Poster presented at the XXV General Assembly of the European Geophysical Society, Nice, April 2000.

Upper Ocean Mixing in the NW Weddell Sea During Winter


Robin Muench, Susan Howard, and Laurie Padman

Introduction

Dovetail CTD Stations


The figure above depicts the study area covered by the 1997 Dovetail cruise. In the present study, we focus our attention on Section 1, a combined meridional (48-50°W) and zonal (63°S) transect. Numbered points indicate locations of oceanographic stations.

The following measurements were obtained at each oceanographic station:

  • a full-depth CTD cast made using a SeaBird SBE 911 system, and
  • upper ocean (20-350 m) vertical current profile measurements made using an RDI 150-kHz hull-mounted ADCP.



Regional Mean circulation


The mean circulation, shown schematically above (continuous gray arrows), is reflected in the upper ocean ADCP observations (black arrows). The vectors reflect the average on-station ADCP measurement (103m - 263m). Vectors of magnitude smaller than the CATS model RMS tidal current speed (shown in color) are not included. Future work will include using an improved CATS model to remove the tidal signal from ADCP measurements.

The mean circulation south of about 61° S is dominated by the western Weddell boundary current and reflects strong topographic control. Initially flowing northeastward, this current bifurcates east of Joinville I. One branch continues eastward south of the South Orkney Plateau, while a second branch veers sharply left, toward the northwest, and is the primary contributor to a cyclonic circulation about Powell Basin. A portion of this latter branch flows southward then eastward around the Plateau.

Return to index

Regional Tidal Currents

Moored current observations and numerical model output provide information on tidal currents in the study area. Modeled tidal currents in the basins are less than 5 cm s-1, however, tides interact with the shelf and bank topography to produce localized currents that can exceed 40 cm s-1 in shallow water (see mean circulation figure).

The topographic dependence of strong tides is shown clearly in current meter data obtained from morrings located across the shelf break to the southwest of Joinville Island. The major axes of both the semidiurnal and diurnal tidal constituents increase with decreasing bottom depth (see figure below). Estimates from the CATS 99.2 model are also shown for comparison. CATS 99.2 overpredicts diurnal tidal energy near the shelf break, whereas a preliminary inverse model solution for Weddell Sea tides underpredicts the same signal. Further inverse modeling using current meters and tide height data will be used to improve the model-data fit.

Return to index

Semidiurnal Currents in Powell Basin

Stations 60-65 were taken along the section across Powell Basin with approximately 6-hour intervals between adjacent stations. Current profiles at these stations showed semidiurnal flow reversals of order 10 cm s-1 (see figure below) between 125-225 m. The greatest shear appeared to be associated with the semidiurnal fluctuations below the base of the upper mixed layer.

These fluctuations are reminiscent of the semidiurnal internal waves noted in CTD profiles from the southeast part of our study area by Foster [1994] and may reflect tidal or inertial motions originating on the banks and shelves to the east and west.

Return to index

Mixing Considerations

Using the ADCP, we measured large shears in the stratified regions of the water column, suggesting that shear instabilities may be active and provide a mechanism for producing vertical mixing. Possible sources of this shear include baroclinic tides that may be generated at the continental slopes, and near- inertial waves forced by flow over topography or by wind stress. We use the gradient Richardson number (Ri) and Vertical Diffusivity (Kv) estimates as guides to the possibility of mixing occurring in the stratified part of the water column.

Richardson Number

To estimate the Richardson number, we used Ri=N2/S2, where N is the buoyancy frequency obtained from CTD data and S is the magnitude of the vertical shear of horizontal velocity. From ADCP data:

S2=(D á uñ /D z)2+(D á vñ /D z)2

Here, á .ñ indicates averaging over the ADCP depth bin (height D z=8 m) and the on-station time period T (typically ~2 h).

With D z=8 m and T=2 h, we do not fully resolve velocity shear and N(z) variations [Wijesekera et al., 1993], thus Ri is generally overestimated and is only a rough guide to mixing. With this caveat, we assume that

Ri < 0.25 => Mixing is likely (shear instabilities)

Ri < 1.0 => Mixing is probable (shear instabilities on scales < D z & T, or advective instabilities).

As an example, detailed results of our mixing calculations are shown for station 60 (see figure below). Here we show the vertical profiles of temperature and density obtained from the CTD data along with calculated values for N, S, and Ri. A strong shear layer is seen between 175 m and 240 m with associated low Ri numbers between 0.25 and 1.0. The effects of mixing associated with the low Ri numbers can be seen in the temperature and density profiles.

Based on calculations such as these for all stations:

  • Ri< 1 at more locations in Powell Basin and on the Peninsula slope and shelf than elsewhere.
  • Low Ri coincided with strong tidal currents seen over the inner peninsula shelf and with the semidiurnally varying mid-water column shear seen in central Powell Basin.

Vertical Diffusivity

To estimate the vertical diffusivity, Kv, we use the bulk diffusivity parameterization of Pacanowsky and Philander [1981]:

Kv = [5 x 10-3 + 10-4(1 + 5 Ri)2 ]/[(1+ 5 Ri)3 ] + 10-5

This estimation of Kv (m2/s) represents a bulk parameterization indicating where regions of weak stratification and strong shear are more likely to correspond to regions of high vertical diffusivities.

The estimates of Kv are shown for Section 1 along with temperature (T), dissolved oxygen (DO), density (s o), and bottom depth (see figure below).



Return to index

Heat Flux

To assess the effects of mixing on vertical heat fluxes in the region, we used the vertical diffusivity values to estimate heat fluxes between layers. Heat fluxes between adjacent bins at each station were calculated using:

Hf = rCp Kv DT/ Dz

where Cp is the heat capacity of sea water, Dz= 8 m , and T and rare the temperature and density obtained from CTD measurements.

Heat flux estimates between several layers are shown below for the entire DOVETAIL study area (see figure below). The mean heat flux over the Powell Basin and the adjacent banks and shelves is estimated at ~4 W m-2, compared with ~2 W m-2 over the other regions. For comparison, Robertson et al. [1995] found ~2 W m-2 in the western Weddell Sea, and the large-scale mean heat flux form WDW in the Weddell Gyre is estimated at ~19 W m-2 [Fahrbach et al., 1994].

Return to index

Conclusions

  • Tidally generated currents, enhanced by the complex topography associated with shelves and banks east and northeast of the Antarctic Peninsula, are a primary regional source for upper ocean mixing.


  • Winter water mass modification on the inner Antarctic Peninsula shelf was probably enhanced by benthic tidal stirring.


  • Diapycnal mixing below the mixed layer, subsequent to interleaving, can be driven by tidal (or possibly inertial) shear.


  • A significant baroclinic semidiurnal signal persists in weakly stratified areas away from the regions of topographically enhanced currents. The source of the signal may be the generation of internal tidal waves as tides flow over the adjacent continental slope.


  • Based on the Pacanowsky and Philander [1981] vertical diffusivity parameterization, the average vertical heat flux through the thermocline in Powell Basin is 2-3 times greater than the average vertical heat flux (2 W m-2) elsewhere in the region

Return to index

References

Fahrbach, E., G. Rohardt, M. Scröder, and V. Strass, 1994: Transport and structure of the Weddell Gyre, Annales Geophysicae, 12, 840-855.


Foster, T.D., 1994: Large, steplike temperature and salinity structures observed in the central Weddell Sea. Antarctic J. of the United States, XXIX (5), 99-100.


Pacanowsky, R.C. and S.G.H. Philander, 1981: Parameterization of vertical mixing in numerical models of tropical oceans. J. Phys. Oceanog., 11, 1443-1451.


Robertson, R., L. Padman, and M. D. Levine , 1995: Finestructure, microstructure, and vertical mixing processes in the upper ocean in the western Weddell Sea. J. Geophys. Res., 100, 18,517-18,536.


Wijesekera, H., L. Padman, T. Dillon, M. Levine, C. Paulson, and R. Pinkel, 1993: The application of internal-wave dissipation models to a region of strong mixing. J. Phys. Oceanog., 23, 269-286.
Return to index

Acknowledgements

The research reported here has been carried out with funding from NSF grant OPP-9527667. We are indebted to Eberhard Fahrbach of the Alfred-Wegener-Institute in Bremerhaven and Marc Garcia of the UPC in Barcelona for providing us with the moored current observations from the NW Weddell Sea.
Return to index

Questions or comments regarding this site may be sent to: Susan Howard
Site last updated on April 26, 2000.