Rapidly Increasing Artificial Iodine Highlights Pathways of Iceland-Scotland Overflow Water and Labrador Sea Water

Introduction

The water circulation in the subpolar North Atlantic (SPNA) plays a key role on the conduit of greenhouse gases and other climate properties from the sea surface down to the ocean interior (e.g., Perez et al., 2018). Climate signals are then exported southward through the lower limb of the Atlantic Meridional Overturning Circulation (AMOC). This limb is mainly composed by the Labrador Sea Water (LSW) and the two dense overflows supplied from the Nordic Seas, namely, the Denmark Strait Overflow Water (DSOW) and the Iceland-Scotland Overflow Water (ISOW). The circulation of DSOW is being investigated in the Labrador Sea and further south off the north American shore using time series observations of artificial iodine (¹²⁹I) that started in the early 1990’s (Smith et al., 2005; Orre et al., 2010; Smith et al., 2016). The focus of this study is on exploring the potential of ¹²⁹I to trace the complex pathways of ISOW and LSW in the SPNA and further south into the subtropical North Atlantic Ocean.

Among the two water masses, ISOW is the less well understood due to its complex transport pathways, temporal variability and strong mixing with contiguous water bodies. The historical ISOW circulation scheme describes a single boundary transport (magenta solid lines in Figure 1). In that depiction, ISOW spills from the Nordic Seas into the SPNA through the sills between Iceland and Scotland (Hansen and Østerhus, 2007; Beaird et al., 2013), and follows a westward journey along the perimeter of the Iceland Basin passing primarily through the Charlie – Gibbs Fracture Zone into the Irminger Sea (Saunders, 1994; Bower and Furey, 2017). Then ISOW turns northward as part of the boundary current in the western flank of Reykjanes Ridge and eventually joins the southward flowing Deep Western Boundary Current in the east Greenland Slope (Dickson and Brown, 1994; Schott et al., 1999). However, there is growing evidence from numerical simulations and field observations (e.g., Zou et al., 2020) to portray a more complex circulation scheme in which ISOW travels, more often than previously thought, off boundary currents (magenta dotted lines in Figure 1). For example, a number of Lagrangian floats deployed at the Reykjanes Ridge and programmed to drift at ISOW levels showed trajectories that detach from the boundary current to travel westward into the Irminger Sea, southward along the flanks of the Mid Atlantic Ridge, or eastward to the afar West European Basin (Bower et al., 2002; Lankhorst and Zenk, 2006; Xu et al., 2010; Zou et al., 2017; Racapé et al., 2019; Zou et al., 2020).

Figure 1 Sampling sites during the four cruises (PE424, OVIDE, ATHENA and BATS), the combined ¹²⁹I marine discharge from the nuclear reprocessing plants of Sellafield and La Hague since the 1960s until the recent past (e.g. OSPAR, 2015; UK Environmental Agency, 2015), and the schematic circulation adapted from Daniault et al. (2016) to highlight boundary (solid lines) and interior (dotted lines) pathways of ISOW, LSW and DSOW. Acronyms in figure: Bight Fracture Zone, (BFZ); Charlie – Gibbs Fracture Zone, (CGFZ); Denmark Strait Overflow Water, (DSOW); Deep Western Boundary Current, (DWBC); East Greenland Current, (EGC); Iceland Basin, (IB); Iceland, (IC); Iceland Scotland Overflow Water, (ISOW); Irminger Sea, (IS); Labrador Sea, (LS); Labrador Sea Water, (LSW); Mid Atlantic Ridge, (MAR); North Atlantic Current, (NAC); North Sea, (NS); Reykjanes Ridge, (RR); Sub-Arctic Front, (SAF); Scotland, (SC); and West European Basin, (WEB).

Along the journey in the SPNA, ISOW mixes with other water masses, especially with central waters carried by the North Atlantic Current and LSW (Yashayaev et al., 2007; Beaird et al., 2013; Devana et al., 2021). LSW forms by winter convection in the Labrador Sea and Irminger Sea and constitutes the major intermediate water mass spreading across the North Atlantic (Clarke and Gascard, 1983; Yashayaev et al., 2007; Piron et al., 2017). The transport pathways of LSW are better known, yet similar to ISOW, drifting buoy observations indicate strong recirculation of this water mass eastward into interior basins (e.g., Bower et al., 2009). The number of floats might still be insufficient, especially for ISOW, but the emerging views are already compelling for the export of climate anomalies and suggest that further investigation is needed to provide a better understanding of the structure of the AMOC (Bower et al., 2019).

Radionuclide transient tracers are valuable tools to complement the information obtained from floats and other physical observations. Hydrographic tracer distributions integrate past circulation as if floats were released in great number. A strong tracer candidate for the investigation of the overflows and LSW is ¹²⁹I, which is released by the nuclear fuel reprocessing industry into the surface seawater in the North Sea region and eventually transported into the SPNA (Yiou et al., 1994; Edmonds et al., 2001; Alfimov et al., 2004; Smith et al., 2005; Smith et al., 2011; Alfimov et al., 2013; Gómez-Guzmán et al., 2013; He et al., 2013a; He et al., 2013b; Smith et al., 2016; Casacuberta et al., 2018; Castrillejo et al., 2018; Vivo-Vilches et al., 2018; Wefing et al., 2018). This tracer has already been successfully employed to identify and follow DSOW between the Irminger Sea and Bermuda based on its increasing concentrations that can be measured with high sensitivity, i.e., million atoms per liter of seawater (Smith et al., 2005; Smith et al., 2016). However, measurements of ¹²⁹I conducted prior to Castrillejo et al. (2018) and this study indicated that tracer levels in ISOW and LSW were not large enough to allow the clear differentiation of these water masses in the North Atlantic (Santschi et al., 1996; Edmonds et al., 2001). On the other hand, numerical modelling (Orre et al., 2010) projects a rise in ¹²⁹I concentrations at depths typically occupied by ISOW in the eastern SPNA in response to increased radionuclide discharge rates in recent decades. A glimpse of such tracer increase in ISOW was captured for the first time in 2014 in the Iceland Basin (Castrillejo et al., 2018). Additionally, the ¹²⁹I time series conducted in the AR7W line show increasing tracer concentrations in LSW between 1993 and present (Smith et al., 2005; Orre et al., 2010; Smith et al., 2016). These results imply that if ¹²⁹I was to continue increasing, the tracer could potentially reveal ISOW and LSW transport pathways in parts of the North Atlantic where their identification can be challenging if using properties such as salinity and temperature alone.

The aim of this work is to further explore the potential of ¹²⁹I to trace ISOW and LSW transport pathways in the near future. To that end, we measured ¹²⁹I from over 200 seawater samples collected across the subpolar and subtropical North Atlantic between 2017 and 2019. The new ¹²⁹I dataset captured deep tracer maxima associated with density, salinity and potential temperatures expected for ISOW. Using ¹²⁹I depth profiles we were able to provide an independent validation of ISOW pathways being redrawn by Lagrangian studies. Specifically, we observed ISOW spreading along the boundary current in the eastern flank of the Reykjanes Ridge, over the ridge through the Bight Fracture Zone and off the boundary current following interior routes into the Iceland Basin and the West European Basin. The comparison of the new ¹²⁹I dataset to earlier distributions shows that tracer concentrations increased notably in recent years, allowing for the first time the straightforward tracking of ISOW by using this tracer. Further, we found that ¹²⁹I increased at an unprecedented rate after 2014 intruding other water masses in the SPNA. For example, ¹²⁹I allows distinguishing between different vintages of LSW in the SPNA and following the southward export of LSW through interior pathways into the subtropical region.

Sources of ¹²⁹I

The presence of ¹²⁹I in the North Atlantic is dominated by authorized, marine liquid discharges occurring since the 1960ies until present from the nuclear fuel reprocessing plants of La Hague into the English Channel and Sellafield into the Irish Sea (Figure 1). The ¹²⁹I discharges have been documented over the past decades in reports of the European OSPAR commission, the Environmental Agency of the UK, or by companies running the nuclear installations (e.g., OSPAR, 2015; UK Environmental Agency, 2015). The two facilities alone have discharged more than 6000 kg of ¹²⁹I to regional waters (He et al., 2013a), which is well above the ~90 kg globally dispersed in the 1950ies/60ies as fallout through atmospheric nuclear weapon tests (Wagner et al., 1996; Raisbeck and Yiou, 1999; Hou, 2004). The combined marine discharge rate of the two reprocessing plants increased gradually from < 1 kg/yr to about 100 kg/yr in the first 20 years, then escalated through the 1990s to peak at > 350 kg/yr in the early 2000s and remained above 200 kg/yr in the most recent part (Figure 1). The long residence time of iodine in the ocean water column [over 300,000 yr (Broecker and Peng, 1982; Wong, 1991)] and the limited interaction with organic particles in the photic layer (Schink et al., 1995) suggest that iodine behaves almost conservatively in the ocean. This conservative behavior is confirmed by the long-distance transport of marine ¹²⁹I discharges (e.g. Santschi et al., 1996; Smith et al., 2016) through the North Sea, the Nordic Seas, the Arctic Ocean and as far as the subtropical North Atlantic (see the routes schematically depicted in Figure 1). Consequently, the ¹²⁹I content in waters circulating between the point of discharge and the Arctic Ocean can be 6 to 12 orders of magnitude above the natural levels (< 0.1 ×10⁷ at/kg; atoms per kilogram of seawater) (Snyder et al., 2010) or weapon tests fallout levels (<< 10 ×10⁷ at/kg) (Edmonds et al., 2001). Further transport and mixing of waters from the Arctic, North Sea and northeast Atlantic leads to the entrainment of ¹²⁹I in different water masses and streams within the Nordic Seas. The overall result is a supply of northern waters with ¹²⁹I in the range of ×10⁸ - ×10¹⁰ at/kg via the Greenland – Scotland passages into the SPNA. These northern waters mix with northward flowing southern waters originating from lower latitudes (e.g., tropical and South Atlantic). The lower latitude waters may present small impacts of reprocessing marine discharges due to shallow water recirculation in the North Atlantic (He et al., 2013b), but they are generally affected by weapon tests alone and thus carry 10 to 10,000 times less ¹²⁹I than the northern waters (Castrillejo et al., 2018).

Methods

Sample Collection

Sea water samples were collected strategically to investigate the pathways of the southward lower AMOC in the subpolar and subtropical North Atlantic Ocean between 2017 and 2019. Colored dots in Figure 1 show the locations visited during the four cruises: i) the Iceland Basin and the West European Basin on board the Dutch R/V Pelagia during the PE424 (PE) cruise in July-August 2017; ii) between Lisbon, Portugal, and Cape Farewell, Greenland, on board the French R/V Thalassa during OVIDE (OV) cruise in June-July 2018; iii) south of the Azores Archipelago in the tropical North Atlantic during ATHENA (ATH) cruise on board the German R/V Meteor in October 2018; and iv) the Bermuda Atlantic Time Series (BATS) station on board the North American R/V Atlantic Explorer in June 2019.

In all cruises surface seawater was collected using a surface pump and the water column was sampled using rosettes equipped with Niskin bottles and CTD oxygen-conductivity-temperature-pressure sensors. The seawater was transferred to 500 mL dark plastic bottles after rinsing them 3 times with seawater. Samples were either processed onboard or at land-based laboratories at ETH-Zurich.

Radiochemistry and AMS Measurement of Iodine Isotopes

Iodine was extracted from about 400 mL of seawater at ETH-Zurich following the methods described in Castrillejo et al. (2018). Each sample was spiked with about 1.5 mg of Woodward stable iodine carrier (¹²⁷I). All iodine in seawater was oxidized to iodate upon addition of 2% Ca(ClO)₂ and then reduced to iodide by adding Na₂S₂O₅ and 1M of NH₃O•HCl solution. Columns filled with DOWEX^® 1X8 ion exchange resin were conditioned with deionized water and diluted 0.5 M KNO₃ solution. The loading of the sample in the column was followed by the elution of all the iodine by adding 2.25 M KNO₃ solution and precipitation as AgI using AgNO₃. The precipitate was mixed with about 4 mg of Ag and pressed into cathodes for Accelerator Mass Spectrometry (AMS). Replicates of a seawater sample (n=11) were conducted to check internal consistency. Blanks (n=48) were prepared using deionized water and treated following the same procedure as for the seawater samples.

The calculation of ¹²⁹I concentrations was done based on the measured ¹²⁹I/¹²⁷I ratio and the well-known amounts of ¹²⁷I carrier spiked to each sample. The ¹²⁹I/¹²⁷I atom ratios were measured using the compact 0.5 MV Tandy AMS system at ETH-Zurich (Vockenhuber et al., 2015). The ¹²⁹I/¹²⁷I ratios were normalized with the ETH-Zurich in-house standard D22 with nominal ¹²⁹I/¹²⁷I of (50.35 ± 0.16) × 10^-12 (Christl et al., 2013) and secondary standards with ratios of 5 × 10^-12 that are linked to D22. Blanks presented (2-8) × 10⁵ atoms/kg of ¹²⁹I, corresponding to 1-5% of the total ¹²⁹I measured in seawater samples.

Results

¹²⁹I Concentrations in the SPNA Between 2017 and 2019

All measured ¹²⁹I concentrations are reported in Table S1 along with temperature, salinity and dissolved oxygen data. In the SPNA we sampled 17 depth profiles which are represented in Figure 2. And, collected additionally 10 surface and 6 near bottom samples which are not shown in Figure 2.

Figure 2 (A) Map of ¹²⁹I depth profiles displayed from (B) low-, (C) mid-, to (D) high- concentrations. The legends include all stations sampled during the PE424 cruise in 2017 (PE) and the OVIDE cruise in 2018 (OV). Acronyms in figure: Iceland Basin, (IB); Iceland, (IC); Irminger Sea, (IS); Reykjanes Ridge, (RR); and West European Basin, (WEB).

The ¹²⁹I concentrations in all seawater samples ranged between (0.05 ± 0.05) ×10⁷ at/kg and (275 ± 4) ×10⁷ at/kg. The lowest ¹²⁹I concentrations represent natural waters without anthropogenic influence or that contain a small amount of nuclear weapon test fallout. On the other end, the highest ¹²⁹I concentrations indicate a strong influence of marine discharges from Sellafield and La Hague. Figure 2 displays a map of ¹²⁹I depth profiles along the OVIDE line (Figure 2A) which are arranged in three panels from low- (Figure 2B), mid- (Figure 2C), to high- (Figure 2D) tracer concentrations. The ¹²⁹I concentrations generally increased from east to west. We found the lowest values at depths greater than 3000 m in the West European Basin (OV7-33, Figure 2B). The Iceland Basin (OV43-104, Figure 2C) represented a region of transition where we began to observe higher ¹²⁹I concentrations in the intermediate range of (10 - 40) ×10⁷ at/kg. The bottom layer of the Irminger Sea and the east Greenland shelf (OV76-93, Figure 2D) presented the highest ¹²⁹I concentrations in the order of ×10⁸ - ×10⁹ at/kg. The ¹²⁹I concentrations from PE stations were similar to those found in the Iceland Basin and the West European Basin during OVIDE (Figure 2C).

All acronyms used hereinafter to denote water masses, currents and geographic locations are listed in Table 1 and explained in the figure captions.

Table 1 Acronyms used to denote water masses, currents and geographic locations.

Relationship Between ¹²⁹I and Water Masses in the SPNA

To discuss the transport pathways of ISOW and LSW on basis of ¹²⁹I concentrations it is first necessary to understand how tracer concentrations relate to the water mass structure. Here we focus on OVIDE data because the location of the section is representative for ¹²⁹I concentrations and water masses that reign in the SPNA (OVIDE+PE), and that are to large extent, present in lower latitudes of the Atlantic Ocean (BATS+ATH). Firstly, we plot the radionuclide concentrations from the OVIDE line versus potential temperature (θ) and salinity (S) in θ - S diagrams (Figure 3A). The water mass endmembers (diamond symbols in Figure 3A) are chosen based on the most recent water mass classification for the OVIDE line (García-Ibáñez et al., 2018). Secondly, we represent the zonal distribution of ¹²⁹I with the overlaid water mass structure (Figure 3B). The water mass structure was inferred from the θ - S diagrams (Figure 3A) and zonal distributions of S, θ and O₂ (Figure S1).

Figure 3 Relationship of ¹²⁹I with water masses in the SPNA. (A) Potential temperature (θ) – salinity diagrams. The color bar indicates the ¹²⁹I concentrations. (B) Zonal section of ¹²⁹I concentrations. The color coding is the same as for (A). The acronyms of the water masses shown in (A) are also represented in (B). Data correspond to the OVIDE cruise in 2018. Acronyms in figure: Denmark Strait Overflow Water, (DSOW); East North Atlantic Central Water, (ENACW); Iceland Scotland Overflow Water, (ISOW); Labrador Sea Water, (LSW); Mediterranean Water, (MW); North East Atlantic Deep Water, (NEADW, lower and upper); Polar Intermediate Water, (PIW); Sub-Arctic Front, (SAF); Subpolar Mode Water, (SPMW).

The θ - S diagrams (Figure 3A) show the purest ISOW east of the Reykjanes Ridge characterized by salinity above 34.95, potential temperature of 2-3°C and density in the range of 27.85 – 27.90 kg/m³. In the SPNA, ISOW entrains and mixes mainly with LSW (McCartney and Talley, 1982; Yashayaev et al., 2007) to form the upper North East Atlantic Deep Water (NEADW_U). The lower branch (NEADW_L) has less ISOW and a larger contribution of Lower Deep Water from Antarctica (van Aken, 2000a). The lower and upper branches of NEADW fall roughly in a similar density range as for ISOW but are comparably fresher and colder. West of Reykjanes Ridge ISOW further mixes with underlying DSOW and overlying LSW. Compared to ISOW, DSOW presents a lower salinity (~ 34.90) and potential temperature (< 2°C), and a density greater than 27.90 kg/m³. The LSW present in the Irminger Sea is fresher (S < 34.90), warmer (θ ~ 3°C) and lighter (27.75 kg/m³) than the two overflows. This LSW probably includes waters formed during strong winter convection events in 2013-2014 that have been further renewed with LSW produced during 2015-2017 in the Labrador Sea and the Irminger Sea (Yashayaev and Loder, 2016; Piron et al., 2017). For simplicity we will differentiate between this LSW (LSW_new) and use the term LSW_old for the LSW that left the convection regions and has travelled further (south)east in the SPNA suffering significant entrainment and mixing during its downstream journey. Other water masses in the SPNA include: the southward flowing Subpolar Mode Waters (SPMWs, other than LSW) that result from cooling and freshening of Eastern North Atlantic Central Water (ENACW) through convection events in the SPNA; Polar Intermediate Water (PIW), the freshest and coldest water originating from the Arctic Ocean; and the Mediterranean Water (MW) characterized by the highest salinity along the OVIDE line.

The vertical distribution of ¹²⁹I along the OVIDE section (Figure 3B) allows the straightforward distinction between northern and southern origin water masses. Waters of northern origin generally presented ¹²⁹I concentrations above 20 ×10⁷ at/kg and occupied the region west of 22.5°W which is geographically delimited by the position of the Sub-Arctic Front. In this group we observe ISOW with ¹²⁹I concentrations in the (25-40) ×10⁷ at/kg range, PIW along the east Greenland shelf carrying the highest ¹²⁹I concentrations (> 100 ×10⁷ at/kg), the core of DSOW filling the bottom of the Irminger Sea with ¹²⁹I concentrations slightly above 100 ×10⁷ at/kg, and LSW and SPMWs with concentrations in the range of (15 – 40) ×10⁷ at/kg. LSW_new filling the Irminger Sea water column carries more ¹²⁹I than the older LSW that is present east of Reykjanes Ridge. Southern waters dominated the region east of the Sub-Arctic Front and carried little ¹²⁹I (< 15 ×10⁷ at/kg) in comparison to northern waters. Because of this reason, the identification of ISOW and LSW was easily accomplished by searching for ¹²⁹I spikes at deep and intermediate depths (e.g., eastern flank of Reykjanes Ridge in Figure 3B), even when these water masses were difficult to distinguish by using S, θ and O₂ alone (Figure S1).

Discussion

Observed ISOW Pathways in 2017-2018

Here we provide an independent means of validating transport pathways of ISOW. Our approach is based on the identification of deep, high ¹²⁹I concentrations (> 15 ×10⁷ at/kg) captured at the layer with 27.85 – 27.90 kg/m³ potential density, S >34.95 and θ of 2-3°C. We focus on cases where there is a large ¹²⁹I concentration gradient between ISOW and surrounding water masses. These conditions are met in the Reykjanes Ridge, Iceland Basin and the West European Basin, but not in the Irminger Sea because there the ¹²⁹I concentrations of ISOW are surpassed by higher tracer amounts carried by the underlying DSOW.

We found seven ¹²⁹I depth profiles from OVIDE and PE424 cruises that captured the mentioned deep tracer spike in 2017/2018 (yellow and blue data in Figs. 4A-4C). Station positions are shown in Figure 4D. The largest ¹²⁹I spike, of (36.8 ± 0.8) ×10⁷ at/kg, was measured at 2440 m depth in station OV63 located in the eastern flank of Reykjanes Ridge (Figure 4A). The result is consistent with a substantial amount of ISOW following the counterclockwise boundary current in the Iceland Basin towards the ridge (e.g., Daniault et al., 2016). The crossing of ISOW from the Iceland Basin into the Irminger Sea occurs primarily via fracture zones (Petit et al., 2018). Traditionally it was thought that the principal origin of ISOW found in the boundary current along the western flank of the Reykjanes Ridge (Daniault et al., 2016) was the Charlie Gibbs Fracture Zone. However, there is growing evidence from modelled and observed float trajectories that Bight Fracture Zone is a very important pathway of ISOW into the western flank of the ridge (Bower et al., 2002; Xu et al., 2010; Zou et al., 2017; Zou et al., 2020). For example, some floats deployed at ISOW densities in the Bight Fracture Zone (Bower et al., 2002; Lankhorst and Zenk, 2006; Zou et al., 2017) have described northward flows connecting with the boundary current while the majority of RAFOS and Deep Argo floats released at the Charlie Gibbs Fracture Zone followed west-northwest interior pathways or took a southward direction along the Mid Atlantic Ridge (Racapé et al., 2019; Zou et al., 2020). To check whether there was ISOW at the Bight Fracture Zone during OVIDE 2018 we collected a near-bottom seawater sample at 2280 m depth in station OV104 (Figure 4A). The unequivocal ¹²⁹I spike [(28.8 ± 0.4) ×10⁷ at/kg] was observed again indicating the presence of ISOW that was warmer and lighter than the one found in the eastern flank of Reykjanes Ridge (Table S1).

Figure 4 ¹²⁹I in the Northeastern Atlantic Ocean. Depth profiles of ¹²⁹I obtained during OVIDE cruise in 2018 (OV, blue) and the PE424 cruise in 2017 (PE, yellow) at (A) Reykjanes Ridge, (B) the Iceland Basin, and (C) the West European Basin, are represented and compared to nearby depth profiles available in the literature (grey profiles): HE sites measured in 1993 by Edmonds et al. (2001) and GV sites from the GEOVIDE cruise in 2014 (Castrillejo et al., 2018) revisited during OVIDE in 2018. (D) Map showing the stations (*) and boundary (solid lines) and interior pathways of ISOW (dotted lines). *Stations GV1 and GV32 were at the same location as OV7 and OV57 respectively. (E) Temporal evolution of mean ¹²⁹I concentrations in ISOW at the Reykjanes Ridge, Iceland Basin and the Irminger Sea. Data used for (E) are summarized in Table S2. Acronyms in figure: Bight Fracture Zone, (BFZ); Charlie – Gibbs Fracture Zone, (CGFZ); Iceland Basin, (IB); Iceland, (IC); Iceland Scotland Overflow Water, (ISOW); Irminger Sea, (IS); Maury Channel, (MC); Mid Atlantic Ridge, (MAR); Reykjanes Ridge, (RR); Rockall Plateau, (RP); Rockall Through, (RT); and West European Basin, (WEB).

In the eastern North Atlantic, ¹²⁹I observations suggest that several veins of ISOW spread following interior routes. The ¹²⁹I peak of about (26 – 30) ×10⁷ at/kg was easily identified in near bottom waters collected at stations PE28 and OV57 (Figure 4B) pointing to ISOW passages near Maury Channel in the eastern Iceland Basin (Figure 4D). This result agrees with modelled and deployed float trajectories (Xu et al., 2010; Zou et al., 2017; Zou et al., 2020) as well as with hydrographic surveys (Daniault et al., 2016), although the limited sampling resolution in this study does not allow confirming if ISOW captured at PE28 and OV57 corresponds to a southward branch detached from the boundary current (as in Zou et al., 2017) or to water that recirculated eastward from the Mid Atlantic Ridge. In comparison to the Iceland Basin, in the West European Basin deep tracer maxima were more diluted due to greater mixing of ISOW with southern components of NEADW (Figure 4C). Yet, one could still use ¹²⁹I to distinguish ISOW contributions to NEADW even when the two water masses display similar salinities and potential temperatures (García-Ibáñez et al., 2018). For example, PE22 captures a small peak of ¹²⁹I ((16.6 ± 0.2) ×10⁷ at/kg) at 2200 m depth in the intersect between Rockall Through and the West European Basin consistent with a small southward flow of ISOW (Sherwin et al., 2008; Chang et al., 2009; Zou et al., 2017). And further south and below 3000 m depth, station PE17 and OV7 presented tracer peaks of (17.9 ± 0.2) ×10⁷ at/kg and (7.8 ± 0.2) ×10⁷ at/kg, respectively. These tracer maxima can only be explained by the lateral advection of dense overflows like ISOW into the West European Basin in the absence of other sources of ¹²⁹I at low latitudes (van Aken, 2000a; Fleischmann et al., 2001; Zou et al., 2017; Xu et al., 2018), but their travel path is still unrevealed.

Temporal Evolution of ¹²⁹I in ISOW

In the 1990s sufficient ¹²⁹I was found only in DSOW while the tracer content was close to the natural background in other waters of the North Atlantic Ocean (Santschi et al., 1996; Edmonds et al., 2001; Smith et al., 2005). Thus, when did ¹²⁹I begin to reveal ISOW pathways? To answer this question, we compare the ¹²⁹I depth profiles from 2017-2018 to previous observations (grey profiles) in the SPNA (Figure 4). Edmonds et al. (2001) reported the first ¹²⁹I depth profiles (stations labelled as ‘HE’ in Figure 4D) in the eastern (HE5) and western (HE6) flanks of the Reykjanes Ridge and southwest of the Rockall Plateau in the Iceland Basin (HE4) for samples collected in 1993. Their ¹²⁹I profiles (Figures 4A, B) showed very low tracer concentrations throughout the water column. On the contrary, the overflow water begun to carry sufficient ¹²⁹I in the Iceland Basin (station GV32) in 2014 (Castrillejo et al., 2018). And it was not until 2018 that a strong tracer signal unraveled the passage of ISOW in the Iceland Basin (OV 57) and other parts of the SPNA.

To provide a more comprehensive view on the temporal evolution of ¹²⁹I in ISOW, we calculated the mean tracer concentrations using all available measurements in the overflow found east of Greenland and west of the West European Basin. We exclude the West European Basin from the calculations because tracer data are insufficient to infer any temporal trend in that region. West of the West European Basin, ¹²⁹I observations are also limited (Table S2), yet Figure 4E convincingly shows a clear increase in tracer values between 1993 and 2018. Whether tracer concentrations increased linearly or stepwise cannot be inferred based on the limited tracer measurements. But it is clear that during the 25-year time period the ¹²⁹I concentration in ISOW was multiplied 10 times on average, from lower than 5 ×10⁷ at/kg in the 1990s to higher than 30 ×10⁷ at/kg in the 2010s. Half of the ¹²⁹I increase occurred between 1993 and 2014 while tracer levels doubled in the short time span between 2014 and 2018. Will ¹²⁹I continue labelling ISOW? The answer is: very likely yes. It is reasonable to think that in comparison to present, future ISOW will carry similar quantities of ¹²⁹I, or more, because of two reasons: firstly, the repeated ¹²⁹I measurements in DSOW show that its tracer content is still rising (see Figure 3 in Smith et al., 2016); secondly, the marine radionuclide discharge from the reprocessing facilities in Europe has either increased or levelled off during the last decades (see Sources of ¹²⁹I). The expected implications are that ¹²⁹I will stand out in ISOW and that the tracer has potential to shed light on time scales of ISOW circulation between the source and the SPNA and further downstream provided there is a continuation of ¹²⁹I observations to extend the so far limited time series displayed in Figure 4E.

Spreading of ¹²⁹I in LSW

To investigate whether ¹²⁹I was increasing rapidly in other water masses of the SPNA we compared the vertical distribution of ¹²⁹I along the OVIDE line between 2014 and 2018 (Figure 5A). The colored dots and contour lines (in spite of the excessive interpolation) show that ¹²⁹I concentrations in the region occupied by northern waters were in all cases higher (i.e., positive values) in 2018 than in 2014, while southern waters showed small differences in the tracer field (< 5 ×10⁷ at/kg). The influence of increased ¹²⁹I releases from the European reprocessing plants (see Sources of ¹²⁹I) is most notorious in the area delimited by difference contour lines of (5-10) ×10⁷ at/kg. Apart from the overflows and shallow waters, the area comprised by these contour lines is mainly filled by LSW. In Figure 3 we showed that it is possible to distinguish young LSW vintages in the Irminger Sea by their higher ¹²⁹I concentrations from the tracer poorer and older LSW that spread (south)eastward passing the Reykjanes Ridge. This is consistent with the formation process of LSW that incorporates ¹²⁹I-rich Arctic-origin fresh water during convective events in the Labrador Sea (van Aken, 2000b). Following that reasoning, we suggest that the spreading of LSW plays an important role in setting intermediate and deep ¹²⁹I values of the SPNA (Figure 5A). For example, ¹²⁹I concentrations increased more than 10 ×10⁷ at/kg at depths occupied by LSW_new in the Irminger Sea. And the eastward tracer increases delimited by the 5 ×10⁷ at/kg contour line possibly indicates the transport of older and tracer-poorer LSW.

Figure 5 (A) Temporal change of ¹²⁹I concentrations along the OVIDE line between 2014 (Castrillejo et al., 2018) and 2018 (this study) estimated as the subtraction of 2014 tracer values from 2018 observations overlapping at similar locations. (B) Map with the section plotted in (A) and the location of ¹²⁹I depth profiles at (C) BATS and (D) ATHENA stations in the subtropical North Atlantic. Plot (C) includes the two ¹²⁹I depth profiles sampled in 1993-1994 near the northeastern American slope at stations 13 (PS13, 36.3981°N, 74.2358°E) and 11 (PS11, 36.5081°N, 74.0908°E) reported by Peter H. Santschi et al. (1996). Acronyms in figure: Deep Western Boundary Current, (DWBC); Iceland Basin, (IB); Irminger Sea, (IS); Labrador Sea, (LS); Labrador Sea Water, (LSW); Mid Atlantic Ridge, (MAR); North Atlantic Deep Water, (NADW); North East Atlantic Deep Water, (NEADW); Reykjanes Ridge, (RR); Sub-Arctic Front, (SAF); and West European Basin, (WEB).

To further explore if ¹²⁹I could unveil off-boundary pathways of LSW beyond the SPNA (turquoise dotted lines in Figure 5B) we visited three sites in the western and eastern subtropical North Atlantic Ocean (Figs. 5C and 5D). In the subtropical region, consistently with the literature, we preferred to call NADW_U and NEADW_U the water-masses derived from LSW far from its formation region, and NADW_L for DSOW. To the best of our knowledge, the tracer observations presented here for BATS (Figure 5C) constitute the first full-depth profile of ¹²⁹I in the low latitude North Atlantic aside from the two full depth profiles measured off the North American Slope in 1993/1994 (Santschi et al., 1996). In their ¹²⁹I dataset, a near bottom tracer peak could be observed associated to the transport of NADW_L via the Deep Western Boundary Current (Figure 5C). More recently another study by Smith et al. (2016) focused on the off-boundary transport of the same lower branch in our study area east of Bermuda (magenta dotted lines in Figure 5B). But they did not report ¹²⁹I values for shallower depths occupied by NADW_U. In our BATS profile (Figure 5C) we observed a very clear peak of ¹²⁹I centered at about 1500 m depth coinciding with potential densities (σ_θ ~27.75 kg/m³, see Table S1) and a local salinity minimum (S~35) that characterize the LSW in the North Atlantic (McCartney and Talley, 1982). This tracer peak at 1500 m depth therefore shows that ¹²⁹I data allow the straightforward identification of NADW_U taking off-boundary pathways near Bermuda (turquoise dotted lines in Figure 5B). We also sampled two partial depth profiles south of the Azores Archipelago during ATHENA (Figure 5D). This far east ¹²⁹I concentrations were lower than at BATS. The depths occupied by the NEADW_U only showed slightly more elevated ¹²⁹I concentrations at station ATH95 and a little further south at ATH96 the ~1500 m depth tracer peak could not be recognized. We learnt from other transient tracer observations, i. e. chlorofluorocarbons (CFCs), that highest fractions of LSW in the subtropical region are found in the DWBC and that the presence of young vintages is very limited east of the Mid Atlantic Ridge (e. g. Rhein et al., 2015). The three ¹²⁹I depth profiles in Figures 5C, D show an eastward decrease in tracer concentrations that generally agrees with the spreading of LSW (or NEADW_U) inferred from CFC tracer fields. Yet, more ¹²⁹I data are certainly needed to alleviate the sampling gap between the western and eastern subtropical North Atlantic and to confirm the relationship between the tracer distribution and the spreading of intermediate waters in the subtropical gyre.

The new ¹²⁹I observations presented here show that deep ventilation in the subpolar region plays an important role on the sequestration of the tracer from shallow to deeper water layers supporting the modelling work of Orre et al. (2010) and Snyder, Aldahan and Possnert et al. (2010). While ¹²⁹I has been used to infer mixing and transport timescales of DSOW in the SPNA (Smith et al., 2005) and down to Bermuda (Santschi et al., 1996; Smith et al., 2016), no published work has used ¹²⁹I for investigating LSW or ISOW advection yet. This work clearly demonstrates that ¹²⁹I is now being supplied at unprecedented rate (Figure 5A) through the Greenland-Scotland passages making more northern water masses traceable in the North Atlantic and that measurable tracer peaks extend at least as far as 30°N (Figures 4, 5C). The tracer intrusion in the North Atlantic shows that LSW and the overflows commonly take off-boundary pathways in agreement with drifting float (Bower et al., 2009; Bower et al., 2019) and CFC observations (Rhein et al., 2015). The clarity of ¹²⁹I peaks captured at the stations sampled afar from the boundary regions (Figures 4, 5C) indicates that recirculation through interior pathways must be important for the southward lower limb of the AMOC. Deflections of LSW and ISOW off the boundary currents appear to be related to latitudinal and meridional displacements of the North Atlantic Current (Bower et al., 2009; Bower and Furey, 2017; Xu et al., 2018) and its deep-reaching eddies and meanders (e.g., Zou et al., 2020). The partition between northern and southern waters also depends on the positioning of the North Atlantic Current, with southern (northern) waters with low (high) ¹²⁹I concentrations generally occupying the regions east (west) of the main North Atlantic Current front. In this regard, a higher density sampling of ¹²⁹I across North Atlantic Current boundaries in the subtropical and subpolar North Atlantic may be of interest to further investigate the driving forces of LSW and ISOW pathways. We think that the findings presented here are encouraging for the radionuclide tracer community. And thus, we call for a larger effort to build continued ¹²⁹I time series which can help us better understand the structure and spreading rates of the AMOC.

Data Availability Statement

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author. The iodine-129 data from the OVIDE 2018 cruise are available at SEANOE (Lherminier et al., 2022).

Author Contributions

MC collected the samples during OVIDE 2018 cruise and processed all seawater samples, he led the interpretation of data and the drafting of the manuscript. MC also raised part of the funds. NC managed the seawater sampling of PE424, BATS and ATHENA cruises, raised part of the funds for research and conducted the AMS measurements of iodine isotopes together with CV. PL led the OVIDE 2018 cruise facilitating the collection of samples and assessed the physical oceanographic aspects discussed in this work. All authors contributed to the article and approved the submitted version.

Funding

The research of MC was partly funded by the ETH Career Seed Grant (SEED-06 19-2) and the Swiss National Science Foundation Postdoc Mobility Program (P400P2_199289). NC’s funding came from the European Research Council (ERC Consolidator GAP-101001451) and the Swiss National Science Foundation (AMBIZIONE PZ00P2_154805 and PRIMA PR00P2_193091). The participation in the OVIDE 2018 Cruise onboard R/V Thalassa (https://doi.org/10.17600/18000510) was supported by PL and funds from the Laboratory of Ion Beam Physics. Support to process samples from PE424 cruise onboard R/V Pelagia and to participate on BATS monitoring onboard R/V Atlantic Explorer came from Swiss National Science Foundation Ambizione grant of NC. Funding for ATHENA Cruise (M151) onboard R/V Meteor came from Deutsche Forschungsgemeinschaft. The analysis of long-lived isotopes of iodine was possible thanks to internal funds of the Laboratory of Ion Beam Physics at ETH-Zurich and its consortium partners EAWAG, EMPA, and PSI. Open access funding provided by ETH Zürich.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher’s Note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Acknowledgments

Captains, crew and scientific staff of the four cruises are thanked for their dedication in collecting the samples used in this study. Thanks go to A. M. Wefing for her assistance in sample collection and comments on the manuscript and to A. Schlatter and K. Kündig for help in the laboratory. Norbert Frank is thanked for inviting us to participate in cruise ATHENA (M151). Acknowledgements go to Loes Gerringa and Rob Middag (NIOZ) who collected samples during the PE424 cruise. Rod Johnson (BIOS) and Anita Schlatter (EAWAG) collected the ¹²⁹I profile during one of their monitoring cruises at BATS station. The analysis of waters masses would not have been possible without the hydrographic data that was kindly produced by different groups from many nationalities, thus, a big thank you to all. The constructive comments from the two reviewers are greatly appreciated.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmars.2022.897729/full#supplementary-material

References