PAHs in the North Atlantic Ocean and the Arctic Ocean: Spatial PAHs in the North Atlantic Ocean and the Arctic Ocean: Spatial Distribution and Water Mass Transport Distribution and Water Mass Transport

Abstract


Introduction
The Arctic Ocean, linking the Atlantic Ocean and the Pacific Ocean, is a fundamental node in the global hydrological cycle and thermohaline circulation (Talley et al., 2011;Carmack et al., 2016).A major inflow to the Arctic Ocean comes from the Atlantic through the Fram Strait and the Barents Sea, with a minor inflow from the Pacific side through Bering Strait (Ma et al., 2018;Liu et al., 2021a).The outflow from the Arctic Ocean, being cold and fresh, is mainly conveyed by East Greenland Current through the Fram Strait, which is the only deep channel allowing energy and materials to exchange between the Arctic Ocean and other oceans (Wang et al., 2021).The Arctic is the most sensitive area worldwide and is undergoing visible and less visible changes such as warming, refreshing, and sea ice loss (Morison et al., 2012;Dai et al., 2019;Ko et al., 2020).External to the Arctic Ocean, the bordering subarctic oceans are undergoing substantial changes in heat, salt, and biogeochemical properties, therefore amplifying the climate response of the Arctic Ocean (Steele and Boyd, 1999).
Semivolatile organic compounds (SVOCs) exist in the environments ubiquitously (Lohmann et al., 2007;Hung et al., 2010;Xue et al., 2016).Among them, polycyclic aromatic hydrocarbons (PAHs) are of great concern because of their toxicity or carcinogenic effects and the ongoing emissions from pyrogenic or petrogenic sources (Shen et al., 2013;Balmer et al., 2019;Du et al., 2020;Wang et al., 2020).Similar to persistent organic pollutants (POPs) with long-range transport potential, PAHs could arrive in the Arctic Ocean through the atmospheric or oceanic pathway, but their fate remains unclear (Ma et al., 2013;Liu et al., 2021a).On the one hand, climate changes may lead to PAHs' re-volatilization from the Arctic (Nizzetto et al., 2010;Yu et al., 2019).On the other hand, their biogeochemical processes, e.g., biodegradation, photodegradation, and vertical sinking, would be further impacted due to the changes in temperature, light, and degrader species (Keyte et al., 2013;Deyme et al., 2011;González-Gaya et al., 2019).Hence, it is challenging but necessary to better understand the transport and fate in the changing Arctic.
Hydrological processes, such as ocean current transport, would make a crucial contribution to the long-range transport of POPs, especially in the deep ocean (Lohmann and Belkin, 2014).Inflowing Atlantic and Pacific waters eventually flow out to the North Atlantic after experiencing cooling and freshwater input in the Arctic Ocean; thus, the pollutants in the Arctic Ocean are likely to be transported via ocean currents on a larger ocean scale (Carmack et al., 2016).According to the limited studies, the outflows of hexachlorocyclohexanes (HCHs) and hexachlorobenzene (HCB) are detected in the deep waters from the Arctic Ocean through Fram Strait, while polychlorinated biphenyls (PCBs) are still loaded from the Atlantic Ocean to the Arctic Ocean (Ma et al., 2018).The previous study has suggested the Atlantic meridional overturning circulation and deep ocean transport reduce perfluorooctane sulfonate (PFOS) accumulated in the Arctic Ocean (Zhang et al., 2017).For the nonvolatile perfluorooctanoic acid (PFOA), oceanic transport would make a more significant contribution to pollutants in the Arctic (Stemmler and Lammel, 2010).
Those results all implied that oceanic transport in the deep ocean is important for their distribution and storage.However, for POPs with different emission trends, it is unclear whether their export from the Arctic Ocean via ocean currents exceeds their input, or whether the Arctic Ocean is still a net sink for pollutants (Sobek and Gustafsson, 2014).
Oceanic transport of POPs is not only a physical mixing process, but also probably influences their biogeochemical processes, such as biodegradation and particulate settling, by changing temperature, dissolved oxygen, and seawater nutrients (Liu et al., 2021b).The strong-stratified Arctic Ocean undergoes complex inputs and outputs of PAHs from the neighboring oceans and continents.The upper manuscript submitted to JGR Oceans ocean plays a crucial role in transporting PAHs from the surface to their major sinks.
To better understand PAHs' transport processes and their contribution to high-latitude oceans, we investigated surface seawater and water-column samples in the North Atlantic Ocean and the Arctic Ocean.The objectives of this study were (i) to obtain broad-scale spatial distributions and sources of PAHs in the surface seawater of the North Atlantic and the Arctic Ocean, (ii) to obtain the vertical profiles of PAHs in different water masses and investigate the potential biogeochemical influences, (iii) to estimate mass flows of PAHs between the North Atlantic and the Arctic Ocean, and to evaluate the role of oceanic transport in PAHs' cycling.

Sample collection
Samples were conducted in the Arctic (Canada Basin, Amundsen Basin, East Siberian Sea, and Barents Sea) and the North Atlantic (Iceland Sea, Greenland Sea, and Norwegian Sea) during the 5th Chinese National Arctic Research Expedition between July and September 2012, onboard RV Xuelong (Figure 1).Surface seawater samples were taken from 18 sites, and in addition, water-column samples were taken from four sites (AT06, BB04, IS02, and SR18).These four sites were located in different areas with different hydrodynamics and bottom depths (821-3409 m), as AT06 and BB04 were in the Norwegian Sea, IS02 in the Greenland Sea, and SR18 in the central Arctic.Details of the sampling sites are shown in the Supporting Information in Table S1, and the sampling volume for each sample was approximate 4 L. Briefly, surface seawater samples were collected using the vessel's seawater intake system (stainless steel pipe), and water-column samples were collected in Niskin bottles mounted on a CTD rosette (Seabird 911/17) at multiple layers.The collected samples were immediately filtered through a pre-combusted (450 o C, 4 h) Whatman glass microfiber filter (GF/F; 47 mm diameter, 0.7μm pore size) and stored in amber glass bottles.The solid-phase extraction (SPE) C18 cartridges were pre-cleaned with dichloromethane, followed by acetone and hexane on land, and were pre-conditioned with methanol followed by ultrapure water onboard.Spiked with surrogate standards (acenaphthylene-d10, phenanthrene-d10, chrysene-d12, and perylene-d12), the filtrate was passed through pre-conditioned cartridges at a flow rate of 6 mL min −1 .After the extraction, the cartridges were well wrapped with pre-combusted aluminum foil and stored at −20°C until sample pretreatment.

Pretreatment and instrumental analysis
The pretreatment for dissolved PAHs was conducted as described earlier (Liu et al., 2021a).In brief, for dissolved PAHs, SPE cartridges were eluted with 10 mL of ethyl acetate and anhydrated by pre-combusted anhydrous sodium sulfate.The eluents were solvent-exchanged to n-hexane and concentrated to 1 mL using a rotary evaporator.After further evaporation under a gentle nitrogen stream, the mixture was spiked with 5 ng of internal standard (pyrene-d10), and stored at −4°C before the instrumental analysis.16 USEPA priority PAHs were analyzed by gas chromatography coupled with double mass spectrometry (GC-MS-MS, Agilent 7890A-7000B).Details of the instrumental analysis are shown in Text S1. manuscript submitted to JGR Oceans 2.3 QA/QC Field blanks, lab blanks, and spiked surrogates were performed to control data quality.Each field blank (Table S3) was performed using 4 L of distilled water and shared the same treatment from storage in an amber glass bottle and filtration to the final instrumental analysis.Laboratory blanks (Table S3) were performed on cleaned SPE C18 cartridges in the same manner as the laboratory samples to prevent any contamination originating from the described experimental processes.Instrumental detection limits and the ions monitored are shown in Table S2.Among the 16 USEPA PAHs analyzed here, 10 PAHs showed detectable concentrations, namely, naphthalene (Nap), acenaphthylene (Acpy), acenaphthene (Acp), fluorene (Flu), phenanthrene (Phe), anthracene (Ant), fluoranthene (FluA), pyrene (Pyr), benzo(a)anthracene (BaA), and chrysene (Chr).Method detection limits were derived from the mean field blanks plus three times the standard deviation of the field blanks, and ranged from <0.01 ng L −1 (Chr) to 4.30 L −1 (Nap), as shown in Table S3.Nap was not discussed in this study because of the potential influence in the field and lab, hence nine PAH compounds were discussed in this study.
Average recoveries of surrogate internal standards Acp-d10 and Phe-d10 were 61 ± 20 and 58 ± 21%.Five-or six-ring PAHs were not reported due to their low recoveries (Chr-d12: 43 ± 15, and perylene-d12: 39 ± 17%), and most of them were under the method detection limits.Although the polymeric sorbents produced lower recoveries of the more volatile compounds, the C18 cartridge is considered a preferred option for PAHs in water with good performance (Martinez et al., 2004).Sample results were corrected for field blank values but not corrected for recoveries.

Data Analysis
The statistical analysis was performed using SPSS (version 25).The significant difference test was using the Mann-Whitney U test and Kruskal-Wallis test, with p <0.05 indicating statistical significance.Principal component analysis (PCA) was performed with varimax rotation, and principal components with eigenvalues >1 were extracted.The calculation was performed using Microsoft Excel (2019 Pro Plus), and figures were produced by software Ocean Data View (version 5.1.5)and Grapher (version 15.3.339).
Positive matrix factorization (PMF) analysis was performed using EPA PMF 5.0, and its concept and application have been described in detail in EPA PMF 5.0 Fundamental and User Guide (USEPA, 2014).The undetectable value was replaced with one-half of its PAH method detection limits, and the uncertainty was set as 20% for each PAH dataset.The model was run for 3-6 factors with random seeds, and the output stability and reliability were checked according to Q value, residual analysis, and correlation coefficients.

Concentration and sources
For surface seawaters collected from the Arctic and North Atlantic, the concentrations of 9 PAHs (∑9PAH) in the dissolved phase ranged from 0.3 to 10.2 ng L −1 (mean 4.3 ng L −1 ) (Figure 2, Table S4).The average concentration of 9 PAHs in the Arctic Ocean (4.9 ± 2.8 ng L −1 ) was higher than in the North Atlantic Ocean (3.4 ± 2.6 ng L −1 ).The highest concentration was observed in the Barents Sea, near the Novaya Zemlya (AO04), while the lowest was found in the Norwegian Sea, North Atlantic (BB09).Three-ring compounds averagely contributed 86% to the total concentrations, among which Phe and Flu accounted for 34% and 31% of ∑9PAH.
PAH concentrations observed in this study were higher than those in the West Atlantic Ocean (∑7PAH: ND-8.1 ng L −1 , the concentrations were recalculated for our target compounds and the same below) (Lohmann et al., 2021) S5.Paired diagnostic ratios of certain individual PAHs have been widely used in the source estimation of PAHs (Yunker et al., 2002;Tobiszewski and Namiesnik, 2012).Considering PAHs in the dissolved phase dominated by low/medium molecule-weight compounds, we used FluA/(FluA + Pyr) and Ant/(Ant + Phe) to assess their possible sources.The FluA/(FluA + Pyr) ratio of less than 0.4 indicates a petrogenic source, and a value between 0.4 and 0.5 indicates mixed sources, while a value of more than 0.5 suggests a combustion source (Yunker et al., 2002).As shown in Figure S1, the FluA/(FluA + Pyr) ratios were mainly more than 0.4, implying that combustion and petroleum combustion were the major sources of PAHs in seawater.
Besides, the ratio Ant/(Ant + Phe) could distinguish petrogenic and pyrogenic sources for PAHs (Yunker et al., 2002).Ant/(Ant + Phe) ratios, mostly more than 0.1, indicated the combustion source.It is worth noting that paired compounds usually have different reactivities to photodegradation and biodegradation, which may lead to a variance in the ratio value after long-range transport (Yunker et al., 2002;Liu et al., 2021b).
We further applied the PMF receptor model to estimate source profiles of PAHs in surface seawater by inputting 18 objects (samples), nine variables (PAHs), and their uncertainty data.Three factors were finally chosen according to Q robust and Q true values.The average contributions of nine PAHs to the three PMF factors are shown in Figure 3a, and these factor profiles are shown in Figure S2.Factor 1 made the least contribution (5%) to the total measured PAHs and was dominated by Pyr, a typical product of wood/coal combustion (Li et al., 2016).Factor 2 explained 15% of the PAH concentrations, with high loadings of Phe, Ant, FluA, BaA, and Chr.These PAH compounds have been regarded as typical products of wood/coal combustion and emission from gasoline and diesel vehicles, hence the profile of Factor 2 was indicative of combined combustion sources (Bzdusek et al., 2004;Zhang et al., 2021).Factor 3, accounting for the highest proportion (80%) of PAH concentrations, was mainly composed of Acpy, Acp, and Flu.This factor profile was of high consistency with PAH emission characteristics from gasoline and diesel combustion (Zeng et al., 2018).As shown in Figure S3, the PAH profile of the field blank showed some similarity to Factor 3, revealing the onboard contamination may partly come from the emission from the vessel's diesel-based engines.It is worth noting that our limited input samples might lead to certain uncertainty in the PMF results.However, we suggested vehicle emissions and biomass combustion being the major PAHs sources in the surface seawater.factor contribution to PAHs levels at each sampling site.

Distribution Patterns of PAHs
The spatial distribution of PAH concentrations in surface seawater showed an "Arctic Shelf > Atlantic Ocean > Arctic Basin" pattern, although the differences among these three areas were not statistically significant (p >0.05).The average ∑9PAH concentrations were 5.5, 3.4, and 2.4 ng L −1 in the Arctic Shelf, the Atlantic Ocean, and the Arctic Basin.The average concentrations of 3-ring PAHs and FluA were highest in the Arctic shelf area (Figure S4), while the average concentrations of PAHs with higher molecular weights were highest in the Atlantic Ocean.The distribution of higher levels in the shelf area and lower levels in the basin area was similar to those of other SVOCs, such as PCBs and PBDEs (Carrizo et al., 2017).
Previous studies have suggested the crucial role of long-range transport for SVOCs in high-latitude environments according to their latitudinal distribution (Lohmann et al., 2009;Zheng et al., 2021).However, no significant latitudinal trends were founded for individual PAHs or their sum concentrations (p: 0.156-0.902)except Chr (p <0.05), which indicates the local PAH input was also an important factor influencing their distribution trend.A significant decrease of PAHs was observed from the shelf area to the central basin of the Arctic Ocean (p <0.05), indicating a decreased input or enhanced depletion occurring in the upper ocean.Especially regarding the depletion mechanism, a strong particulate export could give rise to PAHs reducing during their lateral transport in the Arctic margins as the "shelf sink effect" (Liu et al., 2021a).
Regarding PAH compositions, only Acpy showed significant spatial distinctions (p <0.05).Figure 3b shows the relative contributions of PMF factors to ∑9PAH levels in each surface seawater sample.Site AT06 in the Atlantic Ocean and site AO10 in the Arctic shelf were of the highest fraction of factor 1, while other samples were less impacted by factor 1. The fractions of factor 2 showed evident fluctuations in the Atlantic Ocean and the Arctic Shelf, with the highest value occurring at sites IS01 and AO04.As we previously discussed, factor 1 and factor 2 were representative of wood/coal combustion and vehicle emissions.Surface seawater in the Atlantic Ocean and Arctic Shelf areas were more impacted by wood/coal combustion and vehicle emissions, while they influenced the Arctic basin more evenly.Factor 3, regarded as a vehicle-emission source of PAHs, showed high fractions in these three areas such as AO07, AO10, and AT02.Therefore, land-based wood/coal combustion mainly impacted the Atlantic Ocean and Arctic Shelf areas, while vehicle emission from the terrestrial influenced the whole high-latitude areas.
The Arctic Ocean is surrounded by important source regions of PAHs (30% of the global emission), Eurasia, and North America, and is characterized by its broad continental shelf area and large river runoff (Gustafsson and Andersson, 2012).
Atmospheric emissions of PAHs are expected to decrease by 46-71% and 48-64% in developed and developing countries before 2030 (Shen et al., 2013).However, as the most sensitive area to global warming, the Arctic is suggested to experience less magnitude of PAH decline according to modeling results (Balmer et al., 2019).
Besides the re-volatilization of PAHs in the warming Arctic, recently increasing wildfires also lead to more PAH emissions to the atmosphere (Ma et al., 2011;McCarty et al., 2020).Therefore, we suggested that local inputs, such as snow/ice melting, permafrost thawing, and subsequent river runoff, were crucial for PAHs in the Arctic Ocean.

Hydrological Properties in the water column
The vertical profiles of temperature and salinity of seawater at sites AT06, BB04, IS02, and SR18 were shown in Figure 4  Water, Arctic Ocean Deep Water, and Bering Sea Deep Water), and high temperature with high salinity (Bering Sea Surface Water and Bering Sea Intermediate Water).As a result of riverine inflow, seasonal sea ice melting, and dense saline water from the Atlantic Ocean, the Arctic Ocean is strongly stratified.Its major water masses could be simply divided into Arctic Ocean Surface Water (<100 m), Arctic Ocean Intermediate Water (100-500 m), and Arctic Ocean Deep Water (Jakobsson, 2002).
In the Norwegian Sea (represented by AT06 and BB04) and Greenland Sea (represented by IS02), the differences in hydrological properties reflected water exchange between the Arctic Ocean and the North Atlantic Ocean.In the upper layer, the temperature and salinity of Atlantic seawater were higher than those in the central Arctic Ocean, with the highest value at AT06 (10.2°C).Seawater temperature showed a deeper mixing depth on the surface of the Norwegian Sea (100 m), followed by a sharp decrease at the 500-800 m depth layer.Seawater salinity showed higher values at the surface of the Norwegian Sea, while it was more stable in the deeper layer of the Greenland Sea.Water masses in the Norwegian Sea and the Greenland Sea are primarily divided into six classes, namely, surface water, Atlantic Water, dense Water (Rudels et al., 2005).Here we simply divided the water column in the North Atlantic Ocean into three layers: Modified North Atlantic Water (<800 m), Arctic Intermediate Water (800-1500 m), and Arctic Deep Water (>1500 m).

PAH profiles in water masses
In the water columns of the Greenland Sea, the Norwegian Sea, and the central Arctic Ocean, PAH concentrations (∑9PAH) ranged from 0.2 to 9.9 ng L −1 , with a mean value of 3.3 ng L −1 (Figure 5).The maximum concentration was found at a depth of 300 m at site SR18, near the north pole, and the minimum was located at 1000 m depth at site AT06 in the Norwegian Sea.PAHs in the water columns (Figure 5) showed a "surface-enrichment and depth-depletion" pattern similar to a previous study (Dachs et al., 1997).Specifically, PAHs were firstly enriched in the subsurface layer (50-300 m) with a mean value of 3.6 ng L −1 , after reaching a subsequent maximum at approximately 200-300 m depth, they eventually decreased and kept steady in the intermediate and deep water.For the component contribution, noticeable differences occurred among different water columns, while their variances with depth in one water column are not as noticeable.Such component coherence in one water column indicated, on an ocean scale, that PAHs are probably less influenced by the lateral transport compared with vertical transport processes.The "surface-enrichment and depth-depletion" profile pattern reflected PAHs' biogeochemical processes occurring in the water column.Microbial degradation is considered the dominant depletion mechanism for dissolved PAHs in the open ocean, and this biodegradation process is weaker on the surface compared to the deep chlorophyll maximum (DCM) depth (González-Gaya et al., 2019).Taking the atmospheric input into consideration, it was reasonable to find the enrichment of PAHs at the surface.Besides, part of dissolved PAHs would partition to the particulate phase and subsequently sink to the deep-sea sediment, although the particulate-dissolved partitioning percentages (% on particles) were relatively low in the open ocean (Lohmann et al., 2021).As for the continental shelf, the flux of heat, freshwater, and nutrients through surrounding land has a significant effect on the thermohaline properties, giving rise to enhanced biological productivity (Krembs et al., 2011;Underwood et al., 2019).The dissolved organic carbon could further accelerate the photodegradation of small PAHs such as Phe by enhancing the formation of reactive intermediates (Shang et al., 2015).While in the intermediate and deep waters, where the primary productivity dropped, most dissolved PAHs are in low concentrations without considerable variations.
To further analyze PAH patterns in different water masses, we performed principal component analysis on individual PAH components after autoscaling the data, and three principal components (PCs) were derived from the analysis.The Keiser-Meyer-Olkin value of sampling adequacy was 0.703.PC1 contributed 41.4% to the total variance, heavily weighted in Flu, Phe, Ant, Acp, and FluA, mainly composed of three-ring PAHs with lower molecular weights.PC2, mainly composed of four-ring PAHs with higher molecular weights, contributed 25.9% to the total variance and is mainly weighted in BaA, Acpy, Chr, and FluA.PC3, explaining 11.7% of the total variance, was dominated by four-ring Pyr.We further divided seawater samples into four classes as surface and water-column samples in the Arctic Ocean and the North Atlantic Ocean (Figure 6).Their scores on PC1 and PC2 showed spatial variances at different depths, where PAHs in water columns showed higher loadings of PC2 (p <0.05).As four-ring PAHs are less biodegradable than three-ring PAHs, we suggested that the less degradable pattern of PAH compositions occurred in the deep layers.
where we used the water volume fluxes based on long-term observation (Stöven et al., 2016).
The annual average values of water volume fluxes are 4.4 ± 3.2 Sv and −1.4 ± 0.8 Sv for the Norwegian Atlantic Current and East Greenland Current, respectively.Recirculating Atlantic Water and Arctic Atlantic Water, whose volume fluxes are about −3.5 ± 1.9 Sv, were not discussed due to being indistinguishable in this study.Positive fluxes describe the northward fluxes into the Arctic Ocean, and negative values describe the southward fluxes from the Arctic Ocean (Stöven et al., 2016).The depth-averaged concentration (  ) was calculated using the trapezoidal integral equation (2): where the bottom depths for the East Greenland Current and the Norwegian Atlantic Current were set as 300 m and 800 m, respectively.  is the PAH concentration at depth  according to the depth profiles of PAHs at AT06 and IS02.
As shown in Figure 7, PAH individuals' estimated transport mass flux ranged from 5.4 ± 3.9 to 58 ± 42 tons year −1 through the northward Norwegian Atlantic Current, with a sum of 110 ± 79 tons year −1 .The fluctuation of PAH individuals went from −1.6 ± 0.9 to −14 ± 8.0 tons year −1 through the southward East Greenland Current, with a sum of −45 ± 26 tons year −1 .For

Contribution of oceanic transport for PAHs
The Arctic Ocean Boundary Current could be regarded as a relatively rapid and concentrated passage for Atlantic water to reach the western Arctic Ocean, with a larger spatial scale and more powerful driving force, while the discrete path of Pacific inflow in the Chukchi Sea slowed down the transit transport of Pacific water (Rudels et al., 1994;Mauldin et al., 2010).
However, when comparing the oceanic transport with the atmospheric pathway, the latter has been considered a more efficient pathway for delivering pollutants to the Arctic, while the Atlantic oceanic transport was less important for PAHs in the seawater of Arctic fjords (Pouch et al., 2021).The oceanic advective time to the central basin is estimated to be years, while it takes only days for PAHs to arrive at the Arctic by air (Mauldin et al., 2010).Regarding the removal processes, including degradation and deposition/settling, they are usually slower in the seawater than in the atmosphere.The half-lives of PAHs in the oceans are usually in the order of tens or hundreds of days, while those in the atmosphere range from tens to thousands of hours (González-Gaya et al., 2019;Halsall et al., 2001;Liu et al., 2021a).Hence, the extent of PAH degradation or settling is expected to be higher during the oceanic transport to the Arctic, and we suggested that the ocean current was a less-dominant pathway for PAHs entering the Arctic Ocean.
The oceanic response to POPs change is slower but more complicated.Although the timescale of water mass transport is decades in the Atlantic and Arctic, which is longer than the half-lives of the dominant PAHs, oceanic fronts between two water masses could have an instantaneous effect on PAHs.Besides the direct delivery, the role of ocean currents should be further considered in their indirect impact on PAHs' air-sea interactions.Ocean currents could influence sea surface properties and the biogeochemical processes, subsequently modulating the air-sea exchange of PAHs (Lohmann and Belkin, 2014).To better understand the fate of POPs comprehensively, we suggest that further study should consider the oceanic modulations in both direct and indirect ways based on PAH data in multi-environments.

Conclusion
We investigated PAHs from surface seawater and full-depth water columns in the North Atlantic Ocean and the Arctic Ocean to better understand PAHs' transport processes and their contribution to high-latitude oceans.For surface seawaters collected from the Arctic and North Atlantic, the concentrations of 9 PAHs (∑9PAH) in the dissolved phase ranged from 0.3 to 10.2 ng L −1 (mean 4.3 ng L −1 ).Their spatial distribution showed an "Arctic Shelf > Atlantic Ocean > Arctic Basin" pattern, and inputs from the surrounding margins were suggested to be crucial for

Figure 1 .
Figure 1.Locations of sampling sites for seawater samples in the North Atlantic Ocean and the Arctic Ocean.The dark red dots represent sites for surface seawater samples, and the bright red squares represent sites for water-column samples.

Figure 2 .
Figure 2. Spatial distribution of dissolved PAHs in the surface water from the North Atlantic Ocean and the Arctic Ocean.

Figure 3 .
Figure 3. Results of the PMF model: (a) Source profiles of each PMF factor and (b) , which generally represented the hydrological properties of the Greenland Sea, Norwegian Sea, and the central Arctic Ocean.Sectional distributions of temperature, salinity, dissolved oxygen, and fluorescence in the North Atlantic Ocean are shown in Figure S5.
manuscript submitted to JGR Oceans Atlantic Water, intermediate water, deep water I including Canadian Basin Deep Water and the lightest part of the Nordic Seas Deep Water, and deep water II including Eurasian Basin Deep Water and the deeper part of the Nordic Seas Deep

Figure 6 .
Figure 6.Results of principal compound analysis for PAHs in the Arctic Ocean and the North Atlantic Ocean.

Flu
, and Phe, mass flux values were close in different directions, while the mass flux was larger in the southward current for Acp but smaller for Ant, FluA, and Pyr.The net transport mass flux of PAH individuals ranged from −4.4 ± 1.7 to 53 ± 39 tons year −1 to the Arctic Ocean, with a sum of 63 ± 53 tons year −1 .It is worth noting that the estimation results were of certain error due to the seasonal variations.The transport volume has seasonal variations, with a maximum in March and a minimum in July, and the temperature of upper-layer waters also has strong seasonal signals (Beszczynska-Möller et al., 2012).Since the seawater samples were only collected during summer, the estimated annual flux was of uncertainty but provided the information on the magnitude of PAH transport flux.Compared with the transport fluxes estimated for other POPs through the Fram Strait, PAH fluxes showed the same magnitude of HCHs, but a higher magnitude than PCBs and perand polyfluoroalkyl substances (PFAS)(Ma et al., 2018;Joerss et al., 2020).Except for Acp, whose net flux value was negative, there were net inflows to the Arctic Ocean for the other 5 PAHs, and Pyr contributed the highest load (53 ± 39 tons year −1 ), making up about 83% of the total mass flux.We found that with the increase of PAH molecular weight, the net mass flux to the Arctic Ocean increased.Previous studies reveal similar trends for PFAS, as a net outflow for shorter-chain PFASs and HCHs, while a net inflow for the PFASs with ≥ eight perfluorinated carbons or high-molecule weighted PCBs(Ma et al., 2018;Joerss et al., 2020).

Figure 7 .
Figure 7. PAH mass transport through the northward Norwegian Atlantic Current (Atlantic inflow, positive value) and the southward East Greenland Current (Arctic outflow, negative value).
PAHs in the Arctic Ocean.According to the results of PMF modeling, vehicle emissions and biomass combustion were the major sources of PAHs in the surface seawater.Besides, PAHs showed unique profiles indicating their different origins.Carried by East Greenland Current and the Norwegian Atlantic Current, PAH individuals' net transport mass flux ranged from −4.4 ± 1.7 to 53 ± 39 tons year −1 to the Arctic Ocean, indicating ocean current was a less-dominant pathway for PAHs entering the Arctic Ocean.Although we suggested a limited contribution of ocean currents on PAHs' delivery to the Arctic Ocean, their role in modulating PAHs' air-sea interactions and other biogeochemical processes needs further studies.