- Email: [email protected]
Summer sea surface temperature fronts and elevated chlorophyll-a in the entrance to Spencer Gulf, South Australia
Continental Shelf Research 31 (2011) 849–856
Contents lists available at ScienceDirect
Continental Shelf Research journal homepage: www.elsevier.com/locate/csr
Research papers
Summer sea surface temperature fronts and elevated chlorophyll-a in the entrance to Spencer Gulf, South Australia Peter Petrusevics a,1, John Bye b,c,n, John Luick c, Carlos E.P. Teixeira c,d a
School of Chemistry, Physics and Earth Sciences, Flinders University of South Australia, South Australia 5001, Australia School of Earth Sciences, University of Melbourne, Victoria 3010, Australia c South Australian Research and Development Institute, West Beach, South Australia 5024, Australia d School of Mathematics and Statistics, University of New South Wales, New South Wales 2052, Australia b
a r t i c l e i n f o
abstract
Article history: Received 1 June 2010 Received in revised form 9 November 2010 Accepted 13 February 2011 Available online 4 March 2011
Historical and recent oceanographic cruise data, MODIS chlorophyll-a satellite data, and an analytical model are used to examine SST fronts in the entrance to Spencer Gulf, South Australia. The fronts (2–3 1C) due to the contrast between warm Spencer Gulf waters and cooler waters of the continental shelf are readily observable on satellite imagery. Three water masses: cool, fresh upwelled shelf water; warm, salty Great Australian Bight water; and very warm and salty Spencer Gulf bottom water occupy the area. In consequence a summer density minimum is formed at the entrance to Spencer Gulf. The analytical model predicts that this thermohaline structure sets up an ageostrophic circulation, which favours upwelling in the central portion of the entrance. This is confirmed by the satellite data which show an increased chlorophyll-a concentration in the vicinity of the upwelling. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Spencer Gulf Sea surface temperature fronts Density minima Upwelling Chlorophyll-a MODIS satellite data
1. Introduction Spencer Gulf (SG), South Australia (Fig. 1), is classed as an inverse estuary (in which seawater is concentrated by the removal of freshwater). It penetrates into the arid zone of Australia to about 321 300 S with an average net evaporation rate of about 1.2 m/year (Nunez Vaz et al., 1990). Throughout the year there is little freshwater input from rivers or precipitation. In the summer at the head of SG the salinities exceed 48 (Nunes and Lennon, 1986) due to excessive evaporation. During austral winter a bottom-residing high density outflow occurs along the eastern side of SG (Lennon et al., 1986). A similar outflow of high density waters occurs in the Great Australian Bight (GAB) eastward of the head of the GAB (Petrusevics et al., 2009). In both cases, the discharge occurs as a result of their acting as inverse estuaries, which expel high density bottom waters formed by cooling of high salinity waters in the winter. Both discharges from the GAB and SG eventually cascade over the continental shelf south-west of Kangaroo Island.
n Corresponding author at: School of Earth Sciences, University of Melbourne, Victoria 3010, Australia. Tel.: þ 61 3 9354 1938. E-mail addresses: [email protected] (J. Bye), [email protected] (J. Luick), [email protected] (C.E. Teixeira). 1 Deceased.
0278-4343/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.csr.2011.02.009
During summer the high density discharges in both SG and the GAB cease, and sea surface temperature (SST) fronts occur at the entrance to SG (Nunez Vaz et al., 1990; Petrusevics, 1993), which is the focus of this paper. Such fronts exist in many shelf seas, (Simpson and Hunter, 1974; Simpson and Bowers, 1981; Yanagi and Yoshikawa, 1987), and they were first documented in SG in the mid-1960s in CSIRO oceanographic cruise reports (Anon, 1965, 1968a, b, 1969) of the tuna fishing grounds, and the Flinders University of South Australia cruises in the late 1980s (Petrusevics, 1993). The SST fronts form about December, intensify in March and collapse in April, and are the result of warmer surface waters of the relatively shallow SG abutting cooler surface shelf waters at the entrance to SG. These surface features have been observed on satellite imagery since the mid-1980s using NOAA Advanced Very High Resolution Radiometer imagery (Petrusevics, 1993) and more recently using Moderate Resolution Imaging Spectradiometer (MODIS) satellite data (Petrusevics and Bye, 2008) and are described using climatological data in Middleton and Bye (2007). Temperature fronts are often reported to be associated with increased biological activity represented by increased chlorophyll-a levels at the surface, for example in the Celtic and Irish Seas (Savidge and Foster, 1978). Worm et al. (2005) examined worldwide patterns of tuna and billfish diversity over the past 50 years and concluded that diversity was positively correlated with
850
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
Fig. 1. Study region showing current meter locations during MV Boobook cruise March 1988. NS and NB north surface and bottom moorings, and SS and SB south surface and bottom mooring.
thermal fronts and dissolved oxygen, which appeared to hold generally for other predators and zooplankton. In the local context Bruce and Short (1990) reported a discontinuity in larval distribution and an increase in larval diversity within the SG temperature frontal zone, and since 2002 MODIS satellite observations have revealed co-located seasonally persistent regions of elevated chlorophyll-a levels. This paper examines in detail the interaction between the SST fronts and increased biological activity represented by increasing seasonal concentrations of chlorophyll-a levels in SG. In Section 2, the origin of the density minimum in the entrance to SG in the context of the regional oceanography is discussed, and Section 3 presents the results from two historic summer cruises (1988 and 1989), which describe the fine structure of the density minimum, and have not been previously reported. In Section 4, MODIS satellite observations of the monthly averaged SST and chlorophyll-a structure over an annual cycle (July 2002–June 2003) along a longitudinal section through the entrance to SG are presented, which show the seasonal onset, persistence and decay of chlorophyll-a in the frontal zone, and in Section 5 the results of a recent summer cruise (2009) in the entrance to SG are discussed, in which for the first time, coincident MODIS data, and also shipborne observations of fluorescence were available. Finally in Section 6, a simple analytical model of the density minimum in the entrance to SG is derived, which predicts the uplift of nutrient rich waters to the surface that give rise to the elevated concentrations of chlorophyll-a, and Section 7 is a brief conclusion.
2. The regional oceanographic origin of the summer density minimum in the entrance to Spencer Gulf The SST fronts are formed by the warmer surface waters of the relatively shallow SG abutting cooler surface shelf waters at the entrance to SG. The regional and seasonal picture is shown in the CARS surface climatology (Ridgway et al., 2002) for the austral summer (Middleton and Bye, 2007, Fig. 20), which shows that in the western and central GAB, warm high salinity water occurs inshore along the southern coast of Australia, with cold low
salinity water in the Southern Ocean, which mix together laterally to form a region of lower density water. The density minimum appears to be due to the summer heating, which reduces the surface density of the oceanic water as it approaches the coast, until it is sufficiently inshore for the increase in salinity to become dominant, and cause the surface density to rise. In the eastern GAB the density minimum extends south-eastward in an arm of the anti-cyclonic gyre centred on the wide continental shelf (Herzfeld and Tomczak, 1999). This warm saline stream, which flows between the cooler upwelled water along the west coast of Eyre Peninsula and the cooler Southern Ocean water to the south, has been called the Great Australian Bight Plume (GABP) (Herzfeld and Tomczak, 1999; Richardson et al., 2009). To the south of SG the progress of the GABP is interrupted by a corridor of higher density water, which is due to an onshore arm of the mesoscale structure of the Flinders Current (Bye, 1983, Middleton and Bye, 2007), brought about by the coastal wind structure, as is evident in the fields at 50 and 40 m (Fig. 2), obtained by CSIRO during February 1965 (Anon, 1965). At 50 m depth (Fig. 2a–c) the GAB water extends as a warm saline low density intrusion at the surface south of Eyre Peninsula. This intrusion is also evident in the CARS climatological atlas. The temperature (Fig. 2b) and salinity (Fig. 2c) charts show that the intrusion meets a region of cooler fresher shelf water to the west of Kangaroo Island, which is due to the Kangaroo Island pool of water upwelled from over the shelf break (Middleton and Bye, 2007). At the 40 m level, the density minimum propagates into the entrance to SG (Fig. 2d), with warmer saltier waters on the eastern side and cooler fresher waters on the western side (Fig. 2e–f), until it meets a density front formed by the saline high temperature gulf waters. These three interacting water masses (shelf water, gulf water, and GAB water) were first identified by Hahn (1986). The TS properties described above are also discussed in Petrusevics (1993) in which the three water masses are identified in a TS diagram.
3. The structure of the density minimum in entrance to Spencer Gulf 3.1. The MV Boobook survey (summer 1988) The first, and most detailed, survey of the structure of the density minimum in the entrance to SG was undertaken during March 1988. The purpose of the investigation was to relate SST fronts, readily observed on satellite imagery (Petrusevics, 1993) to underlying water properties. The commercial vessel, MV Boobook, was used to conduct the survey (Fig. 3) in water depths ranging from about 60 m on the shelf to about 40 m in SG (Boobook 03/88). A Seabird CTD logger with 0.5 s sampling rate and a descent rate of 1 m s 1 was used. SST was measured with a standard mercury-in-glass thermometer and water samples (top and bottom) were analysed with an autolab inductive salinometer. A horizontal section at 10 m depth shows a region of lower density (Fig. 4a) in the centre of the entrance to SG and almost zonal temperature and salinity fronts (Fig. 4b and c) at about 341550 S. A similar structure was observed at 40 m. The wind rose for the summer season (December 1987– February 1988) at Neptune Island (351200 S, 136170 E) which lies about 40 km to the south-west of Wedge Island (Fig. 1), indicated that during the study period, the winds were mainly south southeasterly with speeds of 5–10 m s 1, indicating that the components of mean wind stress were similar to the climatological mean values of 0.029 and 0.058 N m 2.
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
851
Fig. 2. Cruise data from Anon (1965). (a) st (kg m 3) distribution at 50 m. (b) Temperature (1C) distribution at 50 m. (c) Salinity distribution at 50 m. (d) st (kg m 3) distribution at 40 m. (e) Temperature (1C) distribution at 40 m. (f) Salinity distribution at 40 m.
The horizontal sections (Fig. 4a–c) show that the thermohaline structure is three dimensional, and indeed except in the far southwest it has an almost circular pattern, which occupies the whole of the entrance to SG. The 1988 campaign also deployed four current meter moorings (Fig. 1) spanning February–March 1988, the period of the MV Boobook cruise (Table 1). Due to strong tidal currents, the mean vector velocities are all much less than the root mean square velocities (Table 1). Over the two months, however, the vector mean current data were consistent with the existence of a gyral circulation, anti-cyclonic near the surface and cyclonic near the bottom, and on which was superimposed a westward or northwestward surface wind-forced circulation that was the strongest at the southern mooring (Table 1). We return to the gyral circulation in Section 6. The tidal currents (Table 2) are almost barotropic with little change in phase from surface to bottom, and the phases are highly coherent across all four moorings, with a reduction in velocity at the bottom due to bottom friction, The semi-diurnal and diurnal
constituents are of similar magnitude denoting a mixed tide, with their major axes directed approximately along the axis of the gulf. The tidal currents at the north moorings rotate clockwise at the surface and counter clockwise at the bottom. Finally, the diurnal ellipses (in particular K1) are more circular than the semi-diurnal ellipses.
3.2. The RV Franklin survey (summer 1989) During summer 1989, the School of Earth Sciences of Flinders University conducted a survey, Fr 04/89, which included a traverse of the entrance to SG using RV Franklin. The transect ran parallel to and about 15 km to the west of the MV Boobook transect (Fig. 3). The Franklin survey data was consistent with the earlier Boobook results. Fig. 5 shows that surface density minima occur at station 168 and between stations 170 and 171, the density minimum at the latter location being more pronounced and
852
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
The subsequent evolution, which maintains chlorophyll-a levels greater in the gulf than on the shelf throughout the winter, is significant to the ecology within the wider SG. The decrease in spring may be due to utilisation in the surface layers or alternatively to the sinking of chlorophyll-a into the dense saline outflow beneath, but this remains to be investigated
5. The RV Ngerin survey (summer 2009)
Fig. 3. Portions of MV Boobook 03/88, RV Franklin Fr 04/89 and RV Ngerin cruise tracks and MODIS transect (July 2002–June 2003).
evident at all depths. st ranges between 26.40 kg m 3 near the bottom and 26.05 kg m 3 near the surface, and at the former location (station 168) the density minimum is limited to a depth of about 25 m. The position of the major minimum (between stations 170 and 171) occurs about 25 km south of that reported for the MV Boobook cruise in March 1988.
4. MODIS satellite observations (2002–2003) MODIS Aqua chlorophyll-a and SST data for the period July 2002–June 2003 were processed using SeaWIFS Data Analysis System (SEADAS for processing and display of ocean colour data) software. A transect along 1361200 E (Fig. 1) of length 50 km with 25 sampling points at approximately 2 km spacing was used to obtain monthly mean values between 351470 S (inside the entrance to SG, water depth 40 m) and 341240 S (on the shelf, water depth 75 m). It is convenient to present the results using ¨ Hovmoller diagrams spanning the annual cycle, which show the progression of the signals in a space–time graph. The SST gradient (Fig. 6a) undergoes a seasonal cycle. Gulf water was cooler than shelf waters in winter–spring (June– October), followed by a reversal in November after which the gulf waters were warmer than shelf waters during December– April. The SST fronts were the strongest in February–March with SST contrasts of 2–3 1C centred on about 341550 S. These features are a regular aspect of the SST dynamics in the entrance to SG (Petrusevics, 1993). The chlorophyll-a concentrations (Fig. 6b) were the lowest ( E0.2 mg m 3) in spring, following which in December a maximum in concentration formed near 351100 S (which is 150 south of the centre of the SST front) intensified, reaching a maximum of about 0.8 mg m 3 in March, and moved northwards during January–April. This distinctive feature observed in 2003, which was called a ‘chlorophyll-a mound’ in Petrusevics and Bye (2008), suggests that 351100 S is a summer source region for chlorophyll-a which subsequently migrates northwards into SG by about 150 of latitude in two months (mean speed, 5 mm s 1).
Recently additional surveys in the entrance were undertaken as part of the South Australian Integrated Marine Observing System (SAIMOS, 2010) using RV Ngerin. SAIMOS C2009_03 included a transect of seven stations adjacent to the MV Boobook survey of March 1988 (Fig. 3). The latter survey used a SBE 19 þ data logger with conductivity, temperature, pressure, oxygen, and fluorescence sensors. The SAIMOS data confirm the presence of a summer density minimum. The longitudinal section of st between the shelf station JL10 and gulf station JL03 (Fig. 7a) indicates a minimum between stations JL04 and JL07 with a near bottom value of about 26.25 kg m 3 and between stations JL04 and JL05 a minimum of 26.20 kg m 3. The separation of the density minimum positions is about 20 km, the minimum at the former location being more pronounced and evident at all depths. On either sides of the extended minimum the density at the bottom increased, with st rising to 26.30 kg m 3. The extended minimum is centred about 15 km south of that reported in March 1988 (Boobook 03/88). MODIS satellite data captured coincident with the RV Ngerin cruise on 10 March 2009 comprised both MODIS-Terra (southward morning overpass) and MODIS-Aqua (northward afternoon overpass). On both occasions a distinct maximum in the chlorophyll-a level may be observed co-located with the position of the density minimum regions. However an interesting aspect of this is the reduction of chlorophyll-a levels between the morning and afternoon overpasses. The values vary from about 1.6 mg m 3 at station JL07 for the morning overpass, to about 1.0 mg m 3 for the afternoon overpass. A similar reduction is noted for stations JL04-05, 1.1 mg m 3 (morning) and 0.8 mg m 3 (afternoon). This may be due to uptake of the chlorophyll-a as a result of diurnal increase in the photo-period and water temperature. The corresponding fluorescence section (Fig. 7b) shows an extended maximum of about 0.45 mg m 3 coincident with the density minimum and extending throughout much of the water column. These direct observations of the geo-positioning of the longitudinal density and chlorophyll-a variation within the water column in the entrance to SG indicates the high spatial coincidence between these parameters, and together with previous time series of CTD (MV Boobook March 1988 and RV Franklin March 1989 and MODIS satellite chlorophyll-a levels since 2002) do indeed form a sound basis for the association of seasonally occurring physical properties and the upwelling of nutrient rich waters from the shelf.
6. The upwelling model The important question to be addressed is whether the presence of a seasonal density minimum centred in the entrance to SG will promote the uplift responsible for the accumulation of chlorophyll-a, which is essential for the wellbeing of the aquaculture industry of the region (Bierman et al., 2008). The main feature of the density structure (Fig. 4d) is the density front, which impinges on the central region of lower density. The current measurements to the north of this region suggest that
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
853
Fig. 4. Cruise data from MV Boobook (Boobook 03/88) survey, showing cruise track. (a) st (kg m 3) distribution at 10 m. (b) Temperature (oC) distribution at 10 m. (c) Salinity distribution at 10 m. (d) st (kg m 3) vertical section shown in Fig. 4a. (e) Temperature (oC) vertical section shown in Fig. 4a. (f) Salinity vertical section shown in Fig. 4a.
the velocity profile consists of an anti-cyclonic surface flow which is superimposed on a deep cyclonic flow, rather in the nature of a tropical cyclone in the atmosphere. In this system, it would be expected that upward vertical motion occurs in the transition region from inner low density to outer high density. This scenario can be modelled in a simple way by considering a segment along ox, assuming no net zonal transport, and that the density anomaly is independent of depth, then it is easily shown (Bullock, 1975) that u ¼ g=ðf r0 Þ@sT =@yðz þ H=2Þ
ð1Þ
where ox, oy is any right handed horizontal co-ordinate system, and oz is vertically upwards, u is the geostrophic current along ox, g is the acceleration of gravity, ro is a standard density for seawater, z ¼0 is the mean sea level, H is the water depth and f¼2O sin j is the Coriolis parameter in which O is the angular speed of rotation of the Earth, and j is latitude. In terms of the surface current (u0), (1) yields, u¼u0(2zþH)/H. The frictional circulation associated with this geostrophic flow is v ¼ Au in which A 40, is the tangent of the angle of turning to the right hand side. Hence from the continuity equation, assuming that qu/qx ¼0, qw/qz ¼Aqu/qy, where w is the vertical velocity,
854
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
Table 1 Current measurements in the entrance to Spencer Gulf (2 February–6 May 1988). Position and period
Water depth (m)
Current meter depth (m)
a (deg.)
oV4 (cm s 1)
Vs (cm s 1)
North surface (341490 S, 1361330 E) February 2–February 29 February 2–March 31 February 2–May 6
44 44 44
5 5 5
67 70 70
10.30 8.54 6.00
25.01
North bottom (341570 S, 1361310 E) February 2–February 29 February 2–March 31 February 2–May 6
44 44 44
39 39 39
122 114 124
1.36 1.28 1.11
13.83
South surface (351170 S, 1361230 E) February 2–February 29 February 2–March 31 February 2–May 6
79 79 79
5 5 5
97 97 105
10.10 8.26 5.39
20.16
79 79 79
74 74 74
48 42 36
0.74 0.85 0.42
7.15
0
0
South bottom (35117 S, 136123 E) February 2–February 29 February 2–March 31 February 2–May 6
o V 4: vector mean current velocity, a: direction (relative to north), and Vs: root mean square current speed.
Table 2 Tidal current ellipses. Major axis (cm s 1)
Minor axis (cm s 1)
Inclination (deg.)
Phase (g) (deg.)
North surface (67%) O1 8.26 K1 17.32 M2 12.22 S2 12.66
1.38 6.37 1.02 0.60
64 73 53 57
279 317 353 47
North bottom (80%) 4.43 O1 K1 7.72 M2 8.28 S2 9.14
-1.52 -3.45 -0.08 -0.61
50 30 67 66
298 355 354 53
South surface (47%) 4.86 O1 K1 12.59 M2 6.41 S2 7.63
0.57 6.03 0.18 -0.94
64 77 50 48
288 325 351 52
South bottom (61%) 1.61 O1 K1 2.93 M2 4.05 S2 4.13
0.12 -0.29 -0.02 -0.13
19 17 37 36
301 350 355 58
Fig. 5. st (kg m 3) vertical section from the RV Franklin (Fr 04/89) survey.
in which a ¼ sT(0), and for c40, at and beyond the rim (y¼R) of the low density core, st ¼ st(R). On substituting (3) in (1), we obtain, c¼ Dst/R2 in which Dst ¼ st(R)–st(0), and hence uRIM ¼ Dst =R gH=ðf r0 Þ where uRIM ¼u0 (R) is the surface rim velocity, and
Inclination is the angle of the major axis to North. Positive and negative minor axes denote, respectively, clockwise and anticlockwise rotations of the tidal currents in the ellipse. [xx%]: percentage of the variance predicted by the tidal analysis. Note that the north surface and bottom data are from different moorings, see Table 1.
and on integrating with respect z using (1), and assuming a flat bottom we obtain w ¼ A zðz þ HÞ=H @u0 =@y
ð2Þ
which has a maximum, wmax ¼ AH/4qu0/qy at mid-depth (z¼ ½H). Thus, for an anti-cyclonic surface flow at the rim of the low density core (qu0/qyo0 along ox), v is radially inward at the bottom and outward at the surface, and w is upwards. On applying this model to the density minimum in the entrance to SG assuming a quadratic variation of st outwards from the central co-ordinate (which is representative of the st distribution shown in Figs. 4d, 5, and 7a), we have 2
sT ðyÞ ¼ a þcy y o R sT ðyÞ ¼ sT ðRÞ y4 R
ð4Þ
ð3Þ
wmax ¼ A Dst =R2 gH2 =ðf r0 Þ
ð5Þ
which is a constant throughout the low density core. On evaluating (4) and (5) using r0 ¼103 kg m 3, g ¼10 m s 2, f¼ 10 4 s 1 and from the distributions of st (Figs. 4d, 5, and 7a), Dst ¼ 0.3 kg m 3, H¼50 m, and R¼30 km, we obtain, uRIM ¼ 5 cm s 1 which indicates a vertical baroclinic shear of 10 cm s 1 at the rim in reasonable agreement with observations (Table 1), and for A¼0.1, which corresponds with an angle of turning of 61, wmax ¼ 2 10 5 m s 1. This analysis suggests that the density minimum is a very efficient mechanism for uplifting nutrient rich water in the entrance to SG. Note that for a sustained upward velocity of 2 10 5 m s 1, a particle would rise from the bottom to the top of the water column (50 m) in about 1 month. 7. Conclusions The mechanism for the key formation of chlorophyll-a maxima has been investigated in this paper using a variety of cruise, current
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
855
¨ Fig. 6. Hovmoller diagrams of the monthly sequence of (a) SST (oC) and (b) chlorophyll-a (mg m 3) data from the MODIS transect shown in Fig. 1.
meter, and satellite data. The formation of the summer density minimum in the entrance to Spencer Gulf by the interaction of gulf water, shelf water, and Great Australian Bight water creates a novel mechanism for the upwelling of nutrient-rich water. This mechanism which depends on the GAB water to form the density minimum through an intrusion into the centre of the entrance to SG differs from the simpler model of Nunez Vaz et al. (1990) in which only the interaction shelf and gulf waters was considered. It appears that the dynamics are robust, as the density minimum has been observed in
several years; the earliest depiction (Bullock, 1975) being from a cruise in March 1963. The chlorophyll-a in lower SG increases through summer to an autumn maximum and then decreases to the spring minimum (Fig. 6b). This is consistent with field observations reported by Van Ruth et al. (2009) in the lower section of SG in which the highest gross phytoplankton growth rates occurred in March and May 2007 with rates 30% lower in July and October. The analytical upwelling model suggests the need for a linked biochemical-hydrodynamic numerical model of the region, which
856
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
Centre, Greenbelt MD 20771 are also gratefully acknowledged for the production and distribution of these data. Carlos Teixeira was supported by the Brazilian Research Council (CAPES 4012-05-4).
References
Fig. 7. Cruise data from RV Ngerin (SAIMOS C2009_03) survey. (a) st (kg m 3) vertical section. (b) chlorophyll-a (mg m 3) vertical section.
encompasses accumulation, utilisation, and dispersion of nitrogen and phosphates. This is of vital importance for improved understanding of marine productivity in Spencer Gulf in support of aquaculture and fisheries.
Acknowledgements The authors are grateful to the crew of the MV Boobook and colleagues from the School of Earth Sciences, the Flinders University of South Australia (especially Professor Geoff. Lennon and Dr Rick Nunes Vaz) and also the CSIRO Marine data section in Hobart for the supply of the 1989 RV Franklin cruise data, and to Dr. John Middleton of SARDI and the crew of MV Ngerin for the 2009 cruise data. The SeaWIFS and MODIS Project and the Distributed Active Archive Centre at the Goddard Space Flight
Anon, 1965. H.M.A.S. Gascoyne Cruise Anon (1965) Oceanogr. Cruise Rep. 43 Oceanographical Observations in the Indian Ocean in G2/65. Anon, 1968a. Investigations by FRV Investigator on the South Australian Tuna Grounds in 1961 CSIRO Oceanographical Station List Cruise No. 64. Anon, 1968b. Investigations by FRV Investigator on the South Australian Tuna Grounds in 1964 CSIRO Oceanographical Station List Cruise No. 67. Anon, 1969. Oceanographical observations in the Indian Ocean in 1966. HMAS Diamantina Cruise Dm 2/66. CSIRO Oceanographical Cruise Report No. 54. Bierman, P., Lewis, M., Tanner, J., Ostendorf, B., 2008. Understanding phytoplankton variability throughout Spencer Gulf, South Australia via satellite derived chlorophyll-a. In: IEEE International Proceedings of Geoscience and Remote Sensing Symposium IGARSS 2008, 4, Boston, MA, pp. 902–905. Bruce, B.D, Short, D.A., 1990. Observations on the distribution of larval fish in relation to a frontal zone at the entrance to Spencer Gulf, South Australia. In: Bureau of Rural Resources Proceedings No. 15. Bullock, D.A., 1975. The general water circulation of Spencer Gulf, South Australia, in the period February–May. Trans. R. Soc. S. Aust. 99, 43–53. Bye, J.A.T., 1983. The general circulation in a dissipative basin with longshore wind stresses. J. Phys. Oceanogr. 13, 1553–1563. Hahn, S.D., 1986. Physical structure of the waters of the South Australian continental shelf. Ph.D. Thesis, School of Earth Sciences, The Flinders University of South Australia, 285pp. Herzfeld, M., Tomczak, M., 1999. Bottom-driven upwelling generated by eastern intensification in closed and semi-enclosed basins. Aust. J. Mar. Freshwater Res. 50, 613–628. Middleton, J.F., Bye, J.A.T., 2007. A review of the shelf-slope circulation along Australia’s southern shelves: Cape Leeuwin to Portland. Progr. Oceanogr. 75, 1–41. Nunes, R.A., Lennon, G.W., 1986. Physical property distributions and seasonal trends in Spencer Gulf, South Australian inverse estuary. Aust. J. Mar. Freshwater Res. 37, 39–53. Nunez Vaz, R.A., Lennon, G.W., Bowers, D.G., 1990. Physical behaviour of a large, negative or inverse estuary. Cont. Shelf Res. 10, 277–304. Petrusevics, P.M., 1993. SST fronts in inverse estuaries, South Australia—indicators of reduced gulf-shelf exchange. Aust. J. Mar. Freshwater Res. 44, 305–323. Petrusevics, P, Bye, J, 2008. Satellite observations of SST and chlorophyll-a time series at the gateway to the inverse estuary Spencer Gulf. In: Proceedings of the Australian Marine Science Association (SA Branch), October 2008. Petrusevics, P., Bye, J.A.T., Fahlbusch, V., Hammat, J., Tippins, D.R., van Wijk, E., 2009. High salinity winter outflow form a mega-inverse-estuary, the Great Australian Bight. Cont. Shelf Res. 29, 371–380. Ridgway, K.R., Dunn, J.R., Wilkin, J.L., 2002. Ocean interpolation by four-dimensional least squares—application to the waters around Australia. J. Atmos. Ocean. Tech. 19, 1357–1375. Richardson, L.E., Kyser, T.K., James, N.P., Bone, Y., 2009. Analysis of hydrographic and stable isotope data to determine water masses, circulation, and mixing in the eastern Great Australian Bight. J. Geophys. Res. 114, C10016. doi:10.1029/ 2009JC005407. SAIMOS, 2010. Cruise files. South Australian Integrated Marine Observing System. SARDI Aquatic Sciences West Beach South Australia 5022, Australia. Savidge, G., Foster, P., 1978. Phytoplankton biology of a thermal front in the Celtic Sea. Nature 271, 155. Simpson, J.H., Hunter, J.R., 1974. Fronts in the Irish Sea. Nature 250, 404–406. Simpson, J.H., Bowers, D.G., 1981. Models of stratification and frontal movement in shelf seas. Deep-Sea Res. A28, 727–738. Van Ruth, P., Thompson, P., Bonham, P., Jones, E., 2009. Primary production and zooplankton ecology in Port Lincoln Tuna Farming Zone. Aquafin CRC Project 4.6 March 2009. FRDC Project No. 2005/059. Worm, B., Sandow, M., Oschlies, A., Lotze, H.K., Myers, R.A., 2005. Global patterns of predator diversity in the open oceans. Science 309, 1365–1369. Yanagi, T., Yoshikawa, K., 1987. Tidal fronts in Hiuchi-nada and Osaka Bay. Bull. Jpn. Soc. Fish. Oceanogr. (Sisan Kaiyo Kenkyukaiho) 51, 115–119.
Contents lists available at ScienceDirect
Continental Shelf Research journal homepage: www.elsevier.com/locate/csr
Research papers
Summer sea surface temperature fronts and elevated chlorophyll-a in the entrance to Spencer Gulf, South Australia Peter Petrusevics a,1, John Bye b,c,n, John Luick c, Carlos E.P. Teixeira c,d a
School of Chemistry, Physics and Earth Sciences, Flinders University of South Australia, South Australia 5001, Australia School of Earth Sciences, University of Melbourne, Victoria 3010, Australia c South Australian Research and Development Institute, West Beach, South Australia 5024, Australia d School of Mathematics and Statistics, University of New South Wales, New South Wales 2052, Australia b
a r t i c l e i n f o
abstract
Article history: Received 1 June 2010 Received in revised form 9 November 2010 Accepted 13 February 2011 Available online 4 March 2011
Historical and recent oceanographic cruise data, MODIS chlorophyll-a satellite data, and an analytical model are used to examine SST fronts in the entrance to Spencer Gulf, South Australia. The fronts (2–3 1C) due to the contrast between warm Spencer Gulf waters and cooler waters of the continental shelf are readily observable on satellite imagery. Three water masses: cool, fresh upwelled shelf water; warm, salty Great Australian Bight water; and very warm and salty Spencer Gulf bottom water occupy the area. In consequence a summer density minimum is formed at the entrance to Spencer Gulf. The analytical model predicts that this thermohaline structure sets up an ageostrophic circulation, which favours upwelling in the central portion of the entrance. This is confirmed by the satellite data which show an increased chlorophyll-a concentration in the vicinity of the upwelling. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Spencer Gulf Sea surface temperature fronts Density minima Upwelling Chlorophyll-a MODIS satellite data
1. Introduction Spencer Gulf (SG), South Australia (Fig. 1), is classed as an inverse estuary (in which seawater is concentrated by the removal of freshwater). It penetrates into the arid zone of Australia to about 321 300 S with an average net evaporation rate of about 1.2 m/year (Nunez Vaz et al., 1990). Throughout the year there is little freshwater input from rivers or precipitation. In the summer at the head of SG the salinities exceed 48 (Nunes and Lennon, 1986) due to excessive evaporation. During austral winter a bottom-residing high density outflow occurs along the eastern side of SG (Lennon et al., 1986). A similar outflow of high density waters occurs in the Great Australian Bight (GAB) eastward of the head of the GAB (Petrusevics et al., 2009). In both cases, the discharge occurs as a result of their acting as inverse estuaries, which expel high density bottom waters formed by cooling of high salinity waters in the winter. Both discharges from the GAB and SG eventually cascade over the continental shelf south-west of Kangaroo Island.
n Corresponding author at: School of Earth Sciences, University of Melbourne, Victoria 3010, Australia. Tel.: þ 61 3 9354 1938. E-mail addresses: [email protected] (J. Bye), [email protected] (J. Luick), [email protected] (C.E. Teixeira). 1 Deceased.
0278-4343/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.csr.2011.02.009
During summer the high density discharges in both SG and the GAB cease, and sea surface temperature (SST) fronts occur at the entrance to SG (Nunez Vaz et al., 1990; Petrusevics, 1993), which is the focus of this paper. Such fronts exist in many shelf seas, (Simpson and Hunter, 1974; Simpson and Bowers, 1981; Yanagi and Yoshikawa, 1987), and they were first documented in SG in the mid-1960s in CSIRO oceanographic cruise reports (Anon, 1965, 1968a, b, 1969) of the tuna fishing grounds, and the Flinders University of South Australia cruises in the late 1980s (Petrusevics, 1993). The SST fronts form about December, intensify in March and collapse in April, and are the result of warmer surface waters of the relatively shallow SG abutting cooler surface shelf waters at the entrance to SG. These surface features have been observed on satellite imagery since the mid-1980s using NOAA Advanced Very High Resolution Radiometer imagery (Petrusevics, 1993) and more recently using Moderate Resolution Imaging Spectradiometer (MODIS) satellite data (Petrusevics and Bye, 2008) and are described using climatological data in Middleton and Bye (2007). Temperature fronts are often reported to be associated with increased biological activity represented by increased chlorophyll-a levels at the surface, for example in the Celtic and Irish Seas (Savidge and Foster, 1978). Worm et al. (2005) examined worldwide patterns of tuna and billfish diversity over the past 50 years and concluded that diversity was positively correlated with
850
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
Fig. 1. Study region showing current meter locations during MV Boobook cruise March 1988. NS and NB north surface and bottom moorings, and SS and SB south surface and bottom mooring.
thermal fronts and dissolved oxygen, which appeared to hold generally for other predators and zooplankton. In the local context Bruce and Short (1990) reported a discontinuity in larval distribution and an increase in larval diversity within the SG temperature frontal zone, and since 2002 MODIS satellite observations have revealed co-located seasonally persistent regions of elevated chlorophyll-a levels. This paper examines in detail the interaction between the SST fronts and increased biological activity represented by increasing seasonal concentrations of chlorophyll-a levels in SG. In Section 2, the origin of the density minimum in the entrance to SG in the context of the regional oceanography is discussed, and Section 3 presents the results from two historic summer cruises (1988 and 1989), which describe the fine structure of the density minimum, and have not been previously reported. In Section 4, MODIS satellite observations of the monthly averaged SST and chlorophyll-a structure over an annual cycle (July 2002–June 2003) along a longitudinal section through the entrance to SG are presented, which show the seasonal onset, persistence and decay of chlorophyll-a in the frontal zone, and in Section 5 the results of a recent summer cruise (2009) in the entrance to SG are discussed, in which for the first time, coincident MODIS data, and also shipborne observations of fluorescence were available. Finally in Section 6, a simple analytical model of the density minimum in the entrance to SG is derived, which predicts the uplift of nutrient rich waters to the surface that give rise to the elevated concentrations of chlorophyll-a, and Section 7 is a brief conclusion.
2. The regional oceanographic origin of the summer density minimum in the entrance to Spencer Gulf The SST fronts are formed by the warmer surface waters of the relatively shallow SG abutting cooler surface shelf waters at the entrance to SG. The regional and seasonal picture is shown in the CARS surface climatology (Ridgway et al., 2002) for the austral summer (Middleton and Bye, 2007, Fig. 20), which shows that in the western and central GAB, warm high salinity water occurs inshore along the southern coast of Australia, with cold low
salinity water in the Southern Ocean, which mix together laterally to form a region of lower density water. The density minimum appears to be due to the summer heating, which reduces the surface density of the oceanic water as it approaches the coast, until it is sufficiently inshore for the increase in salinity to become dominant, and cause the surface density to rise. In the eastern GAB the density minimum extends south-eastward in an arm of the anti-cyclonic gyre centred on the wide continental shelf (Herzfeld and Tomczak, 1999). This warm saline stream, which flows between the cooler upwelled water along the west coast of Eyre Peninsula and the cooler Southern Ocean water to the south, has been called the Great Australian Bight Plume (GABP) (Herzfeld and Tomczak, 1999; Richardson et al., 2009). To the south of SG the progress of the GABP is interrupted by a corridor of higher density water, which is due to an onshore arm of the mesoscale structure of the Flinders Current (Bye, 1983, Middleton and Bye, 2007), brought about by the coastal wind structure, as is evident in the fields at 50 and 40 m (Fig. 2), obtained by CSIRO during February 1965 (Anon, 1965). At 50 m depth (Fig. 2a–c) the GAB water extends as a warm saline low density intrusion at the surface south of Eyre Peninsula. This intrusion is also evident in the CARS climatological atlas. The temperature (Fig. 2b) and salinity (Fig. 2c) charts show that the intrusion meets a region of cooler fresher shelf water to the west of Kangaroo Island, which is due to the Kangaroo Island pool of water upwelled from over the shelf break (Middleton and Bye, 2007). At the 40 m level, the density minimum propagates into the entrance to SG (Fig. 2d), with warmer saltier waters on the eastern side and cooler fresher waters on the western side (Fig. 2e–f), until it meets a density front formed by the saline high temperature gulf waters. These three interacting water masses (shelf water, gulf water, and GAB water) were first identified by Hahn (1986). The TS properties described above are also discussed in Petrusevics (1993) in which the three water masses are identified in a TS diagram.
3. The structure of the density minimum in entrance to Spencer Gulf 3.1. The MV Boobook survey (summer 1988) The first, and most detailed, survey of the structure of the density minimum in the entrance to SG was undertaken during March 1988. The purpose of the investigation was to relate SST fronts, readily observed on satellite imagery (Petrusevics, 1993) to underlying water properties. The commercial vessel, MV Boobook, was used to conduct the survey (Fig. 3) in water depths ranging from about 60 m on the shelf to about 40 m in SG (Boobook 03/88). A Seabird CTD logger with 0.5 s sampling rate and a descent rate of 1 m s 1 was used. SST was measured with a standard mercury-in-glass thermometer and water samples (top and bottom) were analysed with an autolab inductive salinometer. A horizontal section at 10 m depth shows a region of lower density (Fig. 4a) in the centre of the entrance to SG and almost zonal temperature and salinity fronts (Fig. 4b and c) at about 341550 S. A similar structure was observed at 40 m. The wind rose for the summer season (December 1987– February 1988) at Neptune Island (351200 S, 136170 E) which lies about 40 km to the south-west of Wedge Island (Fig. 1), indicated that during the study period, the winds were mainly south southeasterly with speeds of 5–10 m s 1, indicating that the components of mean wind stress were similar to the climatological mean values of 0.029 and 0.058 N m 2.
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
851
Fig. 2. Cruise data from Anon (1965). (a) st (kg m 3) distribution at 50 m. (b) Temperature (1C) distribution at 50 m. (c) Salinity distribution at 50 m. (d) st (kg m 3) distribution at 40 m. (e) Temperature (1C) distribution at 40 m. (f) Salinity distribution at 40 m.
The horizontal sections (Fig. 4a–c) show that the thermohaline structure is three dimensional, and indeed except in the far southwest it has an almost circular pattern, which occupies the whole of the entrance to SG. The 1988 campaign also deployed four current meter moorings (Fig. 1) spanning February–March 1988, the period of the MV Boobook cruise (Table 1). Due to strong tidal currents, the mean vector velocities are all much less than the root mean square velocities (Table 1). Over the two months, however, the vector mean current data were consistent with the existence of a gyral circulation, anti-cyclonic near the surface and cyclonic near the bottom, and on which was superimposed a westward or northwestward surface wind-forced circulation that was the strongest at the southern mooring (Table 1). We return to the gyral circulation in Section 6. The tidal currents (Table 2) are almost barotropic with little change in phase from surface to bottom, and the phases are highly coherent across all four moorings, with a reduction in velocity at the bottom due to bottom friction, The semi-diurnal and diurnal
constituents are of similar magnitude denoting a mixed tide, with their major axes directed approximately along the axis of the gulf. The tidal currents at the north moorings rotate clockwise at the surface and counter clockwise at the bottom. Finally, the diurnal ellipses (in particular K1) are more circular than the semi-diurnal ellipses.
3.2. The RV Franklin survey (summer 1989) During summer 1989, the School of Earth Sciences of Flinders University conducted a survey, Fr 04/89, which included a traverse of the entrance to SG using RV Franklin. The transect ran parallel to and about 15 km to the west of the MV Boobook transect (Fig. 3). The Franklin survey data was consistent with the earlier Boobook results. Fig. 5 shows that surface density minima occur at station 168 and between stations 170 and 171, the density minimum at the latter location being more pronounced and
852
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
The subsequent evolution, which maintains chlorophyll-a levels greater in the gulf than on the shelf throughout the winter, is significant to the ecology within the wider SG. The decrease in spring may be due to utilisation in the surface layers or alternatively to the sinking of chlorophyll-a into the dense saline outflow beneath, but this remains to be investigated
5. The RV Ngerin survey (summer 2009)
Fig. 3. Portions of MV Boobook 03/88, RV Franklin Fr 04/89 and RV Ngerin cruise tracks and MODIS transect (July 2002–June 2003).
evident at all depths. st ranges between 26.40 kg m 3 near the bottom and 26.05 kg m 3 near the surface, and at the former location (station 168) the density minimum is limited to a depth of about 25 m. The position of the major minimum (between stations 170 and 171) occurs about 25 km south of that reported for the MV Boobook cruise in March 1988.
4. MODIS satellite observations (2002–2003) MODIS Aqua chlorophyll-a and SST data for the period July 2002–June 2003 were processed using SeaWIFS Data Analysis System (SEADAS for processing and display of ocean colour data) software. A transect along 1361200 E (Fig. 1) of length 50 km with 25 sampling points at approximately 2 km spacing was used to obtain monthly mean values between 351470 S (inside the entrance to SG, water depth 40 m) and 341240 S (on the shelf, water depth 75 m). It is convenient to present the results using ¨ Hovmoller diagrams spanning the annual cycle, which show the progression of the signals in a space–time graph. The SST gradient (Fig. 6a) undergoes a seasonal cycle. Gulf water was cooler than shelf waters in winter–spring (June– October), followed by a reversal in November after which the gulf waters were warmer than shelf waters during December– April. The SST fronts were the strongest in February–March with SST contrasts of 2–3 1C centred on about 341550 S. These features are a regular aspect of the SST dynamics in the entrance to SG (Petrusevics, 1993). The chlorophyll-a concentrations (Fig. 6b) were the lowest ( E0.2 mg m 3) in spring, following which in December a maximum in concentration formed near 351100 S (which is 150 south of the centre of the SST front) intensified, reaching a maximum of about 0.8 mg m 3 in March, and moved northwards during January–April. This distinctive feature observed in 2003, which was called a ‘chlorophyll-a mound’ in Petrusevics and Bye (2008), suggests that 351100 S is a summer source region for chlorophyll-a which subsequently migrates northwards into SG by about 150 of latitude in two months (mean speed, 5 mm s 1).
Recently additional surveys in the entrance were undertaken as part of the South Australian Integrated Marine Observing System (SAIMOS, 2010) using RV Ngerin. SAIMOS C2009_03 included a transect of seven stations adjacent to the MV Boobook survey of March 1988 (Fig. 3). The latter survey used a SBE 19 þ data logger with conductivity, temperature, pressure, oxygen, and fluorescence sensors. The SAIMOS data confirm the presence of a summer density minimum. The longitudinal section of st between the shelf station JL10 and gulf station JL03 (Fig. 7a) indicates a minimum between stations JL04 and JL07 with a near bottom value of about 26.25 kg m 3 and between stations JL04 and JL05 a minimum of 26.20 kg m 3. The separation of the density minimum positions is about 20 km, the minimum at the former location being more pronounced and evident at all depths. On either sides of the extended minimum the density at the bottom increased, with st rising to 26.30 kg m 3. The extended minimum is centred about 15 km south of that reported in March 1988 (Boobook 03/88). MODIS satellite data captured coincident with the RV Ngerin cruise on 10 March 2009 comprised both MODIS-Terra (southward morning overpass) and MODIS-Aqua (northward afternoon overpass). On both occasions a distinct maximum in the chlorophyll-a level may be observed co-located with the position of the density minimum regions. However an interesting aspect of this is the reduction of chlorophyll-a levels between the morning and afternoon overpasses. The values vary from about 1.6 mg m 3 at station JL07 for the morning overpass, to about 1.0 mg m 3 for the afternoon overpass. A similar reduction is noted for stations JL04-05, 1.1 mg m 3 (morning) and 0.8 mg m 3 (afternoon). This may be due to uptake of the chlorophyll-a as a result of diurnal increase in the photo-period and water temperature. The corresponding fluorescence section (Fig. 7b) shows an extended maximum of about 0.45 mg m 3 coincident with the density minimum and extending throughout much of the water column. These direct observations of the geo-positioning of the longitudinal density and chlorophyll-a variation within the water column in the entrance to SG indicates the high spatial coincidence between these parameters, and together with previous time series of CTD (MV Boobook March 1988 and RV Franklin March 1989 and MODIS satellite chlorophyll-a levels since 2002) do indeed form a sound basis for the association of seasonally occurring physical properties and the upwelling of nutrient rich waters from the shelf.
6. The upwelling model The important question to be addressed is whether the presence of a seasonal density minimum centred in the entrance to SG will promote the uplift responsible for the accumulation of chlorophyll-a, which is essential for the wellbeing of the aquaculture industry of the region (Bierman et al., 2008). The main feature of the density structure (Fig. 4d) is the density front, which impinges on the central region of lower density. The current measurements to the north of this region suggest that
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
853
Fig. 4. Cruise data from MV Boobook (Boobook 03/88) survey, showing cruise track. (a) st (kg m 3) distribution at 10 m. (b) Temperature (oC) distribution at 10 m. (c) Salinity distribution at 10 m. (d) st (kg m 3) vertical section shown in Fig. 4a. (e) Temperature (oC) vertical section shown in Fig. 4a. (f) Salinity vertical section shown in Fig. 4a.
the velocity profile consists of an anti-cyclonic surface flow which is superimposed on a deep cyclonic flow, rather in the nature of a tropical cyclone in the atmosphere. In this system, it would be expected that upward vertical motion occurs in the transition region from inner low density to outer high density. This scenario can be modelled in a simple way by considering a segment along ox, assuming no net zonal transport, and that the density anomaly is independent of depth, then it is easily shown (Bullock, 1975) that u ¼ g=ðf r0 Þ@sT =@yðz þ H=2Þ
ð1Þ
where ox, oy is any right handed horizontal co-ordinate system, and oz is vertically upwards, u is the geostrophic current along ox, g is the acceleration of gravity, ro is a standard density for seawater, z ¼0 is the mean sea level, H is the water depth and f¼2O sin j is the Coriolis parameter in which O is the angular speed of rotation of the Earth, and j is latitude. In terms of the surface current (u0), (1) yields, u¼u0(2zþH)/H. The frictional circulation associated with this geostrophic flow is v ¼ Au in which A 40, is the tangent of the angle of turning to the right hand side. Hence from the continuity equation, assuming that qu/qx ¼0, qw/qz ¼Aqu/qy, where w is the vertical velocity,
854
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
Table 1 Current measurements in the entrance to Spencer Gulf (2 February–6 May 1988). Position and period
Water depth (m)
Current meter depth (m)
a (deg.)
oV4 (cm s 1)
Vs (cm s 1)
North surface (341490 S, 1361330 E) February 2–February 29 February 2–March 31 February 2–May 6
44 44 44
5 5 5
67 70 70
10.30 8.54 6.00
25.01
North bottom (341570 S, 1361310 E) February 2–February 29 February 2–March 31 February 2–May 6
44 44 44
39 39 39
122 114 124
1.36 1.28 1.11
13.83
South surface (351170 S, 1361230 E) February 2–February 29 February 2–March 31 February 2–May 6
79 79 79
5 5 5
97 97 105
10.10 8.26 5.39
20.16
79 79 79
74 74 74
48 42 36
0.74 0.85 0.42
7.15
0
0
South bottom (35117 S, 136123 E) February 2–February 29 February 2–March 31 February 2–May 6
o V 4: vector mean current velocity, a: direction (relative to north), and Vs: root mean square current speed.
Table 2 Tidal current ellipses. Major axis (cm s 1)
Minor axis (cm s 1)
Inclination (deg.)
Phase (g) (deg.)
North surface (67%) O1 8.26 K1 17.32 M2 12.22 S2 12.66
1.38 6.37 1.02 0.60
64 73 53 57
279 317 353 47
North bottom (80%) 4.43 O1 K1 7.72 M2 8.28 S2 9.14
-1.52 -3.45 -0.08 -0.61
50 30 67 66
298 355 354 53
South surface (47%) 4.86 O1 K1 12.59 M2 6.41 S2 7.63
0.57 6.03 0.18 -0.94
64 77 50 48
288 325 351 52
South bottom (61%) 1.61 O1 K1 2.93 M2 4.05 S2 4.13
0.12 -0.29 -0.02 -0.13
19 17 37 36
301 350 355 58
Fig. 5. st (kg m 3) vertical section from the RV Franklin (Fr 04/89) survey.
in which a ¼ sT(0), and for c40, at and beyond the rim (y¼R) of the low density core, st ¼ st(R). On substituting (3) in (1), we obtain, c¼ Dst/R2 in which Dst ¼ st(R)–st(0), and hence uRIM ¼ Dst =R gH=ðf r0 Þ where uRIM ¼u0 (R) is the surface rim velocity, and
Inclination is the angle of the major axis to North. Positive and negative minor axes denote, respectively, clockwise and anticlockwise rotations of the tidal currents in the ellipse. [xx%]: percentage of the variance predicted by the tidal analysis. Note that the north surface and bottom data are from different moorings, see Table 1.
and on integrating with respect z using (1), and assuming a flat bottom we obtain w ¼ A zðz þ HÞ=H @u0 =@y
ð2Þ
which has a maximum, wmax ¼ AH/4qu0/qy at mid-depth (z¼ ½H). Thus, for an anti-cyclonic surface flow at the rim of the low density core (qu0/qyo0 along ox), v is radially inward at the bottom and outward at the surface, and w is upwards. On applying this model to the density minimum in the entrance to SG assuming a quadratic variation of st outwards from the central co-ordinate (which is representative of the st distribution shown in Figs. 4d, 5, and 7a), we have 2
sT ðyÞ ¼ a þcy y o R sT ðyÞ ¼ sT ðRÞ y4 R
ð4Þ
ð3Þ
wmax ¼ A Dst =R2 gH2 =ðf r0 Þ
ð5Þ
which is a constant throughout the low density core. On evaluating (4) and (5) using r0 ¼103 kg m 3, g ¼10 m s 2, f¼ 10 4 s 1 and from the distributions of st (Figs. 4d, 5, and 7a), Dst ¼ 0.3 kg m 3, H¼50 m, and R¼30 km, we obtain, uRIM ¼ 5 cm s 1 which indicates a vertical baroclinic shear of 10 cm s 1 at the rim in reasonable agreement with observations (Table 1), and for A¼0.1, which corresponds with an angle of turning of 61, wmax ¼ 2 10 5 m s 1. This analysis suggests that the density minimum is a very efficient mechanism for uplifting nutrient rich water in the entrance to SG. Note that for a sustained upward velocity of 2 10 5 m s 1, a particle would rise from the bottom to the top of the water column (50 m) in about 1 month. 7. Conclusions The mechanism for the key formation of chlorophyll-a maxima has been investigated in this paper using a variety of cruise, current
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
855
¨ Fig. 6. Hovmoller diagrams of the monthly sequence of (a) SST (oC) and (b) chlorophyll-a (mg m 3) data from the MODIS transect shown in Fig. 1.
meter, and satellite data. The formation of the summer density minimum in the entrance to Spencer Gulf by the interaction of gulf water, shelf water, and Great Australian Bight water creates a novel mechanism for the upwelling of nutrient-rich water. This mechanism which depends on the GAB water to form the density minimum through an intrusion into the centre of the entrance to SG differs from the simpler model of Nunez Vaz et al. (1990) in which only the interaction shelf and gulf waters was considered. It appears that the dynamics are robust, as the density minimum has been observed in
several years; the earliest depiction (Bullock, 1975) being from a cruise in March 1963. The chlorophyll-a in lower SG increases through summer to an autumn maximum and then decreases to the spring minimum (Fig. 6b). This is consistent with field observations reported by Van Ruth et al. (2009) in the lower section of SG in which the highest gross phytoplankton growth rates occurred in March and May 2007 with rates 30% lower in July and October. The analytical upwelling model suggests the need for a linked biochemical-hydrodynamic numerical model of the region, which
856
P. Petrusevics et al. / Continental Shelf Research 31 (2011) 849–856
Centre, Greenbelt MD 20771 are also gratefully acknowledged for the production and distribution of these data. Carlos Teixeira was supported by the Brazilian Research Council (CAPES 4012-05-4).
References
Fig. 7. Cruise data from RV Ngerin (SAIMOS C2009_03) survey. (a) st (kg m 3) vertical section. (b) chlorophyll-a (mg m 3) vertical section.
encompasses accumulation, utilisation, and dispersion of nitrogen and phosphates. This is of vital importance for improved understanding of marine productivity in Spencer Gulf in support of aquaculture and fisheries.
Acknowledgements The authors are grateful to the crew of the MV Boobook and colleagues from the School of Earth Sciences, the Flinders University of South Australia (especially Professor Geoff. Lennon and Dr Rick Nunes Vaz) and also the CSIRO Marine data section in Hobart for the supply of the 1989 RV Franklin cruise data, and to Dr. John Middleton of SARDI and the crew of MV Ngerin for the 2009 cruise data. The SeaWIFS and MODIS Project and the Distributed Active Archive Centre at the Goddard Space Flight
Anon, 1965. H.M.A.S. Gascoyne Cruise Anon (1965) Oceanogr. Cruise Rep. 43 Oceanographical Observations in the Indian Ocean in G2/65. Anon, 1968a. Investigations by FRV Investigator on the South Australian Tuna Grounds in 1961 CSIRO Oceanographical Station List Cruise No. 64. Anon, 1968b. Investigations by FRV Investigator on the South Australian Tuna Grounds in 1964 CSIRO Oceanographical Station List Cruise No. 67. Anon, 1969. Oceanographical observations in the Indian Ocean in 1966. HMAS Diamantina Cruise Dm 2/66. CSIRO Oceanographical Cruise Report No. 54. Bierman, P., Lewis, M., Tanner, J., Ostendorf, B., 2008. Understanding phytoplankton variability throughout Spencer Gulf, South Australia via satellite derived chlorophyll-a. In: IEEE International Proceedings of Geoscience and Remote Sensing Symposium IGARSS 2008, 4, Boston, MA, pp. 902–905. Bruce, B.D, Short, D.A., 1990. Observations on the distribution of larval fish in relation to a frontal zone at the entrance to Spencer Gulf, South Australia. In: Bureau of Rural Resources Proceedings No. 15. Bullock, D.A., 1975. The general water circulation of Spencer Gulf, South Australia, in the period February–May. Trans. R. Soc. S. Aust. 99, 43–53. Bye, J.A.T., 1983. The general circulation in a dissipative basin with longshore wind stresses. J. Phys. Oceanogr. 13, 1553–1563. Hahn, S.D., 1986. Physical structure of the waters of the South Australian continental shelf. Ph.D. Thesis, School of Earth Sciences, The Flinders University of South Australia, 285pp. Herzfeld, M., Tomczak, M., 1999. Bottom-driven upwelling generated by eastern intensification in closed and semi-enclosed basins. Aust. J. Mar. Freshwater Res. 50, 613–628. Middleton, J.F., Bye, J.A.T., 2007. A review of the shelf-slope circulation along Australia’s southern shelves: Cape Leeuwin to Portland. Progr. Oceanogr. 75, 1–41. Nunes, R.A., Lennon, G.W., 1986. Physical property distributions and seasonal trends in Spencer Gulf, South Australian inverse estuary. Aust. J. Mar. Freshwater Res. 37, 39–53. Nunez Vaz, R.A., Lennon, G.W., Bowers, D.G., 1990. Physical behaviour of a large, negative or inverse estuary. Cont. Shelf Res. 10, 277–304. Petrusevics, P.M., 1993. SST fronts in inverse estuaries, South Australia—indicators of reduced gulf-shelf exchange. Aust. J. Mar. Freshwater Res. 44, 305–323. Petrusevics, P, Bye, J, 2008. Satellite observations of SST and chlorophyll-a time series at the gateway to the inverse estuary Spencer Gulf. In: Proceedings of the Australian Marine Science Association (SA Branch), October 2008. Petrusevics, P., Bye, J.A.T., Fahlbusch, V., Hammat, J., Tippins, D.R., van Wijk, E., 2009. High salinity winter outflow form a mega-inverse-estuary, the Great Australian Bight. Cont. Shelf Res. 29, 371–380. Ridgway, K.R., Dunn, J.R., Wilkin, J.L., 2002. Ocean interpolation by four-dimensional least squares—application to the waters around Australia. J. Atmos. Ocean. Tech. 19, 1357–1375. Richardson, L.E., Kyser, T.K., James, N.P., Bone, Y., 2009. Analysis of hydrographic and stable isotope data to determine water masses, circulation, and mixing in the eastern Great Australian Bight. J. Geophys. Res. 114, C10016. doi:10.1029/ 2009JC005407. SAIMOS, 2010. Cruise files. South Australian Integrated Marine Observing System. SARDI Aquatic Sciences West Beach South Australia 5022, Australia. Savidge, G., Foster, P., 1978. Phytoplankton biology of a thermal front in the Celtic Sea. Nature 271, 155. Simpson, J.H., Hunter, J.R., 1974. Fronts in the Irish Sea. Nature 250, 404–406. Simpson, J.H., Bowers, D.G., 1981. Models of stratification and frontal movement in shelf seas. Deep-Sea Res. A28, 727–738. Van Ruth, P., Thompson, P., Bonham, P., Jones, E., 2009. Primary production and zooplankton ecology in Port Lincoln Tuna Farming Zone. Aquafin CRC Project 4.6 March 2009. FRDC Project No. 2005/059. Worm, B., Sandow, M., Oschlies, A., Lotze, H.K., Myers, R.A., 2005. Global patterns of predator diversity in the open oceans. Science 309, 1365–1369. Yanagi, T., Yoshikawa, K., 1987. Tidal fronts in Hiuchi-nada and Osaka Bay. Bull. Jpn. Soc. Fish. Oceanogr. (Sisan Kaiyo Kenkyukaiho) 51, 115–119.