Lagged effect of Southern Annular Mode on chlorophyll-a in the mid-latitude South Pacific and Indian Oceans

This study investigates the influence of the Southern Annular Mode (SAM) on chlorophyll-a (Chl-a) concentrations and the underlying mechanisms governing their associated environmental variations in the mid-latitude (35–50° S) ocean from 1998 to 2021. The intensification of westerly winds during positive SAM phases influences meridional water transport and mixed layer depth (MLD), which are both critical factors that affect surface nutrient availability. A marked contrast in the relationship between the meridional current anomaly and the SAM was observed, with reduced northward transport of nutrient-rich water in regions north of 50° S during positive SAM phases. This reduction could be attributed to the poleward migration of the westerly winds, which impeded the meridional current from reaching the mid-latitudes. The relationship between SAM and MLD south of 50° S was positive whereas that in the mid-latitude eastern (60–110° E) South Indian Ocean and eastern (90–140° W) South Pacific Ocean was negative or weak. The immediate effect of a more positive SAM on Chl-a in the mid-latitude ocean was reduced productivity caused by enhanced nutrient depletion. However, in the mid-latitude eastern South Pacific Ocean, the northward migration of the zonal mean meridional current anomaly closely aligned with the lagged correlation pattern between SAM variability and Chl-a over time, suggesting that the delayed northward transport of nutrient-rich waters may partially counterbalance the immediate effects of the SAM on ocean productivity. This mechanism was not present in the mid-latitude eastern South Indian Ocean, implying that future climate change may variably affect these regions. Our findings emphasize the importance of considering regional differences and temporal lags when evaluating the influence of SAM variability on ocean productivity and nutrient dynamics in the context of climate change.


Introduction
The Southern Annular Mode (SAM) dominates intraseasonal to interannual climate variability across the Southern Hemisphere (Thompson andWallace 2000, Thompson andSolomon 2002).It is characterized by a pattern of atmospheric circulation that involves changes in the intensity and location of westerly winds (Hartmann and Lo 1998).Given the significance of westerly winds in driving Southern Ocean circulation, their meridional shifts, which are associated with the SAM, can significantly alter the environments of the Southern Ocean and its connected regions (Hall andVisbeck 2002, Oke andEngland 2004).Goyal et al (2021) reported the influence of SAM variations and westerly winds on the amplified warming of western boundary current regions in the Southern Hemisphere.In the context of biogeochemical properties, the Southern Ocean overturning circulation is instrumental in transporting freshwater, nutrients, heat, and anthropogenic carbon dioxide (Rintoul et al 2001, Sarmiento et al 2004).
A strong SAM causes increased northward Ekman transport at the latitude of the southern Antarctic Circumpolar Current, inducing stronger upwelling that brings elevated concentrations of dissolved inorganic carbon to the surface (Gruber et al 2019).In response to the SAM, the mixed layer depth (MLD) in the Southern Ocean exhibits distinct variability from seasonal to interannual scales, as observed through a combination of satellite and in-situ measurements (Buongiorno Nardelli et al 2017).The strong phases of the SAM are associated with an intensification of ocean latent heat loss, leading to colder sea surface temperature (SST) and deeper MLD, whereas the opposite is observed during the weak phases of the SAM (Li et al 2019).Throughout recent decades, a discernible SAM intensification has been observed, denoted by its increasingly positive trend.This trajectory is widely attributed to two key factors: the degradation of the Antarctic ozone layer and escalating concentrations of greenhouse gases (Marshall 2003, Son et al 2008, Swart and Fyfe 2012).Although the impacts attributed to ozone layer depletion are projected to diminish, rapidly increasing greenhouse gas emissions may continue to drive the progressive strengthening of the SAM in the future (Swart et al 2018, Hartmann 2022, Li et al 2022).
As the Southern Ocean is the most significant sink for atmospheric carbon dioxide (Peylin et al 2005, Gruber et al 2019), assessing the influence of the SAM on its biological carbon sequestration is necessary.Its impact on marine productivity is subject to regional, temporal, and prevailing physical and biogeochemical conditions in the affected areas.The SAM is directly related to wind patterns; thus, alterations in wind intensity and direction can reshape surface currents and modulate vertical mixing intensity (Hartmann andLo 1998, Behrens andBostock 2023).For instance, an intensification of wind stress, that is, positive SAM, within the sub-Antarctic Front zone, which coincides with the Ekman convergence zone, can deepen the ocean mixing layer.This phenomenon is primarily driven by enhanced seawater convergence and turbulent mixing, which simultaneously amplify the nutrient supply but reduce light availability (Sallée et al 2010, Chapman et al 2020).Under such conditions, the effect of the SAM on phytoplankton productivity in regions under its influence can vary significantly depending on whether nutrient or light availability is a limiting factor (Kahru et al 2010).Supporting this notion, Noh et al (2021) demonstrated the zonally asymmetric response of phytoplankton biomass (as indicated by chlorophyll-a concentration; hereafter Chl-a) to the SAM, owing to varying limiting factors in different regions.This observation suggested contrasting relationships between westerly winds and Chl-a concentrations in the western and eastern Antarctic regions south of 58 • S, which are characterized by negative and positive correlations, respectively.
The SAM may have also contributed to the influx of tropical water into the mid-latitudes (Arblaster and Meehl 2006, Thompson et al 2011, Yang et al 2016).The western boundary regions of the South Pacific Ocean also show significant sensitivity to precipitation in response to the SAM (Fogt and Marshall 2020).An example of such an indirect impact of SAM-associated precipitation changes is the massive bloom of marine phytoplankton that occurred in the eastern South Pacific Ocean (Tang et al 2021).This phytoplankton bloom was triggered by atmospheric aerosol plumes that originated from large-scale wildfires in Australia in 2019-2020, which transported essential nutrients such as nitrogen and iron.During this period, the SAM was strongly negative (along with a positive Indian Ocean Dipole), creating hotter and drier conditions, and thus more favorable conditions for wildfire occurrences.
Previous studies relating the SAM and Chl-a on an interannual timescale have mainly been conducted in the marginal seas around Antarctica (Noh et al 2021) and near the sub-Antarctic zone centered on the Antarctic Circumpolar Current (Sallée et al 2008(Sallée et al , 2010)).In the Southern Hemisphere, biological productivity in mid-latitude oceans is influenced by the nutrients provided by the high-latitude Southern Ocean.This is because nutrient-rich deep waters upwell in the high-latitude Southern Ocean and are transported equatorward until they reach the Subtropical Front (Sarmiento et al 2004, Lauderdale et al 2013, Deppeler and Davidson 2017, Henley et al 2020).This equatorward movement of nutrient-rich waters is driven primarily by the prevailing westerly winds in the Southern Ocean because of the effect of Ekman transport within the surface Ekman layer (Lovenduski and Gruber 2005, Gruber et al 2019).
A recent study revealed that intensified or northward (weaker or southward) shifted westerly winds lead to enhanced northward Ekman transport over large parts of the Southern Ocean, resulting in a northward (southward) shift of the Subtropical Front (Behrens and Bostock 2023).While previous studies have been conducted on the influence of SAM on marine productivity in areas surrounding the Antarctic continent, research regarding the impact of SAM on productivity in northern waters beyond the Southern Ocean has been relatively scarce.Therefore, we extended the study area toward the mid-latitude ocean, which is the northern limb of the Southern Ocean, where the impact of the SAM is likely to be significant (Lovenduski and Gruber 2005).Specifically, this study investigated the relationship between SAM and Chl-a in the eastern South Pacific and Indian Oceans.These regions, located in subtropical areas beyond the sub-Antarctic zone, were selected because one of the main goals of this research is to analyze the influence of polar climate variations, including SAM, on the marine environment at mid-latitudes.Furthermore, the remoteness of these regions from continents helps minimize the influence of terrestrial sources, such as dust and wildfires.In this regard, our study could contribute to filling the knowledge gap concerning the connection between SAM variations and westerly winds over the Southern Ocean and biogeochemical alternations in the mid-latitude ocean.

Data and methods
We classified the SAM time series into two distinct phases, strong and weak years, to highlight their impacts on oceanic physical and biogeochemical conditions.To define the strong and weak SAM phases, empirical orthogonal function (EOF) was applied to the austral summer mean (December/January/February; DJF) geopotential height at 500 hPa over the Southern Hemisphere (30 • S-85 • S, 0-360 • E), acquired from the fifth generation of the ECMWF atmospheric reanalysis (ERA5) with a horizontal resolution of 1.5 • × 1.5 • (figure 1(a)).The first leading EOF mode (EOF1) represented the structure of the SAM, explaining approximately 50% of the total variation, and its standardized principal component time series was used as an index for the SAM (Hersbach et al 2020, Wachter et al 2020).Prior to EOF analysis, the geopotential height data were first detrended to remove long-term climate change effects, and then weights were applied based on the size of the grid cells.Based on the SAM index values, strong (1999, 2001, 2007, 2011, 2020) and weak (2000, 2003, 2005, 2009, 2016, 2019) SAM years were defined.The former are years when the SAM index is greater than +1, and the latter are years when it is less than −1 (figure 1(b)).
We compared the composited values of strong and weak SAM and examined the relationship between Chl-a, meridional ocean current, and MLD with the SAM index within each group.Chl-a concentrations were obtained from monthly datasets (horizontal resolution: 0.25 • × 0.25 • ) provided by the European Service for Ocean Color (www.globcolour.info); in this study, we used the common logarithm (base 10) of Chl-a.The monthly mean ocean currents and MLD datasets (horizontal resolution: 1 • × 0.3 • ) were provided by the National Center for Environmental Prediction Global Ocean Data Assimilation System (Behringer and Xue 2004).Given the recognized uncertainties in MLD data (Toyoda et al 2017), we conducted a verification analysis using the independent MLD data obtained from EN4 ocean analysis dataset provided by the UK Met Office (www.metoffice.gov.uk/hadobs/en4)(Good et al 2013).To ensure compatibility among datasets with different spatial resolutions, all data were re-gridded to a uniform spatial resolution of 1.5 • × 1.5 • .As satellitebased Chl-a data were available between 1998 and 2021, we chose this period for the study.
Linear regression and correlation analyses were applied to determine the relationships between the SAM index and the other variables (Chl-a, ocean current, MLD, and wind).To evaluate the significance of these statistics, we employed a bootstrap method, performing 1000 resamplings of the data.The linear regression coefficients indicate a change in the variables for a unit change in the SAM index.We compared the composited values of strong and weak SAM years and examined the relationship of Chl-a, meridional ocean current, MLD, and winds with the SAM index within each group.Moreover, we performed a correlation analysis that accounted for lag effects, spanning a duration of up to six months, to investigate the lagged responses to SAM-induced variability.

Influence of SAM on oceanic conditions
To the south of approximately 50 • S, the Southern Ocean exhibited regionally asymmetric correlations between the SAM and Chl-a, with a prominent negative correlation observed in the western Antarctic Amundsen-Ross Sea region and a positive correlation over East Antarctic waters (figure 1(c)).These findings are consistent with the results reported by Noh et al (2021).The study determined that while the overall intensification of westerly winds associated with a positive SAM partially alleviated nutrient constraints, the Amundsen-Ross Sea region exhibited a decline in Chl-a due to diminished light availability stemming from increased sea ice formation.However, to the north of the region, exhibiting an asymmetrical pattern, another extensive and consistent negative correlation between the latitudes of 35 • S-50 • S was identified.To elucidate the underlying mechanisms responsible for the broad-scale contrast observed across the 50 • S boundary, we investigated the meridional ocean current, MLD, and winds, all of which can significantly influence surface nutrient availability within these oceanic domains.
North and south of the 50 • S boundary, a marked contrast was observed in the relationship between the meridional current anomaly and the SAM index (figures 2(c) and (d)).Specifically, north of 50 • S, an increase in the strength of the SAM corresponded to a reduced northward water flow (negative anomaly) compared with the average condition, whereas the opposite phenomenon was detected towards the Antarctic continent from 50 • S. The reduced northward transport of nutrient-rich water, as revealed by the relationship between the SAM index and meridional current anomalies, may provide a plausible explanation for the negative correlations observed between the SAM index and the Chl-a anomaly over the mid-latitude ocean (figures 2(a) and (b)), including our primary focus areas in the eastern South Pacific Ocean and eastern South Indian Ocean.These patterns can be attributed to the poleward migration of the westerly winds, which typically occur during positive SAM phases (figures 2(g) and (h)).Specifically, the strengthening of the westerly winds and the accompanying southward displacement of the upwelling zone may have impeded the meridional current from reaching the mid-latitudes.Several studies have reported physical environmental changes (e.g.SST) in mid-latitude regions caused by variations in westerly winds (Goyal et al 2021, Hartmann 2022).According to Hartmann (2022), the strengthening of the westerly winds due to the ozone hole has contributed to the synchronous temporal cooling of the Southern Ocean and the eastern tropical South Pacific Ocean since the 1980s.The SAMrelated winds can directly drive surface temperature variability through air-sea heat fluxes via oceanic latent heat exchange, and indirectly through winddriven circulation (Li et al 2019).A positive SAM phase is associated with a significant increase in SST in the mid-latitude Southern Ocean, except in the Pacific sector, consistent with the southward anomaly in Ekman transport (Lovenduski and Gruber 2005).Conversely, the positive SAM phase corresponds to negative SST anomalies in the Pacific sector, which may be due to other Southern Ocean climate variations that offset the impact of SAM (White and Peterson 1996, Garreaud and Battisti 1999, Liu et al 2002, Li et al 2019).In accordance with these studies, our results show that a positive SAM is associated with an increase in SST in the eastern South Indian Ocean and a decrease in SST in the eastern South Pacific Ocean (figure S1).
While the meridional current supplies nutrients through horizontal advection, the MLD in the surface layer is a critical determinant of the vertical entrainment of nutrients from deep to shallow waters.As the MLD is subjected to direct wind forcing, it likely exhibits variations associated with the SAM.Although the relationship between the SAM and MLD was less clear than that observed between the SAM and Chl-a or meridional current, a positive association between the SAM and MLD was observed south of 50 • S.However, the strength of this relationship varied between the eastern South Indian Ocean and eastern South Pacific Ocean at mid-latitudes.In the eastern South Indian Ocean, the MLD anomaly tended to become negative during positive SAM phases owing to weakened westerly winds, varying by tens to hundreds of meters.In contrast, in the mid-latitude eastern South Pacific Ocean, both positive and negative MLD anomalies coexisted during both strong and weak SAM years, and the spatial heterogeneity was relatively high (figures 2(e) and (f)).The MLD in the ocean reanalysis data may have biases due to both model errors and discrepancies associated with differences between assimilation schemes (Buongiorno Nardelli et al 2017, Toyoda et al 2017); therefore, we repeated the calculation of MLD composite differences using the EN4 ocean analysis dataset (figure S2) and confirmed that both datasets exhibit identical spatial patterns in the study region during the strong and weak phases of the SAM.
During strong SAM years, the reduction in the northward transport of nutrient-rich water in the eastern South Indian Ocean likely contributed to a decrease in Chl-a (or a negative Chl-a anomaly).Conversely, the decrease in MLD under the same conditions and in the same region should have increased Chl-a from a light-limitation perspective and decreased it from a nutrient-limitation perspective.Ultimately, the negative Chl-a anomaly in this region during positive SAM years suggests that changes in light conditions had a minimal impact, and the area could be considered a region of nutrient depletion.In the eastern South Pacific Ocean, where MLD changes are relatively weak and not homogenous in response to SAM variability, the northward transport of nutrients appeared to be a key factor in determining ocean productivity.Therefore, the immediate consequence of the anticipated increase toward more positive SAM conditions due to future climate change is expected to be an overall reduction in ocean productivity within the range of 35 • -50 • S.
The nutrient-rich Antarctic Circumpolar Current is divided into two branches that move northward and southward.The strengthening of westerly winds during positive SAM phases has been observed to increase the upwelling of this current.Thus, even if the westerly winds migrated southward, delayed northward nutrient transport associated with strengthened upwelling could occur.These delayed effects, which are linked to SAM variability, may partially counterbalance the immediate effects of the SAM on ocean productivity in mid-latitude oceans.

SAM variability and lagged effects on Chl-a
Through a lagged correlation analysis in the two research areas of the Indian and Pacific Oceans, we determined whether lagged effects on Chl-a were evident in response to SAM variability (figures 3, 4 and S3).Lagged effects of up to two months were consistently observed in both study areas.The association between SAM variability and the Chl-a anomaly demonstrated a diminishing trend in both regions as the lag increased to two months (figure 3).However, the patterns in the two regions visibly diverged for lags exceeding three months.In the eastern South Indian Ocean, the relationship between SAM variability and the Chl-a anomaly persisted until two months (figures 5 and S3).Conversely, in the eastern South Pacific Ocean study area, the regression coefficients became positive for lags surpassing three months, with the maximum regression coefficient detected at a four-month lag (figures 5 and S4-S7).To investigate the cause of the lagged increase in Chl-a in the South Pacific Ocean study area, we analyzed temporal changes in the zonal mean meridional current anomaly, considering the possibility of delayed northward transport.In the strong SAM year group during the DJF period, when the direct influence of SAM variability was examined, a positive meridional current anomaly developed in the surface layer (0-50 m) of the oceanic region south of approximately 60 • S in the zonal band between  6).As time progressed, the center of this positive anomaly gradually moved northward, reaching its northernmost position at approximately three or four months.After that, the positive zonal mean meridional current anomaly rapidly weakened.The horizontal patterns of the lag regressions between the SAM and zonal winds demonstrated the northward extension of the zonal winds at a lag of 2-3 months in the eastern south Pacific Ocean, which spans from 70 • S to 40 • S (figure S7).The stronger zonal winds can intensify meridional ocean currents by equatorward Ekman transports.Over time, its northward migration closely aligned with the lagged correlation pattern observed between the SAM index and the Chl-a anomaly in the South Pacific Ocean.In addition, during weak SAM years in the same area, the negative zonal mean meridional current anomaly exhibited a pattern of northward movement that disappeared five months after the DJF period.As the lag month increases to +3 months, the areas showing a positive relationship between anomalies in the meridional current and zonal wind speed with the SAM index are gradually shifting northward (figures 5(b), (d), S5 and S7).During periods of a strong SAM, the delayed northward propagation of the intensification of eastward wind speed appears to have enhanced the equatorward Ekman transport.
In the South Indian Ocean study area (spanning 60 • E-110 • E), during the DJF interval, a positive zonal mean meridional current anomaly manifested in the surface layer (0-50 m) of the Southern Ocean south of 60 • S during strong SAM years (figure S8).However, this positive anomaly failed to migrate northward over time, with traces almost disappearing after three months.Concurrently, within the same time frame and region, a negative zonal mean meridional current anomaly developed in the oceanic domain between 20 • S and 50 • S, separating it from the South Pacific Ocean study area.The hindered northward progression of the positive anomaly south of 60 • S in the eastern South Indian Ocean was likely due to the presence of a negative anomaly within the mid-latitude region.Based on the precise correspondence between the presence/absence of the northward movement of the positive meridional current anomaly and the presence/absence of a lagged correlation, we ascertained that during positive SAM phases, the deferred northward transport of nutrientrich waters originating in the southern part of the Southern Ocean may lead to a subsequent enhancement of productivity within the mid-latitude eastern South Pacific Ocean.The absence of such a mechanism in the eastern South Indian Ocean implies that SAM variability under future climate change conditions may have divergent impacts on the eastern South Pacific and eastern South Indian Oceans.The intensification of zonal winds at 60 • S, associated with the SAM, is known to influence the influx of cold water into the Peru Current, which flows along the west coast of South America (Liau and Chao 2017).This intrusion of cold water may be related to the northward propagation of the zonal mean meridional current anomaly, as found in our study.
We investigate the impacts of wind stress curl, which can influence nutrient supply from the subsurface through the process of Ekman suction or pumping, potentially modulating Chl-a concentrations.In the eastern South Pacific Ocean, during periods of strong SAM, we observed negative wind stress curl at zero lag, indicating Ekman divergence (figure S9(a)).This implies favorable conditions for the upwelling of nutrients.However, Chl-a concentrations decreased during this period (figure 4(a)).Additionally, at a lag of 3-4 months, the sign of wind stress curl changed to positive, indicating downwelling-favorable conditions (figures S9(d) and (e)), while Chl-a concentrations increased.In the eastern South Indian Ocean, specifically near 40 • S where we observed a significant decrease in Chl-a during the strong SAM (figure 4(a)), there was a significant negative wind stress curl (figure S9(a)), suggesting upwelling favorable conditions.This relationship suggests that the local effects of wind stress curl-induced Ekman suction or pumping are unlikely to be the primary drivers of ocean productivity in the mid-latitude Southern Ocean, and supports our conclusion that meridional Ekman transport plays a more significant role in explaining the delayed Chl-a peaks with lags.

Conclusions
We discovered a consistent negative correlation between the SAM index and Chl-a anomaly across mid-latitude oceans, which can be ascribed to the poleward shift of the westerly winds and associated southward movement of the upwelling zone during positive SAM phases, as well as the reverse effects during negative SAM phases.We conducted a lagged correlation analysis to investigate the delayed impacts of SAM variability on ocean productivity in the eastern South Indian and eastern South Pacific Ocean regions.Although both regions consistently exhibited no lagged effects for up to two months, their patterns diverged significantly for lags beyond three months.In the eastern South Pacific Ocean region, the northward progression of the zonal mean meridional current anomaly corresponded closely with the lagged correlation pattern observed between the SAM index and Chl-a anomaly, indicating that the delayed northward transport of nutrient-rich waters could partially offset the immediate negative effects of the SAM on ocean productivity in this area during positive SAM phases.Our research underscores the necessity of accounting for regional disparities and temporal lags when evaluating the influence of SAM variability on ocean productivity as well as the potential implications of future climate change on nutrient dynamics and productivity within the Indian and Pacific Oceans.

Figure 1 .
Figure 1.(a) Spatial pattern of the first leading EOF mode (EOF1) estimated using the geopotential height at 500 hPa in the Southern Ocean from 1998 to 2021 (December to February only).(b) Southern Annular Mode (SAM) index, a standardized and detrended principal component time series of EOF1.Red and blue dots represent strong and weak SAM years, respectively.(c) Map of correlation coefficients between anomalies of log10 [Chl-a (mg m −3 )] and the SAM index during austral summer (December-February) from 1998 to 2021.Dots on the map indicate statistically significant correlation values at a 90% confidence level.The two pink boxes represent study areas of interest in the mid-latitudes.

Figure 3 .
Figure 3. Areal-averaged lagged correlation coefficients between the SAM index and log10 [Chl-a (mg m −3 )].Values were averaged over two study areas during the austral summer from 1998 to 2021.The lag correlation coefficient at each grid was calculated by comparing the SAM index with the time series of log10 [Chl-a (mg m −3 )] shifted forward by 1-6 months.Figure S3 shows the spatial distribution of lagged correlation coefficients.The error bars on each bar indicate the 95% confidence intervals for the mean values.

Figure 4 .
Figure 4. Regression maps of anomalies in (a), (b) log10 [Chl-a (mg m −3 )], (c), (d) meridional current (m s −1 ), (e), (f) mixed layer depth (m), and (g), (h) surface zonal wind (m s -1 ) on the SAM index.The left panels are for the onset (a), (c), (e), and (g) and the right panels (b), (d), (f), and (h) are for those shifted forward by 4 months.Dots on the map indicate statistically significant regression coefficients at a 95% confidence level.The two pink boxes represent study areas of interest in the mid-latitude.

Figure 5 .
Figure 5. Areal-averaged lagged regression coefficients for the period of 1998-2021 of anomalies in (a) log10 [Chl-a (mg m −3 )], (b) meridional current (m s −1 ), (c) mixed layer depth (m), and (d) surface zonal wind (m s −1 ) on the austral summer SAM index.The values were averaged over two study areas.Positive lag means that the log10 [Chl-a (mg m −3 )], meridional current (m s −1 ), mixed layer depth (m), and surface zonal wind (m s -1 ) are shifted forward by 1-6 months.Figures S4-S7 show the spatial distribution of lagged regression coefficients.The error bars on each bar indicate the 95% confidence intervals for the mean values.

Figure 6 .
Figure 6.Zonal mean meridional current anomaly (m s −1 ) averaged between 140 • W and 90 • W of the eastern South Pacific Ocean during austral summer (DJF) for (a) strong and (b) weak SAM years.Each latitude-depth panel, arranged from top to bottom and right, shows the monthly evolution of the zonal mean meridional current anomaly for six months.Dots on the map indicate statistically significant anomalies at a 90% confidence level.
21 3020-39 Sarmiento J L, Gruber N, Brzezinski M A and Dunne J P 2004 High-latitude controls of thermocline nutrients and low latitude biological productivity Nature 427 56-60 Son S, Polvani L M, Waugh D W, Akiyoshi H, Garcia R, Kinnison D, Pawson S, Rozanov E, Shepherd T G and Shibata K 2008 The impact of stratospheric ozone recovery on the Southern Hemisphere westerly jet Science 320 1486-9 Swart N C and Fyfe J C 2012 Observed and simulated changes in the Southern Hemisphere surface westerly wind-stress Geophys.Res.Lett.39 L16711 Swart N C, Gille S T, Fyfe J C and Gillett N P 2018 Recent Southern Ocean warming and freshening driven by greenhouse gas emissions and ozone depletion Nat.Geosci.11 836-41 Tang W et al 2021 Widespread phytoplankton blooms triggered by 2019-2020 Australian wildfires Nature 597 370-5 Thompson D W J, Solomon S, Kushner P J, England M H, Grise K M and Karoly D J 2011 Signatures of the Antarctic ozone hole in Southern Hemisphere surface climate change Nat.Geosci.4 741-9 Thompson D W J and Wallace J M 2000 Annular modes in the extratropical circulation Part I: month-to-month variability J. Clim.13 1000-16 Thompson D W and Solomon S 2002 Interpretation of recent Southern Hemisphere climate change Science 296 895-9 Toyoda T et al 2017 Intercomparison and validation of the mixed layer depth fields of global ocean syntheses Clim.Dyn.49 753-73 Wachter P, Beck C, Philipp A, Höppner K and Jacobeit J 2020 Spatiotemporal variability of the Southern Annular Mode and its influence on Antarctic surface temperatures J. Geophys.Res.Atmos.125 e2020JD033818 White W B and Peterson R G 1996 An Antarctic circumpolar wave in surface pressure, wind, temperature and sea-ice extent Nature 380 699-702 Yang H, Lohmann G, Wei W, Dima M, Ionita M and Liu J 2016 Intensification and poleward shift of subtropical western boundary curents in a warming climate J. Geophys.Res.Oceans 121 4928-45 Hall A and Visbeck M 2002 Synchronous variability in the Southern Hemisphere atmosphere, sea ice, and ocean resulting from the annular mode J. Clim.15 3043-57 Hartmann D L 2022 The Antarctic ozone hole and the pattern effect on climate sensitivity Proc.Natl Acad.Sci.119 e2207889119 Hartmann D L and Lo F 1998 Wave-driven zonal flow vacillation in the Southern Hemisphere J. Atmos.Sci.55 1303-15 Henley S F et al 2020 Changing biogeochemistry of the Southern Ocean and its ecosystem implications Front.Mar.Sci. 7 581 Hersbach H et al 2020 The ERA5 global reanalysis.Q.J. R. Lim H G and Kug J S 2021 Zonally asymmetric phytoplankton response to the Southern annular mode in the marginal sea of the Southern ocean Sci.Rep. 11 10266 Oke P R and England M H 2004 Oceanic response to changes in the latitude of the Southern Hemisphere subpolar westerly winds J. Clim.17 1040-54 Peylin P, Bousquet P, Le Quéré C, Sitch S, Friedlingstein P, McKinley G, Gruber N, Rayner P and Ciais P 2005 Multiple constraints on regional CO2 flux variations over land and oceans Glob.Biogeochem.Cycles 19 GB1011 Rintoul S R, Hughes C W and Olbers D 2001 The Antarctic Circumpolar Current system Int.Geophys.77 271-302 Sallée J B, Speer K G and Rintoul S R 2010 Zonally asymmetric response of the Southern Ocean mixed-layer depth to the Southern Annular Mode Nat.Geosci.3 273-9 Sallée J B, Speer K and Morrow R 2008 Response of the Antarctic Circumpolar Current to atmospheric variability J. Clim.