Variability in Labrador Sea Water formation

The Labrador Sea is located in the northwestern corner of the North Atlantic Ocean. Although small in size, this sea is very important for the circulation in the ocean, because it is one of the few locations where deep convection takes place. Deep convection is vertical mixing of water over a large depth in the ocean, occasionally down to more than two kilometers depth in the Labrador Sea area.
This process is the only way in which the water in the deep ocean is exposed to the atmosphere, and the only rapid interaction between water in the deep ocean and in the surface layer.

Deep convection is a typical winter process. The large-scale ocean circulation brings warm water from the Southern Hemisphere to the North Atlantic Ocean in the upper few hundred meters. In the subtropical region, large amounts of evaporation make the water saltier. This saline and warm water is then carried to the subpolar region, where part of it flows into the Labrador Sea. Here, strong winds from the North American continent bring in cold and dry air. The still relatively warm ocean water is strongly cooled down by the cold air, which makes the surface water denser. When the surface water is denser than the water underneath, it will mix (vertically) with the subsurface water. The much colder water returns to the Southern Hemisphere in the abyssal ocean.

While the atmosphere cools down the ocean, the ocean warms up the atmosphere. The warmed air is brought to the European continent by the prevailing
westerly winds, which is why the winters in western Europe are relatively mild compared to other places at a similar latitude. The large-scale ocean circulation that transports heat to the North Atlantic Ocean shows a large variability, and studies have shown that the variability in this circulation is linked to variability in the formation of Labrador Sea Water on time scales of years to decades. Therefore, it is important to understand the variability in Labrador Sea Water formation.

Two of the processes that play an important role in this variability are studied in this dissertation. The first study focuses on the spring and summer period after a deep convection event. In these months the typical layered structure of the ocean, which is destroyed by the deep convective mixing, is quickly restored. The dense water in the area affected by deep convection is then (partly) replaced by lighter (warmer or fresher) water, originating from currents that encircle the interior
Labrador Sea (called boundary currents). This process, known as ’restratification’, is governed by eddies, which are coherent structures in the ocean in a swirling motion.

In the Labrador Sea, three different types of eddies are known to play role in this ’restratification’ process. First, Convective Eddies are relatively small eddies (a radius of about 10 km) which form around the area where deep convection takes place. During and directly after deep convection, they mix the water in the convected
area with the surrounding water. Second, Boundary Current Eddies, also relatively small, mix water from the boundary currents into the interior basin. Third, Irminger Rings are large (a radius of about 25 km) eddies that are formed off the west coast of Greenland due to a steepening of the slope at this location. These eddies bring the light water from the boundary current quickly to the deep convection
area. Furthermore, they very efficiently stir the interior Labrador Sea which may enhance the efficiency of the smaller eddies in the restratification process.

Using a numerical model in a highly idealized configuration for the Labrador Sea, the contributions of these three eddy types to the restratification process after deep convection were quantified. The Convective Eddies and Boundary Current Eddies together replenished 30% of the heat that was lost during deep convection. Irminger Rings added another 45% to this number. The presence of Irminger Rings is thus essential for a realistic amount of restratification in this area.

The second process study focuses on the effects of a very fresh surface layer on deep convection. Fresh water is light, and therefore the wintertime cooling needs to be very strong to make the surface water denser than the subsurface water. A fresh
surface layer can thus inhibit deep convection, a phenomenon that is well known. From 1969 to 1971, convection was restricted to the upper 300 m. This shutdown has been attributed to a substantial surface freshening. The abrupt resumption of
convection in 1972, in contrast, is attributed to the exceptionally cold winter. Using the data collected at the Ocean Weather Station Bravo in the central Labrador Sea, the causes of the shutdown and resumption of deep convection are studied in detail.

The shutdown started as a result of the combined effects of the fresh surface water and the very mild winter conditions. Once convection was shut down, the wintertime temperature of the surface water decreased, which caused a reduction in the heat loss to the atmosphere. Thus, while a large heat loss was required for deep convection, a side effect of having no deep convection is a smaller heat loss. A second effect of the lower surface water temperature was that the thermal expansion
coefficient, which gives the change in density as a function of temperature, was reduced. As a consequence, a certain heat loss gives a smaller density change of the surface water. Both effects reinforce the shutdown state. In 1972, deep convection returned both because of the exceptionally harsh winter as well as advection of
saltier waters into the convection region.

Apart from understanding the variability in Labrador Sea Water formation, it is also important to monitor this variability. In situ monitoring is however severely hampered by the harsh winter conditions. Furthermore, the monitoring programs
do not cover the entire Labrador Sea and are often summer observations. The network of satellite altimeters does not suffer from these limitations and could therefore give valuable additional information. Altimeters can in theory detect deep water formation, because the water column becomes denser during convection and therefore the sea surface becomes lower. Because this signal is small compared to variability in sea surface height induced by other processes, the local depth of deep
convection at a certain location and certain time does not correspond well to the sea surface height anomaly at that location and time. However, from the satellite altimetry data the approximate depth of deep convection (less than 1000 m, between
1000 and 1500 m or more than 1500 m depth) and the location of the convection area at a larger scale can be determined.