• American Meteorological Society, 2017: Thermally indirect. Glossary of Meteorology, http://glossary.ametsoc.org/wiki/thermally_indirect.

  • Badin, G., R. G. Williams, Z. Jing, and L. Wu, 2013: Water mass transformations in the Southern Ocean diagnosed from observations: Contrasting effects of air–sea fluxes and diapycnal mixing. J. Phys. Oceanogr., 43, 14721484, doi:10.1175/JPO-D-12-0216.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Belcher, S. E., and et al. , 2012: A global perspective on Langmuir turbulence in the ocean surface boundary layer. Geophys. Res. Lett., 39, L18605, doi:10.1029/2012GL052932.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cerovečki, I., L. D. Talley, M. R. Mazloff, and G. Maze, 2013: Subantarctic Mode Water formation, destruction, and export in the eddy-permitting Southern Ocean state estimate. J. Phys. Oceanogr., 43, 14851511, doi:10.1175/JPO-D-12-0121.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Danabasoglu, G., R. Ferrari, and J. C. McWilliams, 2008: Sensitivity of an ocean general circulation model to a parameterization of near-surface eddy fluxes. J. Climate, 21, 11921208, doi:10.1175/2007JCLI1508.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dong, S., S. T. Gille, and J. Sprintall, 2007: An assessment of the Southern Ocean mixed layer heat budget. J. Climate, 20, 44254442, doi:10.1175/JCLI4259.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dong, S., J. Sprintall, S. T. Gille, and L. Talley, 2008: Southern Ocean mixed-layer depth from Argo float profiles. J. Geophys. Res., 113, C06013, doi:10.1029/2006JC004051.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Downes, S. M., A. S. Budnick, J. L. Sarmiento, and R. Farneti, 2011: Impacts of wind stress on the Antarctic Circumpolar Current fronts and associated subduction. Geophys. Res. Lett., 38, L11605, doi:10.1029/2011GL047668.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrari, R., J. C. McWilliams, V. M. Canuto, and M. Dubovikov, 2008: Parameterization of eddy fluxes near oceanic boundaries. J. Climate, 21, 27702789, doi:10.1175/2007JCLI1510.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hanawa, K., and L. D. Talley, 2001: Mode waters. Ocean Circulation and Climate, International Geophysical Series, Vol. 77, Academic Press, 373–386.

    • Crossref
    • Export Citation
  • Herraiz-Borreguero, L., and S. R. Rintoul, 2011: Subantarctic mode water: Distribution and circulation. Ocean Dyn., 61, 103126, doi:10.1007/s10236-010-0352-9.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hogg, A. McC., 2010: An Antarctic Circumpolar Current driven by surface buoyancy forcing. Geophys. Res. Lett., 37, L23601, doi:10.1029/2010GL044777.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holte, J., and L. Talley, 2009: A new algorithm for finding mixed layer depths with applications to Argo data and Subantarctic Mode Water formation. J. Atmos. Oceanic Technol., 26, 19201939, doi:10.1175/2009JTECHO543.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holte, J., L. Talley, T. K. Chereskin, and B. M. Sloyan, 2012: The role of air–sea fluxes in Subantarctic Mode Water formation. J. Geophys. Res., 117, C03040, doi:10.1029/2011JC007798.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holte, J., J. Gilson, L. Talley, and D. Roemmich, 2016: Argo mixed layers. Scripps Institution of Oceanography, http://mixedlayer.ucsd.edu.

  • Jones, D. C., A. J. S. Meijers, E. Shuckburgh, J. B. Sallée, P. Haynes, E. K. McAufield, and M. R. Mazloff, 2016: How does Subantarctic Mode Water ventilate the Southern Hemisphere subtropics? J. Geophys. Res. Oceans, 121, 65586582, doi:10.1002/2016JC011680.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and et al. , 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kuo, A., R. A. Plumb, and J. Marshall, 2005: Transformed Eulerian-mean theory. Part II: Potential vorticity homogenization and the equilibrium of a wind- and buoyancy-driven zonal flow. J. Phys. Oceanogr., 35, 175187, doi:10.1175/JPO-2670.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kuo, H.-L., 1956: Forced and free meridional circulations in the atmosphere. J. Meteor., 13, 561568, doi:10.1175/1520-0469(1956)013<0561:FAFMCI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., and S. G. Yeager, 2009: The global climatology of an interannually varying air–sea flux data set. Climate Dyn., 33, 341364, doi:10.1007/s00382-008-0441-3.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., J. C. McWilliams, and S. C. Doney, 1994: Oceanic vertical mixing: A review and a model with a nonlocal boundary layer parameterization. Rev. Geophys., 32, 363403, doi:10.1029/94RG01872.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., G. Danabasoglu, S. C. Doney, and J. C. McWilliams, 1997: Sensitivity to surface forcing and boundary layer mixing in a global ocean model: Annual-mean climatology. J. Phys. Oceanogr., 27, 24182447, doi:10.1175/1520-0485(1997)027<2418:STSFAB>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lee, M. M., A. J. G. Nurser, I. Stevens, and J. B. Sallée, 2011: Subduction over the Southern Indian Ocean in a high-resolution atmosphere–ocean coupled model. J. Climate, 24, 38303849, doi:10.1175/2011JCLI3888.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Qian, S. Lee, and A. Griesel, 2016: Eddy fluxes and jet-scale overturning circulations in the Indo–western Pacific Southern Ocean. J. Phys. Oceanogr., 46, 29432959, doi:10.1175/JPO-D-15-0241.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Qing, A. Webb, B. Fox-Kemper, A. Craig, G. Danabasoglu, W. G. Large, and M. Vertenstein, 2016: Langmuir mixing effects on global climate: WAVEWATCH III in CESM. Ocean Modell., 103, 145160, doi:10.1016/j.ocemod.2015.07.020.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Qing, B. Fox-Kemper, ∅. Breivik, and A. Webb, 2017: Statistical modeling of global Langmuir mixing. Ocean Modell., 113, 95114, doi:10.1016/j.ocemod.2017.03.016.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Maltrud, M. E., F. Bryan, and S. Peacock, 2010: Boundary impulse response functions in a century-long eddying global ocean simulation. Environ. Fluid Mech., 10, 275295, doi:10.1007/s10652-009-9154-3.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mazloff, M., P. Heimbach, and C. Wunsch, 2010: An eddy-permitting Southern Ocean state estimate. J. Phys. Oceanogr., 40, 880899, doi:10.1175/2009JPO4236.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McCartney, M. S., 1977: Subantarctic mode water. A Voyage of Discovery, M. Angel, Ed., Elsevier, 103–119.

  • McCartney, M. S., 1982: The subtropical recirculation of Mode Water. J. Mar. Res., 40, 427464.

  • Morrison, A. K., O. A. Saenko, A. M. Hogg, and P. Spence, 2013: The role of vertical eddy flux in Southern Ocean heat uptake. Geophys. Res. Lett., 40, 54455450, doi:10.1002/2013GL057706.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pfeffer, R. L., 1981: Wave-mean flow interactions in the atmosphere. J. Atmos. Sci., 38, 13401359, doi:10.1175/1520-0469(1981)038<1340:WMFIIT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Phillips, H. E., and S. R. Rintoul, 2000: Eddy variability and energetics from direct current measurements in the Antarctic Circumpolar Current south of Australia. J. Phys. Oceanogr., 30, 30503076, doi:10.1175/1520-0485(2000)030<3050:EVAEFD>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ribbe, J., 1999: On wind-driven mid-latitude convection in ocean general circulation models. Tellus, 51A, 517525, doi:10.3402/tellusa.v51i4.14116.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rintoul, S. R., and M. H. England, 2002: Ekman transport dominates local air–sea fluxes in driving variability of Subantarctic Mode Water. J. Phys. Oceanogr., 32, 13081321, doi:10.1175/1520-0485(2002)032<1308:ETDLAS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Robinson, W. A., 2006: On the self-maintenance of midlatitude jets. J. Atmos. Sci., 63, 21092122, doi:10.1175/JAS3732.1.

  • Sabine, C. L., and et al. , 2004: The oceanic sink for anthropogenic CO2. Science, 305, 367371, doi:10.1126/science.1097403.

  • Sallée, J. B., N. Wienders, K. Speer, and R. Morrow, 2006: Formation of subantarctic mode water in the southeastern Indian Ocean. Ocean Dyn., 56, 525542, doi:10.1007/s10236-005-0054-x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sallée, J. B., K. Speer, and S. Rintoul, 2010: Zonally asymmetric response of the Southern Ocean mixed-layer depth to the Southern Annular Mode. Nat. Geosci., 3, 273279, doi:10.1038/ngeo812.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sallée, J. B., R. Matear, S. R. Rintoul, and A. Lenton, 2012: Localized subduction of anthropogenic carbon dioxide in the Southern Hemisphere oceans. Nat. Geosci., 5, 579584, doi:10.1038/ngeo1523.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sarmiento, J. L., N. Gruber, M. A. Brzezinski, and J. P. Dunne, 2004: High-latitude controls of thermocline nutrients and low latitude biological productivity. Nature, 427, 5660, doi:10.1038/nature02127.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schiller, A., and K. R. Ridgway, 2013: Seasonal mixed-layer dynamics in an eddy-resolving ocean circulation model. J. Geophys. Res. Oceans, 118, 33873405, doi:10.1002/jgrc.20250.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schmitz, W. J., Jr., 1996: On the eddy field in the Agulhas retroflection, with some global considerations. J. Geophys. Res., 101, 16 25916 271, doi:10.1029/96JC01143.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Smith, R. D., J. K. Dukowicz, and R. C. Malone, 1992: Parallel ocean general circulation modeling. Physica D, 60, 3861, doi:10.1016/0167-2789(92)90225-C.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Treguier, A. M., I. M. Held, and V. D. Larichev, 1997: Parameterization of quasigeostrophic eddies in primitive equation ocean models. J. Phys. Oceanogr., 27, 567580, doi:10.1175/1520-0485(1997)027<0567:POQEIP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weller, R. A., P. W. Furey, M. A. Spall, and R. E. Davis, 2004: The large-scale context for oceanic subduction in the northeast Atlantic. Deep-Sea Res., 51, 665699, doi:10.1016/j.dsr.2004.01.003.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wenegrat, J. O., and M. J. McPhaden, 2016: Wind, waves, and fronts: Frictional effects in a generalized Ekman model. J. Phys. Oceanogr., 46, 371394, doi:10.1175/JPO-D-15-0162.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Williams, R. G., C. Wilson, and C. W. Hughes, 2007: Ocean and atmosphere storm tracks: The role of eddy vorticity forcing. J. Phys. Oceanogr., 37, 22672289, doi:10.1175/JPO3120.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • View in gallery

    June MLD (color; m) simulated by the POP model. Contours show Gaussian-boxcar-smoothed (2.5° × 2.5°) and depth-averaged (above 1 km) zonal current speed (cm s−1) that exceeds 6 cm s−1 (hatched) and 11 cm s−1 (thick) with an interval of 5 cm s−1. The black box indicates the domain defined as the Indo–western Pacific Southern Ocean (30°–60°S, 50°–190°E) in this study.

  • View in gallery

    Monthly MLD (color; m) from the POP model simulation for (a)–(f) May–October in the Indo–western Pacific Southern Ocean (30°–60°S, 50°–190°E). Contours show Gaussian-boxcar-smoothed (2.5° × 2.5°) and depth-averaged (above 1 km) zonal current speed (cm s−1) that exceeds 6 cm s−1 (hatched) and 11 cm s−1 (thick) with an interval of 5 cm s−1. The red box indicates the sector (40°–55°S, 130°–142°E) where the current is especially zonal and persistent and the MLD is deepest.

  • View in gallery

    Monthly MLD (color; m) based on the density criterion (Δσ = 0.03 kg m−3) from the Argo float data for (a)–(f) May–October in the Indo–western Pacific Southern Ocean (50°–190°E, 30°–60°S). Contours show Gaussian-boxcar-smoothed (2.5° × 2.5°) and depth-averaged (above 1 km) zonal current speed (cm s−1) that exceeds 6 cm s−1 (hatched) and 11 cm s−1 (thick) with an interval of 2.5 cm s−1. The velocity data are from the SOSE.

  • View in gallery

    (a)–(d) Vertical cross sections of sector-zonal-mean (130°–142°E) zonal current speed (color; cm s−1) and MLD (lines; m) from POP for austral summer (DJF) through austral spring (SON). (e),(f) As in (c),(d) but for zonal current speed from SOSE and MLD from Argo. The vertical black dashed (dashed–dotted) line indicates the latitude of the SAF core from POP (SOSE).

  • View in gallery

    Monthly vertical cross sections of sector-zonal-mean (130°–142°E) buoyancy frequency N2 (color; 10−5 s−2). The white solid line shows the MLD (m) for each month. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

  • View in gallery

    Monthly vertical cross sections of sector-zonal-mean (130°–142°E) tendency of buoyancy frequency ∂N2/∂t (color; 10−6 s−2 month−1). The white solid line shows the MLD (m) for each month. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

  • View in gallery

    (a) Schematic diagram of eddy momentum flux–driven jet (), JSOC (→), and JSOC-driven equatorward potential density σ (contours) gradient north of the SAF. Red arrows (→) denote the eddy momentum flux. (b) Vertical cross section of November–April mean potential density (contours; 10−2 kg m−3) and the change in potential density (color; kg m−3) caused by the sector-zonal-mean (130°–142°E) vertical advection during the period of November–April. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S). The white box indicates the region of the diagram shown in (a).

  • View in gallery

    As in Fig. 6, but for the tendency of buoyancy frequency ∂N2/∂t contributed from the JSOC.

  • View in gallery

    Monthly vertical cross sections of sector-zonal-mean (130°–142°E) tendency of buoyancy frequency ∂N2/∂t (color; 10−6 s−2 mon−1) contributed from the eddy flux and vertical cross sections of potential density (contours; kg m−3) with an interval of 0.05 kg m−3. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

  • View in gallery

    As in Fig. 6, but for the tendency of buoyancy frequency ∂N2/∂t contributed from the residual.

  • View in gallery

    (a) Monthly MLD (m) and net air–sea heat flux (Qnet; W m−2) from the (b) POP and (c) NCEP–NCAR reanalysis for the sector (130°–142°E) during austral winter (JJA). Positive fluxes in (b) and (c) are directed into the ocean. The sector-zonal-mean (130°–142°E) and JJA-mean zonal current speed (cm s−1; gray line) at 918 m are overlaid to show the location of the SAF. The vertical black dashed line represents the latitude of the SAF core (~48.7°S). The location of the MLD wedge is highlighted by the pink shading.

  • View in gallery

    (top half) Monthly net air–sea heat flux (Qnet;W m−2) from the POP for the sector (130°–142°E). (bottom half) Vertical cross sections of sector-zonal-mean (130°–142°E) monthly potential temperature (color; °C) and vertical velocity (vectors; 10−3 cm s−1). Green (white) vectors indicate the downwelling (upwelling). The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

  • View in gallery

    Monthly (a) Ekman meridional velocity (cm s−1), (b) depth-averaged (above 100 m) potential density (kg m−3), and (c) Ekman meridional potential density advection (10−2 kg m−3 month−1) for the sector (130°–142°E) during austral winter (JJA). The sector-zonal-mean (130°–142°E) and JJA-mean zonal current speed (cm s−1; gray line) at 918 m is overlaid to show the location of the SAF. The vertical thick dashed line indicates the latitude of the SAF core (~48.7°S).

  • View in gallery

    Monthly vertical cross sections of sector-zonal-mean (130°–142°E) monthly meridional eddy buoyancy flux (color; 10−1 cm2 s−3) and potential density (contours; kg m−3) with an interval of 0.05 kg m−3. For the eddy buoyancy flux, positive (negative) values indicate equatorward (poleward) transport. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

  • View in gallery

    Schematic diagram of the processes that contribute to the Southern Ocean mixed layer destratification during austral winter (JJA): net air–sea heat flux (Qnet), meridional Ekman advection, eddy buoyancy flux, and downward branch of the JSOC.

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 73 73 5
PDF Downloads 41 41 4

A Mechanism of Mixed Layer Formation in the Indo–Western Pacific Southern Ocean: Preconditioning by an Eddy-Driven Jet-Scale Overturning Circulation

View More View Less
  • 1 Department of Meteorology and Atmospheric Science, The Pennsylvania State University, University Park, Pennsylvania
© Get Permissions
Full access

Abstract

The formation of a narrow band of the deep winter mixed layer (hereinafter “mixed layer wedge”) in the Indo–western Pacific Southern Ocean is examined using an eddy-resolving Parallel Ocean Program (POP) model simulation. The mixed layer wedge starts to deepen in June, centered at 47.5°S, with a meridional scale of only ~2° latitude. Its center is located ~1° north of the model’s Subantarctic Front (SAF). The Argo-based observed mixed layer is similarly narrow and occurs adjacent to the observed SAF. In the small sector of 130°–142°E, where the SAF is persistent and the mixed layer is deepest, the formation of the narrow mixed layer wedge coincides with destratification underneath the mixed layer. This destratification can be attributed primarily to the downwelling branch of a jet-scale overturning circulation (JSOC). The JSOC, which was reported in an earlier study by the authors, is driven by eddy momentum flux convergence and is therefore thermally indirect: its descending branch occurs on the warmer equatorward flank of the SAF, promoting destratification during the warm season. The model-generated net air–sea heat flux reveals a similar wedge-like feature, indicating that the flux contributes to the mixed layer depth wedge, but again this feature is preconditioned by the JSOC. Ekman advection contributes to the formation of the mixed layer, but it occurs farther north of the region where the mixed layer initially deepens. These findings suggest that the eddy-driven JSOC associated with the SAF plays an important role in initiating the narrow, deep mixed layer wedge that forms north of the SAF.

Publisher’s Note: This article was revised on 22 November 2017 to correct an error in the presentation of Eq. (1) when originally published.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Qian Li, qvl5065@psu.edu

Abstract

The formation of a narrow band of the deep winter mixed layer (hereinafter “mixed layer wedge”) in the Indo–western Pacific Southern Ocean is examined using an eddy-resolving Parallel Ocean Program (POP) model simulation. The mixed layer wedge starts to deepen in June, centered at 47.5°S, with a meridional scale of only ~2° latitude. Its center is located ~1° north of the model’s Subantarctic Front (SAF). The Argo-based observed mixed layer is similarly narrow and occurs adjacent to the observed SAF. In the small sector of 130°–142°E, where the SAF is persistent and the mixed layer is deepest, the formation of the narrow mixed layer wedge coincides with destratification underneath the mixed layer. This destratification can be attributed primarily to the downwelling branch of a jet-scale overturning circulation (JSOC). The JSOC, which was reported in an earlier study by the authors, is driven by eddy momentum flux convergence and is therefore thermally indirect: its descending branch occurs on the warmer equatorward flank of the SAF, promoting destratification during the warm season. The model-generated net air–sea heat flux reveals a similar wedge-like feature, indicating that the flux contributes to the mixed layer depth wedge, but again this feature is preconditioned by the JSOC. Ekman advection contributes to the formation of the mixed layer, but it occurs farther north of the region where the mixed layer initially deepens. These findings suggest that the eddy-driven JSOC associated with the SAF plays an important role in initiating the narrow, deep mixed layer wedge that forms north of the SAF.

Publisher’s Note: This article was revised on 22 November 2017 to correct an error in the presentation of Eq. (1) when originally published.

© 2017 American Meteorological Society. For information regarding reuse of this content and general copyright information, consult the AMS Copyright Policy (www.ametsoc.org/PUBSReuseLicenses).

Corresponding author: Qian Li, qvl5065@psu.edu

1. Introduction

The ocean mixed layer is a gateway for atmosphere–ocean exchange. Some of the deepest mixed layers develop in the Southern Ocean, particularly on the equatorward flank of the Antarctic Circumpolar Current (ACC; McCartney 1977). It has been shown that the development of this deep mixed layer tends to occur simultaneously with Subantarctic Mode Water (SAMW) and Antarctic Intermediate Water (AAIW) formation (McCartney 1977; 1982; Hanawa and Talley 2001; Rintoul and England 2002; Sallée et al. 2006; Holte and Talley 2009; Herraiz-Borreguero and Rintoul 2011; Holte et al. 2012; Cerovečki et al. 2013). North of the Subantarctic Front (SAF), the northernmost front within the ACC, the deepest mixed layers are congruent with the subsurface salinity minimum, a signature of SAMW in the southeast Indian Ocean (Sallée et al. 2006) and of AAIW in the southeast Pacific Ocean (Holte and Talley 2009). These water masses play key roles for the oceanic transport of heat, carbon, and nutrients (Sarmiento et al. 2004; Sabine et al. 2004; Sallée et al. 2012; Jones et al. 2016). Therefore, it is important to understand the mechanism of mixed layer formation.

Numerous observation-based studies concluded that air–sea buoyancy loss and Ekman advection of buoyancy are two dominant factors in forming the deep mixed layers in the Southern Ocean (Rintoul and England 2002; Sallée et al. 2006; Dong et al. 2007, 2008; Downes et al. 2011; Holte et al. 2012). Buoyancy forcing in particular has been considered to play the primary role (Sallée et al. 2010; Hogg 2010). Using scaling analysis, Sallée et al. (2010) concluded for the mixed layer north of the ACC that buoyancy forcing is an order of magnitude greater than wind forcing. The Langmuir turbulence associated with the Stokes drift was found to deepen the mixed layer in the Southern Ocean, particularly during the austral summer (Belcher et al. 2012; Wenegrat and McPhaden 2016; Qing Li et al. 2016, 2017).

The results from recent modeling studies suggest that the mixed layer is appreciably more accurate in eddy-permitting or eddy-resolving ocean model simulations than in simulations at coarser resolution where the dynamics is laminar, suggesting that mesoscale eddies also play an important role in determining vertical stratification in the southern Indian Ocean (Lee et al. 2011) and seasonal mixed layer dynamics (Schiller and Ridgway 2013). In addition, the mesoscale eddy fluxes remain important in the mixed layer (Schmitz 1996; Treguier et al. 1997; Phillips and Rintoul 2000; Weller et al. 2004; Kuo et al. 2005). Initial stratification prior to the cooling season may also be important (Dong et al. 2008).

There are two major regions of deep mixed layers in the Southern Ocean, one in the Indo–western Pacific Southern Ocean (IWPSO) and the other in the eastern Pacific Southern Ocean (EPSO; see, e.g., Dong et al. 2008; Cerovečki et al. 2013). The June mixed layer depth (MLD; color) climatology, shown in Fig. 1, reveals that these two regions show different meridional scales and relationships with the ACC jets (contours). In the EPSO, the MLD is relatively broad (more than ~10° latitude wide) and encompasses the entire SAF. In contrast, the MLD in the IWPSO is extremely narrow (less than ~2° latitude) and is located immediately north of the SAF. This stark difference in the MLD structure suggests that there are also differences in driving mechanisms between the two regions. Indeed, Dong et al. (2008) showed that heat loss to the atmosphere is responsible for the development of the mixed layers in the EPSO region, while horizontal Ekman advection is more important in the IWPSO region. However, the meridional scale of the zonal wind stress curl is essentially that of the atmospheric midlatitude westerlies, which is much greater than ~2° latitude. Therefore, at least during the initial development stage of the IWPSO mixed layer (Fig. 1), processes associated with the SAF itself could play an important role. This possibility is in line with the conclusion of the aforementioned eddy-resolving model studies (Lee et al. 2011; Schiller and Ridgway 2013; Morrison et al. 2013) because the IWPSO ACC jets are driven by baroclinic eddy fluxes (Qian Li et al. 2016). In this study, we will present analysis results that support this possibility.

Fig. 1.
Fig. 1.

June MLD (color; m) simulated by the POP model. Contours show Gaussian-boxcar-smoothed (2.5° × 2.5°) and depth-averaged (above 1 km) zonal current speed (cm s−1) that exceeds 6 cm s−1 (hatched) and 11 cm s−1 (thick) with an interval of 5 cm s−1. The black box indicates the domain defined as the Indo–western Pacific Southern Ocean (30°–60°S, 50°–190°E) in this study.

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

The remainder of this paper is organized as follows. In section 2, we will describe the model simulation used in this study. The results will be presented in section 3, and the discussion and conclusions will be followed in section 4.

2. Model description and methodology

a. The eddy-resolving ocean model

This study uses the data output from an eddy-resolving simulation of the global ocean circulation carried out with the Los Alamos Parallel Ocean Program (POP) model (Smith et al. 1992). A Mercator grid is employed with a latitudinal grid spacing of 1/10° cos(ϕ), where ϕ is latitude. Over the Southern Ocean, the corresponding horizontal resolution ranges between 4 and 9 km. In the vertical, the POP model has tripole grid and 42 levels with thickness ranging from 5 m at the surface to 250 m at 5500-m depth. The POP model run analyzed here was restarted with an initial spinup period of 15 years between 1979 and 1993, in which the model was found to have reached a quasi-equilibrium state in the upper ocean (Maltrud et al. 2010). The results described in this paper are based on daily output from the subsequent period between 1994 and 2007 of the same model simulation. As such, the model used in this study is identical to that of Maltrud et al. (2010), except that our experiment was driven by synoptic forcing rather than by a repeating monthly climatology.

The atmospheric state was based on the National Centers for Environmental Prediction–National Center for Atmospheric Research (NCEP–NCAR) reanalysis dataset (Kalnay et al. 1996), but these data were adjusted based on comparisons with observations, especially in high latitudes (Large and Yeager 2009). The ocean was forced by the interannually varying atmospheric fluxes from the Co-ordinated Ocean–Ice Reference Experiments version 2.0 (CORE.v2) forcing dataset (Large and Yeager 2009). The shortwave radiation and precipitation fields were taken from the CORE.v2 forcing dataset. Longwave radiation, sensible heat flux, evaporation, and wind stress were calculated online using bulk formulae in terms of the near-surface atmospheric and oceanic states (Large and Yeager 2009). Latent heat flux was derived from evaporation. The online-calculated air–sea heat fluxes output, available for us to use in this study, is from the period of 1995–2003.

The parameterizations in the POP model are described in Maltrud et al. (2010). The K-profile parameterization (KPP; Large et al. 1994) was implemented in the calculation of the vertical tracer diffusion coefficients. The values range from 10−5 m2 s−1 near the surface to 10−4 m2 s−1 at depths. A much larger diffusion coefficient of 0.1 m2 s−1 was used to simulate convection.

b. Definition of the mixed layer depth

In our study region of the IWPSO, because the temperature and salinity variability is density-compensating (Rintoul and England 2002; Sallée et al. 2006), a buoyancy/density-based MLD definition is more appropriate than a temperature-based definition. The POP model–simulated MLD is defined as the shallowest depth at which the local buoyancy gradient equals the maximum buoyancy gradient between the surface and any discrete depth within that water column (Large et al. 1997). This MLD definition was found to be useful in describing a boundary layer and associated transition layer (Ferrari et al. 2008; Danabasoglu et al. 2008).

To test the sensitivity of our main finding to the definition of the MLD, we also consider another density-based definition where the base of the mixed layer is defined as the depth at which the potential density in the layer changes by 0.03 kg m−3 relative to the ocean surface density (Δσ = 0.03 kg m−3). In addition, the Argo-float-based MLD is also examined for comparison. As will be shown, the key MLD structure associated with the SAF (Fig. 1) is evident in all of these different definitions.

c. Depth-dependent buoyancy frequency budget analysis

The temporal evolution of stratification is examined in this study by considering the depth-dependent potential temperature and salinity budgets that are prognostic in the POP model. These two budget equations are expressed as
e1
e2
with boundary conditions
eq1
eq2
where θ is the potential temperature, S is the salinity, q is the penetrative shortwave flux (positive into the ocean) distributed vertically over several upper model layers, ρ is the density, cw is the specific heat capacity of the seawater, u = (u, υ, w) is the three-dimensional current velocity, and ∇ = (∂/∂x, ∂/∂y, ∂/∂z). The potential temperature diffusion and salinity diffusion terms include horizontal and vertical components, where κH and κZ are the horizontal and vertical diffusivities, respectively; ∇H = (∂/∂x, ∂/∂y); an overbar denotes a time average; and the subscript 0 indicates the variable at the ocean surface. Here, the vertical turbulent fluxes of potential temperature and salinity are applied with a nonlocal KPP parameterization (Large et al. 1994), in which γθ and γS are the nonlocal KPP transport of temperature and salinity respectively. Parameter Qnet is the net air–sea heat flux (from the atmosphere to the ocean is defined as positive), which is composed of four components:
e3
Parameter ε0 is the net surface freshwater flux (from the ocean to the atmosphere is defined as being positive), including evaporation, precipitation, river runoff, and ice formation or melting.
Because of the aforementioned density-compensating property in the Southern Ocean, we further derive an equation of the depth-dependent buoyancy budget by combining (1) and (2). Here, the buoyancy is defined as b ≡ −g(δσ/ρ0), where σ is the potential density, g is the gravity, and ρ0 is a reference density (1027 kg m−3). Then, the buoyancy budget can be expressed as
e4
where αθ ≡ ∂σ/∂θ is the thermal expansion coefficient, βS ≡ ∂σ/∂S is the haline contraction coefficient, and is the buoyancy variation induced by the diabatic heating from the penetrative shortwave flux into the ocean. Parameter is the buoyancy diffusion term, and cab is the cabbeling term due to the nonlinear effect. Our main focus is the advection term, which is the first term on the rhs of (4). Thus, the last three terms on the rhs of (4) are regarded as a residual term, and we express them collectively as . To examine the vertical stratification, we finally compute the buoyancy frequency squared (N2) by taking the vertical derivative of buoyancy:
e5
By combining (4) and (5), the equation of buoyancy frequency budget can be expressed as
e6
where is the contribution from the residual term in (4). In the POP model simulation, the mesoscale eddies are resolved, but submesoscale processes need to be parameterized. The first term on the rhs of (6) expresses the temporal variation of buoyancy frequency caused by the resolved oceanic advection of buoyancy. The second term on the rhs of (6) is a residual term that can be attributed to the parameterized subgrid-scale diffusion of buoyancy, cabbeling, and air–sea exchange. This residual term is estimated by subtracting the resolved advection term from the local rate of change in N2. Because these latter two terms are diagnosed using daily output rather than output at every time step, the residual term is contaminated by sampling error. Nevertheless, as will be shown, the estimated structure of the residual term is consistent with what is expected of the subgrid-scale parameterization and the surface fluxes.

3. Results

a. Evolution of mixed layer depth in the Indo–western Pacific Southern Ocean

We first examine the monthly evolution of the MLD distribution and assess the key features. Figure 2 displays the May–October MLD and depth-averaged (above 1 km) zonal current speed fields in the IWPSO. Because raw zonal current field is complex, for the purpose of display in Fig. 2, we applied a spatial Gaussian boxcar (2.5° × 2.5°) smoothing (e.g., as in Williams et al. 2007) to the raw data. From May to October, the MLD exhibits a marked seasonal march: the MLD begins to deepen in June (Fig. 2b), reaching its maximum in September (Figs. 2c–e), and then shoals during October (Fig. 2f). The deep MLD develops along the northern flank of the SAF with a meridional width of only ~2° latitude. In particular, the deepest MLD occurs in a small sector of 130°–142°E (the red box shown in Fig. 2), where the jets are especially zonal and persistent. Therefore, we will examine in detail the development of the deep mixed layer in this sector.

Fig. 2.
Fig. 2.

Monthly MLD (color; m) from the POP model simulation for (a)–(f) May–October in the Indo–western Pacific Southern Ocean (30°–60°S, 50°–190°E). Contours show Gaussian-boxcar-smoothed (2.5° × 2.5°) and depth-averaged (above 1 km) zonal current speed (cm s−1) that exceeds 6 cm s−1 (hatched) and 11 cm s−1 (thick) with an interval of 5 cm s−1. The red box indicates the sector (40°–55°S, 130°–142°E) where the current is especially zonal and persistent and the MLD is deepest.

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

There is evidence that the aforementioned MLD wedge–jet relationship also exists in nature. Figure 3 shows the Argo-float-based MLD estimated using the density threshold method (Holte et al. 2016). Superimposed in this figure is the zonal current speed from the Southern Ocean State Estimate (SOSE), which is considered as an optimal fit to Southern Ocean observations (Mazloff et al. 2010). For display purposes, the zonal current speed field from the SOSE was also subjected to the Gaussian boxcar (2.5° × 2.5°) smoothing. It can be seen that the thickest MLD from Argo occurs along the northern flank of the strong, localized jets. The observed MLD wedge–jet relationship is consistent with that simulated by the POP model (Fig. 2), although it is evident in several different sectors among 90°–150°E, and the observed MLD is relatively shallower than that from the POP. Therefore, we are encouraged to examine the POP model simulation to investigate the mechanism of the MLD development.

Fig. 3.
Fig. 3.

Monthly MLD (color; m) based on the density criterion (Δσ = 0.03 kg m−3) from the Argo float data for (a)–(f) May–October in the Indo–western Pacific Southern Ocean (50°–190°E, 30°–60°S). Contours show Gaussian-boxcar-smoothed (2.5° × 2.5°) and depth-averaged (above 1 km) zonal current speed (cm s−1) that exceeds 6 cm s−1 (hatched) and 11 cm s−1 (thick) with an interval of 2.5 cm s−1. The velocity data are from the SOSE.

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

The vertical cross section of the MLD and the zonal current speed reveal additional details that are not evident in the horizontal distribution. Figures 4a–d show the annual evolution of the zonal current speed and MLD zonally averaged over the sector (130°–142°E). The primary jet centered at the latitude of ~48.7°S is the SAF. The SAF core is indicated by the vertical dashed line. It can be seen that the mixed layer is very shallow during the austral summer and fall (Figs. 4a,b) and that it starts to develop in June (Fig. 4c) and persists until October (Fig. 4d). As one expects from the horizontal structure (Figs. 2a–e), during this period of austral winter and spring (Figs. 4c,d), the deepest MLD occurs on the equatorward flank of the SAF core. The meridional MLD profile further reveals that during June and July the deepening first occurs very close to the north of the SAF core, and over a limited meridional extent of only ~2° latitude. This initial mixed layer is much narrower than the fully developed mixed layer (~6° latitude wide) that occurs during August and September.

Fig. 4.
Fig. 4.

(a)–(d) Vertical cross sections of sector-zonal-mean (130°–142°E) zonal current speed (color; cm s−1) and MLD (lines; m) from POP for austral summer (DJF) through austral spring (SON). (e),(f) As in (c),(d) but for zonal current speed from SOSE and MLD from Argo. The vertical black dashed (dashed–dotted) line indicates the latitude of the SAF core from POP (SOSE).

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

The austral winter and spring MLD characteristics just described are also evident in the Argo-float-based MLD and SOSE-based zonal current speed (Figs. 4e,f). The SOSE-based SAF core is indicated by the vertical dash–dotted line. Once again, the Argo-float-based MLD also exhibits a wedge-like shape only ~2° latitude north of the SAF core. Compared with the POP-simulated SAF, the SOSE-based SAF is relatively weak and located farther to the south. In addition, there is a slight meridional shift with seasons in the SOSE-based SAF, which is centered at ~49.7°S in austral winter and ~50.4°S in austral spring. In spite of these differences, the observed wedge is even more persistent than that in the POP and lasts from June to November. This finding shows that the MLD wedge–jet relationship in nature and is neither an artifact of the POP model simulation nor sensitive to the definition of the MLD.

b. Evolution of the stratification

To further explore key processes that contribute to the MLD development, we first diagnose the vertical stratification in our study sector. By definition, a mixed layer is characterized by a weak stratification, and as a measure of the stratification we use N2 in (2). In general, the climatological N2 is positive, meaning that the water column is stable. The vertical cross sections of the monthly N2 are presented in Fig. 5. During the warming period of November–February (NDJF), the surface layer becomes increasingly stable, resulting in a thin but strongly stratified layer (above 100 m). This layer lies on top of the less stratified water column (between 100 and 500 m). When the surface starts to cool during the cooling period from May through August, the surface layer becomes less stable. Figure 5 shows that a break of this thin stable layer first appears in June immediately north of the SAF core and over a limited meridional extent of only ~2° latitude. This location coincides with the MLD wedge that was described earlier. Therefore, it is likely that the process that causes the destratification at this particular location is also responsible for the formation of the MLD wedge.

Fig. 5.
Fig. 5.

Monthly vertical cross sections of sector-zonal-mean (130°–142°E) buoyancy frequency N2 (color; 10−5 s−2). The white solid line shows the MLD (m) for each month. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

To investigate why the destratification first occurs immediately north of the SAF, we next examine the vertical distributions of the monthly N2 tendency (∂N2/∂t). As expected, Fig. 6 shows that the N2 tendency undergoes a marked seasonal change with negative values within the deepening mixed layer. The May N2 tendency is uniformly negative above ~125 m, presumably due to surface cooling. The surface layer N2 tendency becomes positive during September and October as spring warming ensues. During the intermediate months, Fig. 6 reveals a rich spatial structure, especially on the equatorward flank of the SAF core. During June and July, the strongest negative N2 tendency occurs in a narrow latitudinal band (~46°–48°S) immediately north of the SAF core. This destratification during the early winter is crucial in the development of the deep mixed layer because a weakly stratified water column is conducive to vertical mixing. In fact, it is noteworthy that the narrow region of negative N2 tendency coincides with the narrow wedge of a developing mixed layer. What drives the destratification in the first place? To answer this question, we next examine destratification caused by buoyancy advection.

Fig. 6.
Fig. 6.

Monthly vertical cross sections of sector-zonal-mean (130°–142°E) tendency of buoyancy frequency ∂N2/∂t (color; 10−6 s−2 month−1). The white solid line shows the MLD (m) for each month. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

c. Destratification and preconditioning by a jet-scale overturning circulation

Our hypothesis is that the initial destratification is driven by the jet-scale overturning circulation (JSOC), which was first described by Qian Li et al. (2016). The concept of the JSOC is illustrated in Fig. 7a: at least in the present POP model simulation, the Indo–western Pacific ACC jets are driven by eddy momentum flux convergence (Qian Li et al. 2016). As indicated by the red arrows in Fig. 7a, the meridional flux of zonal momentum by the eddies converges toward the core latitude of the SAF. This convergence increases with height and thus exerts an eastward acceleration that increases with height. In this manner, the action of the eddy momentum flux generates a positive vertical shear of the zonal current, which helps to account for the jet structure. Since the leading force balances in the ocean are hydrostatic in the vertical direction and geostrophic in the horizontal direction (which combine to form the thermal wind balance), the density field must adjust to the vertical shear of the zonal current. The thermally indirect1 JSOC can be understood as being a manifestation of this adjustment process. In the ocean, this indirect circulation brings down relatively more buoyant water from the surface on the equatorward jet flank and brings up heavier water on the poleward jet flank. These actions result in an increase in the meridional buoyancy gradient, which indeed achieves thermal wind balance with the positive vertical shear of the zonal current. Unlike the buoyancy-flux-driven portion of the Eulerian Ferrel cell (Kuo 1956; Pfeffer 1981), the momentum-driven portion of the indirect cell remains even in the transformed Eulerian mean (TEM) circulation (Robinson 2006).

Fig. 7.
Fig. 7.

(a) Schematic diagram of eddy momentum flux–driven jet (), JSOC (→), and JSOC-driven equatorward potential density σ (contours) gradient north of the SAF. Red arrows (→) denote the eddy momentum flux. (b) Vertical cross section of November–April mean potential density (contours; 10−2 kg m−3) and the change in potential density (color; kg m−3) caused by the sector-zonal-mean (130°–142°E) vertical advection during the period of November–April. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S). The white box indicates the region of the diagram shown in (a).

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

The vertical buoyancy advection induced by the JSOC can be expressed as , where the bracket denotes the sector zonal mean and the overbar represents the monthly mean. The N2 tendency contributed by this vertical buoyancy advection, , was obtained by computing . The result is shown in Fig. 8. Throughout the entire year, the is negative (blue shading) over a narrow region immediately north of the SAF core, indicating that the JSOC-driven downward motion acts to destratify the water masses there. Because this JSOC-driven destratification is present during the warm season, we can conclude that in this narrow latitudinal band the JSOC acts to precondition the MLD development.

Fig. 8.
Fig. 8.

As in Fig. 6, but for the tendency of buoyancy frequency ∂N2/∂t contributed from the JSOC.

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

The effect of the JSOC is noticeable in the climatological density field. As shown with the contours in Fig. 7b, the climatological potential density has a local minimum at the latitudes of the JSOC’s descending branch. To evaluate if the local minimum is caused by the JSOC, we computed the change in potential density caused by the vertical advection. An examination of potential density cross sections indicates that the equatorward density gradient develops during the warm months, starting in November and reaching its maximum in April. Therefore, the change in potential density caused by vertical advection can be estimated by computing , where Δt is the time period of 1 November to 30 April, the bracket denotes the sector zonal mean, and the overbar represents the time mean over the same period. The result is shown in Fig. 7b. It can be seen that the center of negative density advection (blue shading) by the JSOC occurs at ~48.5°S and coincides with the local minimum in the climatological potential density (the contour). For instance, at (~48.5°S, 300 m), the potential density decline due to this downward advection is about 5 × 10−2 kg m−3, and this compares reasonably well with the difference in potential density between 48.5° and ~47.5°S, which is just north of the downward branch of the JSOC. This negative density advection results in a counterintuitive equatorward density gradient north of the SAF core. The result of this analysis indicates that the JSOC pumps down the buoyant surface water to depths during the warm season; this action destratifies the water column where the JSOC’s sinking branch is located, but at the same time, it also generates a local baroclinic zone to its north with its equatorward potential density gradient. This baroclinic zone is favorable to generating eddy buoyancy flux, and its effect on additional destratification will be explored in the next section.

d. Destratification by the eddy buoyancy flux

In tune with the earlier studies by Lee et al. (2011) and Schiller and Ridgway (2013), it turns out that eddy buoyancy flux also contributes to the destratification. Figure 9 presents the contribution of daily eddy buoyancy flux to the monthly N2 tendency averaged over the study sector:
eq3
where the bracket and asterisk denote the sector zonal mean and the departure from the sector zonal mean, respectively, and the overbar and prime present the monthly mean and the departure from the annual mean, respectively. Superimposed in this figure is the potential density profile over the same sector. During April–June there is eddy buoyancy flux that acts to destratify the water column over a narrow latitudinal band immediately north of the SAF core. Since the equatorward meridional potential density gradient is strongest during the same time period and at the same location, we suspect that the destratification is associated with baroclinic eddies that arise from the JSOC-driven equatorward density gradient. It is possible that the equatorward density gradient is also contributed by the processes that form the SAMW to the north of the SAF. Starting in July, the eddy flux effect in the key region becomes complex, but by July the mixed layer has already matured.
Fig. 9.
Fig. 9.

Monthly vertical cross sections of sector-zonal-mean (130°–142°E) tendency of buoyancy frequency ∂N2/∂t (color; 10−6 s−2 mon−1) contributed from the eddy flux and vertical cross sections of potential density (contours; kg m−3) with an interval of 0.05 kg m−3. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

e. Destratification by other processes

In addition to the buoyancy advection by the JSOC and the eddies, other processes, such as air–sea exchange of heat and salinity and subgrid-scale processes, could have contributed to the preconditioning. As was described in section 2c, the sum of these terms was estimated using (6) by subtracting the resolved advection term from the left-hand side. The contribution to the N2 tendency by this residual term is shown in Fig. 10. During April–September, the destratification caused by the residual term is significant mostly within the mixed layer, and it is noticeably smaller underneath the mixed layer. Therefore, we conclude that the processes represented by the residual term are not responsible for the preconditioning.

Fig. 10.
Fig. 10.

As in Fig. 6, but for the tendency of buoyancy frequency ∂N2/∂t contributed from the residual.

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

Since previous studies have shown that net air–sea heat flux and Ekman advection are two important drivers of MLD variability in the Southern Ocean, we further explore their roles.

1) Role of air–sea fluxes

During austral fall and winter, cold, dense water forms near the surface. The resulting destratification deepens the mixed layer by driving convection (Sallée et al. 2006; Dong et al. 2008; Badin et al. 2013). We examined the online-computed Qnet as expressed in (3) and ε0 from our POP simulation. Consistent with the aforementioned density compensation in the IWPSO region (Rintoul and England 2002), the buoyancy forcing driven by Qnet tends to negate that by ε0. Because the latter plays a relatively negligible role in our study sector (not shown), we will only consider the role of Qnet in the subsequent analysis.

Figure 11 shows the monthly MLD and the online-computed Qnet during the austral winter of June–August (JJA). As was mentioned earlier, the MLD wedge develops immediately north of the SAF core (Fig. 11a). In the winter season, the ocean surface loses heat to the atmosphere. Accordingly, Qnet is negative and acts to weaken the stratification. Interestingly, the most negative Qnet in JJA occurs right on the MLD wedge (Fig. 11b). This wedge-like Qnet profile is persistent throughout the entire JJA, although the meridional gradient of Qnet in a broad latitudinal band (~42°–48°S) slightly increases over time. We infer that this wedge-like Qnet structure is caused by mesoscale ocean dynamics simulated by the POP model because it barely exists in the Qnet profiles of the NCEP–NCAR reanalysis (Fig. 11c). As shown in Fig. 11c, with the coarser resolution (1.875°), the meridional gradient of Qnet in the same latitudinal band is nearly zero during June and July, but it increases significantly in August.

Fig. 11.
Fig. 11.

(a) Monthly MLD (m) and net air–sea heat flux (Qnet; W m−2) from the (b) POP and (c) NCEP–NCAR reanalysis for the sector (130°–142°E) during austral winter (JJA). Positive fluxes in (b) and (c) are directed into the ocean. The sector-zonal-mean (130°–142°E) and JJA-mean zonal current speed (cm s−1; gray line) at 918 m are overlaid to show the location of the SAF. The vertical black dashed line represents the latitude of the SAF core (~48.7°S). The location of the MLD wedge is highlighted by the pink shading.

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

To explore the relationship between Qnet and the upper-ocean structure, we examine the monthly Qnet (Fig. 12; top) and vertical distributions of potential temperature and vertical velocity (Fig. 12; bottom) for the study sector. At depths between ~100 and 500 m, there is a narrow warm tongue isolated within only ~2° latitude on the equatorward flank of the SAF core throughout the entire year. This warm tongue is collocated with the downwelling branch of the JSOC, which transports down warm buoyant water during the warming period of NDJF (Figs. 12a,b,k,l). As the fall season begins (starting in March) the surface cools and the temperature in the top layer (0–100 m) becomes lower than that of the water below. Therefore, the vertical potential temperature gradient can become negative (temperature inversion). This causes heat to flux upward to the surface layer where it is lost to the atmosphere. Figure 12 shows that within this narrow latitudinal band, due to the JSOC-driven warm tongue, the inversion is stronger than the surrounding regions (Figs. 12c–e). We computed the term in (1), and the result indicates that the upward heat flux, which first appears in March, is indeed strongest where the warm tongue and the Qnet wedge are located (not shown). From these results, we conclude that the JSOC-driven warm tongue causes the negative Qnet wedge to develop, which in turn enables the mixed layer to deepen where the downwelling branch of the JSOC is located. At the same time, the temperature inversion persists (Fig. 12) and contributes to the weak stratification in the top layer (Fig. 6), further preconditioning for the deepening of the mixed layer and upward heat flux. This interpretation is also supported by the fact that the negative Qnet wedge peaks during the winter season (Figs. 12f–h) when the mixed layer is already fully developed.

Fig. 12.
Fig. 12.

(top half) Monthly net air–sea heat flux (Qnet;W m−2) from the POP for the sector (130°–142°E). (bottom half) Vertical cross sections of sector-zonal-mean (130°–142°E) monthly potential temperature (color; °C) and vertical velocity (vectors; 10−3 cm s−1). Green (white) vectors indicate the downwelling (upwelling). The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

It is worth noting that in the observational study of mixed layer heat balance, Dong et al. (2007) found that there are substantial imbalances north of the SAF. They attributed these imbalances to uncertainty in Qnet. In that study, NCEP–NCAR reanalysis Qnet was used. As we showed here, the finescale Qnet wedge (Fig. 11b), caused by the eddy-driven jet-scale circulation, is almost absent in the coarser resolution Qnet (Fig. 11c). The Qnet wedge occurs north of the SAF core where the heat budget imbalance is severe. Therefore, it is probable that the heat imbalance reported by Dong et al. (2007) is caused by the fact that the mesoscale JSOC effect is unaccounted for in the Qnet data that they used.

2) Role of Ekman transport

The wind stress in the Indo–western Pacific Southern Ocean is much stronger than that in the eastern Pacific Southern Ocean (EPSO), where the MLD is also very deep (Dong et al. 2008). Eastward wind stress drives equatorward Ekman transport, which advects cold surface water from high to low latitudes. This advection can help destratify the water column. Therefore, advection has been regarded as the key player in the development of the deep IWPSO mixed layers (Ribbe 1999; Rintoul and England 2002; Dong et al. 2008). Therefore, it is possible that the Ekman transport may also contribute to the development of the MLD wedge. To test this possibility, we computed Ekman meridional advection of potential density (), where the bracket denotes the sector zonal mean, the overbar represents the monthly mean, and υe is the meridional component of the Ekman velocity, which is defined as −τx/(ρ0fhm). For the zonal wind stress τx we used online-computed values from the POP simulation and hm was taken to be the mixed layer depth. The Ekman meridional velocity is overall equatorward, as was expected, but it is weak in the vicinity of the SAF (Fig. 13a). The meridional gradient of potential density is positive farther north of the SAF core (Fig. 13b). As a result, within the narrow latitude band (~46°–49°S), there is negative Ekman meridional advection of potential density that enhances the local stratification, thus impeding the deepening of the mixed layer (Fig. 13c). Therefore, we conclude that the MLD wedge development cannot be attributed to Ekman transport. Farther north of the SAF, between ~40° and 46°S, the Ekman advection is positive and thus can help explain the development of the deep mixed layer.

Fig. 13.
Fig. 13.

Monthly (a) Ekman meridional velocity (cm s−1), (b) depth-averaged (above 100 m) potential density (kg m−3), and (c) Ekman meridional potential density advection (10−2 kg m−3 month−1) for the sector (130°–142°E) during austral winter (JJA). The sector-zonal-mean (130°–142°E) and JJA-mean zonal current speed (cm s−1; gray line) at 918 m is overlaid to show the location of the SAF. The vertical thick dashed line indicates the latitude of the SAF core (~48.7°S).

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

4. Summary and discussion

We examined the POP-simulated MLD in the Indo–western Pacific Southern Ocean, where the ACC jets are relatively strong, zonal, and persistent. In our study sector (130°–142°E) where the vertical structures and dynamics of the jets are examined (Qian Li et al. 2016), the mixed layer in early winter is remarkably narrow and occurs immediately north of the SAF, the primary and northernmost ACC jet. The meridional extent of the early winter mixed layer wedge is only about 2° latitude wide. Our analysis suggests that initially the weak stratification associated with the narrow mixed layer is due to the downward transport of buoyant surface water driven by the descending branch of the JSOC. This buoyancy transport also generates equatorward potential density gradient north of the SAF. It turns out that eddy buoyancy flux contributes to the destratification in the same region. Figure 14 shows monthly potential density and meridional eddy buoyancy flux , where the bracket and asterisk denote the sector (130°–142°E) zonal mean and the departure from the sector zonal mean respectively, and the overbar represents the monthly mean. It shows that, corresponding to the equatorward potential density gradient north of the SAF core (~46°–48°S), there is equatorward meridional eddy buoyancy flux throughout the entire year. This meridional eddy buoyancy flux acts to decrease this strong gradient, which is developed by the downward branch of the JSOC. During austral winter (December–May), when the potential density gradient is well established, the meridional eddy buoyancy flux is relatively stronger, whereas it is much weaker during austral summer when the potential density gradient is weaker. Given this evidence, we conclude that the eddy flux arises from baroclinic instability associated with the equatorward potential density gradient. This physical picture is summarized schematically in Fig. 15.

Fig. 14.
Fig. 14.

Monthly vertical cross sections of sector-zonal-mean (130°–142°E) monthly meridional eddy buoyancy flux (color; 10−1 cm2 s−3) and potential density (contours; kg m−3) with an interval of 0.05 kg m−3. For the eddy buoyancy flux, positive (negative) values indicate equatorward (poleward) transport. The vertical black dashed line indicates the latitude of the SAF core (~48.7°S).

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

Fig. 15.
Fig. 15.

Schematic diagram of the processes that contribute to the Southern Ocean mixed layer destratification during austral winter (JJA): net air–sea heat flux (Qnet), meridional Ekman advection, eddy buoyancy flux, and downward branch of the JSOC.

Citation: Journal of Physical Oceanography 47, 11; 10.1175/JPO-D-17-0006.1

Our examination of the monthly evolution of the net air–sea heat flux reveals that while it contributes to the deepening of the mixed layer, the net heat flux itself is substantially modified by the downwelling action of the JSOC. Consistent with previous studies, Ekman advection indeed contributes to the formation of the deep mixed layer. However, its contribution occurs farther to the north of the SAF and hence does not help to explain the formation of the deepest mixed layer wedge immediately north of the SAF. In nature, it is plausible that small-scale processes such as the Stokes drift could contribute to the destratification in our study area (Wenegrat and McPhaden 2016). However, because Stokes drift is not explicitly simulated by the POP model, at least in the model, the key MLD structure considered in this study cannot be attributed to the Stokes drift effect.

We conclude that in the region where the mixed layer is deepest, SAF dynamics is an integral part of the mixed layer formation process because the thermally indirect JSOC is a response to eddy momentum flux convergence that drives the SAF in this particular sector (Qian Li et al. 2016). Since the eddy momentum flux–driven JSOC cannot be simulated at coarser resolution, we believe that the findings of this study provide additional insight into why mixed layers are appreciably more accurate in an eddy-permitting or eddy-resolving ocean model simulation than in simulations at coarser resolution.

Acknowledgments

We thank Alexa Griesel for providing the detailed information about the POP model experiment, and we also thank Matthew Mazloff for his comments on using the SOSE. Q.L. and S.L. were supported by the National Science Foundation under Grant ATM-1139970. An allocation of advanced computing resources OCE960914 provided by the National Science Foundation was instrumental in performing this research. The computations were performed on Kraken at the National Institute for Computational Sciences (http://www.nics.tennessee.edu/).

REFERENCES

  • American Meteorological Society, 2017: Thermally indirect. Glossary of Meteorology, http://glossary.ametsoc.org/wiki/thermally_indirect.

  • Badin, G., R. G. Williams, Z. Jing, and L. Wu, 2013: Water mass transformations in the Southern Ocean diagnosed from observations: Contrasting effects of air–sea fluxes and diapycnal mixing. J. Phys. Oceanogr., 43, 14721484, doi:10.1175/JPO-D-12-0216.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Belcher, S. E., and et al. , 2012: A global perspective on Langmuir turbulence in the ocean surface boundary layer. Geophys. Res. Lett., 39, L18605, doi:10.1029/2012GL052932.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Cerovečki, I., L. D. Talley, M. R. Mazloff, and G. Maze, 2013: Subantarctic Mode Water formation, destruction, and export in the eddy-permitting Southern Ocean state estimate. J. Phys. Oceanogr., 43, 14851511, doi:10.1175/JPO-D-12-0121.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Danabasoglu, G., R. Ferrari, and J. C. McWilliams, 2008: Sensitivity of an ocean general circulation model to a parameterization of near-surface eddy fluxes. J. Climate, 21, 11921208, doi:10.1175/2007JCLI1508.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dong, S., S. T. Gille, and J. Sprintall, 2007: An assessment of the Southern Ocean mixed layer heat budget. J. Climate, 20, 44254442, doi:10.1175/JCLI4259.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Dong, S., J. Sprintall, S. T. Gille, and L. Talley, 2008: Southern Ocean mixed-layer depth from Argo float profiles. J. Geophys. Res., 113, C06013, doi:10.1029/2006JC004051.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Downes, S. M., A. S. Budnick, J. L. Sarmiento, and R. Farneti, 2011: Impacts of wind stress on the Antarctic Circumpolar Current fronts and associated subduction. Geophys. Res. Lett., 38, L11605, doi:10.1029/2011GL047668.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ferrari, R., J. C. McWilliams, V. M. Canuto, and M. Dubovikov, 2008: Parameterization of eddy fluxes near oceanic boundaries. J. Climate, 21, 27702789, doi:10.1175/2007JCLI1510.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hanawa, K., and L. D. Talley, 2001: Mode waters. Ocean Circulation and Climate, International Geophysical Series, Vol. 77, Academic Press, 373–386.

    • Crossref
    • Export Citation
  • Herraiz-Borreguero, L., and S. R. Rintoul, 2011: Subantarctic mode water: Distribution and circulation. Ocean Dyn., 61, 103126, doi:10.1007/s10236-010-0352-9.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Hogg, A. McC., 2010: An Antarctic Circumpolar Current driven by surface buoyancy forcing. Geophys. Res. Lett., 37, L23601, doi:10.1029/2010GL044777.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holte, J., and L. Talley, 2009: A new algorithm for finding mixed layer depths with applications to Argo data and Subantarctic Mode Water formation. J. Atmos. Oceanic Technol., 26, 19201939, doi:10.1175/2009JTECHO543.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holte, J., L. Talley, T. K. Chereskin, and B. M. Sloyan, 2012: The role of air–sea fluxes in Subantarctic Mode Water formation. J. Geophys. Res., 117, C03040, doi:10.1029/2011JC007798.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Holte, J., J. Gilson, L. Talley, and D. Roemmich, 2016: Argo mixed layers. Scripps Institution of Oceanography, http://mixedlayer.ucsd.edu.

  • Jones, D. C., A. J. S. Meijers, E. Shuckburgh, J. B. Sallée, P. Haynes, E. K. McAufield, and M. R. Mazloff, 2016: How does Subantarctic Mode Water ventilate the Southern Hemisphere subtropics? J. Geophys. Res. Oceans, 121, 65586582, doi:10.1002/2016JC011680.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kalnay, E., and et al. , 1996: The NCEP/NCAR 40-Year Reanalysis Project. Bull. Amer. Meteor. Soc., 77, 437471, doi:10.1175/1520-0477(1996)077<0437:TNYRP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kuo, A., R. A. Plumb, and J. Marshall, 2005: Transformed Eulerian-mean theory. Part II: Potential vorticity homogenization and the equilibrium of a wind- and buoyancy-driven zonal flow. J. Phys. Oceanogr., 35, 175187, doi:10.1175/JPO-2670.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Kuo, H.-L., 1956: Forced and free meridional circulations in the atmosphere. J. Meteor., 13, 561568, doi:10.1175/1520-0469(1956)013<0561:FAFMCI>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., and S. G. Yeager, 2009: The global climatology of an interannually varying air–sea flux data set. Climate Dyn., 33, 341364, doi:10.1007/s00382-008-0441-3.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., J. C. McWilliams, and S. C. Doney, 1994: Oceanic vertical mixing: A review and a model with a nonlocal boundary layer parameterization. Rev. Geophys., 32, 363403, doi:10.1029/94RG01872.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Large, W. G., G. Danabasoglu, S. C. Doney, and J. C. McWilliams, 1997: Sensitivity to surface forcing and boundary layer mixing in a global ocean model: Annual-mean climatology. J. Phys. Oceanogr., 27, 24182447, doi:10.1175/1520-0485(1997)027<2418:STSFAB>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Lee, M. M., A. J. G. Nurser, I. Stevens, and J. B. Sallée, 2011: Subduction over the Southern Indian Ocean in a high-resolution atmosphere–ocean coupled model. J. Climate, 24, 38303849, doi:10.1175/2011JCLI3888.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Qian, S. Lee, and A. Griesel, 2016: Eddy fluxes and jet-scale overturning circulations in the Indo–western Pacific Southern Ocean. J. Phys. Oceanogr., 46, 29432959, doi:10.1175/JPO-D-15-0241.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Qing, A. Webb, B. Fox-Kemper, A. Craig, G. Danabasoglu, W. G. Large, and M. Vertenstein, 2016: Langmuir mixing effects on global climate: WAVEWATCH III in CESM. Ocean Modell., 103, 145160, doi:10.1016/j.ocemod.2015.07.020.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Li, Qing, B. Fox-Kemper, ∅. Breivik, and A. Webb, 2017: Statistical modeling of global Langmuir mixing. Ocean Modell., 113, 95114, doi:10.1016/j.ocemod.2017.03.016.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Maltrud, M. E., F. Bryan, and S. Peacock, 2010: Boundary impulse response functions in a century-long eddying global ocean simulation. Environ. Fluid Mech., 10, 275295, doi:10.1007/s10652-009-9154-3.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Mazloff, M., P. Heimbach, and C. Wunsch, 2010: An eddy-permitting Southern Ocean state estimate. J. Phys. Oceanogr., 40, 880899, doi:10.1175/2009JPO4236.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • McCartney, M. S., 1977: Subantarctic mode water. A Voyage of Discovery, M. Angel, Ed., Elsevier, 103–119.

  • McCartney, M. S., 1982: The subtropical recirculation of Mode Water. J. Mar. Res., 40, 427464.

  • Morrison, A. K., O. A. Saenko, A. M. Hogg, and P. Spence, 2013: The role of vertical eddy flux in Southern Ocean heat uptake. Geophys. Res. Lett., 40, 54455450, doi:10.1002/2013GL057706.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Pfeffer, R. L., 1981: Wave-mean flow interactions in the atmosphere. J. Atmos. Sci., 38, 13401359, doi:10.1175/1520-0469(1981)038<1340:WMFIIT>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Phillips, H. E., and S. R. Rintoul, 2000: Eddy variability and energetics from direct current measurements in the Antarctic Circumpolar Current south of Australia. J. Phys. Oceanogr., 30, 30503076, doi:10.1175/1520-0485(2000)030<3050:EVAEFD>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Ribbe, J., 1999: On wind-driven mid-latitude convection in ocean general circulation models. Tellus, 51A, 517525, doi:10.3402/tellusa.v51i4.14116.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Rintoul, S. R., and M. H. England, 2002: Ekman transport dominates local air–sea fluxes in driving variability of Subantarctic Mode Water. J. Phys. Oceanogr., 32, 13081321, doi:10.1175/1520-0485(2002)032<1308:ETDLAS>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Robinson, W. A., 2006: On the self-maintenance of midlatitude jets. J. Atmos. Sci., 63, 21092122, doi:10.1175/JAS3732.1.

  • Sabine, C. L., and et al. , 2004: The oceanic sink for anthropogenic CO2. Science, 305, 367371, doi:10.1126/science.1097403.

  • Sallée, J. B., N. Wienders, K. Speer, and R. Morrow, 2006: Formation of subantarctic mode water in the southeastern Indian Ocean. Ocean Dyn., 56, 525542, doi:10.1007/s10236-005-0054-x.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sallée, J. B., K. Speer, and S. Rintoul, 2010: Zonally asymmetric response of the Southern Ocean mixed-layer depth to the Southern Annular Mode. Nat. Geosci., 3, 273279, doi:10.1038/ngeo812.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sallée, J. B., R. Matear, S. R. Rintoul, and A. Lenton, 2012: Localized subduction of anthropogenic carbon dioxide in the Southern Hemisphere oceans. Nat. Geosci., 5, 579584, doi:10.1038/ngeo1523.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Sarmiento, J. L., N. Gruber, M. A. Brzezinski, and J. P. Dunne, 2004: High-latitude controls of thermocline nutrients and low latitude biological productivity. Nature, 427, 5660, doi:10.1038/nature02127.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schiller, A., and K. R. Ridgway, 2013: Seasonal mixed-layer dynamics in an eddy-resolving ocean circulation model. J. Geophys. Res. Oceans, 118, 33873405, doi:10.1002/jgrc.20250.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Schmitz, W. J., Jr., 1996: On the eddy field in the Agulhas retroflection, with some global considerations. J. Geophys. Res., 101, 16 25916 271, doi:10.1029/96JC01143.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Smith, R. D., J. K. Dukowicz, and R. C. Malone, 1992: Parallel ocean general circulation modeling. Physica D, 60, 3861, doi:10.1016/0167-2789(92)90225-C.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Treguier, A. M., I. M. Held, and V. D. Larichev, 1997: Parameterization of quasigeostrophic eddies in primitive equation ocean models. J. Phys. Oceanogr., 27, 567580, doi:10.1175/1520-0485(1997)027<0567:POQEIP>2.0.CO;2.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Weller, R. A., P. W. Furey, M. A. Spall, and R. E. Davis, 2004: The large-scale context for oceanic subduction in the northeast Atlantic. Deep-Sea Res., 51, 665699, doi:10.1016/j.dsr.2004.01.003.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Wenegrat, J. O., and M. J. McPhaden, 2016: Wind, waves, and fronts: Frictional effects in a generalized Ekman model. J. Phys. Oceanogr., 46, 371394, doi:10.1175/JPO-D-15-0162.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
  • Williams, R. G., C. Wilson, and C. W. Hughes, 2007: Ocean and atmosphere storm tracks: The role of eddy vorticity forcing. J. Phys. Oceanogr., 37, 22672289, doi:10.1175/JPO3120.1.

    • Crossref
    • Search Google Scholar
    • Export Citation
1

Thermally indirect circulation is defined as a circulation by which kinetic energy is converted to potential and/or internal energy (American Meteorological Society 2017).

Save