Hydrodynamics Along Meridional Transects in the Southwest Indian Ocean Sector of the Southern Ocean During Austral Summer 2007

73

Twenty Sixth Indian Antarctic Expedition 2006-2008 Ministry of Earth Sciences, Technical Publication No. 24, pp 73-99

Hydrodynamics along Meridional Transects in the Southwest Indian Ocean Sector of the Southern Ocean during Austral Summer 2007

Alvarinho J. Luis National Centre for Antarctic and Ocean Research, Earth System Science Organization, Ministry of Earth Sciences, Headland Sada, Goa

ABSTRACT Preliminary analysis of expendable CTD profiles collected during the 26th Indian Antarctic Expedition along the ship route from Durban to India Bay, Antarctica (Track-1) and Prydz Bay to Mauritius (Track-2) during February– March 2007 are discussed. Vertical thermohaline structure reveal that the northern and southern branches of Subantarctic Front on both Tracks merge; likewise, the Agulhas Retroflection Front (AF) and South Subtropical Front (SSTF) merge between 43° and 44°S on Track-2. The southern branch of the Polar Front (PF2) meanders by 550 km southward towards east. The Subtropical Surface Water, Central Water and Mode Water are located north of 43.5°S, while the Subantarctic Surface Water, Antarctic Surface Water, Antarctic Intermediate Water and Circumpolar Deep Water are encountered in the Antarctic Circumpolar Current (ACC) region. Baroclinic transport relative to 1000 db reveals that the ACC volume is enhanced by 10 × 106 m3s–1 eastward, and a four- fold increase in transport occurs south of ACC. Nearly half of the ACC transport occurs in the 100–500 m slab.

INTRODUCTION The wind driven circulation in the Indian sector of the Southern Ocean (SO) consists of the anticyclonic subtropical gyre (STG), the eastward flowing Antarctic Circumpolar Current (ACC) and the westward flowing Antarctic Coastal Current. The STG comprises of the Agulhas Current (AC) to the west, the Agulhas Return Current (ARC) to the southwest, the South Indian Ocean Current to the south, the equatorward flowing West Australian Current to the east and the South Equatorial Current to the north. This gyre differs from its counterparts elsewhere in that most of the water re-circulates in the western and central parts of the 74 Alvarinho J. Luis ocean basin (Stramma and Lutjeharms, 1997). The AC, which is the strongest western boundary current in the southern hemisphere, transports southwards warm and saline water of tropical and subtropical IO origin amounting to 65 × 106 m3 s-1 (Stramma and Lutjeharms, 1997). These characteristics promote strongest eddy activity in the southwestern IO and generate quasi-permanent meanders and eddies (Weeks and Shillington, 1994). The Antarctic Circumpolar Current (ACC) is organized into a small number of relatively narrow and deep-reaching jets associated with the baroclinic shear (Nowlin and Klinck, 1986). Spanning 45° to 55°S, the ACC flows uninterrupted from west to east (Orsi et al., 1995), linking the different ocean basins, thereby allowing the exchange of climate relevant properties, which constitutes an important part of the global overturning circulation (Schmitz, 1996). The IO sector of the SO has an intricate frontal system of quasi-zonal fronts that merge, split and steer over the uneven bottom topography in the Crozet region (Park et al., 1993) and Kerguelen region (Belkin and Gordon, 1996 (hereafter BG96); Holliday and Reed, 1998). The identification of fronts and their meandering properties are essential elements in tracing the upper-level ocean circulation. At the choke points, such as the Drake Passage, south of Africa, south of Tasmania, the meridional spread of the SO dynamics is constrained, meaning that the ACC transport, water mass and frontal characteristics can be accurately monitored. Hydrodynamic characterization of the southwestern IO is warranted on the basis of a number of meteorological and oceanographic features. First, the positive wind stress curl promotes downwelling throughout the year at the rate of 0–20 cm-1 day (Hellerman and Rosenstein, 1993). Second, the coastally trapped waves are formed in the atmosphere around the south African coast, which propagate eastward as coastal low pressure system during the passage of synoptic pressure systems to the south (Gill, 1977); therefore the winds blow parallel to the east African coast. Third, the general wind patterns over the southern IO drive major currents; consequently, the wind-driven circulation is dominant over thermohaline circulation (Hellerman and Rosenstein, 1993). Fourth, the region exchanges a large heat with the atmosphere, which is received largely via the warm (16–26°C) and saline (35.5, salinity is unitless because it is measured from the ratio of conductivities) AC, which Hydrodynamics along Meridional Transects in the Southwest ... 75 becomes trapped in the Agulhas Retroflection. Fifth, among the world oceans, about 67% of total water volume with temperature between –2° and 2°C is present in the southwest IO, as this region lies immediately downstream of the Weddell Sea, where most of this water is formed (Emery and Meincke, 1986). Despite the above features, the study of hydrodynamics in the southwest IO is hampered by a lack of good quality and spatially- resolved hydrographic data. Although numerical models have been used to simulate specific hydrodynamic features (Wang and Matear, 2001), verification of the numerical solutions requires in-situ measurements. The criteria required to identify large-scale fronts, characterize water masses, and calculate volume transport within the ACC, as well as to determine the spatial and temporal variability of fronts in the Indian ocean sector of the SO, have been outlined in previous studies based on the historical hydrographic data (BG96; Holliday and Read, 1998; Lutjeharms and Valentine, 1984; Orsi et al., 1995; Park et al., 1993, 2001 (hereafter PEP01)Belkin and Gordon (1996). I.M. Belkin and A. Gordon, Southern Ocean fronts from the Greenwich meridian to Tasmania. Journal of Geophysical Research 101 (1996), pp. 3675–3696. Full Text via CrossRef | View Record in Scopus | Cited By in Scopus (235); Sparrow et al., 1996), sea surface temperature (Ts) images obtained by passive radiometer (Kostianoy et al., 2004), and by a combination of data from expendable bathythermograph (XBT) and satellite altimetry (Swart et al., 2008). Anilkumar et al. (2006) (hereinafter AN06) surveyed the meridional sections along 45°E and 57.5°E during the austral summer of 2004. They reported that the merged font, consisting of the Agulhas Retroflection Front (ARF) and Subtropical Front (STF), exhibits meandering by more than 220 km eastwards, and the Polar Front (PF) and Subantarctic Front (SAF) split into northern branch (PF1 and SAF1, respectively) and southern branch (PF2 and SAF2, respectively). Literature survey indicates that the choke point south of Africa has yet to be systematically and repeatedly sampled, except for several synoptic cruises such as KERFIX, SUZIL, WOCE, Antaras, Civa1 and Civa2 (Park et al., 1991; 1993; 1998a; 1998b; PEP01) and the Japanese Antarctic Research Expeditions (Aoki, 1997; Aoki et al., 2003 and references therein). These extensive studies provide a detailed background to spatial trends in the ocean-based synoptic surveys. 76 Alvarinho J. Luis

OBJECTIVES Repeated hydrographic observations are required to compare and quantify changes in the hydrodynamics over a period of many years; however, such observations are impractical because of the high costs of chartering research vessel. To overcome this problem, I have initiated a project to collect hydrographic data in the southwestern IO sector of the SO by using expendable CTD (XCTD) probes which can be launched from a moving ship. The hydrographic surveys were undertaken under the International Polar Year (IPY)-endorsed project (IPY#70) entitled “Monitoring of the Upper Ocean Circulation, Transport and Water Masses between Africa and Antarctica” (http://classic.ipy.org/ development/eoi/index.htm). The data were collected aboard M. V. Emerald Sea chartered for the 26th Indian Antarctic Expedition (IAE). Using the hydrographic data recorded along the ship route from Durban to India Bay and Prydz Bay to Mauritius, the aims of the study are to (1) characterize vertical structures of temperature, salinity and density; (2) identify and compare the location of the hydrological fronts with previous studies; (3) identify various water masses; and (4) compare the baroclinic volume transport between the two sections.

DATA AND METHODS The vertical profiles of temperature and salinity (% in Fig. 1) were recorded using XCTD manufactured by Tsurumi Seiki Company Limited (model: XCTD-3; terminal depth: 1000 m; temperature accuracy: ±0.02°C and salinity accuracy: ±0.03 mS cm–1). In shallow regions, only temperature profiles were recorded using Sippican-make T–7 XBT probes (accuracy: ±0.15°C depth resolution: 0.65 m) (+ in Fig. 1). The section from Durban, South Africa (29.87°S, 31.03°E) to New India Bay, Antarctica (69.9°S, 12.6°E) was occupied during 9–14 February 2007 (hereafter Track-1), while the section from Prydz Bay (69.3°S, 76.35°E) to Port Louis, Mauritius (20.7°S, 57.35°E) (hereafter Track-2) was surveyed during 19–26 March 2007. A comparison of XCTD-3 and Sea Bird CTD profiles reveals that the former is consistent with the manufacturer’s specified accuracy for temperature and salinity (Mizuno and Watanabe, 1998), and the fall rate for XCTD shows no systematic bias in the fall equation provided by the manufacturer (Kizu et al., 2008). The quality control for the profiles was carried out by adopting the following procedure (CSIRO Cookbook for Quality Control of Expendable Hydrodynamics along Meridional Transects in the Southwest ... 77

Fig. 1: Bathymetry from TOPEX/Poseidon (light gray contours, km; Smith and Sandwell, 1997) overlaid with XCTD (•) and XBT (+) stations covered along Track-1 during 9–14 February and along Track-2 during 19–26 March 2007. Also shown are Civa 2 track occupied during February–March 1996. Schematics in thick gray lines indicate the Agulhas Current system

Bathythermography (XBT) Data (1993). CSIRO Cookbook for Quality Control of Expendable Bathythermography (XBT) Data, 1993. CSIRO, Australia, p. 74.Bailey et al., 1993): (1) The geographic coordinates and the launch time of each XCTD were checked against the sampling logs. (2) Each profile was examined to eliminate readily visible malfunctions, such as broken wire and obvious bad profiles due to faulty probes were 78 Alvarinho J. Luis rejected. (3) Each profile was inspected for spikes caused by external electromagnetic interference and insulation penetration, temperature inversion due to wire stretch and leakage etc. (4) Profiles with high frequency spikes and temperature inversions greater than 0.2°C were rejected. (5) Waterfall plots for each track were generated to evaluate further the consistency of the profiles, and those profiles with temperature

Table 1: Axial property indicators for identification of oceanic fronts Frontal structure Property identifier for fronts Reference Temperature (°C) Salinity è Southern branch 0 = 3–2° S0 = 33.8–33.9 Holliday and Reed of Polar Front (1998) (PF2) è Northern branch min layer at 200 m; S0 = 33.8–33.9 Belkin and Gordon of Polar Front northern limit of 2° (1996); Holliday and (PF1) isotherm below 200 m Reed (1998) è Southern branch 0 = 8–9° Consistent S0 = Holliday and Reed è of Subantarctic 200 = 4° 33.85 (1998); Belkin and Front (SAF2) Gordon (1996) è Northern branch 0 = 9–10° S200 = 34.11– Belkin and Gordon è of Subantarctic 200 = 8.4–4.8° 34.47 (1996) è Front (SAF1) 400 = 3.7–6.1° S400 = 34.19– è 600 = 3–4.3° 34.38

S600 = 34.24– 34.28 è Southern branch 0 = 10.3–15.1° S0 = 34.3–35.18 Belkin and Gordon è of Subtropical 200 = 8.0–11.3° S200 = 34.42– (1996) è Front (SSTF) 400 = 6.1–8.6° 34.92 è 600 = 4.2–5.9° S400 = 34.38– 34.63

S600 = 34.30– 34.44 è Agulhas Front 0 = 17–19° S0= 35.54–35.39 Belkin and Gordon è (AF) / Agulhas 300–800 = 10° S200 = 35.57– (1996); Holliday and è Retroflection 200 = 14° 34.9 Reed (1998) Front (ARF) è Northern branch 0 = 21–22° S0 = 34.87–35.58 Belkin and Gordon è of Subtropical 200 = 12.1–15.3° S200 = 34.99– (1996); Holliday and è Front (NSTF) 400 = 9.3–12.2° 35.42 Reed (1998)

S400 = 34.71– 35.07 Hydrodynamics along Meridional transects in the Southwest ... 79

Table 2: Criteria for water mass identification Water mass Water-mass characteristics Reference Temperature (°C) Salinity Subtropical Surface 16–28 > 35.1 Valentine et al. (1993) Water Central Water 1. Southeast Atlantic 6–16 34.5–35.5 Valentine et al. (1993) Ocean 2. Southwest Indian 8–15 34.6–35.5 Valentine et al. (1993) Ocean Mode Water 11–14 35–35.4 Anilkumar et al. (2006) Subantarctic Surface 9 < 34 Park et al. (1993) Water Antarctic Surface < 5 < 34 Anilkumar et al. (2006) Water Antarctic Intermediate 2.2 33.87 Valentine et al. (1993) Water Circumpolar Deep 2 34.77 Anilkumar et al. (2006) Water offset greater than 0.5°C were rejected in this process. (6) High frequency noise in the salinity profiles was minimized by applying a median filter with a 15 m window, following Xiaojun et al. (2004). The hydrological fronts and water masses were identified by assessing the criteria summarized in Table 1 and Table 2, respectively. The geostrophic velocity across a pair of stations was calculated relative to the deepest common level (1000 db) available from the XCTD data using the following the method (Pond and Pickard, 1993):

. . . (1)

where f is the Coriolis parameter (s–1) at a mean latitude, dz is depth interval (m) and ÄÖ is geopotential anomaly (m2s–2) between adjacent pair of XCTD stations. We also used the maps of absolute dynamic topograpy (ADT)” provided on a 1/3°×1/3° Mercator grid by merging 80 Alvarinho J. Luis

TOPEX/Poseidon, JASON-1, ERS-1/2 and Envisat altimetry data, and distributed by CLS/AVISO (http://atoll-motu.aviso.oceanobs. com). Because the ACC is characterized by fine scale structures and variability, “up-to-date” absolute dynamic topography and absolute geostrophic velocity data have been used. Details on the mapping methods and different corrections applied to these fields are available in Le Traon (1998), Le Traon et al. (1998) and Ducet et al. (2000). Sharp vertical gradients were observed in the vertical thermohaline structure down to 1000 m, which suggest that it is inappropriate to assume the level-of-no- motion at 1000 db for estimation of the XCTD-based volume transport; hence, the transport estimated from equation (1) is compared with that estimated from altimeter (MADT) data using, gD T = rSSH, . . . (2) surf f where g is gravity (m s–2), D is the thickness of the water column (m), and rSSH is the sea surface height difference (m). Andres et al. (2008) has demonstrated that rSSH across the Kuroshio can be used as proxy for full-water-column transport. In order to take care of varying distance between stations, volume transport per kilometer for the 0–1000 m slab was also computed. Sea surface convergence was computed by using the altimeter-based geostrophic velocity components to substantiate changes occurring in the vertical structures. Positive (negative) convergence indicates downwelling (upwelling).

RESULTS AND DISCUSSION Hydrological Fronts Variability An oceanic front is a narrow zone marked by enhanced water property gradients of temperature, salinity, chlorophyll, nutrients etc., at various levels and enhanced velocity. In this work, temperature and salinity are used as criteria for identifying the locations of the fronts (Table 1 because these can vary as a result of gradual modification of the adjacent water masses by air-sea interaction and cross-frontal mixing (BG96). It is practical and useful to locate the central position of the front by using the values of water properties at axial locations at a given è è depth; e.g., temperature at surface/200 m ( 0/ 200) and salinity at surface/

200 m (S0/ S200) (PEP01). Several surface criteria, such as surface Ts, Hydrodynamics along Meridional Transects in the Southwest ... 81 salinity and Ts gradients have been proposed to identify ocean frontal structures (Kostianoy et al., 2004), but these can vary with seasons and geographical locations. Focusing on subsurface phenomenological indicators for the front identification, We used the surface features, such as Ts gradient and geostrophic velocity magnitude obtained from MADT data, as guide in the identification process. The interaction between the waters from the Weddell, Scotia, and Bellingshausen Seas and local thermodynamic processes in the southern part of the Scotia Sea gives rise to the Weddell-Scotia Confluence (Gordon, 1967). It is bounded by the Scotia Front (SF) in the north and a weaker front in the south, the Weddell Front (WF). The WF signature is not clearly identifiable from Figure 2, because the last station was occupied at the northernmost sea-ice extension limit. The SF is located at 55°S, based on the extent of 1°C isotherm between 300 and 500 m, which concurs with the location in BG96. The location of SF is characterized by weak surface divergence (2 × 10–9s–1) (Fig. 2e). The PF, which forms a veritable wall between the Agulhas Basin to the north and the Enderby Basin to the south, splits into two branches.

The northern Polar Front (PF1) is located at 49.5°S, based on Tmin layer at 200 m and the northernmost extent of 2°C below 200 m (Park et al., 1993; BG96; Park and Gambéroni, 1995), and its southern branch (PF2) spans 50.5° to 52°S. The surface velocity magnitude associated with PF1 and PF2 is 10 and 14 cms–1, respectively. Surface convergence (14 × 10–9 s–1) coincides with sharp vertical gradients at PF1, while surface divergence (1.5 × 10–9 s–1) and weak vertical gradients occur at PF2 (Fig. 2e). PEP01 identified PF1 as a single front at 51°S on the 30°E section, while AN06 identified PF1 spanning 49° to 50°S and PF2 spanning 52° to 54°S on the 45°E section; the position of PF2 in AN06 coincides with PF location of BG96. The branching of PF has been attributed to the highly convoluted meander or detached cold and warm eddies (Park et al., 1998b), and associated with the baroclinic shears between 48° and 52°S (Fig. 2d). The SAF, which is defined by the location of a rapid descent of the subsurface salinity minimum of 33.8 (Holliday and Reed, 1998; BG96), marks the beginning of the northward spreading of Antarctic Intermediate Water (AAIW) below 400 m (Fig. 4a-c). The southern branch of this 82 Alvarinho J. Luis

Fig. 2: (a) Surface temperature gradient (0.1°C km–1) overlaid on surface geostrophic velocity estimated from altimetry-based absolute dynamic topography (ADT) (cms–1), and vertical structure of (b) temperature (°C), (c) salinity (psu), (d) sigma-t (kgm–3), and (e) sea surface divergence (10–7s–1) computed from ADT data, along Durban–India Bay route (Track-1 in Fig. 1). Positive divergence indicates upwelling. Abbreviations are SF: Scotia Front; PF1 and PF2: North and South Polar Front; SAF1 and SAF2: North and South Subantarctic Front; SSTF: South Subtropical Front, respectively; AF: Agulhas Front; and AC: Agulhas Current Hydrodynamics along Meridional Transects in the Southwest ... 83 front (SAF2) extends from 45.5° to 47°S, which is associated with surface velocity of 13 cms–1, while its northern counterpart (SAF1) spans 44.5° to 45.5°S, which is within 2.5° width of SAF (Lutjeharms and Valentine, 1984). SAF1 is associated with a mean surface velocity of 17 cms–1. The Civa2 section also shows that SAF splits into SAF1 and SAF2, which are located at 47°S and 50°S, respectively (PEP01). On the other hand, AN06 identified SAF1 between 40.2° and 43°E and SAF2 between 47° and 48°S on the 45°E section. It is noted that SAF1 location coincides with strong convergence (7.21 × 10–7s–1) and enhanced gradients in temperature and salinity structure, while SAF2 region is conducive to divergence (8.3 × 10–8s–1). The Subtropical Front (STF) demarcates the boundary between the Subantarctic Surface Water and the Subtropical Surface Waters and defines the northern edge of the ACC (Deacon, 1937). The STF has been identified as a broad Subtropical Frontal Zone consisting of several fronts interspersed by zones of relatively homogenous waters (Lutjeharms et al, 1993). Following the criteria of Orsi et al. (1995), STF is located at 42°S, whereas BG96 identified it at 43°S (Fig. 1). It is pertinent to note that the thermohaline structure (Fig. 2b-d) reveals that the STF has a double structure consisting of a northern branch (NSTF) and a southern branch (SSTF), as recognized previously in all the sectors of the SO (BG96). From the Civa2 data (PEP01) the SSTF is detected at 42°S, which is associated with enhanced surface velocity of 21.5 cms–1; this is in contrast to a merged SAF1+SSTF+ARF on the 45° section reported by AN06. Stramma (1992) also has reported a band of enhanced current in the upper 1000 m, just north of the STF near south Africa. Across the SSTF, Ts falls from 15° to 10°C, salinity drops by 0.4 and weak density gradients are observed in the vertical from surface to 400 m. Pockets of high salinity (34.8) at 100 m depth in the 40°–41.5°S belt are associated with strong interleaving in the underlying waters, as reported by PEP01, to be the most prominent characteristic of the upper-layer thermocline structure in the Subantarctic zone near 30°E. The AC extends down to 39°S from Durban and is marked with an enhanced velocity magnitude of 35 cms–1 at 34°S. The AF is located at 39°S, which is marked by enhanced surface speed of 55 cms–1 (Fig. 2a). Accounting for a distance of 120–220 km from the coast to the first two stations, AC core speed was found to be 17 cms–1, which is 6 cms–1 lower 84 Alvarinho J. Luis than that reported by Bryden et al. (2005). The discrepancy between the two estimates is due to the degradation of altimeter measurements of MADT close to the coast, which is reflected on the velocity magnitude. It is also noted that the vertical velocity shear is smaller in the AC than the rest of the section and that the horizontal density gradient is strongest (6.3 × 10–3 kgm–3 km–1) in AC (Fig. 2a). The water characteristics vary significantly across the AC. Ts decreases from 27° to 19°C and salinity is diluted from 35.5 to 34.8 due to the mixing of warm tropical water and subantarctic water, as a part of the retroflection process from the south African coast to 42°S. Strong stratification above 200 m near the coast (Fig. 2d) is attributed to the presence of high-salinity (>35.5) Red Sea Intermediate Water (RSIW) (Roman and Lutjeharms, 2007). As the RSIW flows southward from Red Sea, it mixes isopycnally with the equatorward-flowing AAIW, and flows into the Mozambique Channel in a series of anticyclonic eddies, eventually entering the AC (de Ruijter et al., 2002). We now discuss the fronts along Track-2. Figure 3 (a-e) depicts the Ts gradient overlaid on surface velocity from MADT, vertical structure of temperature, salinity and density, and surface convergence on Track-2. The following fronts are identified: PF2, PF1, SAF2, SAF1, SSTF, ARF and NTSF. PF2 spans 220 km from 55° to 57°S, associated with increase in Ts by a degree (Fig. 3b), an isohaline surface layer of 33.8 (Fig. 3c), and surface velocity of 6.3 cms–1. The northern limit of 2°C isotherm below 200 m constraints the location of PF1 to 48.5°S, marked by a surface velocity of 6.3 cms–1. PF1, located at 48.5°S, meanders northward by about 110 km when compared to its position on Track-1. Based on the location of 4°C isotherm at 200 m, SAF2 is located between 45° and 46°S, consistent with the location on Track-1. This merged front is marked by strong surface velocity (16 cms–1) and surface divergence (2.87 × 10–7s–1). The steepest Ts gradient (0.1°C km–1) and strong surface velocity (57 cms–1) at the Crozet Plateau correspond to the merging of SSTF and AF over the belt 43.5°–44.5°S. The steeply sloping isolines in the thermohaline structure (Fig. 3b-d) promote a strong geostrophic flow with an average speed of 50 cms–1 in the upper 250 m and an enhanced transport (24 Sv) (Fig. 6a). Hydrodynamics along Meridional Transects in the Southwest ... 85

Fig. 3: Same as in Figure 2, but for the Prydz Bay–Mauritius ship route (Track-2). Additional abbreviations are NSTF: North Subtropical Front, AgNBF-Agulhas Northern Boundary Front 86 Alvarinho J. Luis

Kostianoy et al. (2004) have reported that the ARF, STF, and SAF episodically merge to form the Crozet Front (BG96) in the 35°–60°E belt, whereas in the region of ARF, the ARF meander from STF/ SAF by 3° latitude on 43°E. The merged front (SAF1+STF+ARF) has been reported based on satellite observation of Ts (Kostianoy et al., 2004) and based on hydrographic data it has been placed at 41.5°S along the 45°E section (AN06). While SAF1 (identified between 45.5 and 46.5°S) splits from the merged ARF+STF (identified between 43.5° and 45°S) on 57.5°E section (AN06), Figure 3 reveals that SAF1 and SAF2 merges in the 45°–46°S belt. An enhanced Ts gradient (0.1°Ckm–1), which marks the northern ACC edge and the southern boundary of the subtropical waters, indicates a merged front (SSTF+AF) between 43° and 44°S, marked by a surface velocity of 60 cms–1. The AF is also marked by strong surface velocity (27 cms–1). A pycnostad is identified at 100–150 m depth with potential density anomaly of 25.7–26.2 kgm–3. The presence of high salinity pockets (> 35) in the upper 100 m reflects baroclinic instabilities associated with the AC retroflection, which shedding of eddies which transport warm and saline water into the subtropical IO via several northward branching jets (Lutjeharms and Ansorge, 2001). North of ARF, Ts and salinity increases from 17° to 21°C and from 34.8 to 35.3, respectively, leading to strong stratification above 200 m. With a surface velocity of 14 cms–1, the NTSF (the northernmost front of subtropical zone) is identified at 36°S, 220 km north of the location proposed by BG96. WATER MASSES Figure 6 shows è–S scatter diagrams for Track-1 (left-hand panels) and Track-2 (right-hand panels). The water masses are identified using the è–S criteria listed in Table-2, and the nature of the individual water masses is discussed below. Tropical Surface Water (TSW) in the AC consists of water from the southwest IO anticyclonic gyre (Gordon et al., 1987) and variable flow from the Mozambique Channel (Lutjeharms, 1972). Because of excess evaporation over precipitation in the subtropical zone, the TSW records high temperatures (16–28°C) and salinity exceeding 35.1 (Figure 4a, b). The TSW is transported by the AC into the retroflection region near Hydrodynamics along Meridional transects in the Southwest ... 87

42°S, 18°E where it sinks and forms a salinity maximum throughout the south IO, referred to as Subtropical Surface Water (STSW). STSW is detected north of 40°S and 43.5°S (i.e. the AF region in Figures 2 and 3) on Track-1 and -2, respectively. Central Water (CW) occurs below the STSW. It is formed at the Subtropical Convergence when the mixed Subtropical and Subantarctic Surface Water (SASW) masses sink and spread northward (Orren, 1966). The CW in the AC region receives a contribution from South IO and from the South Atlantic Ocean. The former is characterized by è of 8– 88 Alvarinho J. Luis

Fig. 4: è–S diagram for hydrographic stations occupied on Track-1 and Track-2. (a) and (b) 32° to 39°S, (c) and (b) 40° to 48°S, (e) and (f) 49 to 57°S, and (g) and (h) south of 58°S, on Track -1 (left-hand panels) and Track-2 (right-hand panels). TSW: Tropical Surface Water; STSW: Subtropical Surface Water; MW: Mode Water; CW: Central Water; SASW: Subantarctic Surface Water; AAIW: Antarctic Intermediate Water; AASW: Antarctic Surface Water; CDW: Circumpolar Deep Water

15°C and S of 34.6–35.5, while the latter is characterized by è of 6–16°C and S of 34.5–35.5 (Valentine et al., 1993). The South Atlantic Water enters the Southern Agulhas region as a blend of thermocline water and SASW from the south (Gordon, 1981, 1985); hence, these two water Hydrodynamics along Meridional Transects in the Southwest ... 89 masses have similar è–S ranges. CW is detected to the north of 41°S and 43°S on Track-1 and -2, respectively (Fig. 4a-d), in the depth range of 200–1000 m. The Mode Water (MW) is formed by wintertime convection in the area just to the north of the ACC, and appears as pycnostad or thermostad below the seasonal pycnocline (Park et al., 1993). It contributes volumetrically to the CW of the south IO subtropical gyre by ventilating the upper portion of the permanent thermocline, and is characterized by è of 11°–14°C, S of 35–35.4, and density anomaly of 26.5–26.8 (AN06). MW is identified south of 39° and 43.5°S on Track-1 and -2, respectively (Fig. 4a, b). The features of MW identified on Track-2 are consistent with the findings of Park et al. (1991), wherein they traced its source to the Subtropical Mode Water in the Crozet Basin and noted that the MW is produced locally by frontal mixing to the east of the Crozet Islands (42°–45°S). It is noted that the strong winds, which produce large turbulent heat loss (see Fig. 3c & 3d in “Validation of air–sea interface parameters in the southwestern Indian sector of the southern Ocean,” in this issue), promote convection that leads to the formation of MW, as suggested by Park et al. (1991). The MW identified along Track-1 is fresher by 0.2 than that along Track-2, suggesting that it stems from winter overturning of the subtropical water of Agulhas origin (McCartney, 1977). The water masses of the ACC region include the Subantarctic Surface Water (SASW), Antarctic Surface Water (AASW), AAIW, and Circumpolar Deep Water (CDW). The SASW is found near the southern boundary of the frontal zone, which is characterized by è ~ 9°C and S < 34 (Park et al., 1993). On Track-1, SASW is encountered in the upper 75 m just north of SAF1 at 44.5°S, whereas on Track-2, it is detected in the depth range of 100–300 m at 44°–46.5°S, encompassing SAF1 and SAF2. On 45°E section, it has been detected between 43° and 45°E (AN06). The AASW is encountered at the PF (Fig. 1) and occurs in the upper 100 m south of 45.3°S along Track-1 and south of 46.5°S on Track-2. On the 45°E section, it shows a wider meridional distribution, between 44° and 56°S, possibly due to blocking of ACC by islands and meandering of the flow over uneven bottom topography (AN06). The AASW has two components: Summer Surface Water (SSW) and Winter 90 Alvarinho J. Luis

Water (WW). The WW is detected south of 52.7°S on Track-1 and south of 57.5°S onTrack-2. AAIW is formed continuously near the PF spanning 50° to 55°S, where water with è ~ 2.2°C and S ~ 33.87 sinks and spreads northward (Sverdrup et al., 1942). This water is characterized by a temperature of 4.4°C and a salinity minimum of 34.4. AAIW is generally found at 800– 1200 m depth in the southwest IO (Wyrtki, 1971), and is detected between 38° (at 1000 m) and 44°S (at 600 m) on Track-1 and between 41°S (at 1000 m) to 43.5°S (at 500 m) on Track-2. AAIW is detected between 31° and 41°S at 1150–1200 m on the 45°E section (AN06). Below the AAIW, the CDW occurs, which is a voluminous water mass in the SO characterized by è of 2°C and S of 34.77. This is the critical water mass because of its involvement in the formation of all other water masses via vertical and lateral mixing induced by polar easterlies in the Antarctic slope and shelf regions (Whitworth et al. 1998). The CDW consists of two layers: the upper and lower CDW. Both layers are warmer than the overlying cold surface water. In the southern Agulhas region, Gordon et al. (1987) found traces of CDW based on quasi-synoptic data. Because these observations were restricted to 1000 m, it is not possible to discriminate between the lower and upper CDW (as done by PEP01); however, CDW is encountered south of 49°S on both tracks.

Baroclinic Transport Figure 5 shows baroclinic transport referenced to 1000 db across Track-1. The net transport amounts to 52 Sv (Sv = 106 m3 s–1), which is about 60% of that estimated across a section from Good Hope to Antarctica in the eastern Atlantic Ocean (90±2.4 Sv; Swart et al., 2008). Of the total transport, 35.7 Sv is confined to the ACC region, (spanning 57°S and 42°S), –0.2 Sv flows across the section from 57°S to 55°S and 16.5 Sv is confined to the AC system north of 42°S. The distribution of transport within slabs of different depth in the ACC region is as follows. A transport of 15 Sv is distributed between the 0–100 and 500–1000 m slabs, and 20.7 Sv is confined to the 100–500 m slab, where the slab-mean velocity is 16 cm s–1. Now we focus on transport associated with individual ocean fronts. The AC contributes a southwestward transport of 25 Sv (consistent Hydrodynamics along Meridional transects in the Southwest ... 91

Fig. 5: Baroclinic transport (Sv = 106 m3s–1) relative to the deepest common level between the XCTD stations across Track-1 for different depth layers: (a) 0–1000 m, (b) 0–100 m, (c) 100–500 m, (d) 500–1000 m, and (e) comparison of geostrophic transport (per kilometer distance between stations) computed using XCTD and altimeter- based absolute dynamic topography (ADT). Positive transport occurs eastwards. The locations of fronts identified in this work are shown in the top panel. See Figure 1 for abbreviations. 92 Alvarinho J. Luis with the range of 25–35 Sv reported previously; Harris, 1972; Stramma and Lutjeharms, 1997), with a velocity of 67 cms–1 at 38.6°S (Fig. 2a). Because Track-1 is sub-parallel to the direction of AC, the transport estimate above 1000 m represents just 25% of that estimated by the full- depth Civa2 section (100 Sv), for which the sampling track was oriented normal to the coast (PEP01). The ARC transports 41.5 Sv eastward, representing 40% of that reported (104 Sv) for the full-depth Civa2 section and 70% of the South Indian Ocean Current (60 Sv; Stramma and Lutjeharms, 1997). The largest transport of 20 Sv is associated with AF. The opposing east-west transport near the AF (39°–40°S) arises from meandering of the ARC. In the ACC region, the net transport of 6 Sv is associated with SSTF, 7 Sv with SAF1 and SAF2, 7.5 Sv with PF1, 9 Sv with PF2, and about 3 Sv with PF2 and SF. The total contribution from these fronts represents 85% of the total ACC transport (35.7 Sv). A comparison of XCTD-based volume transport per kilometer width with altimeter data (Fig. 5e) indicates good qualitative agreement between the two independent data sets. For example, both data sets depict eastward transport at AF, PF1 and PF2, and SAF1+SAF2, while westward transport occurs at 47°S, 39.6°S and in the AC region. It is noted that the altimeter-based net transport for the 0–1000 m slab across Track-1 is 2.5 Sv, which is nearly three times larger than that estimated from XCTD data. The net volume transport across Track-2 amounts to 54 Sv, of which 4 Sv is confined to the region south of the ACC (below 59.5°S), 45 Sv to the ACC and 5 Sv in the region north of 44°S (Fig. 6). A slab- wise analysis of transport in the ACC region reveals that 9 Sv is confined to 0–100 m slab, 26 Sv to 100–500 m slab and 10 Sv to the 500–1000 m slab. The transport associated with individual fronts is as follows. A westward transport of 3 Sv is associated with NSTF, which is similar to the transport of 5 Sv that recirculates between 60° and 70°E (Fig. 1), corresponding to the southern limb of the southwest IO sub-gyre (see Fig. 7 of Stramma and Lutjeharms, 1997). As discussed in the previous section (see Fig. 3), cyclonic circulation associated with AgFNBF leads to an eastward transport of 8 Sv, compensated by a westward flow of 7 Sv at the northern edge of the AF. The merged front (SSTF+AF) contributes a large transport of 24 Sv, attributed to a high velocity of 50 cm s–1 Hydrodynamics along Meridional transects in the Southwest ... 93

Fig. 6: Baroclinic transport (Sv = 106 m3s–1) relative to the deepest common level between the XCTD stations across Track-2 for different depth layers: (a) 0–1000 m, (b) 0–100 m, (c) 100–500 m, (d) 500–1000 m, and (e) comparison of geostrophic transport (per kilometer distance between stations) computed using XCTD and altimeter- based absolute dynamic topography (ADT). Positive transport occurs eastwards. The locations of fronts identified in this work are shown in the top panel. See Figure 1 for abbreviations. 94 Alvarinho J. Luis averaged for the 0–250 m slab. The transport associated with SAF1+SAF2 (between 45° and 46°S) amounts to 14 Sv. The westward transport of 8.5 Sv associated with PF1 is nearly compensated by an eastward transport of 7 Sv just to its north. This pattern of opposing transport directions in close proximity indicates the existence of meridional meandering and mesoscale eddy activity, as observed previously along the 30°E section, between 51° and 59°S (PEP01). The transport associated with PF2 on Track-2 is 5 Sv, which is nearly half of its counterpart on Track-1. The altimeter-based full depth net transport amounts to 2.7 Svkm–1, nearly three times larger than that estimated from the XCTD data. DISCUSSION What is the role of bottom topography in meandering of oceanic fronts? Baroclinic flows in theoceans over smoothly varying topography tend to conserve angular momentum by following flow lines of constant potential vorticity represented by (f + î)/H, where f is planetary vorticity (s–1), î is relative vorticity (s–1), and H is ocean depth (m). In the open ocean, f is much larger than î,, hence the mean PPV can be approximated by f/H. In the SO, significant variations in the depth associated with mid-oceanic ridges and uneven bottom topography mean that the ACC jets are unable to follow the circumpolar lines of constant PPV (Koblinsky, 1990). Here we explain the dynamics that dictate the meandering of the fronts based on PPV computation. From west to east, the SSTF meanders from 42° to a mean position of 43.5°S, corresponding to the depth increase of 1000 m east of the Southwest Indian Ridge (Fig. 1). In this situation, f/H decreases eastward; consequently, in order to conserve PPV, which is about 20 × 10–9 s–1, the jets associated with SSTF are displaced poleward. In the case of PF2, depth decreases eastward from 4900 to 2300 m, thereby enhancing PPV and forcing PF2 to shift equatorward from 49.5° to 48.5°S. CONCLUSIONS Hydrodynamics were investigated along two near-meridional tracks in the southwestern Indian ocean sector of the SO occupied during austral summer of 2007 by using hydrographic data collected using XCTD. Employing various criteria from literature, fronts of AC and ACC were identified and their characteristics and locations were compared Hydrodynamics along Meridional Transects in the Southwest ... 95 with those proposed based on historical and synoptic sections. PF splits into PF2 and PF1 and SAF into SAF1 and SAF2, as described previously. Moving eastwards, PF1 is displaced northward by about ~110 km, while PF2 shifts southward by 550 km, SAF1 and SAF2 are identified adjacent to each other in the 44.5°–47°S belt. While AF and SSTF occur as separate fronts on Track-1, they occur as merged fronts on Track-2 and exhibit a southward shift of ~160 km when traced eastwards. Their merging is marked by a steep Ts gradient and enhanced surface velocity. Major water masses such as the STSW, CW and MW are encountered off the southern African coast to 43.5°S, whereas SASW, AASW, AAIW and CDW are encountered in the ACC domain. The net transport across the sections increases eastwards from 52 to 54 Sv, with 70–80% of this volume confined to the ACC belt. About 50% of the ACC transport occurs in the 100–500 m slab. Surface jets with speeds of 55 and 60 cms–1 are identified at the location of AF and AF+SSTF on Track-1 and -2, respectively. Because of insufficient spatial resolution between stations and limitation arising from the level-of-no motion being restricted to 1000 db, these estimates are indicative rather than definitive; however, the additional observations are planned in 2008 and 2009 which will lead to improved understanding of ACC frontal variability and the associated transport. ACKNOWLEDGEMENTS My gratitude to former NCAOR Director, Shri. Rasik Ravindra, for providing the infrastructural facilities for carrying out this study. Ministry of Earth Sciences, Government of India, is acknowledged for financing the XCTD probes. REFERENCES Andres, M., Park, J.-H., Wimbush, M., Zhu, X.-H., Chang K.-I., Ichikawa, H., 2008. Study of the Kuroshio/Ryukyu current system based on satellite-altimeter and in situ measurements, J. Oceanogr. 64, 937–950. Anilkumar, N., Luis, A.J., Somayajulu, Y.K., Babu, V.R., Dash, M.K., Pednekar, S.M., Babu, K.N., Sudhakar, M., Pandey, P.C., 2006. Fronts, water masses, and heat content variability in the western Indian sector of the Southern Ocean during austral summer 2004. J. Mar. Sys. 63, 20–34. Aoki, S., 1997. Trends and interannual variability of surface layer temperature in the Indian sector of the Southern Ocean observed by Japanese Antarctic 96 Alvarinho J. Luis

Research Expeditions. J. Oceanogr. 53, 623–631. Aoki S., Yoritaka, M., Masuyama, A., 2003. Multidecadal warming of subsurface temperature in the Indian sector of the Southern Ocean. J. Geophys. Res. 108, 8081, doi:10.1029/2000JC000307. Belkin, I.M., Gordon, A.L., 1996. Southern Ocean fronts from the Greenwich Meridian to Tasmania. J. Geophys. Res. 101, 3675–3696. Bryden, H.L., Beal, L.M., Duncan, L.M., 2005. Structure and transport of the Agulhas Current and its temporal variability. J. Oceanogr. 61 (3), 479–492. Bailey, R., Gronell, A., Phillips, H., Meyers, G., Tanner, E., 1993. CSIRO Cookbook for Quality Control of Expendable Bathythermograph (XBT) Data, CSIRO Marine Laboratories Report 221, CSIRO, Hobert. Deacon, G. E. R., 1937. The hydrography of the Southern Ocean. Discovery Reports, 15, 1–24. de Ruijter, W.P.M., Ridderinkhof, H., Lutjeharms, J.R.E., Schouten, M.W., Veth, C., 2002. Observations of the flow in the Mozambique Channel. Geophys. Res. Lett. 29 (10), 1502, doi:10.1029/2001GL013714. Ducet, N., Le Traon, P., Reverdin G., 2000. Global high-resolution mapping of ocean circulation from TOPEX/Poseidon and ERS-1 and –2, J. Geophys. Res. 105(C8), 19477–19498. Emery, W.J. and Meincke, J., 1986. Global water masses: summary and review. Oceanologica Acta 9, 383–391. Gill, A.E., 1977. Coastally trapped waves in the atmosphere. Quart. J. R. Meteorol. Soc. 103, 431–440. Gordon, A.I., 1967. Structure of Antarctic waters between 20°W and 170°W in: Bushnell, V. C. (Ed), Antarctic Map Folio Series, Folio 6, American Geographical Society, New York. Gordon, A.L., 1981. South Atlantic thermohaline ventilation. Deep-Sea Res. 28, 1239–1264. Gordon, A.L. 1985. Indian–Atlantic transfer of thermohaline water of the Agulhas retroflection. Science 227, 1030–1033. Gordon, A.L., Lutjeharms, J.R.E., Gründlingh, M.L., 1987. Stratification and circulation at the Agulhas Retroflection. Deep-Sea Res. I 34, 565–599. Harris, T.F.W 1972. Sources of the Agulhas Current in spring of 1964. Deep- Sea Res. 19(9), 633–650. Hellerman, S, Rosenstein, M, 1993. Normal monthly wind stress over the world ocean with error estimates, J. Phys. Oceanogr. 13, 1093–1104. Holliday, N.P., Reed, J.F., 1998. Surface oceanic fronts between Africa and Antarctica, Deep-Sea Res. I 45, 217–238. Hydrodynamics along Meridional Transects in the Southwest ... 97

Kizu, S., Onishi, H., Suga, T., Hanawa, K., Watanabe, T., Iwamiya, H., 2008. Evaluation of the fall rates of the present and developmental XCTDs, Deep-Sea Res. I 55, 571–586. Koblinsky, C.J., 1990. The global distribution of f/H and the barotropic response of the ocean. J. Geophys. Res. 95, 3213–3218. Kostianoy, A.G., Ginbug, A.I., Frankgnoulle, M., Delille, B., 2004. Fronts in the Southern Indian Ocean as inferred from satellite sea surface temperature data, J. Mar. Sys. 45, 55–73. Le Traon, P.-Y., Ogor, F., 1998. ERS-1/2 orbit improvement using TOPEX/ Poseidon: The 2 cm challenge. J. Geophys. Res. 103(C4), 8045–8057. Le Traon, P.-Y., Nadal, P.F., Ducet, N., 1998. An improved mapping method of multisatellite altimeter data, J. Atmos. Oceanic Technol. 15, 522–534. Lutjeharms, J.R.E., 1972. A quantative assessment of year-to-year variability in water movement in the southwest Indian Ocean. Nature 239, 59–60. Lutjeharms, J.R.E., Ansorge, I. 2001. The Agulhas Return Current, J. Mar. Sys. 30(1/2), 115–138. Lutjeharms, J.R.E., Valentine, H.R., 1984. Southern ocean thermal fronts south of Africa. Deep-Sea Res. I 31, 1461–1475. Lutjeharms J.R.E, Valentine, H.R., Van Ballegooyen, R.C., 1993. On the Subtropical Convergence in the south Atlantic Ocean, S. African J. Sci. 89, 552– 559. McCartney, M.S., 1977. Subantarctic Mode Water, in: Angel, M.V. (Ed), A Voyage of Discovery, Pergamon, New York, pp 103–119. Mizuno, K., Watanabe, T, 1998. Preliminary results of in situ XCTD/CTD comparison test. J. Oceanogr. 54, 373–380. Nowlin, W.D., Klinck, J.M., 1986. The physics of the Antarctic Circumpolar Current, Rev. Geophys. 24, 469–491. Orren, M.J., 1966. Hydrology of the South West Indian Ocean. Investigational Report of the Division for Sea Fisheries of S. Africa 55, 1–35. Orsi, A.H., Whitworth III, T., Nowlin Jr., W.D., 1995. On the meridional extent and fronts of the Antarctic Circumpolar Current, Deep-Sea Res. I 45, 641–673. Park, Y-H, Charriaud, E., Craneguy, P., 2001. Fronts, transport, and Weddell Gyre at 30°E between Africa and Antarctica, J. Geophys. Res. 106, 2857–2879. Park, Y-H, Charriaud, E., Fieux, M., 1998b. Thermocline structure of the Antarctic surface water/winter water in the Indian sector of the Southern Ocean, 98 Alvarinho J. Luis

J. Mar. Sys. 17, 5–23. Park, Y-H, Charriaud, E., Ruiz, D., Jeandel, C., 1998a. Seasonal and interannual variability of the mixed layer properties and steric height at station KERFIX, southwest of Kerguelen. J. Mar. Sys. 17, 571–586. Park, Y.-H., Gambéroni, L.,1995. Large-scale circulation and its variability in the south Indian Ocean from TOPEX/POSEIDON altimetry, J. Geophys. Res. 100, 24911–24929. Park, Y-H, Gamberoni, L, Charriaud, E., 1991. Frontal structure and transport of the Antarctic Circumpolar Current in the South Indian Ocean sector, 40-80°E, Mar. Chem. 35, 45–62. Park, Y-H, Gamberoni, L, Charriaud, E., 1993. Frontal structure, water masses, and circulation in the Crozet Basin. J. Geophys. Res. 98, 12361–12385. Pond, S., Pickard, G.L., 1983. Introductory Dynamical Oceanography. Second ed. Pergamon Press, Oxford. Roman, R.E., Lutjeharms, J.R.E., 2007. Red Sea intermediate water at the Agulhas Current termination. Deep-Sea Res. I 54(8), 1329–1340. Schmitz, W.J., Jr., 1996. On the World Ocean Circulation: Volume I, Technical Report WHOI-96-03, Woods Hole Oceanographic Institution, Woods Hole. Smith, W.H.F., Sandwell, D.T., 1997. Global seafloor topography from satellite altimetry and ship depth soundings, Science 277, 1957–1962. Sparrow, M.D., Heywood, K.J., Brown J., Stevens, D.P., 1996. Current structure of the south Indian Ocean, J. Geophys. Res. 101, 6377–6391. Stramma, L., 1992. The South Indian Ocean Current, J. Phys. Oceanogr. 22, 421–430. Stramma, L., Lutjeharms, J.R.E., 1997. The flow field of the subtropical gyre of the south Indian Ocean, J. Geophys. Res. 102, 5513–5530. Sverdrup, H.U., Johnson, M.W., Fleming, R.H., 1942. The oceans: Their physics, Chemistry, and General biology. Prentice-Hall, New Jersey. Swart, S, Speich, S, Ansorge, I.J., Goni, G.J., Gladyshev, S., Lutjeharms, J.R.E., 2008. Transport and variability of the Antarctic Circumpolar Current south of Africa, J. Geophys. Res. 113, C09014, doi:10.1029/2007JC004223. Valentine, H.R., Lutjeharms, J.R.E., Brundrit, G.B., 1993. The water masses and volumetry of the southern Agulhas Current region. Deep-Sea Res. I 40, 1285– 1305. Wang, X., and R. J. Matear (2001), Modeling the upper ocean dynamics in the Hydrodynamics along Meridional Transects in the Southwest ... 99

Subantarctic and Polar Frontal Zones in the Australian sector of the Southern Ocean, J. Geophys. Res., 106(C12), 31511–31524. Weeks, S., Shillington, F.A., 1994. Interannual scales of variation of pigment concentration from coastal zone color scanner data in the Benguela upwelling system and the Subtropical Convergence zone south of Africa, J. Geophys. Res. 99, 7385–7399. Whitworth III, T., Orsi, A.H., Kim, S.J., Nowlin Jr., W.D., Locarnini, R.A., 1998. Water Masses and mixing near the Antarctic Slope Front, in: Jacobs, S.S. and Weiss, R.F. (Eds.). Ocean, Ice and Atmosphere: Interactions at the Antarctic Continental Margin. Antarctic Research Ser. 75. AGU, Washington D.C., pp 1– 27. Wyrtki, K., 1971. Oceanographic Atlas of the International Indian Ocean Expedition. National Science Foundation, Washington, D.C. Xiaojun, Y., Douglas G.M., Zhaoqian, D., 2004. Upper ocean thermohaline structure and its temporal variability in the southeast Indian Ocean Deep-Sea Res. I 51(2), 333–347.