Two Volumetric Approaches to Identification of Water Masses and Their Changes

It is clear from Section 24.5 that the term LSW covers more than one water type; the vintages of LSW convectively formed at different times take on different properties dependent on changes in various sources of freshwater and salt, and variations in the heat loss to the atmosphere. Only by distinguishing accurately between the water types can we truly understand their development. Simple averaging on depth or density levels across the subpolar gyre ignores the spatial and temporal evolution of LSW so presenting an inadequate picture. Here we describe a different approach that avoids those problems - the volumetric method.

The volumetric method is a particularly useful tool for identification and examination of specific water masses and studying their "life cycled. The "life cycle" of each water mass includes production, spreading and mixing. As soon as water mass production starts, a newly formed water mass begins to spread, advect and mix with other waters. The mixing process leads to both spatial and temporal transformation of a water mass, resulting in its structural and property changes. Both mixing and export are responsible for water mass dilution, dissipation and, ultimately, loss. The volumetric methods allow the identification of a water type and tracing of its development in time and space by automatically adjusting to those changes.

Several volumetric applications are discussed in Chapter 21 (Yashayaev and Dickson 2007). Here we summarise the LSW identification technique based on two complementary volumetric approaches (Yashayaev 2007b), the essence of which is reflected in Fig. 24.6. A convectively formed water mass can be reliably identified and monitored firstly by the density layer volumetric method. Each value in Fig. 24.6a represents the basin-mean thickness (in meters) of an individual o2 layer (o2 is potential density anomaly referenced to 2,000db) defined by o2 and time ranges (0.01 kg m-3 year). This o2-time layer thickness plot was constructed by averaging Ao2 = 0.01 kg m-3 layer thicknesses from all hydrographic stations in a

Fig. 24.6 (a) Evolution of LSW in the Labrador Sea: a "volumetric" o2-time plot showing the average thickness (meters) of Ao2 = 0.01 kg m-3 layers in the Labrador Sea (o2 is potential density anomaly referenced to 2,000db) (Reproduced from Yashayaev et al. 2007). (b) Temporal volumetric changes: 1994, 2000 and 2004 "volumetric" potential temperature (9)-sahnity (S) censuses of the Labrador Sea. (c) Spatial volumetric changes: 1995 "volumetric" 9-S censuses of the Labrador, Irminger and Iceland basins (Fig. 24.1). Each value in (b) and (c) represents the average thickness (in meters) of a 0.1 °C x 0.01 A9 x AS layer. The solid and dashed contours are isolines of o2 (kg m-3) defined by 0 and S

Fig. 24.6 (a) Evolution of LSW in the Labrador Sea: a "volumetric" o2-time plot showing the average thickness (meters) of Ao2 = 0.01 kg m-3 layers in the Labrador Sea (o2 is potential density anomaly referenced to 2,000db) (Reproduced from Yashayaev et al. 2007). (b) Temporal volumetric changes: 1994, 2000 and 2004 "volumetric" potential temperature (9)-sahnity (S) censuses of the Labrador Sea. (c) Spatial volumetric changes: 1995 "volumetric" 9-S censuses of the Labrador, Irminger and Iceland basins (Fig. 24.1). Each value in (b) and (c) represents the average thickness (in meters) of a 0.1 °C x 0.01 A9 x AS layer. The solid and dashed contours are isolines of o2 (kg m-3) defined by 0 and S

given year weighted by the distance or area represented by these stations. Two examples of the second approach, volumetric potential temperature (0) - salinity (S) analysis, are presented in Figs. 24.6b and 24.6c. Each analyzed layer in these 0-S diagrams was defined by two-dimensional 0-S intervals: A8 x AS = 0.1 °C x 0.01, set by ^A8 and ^AS in the corresponding directions. This approach was applied individually to all available annual sets of hydrographic data from the Labrador, Irminger and Iceland basins, resulting in annual volumetric 0-S censuses for each of these three basins.

Figure 24.6b shows three annual hydrographic surveys of the Labrador Sea (1994, 2000, 2004), while Fig. 24.6c shows the three basins of interest (Labrador, Irminger, Iceland) "sampled" by AR7 in the same year, 1995. Figure 24.6b, therefore, reflects temporal transformation or evolution of LSW and other waters; while Fig. 24.6c illustrates spatial water mass transformation and change.

Strengthening and deepening convection creates, remixes and modifies LSW, causing its thickness, density and other properties to change. The processes responsible for transformation and losses of LSW, including mixing, entrainment and export, also change the properties of this water mass, altering its core and boundaries. Indeed, the subpolar trans-Atlantic section plots (e.g., Figs. 24.2, 24.3 and 24.14), time series of vertical profiles (e.g., Figs. 24.5 and 24.8), volumetric inventories (e.g., Fig. 24.6) and compilations (e.g., Fig. 24.9) imply that a fixed (static) range of any seawater property or a combination of such ranges can not be used as a universal criterion identifying a specific LSW core, vintage or class (the terms "LSW cord" and "LSW class'' are introduced below). On the contrary, a characteristic property (e.g., density) range and other LSW identification criteria need to comply with the changes in the properties used. The methods that we use in identification and analysis of characteristic water masses are primarily based on o2 and 0-S volumetric censuses. These techniques are capable of automatic adjustment to a specific LSW core, "locking on" its year-to-year transformation and thus revealing its spatial and temporal changes.

0 0

Post a comment