Distribution of 236U in the U.S. GEOTRACES Eastern Pacific Zonal Transect and its use as a 2 water mass tracer

Depósito de investigación de la Universidad de Sevilla https://idus.us.es/ Esta es la versión aceptada del artículo publicado en: This is an accepted manuscript of a paper published in: Chemical Geology (2019): 02/12/2024 DOI: https://doi.org/10.1016/j.chemgeo.2019.04.003 Copyright: © 2019 Elsevier Ltd. All rights reserved. El acceso a la versión publicada del artículo puede requerir la suscripción de la revista. Access to the published version may require subscription. “This is an Accepted Manuscript of an article published by Elsevier in [Chemical Geology] on [2019], available at: https://doi.org/10.1016/j.chemgeo.2019.04.003 ” 1 Distribution of 236U in the U.S. GEOTRACES Eastern Pacific Zonal Transect and its use as a 1 water mass tracer 2 M. Villa-Alfageme1*, E. Chamizo2, T. C. Kenna3, M. López-Lora4, N. Casacuberta5, C. Chang3, P. 3 Masqué6,7,8, M. Christl5 4 5 1 Dpto. Física Aplicada II., ETSIE. Av. Reina Mercedes 4A, Universidad de Sevilla. 41012 Sevilla, Spain 6 2Centro Nacional de Aceleradores (CNA), Universidad de Sevilla. Sevilla, Spain 7 3Lamont-Doherty Earth Observatory. Columbia University. NY. USA 8 4Dpto. Física Aplicada I, Escuela Politécnica Superior, Universidad de Sevilla, Virgen de África 7, 41011 9 Sevilla, Spain 10 5Ion Beam Physics. ETH Zürich. Switzerland 11 6School of Natural Sciences. Centre for Marine Ecosystems Research. Edith Cowan University. Joondalup. WA. 12 Australia 13 7Departament de Física & Institut de Ciència i Tecnologia Ambientals. Universitat Autònoma de Barcelona. 14 08193 Bellaterra. Spain. 15 8Oceans Institute & School of Physics. The University of Western Australia. 35 Stirling Highway. Crawley, WA 16 6009. Australia 17 18 Highlights 19 • 236U inventory in EPZT is one order of magnitude lower than in North Pacific 20 • 236U anthropogenic contribution at the EPZT is due to global fallout contribution 21 • AMS was used to measure dissolved 236U concentrations, which are among the lowest 236U values 22 reported for seawater 23 • Lowest 236U concentrations were 1 × 103 at kg-1, measured between 2000 and 3000 m 24 • 236U distributions were used to identify deep and bottom water masses in EPZT 25 26 Abstract 27 We report dissolved concentrations of the long-lived radioisotope 236U measured in the water column 28 along the 2013 US GEOTRACES Eastern Pacific Zonal Transect (GP16). This transect followed a 10 - 29 15°S line from Manta, Ecuador, to Papeete, Haiti, French Polynesia, crossing the southern East Pacific 30 Rise, intercepting one of the largest hydrothermal plumes as well as a productivity gradient, which 31 2 includes the upwelling zone and associated low oxygen waters offshore from Peru and the oligotrophic 32 sub-tropical gyre. Accelerator Mass Spectrometry was used to measure dissolved seawater 236U 33 concentrations as low as 1 × 103 atom kg-1, which are among the lowest levels reported to date. 34 Differences in upper water column 236U distributions from east to west are a result of variable 35 contributions from different surface and intermediate waters encountered along the transect. 36 The distribution of 236U, both in depth and geographically, provides complementary information to that 37 obtained from 14C and helium isotopes, demonstrating that 236U concentrations are diagnostic in the 38 identification of and contributions from the different deep and bottom water masses encountered along 39 the EPZT (Jenkins et al., 2017). For example, we observe minimum 236U concentrations along the EPZT 40 between 2000 and 3000 m that are consistent with contributions attributed to Pacific Deep Water. We 41 also observe increases in 236U below 3000 m at the eastern and western termini of the EPZT. This is 42 consistent with contributions associated with Antarctic Bottom Water and Lower Circumpolar Deep 43 Water. Our results indicate that 236U may be used in conjunction with 14C and 3He isotopes as an 44 additional tool with which to identify and resolve contributions from different water masses in the Pacific 45 Ocean. 46 47 1. Introduction 48 Over the past ten years, the long-lived radionuclide 236U (T1/2=23.4 × 106 y) has been developed as an 49 ocean circulation tracer. Anthropogenic 236U is produced in nuclear detonations mainly from the reaction 50 238U(n,3n)236U for fast neutrons and in nuclear reactors mostly from thermal neutron activation on 235U 51 (Steier et al., 2008). In contrast to the naturally occurring uranium isotopes it has not reached steady state 52 in the oceans, providing a dynamic signal able to label different water masses. Thanks to the low 53 detection limits and high abundance sensitivity of Accelerator Mass Spectrometry (AMS), 236U can be 54 accurately determined in small-volume (<10 L) seawater samples, making collection and analysis 55 suitable for oceanographic sampling campaigns (Chamizo et al., 2015a; Christl et al., 2015a; Eigl et al., 56 2013; López-Lora et al., 2018). 57 Similar to other tracers such as 14C (Matsumoto, 2007), He3 (Jenkins et al., 2017) and CFCs (Schlitzer, 58 2007), 236U can be characterized as a transient tracer, for which major sources and release histories are 59 3 constrained. However, there are several features that set 236U apart from the others, making it useful as 60 an additional water mass/circulation tracer. 236U exhibits a quasi-conservative behavior in the open-sea, 61 due to its occurrence as a stable carbonate ion complex (Ku et al., 1977). Secondly, while atmospheric 62 delivery and fallout from nuclear weapons testing are common to some of the other tracers, an additional 63 source of 236U to the marine environment are discharges from Nuclear Fuel Reprocessing Plants (NFRPs) 64 located in the North Sea (North Western Atlantic) (Casacuberta et al., 2014). These releases are 65 temporally and geographically constrained, offering the possibility of providing unique information on 66 transport time scales. Similar to work in the Atlantic and Arctic Oceans using 129I and 137Cs (Smith et al., 67 2011), 236U could be used in combination with other transient tracers (e.g. 14C, He, CFCs or 129I) to 68 investigate various transport timing distributions. 69 A total amount of 35 kg of naturally occurring 236U is contained within the upper 1000 m of the Earth’s 70 crust (Christl et al., 2012b; Steier et al., 2008). This is due to neutron activation of natural 235U induced 71 by either cosmic rays and/or by neutrons from (α, n) reactions (Steier et al., 2008). By comparison, an 72 estimation of 1000 kg of anthropogenic 236U has been released to the atmosphere , on a global scale, 73 during the period of nuclear weapons development and testing (Sakaguchi et al., 2009). The subsequent 74 nuclear fallout has increased the 236U/238U atom ratio in seawater by several orders of magnitude above 75 the estimated ratio from naturally occurring sources, which might be below the 10-12 level. The global 76 fallout signal has been also locally altered by regional sources, such as liquid discharges from the NFRP 77 to the North Sea (Casacuberta et al., 2016), or from the Marcoule NFRP (France) to the Mediterranean 78 Sea (Castrillejo et al., 2017; Chamizo et al., 2016). 79 To date, the 236U dataset in the ocean is limited, with most measurements obtained during the last 10 80 years when the AMS techniques improved and allowed routine 236U measurements (Bu et al., 2017; 81 Christl et al., 2012b; Steier et al., 2008). Experimental data have been reported from the Japan Sea 82 (Sakaguchi et al., 2012), Mediterranean Sea (Castrillejo et al., 2017; Chamizo et al., 2016), North Sea 83 (Christl et al., 2015a), North Atlantic Ocean (Casacuberta et al., 2014; Christl et al., 2012a), Arctic Ocean 84 (Casacuberta et al., 2016) and North Pacific Ocean (Eigl et al., 2017). In the Arctic Ocean, 236U has been 85 used in combination with 129I to trace North Atlantic waters, impacted with the releases of Sellafield and 86 La Hague NFRP, into the Arctic Ocean. Additionally, 236U temporal evolution have been reconstructed 87 in the Marshall Islands and the Northwest Pacific Ocean using coral records (Froehlich et al., 2017; 88 Nomura et al., 2017), and very recently, 236U has been determined for the first time in abyssal sediments 89 4 from the North Atlantic Ocean (Villa-Alfageme et al., 2018). Most of the studies have been conducted 90 in the Northern Hemisphere, and no data are yet available for the South Pacific Ocean. 91 To fully exploit the capabilities of 236U as a new ocean tracer it is necessary to determine the sources of 92 anthropogenic and natural 236U in all oceans. Specifically, we need to better constrain the input function 93 of 236U from the NFRPs, because it is still subject to substantial uncertainties (Christl et al., 2015b; 94 Periáñez et al., 2018) due to incomplete or not publicly available data on releases through time. 95 The U.S. GEOTRACES section GP16 (Eastern Pacific Zonal Transect, EPZT) was conducted during 96 autumn 2013 from Manta, Ecuador, to the French Polynesia. The cruise track was designed to sample 97 waters influenced by a variety of processes known to affect the behavior of TEIs and their isotopes, 98 including the long-lived artificial radionuclides (Schlitzer et al., 2018). The western EPZT section is 99 close to the location of the Pacific Proving Grounds (PPG, atmospheric and underwater), where the US 100 conducted nuclear weapon tests from 1946 to 1962 (Hisamatsu and Sakanoue, 1978; Livingston and 101 Povinec, 2000). These tests, unlike those of the French Polynesia (Chamizo et al., 2015b), were an 102 additional anthropogenic source of radioactivity, including 236U, to the atmosphere and thus, via fallout, 103 to the Pacific Ocean (Eigl et al., 2017; Froehlich et al., 2017). Also, the Pacific Deep Water (PDW), one 104 of the oldest water masses in the ocean is found in this area (Matsumoto, 2007), opening the possibility 105 to explore the current limits of the AMS measurement technique for 236U. 106 Here we present the AMS-derived concentration data for five well-resolved profiles of 236U collected 107 along the EPZT (Figure 1). Our work adds to the available dataset with respect to the distribution of 236U 108 in the global ocean; this includes the identification of anthropogenic sources of 236U as well as baseline 109 estimates of naturally occurring 236U concentrations in the Equatorial Pacific Ocean. It also complements 110 the characterization of water masses in the EPZT obtained from other constituents (e.g. nutrients, T-S- 111 Oxygen, He, 14C, etc.), exhibiting key features for the development of 236U as an additional water mass 112 tracer. 113 114 1.1. Regional distribution of water masses. 115 5 Hydrographic and nutrient measurements during GP16 were provided by the Oceanographic Data Facility 116 (ODF, Scripps Institution of Oceanography), and they are accessible through the Biological and Chemical 117 Oceanography Data Management Office website (http://www.bco-dmo.org/project/499723). Peters et al. 118 (2017) provided a thorough discussion of this data, which they used to analyze the distribution of water 119 masses along the EPZT and estimate the fractional contributions of endmember water masses in each 120 sample using Optimum Multi-parameter Analysis (OMPA). The geographic distributions of water 121 masses along with major currents influencing the EPZT are summarized in Figure 1. A T-S plot for the 122 five U-236 profiles as well as water mass end-member T-S values are shown in Figure 2a, and the T-S- 123 oxygen relationship is shown in Figure 2b. Temperature, salinity and oxygen distributions are shown in 124 Figure 3. 125 Subtropical Surface Water (STSW) is salty and warm relative to other water masses contributing to the 126 upper ~200 m of the water column (Llanillo et al., 2013). Equatorial Subsurface Water (ESSW) originates 127 at the equator (Silva et al., 2009) and is found below STSW (north of 15ºS) at ~250 m depth. It is less 128 salty and colder than the overlaying STSW and its contribution and depth varies along the transect (Silva 129 et al., 2009). ESSW is transported eastward to the South American coast, where it is advected into the 130 OMZ. Below ESSW (~250 m), at this latitude there is the Eastern South Pacific Intermediate Water 131 (ESPIW) - also referred to as Subantarctic Waters (SAAW) (Llanillo et al., 2013; Silva et al., 2009). 132 ESPIW originates north of the Subantarctic Front and flows north with the Pacific Coastal Current and 133 northwest within the South Pacific Subtropical Gyre. ESPIW is a fresh and temperate surface water mass 134 that is subducted beneath STSW near 35ºS. South Pacific Central Water (SPCW) occupies the 135 thermocline within the subtropical gyre and is separated from the South American coast by ESPIW and 136 ESSW. Equatorial Pacific Intermediate Water (EqPIW) is formed near the equator and is warmer and 137 saltier than ESPIW. Antarctic Intermediate Water (AAIW) is located beneath the ESPIW at 800-900 m. 138 AAIW is formed by vertical convection of the deep winter mixed layer located north of the Polar Front; 139 it spreads north and westward within the South Pacific subtropical gyre. 140 Pacific Deep Water (PDW) is colder and saltier than AAIW, with its core located between 2500 and 3000 141 m depth (Llanillo et al., 2013). PDW has no surface source, and as a result, it is one of the oldest water 142 masses in the global ocean (Matsumoto, 2007; Talley, 2011). It is formed in the deep waters of the 143 equatorial Pacific by upwelling of Antarctic Bottom Water (AABW) and subsequent vertical mixing. 144 6 Upper Circumpolar Deep Water (UCDW) is found near the bottom, flowing north over the Chile Rise 145 and westward toward the equator. It carries the signal of modified North Atlantic Deep Water (NADW) 146 and flows out of the Antarctic Circumpolar Current (ACC) into the Southeast Pacific Basin. Compared 147 to PDW, UCDW contains higher dissolved oxygen and lower nutrient levels (See Figure 2). Lower 148 Circumpolar Deep Water (LCDW) is formed from transformation of the North Atlantic Deep Water as it 149 flows into the Antarctic Circumpolar Current. Due to its higher salinity, LCDW occurs deeper than 150 UCDW (Peters et al., 2017). 151 Finally, AABW is characterized by a temperature minimum; it is the deepest and densest water found in 152 the low latitude Pacific. It also has a high oxygen content relative to the rest of the oceans' deep waters. 153 AABW is formed at a few sites along the continental margin of Antarctica during the winter months, in 154 polynyas along the shelves of the Antarctic continent. The newly formed highly saline shelf waters sink 155 down to abyssal depths. It is located below 4000 m in all ocean basins that have a connection to the 156 Southern Ocean at that level. The modified NADW upwells in the circumpolar region, as it reaches the 157 Southern Ocean, and contributes to the formation and sinking of the AABW in the coastal seas around 158 Antarctica (Peters et al., 2017). 159 160 2. Materials and Methods 161 2.1. Sampling 162 The U.S. GEOTRACES East Pacific Zonal Transect (EPZT) was sampled from October-December 2013 163 aboard R/V Thomas G. Thomson. The transect from Manta, Ecuador, to Papeete, French Polynesia, along 164 10-15°S (Figure 1) includes the Peru Margin, (Station 1- 12°S 74°W), the upwelling and oxygen 165 minimum zone (OMZ) (Station 11- 12°S 94°W), and the large hydrothermal plume (HP) (Resing et al., 166 2015) originating from the southern East Pacific Rise (Station 18 - 15°S 113°W). Two additional profiles 167 were sampled (Station 26 -at 12°S 128°W and Station 36 - 10°S 152°W). Approximately 25 depths along 168 the water column were sampled per profile for the analysis of 236U (Figure 1, Table 1). 169 Seawater samples were collected in plastic cubitainers using a rosette equipped with CTD sensors and 24 170 12- L Niskin bottles. Subsequently, samples were acidified with 5 ml of suprapure nitric acid to decrease 171 pH to less than 2 as indicated using pH paper. 172 7 The first uranium pre-concentration step was carried out at Lamont-Doherty Earth Observatory (LDEO), 173 Columbia University, New York. The 236U extraction procedure was carried out at Centro Nacional de 174 Aceleradores (CNA), Universidad de Sevilla, Spain. 236U measurements were carried out by Accelerator 175 Mass Spectrometry at two facilities: the 600 kV Tandy AMS system at ETH, Zürich, Switzerland (Christl 176 et al., 2013), and the 1 MV AMS system at CNA (Chamizo and López-Lora, 2018). 177 178 2.2. Sample preparation 179 About 4.5 L seawater samples were acidified with 65% HNO3 and spiked with about 3 pg of 233U. 233U 180 was provided by the ETH-Zürich and contained <0.1 fg 236U, i.e. the measured 236U/233U ratio of the 181 blanks was at a level of 10-5. The use of the spike allows calculating 238U and 236U concentrations. 182 Samples were mixed, settled for homogenization for at least 1 hour and subsequently 250 mg of Fe3+ 183 (previously purified to prevent 236U background contamination) were added to the samples for uranium 184 co-precipitation as Fe(OH)3 (López-Lora et al., 2018). 185 For the 236U separation, different chromatography extraction methods were used for samples measured at 186 CNA or ETH to meet the specific requirements of both facilities. For measurements at the CNA, uranium 187 samples must be plutonium free in order to correct the background caused by 235U on 236U using the 188 239Pu/238U as a proxy, as it is discussed in Chamizo et al. (2015a). To separate the plutonium from the 189 uranium, a TEVA® column was placed on top a UTEVA® column (López-Lora et al., 2018). This step 190 is not necessary in the case of the ETH facility, which provides the necessary abundance sensitivity to 191 suppress background caused by 235U on 236U by more than 11 orders of magnitude (Christl et al., 2015a). 192 Therefore, samples for measurement at ETH were only treated with a single UTEVA® column. The 193 different elution steps are detailed in (López-Lora et al., 2018). AMS cathodes were prepared similarly 194 in both cases using 1 mg of uranium-free Fe3+ for co-precipitation with Fe(OH)3, finally the samples were 195 mixed with 3 mg of Nb powder and pressed into aluminum cathodes (López-Lora et al., 2018). 196 During the process, 236U and 238U contamination was minimized. The 233U primary standard solution used 197 to spike the samples (IRMM-084) has a certified 236U/233U atom ratio at the level of 10-8, and thus the 198 associated 236U contamination was negligible. 199 8 Note that the relationship between 238U concentrations and salinity has been used in the past to calculate 200 238U concentrations (Chen et al., 1986). However, we did not use the global 238U salinity relationship 201 because it is not universal (Owens et al., 2011; Pates and Muir, 2007) . Some deviations in the correlation 202 have been found, especially in depth, where lower 238U activities might be caused by 238U removal 203 processes (Pabortsava, 2014). 204 205 2.3. Measurement of 236U by AMS 206 236U analyses were carried out by Accelerator Mass Spectrometry; details of techniques can be found in 207 Christl et al. (2015a) and in Chamizo et al. (2015a). The measurements were organized according to the 208 expected 236U/238U atom ratios of the samples. The 236U/235U abundance sensitivity that can be currently 209 achieved at the CNA is above the 10-9 level. As a result, a 236U/238U ratio below the 10-10 levels cannot be 210 assessed. In contrast, the Tandy-AMS at ETH offers an abundance sensitivity ~10-12. Therefore, surface 211 samples (i.e. upper 500 m), with highest expected 236U/238U ratios, were measured at the CNA, while 212 deeper samples, expected to have the lowest 236U/238U atom ratios, were measured at the ETH Zürich. 213 For the samples with the lowest counting statistics (i.e. 236U/238U atom ratios below 10-11), 236U was 214 counted for an overall time of about 1 h at the ETH Zürich. 215 For ETH measurements, the measured U isotopic ratios (i.e. 233,236U/238U) were normalized to the ETH 216 Zurich in-house standard ZUTRI, with a nominal 236U/238U value of (4055 ± 200)×10−12 (Christl et al., 217 2013). At the CNA, the normalization was carried out against an in-house secondary standard material, 218 CNA-U236, with a 236U/238U ratio of 3.5 × 10−8 and 236U/233U ratio of 0.71 (Calvo et al., 2015). This 219 standard was also measured at ETH for verification. 220 At both AMS facilities, chemistry blanks (i.e. spiked acid solutions that went through the columns and 221 cathode preparation steps) gave similar 236U count rates as the so-called instrumental blanks (i.e. cathodes 222 with the same matrix as the sample of interest and no additional processing), so a 236U contribution from 223 the radiochemistry could be neglected. The averaged background at ETH was, approximately, 3.5 × 224 104 atoms of 236U in a blank cathode, which corresponds to ~7 × 103 atom per kg of seawater. The blank 225 at CNA would be an order of magnitude higher. The measured background count rates, averaged for a 226 series of 6 cathodes of each type, were subtracted from the count rates of the samples. Uncertainties in 227 15 ± 0.4) ×10-10 and St. 18 (6.8 ± 0.5) ×10-10 increase with increased distance from the mouth of the river, 392 suggesting that the contribution of 236U from the Rio Negro River to our sampling locations is negligible 393 in compared to the influence of contributions from the Humboldt Current. 394 395 4.3. Longitudinal and depth distribution of 236U in deep waters. Identifying deep and bottom water 396 masses. 397 Below 2000 m, the concentrations of 236U in the north Pacific are about an order of magnitude higher 398 than those found along the EPZT (Table 2) (0.003 - 0.075 × 106 at kg-1 at EPZT and (0.094 ± 0.017) × 399 106 at kg-1 (St. 45ºN) and (0.45 ± 0.10) × 106 at kg-1 in the north Pacific). However, we note that Eigl et 400 al. (2017) interpreted their results for the deep samples as an upper limit of the concentrations due to 401 issues related to chemical and/or instrumental backgrounds. Eigl et al. (2017) observed a similar 236U 402 distribution pattern below 2000 m. 403 Along the EPZT, we generally measured low 236U concentrations below ~2000 m, with a minimum 404 concentration measured ~5×103 at kg-1, suggesting that the PDW carries a pre-anthropogenic 236U signal; 405 however, within the present limits of detection, it is not possible to definitively draw this conclusion. We 406 observe the lowest 236U/238U ratios between 2000 and 3000 m (average ratio: (1.5 ± 0.9) × 1012), which 407 is considered as an upper limit for pre-anthropogenic 236U levels. Despite the large uncertainties in the 408 data, we observe 236U-concentrations that are an order of magnitude higher (average concentration; (5 ± 409 2) × 104 at kg-1) at depths below 3000 m (Table 3). In the following sections, we explore contributions 410 from other water masses and physical-chemical behavior of 236U as possible explanations for these 411 observations. 412 4.3.1 Influence of other water masses 413 The first explanation to consider for the 236U increase at depths greater than 3000 m is the influence of 414 contributions from other water masses carrying different concentrations of 236U. Along the EPZT, there 415 are four main water masses below 2000 m (Figure 6): Pacific Deep Water (PDW) flowing South- 416 Eastbound in a complicated turn; the Upper Circumpolar Deep Water (UCDW) flowing towards the 417 equator and that spreads north over the Chile Rise; the Lower Circumpolar Deep Water (LCDW), that 418 16 flows from the Southern Ocean to the subtropical South Pacific; and the Antarctic Bottom Waters 419 (AABW), which is the deepest water in the low latitude Pacific. Both UCDW and LCDW are formed 420 from transformation of the North Atlantic Deep Water, AABW are newly formed waters. 421 The AABW is partly formed in the Pacific sector of the Southern Ocean. AABW is believed to renew 422 PDW from upwelling and subsequent vertical mixing in the Samoan Passage. However, the abyssal 423 waters of the Pacific Ocean are renewed very slowly due to topography and the deep basins of the Pacific 424 Ocean are filled from the west. As a result, PDW, without a direct surface source, is reported to be one 425 of the oldest water masses in all the oceans (Matsumoto, 2007; Talley, 2011). There are three entry routes 426 of AABW into the Pacific basins (Tomczak and Godfrey, 2013). The western route, northward flow east 427 of Australia, is blocked by topography around 20°S and its contribution is found negligible. The eastern 428 route, inflow along the EPR but is blocked by the Chile Rise near 40°S and does not contribute to bottom 429 water renewal. Therefore, the central route is considered to be the major point of supply. This way, 430 AABW, carried by the circumpolar Antarctic currents, enters the Southwest Pacific Basin and spreads 431 gradually northward until it enters the Northwest Pacific Basin through the Samoan Passage, at 9.5ºS 432 170ºW (Voet et al., 2016). This flow mostly takes place as narrow western boundary currents below 3500 433 m. In contrast, Northeast Pacific Basin is believed to be renewed by uniform slow eastward flow 434 (Tomczak and Godfrey, 2013). 435 In the northern hemisphere, water in the depth range 1000 - 3000 m, i.e. PDW, does not participate 436 significantly in the circulation because its properties are determined nearly entirely through the 437 mentioned slow mixing processes and thus very low 236U concentrations are expected for this water mass. 438 At 2000-3000 m, salinity, temperature and oxygen correspond to those expected for the Pacific Deep 439 Waters (Figure 6). The distribution of oxygen, temperature and salinity in the western longitudes of the 440 EPZT indicates the penetration of AABW below 3000 m. The signature of the LCPW and UCPW below 441 the PDW is identified in the eastern longitudes. 442 The values of natural background in the ocean were estimated by Steier et al. (2008), who calculated that 443 approximately 0.5 kg of natural 236U is in the oceans, a concentration of natural dissolved 236U < 103 at 444 kg-1 and 236U/238U <1.2 × 10-13. Casacuberta et al. (2016) reported one of the lowest 236U concentrations 445 and 236U/238U atom ratios in Makarov Basin in the Arctic Ocean, respectively 3 ×104 at kg-1 and 5 × 10- 446 12 at 3500 m. In the North Pacific Eigl et al. (2017) reported concentrations of 7 × 104 at kg-1 and ratios 447 17 of 9 × 10-12 at 2000-3000 m at 50ºN. Our results at 2000-3000 are concentrations of 3 ×103 at kg-1 and 448 236U/238U ratios at the 10-12 order, close to the magnitude of the natural values estimated by Steier et al. 449 (2008), namely 10-13. These concentrations are, up to the present, the lowest 236U measured in seawater. 450 Minimum 236U concentrations values are systematically found at those depths that corresponds to the 451 depth of the identified PDW. 452 PDW waters are old enough to carry a pre-nuclear 236U signal. According to Matsumoto (2007), and using 453 radiocarbon data, certain sections of the Pacific Deep Water has an estimated age of 700-1200 years. For 454 this reason, we explored the possibility that we might be measuring the natural signal of 236U in the ocean 455 in this specific water mass. However, measured 236U/238U ratios are close to the limit of detection of the 456 technique (10-12) and this prevents us from defining a natural background. Instead, we use our 236U 457 values in the PDW as an upper limit. Transit time of the PDW is 1200 years, with the oldest waters found 458 on the Northern East Pacific. Equatorial waters below 1500 m are estimated to be 700-900 years old, i.e. 459 considerably older than the nuclear era. Nevertheless, we cannot know for certain if the 236U minimum 460 detected in the PDW corresponds to pre-nuclear levels, since at this latitude (10ºS) and depth (2000-3000 461 m) 236U concentration is probably affected by the contribution of Antarctic bottom waters flowing 462 northwards and upwelling in the Samoan Passage. 463 The already mentioned increase of 236U below 3000 m changes from 0.003 ×106 to around 0.04 - 0.07 × 464 106 at kg-1 (Figure 4c and Table 1). And these differences at 2000 m and 4000 m point out that with the 465 use of 236U it is possible to differentiate deep and bottom water masses. To further analyze the different 466 concentrations in terms of water masses we used a specific scale to zoom in the deepest 2000 m and depict 467 only the lowest 236U concentrations. We zoomed T-S diagrams versus 236U concentration (Figure 5c) and 468 these higher values correspond to the values associated with UCDW, LCDW and AABW, mixed also 469 with the lowest values, in the 0.006-0.003 ×106 at kg-1 range (Table 1), that correspond to the PDW. 470 To geographically identify 236U differences, we zoomed below 2000 m in the 236U longitudinal and depth 471 distribution (Figure 6a), this distribution clearly shows the minimum concentration around 2000-3000 m 472 (PDW) and the increase in the concentration near the seafloor, that can be observed mainly in the western 473 and easternmost borders (contribution of AABW and LCDW respectively). 474 This in agreement with the description of the water masses in the area done by Peters et al. (2017). The 475 18 OMPA analysis suggests that AABW contributes mainly to the deepest depths of the west longitudes, 476 whereas LCDW penetrates into the deepest Equatorial Pacific through gaps in the eastern Chile Rise and 477 subsequently flows northward into the Peru basin (Tomczak and Godfrey, 2013). 478 Figures 6b, 6c and 6d show higher salinities and oxygen and lower temperatures below 2000 m for the 479 water masses of Antarctic origin, especially in the western longitudes. According to Peters et al. (2017), 480 AABW fractions exceeded 40% on the western end but would be below 20% east of the East Pacific 481 Ridge (EPR). LCDW contributions ranged between 0 and 50% between 2000 m and 5000 m, 482 interestingly the deepest waters samples collected within the Peruvian Trench at 79.1°W (St. 1) 483 demonstrated relatively high LCDW fractions (up to 45%) compared to bottom waters elsewhere along 484 the ridge axis. This is consistent with a northward penetration of deep waters into the Peru and Panama 485 basins. The contribution of UCDW can be seen in shallower depths. 486 Therefore, the slightly higher concentrations of 236U found near the bottom at the EPZT borders is 487 associated with the invasion of relatively younger waters from the south, carrying the modified signal 488 from the North Atlantic Deep Waters (see section 1.1) both in the westernmost part and in the Peru- 489 Chile trench on the east side of the EPZT. In fact, according to Matsumoto (2007), in its paper regarding 490 the circulation age of the oceans radiocarbon-based, a tongue of younger water extends northward from 491 30 to 40°S to nearly the equator around 110–120°W. 492 A similar conclusion is reached by Jenkins et al. (2017), using the concentration of the helium isotopes 493 and radiocarbon. The distribution of Δ14C in the deep Pacific shows a mid-depth (2000–3000 m) 494 minimum, coincident with our 236U minimum and associated with the PDW. Furthermore, higher Δ14C 495 are detected below 3000 m, associated with Antarctic Bottom and North Atlantic Deep Waters (Jenkins 496 et al., 2017). In the western longitudes of the EPZT, they detect bottom waters with relatively less 497 hydrothermal helium and more radiocarbon, indicating relatively younger waters entering in the deep 498 Pacific. Additionally, at the eastern longitudes of the EPZT in the deepest samples we detect the invasion 499 of relatively young bottom waters with high Δ14C low δ3He. They enter the Chile Basin and flow to the 500 Peru Basin. Our 236U results follow a similar pattern. Peters et al. (2017) designated these waters as 501 LCDW. 502 Table 3 summarizes the results from the four water mass tracers used for deep and bottom waters 503 encountered on the EPZT. The four tracers are able to clearly differentiate PDW from AABW in the 504 19 western longitudes. On the contrary, only 236U and 14C tracer is able to clearly detect the LCDW signal 505 entering through the Peru-Chile trench. Only 14C tracer is able to distinguish between results from 506 Eastern and Western longitudes. 507 508 4.3.2 Physical-chemical behavior of 236U 509 A second possible explanation for increased concentrations at depth relates to the physical-chemical 510 behavior of 236U. Below, we discuss 3 possible scenarios: 1) Scavenging and removal of dissolved 236U 511 from the surface by sinking particles coupled with remineralization at depth, 2) in-situ production of 236U 512 from 240Pu that may be present in sediments as a result of scavenging, and 3) rapid downward transport 513 and removal to sediments of fallout debris containing 236U at the time of atmospheric deposition, followed 514 by dissolution and release to bottom waters. 515 Eigl et al. (2017) suggested that the 236U increase near the sea-floor could be due to the remineralization 516 at depth of sinking particles. However, similar to naturally occurring dissolved uranium, dissolved 236U 517 is assumed to be conservative in the open-ocean (Sakaguchi et al., 2009). In this case, the amount of 518 dissolved 236U in surface waters that could be adsorbed by particles has been estimated to be ≤2% of the 519 total (IAEA, 2004).This coupled with the observation that in the open ocean ≤3% of sinking particles at 520 the surface actually reach the sea-floor (Lampitt et al., 2010) suggests that a negligible fraction (≤104 521 atoms/L) of the global fallout inventory of 236U, nearly all of which is present at or near the surface along 522 the EPZT, could be removed to bottom waters as a result of scavenging. This is consistent with Christl et 523 al. (2012a), where they applied a simple conceptual model in the Atlantic Ocean and concluded that 524 scavenging and subsequent remineralization of uranium at depth is not a significant mechanism supplying 525 anthropogenic 236U to the deep ocean. 526 A second scenario is the in-situ production of 236U by alpha decay of 240Pu (T1/2=6564 y), which may 527 have accumulated in the surface layers of the sediments over the past 60 years. A surplus has been 528 observed of about 100% of the 239+240Pu inventories in the central NW Pacific Ocean compared to that 529 estimated from global fallout deposition. This additional Pu inventory is associated with local fallout from 530 the U.S. weapons tests carried out at the PPG at Bikini and Enewetak Atolls between 1946 and 1958. The 531 large number of tests had a large impact on neighboring areas (Livingston et al., 2001). Buesseler (1997) 532 suggested that fallout Pu from the PPG is more rapidly removed from the surface waters than Pu from 533 20 fallout. We assume a 239+240Pu excess inventory of 6 Bq/m2 at the seafloor from the US tests at the 534 Marshall Islands and we assume that 0.18 is the averaged 240Pu/239Pu (Kelley et al., 1999). Thus, about 535 2.7 Bq/m2 or 7.11×1011 atoms/m2 of 240Pu would be accumulated in the abyssal sediments. From 1958 to 536 2013, about 0.6% of these atoms at the seafloor decayed into 236U, according to their half-lives, assuming 537 that 50% is remineralized, the upper limit for 236U concentration would be 1.6×103 atoms/L in the deepest 538 1500 m of the water column. 539 The third scenario relates to fallout debris containing 236U at the time of atmospheric deposition reaching 540 the seafloor as a result of dissolution resistance or rapid sinking due to particle size or density, followed 541 by dissolution and release to bottom waters. Assuming a 236U/239Pu atom ratio in fallout debris of ~0.2 542 (Froehlich et al., 2017) and a 239+240Pu excess inventory of 6 Bq/m2 from the US (discussed above), there 543 will be a 236U excess inventory of 1.2 Bq/m2. If all the 236U is in the dissolved form, we estimate an upper 544 limit on 236U concentration of ~103 atoms kg-1. This estimate is on the same order of magnitude as 545 naturally occurring 236U in the ocean. Therefore, it is not likely that processes related to particle sinking 546 and remineralization are sufficient to account for the increases in 236U measured below 3000 m, which 547 are nearly an order of magnitude higher than our estimates. Nonetheless, further information (e.g. Pu 548 and/or 236U concentration in sea-floor sediments) is needed to further analyze these hypotheses and its 549 implication for 236U concentration in depth. 550 551 5. Conclusions 552 The long-lived radioisotope 236U was measured by AMS in samples collected on the 2013 US 553 GEOTRACES Eastern Pacific Zonal Transect (GP16). The major features of 236U distribution include; 1) 554 The highest concentrations and bulk of the total 236U inventory resides in the upper water column <600 555 m; 2) An increase in maximum concentrations and inventories from east to west along the transect; 3) 556 Minimum concentrations along the transect between 2000 and 3000 m, which are at or near estimates for 557 naturally occurring 236U levels, and, to our knowledge, the lowest 236U values reported for the open ocean. 558 Based on inventories, concentration levels, and distribution profiles, the primary conclusions of our work 559 are: 560 1) Global fallout is the only significant source of 236U to waters encountered along the EPZT. 561 21 2) The observed distributions, both horizontal and vertical, rather than a result of the physical-chemical 562 behavior of 236U, are related to contributions from various water masses containing different levels of 563 236U, most notably in the deep water masses. 564 3) We observe consistent 236U minimum concentrations between 2000 and 3000 m, which we attribute 565 to contributions from Pacific Deep Waters (PDW). The 236U distribution pattern is similar to those 566 observed for 14C and 3He isotopes (Jenkins et al., 2017). Given the age of PDW, this establishes an 567 upper limit on natural 236U concentration values. 568 4) We observe an increase in 236U concentrations below 3000 m towards the seafloor, particularly in 569 western and eastern areas of the EPZT. AABW contributes mainly to the deepest depths of the west 570 longitudes and LCDW penetrates into the deepest Equatorial Pacific through the Peru-Chile trench. This 571 is in agreement with the description of the water masses done by Peters et al. (2017), and it is the most 572 probable reason for the detected increase of 236U in bottom waters. 573 5) Our results add to the global dataset of 236U measured in seawater. They provide baseline values 574 associated with the Pacific deep water masses for the first time. Agreement with information derived 575 from other tracers such as 14C and 3He isotopes demonstrates the potential of 236U as an additional water 576 mass tracer. Features such as a geochemically conservative behavior, transient signal, and the ability to 577 detect low levels via AMS (i.e. below 103 atoms kg-1) make it an attractive option. 578 579 580 Acknowledgements 581 This work was partially supported by project FIS2015-69673-P of the Spanish Ministerio de Economía 582 y Competitividad: Resolución de problemas ambientales marinos y terrestre clave mediante nuevos 583 desarrollos en Espectrometría de Masas con Acelerador de Baja Energía (LEAMS) en el Centro Nacional 584 de Aceleradores (CNA) and by Fundación Cámara Sevilla through a Grant for Graduate Studies of M. 585 López-Lora. This work was made possible with funding from NSF Chemical Oceanography award OCE- 586 1235233. We thank you Marty Fleisher and Leo Pena, ODF group members Susan Becker, Melissa 587 Miller, Rob Palomares, and Mary Johnson. Captain and crew of R/V Thomas G. Thompson (TGT303). 588 Chief Scientist: Dr. James Moffett, and Co-Chief Scientists Dr. Christopher German and Dr. Gregory 589 Cutter. We sincerely thank two anonymous reviewers for their valuable inputs. 590 591 Figure captions 592 22 Figure 1. Study area of the EPZT cruise, showing hydrographic parameters (Oxygen Minimum Zone 593 (OMZ) and hydrothermal plume (HP) and main patters of water circulation (Llanillo et al., 2013; Silva 594 et al., 2009; Yamada and Wang, 2007)). Red lines are associated with surperficial warm currents and 595 blue lines to deep cold currents. Dotted line represents bottom waters, dashed line represents surficial 596 waters and solid line represents intermediate waters (NEC – North Equatorial Current; ECC Equatorial 597 Counter Current- ; SEC – South Equatorial Current; HC – Humboldt Current; STSW - Subtropical Water; 598 ESSW - Equatorial Subsurface Water; SPCW - South Pacific Central Water; ESPIW - Eastern South 599 Pacific Intermediate Water; AAIW - Antarctic Intermediate Water; PDW - Pacific Deep Water; LCDW 600 - Lower Circumpolar Deep Water; AABW - Antarctic Bottom Water). 601 Figure 2. a) Temperature versus Salinity with dissolved U-236 stations identified (1, 11, 18, 26, 36). b) 602 Temperature versus Salinity with dissolved oxygen overlay. Water mass abbreviations are as follows: 603 STSW - Subtropical Water; ESSW - Equatorial Subsurface Water; SPCW - South Pacific Central Water; 604 EqPIW - Equatorial Pacific Intermediate Water; ESPIW - Eastern South Pacific Intermediate Water; 605 AAIW - Antarctic Intermediate Water; UCDW - Upper Circumpolar Deep Water; PDW - Pacific Deep 606 Water; LCDW - Lower Circumpolar Deep Water; AABW - Antarctic Bottom Water. 607 Figure 3 EPZT sections a) Dissolved 236U concentrations, b) Temperature and c) Salinity. 608 609 Figure 4. Dissolved 236U profiles from the EPZT (filled symbols) compared to two profiles from the 610 North Pacific at 32ºN and 45ºN (Eigl et al., 2017) and one profile from the Equatorial Atlantic, at 47ºW 611 (Casacuberta et al., 2014). Colors are referred to those of Figure 1. Concentrations units are 106 atoms 612 kg-1. 613 Figure 5. T-S diagram as a function of 236U concentration from a) surface to sea bottom (all temperatures) 614 b) expanded scale for 10ºC to 30ºC c) expanded scale for 0 to to 10ºC. 615 Figure 6. EPZT section plots for depths greater than 2000 m for a) 236U b) temperature c) oxygen and 616 d) salinity. Main water masses are represented (ESPIW - Eastern South Pacific Intermediate Water; 617 AAIW - Antarctic Intermediate Water; UCDW - Upper Circumpolar Deep Water; PDW - Pacific Deep 618 Water; LCDW - Lower Circumpolar Deep Water; AABW - Antarctic Bottom Waters). 619 620 23 Latitude Longitude Geotraces # Depth 236U/238U 238U 236U m 10-12 at at-1 uncert g kg- 1 uncert 106 at kg-1 uncer 12.0ºS 79.2ºW 2160 20 481 10 3.19 0.06 3.9 0.1 St. 1 2157 65 470 10 3.11 0.06 3.7 0.1 2155 81 517 9 3.09 0.06 4.0 0.1 2153 181 503 10 3.09 0.06 3.9 0.1 2151 212 459 9 3.19 0.06 3.7 0.1 2135 235 436 9 3.08 0.06 3.4 0.1 2180 302 342 8 3.22 0.06 2.8 0.1 2178 453 83 3 3.24 0.06 0.7 0.0 2176 556 38 3 3.10 0.06 0.30 0.02 2174 808 10.0 1.9 3.17 0.06 0.081 0.015 2172 1210 1.8 2.1 3.20 0.06 0.015 0.017 2215 1616 3 2 3.24 0.06 0.02 0.01 2213 2426 1.6 1.2 3.17 0.06 0.013 0.010 2211 2833 2.4 1.5 3.18 0.06 0.019 0.012 2209 3242 6.0 1.6 3.35 0.07 0.051 0.013 2207 3652 0.7 2.5 3.07 0.06 0.005 0.019 2205 4469 2.7 1.8 3.27 0.07 0.023 0.015 2194 4881 4.2 1.9 3.26 0.07 0.035 0.016 2192 5140 4.9 1.9 3.20 0.06 0.040 0.016 2190 5419 4.8 1.5 3.25 0.06 0.040 0.013 2188 5492 5.4 1.1 3.31 0.07 0.045 0.009 2186 5550 3.4 1.0 3.21 0.06 0.027 0.008 2184 5600 2.3 1.4 3.15 0.06 0.018 0.011 12ºS 94ºW 3580 2 578 35 3.37 0.08 5.1 0.3 St. 11 3681 20 606 14 3.31 0.02 5.3 0.1 3678 40 458 32 3.45 0.06 4.2 0.3 3677 75 514 39 3.48 0.10 4.7 0.4 3675 110 558 32 3.31 0.12 4.9 0.3 3673 130 352 27 3.52 0.07 3.3 0.2 3671 180 502 18 3.28 0.05 4.2 0.1 3661 218 410 15 3.34 0.05 3.5 0.1 3659 260 285 12 3.21 0.05 2.3 0.1 3657 320 174 14 3.03 0.05 1.3 0.1 3655 400 104 5 3.25 0.05 0.86 0.04 3653 600 14 2 3.22 0.05 0.12 0.02 24 3651 800 7 2 3.26 0.05 0.06 0.02 3649 1000 0.2 1.0 3.29 0.05 0.002 0.008 3647 1250 1.4 1.2 3.22 0.05 0.011 0.010 3643 1750 5.6 2.0 3.24 0.05 0.046 0.017 3641 2000 1.1 1.3 3.20 0.05 0.009 0.010 3639 2250 0.3 1.4 3.33 0.05 0.003 0.012 3612 2500 1.8 1.2 3.25 0.05 0.015 0.010 3610 2750 <0.8 3.20 0.05 <0.006 3608 3000 9.1 1.8 3.29 0.05 0.075 0.015 3606 3500 5.1 1.8 3.23 0.05 0.042 0.015 3604 3600 3.4 1.4 3.30 0.05 0.028 0.012 3602 3625 3.3 1.6 3.31 0.05 0.027 0.014 14.9ºS 112.7ºW 8478 2 677 56 3.67 0.17 6.6 0.5 St. 18 8580 20 578 33 3.65 0.06 5.6 0.3 8578 60 639 37 3.72 0.13 6.3 0.4 8576 95 443 30 3.63 0.06 4.2 0.3 8574 150 537 40 3.54 0.10 5.0 0.4 8572 200 449 30 3.50 0.14 4.2 0.3 8570 355 257 10 3.29 0.05 2.14 0.08 8558 400 148 8 3.32 0.05 1.24 0.06 8554 600 19 2 3.14 0.05 0.147 0.018 8552 1000 1.8 0.6 3.14 0.05 0.015 0.005 8550 1200 1.9 0.8 3.14 0.05 0.015 0.006 8548 1400 0.2 1.1 3.36 0.05 0.002 0.009 8522 1600 3.1 0.9 3.20 0.05 0.025 0.008 8520 1800 0.7 1.2 3.32 0.05 0.006 0.010 8518 2000 3.9 1.4 3.21 0.05 0.032 0.011 8514 2200 2.5 0.7 3.24 0.05 0.020 0.006 8516 2200 3.8 1.6 3.30 0.05 0.032 0.013 8512 2300 2.3 0.7 3.26 0.05 0.019 0.006 8409 2350 8.2 2.3 3.33 0.05 0.069 0.020 8407 2400 6.3 1.0 3.23 0.05 0.052 0.008 8405 2450 14.3 2.7 3.32 0.05 0.12 0.02 8403 2500 1.3 0.6 3.29 0.05 0.011 0.005 8399 2610 9 1 3.29 0.05 0.073 0.011 11.7ºS 128ºW 9338 2 602 46 3.46 0.11 5.5 0.4 St. 26 9484 20 616 48 3.61 0.13 5.9 0.5 31 813