Keywords

1 Introduction

The portion of the continental shelf of eastern South America between Cape Frío (23°S) and Tierra del Fuego (55°S) is characterized by distinct morphological, climatic, and oceanographic features. This 4300-km-long shelf region encompasses three distinct subregions: the Patagonia continental shelf (PS) (38°S–55°S) to the south; the central shelf (CS), which includes portions of the southern Brazil, Uruguay, and northern Argentina shelves and extends from Cape Santa Marta Grande to Cape Corrientes (28°S–38°S); and the South Brazil Bight (SBB) to the north (23°S–28°S). South of 38°S, the continental shelf is a wide plateau subject to intense westerly winds and high tidal variability. The PS is bounded offshore by the cold, nutrient-rich and relatively fresh waters from the Antarctic Circumpolar Current along the southern boundary, which are advected northward by the Malvinas Current (MC) to about 38°S along the western boundary of the Argentine Basin. Farther north, there is the CS, which is 1350 km long and significantly narrower than the other two shelf regions, reaching about 80 km off Mostardas, near 31°16’S. The CS is strongly influenced by the continental discharge of the Río de la Plata (RP) and the Patos/Mirim Lagoon, and by its proximity to the Brazil Current (BC) and the Brazil/Malvinas Confluence, which bound its offshore limit and promote energetic exchanges between shelf and deep ocean waters. The SBB is a 1000-km-long crescent-shaped bay approximately 200 km wide at the center and 70–80 km at its northern and southern boundaries. This region is bounded offshore by the southward-flowing BC, the western boundary current of the Subtropical Gyre, which exerts a significant influence on the shelf circulation, biology, and biogeochemistry. This chapter describes the water mass, chlorophyll a and CO2 flux distributions, and circulation over the continental shelf of these three shelf regions.

2 Data and Methods

For this review we have analyzed historical hydrographic data from holdings at INIDEP, Argentina (http://www.inidep.edu.ar/oceanografia/PERFILES/ ); CEADO, Argentina (http://www.hidro.gov.ar/ceado/ceado.asp); NOAA’s National Centers for Environmental Prediction World Ocean Data (https://www.nodc.noaa.gov/OC5/WOD/pr_wod.html); and Argo profiling float data (http://www.argo.ucsd.edu/). These data were combined to prepare the updated sea surface salinity distribution presented in Fig. 1a. The hydrographic data are also used to prepare seasonal vertical temperature differences in the upper layer. In addition, we analyzed satellite-derived sea surface temperature (SST) and chlorophyll concentration data. We used NOAA’s Optimum Interpolated SST version 2.0 with daily 0.25 × 0.25° resolution (Reynolds et al. 2007, available at https://www.ncdc.noaa.gov/oisst) to estimate sea surface temperature gradients presented in Fig. 1b. Surface chlorophyll a concentration data are derived from MODIS Aqua collected during 2002–2017. All level 2 data were processed to construct the austral summer climatology of 2 km × 2 km resolution presented in Fig. 1c using the OC3 algorithm (O'Reilly et al. 2000). The data were made available by NASA’s Ocean Biology Processing Group at the Goddard Space Flight Center. Only chlorophyll a data with concentrations within the 0.02–30 mg m−3 range were used in the present analyses. For further details the reader is referred to Marrari et al. (2017). The analysis of surface ocean partial pressure is carried out based on data extracted from the Surface Ocean CO2 Atlas version 4 (SOCAT, https://www.socat.info/). These data include 130,000 observations collected from 1991 to 2015.

Fig. 1
figure 1

Climatological sea surface salinity distribution (a). The red solid (dashed) contour indicates the austral winter (summer) distribution of the 33.5 isohaline, which marks the offshore edge of Plata Plume Water. CF Cape Frio, SBB South Brazil Bight, CSM Cape Santa Marta Grande, LdP Lagoa dos Patos, RP Río de la Plata, CC Cape Corrientes, ER El Rincón. SMG San Matías Gulf, VP Valdes Peninsula, SJG San Jorge Gulf, GB Grande Bay, MS Magellan Strait, LMS Le Maire Strait. The heavy white line indicates the 200 m isobath. (b) Mean summer (orange) and winter (blue) distributions of high sea surface temperature gradients (> 1.5 °C/100 km). The background contours show the bottom topography based on Gebco 30-second-gridded bathymetry. The heavy gray line indicates the 200 m isobath. The insets in a and b display schematic representations of the cross-shelf thermohaline structure in the SBB and the PS. The insets display selected isotherms (light blue) and isohalines (red), surface heat (red wiggly arrow), and freshwater exchanges (orange arrow). The green wiggly arrows indicate vertical mixing. In addition, a schematic surface circulation is shown, adapted from Piola and Matano (2017). (c) Mean summer (21 December–21 March) MODIS surface chlorophyll a in mg m−3 distribution (2002–2017). The heavy white line indicates the 200 m isobath

3 Water Masses and Shelf Fronts

3.1 Patagonian Shelf

In what follows we will use surface salinity, which in this region is not subject to strong seasonal fluctuations, as an identifier of water mass characteristics (e.g., Bianchi et al. 1982; Guerrero and Piola 1997; Piola et al. 2010). Figure 1a presents the climatological near-surface (0–5 meter) salinity distribution based on the analysis of 46,843 observations consisting of bottle, CTD and Argo profiles, hereafter referred to as sea surface salinity (SSS).

The southern portion of the PS is occupied by a modified type of subantarctic waters (SSS < 33.9) that enters the shelf through its southern boundary (e.g., Brandhorst and Castello 1971; Guerrero and Piola 1997). Since the PS is characterized by an excess of evaporation (~2.5 mm day−1, ERA Interim, https://www.ecmwf.int/en/research/climate-reanalysis/era-interim) and only small river runoff, these low-salinity waters reflect the influence of the fresher waters advected from the coastal regions of southern Chile, which are dominated by an excess in precipitation and continental runoff. These low-salinity waters enter the PS via the Magellan Strait (MS, SSS < 32) and the Le Maire Strait (LMS, SSS < 33) (Fig. 1). Slightly saltier waters (SSS < 33.9) enter the PS east of Estados Island. These waters, which are derived from the Cape Horn Current, are diluted along the southern coast of Chile (e.g., Dávila et al. 2002). The low salinity that emanates from the MS and LMS extends northeastward along the coast of Grande Bay, separates from the coast near Cape Tres Puntas (~47°S), and extends to the mid-shelf region at 40°S. This buoyancy plume, which Brandhorst and Castello (1971) called the Patagonian Current , merges with the low-salinity waters derived from the RP near 39°S (Fig. 1a). The Santa Cruz (790 m3 s−1), Negro, and Colorado Rivers (1160 m3 s−1) provide additional freshwater injections that form low-salinity plumes of local significance at 50°S and 40°S (Fig. 1a). Bianchi et al. (1982) defined the coastal waters as those with a SSS < 33.4, shelf waters as 33.4 < SSS < 33.8, and high-salinity shelf water due to local excess evaporation and extended residence time within San Matías Gulf, where SSS > 33.8 (Scasso and Piola 1988; Tonini et al. 2013). Although our SSS analysis is based on a substantially larger number of observations than previously reported (nearly 50% of the data in Fig. 1a were collected after the year 2000), the SSS distribution over the shelf is qualitatively similar to the one described by Bianchi et al. (2005), thus suggesting that the distribution is relatively robust. The most notable differences between our SSS distribution and earlier analyses are subtle: a farther northward extension of the Magellan plume (SSS < 33.6), which merges with the RP plume, and the high-salinity shelf water from San Matías Gulf , which occupies a wider portion of the inner shelf off El Rincón (Fig. 1a). The salty waters derived from the San Matías Gulf extend northeastward during fall–winter and are located east of the gulf’s mouth in spring–summer (Lucas et al. 2005). As will be discussed later, the seasonal wind variability is a dominant forcing of the circulation over the shelf north of about 39°S.

The PS hosts a variety of ocean fronts : upwelling, tidal, shelf break, and cold and temperate estuarine fronts, all of which play a significant ecological role (Acha et al. 2004; Sabatini et al. 2004). The most prominent of these fronts is the shelf break front, which marks the limit between the relatively warm and fresh waters on the shelf from the colder and saltier MC waters. From austral spring to autumn, the shelf break front is readily detected by satellite infrared observations (e.g., Martos and Piccolo 1988; Saraceno et al. 2004; Franco et al. 2008; Rivas and Pisoni 2010). During these seasons, the thermal structure of the shelf break front reveals a transition from the stratified shelf waters to the less stratified waters farther offshore (e.g., Romero et al. 2006). The regional extent of the weakening thermal manifestation of the shelf break front in winter is due to weak thermal stratification over the shelf (<0. 5 °C) compared to the summer stratification (>6 °C, Fig. 2).

Fig. 2
figure 2

Temperature difference (°C) between upper layer (10 m) and lower layer (65 m) from historical data collected during (a) austral winter (July–September) and (b) summer (December–March). Note that only positive temperature differences (T at 10 m > T at 65 m) larger than 4 °C are colored.

Near coastal fronts develop primarily in the summer (Rivas and Pisoni 2010, Fig. 1b). These fronts mark the transition between well-mixed coastal waters and stratified mid-shelf waters (Glorioso 1987, Fig. 2). Tidal-induced mixing is one of the main mechanisms generating these fronts, which are characterized by their sharp SST signature (Palma et al. 2004). The most notable of these fronts are located at ~47–50°S and along the mouth of San Matías Gulf (~42°S) (Palma et al. 2008; Moreira et al. 2011; Tonini et al. 2013). Tidal mixing is also intense along the coast south of 51°S, but the weak thermal stratification in the mid-shelf region (Fig. 2a and inset in Fig. 1a) leads to weaker thermal gradients. This region, which is characterized by the penetration of the Magellan Strait’s buoyant plume and minor contributions from small rivers, is considered a cold estuarine front (Acha et al. 2004). The low-salinity waters, which extend approximately100 km offshore, are moderately warmer/colder than the mid-shelf waters during summer/winter (Rivas and Pisoni 2010). The location of tidal fronts is generally stable although frontal displacements are modulated by the intensity of surface heat flux, the fortnightly tidal cycle, and local mesoscale variability (Pisoni et al. 2015).

3.2 Central Shelf (CS)

The water mass structure of the CS is dominated by the presence of subantarctic shelf waters (SASW) with salinities between 33.6 and 33.8 (Piola et al. 2000). This region is narrower than the PS farther south: 180 km at 38°S, 110 km at 34.5°S, and just under 100 km at 31°S. The oceanic circulation over the CS is driven by seasonally reversing winds (e.g., Castro and Miranda 1998; Palma et al. 2008), which have a substantial imprint on the local water mass distribution on account of their influence on the RP plume (e.g., Piola et al. 2000; Simionato et al. 2001, 2010; Palma et al. 2008; Saraceno et al. 2014; Matano et al. 2014; Strub et al. 2015). Thus, the annual mean SSS distribution in the CS should be interpreted with caution. RP, the second largest river of the southern hemisphere, has an average volume discharge of 23,000 m3 s−1 during 1931–2016, although this value has increased to 25,000 m3 s-1during the past few decades (1997–2016) (Borús et al. 2017). The seasonal variations in RP discharge are small, but its interannual variations are significant due to the impact of El Niño on regional rainfall (Depetris et al. 1996; Robertson and Mechoso 1998). The strong Niños of 1983 and 1998, for example, generated discharges larger than 70,000 m3 s−1. These discharges had a significant impact on the SSS and on the biological and biogeochemical characteristics of the continental shelves of Uruguay and southern Brazil (Ciotti et al. 1995). During the austral fall and winter, the RP plume extends beyond the Cape Santa Marta Grande, near 28°S. In late spring and summer, the plume retracts southward, with its northern limit at approximately the mouth of the Patos Lagoon (32°S) and its southern limit near 38°S (Piola et al. 2000, 2005, 2008b; Möller Jr et al. 2008) (Fig. 1a). Events of strong southwesterly winds and large river outflow lead to exceptionally larger northeastward penetrations of the river plume (Piola et al. 2005). Such events exert a strong biological impact in the coastal areas (Stevenson et al. 1998; Proença et al. 2017). Nagai et al. (2014) show that the RP is the major source of terrigenous sediments for shelf areas between 38 and 27°S and only strong northeasterly winds can reverse this northward flow (Zavialov et al. 2002). Lateral mixing between RP waters and Tropical (TW) and South Atlantic Central waters (SACW) of the BC forms the subtropical shelf waters (Piola et al. 2000, 2008a). During the spring and summer seasons, northeasterly winds push the plume waters toward the south and offshore (Möller Jr et al. 2008; Guerrero et al. 2014; Matano et al. 2014), and most of the northern portion of the CS is occupied by subtropical shelf waters and tropical and central waters derived from the BC (see Souza and Robinson 2004). Under these summer conditions, upwelling events are frequently observed south of Cape Santa Marta Grande (Campos et al. 2013). Similarly, coastal bands of negative sea surface temperature anomalies observed along the southern coast of Uruguay are indicative of upwelling associated with strong easterly winds (e.g., Palma et al. 2008; Pimenta et al. 2008; Simionato et al. 2010). These events are also evident as bands of low coastal sea level in satellite-derived sea surface height anomalies and tide gauge data (Saraceno et al. 2014).

Satellite SST data show two distinct shelf break fronts in the region between 45°S and 25°S (Fig. 1b). South of 38°S the shelf break front marks the transition between the SASW and the cold subantarctic waters within the MC (see Sect. 3.1). In the northern region, the shelf break front marks the transition between the warm BC waters and the relatively colder shelf waters. The SST gradient between these regions is most intense during the austral winter (Fig. 1b; Saraceno et al. 2004), when there is a larger northward penetration of cold subantarctic waters flowing from the southern portion of the CS. The opposing flows reinforce the strong thermohaline gradients in the outer shelf (e.g., Möller Jr et al. 2008; Piola et al. 2008a; Matano et al. 2014). The strong cross-shelf temperature gradients can effectively modulate the atmospheric boundary layer and the surface wind field (Mendonça et al. 2017). In winter on the mid- and outer shelf, the cold subantarctic waters combined with the cold Plata plume waters frequently overlie warm-salty waters influenced by the Brazil Current, creating a sharp thermal inversion over the northern portion of CS (Fig. 2a), which has been referred to as the inverted thermocline (Castello and Möller 1977).

The second most remarkable of the shelf fronts in the southwestern Atlantic region—after the Patagonia shelf break front—is the Subtropical Shelf Front (STSF) , which marks the subsurface limit between the SASW and the STSW (Piola et al. 2000, 2008a). The STSF extends from the inner shelf at around 32°S nearly to 36°S at the shelf break (see Fig. 10 in Piola et al. 2008a). In winter, the density structure is mostly controlled by the salinity distribution preventing isopycnal mixing across the STSF , while in summer the strong thermocline provides common isopycnal layers, allowing cross-front mixing (Piola et al. 2008a).

3.3 South Brazil Bight (SBB)

The southward-flowing BC sweeps the continental slope of the SBB , although only a small fraction of it intrudes on the outer shelf region (Silveira et al. 2000). This western boundary current transports TW (T > 20 °C; S > 36) in the mixed layer and SACW (T < 20 °C; 35 < S < 36) in the pycnocline layer (Emilsson 1961). The low-salinity coastal water (CW; S < 35) mostly occupies the inner shelf (Castro and Miranda 1998). Castro (2014) identified three regions with distinct thermohaline properties in the cross-shelf direction of the SBB (see insets in Fig. 1a, b): outer shelf (OS), mid-shelf (MS), and inner shelf (IS). The IS is located onshore of a bottom temperature front, also referred to as bottom thermal boundary (BTB) , which is defined by the bottom intersection of the 18 °C isotherm. The BTB separates the near-bottom intrusion of SACW from the inshore mixed waters. The IS width varies seasonally between 10–30 km in summer and 40–80 km in winter. The IS waters are usually weakly stratified compared to the MS, except near the medium estuarine systems present in the SBB, and are dominated by relatively low-salinity CW. The MS is located between the BTB and a sharp surface salinity front, also referred to as surface haline boundary (SHB) . The SHB is marked by the surface outcrop of the 36 isohaline which determines the transition between TW and inshore mixed waters. Along most of the SBB, the SHB is found ~ 80 km from shore in summer and 130 km in winter.

The volume of each of the three SBB water masses changes seasonally (Cerda and Castro 2014). During autumn, the volume of SACW in the SBB is approximately equal to the volume of TW, but progressively increases, reaching a maximum in summer, when it is almost twice that of the TW. The CW is the product of mixing of TW with small amounts of continental discharges within the SBB. Its volume and physical characteristics change seasonally. In summer, the volume of CW is larger than in the other seasons, but in summer the CW is somewhat saltier (S > 34.5) than in winter. The main seasonal change in characteristics of the CW is due to northward intrusions in winter of low-salinity waters originating south of the SBB, off the RP and the Patos and Mirim Lagoons (Campos et al. 1995; Stevenson et al. 1998; Lentini et al. 2001; Piola et al. 2000; Acha et al. 2004; Möller Jr et al. 2008; Campos et al. 2013). Interannual variability of the RP intrusions is associated with large outflow events during El Niño (Campos et al. 1999) and, most importantly, with anomalous southwesterlies over the continental shelf (Piola et al. 2005).

Onshore intrusions of nutrient-rich SACW play a significant role in fertilization of the SBB (Castro et al. 2006). There are two main intrusion paths: local, when the oceanic water reaches the MS directly from the shelf break, and remote, when SACW intrudes on the shelf at upstream regions, usually near Cape Frio, and then flows southward along the coast (Cerda and Castro 2014). Intermittent coastal upwelling events of SACW have been observed around Cape Frio, especially in spring and summer, when the region is under the influence of strong and persistent easterly and northeasterly winds (Miranda 1985; Franchito et al. 2008).

Other processes also contribute to the development of upwelling events. Rodrigues and Lorenzetti (2001) showed that the abrupt change in the coastline orientation near Cape Frio favors the effect of wind-induced upwelling. Cerda and Castro (2014) showed that the divergence of the 100 m and 150 m isobaths near Cape Frio favors SACW intrusions toward the coast. Several studies suggested that cyclonic meanders from the BC induce upward motions that inject SACW into the OS (Campos et al. 1995; Cirano and Campos 1996; Campos et al. 2000; Castelão et al. 2004). In addition, the wind stress curl contributes to force local SACW intrusions near Cape Frio (Amor 2004; Castelão and Barth 2006; Castelão 2012). Using numerical models, Palma and Matano (2009) and Matano et al. (2010) showed that the along-shelf pressure gradient near the shelf break, generated by interactions between the BC and the continental slope topography, forces geostrophic currents toward the coast in the bottom boundary layer, elevating the isopycnals toward the sea surface and favoring the onshore penetration of SACW intrusions.

4 Circulation

4.1 Patagonian Shelf (PS)

Numerical simulations suggest that the mean circulation over the continental shelf south of 40°S consists of a broad northeastward flow that intensifies toward the outer shelf (Palma et al. 2008; Combes and Matano 2014). Short-term current records at 43°S indicate mean velocities of 5.9 cm s−1 at 17 m and 3.1 cm s−1 at 67 m (Rivas 1997) that increase toward the shelf break, reaching values of 40 cm s−1 at 10 m depth and 26 cm s−1 at 160 m depth (Valla and Piola 2015). This velocity increase is associated with the proximity of the northward-flowing MC (Palma et al. 2008). In the inner shelf, the circulation is modulated by the coastal geometry, which is characterized by the presence of several bays and gulfs. Two of the largest bays are Grande Bay and the San Jorge Gulf. Numerical models indicate that the circulation in these regions is dominated by relatively weak anticyclonic gyres and poleward coastal currents. South of 49°S, the down shelf development of the Magellan low-salinity plume is highly influenced by tides and wind forcing (Palma and Matano 2012). The circulation features produced by models are in good agreement with geostrophic flows derived from altimeter data at seasonal scales (Ruiz-Etcheverry et al. 2016). The southern shelf exports water northward and eastward in the surface layer that is largely compensated by inflow from the south at all depths and to a lesser extent by subsurface entrainment from the MC (e.g., Combes and Matano 2014). The onshore flow of MC waters is also apparent from the frequent observation of cold-salty subsurface intrusions in the outer shelf (insets in Fig. 1a, b).

Model simulations indicate a moderate strengthening of the northeastward flow during fall and a weakening during spring. North of 48°S, these seasonal variations are mainly driven by the wind forcing over the inner and mid-shelf, while further south, and offshore of the 100 m isobath, the flow is strongly modulated by MC transport variations (Palma et al. 2008). Direct current observations are too short to display a seasonal cycle, but the altimeter-derived currents also present a seasonal pattern. However, in contrast with model results, the intensification and weakening of the altimeter circulation shift to late summer and late winter, respectively (Ruiz-Etcheverry et al. 2016).

The remarkable phytoplankton blooms of the Patagonia shelf break are symptomatic of the upwelling of deep, nutrient-rich waters to the photic zone. It has been postulated and verified using simplified models that the upwelling might be associated with frictionally driven intrusions of the MC onto the shelf (Matano and Palma 2008; Miller et al. 2011). These results have been corroborated by realistic numerical simulations (Palma et al. 2008; Combes and Matano 2014) and observational studies (Valla and Piola 2015). It has also been proposed that the observed variability of the blooms can be modulated by continental shelf waves at seasonal time scales (Saraceno et al. 2005) and by wind variability at intra-seasonal periods (< 15 days, Carranza et al. 2017).

4.2 Central Shelf (CS)

The northward mean flow derived from models extends to about 38°S. This flow strengthens during fall and weakens during spring (Palma et al. 2008; Combes and Matano 2014). The relatively weak seasonal variations predicted by models are in fairly good agreement with those from altimeter observations (Strub et al. 2015; Ruiz-Etcheverry et al. 2016). In contrast, farther north, between 38°S and 28°S, models show a general southward flow on the outer shelf, which is largely controlled by the BC. The inner and mid-shelf regions exhibit significant seasonal changes in circulation and in temperature and salinity fields (Palma et al. 2008; Combes and Matano 2014). The circulation in this region is primarily driven by the alongshore component of the wind stress, which generates a positive sea level coastal anomaly; a geostrophically balanced northward-flowing coastal jet during fall and winter; and a depression of sea level with associated southward coastal flow during spring and summer. These seasonal coastal sea level anomalies are also confirmed by satellite altimeter and tide gauge data (Saraceno et al. 2014). The remarkable seasonal changes in surface salinity are associated with redistribution of the large freshwater discharges from the RP derived from in situ and satellite observations (Piola et al. 2000, 2008b; Möller Jr et al. 2008; Guerrero et al. 2014). During the winter months, northward, downwelling favorable winds generate northward currents and onshore Ekman transport (Meccia et al. 2013; Saraceno et al. 2014; Strub et al. 2015) that traps the RP plume near the coast (Palma et al. 2008). Though the northward plume penetrations in winter exhibit significant interannual variations, on average, the plume is positioned near Cape Santa Marta Grande (Fig. 1a). Relaxation of the winter winds allows some of the freshwater to leave the shelf in filaments that extend into the BC (Piola et al. 2008a; Matano et al. 2014), where the low-salinity signal is rapidly diluted by mixing with the salty upper layer carried by the BC. During summer, poleward winds create upwelling and poleward currents, causing freshwater to move south and offshore into the Brazil/Malvinas Confluence (Guerrero et al. 2014; Matano et al. 2014). The RP plume strengthens the inner and mid-shelf currents during fall and early winter. In summer, models show the development of a subsurface northward countercurrent in the mid-shelf, which flows against the prevailing winds (Palma et al. 2008). This countercurrent is presumably responsible for the northward extension of SASW to about 33°S in summer required to sustain the sharp thermohaline gradients across the STSF (Piola et al. 2008a). The summer geostrophic circulation derived from altimetry also shows a southward flow along the inner shelf and a weak northward flow along the outer shelf extending to about 34°S (Saraceno et al. 2014; Strub et al. 2015). Palma et al. (2008) attribute this northward flow to cross-shelf pressure gradients established farther south by the MC that spread northward well beyond the northernmost reach of the MC, as an arrested topographic wave (Csanady 1978).

Off the RP estuary, the CS is characterized by a persistent off-shelf mass flux. Model-derived volume balances of the shelf region between 34°S and 38°S show an annual mean off-shelf transport of ~1.21 Sv (Matano et al. 2014; see also Mendonça et al. 2017). Most of this transport is drawn from SASW from the Patagonian Shelf (1.15 Sv), with very small contributions from the RP discharge and the northern shelf region. The seasonal variations of the off-shelf transport are relatively small and out of phase with the variations of the southern mass flux; it decreases during the winter and increases during the summer (Matano et al. 2014). These results are in agreement with flow patterns inferred using satellite salinity (Guerrero et al. 2014) and the geostrophic velocities derived from altimetry (Strub et al. 2015).

4.3 South Brazil Bight (SBB)

The mean circulation in the middle and outer shelf north of Cape Santa Marta Grande is mostly southwestward and presents significant cross-shelf and seasonal variations in magnitude (Castro and Miranda 1998; Palma and Matano 2009). Currents in the mid- and outer shelf are stronger in summer and weaker in winter (Matano et al. 2010). These seasonal variations are particularly strong in the inner shelf, where the circulation is driven by baroclinic effects and local winds. Numerical models suggest that the circulation is primarily driven by the poleward flow of the BC in the middle and outer shelf (Palma and Matano 2009), while wind forcing is also important in the inner and middle shelf (Stech and Lorenzzetti 1992). Observations also indicate that the wind stress is a major forcing in the middle shelf region (Dottori and Castro 2009; Dottori and Castro 2018). The influence of remote forcing is manifested in the SST gradients near the shelf break (Campos et al. 2000; Castelão et al. 2004) but is not confined to the surface or to the shelf break, as SACW intrusions and continental shelf waves are additional sources of variability (Castro and Lee 1995; Campos et al. 2010; Filippo et al. 2012; Dottori and Castro 2018). The strongest influence of the BC is observed at depth, where the bottom layer of the outer and middle shelf is permeated by the inflow of SACW throughout the year (Castro and Miranda 1998; Matano et al. 2010). In summer the SACW intrusions extend further onshore (Castro 2014).

The characteristics of the upwelling regime in the SBB are also confirmed by the distribution and variability of passive tracers included in model simulations (Combes and Matano 2014). A large amount of a tracer initially located in the offshore portion of the BC between 25 and 29°S stays offshore and moves south until it meets the MC and leaks offshore in the Brazil/Malvinas Confluence. The remaining part reappears next to the coast as far south as 32°S, transported by the combined action of bottom-layer shelf break upwelling and northeasterly winds (Palma and Matano 2009; Campos et al. 2013). The surface concentration increases at the end of winter when the surface wind stress shifts from downwelling to upwelling favorable and reaches its maximum during spring and summer. A tracer released over the shelf north of ~27°S stays over the shelf but extends just slightly past the 200 m isobath moving south to ~32°S, where it mixes with water from the RP and moves offshore into the Brazil/Malvinas Confluence (Combes and Matano 2014).

5 Surface Chlorophyll a Distribution

Satellite observations of chlorophyll a distributions reveal the high primary productivity levels of the southwestern South Atlantic shelf (Fig. 1c, see Longhurst 1998; Saraceno et al. 2005; Rivas et al. 2006; Romero et al. 2006; Signorini et al. 2006; Marrari et al. 2013). The largest concentrations are observed close to sediment laden continental discharges such as the RP and along the PS shelf break. Though the RP is a source of nutrients, which may lead to significant blooms in the outer estuary (Calliari et al. 2005; Carreto et al. 2008), global chlorophyll a algorithms, such as the one used in this study, frequently overestimate the chlorophyll a concentration in these optically complex waters (e.g., Garcia et al. 2005). Thus, the satellite-derived chlorophyll a distribution in these regions should be interpreted with caution. South of 38°S, the austral summer chlorophyll a distribution suggests a close association between productive regions and semipermanent frontal systems . In austral summer, the shelf break front emerges as an extended band of high concentrations (generally higher than 3 mg m−3, Fig. 1c). In situ chlorophyll a estimates in the shelf break front indicate concentrations exceeding 15 mg m−3 and are in overall agreement with the high concentrations indicated by satellite observations (e.g., Garcia et al. 2008; Lutz et al. 2010; Carreto et al. 2016). On average, the shelf break bloom initiates in September and remains high (~ 3.5 mg m−3) until March, though there is substantial interannual variability (Saraceno et al. 2005; Romero et al. 2006; Signorini et al. 2006; Piola et al. 2010). In contrast, the spring bloom in mid-shelf regions decays monotonically after it peaks in October–November (e.g., Carreto et al. 1995; Romero et al. 2006). Enhanced vertical circulation and mixing intense enough to overcome the strong summer stratification, such as might be expected close to ocean fronts , appear to be essential to maintain productivity after the spring bloom. Consequently, in summer regions of enhanced surface chlorophyll concentration (Fig. 1c) are closely associated with weakened vertical stratification (Fig. 2b).

The high chlorophyll a concentrations associated with the Valdes tidal front reach a spring-summer maximum of ~3.4 mg m−3. This frontal bloom peaks in January and decays monotonically until May (Romero et al. 2006). Chlorophyll blooms are also observed along an extensive band offshore from the 100 m isobath at 85–150 km from shore between 46 and 52°S. This high chlorophyll a band is closely associated with the cold estuarine front (see Fig. 1b, c). The highest satellite chlorophyll a concentrations in the latter region can exceed 30 mg m−3, and similarly high in situ estimates have been reported off Grande Bay near 51°S (Lutz et al. 2010; Gómez et al. 2011). This extensive mid-shelf region in southern Patagonia blooms in January, and the chlorophyll a concentration decays at about the same rate as the Valdes and shelf break fronts (Romero et al. 2006).

In the SBB, onshore intrusions of SACW combined with wind- and BC eddy-induced upwelling are the primary sources of nutrients leading to nearshore blooms (see Sect. 1.3). In addition to wind-induced upwelling (e.g., Campos et al. 2013), the northward intrusions of RP waters in winter appear to play a significant role in fertilizing the CS (e.g., Ciotti et al. 1995). In contrast to the PS, where shelf break upwelling and tidal and wind mixing are permanent sources of nutrients to the upper layer at specific locations, the nutrient sources in the CS and the SBB are modulated by synoptic atmospheric and oceanic processes. The impact of these episodic upwelling events is not readily apparent in long-term surface chlorophyll a climatologies, such as the one presented in Fig. 1c. Consequently, in the two northern regions, the satellite-derived chlorophyll a concentration is significantly lower than in the PS. Given the seasonality of alongshore component of wind stress in the CS, the wind-induced upwelling is more frequent in austral summer, when the region is under the influence of upwelling-favorable northeasterly winds (see Campos et al. 2013; Ito et al. 2016). The contrasting chlorophyll a distributions between the different shelf regions are displayed by the long-term annual mean concentrations in the SBB, CS, and PS, 1.37 ± 0.48, 3.58 ± 0.81, and 2.15 ± 0.39 mg m−3, respectively. The high areal mean chlorophyll a estimated over the CS reflects the influence of high concentration associated with the RP plume, even after limiting the calculation to concentrations lower than 5 mg m−3. Though in situ data support a high chlorophyll a associated with the RP plume (Garcia et al. 2006; Martinez et al. 2005), the mean value in the CS should be interpreted with caution.

6 Surface Ocean CO2 Fluxes

Analysis of historical partial pressure of CO2 (pCO2) is based on data extracted from the SOCAT.v4 data set. Based on the partial pressure and wind data, and using the Wanninkhof (2014) flux parameterization, the sea-air CO2 fluxes were estimated in the PS (see Kahl et al. 2017). Preliminary estimates in the PS and the MC regions indicate an oceanic annual averaged CO2 uptake of −8.1 × 10−3 mol m−2 d−1 (the minus sign indicates flux from the atmosphere to the ocean). The flux is maximum during the austral spring, with an average of −18.6 × 10−3 mol m−2 d−1. These fluxes are among the highest CO2 uptake per unit area in the world ocean, for instance, doubling the CO2 sink in the North Sea (Thomas et al. 2004). In particular, the PS shelf break front presents intense CO2 fluxes into the ocean, with an annual mean of −15 × 10−3 mol m−2 d−1. The PS shelf break also presents a large seasonal flux variability, reaching −22.3 × 10−3 mol m−2 d−1 during spring and − 4.8 × 10−3 mol m−2 d−1 in winter. These new estimates of surface CO2 fluxes in the PS are higher than previously reported (e.g., Bianchi et al. 2009; Padin et al. 2010; Kahl et al. 2017) due to the substantial increase in observations collected during the past decade. Kahl et al. (2017) suggest that biological processes dominate the CO2 variability in the shelf break region and that the decreased CO2 sink during winter is mainly due to the decline in biological activity combined with winter convection. The studies in the PS therefore concluded that vertical stratification plays a key role in CO2 dynamics .

In contrast with the PS, the subtropical shelf off the Brazilian coast acts as a source of CO2 to the atmosphere (Ito et al. 2005; Padin et al. 2010). Observations conducted in the SBB in November 1997 and January/February and June 1998 indicate net sea-air CO2 fluxes between 0.3 and 9.8 × 10−3 mol m−2 day−1 which decrease to ~0.4 and 3.7 × 10−3 mol m−2 day−1 over the slope and neighboring deep ocean (Ito et al. 2005). Upwelling events observed primarily in summer and biological regenerative processes are thought to play a key role in the CO2 dynamics in this region (Ito et al. 2005). Observations in the mid- and outer CS collected during 2000–2008 suggest that in austral spring, most of the region acts as a CO2 sink, while during the fall, the region is characterized by the sharp transition from atmospheric CO2 source along the coast of southern Brazil to CO2 sink off the RP mouth (Padin et al. 2010). More recent observations collected in austral spring 2010 and early summer 2011 in the southern SBB and northern CS indicate that shelf waters are a source of CO2 to the atmosphere (Ito et al. 2016). In late spring the surface CO2 fluxes in the northern CS vary between 0.5 and 1 × 10−3 mol CO2 m−2 day−1 and decrease offshore. The highest fluxes in this region exceed 3 × 10−3 mol CO2 m−2 day−1 in the upwelling region off Cape Santa Marta Grande (Ito et al. 2016). The late summer observations in the northern SBB present much higher CO2 flux estimates, exceeding 15 × 10−3 mol CO2 m−2 day−1 (Ito et al. 2016).

Results from an eddy-resolving, regional ocean biogeochemical model are in good overall agreement with observations, confirming that south of 30° S the western South Atlantic shelf acts as a sink of atmospheric CO2 (Arruda et al. 2015). The model indicates a weak source of CO2 to the atmosphere in the SBB. Observed near shore to offshore and meridional pCO2 gradients are well represented in the simulation. A sensitivity analysis shows that the biological production and solubility are the main processes regulating the model pCO2, with biological production being particularly important over the continental shelves.