1. Introduction
Satellite altimeter measurements over the last three decades have revealed that the Southern Ocean (SO) is abundant with energetic oceanic mesoscale eddies with spatiotemporal scales of O(50–300) km and O(10–100) days (Chelton et al. 2011; Frenger et al. 2015; Martínez‐Moreno et al. 2019). These mesoscale eddies can induce an anticlockwise meridional overturning cell, which opposes the westerly wind–driven clockwise cell that eventually shapes the residual meridional overturning circulation in the SO (Marshall and Speer 2012; Thompson et al. 2014). Accompanying mesoscale-eddy-induced overturning circulation, they transport a huge amount of heat poleward (i.e., southward) across the Antarctic Circumpolar Current (ACC; Jayne and Marotzke 2002; Volkov et al. 2008; Wunsch 1999). Given that in the SO, the large-scale geostrophic currents like the ACC are generally zonal and the Ekman heat transport driven by westerly winds is intrinsically equatorward (i.e., northward), the mesoscale-eddy-induced meridional heat transport (EHT) is believed as a crucial pathway for transporting warmer water from subtropical to subpolar regions in the SO (de Szoeke and Levine 1981; Meijers et al. 2007; Volkov et al. 2010). Therefore, the variability of EHT significantly modulates the heat budget in the SO and profoundly influences the melting and refreezing of sea ice around Antarctica (Dufour et al. 2015; Griffies et al. 2015; McKee et al. 2019; Rackow et al. 2022; Stewart et al. 2018).
Given the importance of EHT in the SO, large efforts have been made to quantify its time-mean values and examine its spatiotemporal variations over the past several decades (Foppert et al. 2017 and references therein). As the EHT is the area integration of eddy-induced meridional heat flux (EHF) over a specific zonal and vertical transect, direct quantification of the time-mean EHT in the whole SO from observations is a very challenging issue. The existing in situ observation–based studies primarily focus on EHFs in some specific regions such as the Drake Passage (Bryden 1979; Ferrari et al. 2014; Nowlin et al. 1985; Sciremammano 1980; Watts et al. 2016), the Shag Rocks Passage (Walkden et al. 2008), and the regions southeast of New Zealand and south of Tasmania (Bryden and Heath 1985; Phillips and Rintoul 2000). However, because the EHFs calculated from these localized in situ observations contain a significant rotational component that makes no contribution to the heat budget, it will bring large uncertainties to use them to quantify the mean EHT, even in regional scopes (e.g., Ferrari et al. 2014; Gille 2003; Marshall and Shutts 1981; Wunsch 1999).
Due to the scarcity of in situ observations, researchers turned to adopt satellite data, eddy-resolving simulations, and their combinations with in situ observations to investigate the EHF and EHT in the SO or global ocean (Dong et al. 2014; Foppert et al. 2017; Frenger et al. 2015; Guo et al. 2022; Hausmann and Czaja 2012; Hogg et al. 2008; Jayne and Marotzke 2002; Laxenaire et al. 2020; Meijers et al. 2007; Stammer 1998; Sun et al. 2019; Volkov et al. 2008, 2010; Wang et al. 2021; Zhang et al. 2014). These studies found that the EHFs are highly nonuniform in the SO and their magnitudes are notably enhanced in the strong-current regions such as the ACC, the Agulhas Current and its return current, the East Australian Current, and the Brazil Current. These studies also suggested that the Indian Ocean sector contributes significantly to the poleward EHT in the SO; in the core band of the ACC between 40° and 60°S, the poleward EHT (circumpolar integration) is generally between 0.05 and 0.8 PW (1 PW = 1015 W), which is comparable in magnitude to the mean heat transport.
Compared with the mean values and spatial distributions of EHT, investigating its temporal variations (in the long term, in particular) is more challenging. Based on the 1/4° Estimating the Circulation and Climate of the Ocean, Phase II (ECCO2) data, Volkov et al. (2008) explored the interannual variability of EHT in the SO, but the time was not long enough to study decadal changes and trends. More recently, Foppert et al. (2017) estimated the depth-integrated EHF (DI-EHF) in ACC using the standard deviation of sea surface height (SSH) as a proxy. They found that three DI-EHF hotspots showed significant increasing trends between 1993 and 2016. However, extrapolating the statistical relation obtained in the Drake Passage to the entire SO by the authors may have resulted in large uncertainties in their results. Based on satellite observations, the most recent study by Guo et al. (2022) showed that surface EHF had an increasing trend in the ACC over the period of 1993–2020. However, their surface-only results cannot represent the DI-EHF, which truly matters in the meridional heat transport and heat budget in the SO (Jayne and Marotzke 2002; Meijers et al. 2007; Watts et al. 2016; Zhang et al. 2023). Therefore, the decadal variations of DI-EHF and EHT (i.e., the zonal integration of DI-EHF) in the SO are still obscure now.
In this study, decadal trends in the EHF and EHT in the SO and their associated mechanisms are investigated by combining 24-yr-long ECCO2 data and satellite data between 1993 and 2016. The remainder of this paper is organized as follows. Section 2 shows the details of the data and methodology used in this study. Section 3 describes the spatial pattern and decadal trends in the EHF and EHT. In section 4, we reveal the mechanisms of the decadal trends in the EHF and EHT. Finally, summary and discussion of the paper are given in section 5.
2. Data and methods
a. ECCO2 data
To investigate the decadal variations of EHF and EHT, the 24-yr ECCO2 reanalysis data between January 1993 and December 2016 were analyzed. The ECCO2 reanalysis data were produced by synthesizing the available global ocean and sea ice observational data in the Massachusetts Institute of Technology general circulation model with a global full-depth ocean and sea ice configuration using the Green’s function approach (Menemenlis et al. 2005a, 2008). The horizontal and temporal resolutions of the ECCO2 data are 1/4° and 3 days, respectively. In the vertical, the ECCO2 has 50 levels with a maximum model depth of 5906 m, and the level thicknesses vary from 10 m at the surface to 456 m near the maximum depth. The vertical eddy diffusivity and viscosity are parameterized using the K-profile parameterization scheme (Large et al. 1994) and the horizontal subgrid viscosity is parameterized using the Leith scheme (Fox-Kemper and Menemenlis 2008; Leith 1996). In this study, the three-dimensional variables including current velocities, potential temperature, and salinity in the SO were downloaded and analyzed. To help investigate the mechanisms, the two-dimensional variables of SSH, sea surface temperature (SST), and zonal and meridional surface wind stress were also used. Detailed descriptions of the ECCO2 data can be found in Menemenlis et al. (2008).
With a 1/4° horizontal resolution, the ECCO2 data are eddy permitting and they can to a certain degree resolve the larger mesoscale eddies in the SO. Compared with other eddy-permitting or eddy-resolving ocean models, the advantage of ECCO2 is that it runs forward with optimized control parameters obtained through minimizing model/observation misfit based on the Green’s function approach (Menemenlis et al. 2005a,b). Therefore, the ECCO2 data are dynamically and physically consistent, which are suitable to be used to perform diagnostic analysis in aspects of both phenomena and mechanisms (Menemenlis et al. 2008). Indeed, the ECCO2 data have been widely used in the past decade to study mesoscale eddies and the relevant issues in different regions of world oceans (e.g., Chen et al. 2014; Fu 2009; Qiu et al. 2017; Volkov et al. 2008; Yang et al. 2018; Zemskova et al. 2015; Zhu et al. 2018). In particular, the studies of Volkov et al. (2010) and Zhu et al. (2018) have demonstrated that the ECCO2 data has a good performance in simulating spatiotemporal variations of mesoscale eddies and the corresponding EHT in the SO. The reason why the earlier version of ECCO2 data between 1993 and 2016 is analyzed here is that the new version of the ECCO2 model has changed the atmospheric forcing fields after 2009, which may result in fictitious trends in eddy activities (Zhang 2020).
b. Satellite data
To examine the surface EHF trend and validate the performance of ECCO2 data, the concurrent satellite altimeter data and SST data (i.e., 1993–2016) are also used in this study. For the altimeter data, they are the multimission merged gridded product provided by the Copernicus Marine Environment Monitoring Service. The data include the absolute dynamic topography and surface geostrophic velocity in the SO, whose spatial and temporal resolutions are 1/4° and 1 day, respectively. With respect to the satellite SST, the daily Optimum Interpolation Sea Surface Temperature (OISST) dataset distributed by the National Oceanic and Atmospheric Administration is used. Here, the Advanced Very High Resolution Radiometer (AVHRR)-only version 2 data were downloaded and analyzed (Banzon et al. 2016; Reynolds et al. 2007), whose spatiotemporal resolutions are the same as the altimeter data.
c. Calculation of EHF and EHT
After obtaining the EHF from the ECCO2 data at each grid point, we further calculated the DI-EHF in the upper 1000 m; then, we obtained the EHT by circumpolar integration of the DI-EHF at each specific latitude. The upper 1000-m integration was chosen because previous studies have found that the EHF in the SO is primarily confined to the upper 1000 m (Jayne and Marotzke 2002; Meijers et al. 2007; Watts et al. 2016; Zhang et al. 2023). Note that for the surface-data-derived results (for both satellite and ECCO2 data), the term surface EHT actually refers to heat transport per unit meter (i.e., with the unit of W m−1). The geostrophic current used to calculate the surface EHF is calculated from SSH using geostrophic relation. The recent study by Zhu et al. (2022) has found that the AVHRR-only version 2 OISST data have significantly underestimated the mesoscale SST anomaly before 2007 due to the change in satellite AVHRR instrument in 2007. To avoid the fictitious trend, before the calculation of EHF using satellite data, we first corrected the SST anomaly data before 2007 by multiplying it by 1.3 according to Zhu et al. (2022) (Fig. S1 in the online supplemental material). In the following, the positive and negative values of EHF and EHT denote poleward and equatorward directions (i.e., southward and northward), respectively.
d. EKE budget analysis
3. Decadal trends in EHF and EHT
a. Surface EHF and surface EHT
Before showing the surface EHF and surface EHT, we first examine the distributions of the root-mean-square (RMS) υ′ at the surface and mesoscale anomaly of SST (θ′) derived from the ECCO2 and satellite data (Fig. 1). It is well expected that both the observed RMS υ′ and θ′ show large values in areas of intense currents such as the ACC, the Agulhas Current and its return current, the East Australian Current, and the Brazil–Malvinas Confluence, which is consistent with previous findings (e.g., Zhang et al. 2023). For the observed RMS θ′, it also shows a moderately high magnitude in the subtropical frontal regions north of 35°S. The results derived from the ECCO2 data display very similar distributions to the satellite-observed ones, demonstrating that ECCO2 can reasonably capture the mesoscale processes at the sea surface. However, it should also be noted that the magnitude of the ECCO2-derived results is slightly weaker than that of satellite observations. In the core band of the ACC between 40° and 60°S, the region-averaged RMS υ′ and θ′ from ECCO2 data are 21% and 32% smaller than the satellite results, respectively. This is consistent with the previous studies (Volkov et al. 2010; Zhu et al. 2018). Specifically, Volkov et al. (2010) found that due to the limited resolution of ECCO2, the ECCO2-derived EKE is on average 28% smaller than altimeter-derived EKE in the SO between 30° and 60°S.
The RMS sea surface (a) υ′ and (c) θ′ in the SO derived from satellite data between 1993 and 2016. Black lines are contours of mean SSH at 0.25-m intervals. Black thick lines are contours of −1.0 and 0.3 m that roughly indicate the scope of the ACC. (b),(d) As in (a) and (c), but for the ECCO2 data. For a fair comparison, the meridional geostrophic velocity anomaly is shown here for both the altimeter and ECCO2 data. The black thick lines in (b) and (d) are SSH contours of −1.5 and −0.25 m.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
By combining the concurrent SST and surface geostrophic velocity, we calculate the time-mean surface EHFs. The results derived from satellite and ECCO2 data are compared in Fig. 2. As expected, the surface EHF is dominantly poleward (i.e., southward) and its spatial distribution is overall similar to the RMS υ′ and θ′, that is, showing elevated magnitudes in the strong current regions and subtropical frontal regions (Martínez-Moreno et al. 2021; Volkov et al. 2008). Overall, the Indian and Atlantic Ocean sectors make the dominant contribution to the poleward surface EHF corresponding to the strong mesoscale variabilities associated with strong currents. The only evident negative values (i.e., equatorward) of surface EHF occur in the region southwest of Africa, which are linked to the Agulhas rings shedding from the retroreflection of Agulhas Current that take warm waters northward into the southeast Atlantic Ocean (Laxenaire et al. 2020; Souza et al. 2011). Similar to the comparisons in Fig. 1, spatial patterns of the satellite- and ECCO2-derived surface EHFs are very similar, but the magnitude of the latter is weaker. For example, in the core band of the ACC (i.e., 40°–60°S), the region-averaged mean surface EHFs from the satellite and ECCO2 data are 12.8 and 7.2 kW m−2, respectively. It means that in this ACC latitude band, the surface EHF from the ECCO2 data is 44% weaker than that from the satellite data.
Sea surface EHF calculated from (a) satellite data and (b) ECCO2 data averaged between 1993 and 2016 in the SO. Positive values denote poleward (i.e., southward) surface EHF.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
In Figs. 3a and 3b, we show spatial distributions of the linear trend of surface EHF in the SO derived from satellite and ECCO2 data, respectively. Here, the trend was obtained by applying least squares linear fitting to the time series of surface EHF between 1993 and 2016 at each grid point. The results reveal that both the satellite- and ECCO2-derived surface EHFs generally have increasing trends (i.e., poleward surface EHFs increase) in the SO, particularly in the ACC and the western boundary current regions. The ECCO2-derived surface EHF, however, has a fictitious increasing trend in several regions, which is opposite to the decreasing trend observed in the satellite-derived result. These fictitious-trend regions are primarily in the subtropics including the interiors of subtropical Indian and Atlantic Oceans and the northern part of the East Australian Current. Given that the observed and ECCO2 results have consistently increasing trends in the ACC latitude band between 40° and 60°S, we will mainly focus on this important band in the following analysis.
Spatial distributions of the linear trends in surface EHF between 1993 and 2016 derived from (a) satellite data and (b) ECCO2 data. Positive values denote increasing trend in poleward surface EHF. Two green lines denote the core band of the ACC between 40° and 60°S. (c) Time series of surface EHT averaged between 40° and 60°S. The annual mean (original) surface EHT is denoted using dark (light) color solid line. Red and blue solid lines represent the satellite and ECCO2 results, respectively. Orange and cyan dashed lines denote the linear trends of the associated solid lines. Note that the ECCO2-derived surface EHT has been multiplied by a factor of 1.5 here.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
In Fig. 3c, we compare the time series of satellite- and ECCO2-derived surface EHTs averaged over the ACC latitude band (i.e., 40°–60°S). Although the magnitude of the ECCO2-derived surface EHT is smaller than the observed one, the two time series indeed show similar increasing trends between 1993 and 2016. Specifically, for the satellite (ECCO2)-derived surface EHT, its time-mean value and linear trend are 3.3 × 10−4 PW m−1 (1.8 × 10−4 PW m−1) and 5.4 ± 1.8 × 10−5 PW m−1 decade−1 (1.8 ± 0.9 × 10−5 PW m−1 decade−1), respectively. Both the two linear trends are statistically significant at the 95% confidence level. These results mean that over the 24 years between 1993 and 2016, the satellite- and ECCO2-derived surface EHTs in the ACC band had increased by 46% and 26%, respectively. In sum, the comparisons in Figs. 1–3 validate the performance of ECCO2, which lends us the confidence to further investigate the EHT and its decadal trends using ECCO2 data.
b. DI-EHF and EHT
Given that the ECCO2 data can overall well reproduce the mean pattern and trends of surface EHF, we further use it to calculate the DI-EHF and EHT. Because the EHT over the upper 1000 m constitutes more than 80% of the total EHT integrated over the entire water column in the SO (Zhang et al. 2023), only the upper 1000-m data are used in the vertical integral. If not specified below, both the DI-EHF and EHT refer to the depth-integrated results in the upper 1000 m. The spatial pattern of the time-mean DI-EHF is overall close to the surface one except that the elevated values in the ACC and the western boundary current regions are more prominent compared with the subtropical ocean interior regions (Fig. 4a versus Fig. 2b). Again, the spatial pattern of the ECCO2-derived DI-EHF in the SO is overall in line with previous studies based on independent data (Foppert et al. 2017; Jayne and Marotzke 2002; Stammer 1998; Sun et al. 2019).
Spatial distributions of ECCO2 data-derived (a) time-mean and (b) linear trend of DI-EHF in the upper 1000 m between 1993 and 2016. The four hotspots of DI-EHF are marked using cyan boxes and their names are labeled.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
Figure 4b shows the spatial distribution of DI-EHF trend in the SO. Similar to the surface EHF trend, the DI-EHF also has overall increasing trends in the SO. The difference is that significant increasing trends of DI-EHF are mainly confined in the ACC band and trends in the subtropical ocean interior regions are weak. The hotspots of increasing trends of DI-EHF coincide with the regions with major topographic features such as the Kerguelen Plateau, Southeast Indian Ridge, Pacific Antarctic Ridge, and Drake Passage. The first two topography-related hotspots of DI-EHF increasing trends agree with those found by Foppert et al. (2017), which estimated the DI-EHF using SSH standard deviation. These hotspots are regions where there are strong fronts with high mesoscale activities.
In Fig. 5a, we show the time series of EHT averaged over every 2.5° latitude in the ACC band between 40° and 60°S (i.e., a total of 8 time series). The mean values of these time series range from 0.04 to 0.13 PW, with the maximum value reaching 0.20 PW for that at around 40°S. Note that the EHTs at different latitudes of ACC all have significantly increasing trends, although their magnitudes and increasing rates are latitude dependent (Fig. 5a; Table S1). The increasing rate of the EHT averaged over the ACC band is approximately 16% per decade compared with its time-mean value of 0.07 PW, which is more prominent than the surface EHT. Correspondingly, the ACC-band-averaged EHT has a linear trend of 1.1 ± 0.2 × 10−2 PW decade−1 (95% confidence level).
(a) Time series of the 13-month low-pass-filtered EHTs. The eight thin lines with colors of olive, magenta, purple, pink, green, yellow, orange, and red denote EHTs averaged every 2.5° from 40° to 60°S, respectively. Thick blue line is the mean of the eight thin lines, and the dashed cyan line is its linear trend. (b) The latitudinal distribution of EHTs averaged over the first decade between 1993 and 2002 (blue) and the second decade between 2007 and 2016 (red). Shadings denote the 95% confidence interval computed using the bootstrap method.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
To further explore the decadal increase in EHT at different latitudes, in Fig. 5b, we compare the latitudinal distributions of EHT averaged in the first and last decades (i.e., 1993–2002 versus 2007–16). It reveals that except at 38°–39°S, the EHTs all have a poleward increase throughout the SO. The decadal increase in EHT is more prominent south of 50°S than north of that. The largest EHT increase occurs at 42°S with an increased value of 0.03 PW.
4. Mechanisms of the DI-EHF and EHT trends
a. Diagnostics based on mixing-length theory
Spatial distributions of the time-mean DI-EHFp and its decadal trend are shown in Figs. 6a and 6b, respectively. By comparing these results with Fig. 4, we find that both the mean DI-EHFp and the corresponding decadal trend are overall similar to the directly calculated DI-EHF, in aspects of both spatial patterns and magnitudes. The consistency becomes even better when the ACC-band-averaged time series of EHTp and EHT are compared (Fig. 6c). The linear trend of the ACC-band-averaged EHTp is 1.0 × 10−2 PW decade−1, which is 9% lower than that of the directly calculated EHT. The consistency between the DI-EHFp and EHTp with the directly calculated ones demonstrates that it is feasible to investigate the mechanisms of their decadal trends using the parameterization method in Eqs. (3)–(5).
(a),(b) As in Figs. 4a and 4b, but for the DI-EHFp in the SO. (c) Time series of the 13-month low-pass-filtered EHTs averaged in the ACC band (i.e., 40°–60°S). Blue and red solid lines denote the directly calculated EHT and the EHTp, respectively. Cyan (orange) dashed line denotes the linear trend of the blue (red) solid line.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
Spatial distributions of (a) time-mean and (b) linear trend of depth-integrated EKE in the upper 1000 m during 1993–2016 in the SO. (c),(d) As in (a) and (b), but for the depth-integrated
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
Based on Eq. (6), we quantified the sources of the decadal trend of DI-EHFp by calculating each term. The results reveal that the first and second right-hand side terms (i.e., due to the changes in EKE and
Linear trends of the DI-EHFp due to the changes in (a) EKE, (b)
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
b. Mechanisms of the EKE trend
To investigate the mechanisms of EKE increasing trend, we perform an EKE budget analysis in the upper 1000 m using ECCO2 data (see the method in section 2d). In Fig. 9 and Fig. S4a, we show the long-term mean distributions of the right-hand side terms in Eq. (2) (i.e., source or sink terms of EKE integrated in the upper 1000 m). The results reveal that VEDF and BT are the largest positive and largest negative terms throughout the SO, respectively, but the magnitude of the latter is smaller than the former. Both of them show a spatial pattern similar to the EKE (Figs. 7a and 9a,b). The large positive VEDF term means that mesoscale eddies obtain their kinetic energy primarily from the release of APEme through baroclinic instability (Smith 2007). With respect to the negative BT term, it means that mesoscale eddies lose a certain proportion of their kinetic energy through inverse energy cascade (i.e., feeding back the large-scale kinetic energy). Although the magnitudes of the ADV and PD (i.e., advection of EKE and divergence of pressure work) are comparable to the BT term, their values alternate positive and negative in the SO and therefore make a much smaller contribution to the EKE budget on the whole. The WW is smaller by 1–2 orders of magnitude than the VEDF and BT terms, and therefore, its contribution is also negligible.
Spatial distributions of time-mean EKE budget terms in the SO including (a) VEDF, (b) BT, (c) ADV, and (d) PD. Detailed definitions of these budget components are available in the data and methods section.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
Figures 10a–d show the spatial distributions of linear trends of VEDF, BT, ADV, and PD, respectively. We find that the VEDF term has significant increasing trends in the SO, particularly in the band of ACC. With respect to the other three terms (i.e., BT, ADV, and PD), however, their trends are all alternatingly positive and negative in the SO. Although the absolute magnitudes of their trends are not smaller than those of VEDF, their integrated trends in the ACC band are trivial. The WW term also has overall increasing trends in the SO, but its magnitude is one order smaller than that of VEDF (Fig. 10a versus Fig. S4b). The reason why the WW term (small but positive) does not behave as an EKE sink as revealed by previous studies (Rai et al. 2021; Zhai et al. 2012) is that the wind stress used in ECCO2 does not consider the relative motion between wind and sea surface current. The significant increasing trend in VEDF and trivial trends in BT, ADV, PD, and ET can be seen more clearly from their time series averaged in the ACC band (Fig. 11). Quantitatively, the averaged VEDF term in the ACC band has an increasing rate of 2.8 ± 0.2 × 10−4 W m−2 decade−1 or 19% per decade (95% confidence level) relative to its time-mean value (i.e., 1.5 × 10−3 W m−2). The above results suggest that the enhanced APEme release through strengthened baroclinic instability is the dominant mechanism for the decadal increase in EKE.
As in Fig. 9 but for the linear trends of different EKE budget terms between 1993 and 2016.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
(a) Time series of the 13-month low-pass-filtered EKE budget terms averaged in the ACC band (i.e., 40°–60°S). Red, blue, yellow, pink, and green solid lines denote VEDF, PD, BT, ADV, and ET, respectively. Orange, cyan, brown, light pink, and light green dashed lines denote the linear trends of the associated solid lines. (b) Increasing rates of the EKE budget terms in (a).
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
To understand the strengthening trend of baroclinic instability, we further examine the trends in large-scale WW and depth-integrated baroclinic conversion (BC) from large-scale APE to APEme in Fig. 12. Here, the large-scale WW is calculated using
Spatial distributions of the linear trends in (a) large-scale WW and (b) depth-integrated BC between 1993 and 2016. (c) Time series of the 13-month low-pass-filtered large-scale WW (blue) and depth-integrated BC (red) averaged in the ACC band (i.e., 40°–60°S). Cyan and orange dashed lines denote the linear trends of the associated solid lines.
Citation: Journal of Climate 37, 14; 10.1175/JCLI-D-23-0462.1
It shows that positive large-scale WW has significantly increasing trends in the SO (Fig. 12a and Fig. S5a), which is due to the strengthening westerly winds corresponding to the positive phase of the Southern Annular Mode during the study period (Abram et al. 2014; Fogt and Marshall 2020; Hogg et al. 2015; Huang et al. 2019; Wang et al. 2015; Young and Ribal 2019). As a result, the positive BC term also significantly increases in the SO (Fig. 12b and Fig. S5b), which means that the conversion from large-scale APE to APEme is enhanced. In terms of the ACC-band average, large-scale WW and BC have an increasing rate of 1.1 × 10−3 W m−2 decade−1 and 2.9 × 10−4 W m−2 decade−1 (i.e., 13% and 15% per decade), respectively (Fig. 12c). The above increasing rate of BC is very close to that of VEDF (i.e., 2.8 × 10−4 W m−2 decade−1), suggesting that nearly all of the increased APEme obtained from baroclinic instability has been converted to EKE. The connection between the increasing large-scale WW and BC is that the former strengthens the energy conversion from large-scale KE to large-scale APE through increasing vertical mean density flux via Ekman pumping (Fig. S6), which then strengthens the baroclinic instability.
5. Summary and discussion
Based on ECCO2 reanalysis data verified by satellite observations, this study investigates the decadal trends and the associated mechanisms of EHF and EHT in the SO. We find that EHF and EHT have poleward increasing trends in the SO between 1993 and 2016. The hotspots of increasing DI-EHF occur in the major topographic regions in the ACC band such as the Kerguelen Plateau, Southeast Indian Ridge, Pacific Antarctic Ridge, and Drake Passage. In terms of average in the ACC band, the EHT has a linear trend of 1.1 × 10−12 PW decade−1 or equivalently 16% per decade between 1993 and 2016. Through performing diagnostic analysis using the mixing length theory, we find that the strengthened EKE is the dominant mechanism for the decadal increasing trends of EHT. Further energy budget analysis suggests that the decadal increase of EKE is mainly fed by the enhanced APEme release through strengthened baroclinic instability. The strengthened baroclinic instability here is attributed to the increasing large-scale WW on the large-scale circulation corresponding to the positive phase of the Southern Annular Mode during the study period of 1993–2016.
Given that EHT is of leading-order importance in the meridional heat transport in the SO, the decadal increasing trend in EHT identified here may significantly modulate the decadal variations of heat storage in different regions of the SO (Morrison et al. 2016; Sun et al. 2019; Volkov et al. 2008). In addition, by taking more heat poleward, the increasing EHT may also influence the Antarctic sea ice extent at the decadal time scale (Rackow et al. 2022). We should note that the mesoscale eddies defined here are only based on temporal scale (i.e., period shorter than 150 days) and the standing eddies that do not change over time or change only at low frequency (interannual or longer) are not considered, although they are also demonstrated to have an essential role in the EHT (Bryan et al. 2014; Meijers et al. 2007; Volkov et al. 2010). Whether the low-frequency variation of standing eddies can also contribute to the decadal trend in EHT needs to be evaluated in future studies. According to the “eddy compensation” theory (Downes and Hogg 2013; Marshall and Radko 2003; Meredith et al. 2012), the poleward increasing trend in EHT may compensate for the equatorward increasing trend in wind-driven Eulerian mean heat transport at the decadal time scale. How this compensation will influence the decadal variation of the total meridional heat transport in the SO will be left for future studies. Last, our recent study has pointed out that an inverse energy cascade associated with submesoscale processes can feed mesoscale eddies and therefore enhance EHT in the SO (Zhang et al. 2023). Because the ECCO2 data used here cannot resolve submesoscales, the decadal variations of EHT associated with submesoscale inverse cascade also deserve to be studied when long-term submesoscale permitting data are available in the future.
Acknowledgments.
This study is jointly supported by the National Natural Science Foundation of China and the National Key Research and Development Program of China (42222601, 2022YFC3105003, 92258301, and 42076004). Z. Zhang is also supported by the “Taishan” Talents Program (tsqn202103032) and the Shandong Provincial Natural Science Foundation (ZR2023JQ013). The authors also wish to acknowledge the ECCO team (https://ecco.jpl.nasa.gov) for their contributions to the development and production of ECCO2 data.
Data availability statement.
The ECCO2 data as the key dataset used in the manuscript are accessible at https://ecco.jpl.nasa.gov/drive/files/ECCO2/cube92_latlon_quart_90S90N/(accessed in 2018). The satellite altimeter data and SST data are downloaded from https://data.marine.copernicus.eu/product/SEALEVEL_GLO_PHY_L4_MY_008_047/description and https://psl.noaa.gov/data/gridded/data.noaa.oisst.v2.highres.html, respectively.
APPENDIX
Divergent and Rotational Components Decomposition of EHF
REFERENCES
Abram, N. J., R. Mulvaney, F. Vimeux, S. J. Phipps, J. Turner, and M. H. England, 2014: Evolution of the Southern Annular Mode during the past millennium. Nat. Climate Change, 4, 564–569, https://doi.org/10.1038/nclimate2235.
Banzon, V., T. M. Smith, T. M. Chin, C. Liu, and W. Hankins, 2016: A long-term record of blended satellite and in situ sea-surface temperature for climate monitoring, modeling and environmental studies. Earth Syst. Sci. Data, 8, 165–176, https://doi.org/10.5194/essd-8-165-2016.
Bryan, F. O., P. R. Gent, and R. Tomas, 2014: Can Southern Ocean eddy effects be parameterized in climate models? J. Climate, 27, 411–425, https://doi.org/10.1175/JCLI-D-12-00759.1.
Bryden, H. L., 1979: Poleward heat flux and conversion of available potential energy in Drake Passage. J. Mar. Res., 37, 1–22.
Bryden, H. L., and R. A. Heath, 1985: Energetic eddies at the northern edge of the Antarctic Circumpolar Current in the Southwest Pacific. Prog. Oceanogr., 14, 65–87, https://doi.org/10.1016/0079-6611(85)90006-0.
Chelton, D. B., M. G. Schlax, and R. M. Samelson, 2011: Global observations of nonlinear mesoscale eddies. Prog. Oceanogr., 91, 167–216, https://doi.org/10.1016/j.pocean.2011.01.002.
Chen, R., G. R. Flierl, and C. Wunsch, 2014: A description of local and nonlocal eddy–mean flow interaction in a global eddy-permitting state estimate. J. Phys. Oceanogr., 44, 2336–2352, https://doi.org/10.1175/JPO-D-14-0009.1.
de Szoeke, R. A., and M. D. Levine, 1981: The advective flux of heat by mean geostrophic motions in the Southern Ocean. Deep-Sea Res., 28A, 1057–1085, https://doi.org/10.1016/0198-0149(81)90048-0.
Dong, C., J. C. McWilliams, Y. Liu, and D. Chen, 2014: Global heat and salt transports by eddy movement. Nat. Commun., 5, 3294, https://doi.org/10.1038/ncomms4294.
Downes, S. M., and A. M. Hogg, 2013: Southern Ocean circulation and eddy compensation in CMIP5 models. J. Climate, 26, 7198–7220, https://doi.org/10.1175/JCLI-D-12-00504.1.
Dufour, C. O., and Coauthors, 2015: Role of mesoscale eddies in cross-frontal transport of heat and biogeochemical tracers in the Southern Ocean. J. Phys. Oceanogr., 45, 3057–3081, https://doi.org/10.1175/JPO-D-14-0240.1.
Ferrari, R., C. Provost, Y.-H. Park, N. Sennéchael, Z. Koenig, H. Sekma, G. Garric, and R. Bourdallé-Badie, 2014: Heat fluxes across the Antarctic Circumpolar Current in Drake Passage: Mean flow and eddy contributions. J. Geophys. Res. Oceans, 119, 6381–6402, https://doi.org/10.1002/2014JC010201.
Fogt, R. L., and G. J. Marshall, 2020: The Southern Annular Mode: Variability, trends, and climate impacts across the Southern Hemisphere. Wiley Interdiscip. Rev.: Climate Change, 11, e652, https://doi.org/10.1002/wcc.652.
Foppert, A., K. A. Donohue, D. R. Watts, and K. L. Tracey, 2017: Eddy heat flux across the Antarctic Circumpolar Current estimated from sea surface height standard deviation. J. Geophys. Res. Oceans, 122, 6947–6964, https://doi.org/10.1002/2017JC012837.
Fox-Kemper, B., and D. Menemenlis, 2008: Can large eddy simulation techniques improve mesoscale rich ocean models? Ocean Modeling in an Eddying Regime, Geophys. Monogr., Vol. 177, Amer. Geophys. Union, 319–337.
Frenger, I., M. Münnich, N. Gruber, and R. Knutti, 2015: Southern Ocean eddy phenomenology. J. Geophys. Res. Oceans, 120, 7413–7449, https://doi.org/10.1002/2015JC011047.
Fu, L.-L., 2009: Pattern and velocity of propagation of the global ocean eddy variability. J. Geophys. Res., 114, C11017, https://doi.org/10.1029/2009JC005349.
Gille, S. T., 2003: Float observations of the Southern Ocean. Part II: Eddy fluxes. J. Phys. Oceanogr., 33, 1182–1196, https://doi.org/10.1175/1520-0485(2003)033<1182:FOOTSO>2.0.CO;2.
Griffies, S. M., and Coauthors, 2015: Impacts on ocean heat from transient mesoscale eddies in a hierarchy of climate models. J. Climate, 28, 952–977, https://doi.org/10.1175/JCLI-D-14-00353.1.
Guo, Y., and S. P. Bishop, 2022: Surface divergent eddy heat fluxes and their impacts on mixed layer eddy‐mean flow interactions. J. Adv. Model. Earth Syst., 14, e2021MS002863, https://doi.org/10.1029/2021MS002863.
Guo, Y., S. Bachman, F. Bryan, and S. Bishop, 2022: Increasing trends in oceanic surface poleward eddy heat flux observed over the past three decades. Geophys. Res. Lett., 49, e2022GL099362, https://doi.org/10.1029/2022GL099362.
Hausmann, U., and A. Czaja, 2012: The observed signature of mesoscale eddies in sea surface temperature and the associated heat transport. Deep-Sea Res. I, 70, 60–72, https://doi.org/10.1016/j.dsr.2012.08.005.
Hogg, A. M., M. P. Meredith, J. R. Blundell, and C. Wilson, 2008: Eddy heat flux in the Southern Ocean: Response to variable wind forcing. J. Climate, 21, 608–620, https://doi.org/10.1175/2007JCLI1925.1.
Hogg, A. M., M. P. Meredith, D. P. Chambers, E. P. Abrahamsen, C. W. Hughes, and A. K. Morrison, 2015: Recent trends in the Southern Ocean eddy field. J. Geophys. Res. Oceans, 120, 257–267, https://doi.org/10.1002/2014JC010470.
Holloway, G., 1986: Estimation of oceanic eddy transports from satellite altimetry. Nature, 323, 243–244, https://doi.org/10.1038/323243a0.
Huang, X., X. Cheng, and Y. Qi, 2019: SST front anchored mesoscale feature of surface wind in the southern Indian Ocean. Climate Dyn., 53, 477–490, https://doi.org/10.1007/s00382-018-4598-0.
Jayne, S. R., and J. Marotzke, 2002: The oceanic eddy heat transport. J. Phys. Oceanogr., 32, 3328–3345, https://doi.org/10.1175/1520-0485(2002)032<3328:TOEHT>2.0.CO;2.
Keffer, T., and G. Holloway, 1988: Estimating Southern Ocean eddy flux of heat and salt from satellite altimetry. Nature, 332, 624–626, https://doi.org/10.1038/332624a0.
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, 363–403, https://doi.org/10.1029/94RG01872.
Laxenaire, R., S. Speich, and A. Stegner, 2020: Agulhas ring heat content and transport in the South Atlantic estimated by combining satellite altimetry and Argo profiling floats data. J. Geophys. Res. Oceans, 125, e2019JC015511, https://doi.org/10.1029/2019JC015511.
Leith, C. E., 1996: Stochastic models of chaotic systems. Physica D, 98, 481–491, https://doi.org/10.1016/0167-2789(96)00107-8.
Marshall, J., and G. Shutts, 1981: A note on rotational and divergent eddy fluxes. J. Phys. Oceanogr., 11, 1677–1680, https://doi.org/10.1175/1520-0485(1981)011<1677:ANORAD>2.0.CO;2.
Marshall, J., and T. Radko, 2003: Residual-mean solutions for the Antarctic Circumpolar Current and its associated overturning circulation. J. Phys. Oceanogr., 33, 2341–2354, https://doi.org/10.1175/1520-0485(2003)033<2341:RSFTAC>2.0.CO;2.
Marshall, J., and K. Speer, 2012: Closure of the meridional overturning circulation through Southern Ocean upwelling. Nat. Geosci., 5, 171–180, https://doi.org/10.1038/ngeo1391.
Martínez‐Moreno, J., A. M. Hogg, A. E. Kiss, N. C. Constantinou, and A. K. Morrison, 2019: Kinetic energy of eddy‐like features from sea surface altimetry. J. Adv. Model. Earth Syst., 11, 3090–3105, https://doi.org/10.1029/2019MS001769.
Martínez‐Moreno, J., A. M. Hogg, M. H. England, N. C. Constantinou, A. E. Kiss, and A. K. Morrison, 2021: Global changes in oceanic mesoscale currents over the satellite altimetry record. Nat. Climate Change, 11, 397–403, https://doi.org/10.1038/s41558-021-01006-9.
McKee, D. C., D. G. Martinson, and O. Schofield, 2019: Origin and attenuation of mesoscale structure in circumpolar deep water intrusions to an Antarctic shelf. J. Phys. Oceanogr., 49, 1293–1318, https://doi.org/10.1175/JPO-D-18-0133.1.
Meijers, A. J., N. L. Bindoff, and J. L. Roberts, 2007: On the total, mean, and eddy heat and freshwater transports in the Southern Hemisphere of a ⅛° × ⅛° global ocean model. J. Phys. Oceanogr., 37, 277–295, https://doi.org/10.1175/JPO3012.1.
Menemenlis, D., I. Fukumori, and T. Lee, 2005a: Using Green’s functions to calibrate an ocean general circulation model. Mon. Wea. Rev., 133, 1224–1240, https://doi.org/10.1175/MWR2912.1.
Menemenlis, D., and Coauthors, 2005b: NASA supercomputer improves prospects for ocean climate research. Eos, Trans. Amer. Geophys. Union, 86, 89–96, https://doi.org/10.1029/2005EO090002.
Menemenlis, D., J.-M. Campin, P. Heimbach, C. Hill, T. Lee, A. Nguyen, M. Schodlok, and H. Zhang, 2008: ECCO2: High resolution global ocean and sea ice data synthesis. Mercator Ocean Quarterly Newsletter, No. 31, Mercator Ocean, Toulouse, France, 13–21.
Meredith, M. P., A. C. Naveira Garabato, A. M. Hogg, and R. Farneti, 2012: Sensitivity of the overturning circulation in the Southern Ocean to decadal changes in wind forcing. J. Climate, 25, 99–110, https://doi.org/10.1175/2011JCLI4204.1.
Miao, M., and Coauthors, 2021: On contributions of multiscale dynamic processes to the steric height in the northeastern South China Sea as revealed by moored observations. Geophys. Res. Lett., 48, e2021GL093829, https://doi.org/10.1029/2021GL093829.
Morrison, A. K., S. M. Griffies, M. Winton, W. G. Anderson, and J. L. Sarmiento, 2016: Mechanisms of Southern Ocean heat uptake and transport in a global eddying climate model. J. Climate, 29, 2059–2075, https://doi.org/10.1175/JCLI-D-15-0579.1.
Nowlin, W. D., Jr., S. J. Worley, and T. Whitworth III, 1985: Methods for making point estimates of eddy heat flux as applied to the Antarctic Circumpolar Current. J. Geophys. Res., 90, 3305–3324, https://doi.org/10.1029/JC090iC02p03305.
Patara, L., C. W. Böning, and A. Biastoch, 2016: Variability and trends in Southern Ocean eddy activity in 1/12° ocean model simulations. Geophys. Res. Lett., 43, 4517–4523, https://doi.org/10.1002/2016GL069026.
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, 3050–3076, https://doi.org/10.1175/1520-0485(2000)030<3050:EVAEFD>2.0.CO;2.
Qiu, B., S. Chen, and N. Schneider, 2017: Dynamical links between the decadal variability of the Oyashio and Kuroshio Extensions. J. Climate, 30, 9591–9605, https://doi.org/10.1175/JCLI-D-17-0397.1.
Rackow, T., S. Danilov, H. F. Goessling, H. H. Hellmer, D. V. Sein, T. Semmler, D. Sidorenko, and T. Jung, 2022: Delayed Antarctic sea-ice decline in high-resolution climate change simulations. Nat. Commun., 13, 637, https://doi.org/10.1038/s41467-022-28259-y.
Rai, S., M. Hecht, M. Maltrud, and H. Aluie, 2021: Scale of oceanic eddy killing by wind from global satellite observations. Sci. Adv., 7, eabf4920, https://doi.org/10.1126/sciadv.abf4920.
Reynolds, R. W., T. M. Smith, C. Liu, D. B. Chelton, K. S. Casey, and M. G. Schlax, 2007: Daily high-resolution-blended analyses for sea surface temperature. J. Climate, 20, 5473–5496, https://doi.org/10.1175/2007JCLI1824.1.
Sciremammano, F., Jr., 1980: The nature of the poleward heat flux due to low-frequency current fluctuations in Drake Passage. J. Phys. Oceanogr., 10, 843–852, https://doi.org/10.1175/1520-0485(1980)010<0843:TNOTPH>2.0.CO;2.
Smith, K. S., 2007: The geography of linear baroclinic instability in Earth’s oceans. J. Mar. Res., 65, 655–683, https://doi.org/10.1357/002224007783649484.
Souza, J. M. A. C., C. de Boyer Montégut, C. Cabanes, and P. Klein, 2011: Estimation of the Agulhas ring impacts on meridional heat fluxes and transport using ARGO floats and satellite data. Geophys. Res. Lett., 38, L21602, https://doi.org/10.1029/2011GL049359.
Stammer, D., 1997: Global characteristics of ocean variability estimated from regional TOPEX/POSEIDON altimeter measurements. J. Phys. Oceanogr., 27, 1743–1769, https://doi.org/10.1175/1520-0485(1997)027<1743:GCOOVE>2.0.CO;2.
Stammer, D., 1998: On eddy characteristics, eddy transports, and mean flow properties. J. Phys. Oceanogr., 28, 727–739, https://doi.org/10.1175/1520-0485(1998)028<0727:OECETA>2.0.CO;2.
Stewart, A. L., A. Klocker, and D. Menemenlis, 2018: Circum‐Antarctic shoreward heat transport derived from an eddy‐and tide‐resolving simulation. Geophys. Res. Lett., 45, 834–845, https://doi.org/10.1002/2017GL075677.
Sun, B., C. Liu, and F. Wang, 2019: Global meridional eddy heat transport inferred from Argo and altimetry observations. Sci. Rep., 9, 1345, https://doi.org/10.1038/s41598-018-38069-2.
Taylor, G. I., 1922: Diffusion by continuous movements. Proc. London Math. Soc., 20, 196–212, https://doi.org/10.1112/plms/s2-20.1.196.
Thompson, A. F., K. J. Heywood, S. Schmidtko, and A. L. Stewart, 2014: Eddy transport as a key component of the Antarctic overturning circulation. Nat. Geosci., 7, 879–884, https://doi.org/10.1038/ngeo2289.
Volkov, D. L., T. Lee, and L.-L. Fu, 2008: Eddy-induced meridional heat transport in the ocean. Geophys. Res. Lett., 35, L20601, https://doi.org/10.1029/2008GL035490.
Volkov, D. L., L.-L. Fu, and T. Lee, 2010: Mechanisms of the meridional heat transport in the Southern Ocean. Ocean Dyn., 60, 791–801, https://doi.org/10.1007/s10236-010-0288-0.
Walkden, G. J., K. J. Heywood, and D. P. Stevens, 2008: Eddy heat fluxes from direct current measurements of the Antarctic polar front in shag rocks passage. Geophys. Res. Lett., 35, L06602, https://doi.org/10.1029/2007GL032767.
Wang, X., S. Zhang, X. Lin, B. Qiu, and L. Yu, 2021: Characteristics of 3‐dimensional structure and heat budget of mesoscale eddies in the South Atlantic Ocean. J. Geophys. Res. Oceans, 126, e2020JC016922, https://doi.org/10.1029/2020JC016922.
Wang, Z., X. Zhang, Z. Guan, B. Sun, X. Yang, and C. Liu, 2015: An atmospheric origin of the multi-decadal bipolar seesaw. Sci. Rep., 5, 8909, https://doi.org/10.1038/srep08909.
Watts, D. R., K. L. Tracey, K. A. Donohue, and T. K. Chereskin, 2016: Estimates of eddy heat flux crossing the Antarctic Circumpolar Current from observations in Drake Passage. J. Phys. Oceanogr., 46, 2103–2122, https://doi.org/10.1175/JPO-D-16-0029.1.
Wunsch, C., 1999: Where do ocean eddy heat fluxes matter? J. Geophys. Res., 104, 13 235–13 249, https://doi.org/10.1029/1999JC900062.
Yang, H., B. Qiu, P. Chang, L. Wu, S. Wang, Z. Chen, and Y. Yang, 2018: Decadal variability of eddy characteristics and energetics in the Kuroshio Extension: Unstable versus stable states. J. Geophys. Res. Oceans, 123, 6653–6669, https://doi.org/10.1029/2018JC014081.
Young, I. R., and A. Ribal, 2019: Multiplatform evaluation of global trends in wind speed and wave height. Science, 364, 548–552, https://doi.org/10.1126/science.aav9527.
Zemskova, V. E., B. L. White, and A. Scotti, 2015: Available potential energy and the general circulation: Partitioning wind, buoyancy forcing, and diapycnal mixing. J. Phys. Oceanogr., 45, 1510–1531, https://doi.org/10.1175/JPO-D-14-0043.1.
Zhai, X., H. L. Johnson, D. P. Marshall, and C. Wunsch, 2012: On the wind power input to the ocean general circulation. J. Phys. Oceanogr., 42, 1357–1365, https://doi.org/10.1175/JPO-D-12-09.1.
Zhang, H., 2020: MITgcm model setup and output for “Improved representation of river runoff in ECCO simulations: Implementation, evaluation, and impacts to coastal plume regions”, version 1. Zenodo, accessed 16 October 2020, https://doi.org/10.5281/zenodo.4095613.
Zhang, Z., Y. Zhong, J. Tian, Q. Yang, and W. Zhao, 2014: Estimation of eddy heat transport in the global ocean from Argo data. Acta Oceanol. Sin., 33, 42–47, https://doi.org/10.1007/s13131-014-0421-x.
Zhang, Z., and Coauthors, 2023: Submesoscale inverse energy cascade enhances Southern Ocean eddy heat transport. Nat. Commun., 14, 1335, https://doi.org/10.1038/s41467-023-36991-2.
Zhu, Y., B. Qiu, X. Lin, and F. Wang, 2018: Interannual eddy kinetic energy modulations in the Agulhas return current. J. Geophys. Res. Oceans, 123, 6449–6462, https://doi.org/10.1029/2018JC014333.
Zhu, Y., Y. Li, F. Wang, and M. Lv, 2022: Weak mesoscale variability in the Optimum Interpolation Sea Surface Temperature (OISST)-AVHRR-only version 2 data before 2007. Remote Sens., 14, 409, https://doi.org/10.3390/rs14020409.
