A haline

n\ *

thermal J—1-1-1-1-

\ ■ 1 1;'' r -1-1-1-1—

FIGURE 11.7. The zonally-averaged buoyancy forcing (thick black line) and the thermal (dotted line) andhaline (thin line) components that make it up, Eq. 11-4 in units of m2 s-3. Courtesy of Arnaud Czaja (Imperial College). Note that a heat flux of 50 Wm-2 is (roughly) equal to a buoyancy flux of 2 x 10-8 m2 s-3. Data from Kalnay et al. (1996).

Finally we note that Figs. 11.4, 11.6, and 11.7 only show components of the buoyancy flux associated with air-sea exchange; they ignore the effects of, for example, brine rejection in ice formation which is thought to be a key mechanism in the creation of dense water around Antarctica. Here ice is blown equatorwards in the surface Ekman layers (to the left of the wind in the southern hemisphere) leaving salty and hence dense water behind, which is susceptible to convection.2

11.1.2. Interpretation of surface temperature distributions

The observed SST distribution shown in Figs. 9.3, 9.14, 9.15, and 9.16 is maintained by (i) heat flux through the sea surface, (ii) heat flux through the base of the mixed layer by mixing and upwelling/down-welling, and (iii) horizontal advection.

Let us first consider tropical latitudes. Fig. 11.8 illustrates important aspects of the air-sea flux contribution there. We take a mean air temperature Tair = 27°C which yields a specific humidity of about 15 g kg-1 if the relative humidity is 70%. Assuming a mean wind-speed of 3 ms-1, the contributions of the Qsw, Qlw, Qs, and Ql are plotted as functions of SST using bulk formulae, Eqs. 11-6 and 11-7. The shortwave incoming radiation reaching the surface is assumed to be 341 Wm-2 and is offset by back radiation QLW, sensible heat loss QS (when SST > 27°C) and evaporative loss QL. Because QL rises very steeply with temperature, there is a natural limit on the tropical SST that depends on the radiation available and the wind speed. For the parameters chosen, energy supply and loss balance at an SST of about 30°C. At temperatures only slightly above this limit, evaporative losses far exceed the possible input of radiation, limiting SST. This evaporative feedback explains why tropical temperatures are so stable. In fact some energy is carried down

2At low temperatures typical of polar oceans, aT varies strongly with T and p and becomes smaller at lower temperatures and increases with depth, especially in the Weddell and Greenland Seas (see Table 9.4). The excess acceleration of a parcel resulting from the increase of aT with depth (known as the thermobaric effect), can result in a destabilization of the water column if the displacement of a fluid parcel (as a result of gravity waves, turbulence, or convection) is sufficiently large.


= incoming energy

net loss /


/ /



FIGURE 11.8. Contributions of the air-sea flux terms: Qsw, the incoming solar radiation, and Qlw, Qs, and Ql making up the net loss term (see Eq. 11-5), plotted as a function of SST using bulk formulae, Eqs. 11-6 and 11-7.

into the interior ocean and the solar energy is absorbed in the top few meters rather than right at the surface. Nevertheless, the limit on SST due to surface processes illustrated in Fig. 11.8 is very much at work in regulating surface temperatures in the tropics.

In the subtropics, evaporation is the principal heat loss term (see Fig. 11.3) and varies comparatively little throughout the year. Due to the seasonal cycle of insolation, there is a net deficit of energy in the winter and a net excess in the summer. Along with cooling to the atmosphere, upwelling of deep cold fluid in the subpolar gyres helps to maintain low surface temperatures in polar latitudes. At very high latitudes there is a net buoyancy loss out of the ocean in regions of very weak stratification. This, as discussed below, can trigger deep-reaching convection creating very deep mixed layers that can ventilate the abyss.

Small-scale turbulent mixing in the ocean allows temperature changes at the surface to be communicated to deeper layers, as sketched in Fig. 9.11. Wind-generated turbulence often creates almost isothermal conditions in the top 20-50 m of the ocean with a sharp discontinuity at the base. Another major factor in determining the distribution of surface properties is the pattern of Ekman pumping shown in Fig. 10.11. For example, in the equatorial belt the trade winds drive Ekman transport away from the equator drawing cold water up from below. The surface temperature can be actually lower directly along the equator than immediately to the north or south!

Such considerations of energy balance at the surface and wind-induced Ekman pumping and turbulence, provide a firstorder explanation of many of the major features of the SST distribution and its seasonal variation shown in Fig. 9.3, and reflect variations in the available solar radiation modified by air-sea fluxes of sensible and latent heat, advection (both horizontal and vertical), and mixing of properties with deeper layers. These processes are frequently interpreted in terms of vertical one-dimensional models which attempt to represent the turbulent transfer of heat and buoyancy through the mixed layer, by wind and convectively driven turbulence, and its communication with the ocean below through entrainment and vertical motion.

It is also worthy of note that the North Atlantic is warmer than both the North Pacific and the Southern Ocean at the same latitude in the same season. The warmth of the Atlantic relative to the Pacific is thought to be largely a consequence of differences in the surface wind patterns. In the North Atlantic the zero wind-stress curl line (along which the interior extension of the western boundary currents tends to flow) slants much more than in the Pacific, allowing the Gulf Stream to carry warm surface waters into far northern latitudes in the Atlantic.

11.1.3. Sites of deep convection

A comparison of Figs. 11.4 and 11.6 with Fig. 9.10 shows that there is no direct relationship between the pattern of air-sea buoyancy forcing and the pattern of mixed layer depth. This is because the strength of the underlying stratification plays an important role in "preconditioning" the ocean for convection. The deepest mixed layers are seen in the polar regions of the winter hemisphere and are particularly deep in the Labrador and Greenland Seas of the North Atlantic, where they can often reach depths well in excess of 1 km. Here the ambient stratification of the ocean is sufficiently weak and the forcing sufficiently strong to trigger deep-reaching convection and bring fluid from great depth into contact with the surface. Note that deep mixed layers are notably absent in the North Pacific Ocean. Waters at the surface of the North Pacific are relatively fresh (note, for example, how much fresher the surface of the Pacific is than the Atlantic in Fig. 9.4) and remain buoyant even when cooled. Deep mixed layers over wide areas of the southern oceans in winter are also observed, but they are considerably shallower than their counterparts in the northern North Atlantic.

Evidence from observations of mixed layer depth and interior tracer distributions reviewed below in Section 11.2.1, suggest that convection reaches down into the abyssal ocean only in the Atlantic (in the Labrador and Greenland Seas) and also in the Weddell Sea, as marked in Fig. 11.9. These sites, despite their small areal extent, have global significance in setting and maintaining the properties of the abyss. They are thought to play a major role in climate variability (see Section 12.3.5). Observations suggest that there are certain common features and conditions that predispose these regions to deep-reaching convection. First, there is strong atmospheric forcing because of thermal and/or haline surface fluxes (see Fig. 11.4). Thus open ocean regions adjacent to boundaries are favored, where cold, dry winds from land or ice surfaces blow over water inducing large sensible and latent heat and moisture fluxes. Second, the stratification beneath the surface-mixed layer is weak, made weak perhaps by previous convection. And third, the weakly stratified underlying waters are brought up toward

FIGURE 11.9. The annual-mean stratification of the ocean at a depth of 200 m, as measured by N/fref: i.e., the buoyancy frequency, Eq. 9-6, normalized by a reference value of the Coriolis parameter, fref = 10-4 s-1. Note that N/fref . 20 in regions where deep mixed layers are common (cf. Fig. 9.10). Sites of deep-reaching convection are marked in the Labrador Sea, the Greenland Sea, the Western Mediterranean and the Weddell Sea.

FIGURE 11.9. The annual-mean stratification of the ocean at a depth of 200 m, as measured by N/fref: i.e., the buoyancy frequency, Eq. 9-6, normalized by a reference value of the Coriolis parameter, fref = 10-4 s-1. Note that N/fref . 20 in regions where deep mixed layers are common (cf. Fig. 9.10). Sites of deep-reaching convection are marked in the Labrador Sea, the Greenland Sea, the Western Mediterranean and the Weddell Sea.

the surface so that they can be readily and directly exposed to buoyancy loss from the surface. This latter condition is favored by cyclonic circulation associated with density surfaces, which ''dome up'' to the surface, drawn upward by Ekman suction over subpolar gyres (Fig. 10.11). In places where deep convection is occurring, weak vertical buoyancy gradients are observed (see Fig. 11.9, which plots N, Eq. 9-6, at a depth of 200 m over the global ocean) and isopycnals dome up toward the surface (see Fig. 9.7).

Observations of deep convection

Observations at sea during deep convection are rare because of the inhospitable conditions in which wintertime convection occurs (see the photographs taken from a research vessel in the Labrador Sea in winter, presented in Fig. 11.10). The best-observed region of deep convection is the Labrador Sea. Fig. 11.11 shows sections of a and N through the Labrador Sea just before (fall 1996) and during/after wintertime deep convection. In October we see a near-surface stratified layer, some 500 m or so in depth, overlaying a relatively well-mixed intermediate layer, formed by prior convection. By March of the following year, however, convection triggered by cooling from the surface has broken through the stratified layer, mixing intermediate and surface fluid and leading to a well mixed patch of some 200 km in horizontal extent by, in places, 1500 m in depth. Just as in our studies of convection in water heated from below described in Section 4.2.4, cooling of the ocean from above results in convection which returns the fluid to a state of neutral stability with a well mixed column in which N —► 0. By the following fall (not shown but similar to the October section of Fig. 11.11) the mixed patch has been ''covered up'' by stratified fluid sliding over from the side. The water mass formed by convection in the previous winter, now exists as a subsurface bolus of fluid which in the subsequent months and years is drawn into the interior of the ocean. The process has been likened to a chimney, but is perhaps better described as analogous to the way in which a snake swallows an egg, as sketched in the schematic diagram setting out the phases of deep convection in the ocean shown in Fig. 11.12.

Was this article helpful?

0 0
Solar Power Sensation V2

Solar Power Sensation V2

This is a product all about solar power. Within this product you will get 24 videos, 5 guides, reviews and much more. This product is great for affiliate marketers who is trying to market products all about alternative energy.

Get My Free Ebook

Post a comment