Water Circulation, Temperature, Salinity, and pCO 2 Distribution in the Surface Layer of the East Kamchatka Current

: The ship-borne observations of the temperature, salinity, pCO 2 (1995–2020) and satellite geostrophic velocity ﬁelds, SST, and chlorophyll concentration are used to identify the factors that determine the spatio-temporal variability of seawater parameters on the western boundary of the subarctic North Paciﬁc. In winter, the surface layer of the East Kamchatka Current (EKC) was characterized by two types of water: the waters with a negative temperature ( − 1.0– − 0.5 ◦ C) and salinity of 32.4–32.9 and waters with a positive temperature (0.4–1.7 ◦ C) and salinity of 33.0–33.1. The source of water with negative (positive) temperature and decreased (increased) salinity for the EKC zone is the Bering Sea shelf (Aleutian Basin). The surface waters in the eastern Kamchatka area in winter were close to gas equilibrium with the atmosphere or supersaturated with carbon dioxide (pCO 2 = 380–460 µ atm). In summer, extremely low pCO 2 values (140–220 µ atm) in the surface layer of the eastern Kamchatka and the northern Kuril Islands regions have been associated with the decreased salinity (32.1–32.6) of the waters. The distributions of the temperature, salinity, and pCO 2 in the surface layer of the central Kuril Islands are determined by the location and intensity of the Kuril eddies and the EKC stream jets. The water mixing in the central Kuril Straits and the Kruzenshterna Bank area leads to increased salinity (33.2–33.4) and high values of pCO 2 (480–670 µ atm) in the surface layer of the EKC. The comparison of the pCO 2 data collected in winter demonstrates an increase in pCO 2 between 1998/2001 and 2018/2020 at about 50 µ atm in the surface waters with a salinity of 33.0–33.1, which is in agreement with an increase in pCO 2 in the atmosphere at 46 µ atm (from 368 to 414 µ atm) during this period.


Introduction
The water circulation of the subarctic North Pacific consists of a cyclonic gyre that encompasses the Gulf of Alaska and the northwestern Pacific and enters the Bering and Okhotsk Seas [1]. The Alaskan Stream and the East Kamchatka Current (EKC) are the northern and western boundary currents of the subarctic North Pacific (Figure 1). Portions of the Alaskan Stream flow through the Aleutian Passes and Near Strait and form a cyclonic circulation cell in the Bering Sea [2]. The EKC, flowing southwestward along the east coast of the Kamchatka Peninsula and the Kuril Islands, originates from a flow out of the Bering Sea waters through the Kamchatka Strait. The flow of the EKC in the western Bering Sea and the Kamchatka Strait is strong between November and April and relatively weak in June-September. The strong seasonality in the surface flow of the EKC can be explained by temporal changes in the wind stress over the northern and western Bering Sea continental slope [3,4]. Meanders and eddies are consistent features in the EKC region [5,6]. The EKC area is the main pathway of the Bering Sea origin water to the North Pacific. Enhanced EKC advection is accompanied by an increase in water temperature and decreasing ice area in the Okhotsk Sea in winter [7].
There is a lack of analyses of the EKC circulation in the western subarctic North Pacific. Previous studies have mainly addressed the East Kamchatka and Kuril anticy-clonic eddies [7]. The study conducted during the joint Russia-Japan expedition showed significant spatial variability in the distribution of chemical parameters in the central Kuril Islands region in May-June 2000 [8]. The impact of the water circulation on the temperature, salinity, and chemical parameter distributions in the surface layer of the EKC area has not been described and discussed before. There is a lack of analyses of the EKC circulation in the western subarctic North Pacific. Previous studies have mainly addressed the East Kamchatka and Kuril anticyclonic eddies [7]. The study conducted during the joint Russia-Japan expedition showed significant spatial variability in the distribution of chemical parameters in the central Kuril Islands region in May-June 2000 [8]. The impact of the water circulation on the temperature, salinity, and chemical parameter distributions in the surface layer of the EKC area has not been described and discussed before.
Seawater carbon dioxide partial pressure (pCO2) (partial pressure of CO2 in the gas phase in equilibrium with the seawater) is one of the carbonate parameters of seawater that determines the direction and magnitude of the CO2 flux between the atmosphere and the ocean. pCO2 is an indicator of the physical and biogeochemical processes in seawater. The seasonal maps of pCO2 [9][10][11] demonstrate that the western subarctic Pacific is a source (sink) for atmospheric CO2 in winter (summer). The EKC area is a dynamically active region, which is characterized by strong tidal currents and water mixing in the Kuril Straits, the water supply from the Bering and Okhotsk Sea, eddy formation and migration [12], and enhanced biological activity in spring-summer [8]. These factors affect the temperature, salinity, and pCO2 distribution in the study area. The supply of the Okhotsk Sea and Bering Sea shelf water decreases temperature and salinity in the EKC zone in winterspring. Due to tidal mixing in the Kuril Straits, the surface layer of the EKC is enriched in salinity and seawater pCO2. In winter (summer), the tidal mixing increases (decreases) the temperature in the surface layer due to the temperature maximum (minimum) in the intermediate (surface) layer. The phytoplankton bloom is accompanied by a decrease in seawater pCO2.
In this paper, we analyze the ship-borne and satellite-derived data collected in the zone of the EKC between 1995 and 2020. The processes that influence the spatial and temporal variability of temperature, salinity, and pCO2 in the study region are shown. The main emphasis is on the effect of water circulation on the distribution of salinity, Seawater carbon dioxide partial pressure (pCO 2 ) (partial pressure of CO 2 in the gas phase in equilibrium with the seawater) is one of the carbonate parameters of seawater that determines the direction and magnitude of the CO 2 flux between the atmosphere and the ocean. pCO 2 is an indicator of the physical and biogeochemical processes in seawater. The seasonal maps of pCO 2 [9][10][11] demonstrate that the western subarctic Pacific is a source (sink) for atmospheric CO 2 in winter (summer). The EKC area is a dynamically active region, which is characterized by strong tidal currents and water mixing in the Kuril Straits, the water supply from the Bering and Okhotsk Sea, eddy formation and migration [12], and enhanced biological activity in spring-summer [8]. These factors affect the temperature, salinity, and pCO 2 distribution in the study area. The supply of the Okhotsk Sea and Bering Sea shelf water decreases temperature and salinity in the EKC zone in winter-spring. Due to tidal mixing in the Kuril Straits, the surface layer of the EKC is enriched in salinity and seawater pCO 2 . In winter (summer), the tidal mixing increases (decreases) the temperature in the surface layer due to the temperature maximum (minimum) in the intermediate (surface) layer. The phytoplankton bloom is accompanied by a decrease in seawater pCO 2 .
In this paper, we analyze the ship-borne and satellite-derived data collected in the zone of the EKC between 1995 and 2020. The processes that influence the spatial and temporal variability of temperature, salinity, and pCO 2 in the study region are shown. The main emphasis is on the effect of water circulation on the distribution of salinity, temperature, and pCO 2 of surface waters. First, we discuss the pCO 2 distribution in the winter of 2003 and winter of 2013 during periods, respectively, with low temperature and salinity and increased temperature and salinity in the surface layer of the EKC zone. Then, we compare the data collected in 2003 and 2013 with the data obtained in the winter of 1998/2002, the winter of 2018, and the winter of 2020. To demonstrate the amplitude of the seasonal variability of pCO 2 in the study area, the data collected in the summer of 1995, the summer of 1998, and the summer of 1999 are presented.

Sea Surface Heights, Geostrophic Velocities, SST, and Chlorophyll Concentration Data
Our study is based on data on the sea surface heights (SSHs) and geostrophic current velocities with a spatial resolution of 0.25 • by 0.25 • (for the study area of~30 km in longitude and~20 km in latitude) and a temporal resolution of 1 day obtained from satellite measurements (Copernicus database, http://marine.copernicus.eu, accessed on 10 April 2022) for the period from 1995 to 2020. The spatial and temporal variability of the chlorophyll concentration was investigated using data from the Aqua satellite with a spatial resolution of 4 km (http://oceancolor.gsfc.nasa.gov, accessed on 20 April 2022). The error value of satellite data for the SSH values measured from 2002 until the present is 1-2 cm at distances more than 20-40 km from the shoreline [13]. Based on the accepted SSH error value, the error of the geostrophic current velocities for the off-shore region is 3-6 cm/s. The SST (Sea Surface Temperature) images (spatial resolution of 1 km) were provided by GHRSST (Group for High-Resolution Sea Surface Temperature) (PO.DAAC-GHRSST Level 4 MUR Global Foundation Sea Surface Temperature Analysis). The Argo float data (temperature, salinity) were provided by the National Oceanic and Atmospheric Administration (NOAA) (http://www.nodc.noaa.gov/argo, accessed on 15 May 2022) [14]. The data on wind directions and velocities, as well as sea-level atmospheric pressure, were obtained from the NOAA Earth System Research Laboratories (http://www.esrl.noaa.gov accessed on 15 April 2022, accessed on 15 April 2022).

Underway pCO 2 , Salinity, Temperature Observations and Air-Sea CO 2 Flux Estimation
In this paper, we use ship-borne observations of temperature, salinity in the surface layer, seawater pCO 2 , and pCO 2 in the atmosphere (pCO 2 atm ) in the period from 1995 to 2020 [15][16][17][18][19] (https://www.ncei.noaa.gov/access/ocean-carbon-data-system/oceans/ VOS_Program, accessed on 10 April 2022). The determination of seawater pCO 2 in the surface layer was carried out in accordance with the procedure described in the manual for ocean CO 2 measurements (SOP 5) [20]. The measurements were carried out underway ( Figure 2). Seawater from the surface horizon (5-8 m) was continuously supplied to the gas equilibrator. The measurement of CO 2 concentration in the gas phase was carried out using an NDIR CO 2 analyzer. The accuracy of the pCO 2 sw measurements was ±1-3 µatm. The temperature and salinity measurements were carried out using a thermosalinograph manufactured by Sea-Bird company with a precision of 0.001 • C and 0.01, respectively [15].
The flux of carbon dioxide between the ocean and the atmosphere (F) was computed using the following equation: where ∆pCO 2 is the difference in seawater and atmosphere pCO 2 , and K is the CO 2 exchange coefficient determined by the product of CO 2 solubility in seawater [21] and the coefficient of CO 2 transfer through the seawater-atmosphere interface (gas transfer velocity). The gas transfer velocity (Kw) was calculated using the wind speed data (NCEP/NCAR reanalysis data) and the quadratic relationship between the Kw and wind speed (U) [22].
The equation for calculating the Schmidt number (Sc) for CO 2 was obtained from [23].

Calculation of the Seawater pCO 2 Depth Profiles
To calculate the depth profiles of seawater pCO 2 in the study region, we used hydrocast data (temperature, salinity, nutrients, total alkalinity, and dissolved inorganic carbon concentrations) collected in the northwestern Pacific in winter-spring 2010, 2012, and 2013. The data were provided by the National Center for Environmental Information (https: //www.ncei.noaa.gov/access/ocean-carbon-data-system/oceans/GLODAPv2_2020, accessed on 15 May 2022) [24]. Nutrients (phosphate, nitrate, and silicate) were determined spectrophotometrically with a precision of 1-2%. The total alkalinity (TA) was determined by potentiometric titration in an open cell [20]. The dissolved inorganic carbon (DIC) was measured by a coulometer. DIC and TA were calibrated versus certified reference materials (CRMs) [20]. The repeatability of the DIC and TA measurements [20] was ±2-3 µmol kg −1 . The seawater pCO 2 values were computed from DIC-TA and nutrient data using the CO2SYS Program [25]. The carbonic acid dissociation constants of Mehrbach et al. [26], as refitted by Dickson and Millero [27], were used. The uncertainty in the computed pCO 2 was about ±10 µatm. The flux of carbon dioxide between the ocean and the atmosphere (F) was computed using the following equation: Where ΔpCO2 is the difference in seawater and atmosphere pCO2, and K is the CO2 exchange coefficient determined by the product of CO2 solubility in seawater [21] and the coefficient of CO2 transfer through the seawater-atmosphere interface (gas transfer velocity). The gas transfer velocity (Kw) was calculated using the wind speed data (NCEP/NCAR reanalysis data) and the quadratic relationship between the Kw and wind speed (U) [22].
The equation for calculating the Schmidt number (Sc) for CO2 was obtained from [23].

Calculation of the Seawater pCO2 Depth Profiles
To calculate the depth profiles of seawater pCO2 in the study region, we used hydrocast data (temperature, salinity, nutrients, total alkalinity, and dissolved inorganic carbon concentrations) collected in the northwestern Pacific in winter-spring 2010, 2012, and 2013. The data were provided by the National Center for Environmental Information (https://www.ncei.noaa.gov/access/ocean-carbon-data-system/oceans/GLODAPv2_2020, accessed on 15 May 2022) [24]. Nutrients (phosphate, nitrate, and silicate) were determined spectrophotometrically with a precision of 1-2%. The total alkalinity (TA) was determined by potentiometric titration in an open cell [20]. The dissolved inorganic carbon (DIC) was measured by a coulometer. DIC and TA were calibrated versus certified reference materials (CRMs) [20]. The repeatability of the DIC and TA measurements [20] was ± 2-3 μmol kg −1 . The seawater pCO2 values were computed from DIC-TA and nutrient data using the CO2SYS Program [25]. The carbonic acid dissociation constants of Mehrbach et al. [26], as refitted by Dickson

Water Circulation in the Study Area
The SSH, current velocity maps (    According to Argo float data (2012-2020), the vertical water structure in the southwestern Bering Sea (57-59 • N, 164-166 • E) and near the eastern Kamchatka in the winter and early spring was characterized by a uniform temperature (0.5-1.5 • C) and salinity (32.9-33.1) in the 0-300/400 m layer. On some vertical profiles, the Bering Sea shelf water with a negative temperature (−1.5-0.5 • C) and salinity of 32.2-32.7 was observed in the 0-100 m layer. The water mixing induced by winter convection and wind, covering the upper layer to a depth of 300/400 m, should have led to a significant supersaturation of CO 2 in the surface waters. However, in the winter of 2003 and the winter of 2013, the surface waters near the Kamchatka and northern Kuril Islands were only slightly oversaturated with CO 2 relative to atmospheric CO 2 . Probably, intense gas exchange between the ocean and atmosphere reduced the seawater pCO 2 . Another important factor that reduces the seawater pCO 2 in winter is a decrease in water temperature [28].
A significant reduction in pCO 2 from 400 µatm on 16-18 February 2013 to 250-300 µatm on 21 April 2013 in the zone of eastern Kamchatka ( Figure 4C) was related to spring phytoplankton bloom. The decrease in pCO 2 during phytoplankton growth was accompanied by an increase in the concentration of chlorophyll a ( Figure 4D). The distribution of chlorophyll a concentration (Figure 5a,b) and satellite images (Figures 3f and 5c) demonstrate that spring bloom in early April 2013 was observed in waters with negative SST originating from the Bering Sea and was tied to the boundaries of ice fields. Two weeks later (15)(16)(17)(18)(19)(20)(21) April 2013), a high concentration of chlorophyll was noted at the boundaries of mesoscale anticyclones and the cyclones near southeastern Kamchatka (Figure 5b). Based on the pCO 2 data, it is possible to estimate the consumption of DIC in seawater during photosynthesis.
For the western part of the subarctic North Pacific, the change in the carbonate parameters in the surface waters caused by the formation of carbonates is insignificant [29]. During the synthesis of organic matter, CO 2 and nitrate are consumed, which is accompanied by a decrease in DIC and an increase in TA [30]. The changes in DIC, TA, and pCO 2 during photosynthesis could be quantified using the Redfield ratio (C:N:P = 117:16:1) [31] and CO2SYS Program [25]. For the eastern Kamchatka area, a decrease in DIC concentration by 50 µmol L −1 and an increase in TA by 7 µmol L −1 during the phytoplankton bloom in early spring (t = 0 • C, S = 33.0, DIC = 2126 µmol L −1 , TA = 2240 µmol L −1 ) lead to a decrease in pCO 2 by 120 µatm (from 390 to 270 µatm). Therefore, to reduce pCO 2 from 390 to 270 µatm in the EKC zone in April 2013, the DIC concentration had to decrease by 50 µmol L −1 . The decline of the seawater pCO 2 below the pCO 2 atm value (400 µatm, April 2013) causes a flux of CO 2 from the atmosphere into seawater, which partially compensates for the reduced DIC and pCO 2 due to phytoplankton growth. Applying Equations (1) and (2), we find that the average CO 2 flux from the atmosphere to the ocean over the period under study (4-21 April 2013, wind speed of 5 m/s) was equal to 5 mmol C m −2 day −1 . The increase in DIC in the upper 10 m layer of the study area due to the influx of CO 2 from the atmosphere was approximately 10 µmol L −1 . Therefore, the consumption of DIC related to phytoplankton growth was about 60 µmol L −1 (720 µg C L −1 ). In the period from 4 April to 21 April 2013, the concentration of chlorophyll increased by 5-6 µg L −1 . The ratio between the decrease in the concentration of DIC and the increase in the concentration of chlorophyll a was equal to 120-144 (µg C/µg chlorophyll), without considering the loss of phytoplankton biomass and the decrease in the concentration of chlorophyll a due to phytoplankton mortality and zooplankton grazing.
In late January-early February 1999 and in March 1999 (Figure 6), the distributions of temperature and salinity in the EKC zone (49 • N, 156 • E-54 • N, 162 • N) were similar to those in January-February 2013 ( Figure 4A (Figure 7), the surface layer of the EKC was characterized by positive temperature (0.5-1.5 • C), a salinity of 33.0-33.1, and pCO 2 values equal to 390-460 µatm. During these periods, the source of surface water for the EKC area was the Bering Sea deep basin (Aleutian Basin), characterized by a positive water temperature (1.0-2.0 • C) and a salinity of 33.0-33.2 in winter [32]. In February 2018, the surface water temperature in the northern Kuril and eastern Kamchatka region (155-159 • E) was higher than in February 2020 (1.2 • C and 1.7 • C, respectively), but the salinity and pCO 2 were similar to those observed in February 2020 (33.0 and 450-460 µatm) (Figure 7). For the western part of the subarctic North Pacific, the change in the carbonate parameters in the surface waters caused by the formation of carbonates is insignificant [29]. During the synthesis of organic matter, CO2 and nitrate are consumed, which is accompanied by a decrease in DIC and an increase in TA [30]. The changes in DIC, TA, and pCO2 during photosynthesis could be quantified using the Redfield ratio (C:N:P = 117:16:1) [31] and CO2SYS Program [25]. For the eastern Kamchatka area, a decrease in DIC concentration by 50 µmol l −1 and an increase in TA by 7 µmol l −1 during the phytoplankton bloom in early spring (t = 0 °C, S = 33.0, DIC = 2126 µmol l −1 , TA = 2240 µmol l −1 ) lead to a decrease    The water inflow from the Bering Sea into the Pacific Ocean through the Kamchatka Strait is enhanced in winter and determined by the wind stress along the continental slope of the Bering Sea [3]. In the winter of 1999 and winter of 2013, the low atmospheric pressure area (Aleutian Low) was located over the subarctic Pacific and southern Bering Sea zone. The eastern winds (10-12 m/s, monthly averaged speeds) over the central Bering Sea contributed to the intensification of the southwestern current along the western boundary of the Bering Sea [4] and the inflow of Bering shelf water into the EKC zone.
In The water inflow from the Bering Sea into the Pacific Ocean through the Kamchatka Strait is enhanced in winter and determined by the wind stress along the continental slope of the Bering Sea [3]. In the winter of 1999 and winter of 2013, the low atmospheric pressure area (Aleutian Low) was located over the subarctic Pacific and southern Bering Sea zone. The eastern winds (10-12 m/s, monthly averaged speeds) over the central Bering Sea contributed to the intensification of the southwestern current along the western boundary of the Bering Sea [4] and the inflow of Bering shelf water into the EKC zone.
In winter 1998, winter 2000, winter 2003, winter 2018, and winter 2020, the Aleutian Low was located over the central subarctic Pacific. In the southern part of the Bering Sea, the northeastern winds with a speed of 4-8 m/s were dominated, and, as a result, the Aleutian Basin water with positive temperatures and salinity of 33.0-33.2 entered the EKC zone.
In summer, the EKC is weak, and the water dynamics at the western subarctic Pacific boundary are determined by mesoscale water circulation. In June 1995, July 1998, and July 1999, mesoscale (horizontal size 100-200 km) anticyclonic eddies were observed near the eastern Kamchatka, northern Kuril, and central Kuril Islands. In June 1995 and July 1998, the Kamchatka eddy was weak (the velocities at the margins were 20-40 cm/s). In July 1999, the Kamchatka eddy was intensified, its horizontal dimension was increased to 200 km, and the current velocities at its boundary were increased up to 70 cm/s. The distributions of salinity and pCO 2 across the Kamchatka eddy (July 1998, July 1999) and the eddy southeastern periphery (June 1995) ( Figure 8A,B; 158-160 • E) show that in summer the surface layer of the Kamchatka eddy was composed by low salinity (32.1-32.5) and low pCO 2 (140-220 µatm) waters. In July 1999, low-salinity (32.1-32.6) and low pCO 2 (160-200 µatm) Bering Shelf waters occupied the EKC zone from the eastern Kamchatka to the northern Kuril Islands.

Central Kuril Island Area
The distributions of temperature, salinity, and pCO 2  A principal feature of the western subarctic water structure is the sharp halocline/ pycnocline at the bottom of the upper layer at about 100 m depth ( Figure 9B). In winter, the halocline prevents the direct ventilation of the intermediate water layer, characterized by high concentrations of nutrients and DIC [29]. In the 100-200 m layer, pCO 2 increased from 500 to 1400 µatm ( Figure 9C). In the salinity range of 33.1-33.6, dependence (close to linear) of pCO 2 on salinity is observed ( Figure 9D). Vertical mixing will increase the salinity and a significant increase in pCO 2 in the surface layer. Due to tidal mixing in the Kuril Straits, the halocline is destroyed, and the surface layer is enriched in salinity and seawater pCO 2 . In winter, the water mixing increases the temperature in the surface layer due to the maximum temperature in the intermediate layer ( Figure 9A). In the surface layer at the western boundary of the subarctic North Pacific in the winter of 2003 and the winter of 2013, there is a tendency for pCO 2 to increase from 380 to 660 µatm with an increase in salinity from 32.34 to 33.37.  (1998)(1999)(2000)(2001)(2002)(2003)2013), red squares-summer (1995,1998,1999), green crosses-winter 2018 and winter 2020) and change in the seawater (diamonds-water with a salinity of 33.00-33.05, crosses-water a salinity of 33.05-33.10) and atmospheric pCO2 in the study area (D)).  (1998)(1999)(2000)(2001)(2002)(2003)2013), red squares-summer (1995,1998,1999), green crosses-winter 2018 and winter 2020) and change in the seawater (diamonds-water with a salinity of 33.00-33.05, crosses-water a salinity of 33.05-33.10) and atmospheric pCO 2 in the study area (D)). In June 1995 and July 1998, the Kuril (and Kamchatka) eddies were weak. During these periods, the surface layer in the central Kuril Islands zone was characterized by decreased salinity (32.8-33.0) and pCO 2 equal to 320-380 µatm. In July 1999, the Kuril eddy was intensified, accompanied by an increase in its size and the current velocity at its borders. During this period, the strong northern Kuril eddy initiated the coastal branch of the EKC that led to the increased salinity (33.10-33.27) and pCO 2 (465-490 µatm) in the Kruzenshterna Bank area ( Figure 8A,B: 154.4 • E).

pCO 2 versus Salinity
In winter, at the western boundary of the subarctic North Pacific, there was a tendency for pCO 2 to increase from 380 to 670 µatm with an increase in salinity from 32.3 to 33.4 in the surface layer ( Figure 8C). In a salinity range of 33.1-33.4, the dependence of pCO 2 on salinity agreed well with the relationship between pCO 2 and salinity obtained from vertical profile data (47 • N, 160 • E) ( Figure 9D). This supports our conclusion that the increased salinity and pCO 2 in the study area are due to water mixing in the central Kuril Islands region and the Kruzenshterna Bank zone.
The salinity-pCO 2 diagram ( Figure 8C) shows that during the transition from winter to summer, the largest decrease in pCO 2 (from 370 to 130-230 µatm) was observed for waters with low salinity (32.1-32.5). In winter, the EKC waters with low salinity in the zones of eastern Kamchatka and the northern Kuril Islands were in a state close to CO 2 equilibrium with the atmosphere (pCO 2 = 360-380 µatm). In summer, these areas of the western part of the subarctic North Pacific were significantly undersaturated (∆pCO 2 sw-atm = −(140-240) µatm) with carbon dioxide (a sink for atmospheric CO 2 ). A comparison of the data obtained in 1998-2001 with data for 2018/2020 shows an increase in pCO 2 at the western boundary of the subarctic North Pacific for the surface waters with a salinity of 33.0-33.1 ( Figure 8D). In 1998-2001, in the surface layer, pCO 2 was equal to 402 ± 10 µatm (STD, N = 182) and 409 ± 9 µatm (N = 198) for waters with a salinity of 33.00-33.05 and 33.05-33.10, respectively. In 2018/2020, pCO 2 in the surface layer was 447 ± 7 µatm (N = 92) and 461 ± 7 µatm (N = 198) for waters with a salinity of 33.00-33.05 and 33.05-33.10, respectively. An increase in pCO 2 in the surface layer by 45-50 µatm corresponded to an increase in pCO 2 in the atmosphere by 46 µatm (from 368 ± 3 µatm to 414 ± 3 µatm).

Discussion
An analysis of the seawater parameters showed significant spatial and temporal variability in temperature, salinity, and pCO 2 in the surface layer on the western boundary of the subarctic North Pacific. In winter, in the zone of EKC, waters with a negative temperature (−1.0-−0.5 • C) and low salinity (32.4-32.9) and waters with a positive temperature In winter (1998-2003, 2013, 2018, 2020), the pCO 2 in the surface waters of the northern EKC area was 385-460 µatm. The surface waters were supersaturated with carbon dioxide or were close to equilibrium with the atmosphere. The water mixing due to convection and wind, covering the upper 300-400 m water layer in the western Bering Sea and Kamchatka Strait, should lead to a significant supersaturation of CO 2 surface waters in the eastern Kamchatka region in winter. However, intense gas exchange between the seawater and atmosphere and a decrease in water temperature in the autumn-winter period reduced pCO 2 .
The carbon production estimated by the change in the phosphate and dissolved oxygen concentrations in the surface layer of the eastern Kamchatka during spring diatom bloom changes from 10 mol C m −2 in the nearshore zone to 3-4 mol C m −2 in the pelagic area [34]. Accordingly, the silicate content of the surface layer decreases by 1.5 mol Si m −2 in the nearshore zone and by 0.7 mol Si m −2 in the pelagic area. The enhanced bloom of the diatom phytoplankton near the Kamchatka is related to low salinity (32.0-32.6) water [34]. The data presented in our paper show that extremely low pCO 2 values (130-230 µatm) due to the consumption of DIC during the spring phytoplankton bloom were observed near the eastern Kamchatka and the northern Kuriles in waters with low salinity and negative temperatures (Bering shelf water). The concentration of dissolved iron in the shelf waters of the Bering Sea is an order of magnitude higher than that in the surface waters of the Bering deep basin [35]. Therefore, the supply of the Bering shelf water could enhance the spring bloom of diatom plankton in the zone of eastern Kamchatka. In the period from 4 April to 21 April 2013, the concentration of chlorophyll increased by 5-6 µg L −1 . The calculated ratio between the decrease in the concentration of DIC and the increase in the concentration of chlorophyll a was equal to 120-144 (µg C/µg chlorophyll), without considering the decrease in phytoplankton biomass and the concentration of chlorophyll a due to the phytoplankton mortality and grazing by zooplankton. For coastal diatom phytoplankton, the relationship between biomass and chlorophyll concentration is determined by water temperature, photosynthetically active radiation (PAR), and nutrient availability [36]. At a water temperature of 0 • C, a PAR of 25 mol/(m 2 day) (4-21 April 2013) and the absence of nutrient limitation, the ratio between phytoplankton biomass and chlorophyll concentration should be approximately 150 (µg C/µg chlorophyll) that is consistent with our estimate based on the DIC consumption during phytoplankton growth.
In June 1995, July 1998, and July 1999, near the eastern Kamchatka and the northern Kuril Islands, waters with low salinity (32.1-32.6) and pCO 2 (140-220 µatm) were observed. During these periods, the EKC zone was a sink for atmospheric CO 2 . Previously, it was shown that the surface waters of the Bering Sea shelf are characterized by extremely low pCO 2 values in summer. In August 2002, the surface layer of the northwestern Bering Sea shelf was characterized by low pCO 2 values (213 ± 55 µatm), significantly lower than the pCO 2 atm (365 µatm) [37]. The distribution of salinity and pCO 2 in the central Kuril Islands region was determined by the water dynamics. During periods when the EKC stream jet crossed the Kruzenshterna Bank, increased salinity (33.2-33.3) and high values of pCO 2 (460-600 µatm) were observed in the surface layer of the area (47.6 • N, 154 • E-48.4 • N, 155 • E). The significant increase in pCO 2 (up to 685 µatm) in the central Kuril Islands in winter was related to the coastal branch of the EKC and the anticyclonic eddies, which determined the water exchange between the coastal and pelagic parts of the subarctic Pacific. The main sources of surface waters with high pCO 2 values are the Nadezdy, Rickorda, and Diana Straits (Rickorda Strait area) due to intensive tidal mixing, increased salinity, and the high pCO 2 observed in the surface layer. The zone of the central Kuril Straits is a source of carbon dioxide for the atmosphere in winter. The study conducted in the central Kuril Islands region in May-June 2000 showed significant spatial variability in the distribution of pCO 2 [8]. The seawater pCO 2 varied from 180 µatm to 500 µatm. High salinity (33.36) and high pCO 2 values (502 µatm) were observed in the Rickorda Strait zone. The lowest pCO 2 values (180-200 µatm) and high concentration of chlorophyll a were confined to the coastal branch of the EKC and the northern and eastern peripheries of the Kuril eddy. In June 1995, July 1998, and July 1999, near the central Kuril Islands, pCO 2 varied from 250 µatm (surface waters were undersaturated relative to atmospheric CO 2 ) to 485 µatm (surface waters were supersaturated with CO 2 ).
Our results demonstrate that due to a lack of intensive surveys and significant spatiotemporal variability, the distribution of pCO 2 and sea-air pCO 2 difference [9][10][11] in the EKC zone could be characterized by large uncertainty. To create seasonal maps of pCO 2 and sea-air CO 2 flux at the western boundary of the subarctic North Pacific, more pCO 2 observations and joint analysis of the ship-borne and satellite data are needed.
A comparison of data collected in the winters of 1998/2001 and the data collected in 2018/2020 shows an increase in pCO 2 in the surface layer at the western boundary of the subarctic North Pacific ( Figure 8D). In winter 1998/2001, in the surface layer, pCO 2 was equal to 402 ± 10 µatm (STD, N = 182) and 409 ± 9 µatm (N = 198) for waters with a salinity of 33.00-33.05 and 33.05-33.10, respectively. In winter 2018/2020, in the surface layer, pCO 2 was 447 ± 7 µatm (N = 92) and 461 ± 7 µatm (N = 198) for waters with a salinity of 33.00-33.05 and 33.05-33.10, respectively. An increase in pCO 2 in the surface layer by 45-50 µatm corresponded to an increase in pCO 2 in the atmosphere by 46 µatm (from 368 ± 3 µatm to 414 ± 3 µatm). In the subtropical North Pacific (Hawaii Ocean Timeseries Station) between 2002 and 2019, both seawater pCO 2 and pCO 2 atm show significant increases at rates of 1.7 ± 0.1 µatm yr −1 and 2.2 ± 0.1 µatm yr −1 , respectively [38]. However, the seawater pCO 2 in the central subtropical North Pacific shows a lower increasing rate than that in the western subarctic Pacific (2.3 ± 0.7 µatm yr −1 ).
The physical and chemical parameters of seawater in the surface and intermediate layers at the western boundary of the Pacific subarctic are subject to a 20-year variability associated with the 18.6-year nodal cycle [39]. During periods of increased nodal tides, salinity and density in the surface layer increase. The maximum nodal tides were observed in 2006, and the minimum tides in 1997 and 2015. The comparison of the data collected in periods with the same phase of nodal tides (1998/2001 and 2018/2020) may exclude the influence of nodal tides on the distribution of the salinity and pCO 2 in the surface layer at the western subarctic Pacific boundary.

Conclusions
The ship-borne data (temperature, salinity, and seawater pCO 2 ) and satellite data (SSH, geostrophic velocity, chlorophyll concentrations, and SST) collected between 1995 and 2020 at the western boundary of the subarctic North Pacific (EKC area) are analyzed. A joint analysis of satellite and ship-borne observational data made it possible to establish the processes that determine the spatial and temporal variability of temperature, salinity, and pCO 2 in the study region. The importance of water circulation on the distribution of temperature, salinity, and pCO 2 in surface waters of the EKC has been shown.
In winter, the surface layer of the EKC zone was characterized by two types of waters: the waters with a negative temperature (−1.0-−0.5 • C) and salinity of 32.4-32.9 and the waters with a positive temperature (0.4-1.7 • C) and salinity of 33.0-33.1. The source of water with negative (positive) temperature and decreased (increased) salinity for the EKC zone is the Bering Sea's northern shelf (Aleutian Basin). Strong eastern/northeastern winds in the western central part of the Bering Sea in winter force the inflow of the Bering shelf water to the EKC zone. The surface waters in the northern EKC area in winter were close to CO 2 gas equilibrium with the atmosphere or slightly supersaturated with carbon dioxide (pCO 2 = 380-460 µatm).
In summer, the low pCO 2 values (140-220 µatm) in the surface layer of the eastern Kamchatka and the northern Kuril Islands were connected to low salinity (32.1-32.6) waters. The EKC area was a sink for atmospheric CO 2 . The phytoplankton spring bloom in the EKC zone in early April 2013 was accompanied by a significant decrease in pCO 2 (from 400 µatm to 250-300 µatm) in the waters with negative temperatures supplied from the Bering Sea through the Kamchatka Strait and was tied to the boundaries of ice fields.
The distributions of temperature, salinity, and pCO 2 in the central Kuril Islands region (45-48 • N, 150-155 • E) were related to water dynamics. The water mixing in the central Kuril Straits and the Kruzenshterna Bank leads to increased salinity (33.2-33.4) and high values of pCO 2 (480-670 µatm) in the surface layer of the EKC.
The comparison of the pCO 2 data collected in winter demonstrates an increase in pCO 2 in the surface layer between 1998/2001 and 2018/2020 at about 50 µatm for the waters with a salinity of 33.0-33.1, which is in agreement with increased pCO 2 in the atmosphere at 46 µatm (from 368 to 414 µatm) during this period.
Author Contributions: Conceptualization, A.A. and I.P.; methodology, A.A. and I.P.; software, A.A.; investigation, A.A.; writing-original draft preparation, A.A. and I.P. All authors have read and agreed to the published version of the manuscript.