Draft study of Subantarctic Mode Water and Antarctic Intermediate Water formation

Lynne Talley (ltalley@ucsd.edu) 11/09/00, with some phone comments from K. Speer incorporated
Please feel free to adapt, adopt, modify, etc
Figures included here are from our own publications since they were easily available to me.

Subduction into the southern hemisphere subtropical gyres is dominated volumetrically by Subantarctic Mode Water and the densest of the SAMW's, Antarctic Intermediate Water (AAIW). SAMW refers to the thick mixed layers found directly north of the Subantarctic Front. A map of a proxy of winter time mixed layer depth ( Fig. 1 from Talley, 1999) shows that these layers are thickest in the eastern Indian Ocean and all across the Pacific. The sudden onset of thicker layers in the central Indian Ocean is an important mystery to be solved.

Figure 1. (left) Approximate proxy for winter mixed layer depth based on depth of oxygen saturation of 95% (Talley, 1999). (right) Schematic of mode waters: brown - subantarctic and subpolar MWs; red - subtropical MWs; pink - eastern subtropical MWs. Choice of which streamline to show was somewhat arbitrary. (Hanawa and Talley, 2000).

SEISAMW. In the Indian Ocean, the densest subducting water is also the thickest layer, and is formed in the southeastern Indian Ocean, south of Australia. It is often referred to as Southeast Indian SAMW or SEISAMW. It is the dominant mode of ventilation for the Indian Ocean, leading to a subsurface oxygen maximum.layer that extends northward, even into the tropical and northern Indian Ocean. The spread of this SEISAMW is illustrated by a map of potential vorticity at its core density ( Fig. 2, from McCarthy and Talley, 1999).

Figure 2. Potential vorticity at: (left) 26.88 gamma-n (neutral density), representative of the SEISAMW, with its source in the southeastern Indian Ocean; (right) 27.35 gamma-n, at the core of the Antarctic Intermediate Water salinity minimum in the Indian Ocean, with its low PV source upstream in the Atlantic, and a high PV source in the SEISAMW formation area south of Australia.

AAIW. In the Pacific Ocean, the densest subducting layer is also the densest of the global SAMWs, and is identical there to AAIW. AAIW is formed in the southeastern Pacific, just west of southern Chile, from where it spreads to the entire southern hemisphere and tropics throughout the globe (schematic in Fig. 3, from Talley, 1999). AAIW in the Pacific has a low potential vorticity signature ( Fig. 4, Talley unpublished manuscript), like SEISAMW in the Indian Ocean. The Pacific AAIW salinity minimum originates from this single source region, through northward subduction. The AAIW found in the Atlantic and Indian originates primarily from this source as well, as a portion of the new AAIW flows eastward through Drake Passage, north of the Subantarctic Front. Some modification of the AAIW takes place during this passage and then it spreads into the South Atlantic's subtropical gyre through eddy shedding and mixing out of the Falkland (Malvinas) loop. (This is not a wind-driven subduction process.) Based on oxygen, potential vorticity and salinity in the Atlantic and Indian Oceans, there is no source of AAIW in these oceans other than the southwestern Atlantic. Hence the primary site to study for AAIW formation is the southeastern Pacific, and the primary sites to study for AAIW spreading are Drake Passage and a subtropical section such as the quarterly XBT line at 30S in the Pacific.

Figure 3. Schematic of the global low salinity intermediate waters. Red X's mark the primary formation sites. Cross-hatching indicates regions of vigorous mixing important for setting initial properties. For AAIW, the primary convection site is west of Chile and Pacific AAIW comes from this region. Vigorous mixing of AAIW, which originates from the Chilean source via Drake Passage, occurs in the Brazil-Falkland current confluence east of South America, and differentiates the properties of the Atlantic/Indian AAIW from the Pacific AAIW.

Figure 4. Potential vorticity at sigma_1 = 31.7 (core of the Pacific AAIW) for (left) Pacific Ocean (Talley, unpublished) and (right) globe (Talley, 1996).

Climate impact. Why are these dense and volumetrically large subducting water masses important for climate? They form a large fraction of the subducted water of the subtropical gyres. They spread northward and cross the subtropical/tropical potential vorticity barrier and are found in a weaker form in the tropics. Variations in Southern Ocean surface properties are therefore communicated throughout the subtropical southern hemisphere and into the tropics through these water masses. In the Atlantic, AAIW is a marked part of the warm layer that moves northward, feeding the North Atlantic thermohaline overturning. The freshwater balance is considered to be of importance for long-term climate change (e.g. NADW formation), and AAIW is an important part of the Atlantic's freshwater balance. Thus changes in surface temperature and salinity that might, for instance, be due to the Circumpolar Wave or a stationary Antarctic dipole, or due to longer term changes in Southern Ocean conditions, would be communicated to low latitudes through these water masses.

Formation mechanisms. For several decades there have been two separate ideas about SAMW and AAIW formation. (a) The older concept was that AAIW is formed circumpolarly, through Ekman transport northward across the Subantarctic Front and then sliding northward down isopycnal surfaces.

(b) The newer concept (still dating to the 1970s, McCartney, 1976) is that SAMW is formed north of the Subantarctic Front through subtropical gyre processes and that AAIW is the densest of the SAMWs, and hence localized in its formation area. (SEISAMW has been recently recognized as a similarly important water mass with respect to its ventilation of the Indian Ocean. )

Both mechanisms are likely active. The northward Ekman transport across the SAF occurs into the local SAMW layer and may be important for creating the great thickness of the SAMW layer. WOCE hydrographic sections show intrusions across the SAF into the SAMW. Eastward transport of the SAMWs then eventually leads to both new SEISAMW, subducting into the Indian Ocean, and new AAIW, partiially subducting into the Pacific Ocean and partially passing on to the Atlantic/Indian subtropical gyre. The second source of water for the SAMW and AAIW is subtropical circulation poleward in the western boundary currents, in the traditional model of subtropical subduction.

Proposed observational program.A complete study of SAMW and AAIW formation, subduction and variability should include at least the following:

1. a circumpolar study of Ekman transport relative to the Subantarctic Front, including its strength and properties (temperature and salinity)
2. how much subtropical water enters each ocean's SAMW region north of the Subantarctic Front
3. eddy fluxes of mass, heat and salt in and out of the mode waters
4. improved air-sea flux estimates in the Southern Ocean, which have suffered from the extremely sparse observational network. With improvements in the air-sea fluxes, we can be more comfortable with the circumpolar budgets for SAMW.
5. tracking of SEISAMW and AAIW as they subduct and move northward, and as AAIW moves eastward through Drake Passage
6. intensive study of the formation regions of the deep SEISAMW and AAIW layers, that is, south of Australia and west of southern Chile

A strawman observational plan might thus include:

1. monitoring of Southern Ocean winds, SST and SSS, particularly between the Polar Front and Subantarctic Front and just north of the Subantarctic Front
2. historical data analysis in any regions with enough data to establish climatic changes in AAIW and SAMW: time and space structures, vertical penetration of the changes.
3. direct velocity measurements and broad-scale profiling that can be accomplished by ARGO
4. closely-repeated XBT/XCTD sections, extending to > 1000 m to completely capture the AAIW as well as SAMW, at 30S and at the choke points - Drake Passage, Tasmania and possibly Africa
5. repeat of the 1968 43S hydrographic section and others in the AAIW and SEISAMW formation regions to look at long-term variability throughout the water column (also useful for studying CDW and AABW).
6. analysis of eddy fluxes using altimetry and other satellite measurements
7. air-sea flux buoys appropriately distributed around the Southern Ocean
8. winter-time convection study in the AAIW formation region west of southern Chile, including one or two moored profiling T/S/velocity sites, gliders measuring T/S if available, and a late winter hydrographic survey.
9. a similar winter-time convection study south of Australia in the SEISAMW formation region.

For postscript versions of the plots: Click these with the right mouse button.
Figure 1 (left) postscript (Oxygen saturation map)
Figure 1 (right) postscript (Mode water schematic)
Figure 2 (left) postscript (Indian PV at 26.88 gamma - SEISAMW)
Figure 2 (right) postscript (Indian PV at 27.35 gamma - AAIW)
Figure 3 postscript (AAIW schematic - you might have trouble printing this postscript - gif file from above recommended)
Figure 4 (left) postscript (Pacific PV at 31.70 sigma 1 - AAIW)
Figure 4 (right) postscript (global PV at 31.70 sigma 1 - AAIW)
Additional potential vorticity map Pacific at 26.80 sigma theta (for Kevin Speer)
Pacific 26.80 sigma theta potential vorticity - gif file - click here to view

Sources for figures

Hanawa, K. and L. D. Talley, 2000. Mode Waters. Ocean Circulation and Climate, G. Siedler and J. Church, editors, International Geophysics Series, Academic Press, in press.

McCarthy, M. C. and L. D. Talley, 1999. Three-dimensional potential vorticity structure in the Indian Ocean. J. Geophys. Res., 104, 13251-13267.

Talley, L.D., 1996. Antarctic Intermediate Water in the South Atlantic. The South Atlantic: Present and Past Circulation, ed. G. Wefer, W.H. Berger, G. Siedler and D. Webb, Springer-Verlag, 219-238.