Oceanic Convection Chimneys

Australia: The Land Where Time Began

Oceanic Convection Chimneys

A chimney is a deep, rotating, vertical cylinder of water that transports cold water from the surface of the ocean to great depths, as much as 2,500 m. This contradicts all that is believed about the stability of the ocean, which is thought of as different masses of water in horizontal layers, which are separated by vertical differences in salinity and temperature from which they derive their different densities. As a whole the ocean is stable: low-density water types above water types of greater density, all the way to the bottom. In a very few vital places surface water can be made to sink to 1.5 miles below the surface. One of these places is in the Greenland Sea. It is thought that such features should be unstable, but they can in fact be stable enough to last for years. It is not known why.

The Greenland Sea convection site

The Greenland Sea is an area of open ocean that is of unique importance, which is situated close to Europe and is linked intimately with the climate of Europe. Western Europe is 5-10^oC warmer than the average temperature for this latitude as a result of ocean currents transporting heat into the Greenland Sea. If this heat transport failed the climate of Western Europe and Britain would be more like that of Labrador.

In the Greenland Sea the region of sinking covers an area of less than 1/1000 of the world ocean, yet is vital to the circulation of the ocean, as it is only by this sinking (aka ventilation) that a complete circulation, vertical and horizontal, can be achieved, allowing the gases and nutrients that are dissolved in the surface waters to be cycled back into the depths. By way of this downwelling dissolved CO₂ is also carried down to the depths, and thereby has a big impact on the ability of the ocean to absorb a significant fraction of the CO₂ that is emitted to the atmosphere every year. It has been suggested that some of the rapid climate fluctuations that have occurred in the past, evidence of which has been detected in sediment and ice cores, have been triggered by convection changes in the past. Some models predict a decline in the convection in the Greenland Sea, and also predict a consequent cooling of the climate of Western Europe.

The convection in the Greenland Sea is located at the centre of a cyclonic gyre, i.e. a rotating circulation that is anticlockwise in the Northern Hemisphere, that is bounded by a cold current to the west, the East Greenland Current (EGC) which carries polar ice and water into the system from the Arctic Basin; to the east by a warm current that flows in a northwards direction, the West Spitzbergen Current, which is an extension of the Gulf Stream; and by the Jan Mayen Current to the south, a cold current offshoot of the East Greenland Current, which diverts to the eastward from the main East Greenland Current at about 72^o-73^oN by the presence of the Jan Mayen Ridge, a subsea mountain chain. The main highway of water and heat exchange, between the Arctic Ocean and rest of the world, is represented by the Greenland Sea, as the only deep water entrance to the Arctic Ocean is Fram Strait which connects the Greenland Sea to the rest of the world. Ice that is transported down from the Arctic Ocean into the Greenland Sea melts on its way to the south, therefore, when averaged over a year, the Greenland Sea is an ice sink, and so a source of fresh water. The melting of the ice contributes around 3,000 km³ per year to the Greenland Sea.

Local ice can form in the Greenland Sea itself in winter, within the region of cold water which has been moved to the east in the Jan Mayen Current. As the water comes from the East Greenland Current it is already cold. It leaves behind its polar ice cover, which continues on southwards down the Greenland coast. The cold, though ice free water remains behind where it is exposed to further intense cooling from a cold atmosphere in winter, in particular during climatic phases when the prevailing winter winds come from the west, blowing off the ice cap of Greenland. New sea ice is caused to form on this cold open water by the intense cooling. As a consequence of the huge amount of wave energy in the Greenland Sea in winter new ice cannot for a continuous sheet. It follows instead the classic ‘frazil-pancake cycle’ in which it initially forms as what Wadhams describes as a milk-of-magnesia suspension of frazil ice crystals in the water column, then as small cakes 1-5 m in diameter on which the edges are raised by their frequent collisions. Waves cause the crystals in the frazil in suspension to be squeezed together into clumps to form the cakes. The area of the sea carrying the cold polar water in the Jan Mayen Current is filled by these pancakes and the frazil they float in. A tongue-shaped protrusion, the Odden Ice Tongue, which can cover an area of up to 250,000 km² is formed by the new ice, which can be seen in satellite images. Sealers were the first to discover and name this in the 19^th century, as the small cakes were used by harp seals to give birth to their pups in spring. Norwegian sealers who followed the outer edge of the ice tongue to collect the fur from the seal pups called the area Odden (Norwegian for headland). Early whalers also knew the area, as slow swimming right whales (Bowhead whale) were often found in the bay of open water to the west of the Odden, Nordbukta (‘Northern Bight’).

As pancake ice forms most of the salt that in the sea water as it freezes is expelled back into the ocean. Wadhams and his research group cut up pancakes on the deck of their ship. They found that the salinity of the thinner pancakes was about 10 ppt, compared to the 35 ppt for ocean water, and the thicker pancakes can have salinity levels as low as 4 ppt, so they had excluded almost 90 % of their salt. The density of the surface water is increased by the addition of the brine from the freezing pancakes, adding to the cooling effect that destabilises the surface layer and causes this surface water to sink (Wilkinson & Wadhams, 2003). In the Greenland Sea this effect is much more powerful than in the Labrador Sea because of the impact of the extra salt. A factor that is crucial to the large amount of convection that occurs in the Greenland Sea, and therefore the maintenance of the thermohaline circulation of the Atlantic is the rapid growth rate of the pancake ice, and therefore the rapid increase of the brine content of the surface water. Wadhams says the salt rejection by the pancake ice is exciting; since pancakes grow rapidly making it a rapid process, and this happens in just right spot to have a big impact on the stability of the ocean.

The extent of the Odden has been recorded nearly every year since 1855 because it was so important to the Scandinavian sealers, and this was earlier than the founding and reporting of the Danish Meteorological Institute. In the past it formed in November almost every winter and lasted until April or May, so that it can be assumed that convection occurred throughout this period. Something has happened to disturb this since the 1990s. The Odden failed to develop in 1994-5, and since 1998 to the present. This is a major change in the nature of the Greenland Sea. This change is mainly the result of the climate switching to a new phase in which the prevailing winds over the area of the Odden came from the east and were warmer (this switch between 2 systems of atmospheric circulation is known as the North Atlantic Oscillation (NAO)). Though what is more serious is that when the NAO switched back into its former phase, the Odden didn’t develop, as the air temperatures over the sea had increased enough to prevent its formation as a result of global warming.

Chimneys in the Ocean

In order to determine the effect these changes have on the sinking of the surface water to great depths it is necessary to discover how convection occurs; and another process has been found, not all parts of which are understood, Chimney formation. In 1970 the first chimneys were discovered in a warm part of the ocean, the Gulf of Lion in the northwestern Mediterranean, during the Mediterranean Ocean Circulation Experiment (MEDOC), a large oceanographic experiment (MEDOC Group, 1970). It was found that during winter, at times when the mistral, a northwesterly wind that is intensely cold, blows out to sea from the Alpes Maritimes, the surface water was chilled by the cold air to such a degree that the water sank, not in a random manner, but in the form of small coherent rotating cylinders, called chimneys. Because the winds would change direction, a chimney would last only a few days. In the 1990s it began to be suspected that this was the mechanism by which convection beneath the Odden worked. It has been found that the surface water over a distance of 20 km in diameter forms a tight cylinder that rotates like a solid body in a clockwise direction, which is opposite to the direction of the Greenland Sea gyre as a whole, which moved the water downwards and extended its influence to a depth of 2,500 m, in an ocean that has a maximum depth of 3,500 m (Wadhams et al., 2002). The combination of ice formation and cooling makes the surface water enormously denser so that it sinks to a depth at which it reaches water of similar density, at which point it stops sinking. As the cylinder sinks it cuts through any layers of water whatever its temperature, including a deep layer of warmer water which the sinking water punches through. The chimney can be traced whether it is plotted by temperature, salinity or density. In the case of a smaller chimney that is present near a large chimney, the smaller chimney has not sunk as deep as the larger one.

It has been found by using an acoustic device (an ADCP, or acoustic döppler current sampler) that the water in the cylinder has a high degree of coherence (Budéus et al., 2004). The water within the cylinder is rotating at a speed that is proportional to the distance from the centre – i.e., like a solid rotating mass. The rotation of this cylinder of water is clockwise (anticyclonic in oceanography), which is directly opposite to the currents in the Greenland Sea which have a generally anticlockwise rotation, another reason the scientists were amazed that the cylinder can form and persist.

According to Wadhams the problem with using a small research ship when searching for chimneys is that when they found one it took a long time to map it adequately. Because of the bad weather in the Greenland Sea in winter the researchers often had to stop work and sit out the worst of the weather. The largest number of chimneys that were found in a single survey was 2 (though in most winters only 1 was found, which was located at 75^oN 0^oW), and that survey was carried out in a quiet period in winter when it was believed the stations were close enough to detect any chimney that was present (Wadhams et al., 2004). Therefore, they suspect there were only 2 chimneys in the central Greenland Sea that year. It was found when previous studies were re-analysed that in the past there were many more chimneys: A series of neutrally buoyant floats had been deployed by Jean-Claude Gascard in 1997, which are weighted to float at pre-arranged depths, and it was found that at any one time 4 of them would be turning in tight circles at depths of 240-530 m, which Wadhams et al. realised later must have meant that they were trapped in chimneys. Therefore in the 1990s there were many more chimneys than in the 2000s. Wadhams suggests it is no accident that there was also more ice.

Wadhams and colleagues visited the centre of the gyre during the Convection project for 3 winters (2001-3), while others in the Alfred Wegener Institute visited the gyre in the intervening summers. It was found that a chimney is very long lived. In the first winter an open chimney was found to be in the same position in the subsequent summer, though in the summer it was capped by 50 m of fresher water that was less dense which covers the surface of the Greenland Sea in summer as a result of the melt from sea ice and glaciers. The chimney continued to exist as a submerged rotating cell beneath this cap of freshwater. In the subsequent summer and winter the process was repeated until the end of the project so that it could not be followed any further. This is the longest-lived ocean chimney that has ever been studied (Wadhams, 2004). Such longevity in such as small, tight feature is not known of elsewhere in the ocean, where features the size of eddies lose energy and momentum by friction, ‘running down’ after a few days or weeks. It is not known what maintains a chimney in such a compressed state that is rotating so rapidly. It is not known why it doesn’t run down. It is also not known why a chimney stays in exactly one place, the longest-lived chimney moved 10 km during 3 years, in spite of no feature being present on the seabed to anchor it to one location, such as often occurs in the case of ocean eddies. In many ways chimneys have remained a mystery. According to Wadhams in spite of having made these key discoveries with huge climatic implications, their repeated bids to the Natural Environment Research Council (NERC) in the UK for further support to study chimneys in the field were all unsuccessful.

What is now known is that there are now fewer of these structures, which coincides with loss of ice from the Odden, and this convection decrease in the Greenland Sea will have a serious impact on the world ocean. It has been suggested by models that to account for the amount of deep water formation that is occurring between 6 and 12 chimneys need to form and dissipate each year. Wadhams says it is not known where these chimneys are now, whether they still exist in spite of the difficulty of forming without ice, whether the formation of deep water is slowing or stopping, or is it just happening in a different way, or in a different place.

Sources & Further reading