EVOLUTION OF NORTH ATLANTIC WATER MASSES INFERRED FROM LABRADOR SEA SALINITY

The Labrador Sea is the coldest and freshest basin of the North Atlantic. Winter cooling in this sea produces Labrador Sea Water. This intermediate water plays an important role in the exchange of heat, freshwater, and other substances between the atmosphere and the abyssal ocean, affecting the water masses, circulation, and, ultimately, climate of the subpolar North Atlantic basins. The subpolar gyre of the North Atlantic has exhibited large changes in temperature, salinity, and volume over the past six decades, largely in response to changing winter conditions over the Labrador Sea. The signature of these changes can be seen in the lower limb of the Meridional Overturning Circulation down into the North Atlantic tropics.


INTRODUCTION
salinity, which is a measure of (or, practically, a proxy for) the concentration of the salts dissolved in seawater.

B Y I G O R Y A S H AY A E V A N D A L LY N C L A R K E
This article has been published in Oceanography, Volume 21, Number 1, a quarterly journal of The Oceanography Society. Copyright 2008 by The Oceanography Society. All rights reserved. Permission is granted to copy this article for use in teaching and research. Republication, systemmatic reproduction, or collective redistirbution of any portion of this article by photocopy machine, reposting, or other means is permitted only with the approval of The Oceanography Society. Send all correspondence to: info@tos.org or Th e Oceanography Society, PO Box 1931, Rockville, MD 20849-1931 The large-scale pattern of temperature and salinity set by these processes means that each of the three main subpolar basins-the Iceland Basin, the Irminger Sea, and, particularly, the Labrador Sea ( Figure 1)-exhibit unique ranges of temperature and salinity that allow their waters to be distinguished from each other. This characterization of waters from their temperatures and salinities is an important tool of classical oceanography that underlies water-mass analysis of any level of complexity. In physical oceanography, a "water mass" denotes a large, relatively homogeneous or uniform body or volume of water, formed in the same source or formation region and by the same process (Dobrovolskiy, 1961). Water masses are defined by their physical (and chemical) properties. In this article, we use seawater salinity as the principal identifier and tracer of major water masses of the subpolar North Atlantic.   Seawater becomes denser as it gets colder and more saline. The impact of temperature on seawater density decreases as the water becomes colder and less saline. At a temperature of 10°C and a salinity of 34.85, a 1°C temperature change causes more than twice the density change as does a 0.1 salinity change. At 3°C, the difference is reduced to about 10%.

TE MPER ATURE AND SALINITY (THERMOHALINE) CHANGE S IN THE L ABR AD OR SE A
The Labrador Sea has been reasonably well observed over the past five decades, through periods of exceptionally active LSW formation and also periods when there was virtually none. In this section we give a more detailed, though not complete, overview of this water mass, its production, and associated changes.   .
The most remarkable event of convective water-mass renewal occurred between the mid-1980s and mid-1990s.
Convective cooling and freshening of the Labrador Sea's mid-depths produced a characteristic LSW that by 1994 became the coldest, densest, deepest, and most voluminous 3 in the entire observational record (Lazier et al., 2002;Yashayaev et al., 2003;Yashayaev, 2007aYashayaev, , 2007b.  (Yashayaev, 2007b). The warming and increasing salinity are maintained by the increasing influence of warm, saline 3 The LSW volume was determined via the thickness of the corresponding layer (e.g., Yashayaev, 2007b).

BOX 2. THE ROLE OF THE SUBPOL AR NORTH ATL ANTIC IN THE ATL ANTIC OVERTURNING CIRCUL ATION
Through production of its characteristic intermediate (LSW), deep (ISOW/NEADW), and abyssal (DSOW) water masses, the northern North Atlantic contributes to the meridional overturning circulation (MOC) of the whole Atlantic Ocean. These waters form the lower limb of the great ocean conveyor and subsequently participate in the ventilation of the deep layers of the world ocean. By transporting warm water poleward in the upper layers and cold waters equatorward at depth, this circulation is an important mechanism by which the climate system moves heat from the tropics to higher latitudes. It is also a mechanism to sequester carbon dioxide and other greenhouse gases in the ocean's abyssal waters. Changes in the properties and volumes of these water masses through changes in the processes leading to their formation are likely to be reflected in climate change and variability on the regional if not global scale.  Figure 2 Yashayaev and Dickson, 2008). This figure was constructed by averaging on density surfaces all of the available temperature and salinity data for each year for the central Labrador Sea. The black dashed contours are the σ 2 (potential density anomaly referenced to 2000 dbars) isolines (note the changing density of LSW). The data in the upper 200 m are not displayed because this layer is subject to strong seasonality, which is not adequately resolved by the observations.
Note that up to 1974, most of the data was collected by the US Coast Guard within their Ocean Weather Ship Bravo and Ice Patrol programs. From 1990 on, Bedford Institute of Oceanography has carried out an annual occupation of the AR7W section under the Ocean Climate Monitoring Program of Fisheries and Oceans Canada at the Bedford Institute of Oceanography (Figure 1). These regular sources of data are supplemented by a number of research and fisheries survey cruises conducted by various American, Canadian, Danish, and German groups. There were fewer oceanographic observations in the Labrador Sea during the 1980s with several years (1982, 1983, 1985, 1986, and 1989) Figure 5a and Lazier, 1980). The relationship between the NAO index and LSW production and properties evident on decadal time scales is not that straightforward from one year to next, because significant local processes force the ocean on interannual scales (Yashayaev, 2007b). For example, the water column establishes considerable stability during a sustained period without convection. It takes several consecutive cold winters to overcome this stability and to renew the densest LSW classes. Alternatively, a moderately cold winter following a series of particularly intense winters can still drive deep convection because the water column enters the cooling season with little stability to overcome. In addition, the NAO index is a large-scale regional climate index. The local wind field over the Labrador Sea is not always directly related to the strength of the westerlies over the central North Atlantic (Yashayaev, 2007b).   . As the LSW circulates around the subpolar gyre, its salinity increases because it is always mixing with greater-salinity waters, but it retains more or less the same density because its mixing is largely isopycnal .   1966 (b), 1994 (c), and 2004 (d). Sampling/profiling sites are shown in inserts. The earlier two sections arose from research cruises in the 1960s, the 1962 Erika Dan cruise out of Woods Hole Oceanographic Institution (Worthington and Wright, 1970), and the 1966 Hudson cruise out of Bedford Institute of Oceanography (Grant, 1968). The two later sections are taken from the western (AR7W) and eastern (AR7E) parts of the World Ocean Circulation Experiment (WOCE) repeat section AR7, which currently serves as the key outpost of the Climate Variability and Predictability (CLIVAR) program in the subpolar North Atlantic. LSW appears as a prominent feature on each of these sections although with varying strengths and properties. Some of these waters are likely to represent remnants of the colder and fresher LSW produced in the 1950s (Yashayaev et al., 2003). It appears that 1962 had the same relation to the 1950s as the recent years (e.g., 2004) had to the early-tomid-1990s (Figures 4 and 5). Similar to the recent observations, in 1962 we observe fragments of the warm and saline Icelandic Slope Water        Atlantic. These anomalies also move equatorward within the western boundary current and can be observed down to the tropics Yashayaev, 2000).

THE TR ANS -ATL ANTIC SIGNATURE OF L ABR AD OR SE A WATER
Changes in the strength of winter convection do not just alter the physical properties of the water masses, they also affect the water masses' relative layer thicknesses and volumes, and thus lead to changes in sea-surface height in the interior of the basins. The sea-surfaceheight changes regulate the strength of the subpolar gyre and also lead to changes in the dynamic height gradient between the subpolar gyre and the subtropical gyre, possibly influencing the volume transport of the North Atlantic Current (Curry and McCartney, 2001).
A weak gradient acts to contract the subpolar gyre as well as reduce the transport, pulling the subpolar front eastward and drawing more subtropical water into the eastern subpolar gyre (Hakkinen and Rhines, 2004;Hátún et al., 2005).
LSW contributes greatly to the formation of the water masses that eventually fill the deep basins of the northern North Atlantic and form the lower limb of the MOC. As ISOW and DSOW descend from their shallow sills in the narrow passages through the Greenland-Iceland-Faroe-Shetland-Scotland ridges, they mix briefly with surface mode waters, and, in a more prolonged way, mix with the intermediate LSW (Dickson et al., 2002) and increase their volumes by factors of four to six. If there has been strong convection in the Labrador Sea, then much of this entrained water has been in recent contact with the surface.
In this case, this mixing process is an effective mechanism to carry substances

ACKNOWLED GEMENTS
The authors thank the anonymous reviewer for very valuable comments and suggestions, and we also thank the many colleagues who over several decades have surveyed the subpolar North Atlantic, supported observing programs, and most generously helped in our research.