Identifying the biological control of the annual and multi-year variations in South Atlantic air-sea CO 2 flux

. The accumulation of anthropogenic CO 2 emissions in the atmosphere has been buffered by the absorption of CO 2 by the global ocean which acts as a net CO 2 sink. The CO 2 flux between the atmosphere and the ocean, that collectively results in the oceanic carbon sink, is spatially and temporally variable, and fully understanding the driving mechanisms behind this flux is key to assessing how the sink may change in the future. In this study a time series decomposition analysis 10 was applied to satellite observations to determine the drivers that control the sea-air difference of CO 2 partial pressure (Δ p CO 2 ) and the CO 2 flux on seasonal and interannual time scales in the South Atlantic Ocean. Linear trends in Δ p CO 2 and the CO 2 flux were calculated to identify key areas of change. Seasonally, changes in both the Δ p CO 2 and CO 2 flux were dominated by sea surface temperature (SST) in the subtropics (north of 40 o S) and were correlated with biological processes in the subpolar regions (south of 40° S). In the Equatorial 15 Atlantic, analysis of the data indicated that biological processes are likely a key driver, as a response to upwelling and riverine inputs. These results highlighted that seasonally Δ p CO 2 can act as an indicator to identify drivers of the CO 2 flux. Interannually, the SST and biological contributions to the CO 2 flux in the subtropics were correlated with the Multivariate ENSO Index (MEI) which leads to a weaker (stronger) CO 2 sink in El Niño (La Niña) years. The 16-year time-series identified significant trends in Δ p CO 2 and CO 2 flux, however, these trends were not always 20 consistent in spatial extent. Therefore, predicting the oceanic response to climate change requires the examination of CO 2 flux rather than Δ p CO 2 . Positive CO 2 flux trends (weakening sink for atmospheric CO 2 ) were identified within the Benguela upwelling system, consistent with increased upwelling and wind speeds. Negative trends in the CO 2 flux (intensifying sink for atmospheric CO 2 ) offshore into the South Atlantic Gyre, were consistent with an increase in the export of nutrients from mesoscale features, which drives the biological drawdown of CO 2 . These multi-year trends in the CO 2 flux indicate that the 25 biological contribution to changes in the air-sea CO 2 flux cannot be overlooked when scaling up to estimates of the global ocean carbon sink. the seasonal and interannual drivers of Δ p CO 2 and the air-sea CO 2 flux in the using satellite observations. Seasonally, our results indicated that the subtropics were controlled by SST, and the subpolar regions were correlated with biological processes. Deviations from this trend occurred in the Benguela upwelling where predominately biological processes correlated with variability in the Δ p CO 2 as well as upwelling. The Equatorial Atlantic showed spatially variable drivers associated with the Amazon Plume and Equatorial upwelling which induced a


Introduction
Since the industrial revolution, anthropogenic CO2 emissions have increased unabated and continue to rise atmospheric CO2 concentrations (IPCC, 2021). The global oceans have buffered the rise by acting as a sink for atmospheric CO2 at a rate of 30 between 1 and 3.5 Pg C yr -1 (e.g. Friedlingstein et al., 2020;Landschützer et al., 2014;Watson et al., 2020). The strength of the ocean as a sink for CO2 appears to be increasing with time (Friedlingstein et al., 2020;Watson et al., 2020). Regionally this can vary hugely however, and the ocean can oscillate between a source or sink of atmospheric CO2. The difference in the partial pressure of CO2 (pCO2) between the seawater and atmosphere (ΔpCO2) is used as an indicator or proxy for the net direction of air-sea CO2 flux during gas exchange. 35 In the open ocean, changes in physical and biogeochemical processes that control seawater pCO2 (pCO2 (sw)) also modify ΔpCO2 as the atmospheric pCO2 (pCO2 (atm)) is less variable (e.g. Henson et al., 2018;Landschützer et al., 2016). ΔpCO2 can therefore be controlled by changes in sea surface temperature (SST), because the pCO2 is proportional to the temperature. In addition, plankton net community production (NCP) modifies the concentration of CO2 in the seawater depending on the balance between net primary production (NPP; uptake of CO2 via photosynthesis) and respiration (release of CO2 into the 40 water). The NCP describes the overall metabolic balance of the plankton community, where positive (negative) NCP indicates a drawdown (or release) of CO2 from (or into) the water contributing to a decrease (increase) in ΔpCO2. Physical processes, including riverine input (e.g. Ibánhez et al., 2016;Lefèvre et al., 2020;Valerio et al., 2021), and upwelling (e.g. González-Dávila et al., 2009;Lefèvre et al., 2008;Santana-Casiano et al., 2009) can alter pCO2 (sw) and ΔpCO2 directly through the entrainment of high-CO2 water or indirectly by modifying NCP through nutrient supply (enhancing 45 photosynthesis) and/or organic material supply (enhancing respiration).
The air-sea CO2 flux is more precisely a function of the difference in CO2 concentrations across the mass boundary layer at the ocean's surface, with any turbulent exchange characterised by the gas transfer velocity. The CO2 concentration difference is determined by the pCO2 at the base (pCO2 (sw)) and top (pCO2 (atm)) of the mass boundary layer and the respective solubilities (Weiss, 1974), and must be carefully calculated due to vertical thermo-haline gradients existing across the mass 50

Air-sea CO2 flux data
The air-sea CO2 flux (F) can be estimated using a bulk parameterisation as: Where k is the gas transfer velocity which was estimated from ERA5 monthly reanalysis wind speed (Hersbach et al., 2019) 95 following the parameterisation of Nightingale et al. (2000). αw and αs are the solubility of CO2 at the base and top of the mass boundary layer at the sea surface (Woolf et al., 2016). αw was calculated as a function of SST and sea surface salinity (SSS) (Weiss, 1974) using the monthly Optimum Interpolated SST (Reynolds et al., 2002) and SSS from the Copernicus Marine Environment Modelling Service global ocean physics reanalysis product (GLORYS12V1; CMEMS, 2021). αs was calculated using the same temperature and salinity datasets but included a gradient from the base to the top of mass boundary 100 layer of -0.17 K (Donlon et al., 1999) and +0.1 salinity units (Woolf et al., 2016). pCO2 (atm) was calculated using the dry mixing ratio of CO2 from the NOAA-ESRL marine boundary layer reference, Optimum Interpolated SST (Reynolds et al., 2002) applying a cool skin bias (0.17K; Donlon et al., 1999) and sea level pressure following Dickson et al. (2007).
All of these calculations along with the resulting monthly CO2 flux were carried out using the open source FluxEngine toolbox (Holding et al., 2019;Shutler et al., 2016), for the period between July 2002 and December 2018, assuming 'rapid' 105 transfer (as described in Woolf et al., 2016).

Seasonal and interannual driver analysis
The X-11 analytical econometric tool (Shiskin et al., 1967) was used to decompose the timeseries into seasonal, interannual and residual components following the methodology of Pezzulli et al. (2005). In brief, the X-11 method comprises a three step filtering algorithm: (1) The interannual component (Tt) is initially estimated using an annual centred running mean, 125 which is subtracted from the initial timeseries (Xt) to estimate the seasonal component (St). (2) Tt is revised by applying an annual centred running mean to the Xt minus St. The revised Tt is removed from Xt and the final St calculated. (3) The final Tt is calculated by applying an annual centred running mean to Xt minus the revised St. The analysis has been shown to be effective in the decomposition of environmental time-series (Pezzulli et al., 2005;Vantrepotte & Mélin, 2011;Henson et al., 2018), that allows the seasonal cycle to vary on a yearly basis and, produces an interannual component that results in a 130 robust representation of the longer-term changes in the timeseries.
The approach was applied to monthly 1° fields of ΔpCO2 that were estimated from pCO2 (atm) and SA-FNN pCO2 (sw), on a per pixel basis. The pCO2 (atm) and spatially and temporally varying pCO2 (sw) uncertainties (Table 1; Fig. B1) were propagated through the X-11 analysis, using a Monte Carlo uncertainty propagation approach. The input time series were randomly perturbed 1000 times within the uncertainties of each parameter, and Spearman correlations calculated for each perturbation. 135 The 95% confidence interval was extracted from the resulting distribution of correlations coefficients, and results were deemed significant (α < 0.05) where the confidence interval remained significant. Spatial autocorrelation was tested using the method of field significance (Wilks, 2006). The analysis was then conducted on the CO2 fluxes, on a per pixel basis. The pCO2 (sw), pCO2 (atm), gas transfer velocity, SST and SSS uncertainties (Table 1) were propagated through the flux calculations using the same Monte Carlo uncertainty propagation approach used for ΔpCO2. 140 The potential drivers tested were MODIS-A skin SST, NCP and NPP alongside SSS from the CMEMS global reanalysis product (GLORYSV12; CMEMS, 2021) and two climate indices: Multivariate ENSO Index (MEI) as an indicator of El Niño Southern Oscillation phases, https://www.esrl.noaa.gov/psd/enso/mei (last accessed: 19/12/2019); Southern Annular Mode (SAM) data, which indicate the displacement of the westerly winds in the Southern Ocean, were downloaded from http://www.nerc-bas.ac.uk/icd/gjma/sam.html (last accessed: 19/12/2019). 145

Trend analysis
The linear trend in the interannual components of ΔpCO2 and the CO2 flux were calculated on a per pixel basis using the non 150 parametric Mann-Kendall test (Kendall, 1975;Mann, 1945) and Sen's Slope estimates (Sen, 1968), which are less sensitive to outliers in the timeseries. The input parameter uncertainties (Table 1) were propagated within this trend analysis using a Monte Carlo uncertainty propagation (n = 1000) to extract the 95% confidence interval on the trends. The overall trend was deemed significant if 95% of the trends were significant (α = 0.05), and the uncertainties in these trends are displayed in Appendix B (Fig. B2). 155

Limitations
It should be noted that correlations between the ΔpCO2 and SST/NCP are expected since the SA-FNN estimates pCO2 (sw) (the major determinant of ΔpCO2 variability) using SST and NCP as input parameters which are subsequently interpreted as drivers here. By extension, but to a lesser extent, this also applies to correlations between CO2 flux and SST/NCP since pCO2 (sw) is included in the flux calculations. Different lines of evidence suggest that this is not a major limitation of our study. 160 Firstly, any correlation between ΔpCO2/CO2 flux and SST/NCP is not determined a priori, but is an emerging property of the SA-FNN. Therefore, the driver analysis undertaken here represents an indirect decomposition of the SA-FNN drivers rather than a strict correlation analysis between independent variables. The accurate representation of seasonal pCO2 (sw) cycles across the South Atlantic Ocean (Ford et al., 2022) provides confidence in the SA-FNN. Secondly, conducting the analysis described by Henson et al. (2018) using in situ pCO2 (sw) to estimate ΔpCO2 on a per province basis (Longhurst, 1998) for the 165 South Atlantic Ocean, yielded similar seasonal drivers to the SA-FNN (Appendix A). The interannual drivers displayed some differences however, which may be due to the spatial and temporal averaging that is required to construct the in situ timeseries.

Seasonal drivers of ΔpCO2 and CO2 flux 170
The X-11 analysis conducted on ΔpCO2 indicated significant seasonal correlations (Fig. 1), when the uncertainties are accounted for. The subtropics (10° S to 40° S) showed positive correlations between ΔpCO2, SST and SSS (Fig. 1c, d), as well as negative correlations between ΔpCO2, NCP and NPP (Fig. 1a, b). In contrast the subpolar (south of 40° S) and equatorial regions (10° N to 10° S) displayed negative correlations between ΔpCO2 and SST (Fig. 1c). Correlations between ΔpCO2 and NCP were negative in the subpolar regions and were positive in the Equatorial regions (Fig. 1a). There were no 175 significant correlations observed between ΔpCO2 and MEI or SAM in any of the regions.

Multivariate ENSO index (MEI) and (f) Southern Annular Mode (SAM) seasonal components. White regions indicate no significant correlations, and green regions indicate no analysis was performed due to missing satellite data.
Regional deviations were observed in the Amazon Plume, Benguela upwelling, the South American coast, and a band across Negative correlation between ΔpCO2 and SSS, and positive correlations between NCP, NPP and ΔpCO2 were also observed 190 in the southwestern Atlantic (Fig. 1e). Positive correlations between NCP, NPP and ΔpCO2 were identified in a band across 40° S (Fig. 1a, b). Performing the X-11 analysis on the CO2 flux revealed similar and comparable correlations to ΔpCO2 ( Fig. 2). Significant driver-flux correlations were observed over a larger area however, compared to ΔpCO2.

Interannual drivers of ΔpCO2 and CO2 flux 200
The X-11 analysis identified regionally significant interannual correlations between ΔpCO2 and SST, MEI and to a lesser extent NCP and SSS (Fig. 3). The subtropics displayed positive correlations between SST and ΔpCO2, which extended across the basin from the South American coast (Fig. 3c). Positive correlations were also observed between the MEI and ΔpCO2 (Fig. 3e), with a similar geographic extent as the correlations with SST. In the central South Atlantic gyre spatially variable negative correlations between NCP and ΔpCO2, and positive correlations between SSS and ΔpCO2 were observed 205 ( Fig. 3a, d). The central Equatorial Atlantic displayed spatially variable positive correlations between NCP and ΔpCO2, which extended south-east towards the African coast (Fig. 3a).

(e) Multivariate ENSO index (MEI) and (f) Southern Annular Mode (SAM) interannual components. White regions indicate no significant correlations, and green regions indicate no analysis was performed due to missing satellite data.
Significant interannual correlations for the CO2 flux were also identified by the X-11 analysis (Fig. 4), which generally covered a larger spatial area to the corresponding ΔpCO2 correlations (Fig. 3). Positive correlations between the CO2 flux 215 and SST were observed in the subtropics (Fig. 4c), consistent with the correlations with ΔpCO2 (i.e. by comparing Fig. 4c and Fig. 3c). Nevertheless, negative correlations between the CO2 flux and SST were observed at the border between the equatorial region and subtropics, which was not identified in the ΔpCO2 correlations. Negative correlations between NCP and the CO2 flux were also identified over a spatially larger area (Fig. 4a, 3a). Correlations between the MEI and CO2 flux were positive in the subtropics (Fig. 4e) and included a band of negative correlations to the south between 35° S and 45° S 220 Positive correlations between NCP and CO2 flux were observed in the western equatorial Atlantic, alongside spatially variable negative correlations to SST (Fig. 4a, c). Positive correlations between SSS and CO2 flux were identified in the region of the Amazon plume (Fig. 4d). Weak positive correlations between the SAM and CO2 flux were identified between 30° S and 45° S (Fig. 4f). 225

Trends in interannual ΔpCO2 and CO2 flux
Regions of significant trends in the interannual component of ΔpCO2 were observed (Fig. 5a). Negative trends occurred in the South Atlantic gyre. Positive trends in ΔpCO2 were identified along the South African coast, which switched to strong negative trends moving offshore into the central South Atlantic gyre. Positive trends were also observed in the Equatorial Atlantic consistent with the positions of the Amazon Plume and Equatorial Upwelling.

Seasonal drivers of ΔpCO2 and CO2 flux
Previous studies have explored the seasonal drivers of ΔpCO2 and to a lesser extent the air-sea CO2 flux. In this study, we 245 investigated the drivers of ΔpCO2 and CO2 flux at both seasonal and interannual timescales in the South Atlantic Ocean. In the North Atlantic, Henson et al. (2018) indicated that the seasonal variability in subtropical ΔpCO2 variability is driven by SST, whereas the variability in ΔpCO2 in subpolar regions is biologically driven, similar to previous studies (Takahashi et al., 2002;Landschützer et al., 2013). The X-11 analysis conducted on spatially complete ΔpCO2 and CO2 flux displayed consistent seasonal results (Fig. 1, 2), though for the CO2 flux significant correlations occupied a larger area. These both 250 indicated a similar pattern in seasonal drivers for the South Atlantic Ocean, with subtropical ΔpCO2 and CO2 flux driven by SST, and subpolar correlated with biological controls, although the equatorial region exhibited more complex patterns (Fig.   1).
In the Equatorial Atlantic, the correlations between ΔpCO2, SST and biological production were spatially variable (Fig. 1). Landschützer et al. (2013) suggested that the temperature and non-temperature (i.e. biological and circulation) drivers 255 generally compensated each other. We found positive correlations between the NCP, ΔpCO2 and CO2 flux seasonal components, indicating that biological activity is likely a key driver of seasonal variability in response to the equatorial upwelling. Ford et al. (2022) showed that the SA-FNN improved the seasonal pCO2 (sw) variability in the Equatorial Atlantic compared to the current 'state of the art' SOM-FNN methodology (Watson et al., 2020). Elevated ΔpCO2 associated with elevated NCP in the eastern Equatorial Atlantic was consistent with the seasonal equatorial upwelling (Radenac et al., 2020). 260 Parard et al. (2010) indicated strong negative correlations between SST and ΔpCO2 during the upwelling season (R= -0.76 for June to September), which is also consistent with our results. By contrast, Lefèvre et al. (2016) showed that correlations between pCO2 (sw) and SST were weak across the whole year (R= -0.13), and SSS (R = 0.93) was the primary driver at the same station.
In the western Equatorial Atlantic, negative correlations between NCP and ΔpCO2, and positive correlations between the 265 SSS and ΔpCO2 seasonal component occurred in the vicinity of the Amazon River mouth. The mixing of the Amazon river and oceanic water decreases SSS (Ibánhez et al., 2016;Lefèvre et al., 2020;Bonou et al., 2016;Lefévre et al., 2010), and increases the nutrient supply to the ocean which can in turn enhance NPP and NCP, leading to a decrease in ΔpCO2 within the Amazon plume (Körtzinger, 2003;Cooley et al., 2007). This coupling produces an extensive area of depressed ΔpCO2 which is a CO2 sink (Ibánhez et al., 2016). Lefèvre et al. (2010) indicated that rainfall from the intertropical convergence 270 zone could reduce SSS, with an associated decrease in ΔpCO2. The Eastern Tropical Atlantic is also subject to large river input, especially from the Congo (Hopkins et al., 2013) and Niger rivers, which could produce nutrient-rich plumes that fuel NCP and decrease ΔpCO2 (Lefèvre et al., , 2021. Between 30° S and 45° S, dissolved inorganic carbon and SST exert a similar influence on pCO2 ( Deviations from the expected drivers in the subtropics, occurred within the Benguela upwelling system between 20° S and 35° S. Positive correlations between NCP and the CO2 flux (Fig. 2a) alongside negative correlations between SST, SSS and 280 the CO2 flux (Fig. 2c, d) are indicative of upwelled waters that have both elevated pCO2 (sw) and nutrients, which cause an increase in NPP (Lamont et al., 2014). These upwelled waters move offshore in filaments (Rubio et al., 2009) where NPP decreases, and SST becomes the dominant driver, which is confirmed by the positive correlations between SST and the CO2 flux further offshore. Ford et al. (2021b) indicated a switch in NCP drivers in the Benguela upwelling from wind driven upwelling on the shelf, to filaments that propagate offshore from the upwelling front, which is consistent with the switch in 285 the drivers observed for the CO2 flux as these filaments move offshore.
At between 12° S and 17 °S along the South American coast, there were also deviations from the expected drivers as there were positive correlations between NPP and ΔpCO2 (Fig. 1b) and negative correlations between SSS and ΔpCO2 (Fig. 1d), which are consistent with an upwelling signature that occurs along the coast. Aguiar et al. (2018) also showed intense seasonal upwelling events in this region that are driven by wind and currents. The southern coast of South America is 290 strongly influenced by riverine water input that reduces the total alkalinity and therefore causes an increase in pCO2 (sw) (Liutti et al., 2021). This is associated with an increased supply of nutrients which in turn enhances NPP, though the main drivers of pCO2 (sw) in this region still remain as total alkalinity and SST (Liutti et al., 2021). This potentially explains the positive correlation between ΔpCO2 and both NCP and NPP (Fig. 1a, b), as well as the negative correlations between ΔpCO2 and SSS. The extension offshore of this negative correlation between SSS and ΔpCO2 (Fig. 1d) could be caused by the 295 advection of water masses due to intense mesoscale eddy activity arising from the Brazil-Malvinas confluence (Mason et al., 2017).
The seasonal correlations between the CO2 flux and the drivers were similar to ΔpCO2, but for CO2 flux these occurred over a larger spatial area. The South Atlantic subtropical anticyclone (Reboita et al., 2019) which controls wind speeds across the region, and therefore the gas transfer velocity could enhance the CO2 flux into the subtropical ocean, through higher (or 300 lower) wind speeds in winter (or summer; Xiong et al., 2015). Since seasonal variations in ΔpCO2 largely explain the seasonal variability in the CO2 flux ΔpCO2 can be used as a proxy to understand seasonal variations in the CO2 flux in this region.

Interannual drivers of ΔpCO2 and CO2 flux
The larger geographic region of significant correlations for the air-sea CO2 flux compared to ΔpCO2, and the consistency 305 between the two results (i.e. comparing the smaller regions of ΔpCO2 correlations with their equivalent in the flux results; Fig. 3, 4) suggests that analysing the CO2 flux is the better dataset to investigate drivers of variations in interannual and longer timescales. The results become clearer when analysing the CO2 flux, where the effects of solubility and gas transfer (estimated via wind speed proxy) could reinforce correlations and multi-year trends, which will be retrieved by performing long timeseries analyses on the CO2 flux. Landschützer et al. (2015) showed that variations in the Southern Ocean carbon 310 sink were primarily driven by changes in ΔpCO2, when integrating across basin scales. At localised scales of 1° by 1° as performed in our analysis, changes in surface turbulence and solubility are shown to be important in determining interannual variability, consistent with Keppler and Landschützer (2019). In the North Atlantic Ocean, Henson et al. (2018) showed that the seasonal and interannual drivers of ΔpCO2 are different, which could arise from the necessity to study CO2 fluxes over longer timescales. 315 The interannual component of NCP and the CO2 flux were negatively correlated in the subtropical gyre (Fig. 4a), alongside a positive correlation between SST and CO2 flux (Fig. 4b). El Niño (La Niña) events are known to influence the South Atlantic Ocean, causing an increase (decrease) in SST across the basin (Rodrigues et al., 2015;Colberg et al., 2004), and a decrease (increase) in NPP and NCP (Ford et al., 2021b;Tilstone et al., 2015). Positive correlations between the MEI and CO2 flux ( Fig. 4e) indicate that the MEI partially controls the interannual variability in CO2 flux in the South Atlantic subtropical gyre, 320 through modulations primarily in SST and to a lesser extent NCP. The South Atlantic Subtropical Anticyclone has been observed to strengthen (weaken) and move south (north) during La Niña (El Niño) events. This displacement increases (decreases) wind speeds across the subtropical South Atlantic, which will enhance (weaken) gas exchange, and elevate (depress) NCP (Ford et al., 2021b). These results suggest a more significant role of NCP in controlling the interannual variability in the CO2 flux than has previously been thought. 325 The negative correlation between the CO2 flux and the MEI in a band between 30° S and 45° S (Fig. 4e), indicates that reduced (elevated) wind speeds that occur during La Niña (El Niño) events in this region, suppress (enhance) the gas exchange (Colberg et al., 2004) and therefore acts as a weaker (stronger) CO2 sink. In the equatorial region, neither ΔpCO2 or the CO2 flux were correlated with the MEI, in sharp contrast with  who showed stronger outgassing of CO2 in the western equatorial Atlantic for the year following the 2009 El Niño. In that respect, it should be noted that our 330 analysis would not identify such lagged correlations.
The SAM has known meteorological connections to the MEI (Fogt et al., 2011), where El Niño (La Niña) events generally coincide with negative (positive) SAM phases, resulting in northward (southward) displacement of the westerly winds in the Southern Ocean. Our results showed positive correlations between the CO2 flux and the SAM between 30° S and 45° S (Fig.   4f) indicating stronger (weaker) CO2 drawdown into the oceans during negative (positive) SAM phases. Although no 335 significant correlations were found between ΔpCO2 and the SAM (Fig. 3f), the changes in the gas transfer driven by the displacement of the westerly winds could control the CO2 flux. It should be noted that the effect of the SAM may be more pronounced outside the domain of the present study (i.e south of 45 °S; Keppler and Landschützer, 2019). Landschützer et al. (2015) indicated that the SAM is unlikely to be the main driver of changes in the Southern Ocean CO2 flux, but an observed zonally asymmetric atmospheric pattern could induce changes in the CO2 flux (Keppler and Landschützer, 2019;340 Landschützer et al., 2015). This asymmetric atmospheric pattern, however, may not be captured within the SAM index.

Multi-year trends in ΔpCO2 and CO2 flux
The trends in ΔpCO2 and CO2 flux over 16 years (Fig. 5) showed some similarities to previous trend assessments in the South Atlantic Ocean (Landschützer et al., 2016). Our results indicated a lower number of significant trends however, since uncertainties in the trend analysis were accounted for. The uncertainties in both the pCO2 (sw) estimates from extrapolation 345 techniques and the gas transfer velocity are rarely propagated through previous trend analyses. By accounting for these uncertainties, the trend analyses provide a robust depiction of regions that can confidently be determined as changing. As with the seasonal and interannual analysis, the CO2 flux-based trend analysis showed a greater spatial area of significant trends, when compared to ΔpCO2 (Fig. 5).
The strongest trends in ΔpCO2 and the CO2 flux were observed in the Benguela upwelling system. Arnone et al. (2017)  350 reported positive trends in in situ pCO2 (sw) of 6.1 ± 1.4 µatm yr -1 , between 2005 and 2015. Assuming an atmospheric CO2 increase of 1.5 µatm yr -1 (Takahashi et al., 2002;Zeng et al., 2014), these results are consistent with the ΔpCO2 trends observed in this study (1.5 -3.8 µatm yr -1 , Fig. 5a). Arnone et al. (2017) also suggested that the positive trend was due to a stronger influence of upwelling (Rouault et al., 2010), which injects CO2 and nutrients into the area that is then not completely removed by the enhanced NPP/NCP. Varela et al. (2015) indicated an increase in the strength of the Benguela increases in the Northern Benguela which are consistent with our data that highlights an increasing efflux of CO2 to the atmosphere (Fig. 5b). The CO2 flux trends in this study (0.03 -0.09 mol m -2 yr -1 , Fig. 5b) were also consistent with but slightly lower than the 0.13 ± 0.03 mol m -2 yr -1 trend in CO2 flux observed by Arnone et al. (2017). An increase in the strength of the upwelling that injects CO2 into the surface layer, will be driven by enhanced (upwelling-conducive) winds, 360 that also enhance the gas transfer. This highlights the importance of studying multi-year trends using the CO2 flux, because the enhancement of these trends by meteorological conditions would not be observed using ΔpCO2 alone.
Offshore from the upwelling region negative ΔpCO2 and CO2 flux trends were observed. Rubio et al. (2009) showed that mesoscale filaments and eddies propagate away from the upwelling front, transporting nutrients offshore into the South Atlantic gyre. Ford et al. (2021b)  increase in in situ pCO2 (sw) than pCO2 (atm) (increasing ΔpCO2), but the trend was derived from data from only two research cruises. For the Equatorial upwelling, an increase in ΔpCO2 (as shown here and in Landschützer et al., 2016) is counter intuitive because there is evidence that upwelled water has recently been in contact with the atmosphere (~15 years; Reverdin et al., 1993). Dissolved inorganic carbon in these upwelled waters has been shown to increase at a similar rate to the surface waters (e.g Woosley et al., 2016). Therefore, the trend in ΔpCO2 should be ~0 with increasing pCO2 ( the Amazon Plume region is spatially and temporally variable (Valerio et al., 2021;Ibánhez et al., 2016;Bruto et al., 2017).
The South Atlantic gyre exhibited negative trends in ΔpCO2 and the CO2 flux indicating an increasing drawdown of atmospheric CO2 into the ocean, which were consistent with Landschützer et al. (2016) over the period from 1982 and 2011 though the trends were at the limits of the uncertainties (Fig. B2). Fay and Mckinley (2013) showed weak negative trends in ΔpCO2 using in situ observations over different time series lengths. Using an ensemble of complete pCO2 (sw) fields, Gregor et al. (2019) indicated negative trends in ΔpCO2 however there was low confidence in these trends especially in the South Atlantic gyre. By contrast, Kitidis et al. (2017) reported a mean trend in in situ ΔpCO2 between 1995 and 2013, that was not significantly different from zero. These contradictory trends support the conclusion that ΔpCO2 is unlikely to be 395 representative of the CO2 flux over multi-year timescales. Therefore, we recommend that the CO2 flux should be used to assess multi-year variability in the oceanic CO2 sink, as the importance of changes in solubility and gas transfer velocity (estimated via wind speed) increases (Keppler and Landschützer, 2019).
During the United Nations decade of ocean science (2021)(2022)(2023)(2024)(2025)(2026)(2027)(2028)(2029)(2030) , the Integrated Ocean Carbon Research (IOC-R) highlights that the role of biology is a key issue to understanding the global ocean CO2 sink (Aricò et al., 2021). The biological 400 contribution to both interannual and multi-year variations in the South Atlantic air-sea CO2 flux shown in this study, and supported by Ford et al. (2022), indicates that the biology activity through NCP cannot be assumed to be in steady state. The biological effect of NCP on ΔpCO2 and CO2 flux should therefore not be overlooked when assessing the interannual and multi-year variations in the global ocean carbon sink.

Conclusions 405
In this paper, we have identified the seasonal and interannual drivers of ΔpCO2 and the air-sea CO2 flux in the South Atlantic Ocean using satellite observations. Seasonally, our results indicated that the subtropics were controlled by SST, and the subpolar regions were correlated with biological processes. Deviations from this trend occurred in the Benguela upwelling where predominately biological processes correlated with variability in the ΔpCO2 as well as upwelling. The Equatorial Atlantic showed spatially variable drivers associated with the Amazon Plume and Equatorial upwelling which induced a 410 biological effect. These regions imply a strong biological control on ΔpCO2 through local physical processes. The CO2 flux had similar seasonal drivers to ΔpCO2, but with significant correlations over a larger spatial area. This highlights that ΔpCO2 can be used to indicate the important drivers of the CO2 flux on seasonal timescales, but it is still possible that ΔpCO2 will miss some of the spatial correlations and will likely overestimate the strength of these correlations.
The interannual variability of ΔpCO2 and the CO2 flux was correlated with the MEI through a reduction (increase) of NCP 415 and increase (decrease) in SST during El Niño (La Niña) events, again highlighting the importance of biology to the interannual variability. The CO2 flux response extended over a larger geographical region, indicating that the CO2 flux should be used to assess interannual trends in the oceanic CO2 sink, as opposed to a proxy such as ΔpCO2, which may overestimate the strength of the correlations and does not include variability in the solubility and the gas transfer velocity (estimated via wind speed). The 16 year trends in ΔpCO2 and the CO2 flux were determined with associated uncertainties 420 which identified negative trends in the CO2 flux in the South Atlantic gyre. Positive trends in the CO2 flux were observed in the Benguela upwelling region, which were associated with an increase in the strength and frequency of upwelling. A transition to negative trends offshore were consistent with elevated nutrient export from the upwelling area, and subsequent biological drawdown of CO2. These results highlight, that changes in biological activity in the South Atlantic Ocean can control the interannual and multi-year trends in the oceanic CO2 flux. This emphasises the importance of biology and 425 specifically NCP in assessing the global ocean carbon sink.

Appendices
Appendix A -Driver analysis using in situ ΔpCO2 Henson et al. (2018) performed the X-11 analysis using in situ pCO2 (sw) observations to estimate average ΔpCO2 for the Longhurst provinces (Longhurst, 1998). The in situ pCO2 (sw) observations were obtained from SOCATv2020 430 (https://www.socat.info/; Bakker et al., 2016), and were reanalysed to a temperature dataset representative for a consistent and fixed depth (Reynolds et al., 2002) which is used to represent the base of the mass boundary layer. The reanalysis method used the 'fe_reanalyse_socat.py' routine within FluxEngine (Holding et al., 2019;Shutler et al., 2016), which follows the methodology of Goddijn- Murphy et al. (2015), and as used in Woolf et al. (2019) and Watson et al (2020).
ΔpCO2 was calculated using the reanalysed in situ pCO2 (sw) observations and pCO2 (atm). These ΔpCO2 estimates were used 435 within the driver analysis as described by Henson et al. (2018), using the drivers described in section 2.4, for the South Atlantic Longhurst provinces (Longhurst, 1998). The seasonal drivers of in situ ΔpCO2 (Fig. A1) showed a similar spatial distribution as the SA-FNN ΔpCO2 (Fig. 1). The interannual drivers (Fig. A2) showed some differences to the SA-FNN (Fig.   3). The averaging required to produce the in situ ΔpCO2 timeseries may mask interannual signals, and Ford et al. (2021b) indicated that averaging over large province areas could mask correlations, especially in dynamic regions, and locally these 440 correlations may be significant.    Modelled sea surface salinity from the Copernicus Marine Environment Modelling Service global ocean physics reanalysis product (GLORYS12V1) are available from CMEMS (CMEMS, 2021). ERA5 monthly reanalysis wind speeds are available from the Copernicus Climate Data Store (Hersbach et al., 2019). pCO2 (atm) data are available from v5.5 of the global estimates of pCO2 (sw) dataset (Landschützer et al., 2017(Landschützer et al., , 2016. pCO2 (sw) estimates generated by the SA-FNN are available 475 from Pangaea (Ford et al., 2021a). SOCATv2020 in situ pCO2 (sw) observations (Bakker et al., 2016) are available from https://www.socat.info/index.php/version-2020/.

Author Contributions
DJF, GHT, JDS and VK conceived and directed the research. DJF developed the code and prepared the manuscript. GHT, JDS and VK provided comments that shaped the final manuscript. 480