The SOIREE desktop survey was compiled in late 1998 and early 1999 and published on the WWW to provide background material for planning and executing the SOIREE iron fertilisation experiment in February 1999. During the time it was being prepared, the orientation changed somewhat as the critical planning issues changed. During the earlier stages of planning the main concern was to choose the site, with the two candidates being at 170°W 65°S (northern edge of Ross Sea) and 140°E 61°S (SSW of Tasmania). It soon became apparent that the latter site was more promising because of the much larger zonal separation of the major fronts there. The emphasis then shifted to gaining more detailed information about the 140°E site and anticipating the range of conditions that might be experienced during the experiment. Mixed layer depth was a particular concern: right up until the successful completion of the experiment, the planning team was nervous that the mixed layer depth might be so large that light limitation would prevent a phytoplankton bloom from occuring in response to iron fertilisation. (The climatological analyses suggested that a typical summer mixed layer depth was 60 m; this is very close to what was found and we now know that iron fertilisation does produce a bloom in such conditions.)
So the desktop survey was modified and extended in response to various planning concerns and then left in an unfinished state once the SOIREE experiment was completed. In some respects it has been superseded by an article by Trull et al. (in press), who present a more complete description of the seasonal evolution of mixed layer characteristics in the area. However it has been decided that it is still of sufficient interest to include in the companion CD-ROM for the Deep-Sea Research Part II volume dedicated to SOIREE. In preparation for this I have written this preface and lightly revised the document. These revisions, like this preface, are in italics.
The desktop survey consists of a pair of HTML documents, the first (this one) containing text and the second containing figures and captions. Below there is a section (the Guided Tour) containing links to the figures. The HTML files should be readable on recent WWW browsers (eg Internet Explorer version 4.0 and later, Netscape version 4.04 and later). Older browsers will not be able to view the PNG graphics format used for most of the figures. Note also that the total size of the graphics files linked from the Figures document is rather large at 2.63 MB, so machines with a slow processor or a modest amount of memory may take a long time to display them. If you encounter problems viewing the document please contact me and I will see what I can do.
I would like to acknowledge the input and guidance of other members of the SOIREE team, notably Phil Boyd, Ed Abraham, Steve Rintoul and Tom Trull. As will be apparent to anyone who reads the document, the SOIREE site assessment relied on a large body of published datasets and analyses and in particular on recent published and unpublished results from measurement programmes conducted by Steve Rintoul, Tom Trull and others at CSIRO Marine Research and Antarctic Cooperative Research Centre
Mark Hadfield (m.hadfield@niwa.cri.nz) 29 January 2001.
An iron release experiment is planned for the Southern Ocean south of New Zealand or Australia in February 1999.
The site requirements are:
The main obstructions to the Antarctic Circumpolar Current on its path around the Southern Hemisphere are the Drake Passage (70-40°W), Kerguelen Plateau (70°E) and the continental shelf and mid-ocean ridge system south of Tasmania and New Zealand extending eastwards into the Pacific (140°E-140°W). South of Tasmania and New Zealand, a mid-ocean ridge swings southwards to the Antarctic Continent as the Australian-Antarctic and Pacific-Antarctic Ridges. Although this ridge is deeper than 2000 m, it has a major effect on the position of the Polar Front(s).
The most recent, comprehensive review (Orsi et al. 1995) recognises three deep-reaching ACC fronts that are unbroken around the Southern Hemisphere. From the north they are the Subantarctic Front (SAF), Polar Front (PF) and Southern ACC Front (SACCF). The Southern ACC Front is distinct from the Southern Boundary of the ACC (SB) but the two are very close south of Australia and New Zealand.
The region south of the Polar Front is the one of interest, because water north of the Polar Front is unsuitable for the iron release experiment on two counts, being relatively low in silicate and having large summer mixed layer depths. The Polar Front is the northern boundary of a layer of cold "winter water" (see meridional sections later) at ca 100 m depth, overlying Upper Circumpolar Deep Water (UCDW), which is warmer, saltier and high in nutrients. In winter, the winter water is exposed to the surface, but in summer the surface layer warms by 2-3 °C, leading to a minimum in the temperature profiles.
South of Tasmania, the Polar Front is at ca 55°S but east of here it is deflected to the south, more or less following the Australian-Antarctic Ridge. South of New Zealand the Polar Front is at ca 60°S, which is its globally highest latitude and within a few degrees of the SACCF and SB. The large gap between the Polar Front and Southern ACC Front south of Tasmania (and west of here around to the Kerguelen Plateau) suggests that there is a wide band water with homogenoeus properties. Unfortunately it's not that simple. Recent CTD sections on the WOCE SR3 line have shown a southern branch of the Polar Front, at 59°S.
A quote from an email message from Steve Rintoul:
On the SR3 section [the southern half of which is along 140°E] there are two polar fronts, each is deep-reaching, and each corresponds to one or more Polar Front definitions used in the literature. This structure is not only found here, but also in the Drake Passage (Sievers and Nowlin, 1984) and the southwest Indian Ocean (e.g. Sparrow et al., 1996). The two fronts are sometimes called the "surface" and "subsurface" expressions of the PF, a terminology I'm not fond of since there are two distinct fronts, and both are deep-reaching.
The southern PF near 59°S is somewhat weaker than the northern one. It is very steady in position, based on our repeat CTD and XBT transects. The mixed layer depth and silica concentrations change across this front. North of the [southern branch of the PF] the silica concentration in the mixed layer is drawn down to a few µM by end of summer; south of this front (lat > 60°S), high levels of silica remain in the surface layer year-round. In winter I think the silica levels in the surface layer are about 8-9 µM north of 60°S, and substantially higher south of 60°S.
The southern PF is distinct from Orsi's Southern ACC Front.
You'll see in the paper I am sending you that I think part of the flow that crosses the section at the southern PF turns back to the north a bit further east and may re-cross the section, before turning back to the east again at the northern PF. that is, between 59°S and 53°S the flow is weak and probably more along the section than across it.
If you want high (>5 µM) silica in the surface layer in February, you may need to be south of the southern PF. Summer mixed layer depths in the 55-60°S band are about 80 m or so, and shoal a bit to about 50-60 m at 62°S.
And from Tom Trull:
The Polar Frontal Zone (PFZ) extends south from the SAF to the Polar Front (PF), a subsurface feature marked by the northern-most extent of cold subsurface waters (less than 2šC at about 200 m), typically found at about 54šS south of Australia. The Polar Zone (PZ) extends on southward to the Antarctic Divergence, which marks the transition from prevailing westerlies to coastal Antarctic easterlies, as well as the location of upwelling of circumpolar deep water. South of the Antarctic Divergence, the influence of sea-ice is large and we refer to this region as the seasonal sea-ice zone (SSIZ). South of Australia, the PZ can be further divided into north and south zones by the presence of a southern branch of the Polar Front (PF-S). The southern branch is less clearly defined than the northern but generally occurs between 57°S and 60°S. Indicators of the southern PF include the region where there is a rapid change in depth of the temperature minimum layer, or where there is a steep gradient in salinity in the 150-300 m depth range between water with a salinity of 34.0-34.2 to 34.6, which is characteristic of upper Circumpolar Deep Water.
All of the zones defined above can be characterized as high nutrient, low chlorophyll (HNLC) oceanic environments, although there are significant differences among them in terms of biological, physical and chemical attributes. Levels of chlorophyll a generally decrease southward across the SAZ to the Antarctic Divergence and then increase again south of the Divergence. Algal distributions within the water column also vary. Note the sub-surface chlorophyll maximum in the north Polar Zone along the SR3 section... The different Southern Ocean environments exhibit both overlap and differences in typical algal communities. For example, north of the SAF coccolithophores are relatively common, but are much less so south of the SAF, and particularly in the diatom-dominated Polar and Seasonal Sea Ice Zones. These biological variations are accompanied by changes in nutrient element abundances. Of particular note is that while nutrient levels generally increase southward, nitrate levels increase strongly across the SAF, but silicate levels remain low much further south, until the southern branch of the PF is reached.
Although sea surface temperature gives only an indirect view of upper-ocean processes, SST datasets based on satellite data have the advantage that they have much better data coverage than sub-surface measurements, therefore SST climatologies reveal sub-surface patterns. On a hemispheric scale, the mean SST is clearly related to the position of the deep-reaching fronts, eg in the concentration of the meridional gradient between 40 and 50°S in the southwest Atlantic and southwest Indian Oceans (where the SAF and PF are close to the STF), and the slack gradient at the same latitudes immediately southeast of New Zealand, where the SAF and PF are much further south. The main point of relevance to the present work is the way the isotherms between about 55 and 65°S are deflected sharply northwards as they cross the Australian-Antarctic Ridge and then sweep in a broad arc along the Pacific-Antarctic Ridge. The northwards deflection of the surface isotherms at the Australian-Antarctic Ridge is presumably related to the northwards flow observed by Steve Rintoul (above) between 59°S and 53°S on the SR3 section.
The amplitude of the annual variation is SST is of interest because it is a measure of the seasonality in the upper ocean, and because it is a related to the seasonal thermocline depth. In middle and high latitudes the seasonal variation in SST is driven largely by the annual cycle in insolation. The amplitude of the heat flux forcing associated with this cycle doesn't vary much with latitude poleward from 40°S or 40°N and the amplitude of the seasonal cycle is determined to first order by the seasonal thermocline depth, which determines the thickness of the layer over which the seasonal variation in heat content is distributed. In the Southern Ocean there is a hemispheric band of low annual amplitude in SST (< 1.5°C) associated with the westerly wind maximum. Over most of the Indian and Pacific Oceans there is a narrow band to the south of this with larger amplitude. Immediately next to the Antarctic Continent the annual amplitude is very small again because SST is close to freezing all year round.
Another factor that affects seasonal amplitude in SST is meridional movement of water masses. The SST in winter is tied to the temperature of the water a few hundred metres below the surface, but in summer the SST is determined more by the atmosphere. When water is displaced equatorward it becomes anomalously cold for its new latitude, thus in winter the SST is cold but in summer it warms up to nearer a typical temperature for that latitude; thus the annual amplitude is large. Conversely water displaced poleward is anomalously warm in winter, but less so in summer, so the annual amplitude is small.
The mixed layer depth (MLD) is crucially important for the iron release experiment because it affects both the dilution of the injected material and the ecosystem response. Unfortunately the sparsity of the subsurface data is a problem, because mixed layer depth varies considerably with the seasons and the passage of weather systems. The graphs here show MLD calculated from the summer-mean (December-February) temperature and salinity fields from the World Ocean Atlas. The MLD is taken as the depth at which the density exceeds the surface value by +0.1 kg m-3. (At typical temperatures this corresponds to a temperature difference of -1.1°C or a salinity difference of +0.13 psu. This is a fairly large difference physically, so the results are overestimates of the depth to which surface mixing typically extends.)
There is a belt of large summer MLD (>100 m) at around 55°S. It coincides roughly with the band of large wind speeds (below) and it is probably also related to the stability of the underlying water mass. South of this belt, summer mixed layer depth is smaller. At the candidate sites the summer MLD is around 50 m.
The figure shows mean wind speed for January-March for the Australia-New Zealand sector of the Southern Ocean. Because wind data are sparse there is considerable spurious variability between adjacent grid cells and the data are not contoured. Mean wind speed at the experiment site is in the vicinity of 7 m s-1. There is a band of larger mean wind speeds at 50-55°S.
Maps of mean surface salinity are shown from the World Ocean Atlas and AWI climatologies. They are similar except for some low-salinity features in the AWI dataset near the Antarctic continent at 150°E and in the Ross Sea. The surface salinity pattern does not correspond very closely with the position of the major fronts.
Annual-average nutrient fields near the surface from the World Ocean Atlas show the expected strong meridional gradient. Nitrate increases poleward throughout the Southern Ocean; silicate increases poleward from the Polar Front south. The World Ocean Atlas fields are very smooth; the WOCE SR3 transect below shows the relationship between fronts and nutrient gradients much better.
Temperature and salinity sections at at 140°E show the layer of cold "winter water" at ca 100 m depth south of the Polar Front, overlying warmer, saltier Upper Circumpolar Deep Water (UCDW). The AWI climatology was used for this figure in preference to the World Ocean Atlas because the latter is somewhat over-smoothed.
Like the AWI sections, this shows the Polar Front and the winter water. Also evident on these sections, and of considerable importance in determining the site, is the southern branch of the Polar Front at 59°S (see the section above on Fronts).
XBT transects between Tasmania and Antarctica give an indication of seasonal evolution and variability in the temperature profiles at the site. The layer of winter water at approx. 100 m depth south of 54°S is noticeable in all transects. In summer (Jan-Feb) there is invariably a layer of warmer water above it. Summer mixed layer depth at the site has a maximum of approx. 70 m.
Monthly SeaWIFS ocean colour composites generally show low surface chlorophyll at the site in late summer.
Tracks of FGGE drifters in the vicinity of the experiment site were analysed by Ed Abraham to estimate how far the fertilised patch would drift during the experiment
These figures were included to help the on-board team detect anomalous conditions at the site. From SST anomalies maps in the weeks before the experiment, the SST anomaly at the site was weakly negative. A large negative anomaly might have indicated that the mixed layer depth was anomalously large.