2.3. North Atlantic Deep Water Components

In the high latitudes of the North Atlantic there appear to be two main geographical locations where deep water that directly feeds the lower limb of the thermohaline circulation is formed, northeast of Iceland [cf. Mauritzen, 1994] and in the Labrador Sea [cf. Lazier, 1973]. The newly formed waters carried southward by the DWBC are collectively called North Atlantic Deep Water (NADW). The components of NADW [e.g., Wright and Worthington, 1970] that circulate throughout the North Atlantic [e.g. Reid, 1994], differ in their physical/chemical traits, yet they share common attributes. Newly formed NADW is high in transient tracer and oxygen concentrations, and low in nutrients. Classically, NADW has been thought to be composed of Labrador Sea Water (LSW) and Lower NADW (LNADW). The major components of the deep core of LNADW are Gibbs Fracture Zone Water (GFZW) and Denmark Strait Overflow Water (DSOW) [e.g., Swift, 1984], with a small addition of Southern Hemisphere Antarctic Bottom Water (AABW) [e.g. Amos et al., 1971; Top et al., 1987; Watts, 1991; Broecker et al., 1991]. The DSOW is the primary contributor to the tracer signal in LNADW [ Smethie and Swift, 1989].

It was not until transient tracers were measured that a warmer fourth component of NADW, Shallow LSW (SLSW), was recognized. Due to mixing with oxygen minimum water above, the SLSW is not an oxygen maximum [ Pickart, 1992a]. The SLSW were observed as tritium maxima at 4-5C in the subtropical western North Atlantic by Jenkins and Rhines [1980] and Olson et al. [1986]. Also Weiss et al. [1985] found a CFC maximum at the equator. Using an F11/F12 derived age, they estimated that the CFC maximum water had taken twenty-three years to travel in an undercurrent along the western boundary from the Labrador Sea to the equator with a mean velocity of 1.4 cm/s. Fine and Molinari [1988] documented the existence of the CFC maximum as an additional water mass that is carried by the DWBC between Abaco (26.5N) and Barbados (13N). They pointed out that the water mass was too warm to be classical LSW. Based on T-S characteristics of historical data, Pickart [1992a] identified the most likely region for convective formation of the SLSW to be in the southern Labrador Sea.

The waters, particularly at intermediate depths, in the Greenland/Iceland/Norwegian seas and Arctic Ocean form a system closely linked to the NADW. Bonisch and Schlosser [1995] estimated that 0.77 Sv of the deep water formed in the Greenland Sea and Eurasian Basin contribute to the formation of the NADW. Greenland Sea Deep Water tracer derived ages vary from twenty-four to forty- two years, which could be related to variability in its formation rate [cf. Peterson and Rooth, 1976; Bullister and Weiss, 1983; Smethie et al., 1986; Smethie et al., 1988; Heinze et al., 1990; Schlosser et al., 1991; Rhein, 1991; Schlosser et al., 1995; Bonisch and Schlosser, 1995]. In the Arctic Ocean, Fogelqvist [1985] measured CCl at a station in the Fram Straits. However, Krysell and Wallace [1988] published the first comprehensive oceanographic survey of CCl data, they were collected in 1987 in the Nansen Basin. Then Wallace et al. [1992] expanded on this work with a detailed investigation of the interpretation of CCl/F11, tritium/He, and F11/F12 tracer derived ages for Nansen Basin waters.

A prime concern in evaluating future climate change involves the stability of formation rates of the various components of the NADW. Formation rates and properties [ Brewer et al., 1983] of NADW can change on time scales as short as decades. During the last century, there have been at least two episodes of low salinity water outbreaks in the northern North Atlantic [e.g., Dickson et al., 1988; Lazier, 1973]. Lazier [1973] and Talley and McCartney [1982] documented decadal time scale changes in the LSW formation, showing there were decades in which convection was drastically reduced. Roemmich and Wunsch [1985] suggested that the increase in velocity below 3000 m in 1981 versus 1957 (IGY) data, could be due to the lack of LSW formation during the 1960s. It is expected that changes in the NADW formation will be reflected in the variability of the DWBC transport offshore of North America. Before such variability can be identified, it is first necessary to describe and understand the processes and time scales related to the DWBC circulation.