Home Journals Articles Updates About
CN
image
收藏切换
Characteristics, interannual variation, and influence mechanisms of a double-halocline in the western Arctic Ocean during 2002 to 2022
收藏切换
PDF
Yu Zhang2, Tao Li1, 2, *, Guoping Gao3, Hongxin Zhang4, Yubin Yao2, Xingyuan Zhu2, Yongjun Wang2
Acta Oceanologica Sinica | 2025, 44(3) : 57 - 70
Less
收藏切换
Acta Oceanologica Sinica | 2025, 44(3): 57-70
• Multiscale processes in polar oceans and their climate and ecological implications •
Characteristics, interannual variation, and influence mechanisms of a double-halocline in the western Arctic Ocean during 2002 to 2022
Full
Yu Zhang2, Tao Li1, 2, *, Guoping Gao3, Hongxin Zhang4, Yubin Yao2, Xingyuan Zhu2, Yongjun Wang2
Affiliations
  • 1Frontier Science Center for Deep Ocean Multispheres and Earth System (FDOMES) and Physical Oceanography Laboratory, Ocean University of China, Qingdao 266100, China
  • 2College of Oceanic and Atmospheric Sciences, Ocean University of China, Qingdao 266100, China
  • 3College of Oceanography and Ecological Science, Shanghai Ocean University, Shanghai 201306, China
  • 4Beihai Bureau Support Center, Ministry of Natural Resources, Qingdao 266100, China
Published: 2025-03-25 doi: 10.1007/s13131-025-2463-7
Outline
收藏切换

A halocline in the Arctic Ocean significantly slows the upward heat flux from deep warm water, thereby inhibiting the melting of surface sea ice. The western Arctic Ocean exhibits a double-halocline (DH) structure due to the complexity of the water mass. Using in situ measurements, we analyzed the vertical structural characteristics of DH and its interannual variation. The results indicated that the DH primarily occurs at the Northwind Ridge and the southern Canada Basin, extending westward to the Chukchi Abyssal Plain and northward to the northern boundary of the Canada Basin. From 2002 to 2022, there were changes in water masses that determined the structure of the DH. The significant increase in Pacific Water has resulted in 42% and 65% increases in freshwater and the heat content of the DH, respectively, along with a 14% reduction in stratification. Pacific Winter Water characterized by salinity of 33 has exhibited a gradually decreasing trend, suggesting that the lower halocline may be difficult to ventilate. The combined effects of Ekman pumping, mesoscale eddies, and positive buoyancy forcing (heat and freshwater input) from Pacific Water have altered the thickness and stratification of the DH. This study has enhanced our understanding of the evolution of vertical heat flux in the upper western Arctic Ocean.

western Arctic Ocean  /  double-halocline  /  ocean stratification  /  buoyancy forcing
Yu Zhang, Tao Li, Guoping Gao, Hongxin Zhang, Yubin Yao, Xingyuan Zhu, Yongjun Wang. Characteristics, interannual variation, and influence mechanisms of a double-halocline in the western Arctic Ocean during 2002 to 2022[J]. Acta Oceanologica Sinica, 2025 , 44 (3) : 57 -70 . DOI: 10.1007/s13131-025-2463-7
Full Text
Less
1　Introduction
Less
The Arctic Ocean, the smallest of the world’s oceans, is connected to the Pacific Ocean through the Bering Strait and to the Atlantic Ocean via the Fram Strait, the Barents Sea opening, and the waterways among the Arctic Archipelago. The inflow into the Arctic Ocean is composed of high-salinity Atlantic Water (AW) and low-salinity Pacific Water. The Arctic Ocean is currently undergoing unprecedented warming (Zhao et al., 2015), with the upper 2000 m warming at 2.3 times the global mean rate within this depth range averaged over the 21st century (Shu et al., 2022). This has resulted in a persistent decline in the extent of the summer sea ice, reaching a historic minimum of 3.34 × 106 km2 in 2012 (Liu, 2022). The extent of sea ice in September 2020 was only half the average from 1979 to 2000 (Ye et al., 2024). In addition to solar radiation, the warm water inflow from the Atlantic constitutes a critical source of heat that influences the melting of sea ice (Shimada et al., 2006). After entering the Arctic Ocean, AW moves in a cyclonic manner along the continental slope as a boundary current. The substantial amount of heat the water carries affects the thermal balance of the Arctic Ocean, thereby contributing to the rapid retreat of sea ice (Zhang et al., 1998; Polyakov et al., 2010). As AW flows deeper, its heat progressively dissipates, establishing a pronounced halocline structure between the cold, fresh surface water and the warm, salty AW. This halocline is important for providing a density barrier trapping heat from the AW at depths away from the ice throughout the Arctic (Toole et al., 2010). If the halocline weakens or disappears, a substantial volume of meltwater will enter the North Atlantic, hindering deep water formation and consequently weakening the strength of Atlantic Meridional Overturning Circulation (Lin et al., 2023a). The strength of the halocline may serve as a novel indicator of climate change in the Arctic Ocean (Polyakov et al., 2018). The vertical structure of the halocline and its role in the variation of sea ice have become central topics in Arctic studies.
The Arctic Ocean is divided into the western Arctic Ocean and the eastern Arctic Ocean along 0° and 180° meridians. The eastern Arctic Ocean is characterized by a single vertical permanent halocline, called as cold halocline, whereas the western Arctic Ocean exhibits distinct halocline structure: a summer halocline, an upper halocline (UH), and a lower halocline (LH) in order from top to bottom. The summer halocline, commonly known as the seasonal halocline as it occurs exclusively in summer, is positioned between cold, fresh surface water (comprising sea ice meltwater and river freshwater) and warm, salty Pacific Summer Water (PSW). The UH is located between PSW and Pacific Winter Water (PWW). Pacific Water entering the Arctic Ocean via the Bering Strait sinks beneath the surface water and is subsequently transported laterally into the Beaufort Gyre (Timmermans et al., 2014, 2017). The sinking depth of Pacific Water varies between summer and winter. In winter, PWW becomes denser due to the formation of ice and brine rejection from coastal polynyas, ultimately sinking beneath the PSW (Weingartner et al., 1998; Shimada et al., 2005). The LH is positioned between the PWW and AW. AW from the Fram Strait and the Barents Sea opening is characterized by higher salinity than PWW, and thus, it will sink upon reaching the Beaufort Sea (Steele et al., 2004). Therefore, permanent double-halocline (DH), including UH and LH, formed by the advection mechanism that persists year-round (Aagaard et al., 1981; Shi et al., 2005).
The variation in the DH in the western Arctic Ocean is linked to processes at different spatial scales. Atmospheric forcing, the Arctic Oscillation anomaly, and increased glacial meltwater all contribute to changes in the salinity of surface seawater, subsequently resulting in variation in the halocline. In the southern Canada Basin, the clockwise-rotating, wind-driven Beaufort Gyre effectively traps a substantial amount of freshwater (McPhee et al., 2009). Darby and Bischof (2004) found that the Arctic Oscillation influenced the pathway of the transpolar drift, thereby affecting the drift trajectory of thick ice within the Beaufort Gyre and consequently influencing the freshwater content of the halocline. Climate warming has accelerated the melting of continental glaciers in Alaska, resulting in the freshening of Pacific Water (Chan et al., 2011). Based on long-term observational mooring data deployed at the Bering Strait, PWW is freshening at a rate of 0.03 per year in salinity, a rate that could influence the ventilation of the halocline. In addition, convective processes (Shimada et al., 2006; Woodgate et al., 2010), along with diffusion and double-diffusion phenomena (Bourgain and Gascard, 2011), contribute to alteration in the vertical structure of the halocline.
Significant changes have occurred in the DH due to the rapid decrease in sea ice and the increase in the heat content in the upper ocean. Further investigation is required to understand the mechanisms influencing the changes in the depth and thickness of the DH. Since the stratification of the halocline and the changes in the vertical heat flux through the DH have not been evaluated, we addressed these issues in the present study using comprehensive in situ hydrographic information and reanalysis data from 2002 to 2022. Section 2 describes the data and methods. Section 3 shows the variability of the DH and its corresponding influencing factors. The main conclusions are summarized in Section 4.
2　Data and methods
Less
2.1　 In situ measurements
The study area was divided into four regions based on topographical and circulation characteristics: the Chukchi Abyssal Plain (CAP), the Northwind Ridge (NwR), the northern Canada Basin (NCB), and the southern Canada Basin (SCB) (Fig. 1). The NwR and SCB are located at the center of the Beaufort Gyre, whereas the CAP and NCB are located at the edges of the gyre. We collected various in situ measurements of the western Arctic Ocean from 2002 to 2022 (Table 1). The principal scientific projects included the Chinese National Arctic Research Expedition (CHINARE), the Beaufort Gyre Exploration Project (BGEP), Ice Tether Profilers (ITP), the Unified Database for Arctic and Subarctic Hydrography (UDASH), and the World Ocean Database (WOD). Figure 1 depicts the distribution of the stations involving the DH structure from 2002 to 2022.
2.2　 Reanalysis data
To quantify the interannual variability of the halocline, we used the GLORYS12 monthly global ocean reanalysis product provided by the Copernicus Marine Environment Monitoring Service. This product has a horizontal resolution of (1/12°) and comprises 50 vertical levels. Since the depth of the halocline is directly affected by Ekman pumping, we selected multi-source reanalysis data products to calculate the velocity of Ekman pumping and performed linear interpolation to obtain a monthly averaged dataset with a spatial resolution of 0.25° (latitude) × 0.75° (longitude). The sea ice concentration and surface wind speed at 10 m were obtained from ERA5 released by the European Centre for Medium-Range Weather Forecasts (ECMWF); these data have a spatial resolution of 0.5° × 0.5° and a temporal resolution of 1 month. Sea ice drift data were obtained from Polar Pathfinder Daily 25 km EASE-Grid Sea Ice Motion Vectors, Version 4 released by the National Snow and Ice Data Center; the data have a spatial resolution of 25 km × 25 km and a temporal resolution of 1 month. Dynamic ocean topography (DOT) was obtained by a monthly averaged dataset produced by Armitage et al. (2020) that has a monthly (annual) average uncertainty of 1.1 cm (0.9 cm). The DOT features a spatial resolution of 0.25° (latitude) × 0.75° (longitude) and covers a range from 70.0°N to 81.5°N. The temporal resolution is 1 month, and the temporal range extends from 2003 to 2020.
2.3　 Methods for determining DH boundaries
Although isohalines and isopycnals have commonly been used to delineate the boundaries of the halocline (Rudels et al., 1996; Korhonen et al., 2013; Polyakov et al., 2020), the freshening of Pacific Water (Woodgate and Peralta-Ferriz, 2021) introduces errors using these methods on a time scale of decades. According to Bertosio et al. (2022), the boundaries of the UH are determined based on the core location of the water masses. The in situ observations suggest that the UH can form when the temperature difference between the PSW and PWW exceeds 0.3℃. Therefore, the top boundary of the UH is determined by the core of the PSW (Fig. 2), specifically at the depth of temperature maximum within the salinity of 29–33 and deeper than 40 m. The PSW core is primarily composed of Alaska Coastal Water (ACW, 29–32.2 in salinity), with minor proportions of summer Bering Sea Water (BSW, 29.2–33 in salinity) (Steele et al., 2004). ACW is regarded as the PSW core when both water masses are present, whereas BSW becomes the PSW core only due to a lack of ACW. The bottom boundary of the UH, which also serves as the top of the LH, is defined by the PWW core, identifying the temperature minimum within a salinity of 32.7–33.3.
The base of the LH was determined using the density gradient ratio (R) proposed by Bourgain and Gascard (2011). This ratio is defined as a function of both the temperature and salinity gradients:
$ R=\alpha \Delta T/(\beta \Delta S) , $
where $ \alpha $ represents the thermal expansion coefficient; $ \beta $ denotes the saline contraction coefficient; $ \Delta T $ and $ \Delta S $ indicate changes in temperature and salinity in 1 m intervals. A smaller $ R $ value indicates a stronger salinity gradient but a weaker temperature gradient, a key characteristic of the structure of a halocline. The observational data indicated that $ R $ = 0.05 yielded the best fit for the LH base.
2.4　 Quantification of DH characteristics
The accumulation of freshwater content (FWC) and heat content (HC) within the halocline influences the thermodynamic and dynamic processes in the upper ocean, thereby affecting the large-scale thermohaline circulation. Freshwater content is defined as
$ {\mathrm{FWC}}=\int _{{z}_{1}}^{{z}_{2}}{(S}_{{\mathrm{ref}}}-S)/S{\mathrm{d}}z , $
where $ {z}_{1} $ represents the depth of the halocline bottom; $ {z}_{2} $ denotes the depth of the halocline top; $ {S}_{{\mathrm{ref}}} $ is the reference salinity of 34.8.
Heat content is defined as
$ {\mathrm{HC}}=\int _{{z}_{1}}^{{z}_{2}}\rho {C}_{{\mathrm{p}}}(\theta -{\theta }_{{\mathrm{f}}}){\mathrm{d}}z , $
where $ \rho $ represents the density of seawater; $ \theta $ and $ {\theta }_{{\mathrm{f}}} $ are the potential temperature and freezing point, respectively. $ {C}_{{\mathrm{p}}} $ is the specific heat capacity of seawater, a function of the salinity and potential temperature.
Given that the DH in the western Arctic Ocean comprises multiple water masses, the Brunt-Väisälä Frequency ($ {N}^{2} $), also known as the buoyancy frequency, was calculated to evaluate the stratification of the water column:
$ {N}^{2}=-g \partial \rho /\rho \partial z , $
$ {N}_{{\mathrm{mean}}}^{2}=\int _{{z}_{1}}^{{z}_{2}}{N}^{2}{\mathrm{d}}z/\Delta h , $
where $ g $ is the acceleration of gravity; $ \rho $ indicates the density of seawater; $ \Delta h $ is the difference between $ {z}_{1} $ and $ {z}_{2} $.
Vertical heat flux ($ {F}_{\mathrm{h}} $) caused by diapycnal mixing is defined as
$ {F}_{\mathrm{h}}=\rho {C}_{{\mathrm{p}}}k({\theta }_{1}-{\theta }_{2})/({z}_{1}-{z}_{2}) , $
where $ \rho $ represents the density of seawater; $ {\theta }_{1} $ is the potential temperature at the halocline bottom; $ {\theta }_{2} $ is the potential temperature at the halocline top. $ k $ is the diapycnal diffusivity, here we assume a constant $ k $ = 2 × 10−6 m2/s (Lique et al., 2014).
Ekman pumping can affect the boundaries of the DH via vertical movement of the water parcels; this is numerically equivalent to the stress curl acting on the ocean surface (Hibler III, 1979; Hibler III and Bryan, 1987). In open water, the ocean surface stress $ {\vec{\tau }}_{\mathrm{a}\mathrm{o}} $ is a function of the surface wind speed $ {\vec{U}}_{\mathrm{s}} $. According to Zhong et al. (2018), the ocean surface stress $ {\vec{\tau }}_{\mathrm{i}\mathrm{o}} $ in regions covered by sea ice is a function of the difference between ice motion $ {\vec{U}}_{\mathrm{i}\mathrm{c}\mathrm{e}} $ and geostrophic flow $ {\vec{U}}_{\mathrm{o}\mathrm{c}\mathrm{e}} $ below the Ekman layer. We neglected the vertical variation in Ekman pumping (Kenigson et al., 2021), and the resulting formulas were as follows:
$ {W}_{\mathrm{e}}=\nabla \mathrm{ }\times {\vec{\tau }}_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}/\mathrm{ }\left(f{\rho }_{\mathrm{w}}\right) , $
$ {\vec{\tau }}_{\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l}}={\alpha \vec{\tau }}_{\mathrm{i}\mathrm{o}}+(1-\alpha ){\vec{\tau }}_{\mathrm{a}\mathrm{o}} , $
$ {\vec{\tau }}_{\mathrm{a}\mathrm{o}}={\rho }_{\mathrm{a}}{C}_{\mathrm{d}}\left|{\vec{U}}_{\mathrm{s}}\right|{\vec{U}}_{\mathrm{s}} , $
$ {\vec{\tau }}_{\mathrm{i}\mathrm{o}}={\rho }_{\mathrm{w}}{C}_{\mathrm{i}\mathrm{w}}\left|{\vec{U}}_{\mathrm{i}\mathrm{c}\mathrm{e}}-{\vec{U}}_{\mathrm{o}\mathrm{c}\mathrm{e}}\right| \Big[\big({\vec{U}}_{\mathrm{i}\mathrm{c}\mathrm{e}}-{\vec{U}}_{\mathrm{o}\mathrm{c}\mathrm{e}}\big)\cos\theta +\vec{k}\times\, \big({\vec{U}}_{\mathrm{i}\mathrm{c}\mathrm{e}}-{\vec{U}}_{\mathrm{o}\mathrm{c}\mathrm{e}}\big)\sin\theta \Big] , $
where $ \alpha $ is the ice concentration; the air density $ {\rho }_{\mathrm{a}} $ = 1.25 kg/m3; the atmospheric drag coefficient $ {C}_{\mathrm{d}} $ = 1.25 × 10−3; the seawater density $ {\rho }_{\mathrm{w}} $ = 1 023 kg/m3; and the ice-ocean drag coefficient $ {C}_{\mathrm{i}\mathrm{w}} $ = 5.5 × 10−3. Since $ {\vec{U}}_{\mathrm{o}\mathrm{c}\mathrm{e}} $ is positioned beneath $ {\vec{U}}_{\mathrm{i}\mathrm{c}\mathrm{e}} $, a reversal of the angle (i.e., 25°) is necessary to define the relationship between the two (Zhong et al., 2018).
Ekman pumping gives rise to a baroclinically unstable tilt of isopycnals, generating mesoscale eddies that reduce their slope via along-isopycnal fluxes. Consequently, the total transport in an isopycnal layer is known as the residual mean defined by $ {\varPsi }^{{\mathrm{res}}} = \overline{\varPsi } + {\varPsi }^{\mathrm{*}}$. The residual circulation $ {\varPsi }^{{\mathrm{res}}} $ through a layer has a component eddy-induced circulation $ {\varPsi }^{*} $ in addition to the Eulerian mean $ \overline{\varPsi } $ because there can be correlations between the lateral flow and the thickness of the layer, leading to a significant bolus transport (Timmermans and Marshall, 2020). Meneghello et al. (2017) showed that if the lateral transport of Pacific Water is ignored, the residual circulation within the steady Beaufort Gyre is small enough to maintain an equilibrium between $ \overline{\varPsi } $ and $ {\varPsi }^{*} $. The Eulerian mean circulation can be expressed in terms of Ekman pumping as $ \overline{\varPsi } $=$ \overline{\tau }/{(\rho }_{0}f) $. The eddy-induced circulation can be represented by the horizontal buoyancy flux as $ {\varPsi }^{\mathrm{*}} $ = $ \overline{v^{\mathrm{'}}b^{\mathrm{'}}}/(\partial\overline{b}/\partial z) $. This leads to $ \overline{{v}^{\mathrm{{'}}}{b}^{\mathrm{{'}}}} $ =$ -K_{\mathrm{D}}(\partial\overline{b}/\partial r) $, resulting in $ {\varPsi }^{\mathrm{*}} $ = $ {K}_{{\mathrm{D}}}I. $ Here, $ I $ denotes the slope of the halocline, and $ {K}_{{\mathrm{D}}} $ represents the eddy diffusion when the residual circulation is zero. Ultimately, we could derive a formula
$ I = -\overline{\tau }/{(K}_{{\mathrm{D}}}{\rho }_{0}f) , $
which indicates that the slope of the halocline is jointly determined by surface stress and the eddy buoyancy diffusivity.
3　Results and discussion
Less
3.1　 Spatial distribution of the DH
Figure 1 shows that the spatial coverage of the DH over the western Arctic Ocean is primarily centered around the Northwind Ridge and the southern Canada Basin. The UH is composed of PSW and PWW, whereas the LH is composed of PWW and may also include a minor proportion of AW in certain areas. There is a strong correspondence between the DH and $ {N}^{2} $, with the UH demonstrating low $ {N}^{2} $ values and the LH exhibiting high $ {N}^{2} $ values (Fig. 2). This phenomenon can be attributed to the differences in temperature and salinity between the PSW and PWW. During winter, the PWW becomes denser due to the formation of ice and brine rejection in polynyas, resulting in the formation of the LH. The vertical density gradient of the PWW is more pronounced than that of the PSW, resulting in the water column being less susceptible to displacement.
Given that hydrographic stations in 2008 were distributed widely in the West Arctic Ocean (Fig. 1), we chose this year to examine the differences in the DH across the four regions. The results showed the DH occurred at a depth of 41–257 m in 2008 (Figs 3a–c), with the UH exhibiting a thickness of (97 ± 36) m and the LH being approximately (55 ± 21) m thick. From CAP and NCB to NwR and SCB, the DH deepened from 41–177 m to 98–257 m. These spatial distributions reflected the fundamental characteristics of the Beaufort Gyre. In 2008, the center of the Beaufort Gyre was situated in NwR and SCB, where Ekman transport caused a substantial accumulation of freshwater, leading to increased sea surface height and downwelling (Shimada et al., 2005). Consequently, the halocline was deeper in NwR and SCB than in CAP and NCB. Furthermore, the strength of downwelling was influenced by surface wind, the concentration of sea ice, and geostrophic flow. In 2008, the total stress input led to an Ekman pumping velocity twice the climatological average (Figs 3h and i), resulting in a deeper DH.
The thickness of the DH did not have a consistent spatial distribution with depth. In 2008, both UH and LH reached their maximum thickness at NwR, gradually decreasing toward the surrounding regions (Fig. 4). The discrepancy between the locations of maximum halocline thickness and maximum Ekman pumping velocity indicated that the central position of the Beaufort Gyre was not the main factor influencing halocline thickness. Furthermore, a BSW-type DH structure was present in NCB, leading to a thicker and deeper UH.
3.2　 Interannual variation of the DH
From 2002 to 2022, the regionally averaged depth of the UH showed a generally increasing trend, but there was significant phase variation (Figs 5a and b). Between 2002 and 2008, the UH and LH deepened rapidly by (36 ± 11) m, while both exhibited only minor fluctuations in the following 14 a until 2022. During the same period, the thickness of the UH across the four regions increased by (31 ± 14) m, exhibiting a trend similar to the UH depth (Figs 5c and d). The changes in the thickness of the LH were particularly notable, being characterized by a sharp decrease from 2002 to 2007, a rapid increase from 2007 to 2010, and a gradual decrease from 2010 to 2022. Between 2002 and 2010, the LH thickness exhibited a reverse pattern, whereas the UH thickness consistently demonstrated a trend of stable increase. This indicates that the UH and LH thickness are not in a compensatory relationship, thereby ruling out the possibility of the LH water crossing into the UH.
In the four regions, the freshwater content of the UH was nearly three times that of the LH (Fig. 5e). According to Eq. (2), two factors jointly influence the freshwater content within the DH: seawater salinity relative to 34.8 and the thickness of the halocline. However, the latter was the primary factor, as indicated by the Pearson correlation of 0.91. Lin et al. (2023a) pointed out that from 2003 to 2019, the total freshwater content in the seasonal and permanent halocline increased from 14.6 m to 20.0 m, with an increase of 37%. We found that the average freshwater content of the UH increased from (6.7 ± 2.4) m to (9.5 ± 3.8) m from 2002 to 2022 due to an increase in thickness. This 42% increase reflects the significant freshening of Pacific Water. Although the freshwater content of the LH experienced fluctuations between 2002 and 2010, the level gradually recovered to (2.0 ± 0.2) m in the subsequent period (Fig. 5f). The heat content of the UH increased from (2.1 ± 0.3) × 108 J/m2 in 2002 to approximately (4.0 ± 1.4) × 108 J/m2 in 2022 (Fig. 5g). This nearly 100% increase indicated significant warming of Pacific Water. Timmermans et al. (2018) attributed this warming effect to the melting of sea ice in the Chukchi Sea and the increase in solar radiation entering the surface ocean in summer. The heat content of the LH did not change notably during this period (Fig. 5h).
3.3　 Mechanisms of influencing depth and thickness of the DH
3.3.1　 Equilibrium between Ekman pumping and mesoscale eddy activity
We could estimate the slope of the halocline using surface stress and the eddy buoyancy diffusivity according to Eq. (11). However, $ {K}_{{\mathrm{D}}} $ cannot be measured directly. Zhong et al. (2019a) calculated the actual eddy diffusivity $ {K}_{\lambda } $ in the Beaufort Gyre from 80 m to 400 m using climatological data, and they chose a constant value of 500 m2/s as the typical eddy diffusivity, assuming that there is no additional lateral advection or injection of PWW. Thus, $ \mathrm{\Delta }h $ = $ \overline{\tau }\mathrm{\Delta }L/{(K}_{{\mathrm{D}}}{\rho }_{0}f) $ ≈ 27 m, where $ \overline{\tau } $ = 5 × 10−3 N/m2, $ f $ = 1.5 × 10−4 S−1, $ {K}_{{\mathrm{D}}} $ = 500 m2/s, and $ \mathrm{\Delta }L $ = 400 km (the radius of the Beaufort Gyre). The results were generally consistent with observational data. The difference in the depth of the LH between the central region (NwR and SCB) and the edge region (CAP and NCB) of Beaufort Gyre is (32 ± 16) m (Fig. 6). Therefore, on a climatological timescale, the balance between Ekman pumping and mesoscale eddy activity governs the intensity of depression of the halocline across different regions.
Analysis of the interannual variation of the halocline identified a strong correlation between the Ekman pumping velocity and the upper boundary depth of the UH (Fig. 6). In NwR, NCB, and SCB, the Pearson correlations exceeded 0.60, with SCB reaching as high as 0.81. These strong correlations indicate that Ekman pumping predominates for the upper boundary depth of the UH. Furthermore, the LH is sensitive to upper ocean pressure due to its relative thinness. From 2002 to 2007, the sea ice conditions in the western Arctic transitioned from multiyear ice to first-year ice, accompanied by a strengthening of surface winds. This change increased the total stress curl input into the ocean and enhanced Ekman pumping (Zhong et al., 2019b), leading to a decrease in the thickness of the LH (Figs 5d and 6). After 2007, both sea ice and wind speed gradually returned to normal climatological levels, and the thickness of the LH also recovered to (50 ± 7) m.
From 2002 to 2022, there was a stable deepening trend of the LH in the four regions that cannot be explained by Ekman pumping (Fig. 6). The average deepening rate of the upper boundary was (3.1 ± 0.3) m/a; the lower boundary was (2.7 ± 0.5) m/a. We think this phenomenon may be related to the positive buoyancy forcing of Pacific Water.
3.3.2　 Buoyancy forcing from Pacific Water
The newly ventilated Pacific Water that determines the structure of DH becomes warmed by solar radiation and freshened by melting sea ice. We selected a long-span station (78.0°N, 140.0°W) for the period 2004–2018 to illustrate the impacts of positive buoyancy forcing (Fig. 7). The PWW core deepened from (132 ± 9) m to (173 ± 6) m. In the long term, the newly ventilated Pacific Winter Water (NPWW) became fresher and warmer, resulting in a shallower neutral buoyancy depth. The old Pacific Winter Water (OPWW) demonstrating an oxygen minimum (Fig. 7d) sank due to the intrusion of the NPWW. Also, newly ventilated Pacific Summer Water (NPSW) occupied the position originally held by old Pacific Summer Water (OPSW), leading to the sinking of OPSW.
In fact, not only is the Pacific Water at this station became fresher and warmer, but the entire Western Arctic Ocean was undergoing this change. Based on the trends of the DH thickness, we selected 2003–2007 and 2016–2020 for comparative analysis. The in situ data and GLORYS12 were utilized to calculate the thickness of the UH and LH, with results shown in Table 2. In the four areas, the halocline thickness error between the two data sets is less than 11%. Therefore, we thought GLORYS12 could reflect changes in Pacific Water volume, providing a basis for a more detailed assessment.
For each grid in GLORYS12, the volumes of water masses within 40–400 m were calculated with a resolution of ΔSr = 0.1, and the volumes corresponding to the same salinity were aggregated. We can see an increase in the volume of low-salinity water, which is the manifestation of freshening in the PSW (Fig. 8). The volume series of water masses for the two periods was then differenced. The results showed that the core salinities of OPSW and OPWW we calculated from in situ data (red and blue dots in Fig. 9) corresponded precisely to the minimum volume differences in CAP, NwR, and SCB and that the core salinity of NPSW corresponded to the maximum volume difference. This indicated that GLORYS12 effectively captured the freshened Pacific Water in these three regions. However, for NCB, there was a lack of correspondence between the core salinity and the peak volume difference. Additionally, the volume of the water masses corresponding to 33–34 in salinity exhibited serrated fluctuations, indicating that the simulation results of GLORYS12 in NCB were unreliable. Based on the volume difference, we visually identified the core salinity of NPWW for the period 2016–2020 (the green dots in Fig. 9). Finally, the core characteristics of the DH water masses were obtained by integrating both in situ data and GLORYS12 and are presented in Table 3.
On average, from the periods 2003–2007 to 2016–2020, the salinity of PSW decreased from 31.6 to 30.9, while the salinity of PWW decreased to 32.4 from 33.0. The intensity of freshening was similar in the Pacific Water, with no significant regional differences (Table 3). Warming of the PSW was more pronounced than the PWW, with the former increasing averagely by 0.4℃ on average. There were minimal changes in the PSW and PWW in the earlier period compared with OPSW and OPWW in the later period, indicating that the Pacific Water before freshening had not completely dissipated and remained at a detectable level. Analysis of the volume differences in water masses in the CAP, NwR, and SCB (Fig. 9) revealed that the increases in NPSW and NPWW significantly exceeded the decreases in OPSW and OPWW. This lag effect led to the coexistence of different Pacific Water masses. Consequently, the average UH thickness increased by (34 ± 12) m in 2022 compared with 2002. Additionally, OPWW exhibited varying degrees of loss across the four regions, resulting in a thinning of LH in the later period. In other words, the changes in water masses caused by the positive buoyancy forcing of Pacific Water significantly altered the thickness of the DH.
Furthermore, the LH depth has responded to the interannual pattern of the positive buoyancy forcing. There was a 1-a lag relationship of LH depth between NwR and SCB (Fig. 6), with a correlation of 0.85. Lin et al. (2023b) noted that a branch of the Chukchi Shelf Current flowing into the NwR may eventually shift eastward into SCB, thus becoming part of the Beaufort Gyre. The distance between the centers of NwR and SCB is approximately 300 km. We chose a typical geostrophic current speed of 1.5 cm/s that allowed us to estimate the time required for the water mass to travel between the two locations as 310 d. Therefore, the lateral transport of the newly ventilated Pacific Water affected the spatial variation of LH depth.
In summary, the DH structure underwent continuous changes from 2002 to 2022; these changes were influenced by the synergistic effects of Ekman pumping, mesoscale eddies, and positive buoyancy forcing of Pacific Water. The first two factors collaboratively regulated the strength of downwelling within the Beaufort Gyre, resulting in spatial variation in halocline depth across the different regions. The positive buoyancy forcing promoted the coexistence of multiple Pacific Water masses that further altered the thickness and depth of the DH.
3.4　 Influence of changes in the DH on the upper ocean
The stratification of the DH decreased by 14% from 2002 to 2022, with different trends in different periods. Between 2002 and 2012, the notable freshening of Pacific Water led to a significant thickening of the central UH part with low N2 values (Fig. 7e), consequently causing a decrease in the mean buoyancy frequency (Fig. 10e). The freshening process decelerated from 2012 to 2022, leading to stabilization of the UH thickness and a continued decline in mean buoyancy frequency. Nguyen et al. (2012) noted that the ventilation time of the halocline water in the western Arctic Ocean was 10–21 a. However, our findings indicate that PWW before freshening remained detectable in the DH, with no significant reduction in water volume. This suggests that the LH is difficult to ventilating.
Warming of the Pacific Water has increased the heat content of the DH, from (3.3 ± 0.2) × 108 J/m2 in 2002 to (5.4 ± 1.4) × 108 J/m2 in 2022, a 65% increase. The additional heat will melt 0.9 m of sea ice, taking the latent heat of melting to be 2.67 × 105 J/kg and the density of sea ice to be 900 kg/m3. The average temperature of the OPWW core has increased by 0.15℃ (Table 3), potentially as a result of the mixing of the OPWW with adjacent water masses. In this context, the temperature difference of the LH continued to increase, and the vertical heat flux through the DH exhibited an upward trend, rising from (0.17 ± 0.02) W/m2 in 2002 to (0.21 ± 0.01) W/m2 in 2022.
4　Conclusions
Less
This study has presented a comprehensive analysis of the spatial variation and interannual differences in the DH in the western Arctic Ocean, using in situ measurements and the reanalysis data. We investigated the mechanisms contributing to the changes in depth and thickness of the DH, ultimately offering a quantitative assessment of halocline stratification. The main conclusions are as follows.
(1) The upper western Arctic Ocean is characterized by a DH that contrasts sharply with the single halocline in the eastern Arctic Ocean. The DH is particularly prominent in the Northwind Ridge and the southern Canada Basin. In comparison to the upper halocline, the lower halocline exhibits a pronounced vertical density gradient and stronger stratification.
(2) Between 2002 and 2022, the DH increased in thickness by (31 ± 14) m, with the freshwater and heat contents increasing by 42% and 65%, respectively. Overall, the stratification of the DH has gradually decreased, resulting in an increase in vertical heat flux through the DH that rose from (0.17 ± 0.02) W/m2 in 2002 to (0.21 ± 0.01) W/m2 in 2022.
(3) The depth and thickness of the halocline are influenced by two primary mechanisms: the balance between Ekman pumping and mesoscale eddies, and the positive buoyancy forcing from the Pacific Water. The former mechanism is closely associated with the interannual variability of the halocline depth. Ekman pumping regulates the upper boundary depth of the UH and the depression of the LH across the four regions. The positive buoyancy forcing from the Pacific Water alters the thickness and stratification of the DH. The buoyancy gain leads to a shoaling of New Pacific Water, causing the sinking of the old Pacific Water. Meanwhile, the volume of OPWW has remained relatively stable, and its core thermohaline characteristics have not undergone significant changes, suggesting that the ventilation of the halocline has been significantly inhibited.
The DH in the western Arctic Ocean thickened, warmed, and became less stratified between 2002 and 2022, indicating that the heat content is continuously increasing and is likely to reach the seasonal halocline and mixed layer. The increased heat content will likely accelerate the retreat of sea ice.
Funding
Less
  • The National Key Research and Development Program of China under contract No. 2023YFC2809101; the National Natural Science Foundation of China under contract No. 42276239.
References
Less
Aagaard K, Coachman L K, Carmack E. 1981. On the halocline of the Arctic Ocean. Deep Sea Research Part A. Oceanographic Research Papers, 28(6): 529–545
Armitage T W K, Manucharyan G E, Petty A A, et al. 2020. Enhanced eddy activity in the Beaufort Gyre in response to sea ice loss. Nature Communications, 11(1): 761, doi: 10.1038/s41467-020-14449-z
Bertosio C, Provost C, Athanase M, et al. 2022. Changes in Arctic halocline waters along the East Siberian slope and in the Makarov Basin from 2007 to 2020. Journal of Geophysical Research: Oceans, 127(9): e2021JC018082, doi: 10.1029/2021JC018082
Bourgain P, Gascard J C. 2011. The Arctic Ocean halocline and its interannual variability from 1997 to 2008. Deep-Sea Research Part Ⅰ: Oceanographic Research Papers, 58(7): 745–756, doi: 10.1016/j.dsr.2011.05.001
Chan P, Halfar J, Williams B, et al. 2011. Freshening of the Alaska Coastal Current recorded by coralline algal Ba/Ca ratios. Journal of Geophysical Research: Biogeosciences, 116(G1): G01032
Darby D A, Bischof J F. 2004. A Holocene record of changing Arctic Ocean ice drift analogous to the effects of the Arctic Oscillation. Paleoceanography, 19(1): PA1027
Hibler III W D. 1979. A dynamic thermodynamic sea ice model. Journal of Physical Oceanography, 9(4): 815–846, doi: 10.1175/1520-0485(1979)009<0815:ADTSIM>2.0.CO;2
Hibler III W D, Bryan K. 1987. A diagnostic ice–ocean model. Journal of Physical Oceanography, 17(7): 987–1015, doi: 10.1175/1520-0485(1987)017<0987:ADIM>2.0.CO;2
Kenigson J S, Gelderloos R, Manucharyan G E. 2021. Vertical structure of the Beaufort Gyre halocline and the crucial role of the depth-dependent eddy diffusivity. Journal of Physical Oceanography, 51(3): 845–860, doi: 10.1175/JPO-D-20-0077.1
Korhonen M, Rudels B, Marnela M, et al. 2013. Time and space variability of freshwater content, heat content and seasonal ice melt in the Arctic Ocean from 1991 to 2011. Ocean Science, 9(6): 1015–1055, doi: 10.5194/os-9-1015-2013
Lin Peigen, Pickart R S, Heorton H, et al. 2023a. Recent state transition of the Arctic Ocean’s Beaufort Gyre. Nature Geoscience, 16(6): 485–491, doi: 10.1038/s41561-023-01184-5
Lin Peigen, Pickart R S, Weingartner T J, et al. 2023b. Formation and circulation of newly ventilated winter water in the western Beaufort Sea. Progress in Oceanography, 216: 103068, doi: 10.1016/j.pocean.2023.103068
Lique C, Guthrie J D, Steele M, et al. 2014. Diffusive vertical heat flux in the Canada Basin of the Arctic Ocean inferred from moored instruments. Journal of Geophysical Research: Oceans, 119(1): 496–508, doi: 10.1002/2013JC009346
Liu Zhongfang. 2022. Arctic sea ice decline and its driving mechanisms. Chinese Journal of Nature (in Chinese), 44(1): 39–46
McPhee M G, Proshutinsky A, Morison J H, et al. 2009. Rapid change in freshwater content of the Arctic Ocean. Geophysical Research Letters, 36(10): L10602
Meneghello G, Marshall J, Cole S T, et al. 2017. Observational inferences of lateral eddy diffusivity in the halocline of the Beaufort Gyre. Geophysical Research Letters, 44(24): 12331–312338
Nguyen A T, Kwok R, Menemenlis D. 2012. Source and pathway of the western Arctic upper halocline in a data-constrained coupled ocean and sea ice model. Journal of Physical Oceanography, 42(5): 802–823, doi: 10.1175/JPO-D-11-040.1
Polyakov I V, Pnyushkov A V, Carmack E C. 2018. Stability of the Arctic halocline: A new indicator of Arctic climate change. Environmental Research Letters, 13(12): 125008, doi: 10.1088/1748-9326/aaec1e
Polyakov I V, Rippeth T P, Fer I, et al. 2020. Weakening of cold halocline layer exposes sea ice to oceanic heat in the eastern Arctic Ocean. Journal of Climate, 33(18): 8107–8123, doi: 10.1175/JCLI-D-19-0976.1
Polyakov I V, Timokhov L A, Alexeev V A, et al. 2010. Arctic ocean warming contributes to reduced polar ice cap. Journal of Physical Oceanography, 40(12): 2743–2756, doi: 10.1175/2010JPO4339.1
Rudels B, Anderson L G, Jones E P. 1996. Formation and evolution of the surface mixed layer and halocline of the Arctic Ocean. Journal of Geophysical Research: Oceans, 101(C4): 8807–8821, doi: 10.1029/96JC00143
Shi Jiuxin, Zhao Jinping, Li Shujiang, et al. 2005. A double-Halocline structure in the Canada basin of the Arctic Ocean. Acta Oceanologica Sinica, 24(6): 25–35
Shimada K, Itoh M, Nishino S, et al. 2005. Halocline structure in the Canada Basin of the Arctic Ocean. Geophysical Research Letters, 32(3): L03605
Shimada K, Kamoshida T, Itoh M, et al. 2006. Pacific Ocean inflow: influence on catastrophic reduction of sea ice cover in the Arctic Ocean. Geophysical Research Letters, 33(8): L08605
Shu Qi, Wang Qiang, Årthun M, et al. 2022. Arctic Ocean Amplification in a warming climate in CMIP6 Models. Science Advances, 8(30): eabn9755, doi: 10.1126/sciadv.abn9755
Steele M, Morison J, Ermold W, et al. 2004. Circulation of summer Pacific halocline water in the Arctic Ocean. Journal of Geophysical Research: Oceans, 109(C2): C02027
Timmermans M L, Marshall J. 2020. Understanding Arctic Ocean circulation: A review of ocean dynamics in a changing climate. Journal of Geophysical Research: Oceans, 125(4): e2018JC014378, doi: 10.1029/2018JC014378
Timmermans M L, Marshall J, Proshutinsky A, et al. 2017. Seasonally derived components of the Canada Basin halocline. Geophysical Research Letters, 44(10): 5008–5015, doi: 10.1002/2017GL073042
Timmermans M L, Proshutinsky A, Golubeva E, et al. 2014. Mechanisms of Pacific summer water variability in the Arctic’s Central Canada Basin. Journal of Geophysical Research: Oceans, 119(11): 7523–7548, doi: 10.1002/2014JC010273
Timmermans M L, Toole J, Krishfield R. 2018. Warming of the interior Arctic Ocean linked to sea ice losses at the basin margins. Science Advances, 4(8): eaat6773, doi: 10.1126/sciadv.aat6773
Toole J M, Timmermans M L, Perovich D K, et al. 2010. Influences of the ocean surface mixed layer and thermohaline stratification on Arctic Sea Ice in the Central Canada Basin. Journal of Geophysical Research: Oceans, 115(C10): C10018
Weingartner T J, Cavalieri D J, Aagaard K, et al. 1998. Circulation, dense water formation, and outflow on the Northeast Chukchi Shelf. Journal of Geophysical Research: Oceans, 103(C4): 7647–7661, doi: 10.1029/98JC00374
Woodgate R A, Peralta-Ferriz C. 2021. Warming and freshening of the Pacific Inflow to the Arctic from 1990–2019 implying dramatic shoaling in Pacific winter water ventilation of the Arctic Water Column. Geophysical Research Letters, 48(9): e2021GL092528, doi: 10.1029/2021GL092528
Woodgate R A, Weingartner T, Lindsay R. 2010. The 2007 Bering Strait oceanic heat flux and anomalous Arctic sea-ice retreat. Geophysical Research Letters, 37(1): L01602
Ye Kunhui, Woollings T, Sparrow S N, et al. 2024. Response of winter climate and extreme weather to projected Arctic sea-ice loss in very large-ensemble climate model simulations. npj Climate and Atmospheric Science, 7(1): 20, doi: 10.1038/s41612-023-00562-5
Zhang Jinlun, Rothrock D A, Steele M. 1998. Warming of the Arctic Ocean by a strengthened Atlantic inflow: Model results. Geophysical Research Letters, 25(10): 1745–1748, doi: 10.1029/98GL01299
Zhao Jinping, Shi Jiuxin, Wang Zhaomin, et al. 2015. Arctic amplification produced by sea ice retreat and its global climate effects. Advances in Earth Science (in Chinese), 30(9): 985–995
Zhong Wenli, Steele M, Zhang Jinlun, et al. 2018. Greater role of geostrophic currents in Ekman dynamics in the western Arctic Ocean as a mechanism for Beaufort Gyre stabilization. Journal of Geophysical Research: Oceans, 123(1): 149–165, doi: 10.1002/2017JC013282
Zhong Wenli, Steele M, Zhang Jinlun, et al. 2019a. Circulation of Pacific Winter Water in the western Arctic Ocean. Journal of Geophysical Research: Oceans, 124(2): 863–881, doi: 10.1029/2018JC014604
Zhong Wenli, Zhang Jinlun, Steele M, et al. 2019b. Episodic extrema of surface stress energy input to the western Arctic Ocean contributed to step changes of freshwater content in the Beaufort Gyre. Geophysical Research Letters, 46(21): 12173–12182, doi: 10.1029/2019GL084652
Appendix
Less
Year 2025 volume 44 Issue 3
PDF
14
8
Cite this Article
BibTeX
Article Info
doi: 10.1007/s13131-025-2463-7
  • Receive Date：2024-12-08
  • Online Date：2026-11-05
  • Published：2025-03-25
Article Data
Affiliations
History
  • Received：2024-12-08
  • Accepted：2025-02-16
Funding
The National Key Research and Development Program of China under contract No. 2023YFC2809101; the National Natural Science Foundation of China under contract No. 42276239.
Affiliations
    1Frontier Science Center for Deep Ocean Multispheres and Earth System (FDOMES) and Physical Oceanography Laboratory, Ocean University of China, Qingdao 266100, China
    2College of Oceanic and Atmospheric Sciences, Ocean University of China, Qingdao 266100, China
    3College of Oceanography and Ecological Science, Shanghai Ocean University, Shanghai 201306, China
    4Beihai Bureau Support Center, Ministry of Natural Resources, Qingdao 266100, China

Corresponding:

References
Share
https://castjournals.cast.org.cn/joweb/aos/EN/10.1007/s13131-025-2463-7
Share to
QR

Scan QR to access full text

Cite this article
BibTeX
Citations
表12种不同金属材料的力学参数
科
Family		 属数
Number of
genus		 种数
Number of
species		 占总种数比例
Percentage of
total species (%)		 属
Genus		 种数
Number of
species		 占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏
  • BibTeX
  • EndNote
  • RefWorks
  • TxT
Links
Articles
Latest Articles
Most Read
Collections
Updates
Events
News
Multimedia
About
About Us
Contact
No. 86 Xueyuan South Road, Haidian District, Beijing
100081
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn

Copyright © 2025 China Association for Science and Technology. All rights reserved. For all open access content, the relevant licensing terms apply.
Sponsored by the Office of the Leading Group for Cybersecurity and Informatization of CAST, and supported by Science and Technology Review Publishing House