied from the beginning of the 20th century. Maps produced in the 1960s did not differ much, and the general circulation schemes were widely accepted.
This prompted a re-examination and interpretation of the few available data sets . The main point was to propose a coherent explanation for the episodic departure of water masses from their "normal" path. New hypotheses were first set forth about the Algerian Current and the role of the eddies it generates on the circulation of the water masses in the Western Mediterranean [Millot, 1985]. The first mesoscale-dedicated operation designed to study the algerian eddies in the western part of the Algerian Basin allowed Millot [1987] to propose revisited schemes of the circulation in the western Mediterranean. The major change was the introduction of the mesoscale activity: eddies in the Algerian Basin were supposed to be able to locally modify the water masses "normal" paths, providing coherent explanation for "unexpected" in situ observations.
About 10 years later, additional observations enabled
to complete and refine the circulation schemes [Millot, 1999]. But they
also raised new questions and prompted other hypotheses. It was time for
a new large amplitude investigation in the Algerian Basin: it was
time for ELISA.
1. The general circulation of water masses in the southwestern Mediterranean.The Mediterranean Sea is called an evaporation basin, because loss of water by evaporation exceeds inputs from precipitations and river runoff. Would the Strait of Gibraltar be closed, the Mediterranean sea level would decrease by ~1m/year. To compensate for this water loss, there is an inflow of Atlantic Water through the Strait of Gibraltar in the upper layer.
The general circulation of the water masses in the Mediterranean is mainly determined by exchanges of heat and salt (thermohaline circulation) and by earth rotation effect (Coriolis force). As a result, circulation roughly runs counter-clockwise along the continental slopes. However this circulation, as depicted by schemes such as Nielsen's [1912] or Ovchinnikov's [1966], is more likely to be established if measured over a period of time long enough (several months). Indeed, dynamical activity at mesoscale is superimposed nearly everywhere in the Mediterranean, modifying locally and episodically this mean path.
- time scales: from weeks to > 1 year
- horizontal space scales: from several 10s to a few 100s of km
- vertical space scales: from few 10s to a few 1000s of m (down to the bottom: ~2800m).
Mesoscale phenomena are typically upwellings,
fronts, meanders and eddies. Locally and episodically (up to several months,
Taupier-Letage and Millot [1986], Millot [ 1994]), they can strongly perturb
the general circulation.
Click here to see the current time series recorded ~ 25km off
the Algerian coast at 100m deep (ELISA mooring 3):
- changes in direction and intensity of the current vector are due
to meanders and eddies,
- due to an algerian eddy, which stayed over the mooring for ~5 months
inducing strong currents, the current averaged over 1 year is actually
northwestward (~10cm/s), when general (mean) circulation is eastward.
A main consequence is that, on new circulation schemes, there are no longer any continuous paths figured in the Algerian Basin.
The main water masses of the western Mediterranean are:
More information can be found e.g. in EUROMODEL
[1995], Send
et al. [1997] and Millot [1999].
Throughout its path in the western Mediterranean,
the water of Atlantic origin will be mixed with mediterranean resident
water, and therefore will be called Modified Atlantic Water (MAW).
It first flows through the Alboran Sea and exits
usually forming the Almeria-Oran jet, reaching the Algerian coast near
~0°. If only constrained by the Coriolis force, MAW should flow eastward,
keeping the coast on its right, in a vein 25 to 50km wide. Thus MAW should
flow along the Algerian and Tunisian slopes, and wholly enter the eastern
Mediterranean through the Channel of Sardinia. But MAW is also found in
the whole northwestern Mediterranean, forming the permanent Northern Current
along the Italian, French and Spanish coasts. As well, in situ measurements
show that water of Atlantic origin only little modifed (ie newly
entered) can also be found offshore in the Algerian basin, especially in
the eastern part of the Algerian Basin. Hence previous widely-accepted
circulation schemes displayed a progressively widening MAW vein, "splitting"
or "branching" as it approached the Channel of Sardinia: one "branch" heading
northward into the NW Mediterranean, the other continuing eastward
through the Channel of Sardinia into the eastern Mediterranean. No mechanism
was proposed to explain such a "branching".
Satellite observations becoming more accessible in
the 80s brought information about mesoscale dynamics. Satellite
images analysis showed that the water of atlantic origin flowing along
the Algerian coast does so mainly by way of
meanders and eddies, and as soon as 0°-1°E. This led Millot
[1985] to define an Algerian Current, characterized by its instability
which generates this intense mesoscale activity, and to link the occurrence
of recent MAW in the open basin to Algerian Current- induced eddies, hereafter
called Algerian Eddies (AEs). The MEDIPROD-5
(1986-1987) operation [Millot and Bonin, 1990] was the first mesoscale-dedicated
experiment in this region. Consequences of AEs on hydrology and circulation
were confirmed [Benzora and Millot, 1995 a,b; Millot et al., 1997]. The
Algerian Basin thus acts as a reservoir for MAW, disconnecting the flow
of Atlantic Water coming in at Gibraltar from those exiting eastward through
the Channel of Sardinia and northward between Balearic Islands and Sardinia.
Click on image to see full resolution
MAW
Circulation Scheme
from Millot [1999]
Click on image to see full resolution
LIW
Circulation Scheme
from Millot [1999]
Click on image to see full resolution
WMDW
Circulation Scheme
from Millot [1999]
2. The origin, structure and trajectories of Algerian Eddies.
The Algerian Current (~200m thick) is unstable,
and generates meanders
and eddies
At that stage, eddies (AE) have an associated upwelling cell (U) on their southwestern edge. Meanders and eddies (with attached upwelling cell) propagate eastward along the coast at few km/day.
Only anticyclonic eddies develop. Algerian Eddies (AEs) diameters range from ~50 to 100km, up to ~250km for some. Their vertical extent can be as deep as the sea floor (~2800m). Lifetimes range from months up to ~3 years, as observed for 2 of them at least.
AEs can detach from the coast. This is most likely to happen in the vicinity of the Channel of Sardinia, where the narrowing and shallowing bathymetry will constrain AEs to a northward path along the Sardinian slope. AEs will eventually drift westward in the open basin for months. They often come back in the Algerian coastal zone, where they interact with their parent current, potentially leading to dramatic pertubation of the general circulation [Taupier-Letage and Millot, 1986; Benzohra and Millot, 1995]. Algerian Eddies generally describe a counter-clockwise loop in the Algerian Basin [Fuda et al., 2000].
Before the MEDIPROD-5 experiment in 1986-1987 ( 24, 9-month current and temperature time series) AEs were believed to progressively grow in diameter and depth as they propagated eastward. It appeared that AEs actually get their full vertical extent very rapidly, ie shortly after their creation. However their vertical extent has been observed to vary rapidly (from ~2800m to ~200m), their shape and diameter as well.
The LIW vein might become unstable as it passes the open SW corner of the Sardinian slope. Millot [1999] hypothezised that such instability could produce baroclinic LIW eddies which he named "Leddies", by analogy with Meddies. Leddies would then propagate into the interior of the Algerian Basin, and contribute to the occurrence of recent LIW in the basin.
More info can be found in Millot et al. [1997], and Obaton et al. [2000].
Finally, AEs induce secondary phenomena (limited space and time scales):
- small (diameter a few 10s of km) shear
cyclonic eddies (C)
- and
filaments (up to 100-200km in length, but very limited in depth)
mainly between AEs and coastline breaks.
3. The biological response to mesoscale phenomena.
The dynamical signature of mesoscale phenomena
can be determined by the Sea
Surface Temperature (SST) distribution using satellite thermal infrared
radiometer such as NOAA/AVHRR. The corresponding
biological signature can be determined using ocean color, as it is linked
to algal biomass (roughly the more phytoplankton the greener), using satellites
like past CZCS
(Coastal Zone Colour Scanner) and present SeaWiFS
(Sea-viewing Wide Field-of-view Sensor) .
The close correlation between both dynamical
and biological signatures illustrates
the fact that, at mesoscale, dynamics is driving biology.
The instability of the Algerian Current leads to a relatively richer coastal zone, with richer upwelling cells and entrainments around coastal eddies (however enrichment by upwelling cells is not to be compared with large wind-induced upwelling systems). However, large-sized and long-lived algerian eddies being anticyclonic*, they represent unfavorable situations for biological development. Hence they are signed by very oligotrophic (poor) areas, which can occupy the greatest part of the basin.
Sometimes small-scale shear cyclonic eddies are generated,
that represent the richest spots**. However, as they do not last for long
(days to weeks) and are very limited in size (10s of km), their overall
contribution is not expected to be large.
* anticyclonic rotation induces a depression of the isolines within the eddy. After the thermal stratification is established, the euphotic zone (upper layer where light allows algae life) will become nutrient-depleted.
** cyclonic rotation induces a doming of the isolines
within the eddy, and enables input of nutrients from the lower layer into
the euphotic zone.
Since there is no continental shelf off Algeria (fish is especially abundant on shelves), and no big river (likely to bring nutrients and organic matter into sea) the Algerian Basin is in a situation not very favorable for natural ressources. It is all the more important to study the Algerian Current.
As variability is very high in both dynamical and
biological fields, correct description, and hence, understanding, require
simultaneous sampling at short space and time intervals, over a long period
of time. Due to instrumental limitations (especially in biology) and to
the large effort needed for such experiments, only a few mesoscale-coherent
data sets were available.
The MAST3/MTP2 european programme provided the framework for the large
interdisciplinary effort of the ELISA experiment designed to study the
Algerian Eddies.