Ocean Circulation And Climate Isbn 0126413517

Copyright © 2001 Academic Press All rights of reproduction in any form reserved

North Atlantic Ocean Depths
Fig. 5.7.1 Salinity (a) and temperature (b) contours at 1100 m depth in the North Atlantic from Levitus (1982) climatology.The contour interval for salinity is 0.2%o and for temperature is 1 °C.

The Mediterranean Sea, the Mare Nostrum of the Romans, is a semi-enclosed sea with a broad history intimately related to the development of western civilization, but only recently have oceanog-raphers recognized that the waters of the Mediterranean provide them with a model of the world ocean itself (Lacombe, 1990). Geographically it has a zonal extent of about 4000 km and a mean meridional width of 1000 km, with a mean depth of 1500 m and a maximum depth of up to 5000 m in the Ionian Sea. It is divided by the Strait of Sicily into Western and Eastern basins (Fig. 5.7.2). Mainly an evaporative basin, it acts to transform relatively fresh North Atlantic surface water (salinity of

Adriatic Sea Adcp
Fig. 5.7.2 Map of the Mediterranean Sea indicating names of places mentioned in the text.The shaded areas in the Gulf of Lions, Adriatic, Aegean and Levantine correspond to known regions of water mass formation.The 200,1000, 2000 and 4000 m depth contours are also shown.

36.1 %o) into a dense Mediterranean Water, which has a salinity of 38.4%o, a temperature of 13°C and a density of 1029.7kgm~3, as it outflows at depth through the Strait of Gibraltar. The Mediterranean Sea is often regarded as a 'laboratory basin' for oceanographic circulation studies, especially of the thermohaline circulation, which is particularly active despite the relatively small size of the sea. It is one of the few places in the world where deep convection and water mass formation take place. In the present climate, deep convection occurs only in the Atlantic Ocean; the Labrador, Greenland and Mediterranean Seas; and occasionally also in the Weddell Sea. Convection in these regions feeds the thermohaline circulation, the global meridional-overturning circulation of the ocean responsible for roughly half of the poleward heat transport demanded of the atmosphere-ocean system (Marshall and Schott, 1999; see also Bryden and Imawaki, Chapter 6.1). It is conjectured that the resulting Mediterranean outflow plays an important though indirect role in the North Atlantic circulation (Reid, 1979) and, consequently, in the thermohaline conveyor belt at global scales and on time scales of global climate change (Wu and Haines, 1996). According to Reid (1979) the Mediterranean outflow helps maintain the high salinity of the Norwegian Sea. Without this source of high-salinity water the Norwegian-Greenland

Sea might not provide the denser waters that fill the Arctic Basin and thus contribute a major component of the North Atlantic Deep Water. It is the relatively saline North Atlantic Deep Water, transported by the deep western boundary current, that penetrates into the low-salinity waters of the Weddell Sea, where it is cooled further and enriched with brine to provide the Antarctic Bottom Water - the densest water found in the oceans of the world. To the extent that these suggested linkages are correct, the exchange between the Atlantic Ocean and the Mediterranean Sea is of significant importance.

5.7.2 Formation of Mediterranean Water

There are several places within the Mediterranean where preconditioning (i.e. a cyclonic circulation with convex curvature of isopycnals that bring dense, and usually weakly stratified, waters close to the surface) and air-sea fluxes combine to induce convective processes (for details on the convection mechanisms, see Marshall and Schott, 1999). These are the Gulf of Lions in the Western Mediterranean, the region south of Rhodes, the Levantine Basin, the southern part of the Adriatic Sea, and in recent years the south region of the Aegean Sea (Roether et al., 1996) in the Eastern Mediterranean (Fig. 5.7.2). Of these sites, the Gulf of Lions region is where convective processes reaching depths of more than 2000 m have been extensively documented since the Mediterranean Ocean Convective (MEDOC) experiment (MEDOC Group, 1970). The water mass distribution in the Western Mediterranean comprises three layers. In the upper layer, and originating mostly from the inflow through the Strait of Gibraltar, is the Modified Atlantic Water (MAW). Between 150 to 500 m depth, a warm and salty layer is found, referred to as Levantine Intermediate Water (LIW), which is formed by shallow convection in the Eastern Mediterranean Basin and then slowly spreads into the Western Mediterranean through the Strait of Sicily. Beneath the LIW layer, the basin is filled with near-homogeneous Western Mediterranean Deep Water (WMDW). It is this WMDW that is formed by deep convection processes in the Gulf of Lions and which contributes importantly to the characteristic of the Mediterranean outflow (Stommel et al., 1973; Kinder and Parrilla, 1987). Based on a newly composed hydrographic climatology, Krahmann (1997) estimates a WMDW production rate in the northwestern Mediterranean of 1.8 + 0.6 x 1013m3 yr"1, corresponding to 0.6 +0.2 Sv. The average yearly WMDW production is made up of 1.3 x 1013m3 of LIW and 0.5 x 1013m3 of MAW. Thus, the formation rate of LIW and slight variations of its characteristics have implications on the WMDW formation rate and its variability. Actually, it is becoming clear from recent observations that deep water formation is not a process that recurs every year with certainty and regularity. The intensity of convection shows great variability from one year to the next and from one decade to another. For example, in the Gulf of Lions, 1969, the year of the first MEDOC experiment (MEDOC Group, 1970), was a year of strong convection, but convection in 1971 was not as strong (Gascard, 1978). Vigorous deep convection to 2200 m returned in 1987 causing a very homogeneous water body of potential temperature 12.79°C and salinity 38.45%o (Leaman and Schott, 1991). Convection reached only to 1700 m in 1991 and did not mix the water column as thoroughly (Schott et al., 1996). Fluctuations in the composition and possibly also the volume of the Mediterranean outflow are the result of the variability in WMDW formation, coupled to the LIW formation variability (Nittis and Lascaratos, 1998). Recent findings that in the last decade an influx of Aegean Sea water has replaced 20% of the deep waters of the Eastern Mediterranean, which were previously only formed in the Adriatic Sea (Roether et al., 1996), also affect such fluctuations.

Apart from these observed interannual variations in deep and intermediate water formation, there is a well-documented increase in the salinity of the deep waters of the Western Mediterranean over the past 40 years (Lacombe et al., 1985; Leaman and Schott, 1991). Observations also suggest that in the past 10 years there has been a jump in the salinity of the newly formed deep waters in the Eastern basin (Roether et al., 1996). It has been argued that this salinity increase has resulted from the diversion of the Nile and Russian rivers for irrigation so that the effective net evaporation over the Mediterranean basin has increased. Application of hydraulic control models then project that the overall Mediterranean salinity will increase by about 0.13%o over the next 100 years or so (Rohling and Bryden, 1992). However, it remains unclear what the implications of such an increase in salinity would be on the overall circulation of the Mediterranean and North Atlantic.

5.7.3 Outflow of Mediterranean Water at the Strait of Gibraltar

The Strait of Gibraltar is the Mediterranean's only communication with the World Ocean; it is about 60 km long, 15 km wide at its narrowest section (Tarifa narrows) and only 280 m deep at its main sill (Fig. 5.7.3). There are three main components to the flow (Lacombe and Richez, 1982; Candela, 1991): a tidal, mainly barotropic flow, with magnitudes of up to 2.5ms-1 (Candela et al., 1990); a barotropic subinertial component driven by atmospheric pressure fluctuations within the Mediterranean and with magnitudes close to 0.4 m s^1 (Candela et al., 1989); and a baroclinic subinertial component driven by the internal pressure gradient due to the density difference between the Mediterranean and the Atlantic Waters, with magnitudes of about 0.5 m s^1, and likely to be hydraulically controlled (Armi and Farmer, 1988). Therefore, the Strait is dynamically very energetic with tidal, subinertial and long-term currents all being of significant amplitude. This situation makes studying the exchange flows particularly difficult, requiring long and careful measurements

Major Basins The Mediterranean Sea


Fig. 5.7.3 Map of the Strait of Gibraltar showing the location of the sill mooring indicated in the text (large dot). The 50,100, 200,300,400 and 500 m depth contours are also shown. Depths larger than 400 m are shaded.The small dots distributed along Gibraltar's main sill indicate the positions where current profiles were obtained from a ship-mounted ADCP during a tidal cycle.


Fig. 5.7.3 Map of the Strait of Gibraltar showing the location of the sill mooring indicated in the text (large dot). The 50,100, 200,300,400 and 500 m depth contours are also shown. Depths larger than 400 m are shaded.The small dots distributed along Gibraltar's main sill indicate the positions where current profiles were obtained from a ship-mounted ADCP during a tidal cycle.

not only of currents but also of water characteristics and in particular of the fluctuations of the interface that separates the inflowing Atlantic from the outflowing Mediterranean Waters (Bryden et al, 1994).

The Strait of Gibraltar has been the subject of several field measuring programmes in recent years (Bryden and Kinder, 1991). The longest continuous record, 2 years (October 1994-October 1996) of continuous measurements of the current profile and water properties at a mid-sill location on Gibraltar's main sill (Fig. 5.7.3), comes from a 2.5-year measurement programme that concluded in October 1996. These observations were obtained with an upward-looking, bottom-mounted, broadband 150 kHz Acoustic Doppler Current Profiler (ADCP), capable of measuring the entire 280m water column at this mid-sill location, with a vertical resolution of 10 m. In addition, during two cruises on board the RV Poseidon in April 1996 and in October 1997, consecutive crossings of the strait were performed over the sill section measuring the current profile using a ship-mounted ADCP through a complete semidiurnal tidal cycle (Figure 5.7.3 shows the location of the current profiles measured by the ship over the sill). The April 1996 cruise coincided with a period of neap tides (small amplitude), while those of October 1997 were performed during spring tides (large amplitude), providing an idea of the across-strait current structure during both tidal extremes. From these sections it is clear that the currents at the sill present large cross-strait variability; however, the mid-sill moored ADCP measurements capture the main time variability of the currents and both sets of observations are used here to estimate a time series of the exchange through the Strait. In calculating the exchange through the Strait, it is important to obtain estimates of the quality, as well as the quantity, of the water being exchanged. In order to distinguish Atlantic from Mediterranean waters it is essential to have simultaneous measurements of the density structure of the water column along with those of the currents. In addition, it is mandatory to take into account the contribution to the mean exchange from the high correlation between the barotropic (tidal and subinertial) currents and the depth of the interface separating Atlantic and Mediterranean Water types at the sill (Bryden et al., 1994). For this reason, simultaneously with the bottom-mounted ADCP measurements, an additional mooring was installed that contained several (3 to 5) instruments in the water column, depending on the deployment period, which made it possible to construct time series of the depth of the interface between the Atlantic and Mediterranean Water cores. Based on previous work (Bryden et al., 1994), as well as these observations, it was decided to use the 37 salinity as the characteristic value delimiting the boundary between the two layers. Estimates of Atlantic and Mediterranean Water exchanges were calculated using hourly time series of current velocities at 10 m depth intervals from the surface to the bottom, hourly time series of the depth of the interface, and a realistic cross-section bottom relief together with the cross-strait current structure based on the aforementioned ship surveys done during spring and neap tidal cycles. These, after being low-pass filtered to retain periods longer than 3 months, are shown in Figure 5.7.4.

An important result from these calculations is that both the Atlantic and Mediterranean Water transports show a small, but appreciable, seasonal cycle. Of the two, the outward Mediterranean lower-layer flow is a more reliable estimate, showing an annual transport range of 0.28 Sv with minimum outflow around early summer (July 1995) and a maximum in late winter (February 1995 and 1996). The seasonal cycle on the inflow is not as well represented in these observations, although maximum inflow tends to occur during the summer of 1995 with a second maximum at the beginning of 1996. Minimum inflows occur around late winter (February 1995 and 1996) coinciding with the maximal outflows. Bormans et al. (1986) suggested a seasonal cycle in the inflow of about 6%, with maximum transport in the spring. They attributed the increase to changes in interface depth and argued that winter water mass formation processes raise the interface level within the Mediterranean, while draining of the Levantine Intermediate Water reservoir occurs during the rest of the year, effectively lowering the interface. The

0 0

Post a comment