Volume 96, Issues 1-2, 11 August 2005, Pages 87-97
The partial pressure of carbon dioxide and air–sea fluxes in the northern South China Sea in spring, summer and autumn
|Referred to by:||Erratum to “The partial pressure of carbon dioxide and
air–sea fluxes in the northern South China Sea in spring, summer and
autumn” [Marine Chemistry 96 (2005) 87–97]|
Marine Chemistry, Volume 103, Issues 1-2, 8 January 2007, Page 209
Weidong Zhai, Minhan Dai, Wei-Jun Cai, Yongchen Wang and Huasheng Hong
SummaryPlus | Full Text + Links | PDF (74 K)
The distribution of pCO2 in the surface waters of the northern South China Sea (NSCS) was examined in the summer of 2000, the spring of 2001 and the late fall of 2002. For the offshore region >100 km away from the coastline, surface water pCO2 varied within a range of 360–450 μatm during all the surveys. Nevertheless, they were generally higher than atmospheric pCO2. Sea–air ΔpCO2 ranged in 50–100 μatm in the summer, 0–50 μatm in the spring and 0–90 μatm in the late fall. Average sea-to-air CO2 flux was 7 mmol CO2 m−2 day−1 in the summer and 1–3 mmol CO2 m−2 day−1 in the spring and fall. Nearshore pCO2 showed a very dynamic pattern likely associated with the regional hydrodynamic settings, yet again pCO2 in the surface water overall exceeded the air pCO2. Data from this study thus suggests that the NSCS served as a source of atmospheric CO2. Seasonal variations of the pCO2 in the NSCS outer shelf and slope surface waters were significantly influenced by sea surface temperature.
Keywords: Air–sea flux; Carbon dioxide; Marginal sea; South
Continental shelves and slopes comprise only 7% of the world's ocean surface area, yet they play a disproportionally important role in the global oceanic carbon cycle (Walsh et al., 1981, Rabouille et al., 2001 and Chen, 2003). At present it remains difficult to reliably assess the source/sink terms and the associated air–sea CO2 fluxes of the global ocean margin due primarily to the lack of pCO2 field data with high spatial and temporal resolution in this complex regime (Fasham et al., 2001). Based upon annually integrated data obtained from the East China Sea (ESC) (Chen and Wang, 1999, Tsunogai et al., 1999 and Wang et al., 2000) and the North Atlantic European shelf, Frankignoulle and Borges (2001) argued that middle- and high-latitude continental shelves in the Northern Hemisphere are a net sink of atmospheric CO2. Sea–air fluxes in these shelves were estimated in the range from −0.68 mol m−2 year−1 off New Jersey (Boehme et al., 1998) and −0.9 mol m−2 year−1 in the Baltic Sea (Thomas and Schneider, 1999), to −2.9 mol m−2 year−1 in the ESC (Tsunogai et al., 1999 and Wang et al., 2000) and in the North Atlantic European shelf (Frankignoulle and Borges, 2001). Liu et al. (2000) hypothesized that the world continental margins as a whole would be a weak CO2 sink, 0.1 Gt C year−1 as indicated by the net downward flux between atmosphere and margins. Extrapolated from seasonal field observations in a Northern European shelf sea, the North Sea, Thomas et al. (2004) reported an enhanced uptake of CO2 by global coastal and marginal seas as 20% of the world ocean's uptake of anthropogenic CO2, i.e. 0.4 Gt C year−1. We contend that reliable global margin flux estimates may have to take into account the potential latitudinal difference between world margins (Cai and Dai, 2004). It is important to realize that most of the presently studied shelves are located at middle-latitude associated with a higher level of biological productivity. As revealed in open ocean regimes, the relative importance of the overall biological effect decreases from high latitude to low latitude. For example, the relative importance ratios of temperature to biological effects on surface water pCO2 were typically 0.02 for the Ross Sea at 76°30′S, 0.9 for the North Pacific at 50 °N and 2.7 for the Sargasso Sea at 32°50′N (Takahashi et al., 2002). Therefore, while higher latitude shelves have been reported to behave as an atmospheric CO2 sink, low latitude ocean margins may behave differently (Cai et al., 2003 and Cai and Dai, 2004). Thus, a better understanding of the fluxes associated with the tropical and subtropical shelf waters is required to better constrain the overall source/sink terms of marginal seas.
Located between the equator and 23 °N, and characterized by a tropical and subtropical climate, the South China Sea (SCS) is the world's second largest marginal sea with a deep semi-closed basin and wide continental shelves to the northwest and south (Chen et al., 2001 and Liu et al., 2002). This marginal sea is fed by two world major rivers (the Mekong and the Pearl Rivers) and some smaller rivers featuring tropical and subtropical drainage basins. Thus it provides an interesting contrast to most other CO2 flux studies at marginal seas. Thus far, direct measurements of pCO2 in this region were limited to a survey along the eastern boundary of the SCS in September 1994 (Rehder and Suess, 2001) and two surveys in the Taiwan Strait (north to the SCS, Fig. 1a) in August 1994 and February 1995 (Zhang et al., 2000). pCO2 from the above two studies were measured using semi-continuous or discrete systems based on gas chromatography. Data from Rehder and Suess (2001) suggest that, when SST reached its maximum in the late summer, the SCS was a moderate source of atmospheric CO2 with sea–air CO2 fluxes of 0–1.9 mmol m−2 day−1 in the basin and 0.3–5.5 mmol m−2 day−1 in the southern shelf. Zhang et al. (2000) reported that the southern Taiwan Strait was a weak source in summer and a sink in winter, with the sea–air fluxes of 0.1 and −8 mmol m−2 day−1, respectively. In addition, Chen and Huang (1995) estimated that the SCS contained 0.43 Gt C of anthropogenic CO2 with a limited penetration depth of 500 m based upon seawater carbonate system measurements in the northeastern SCS. In summary, sea–air CO2 flux data are at paucity and results from current studies are far from conclusive whether the SCS acts as a source or sink for atmospheric CO2.
Display Full Size version of this image (189K)
Fig. 1. Maps of the area under study and surface water pCO2 distributions (in μatm) during each cruise. (a) Study area and cruise tracks. Transect A represents the repeated transect during all the three cruises. Transect B was surveyed during the 2000 and 2002 cruises. Transects B′ and ST were surveyed during the 2000 cruise. (b) pCO2 distributions in Jul–Aug 2000. Data from transects B, B′ and ST are not shown because they are not fully corrected. See text for detailed descriptions. (c) pCO2 distributions in May–Jun 2001. The dull green area around the Pearl River estuary represents an area where a spring bloom occurred during the survey, where aqueous pCO2 was very dynamic and the lowest came to 200 μatm. See Fig. 2 for detailed data. (d) pCO2 distributions in Nov 2002.
We report here the distribution of pCO2 in the northern SCS (NSCS) based upon underway determinations during 3 cruises that encompass spring, summer and late fall. These data represent one of the most complete pCO2 data sets obtained thus far for China Marginal Seas, and provide an important case study which implies low latitude marginal sea may act a source of atmospheric CO2.
2.1. Study area and survey transects
Climatic variations in the atmosphere and in the upper ocean of the SCS are primarily dominated by the Asian Monsoon. The rain-bearing southwest monsoon lasts from June to September, but the northeast monsoon, typically with higher wind speed prevails in winter, from November to March (Han, 1998 and Liu et al., 2002). Surface circulation in the SCS changes drastically according to the season in response to the Asian Monsoon (Hu et al., 2000 and references therein). During the southwest monsoon period, seasonal upwelling distributes along the coast (Han, 1998). Runoff in the amount of 3.8×1011 m3 year−1 pours into the NSCS, 90% of which is from the Pearl River and the Hanjiang River (Fig. 1a; Han, 1998).
This study is a part of the CAR-TTT project (CARbon Transfer, Transport and Transformation) in the Pearl River Estuary and the adjacent NSCS. Data were collected on 3 cruises on board R/V Yanping II in July–August of 2000, May–June of 2001 and November of 2002. During these cruises, we performed underway measurements for temperature, salinity and pCO2. Detailed cruise tracks and transects are shown in Fig. 1a. Transect A covers the Pearl River Estuary (114 °E, 22 °N) to the southwest to the Dongsha Islands (115 °48′E, 20 °10′N), crossing through the shelf (<200 m deep within a distance of 200 km) to the slope (with water depth of 200–1000 m within a distance of 100 km). This transect was surveyed during all 3 cruises. Data from them are the major sources for this study. Transect B was surveyed in both 2000 and 2002. Transects B' and ST were investigated only in 2000.
2.2. pCO2 determination
Surface water was pumped from a side intake at a depth of 2 m. SST and conductivity were measured continuously using a SEACAT thermosalinograph system (CTD, SBE21, Sea-Bird) with an in situ temperature sensor. This underway CTD system was calibrated at Sea-Bird, just prior to our 2000 cruise. Surface water pCO2 was determined using an underway system with a continuous flow and cylinder-type equilibrator (9 cm in diameter and 20 cm long) that is filled with plastic balls and has 100 mL of the headspace. Detailed configuration of our underway measurement system has been given in Zhai et al. (2005). The CO2 mole fraction in dry air (xCO2) was detected continuously using a Li-Cor® non-dispersive infrared (NDIR) spectrometer (LI-6262 during the 2000 cruise and LI-6252 during the 2001 and 2002 cruises). The NDIR detectors were calibrated every 8 h against 2–4 CO2 gas standards and one N2 reference. xCO2 of the standards ranged from 273×10−6 to 819×10−6. The overall uncertainty of the xCO2 measurements is <1% as constrained by our standard gases (Zhai et al., 2005). Temperature probes of YSI UPS600 (during the 2000 cruise), YSI 6000UPG3-B-M-T (during the 2001 cruise) and WTW's CellOx 325 DO probe (during the 2002 cruise) were used to continuously measure temperature in the equilibrator. The temperature difference between the equilibrator and the sea surface ranged between −0.35 °C and 0.55 °C during the 2000 and 2001 cruises, and between −0.16 °C and 1.2 °C for 85% data during the 2002 cruise. Air pCO2 was determined every 4–12 h. The bow intake from which atmospheric air was pumped was installed at 15 m (during 2000 and 2001 cruises) and 6 m (during the 2002 cruise) above the water surface to avoid contamination from the ship.
We calculated water saturated pCO2 in the equilibrator from the xCO2 in dry air, atmospheric pressure and equilibrium water vapor (DOE, 1994):
Air pCO2 measurements from the 2000 and 2001 cruises were corrected to 100% humidity at in situ SST and salinity. Those from the 2002 cruise were corrected to in situ air temperature and relative humidity that was measured with an onboard weather station at 10 m height above the water surface.
2.3. Sea–air CO2 fluxes estimation
Net CO2 flux (F) was estimated using F=k×KH×ΔpCO2, where k is the gas transfer velocity of CO2, KH is the solubility of CO2 in seawater (Weiss, 1974), and ΔpCO2 is mean sea–air pCO2 difference. A positive flux value represents a net CO2 exchange from sea to the atmosphere and a negative flux value refers to the net CO2 exchange from the atmosphere to the sea. Since the k value of the NSCS measured on the spot is not available, we use Wanninkhof (1992) function of wind speed (u) to calculate the value:
Field measurements had suggested that the Wanninkhof (1992) quadratic relationship (for short-term winds) is reliable at low/moderate wind speeds, i.e., 0–12 m s−1 (McGillis et al., 2001). This range of wind speed is consistent with our spot-measured wind speeds along transect A, which ranged from 0 to 13 m s−1 during the surveys.
3.1. Surface water pCO2 distribution and its seasonal variability
Surface water pCO2 ranged between 325 and 650 μatm during the 2000 cruise (Fig. 1b). Most areas were over-saturated with respect to the atmospheric CO2. However, undersaturation data (pCO2<360 μatm) were measured in southern Taiwan Strait during the earlier legs of the cruise (Fig. 1b). In the spring of 2001, the shelf and slope waters were generally over-saturated, with a pCO2 level of 370–430 μatm (Fig. 1c). Exceptions existed in the area close to the Pearl River Estuary where pCO2 was undersaturated and around the Hanjiang River estuary where surface water CO2 was nearly in equilibrium with the atmosphere. During the 2002 cruise, surface water pCO2 ranged between 360 and 500 μatm (Fig. 1d) and was again mostly higher than air pCO2 with exceptions in the region located near shore during the terminal legs of the cruise, where air pCO2 increased from 357–371 μatm to 382 μatm when the winter monsoon came in from the continent of Asia, and some aqueous pCO2 were 0–10 μatm lower than the air pCO2.
3.1.1. Variability of surface pCO2 in outer shelf and slope waters
Based on repeated surveys along transect A, surface water pCO2 in the outer shelf (defined here as more than 100 km away from the coastline) and the slope waters varied in a range of 360–450 μatm (Fig. 2a), while surface salinity was mostly within a narrow range of 33.4–34.0 during the three seasons (Table 1), suggesting that water mass exchanges and/or freshwater influx may not be a determining factor on the seasonal variation of surface pCO2 in the region.
Display Full Size version of this image (62K)
Fig. 2. Distributions of salinity, temperature and pCO2 along transect A. (a) Air–sea pCO2. (b) Surface water salinity. (c) Normalized aqueous pCO2 at 26 °C. (d) Surface water temperature. The vertical grey line in the diagrams represents 100 km away from the coastal line, which is suggested in this work as the division between inner shelf/coast and outer shelf/slope areas along the NSCS transect A.
Surface salinity, temperature and pCO2 in the outer shelf (>100 km away from the coastline) and the slope of the northern South China Sea
Observation time Aqueous pCO2 (μatm) Temperature (°C) Salinity Normalized pCO2 (μatm at 26 °C) Average S.D. Average S.D. Average S.D. Average S.D. 19–21 Jul 2000 411 12 28.7 0.08 33.58 0.14 368 10 24–25 May 2001 400 15 28.0 0.56 33.66 0.14 371 13 29 May 2001 391 11 27.5 0.23 33.81 0.14 366 10 13–14 Nov 2002 393 18 26.7 0.37 33.73 0.09 382 17
The average surface water pCO2 was 393±18 μatm in the late fall, 411±12 μatm in the summer and between 391±11 and 400±15 μatm during the spring (Table 1). By comparison, air pCO2 was 350 μatm in summer of 2000, 370 μatm in spring of 2001 and 363 μatm in late fall of 2002. Surface water pCO2 was thus generally higher than the air pCO2 in the offshore waters (Fig. 2a), suggesting a supersaturation of CO2 in surface water with respect to the atmosphere. This supersaturation reached the highest in the summer with the sea–air ΔpCO2 values of 50–100 μatm. This ΔpCO2 range declined to 0–50 μatm in the spring of 2001 and 0–90 μatm in the late fall of 2002.
pCO2 along transect A showed an overall uniform distribution in the outer shelf/slope during the summertime survey. However, during the spring and fall surveys, a general distribution pattern with high pCO2 on the shelf to some extant but generally higher pCO2 in the slope waters was observable (Fig. 1c and d). For example, during the fall survey of 2002, pCO2 in the outer shelf mostly ranged between 370 μatm and 410 μatm , while the slope pCO2 rose to 400–450 μatm at >250 km away from the coastline.
It should be mentioned that during our surveys along transects B, B′ and ST from July 30 to August 1, 2000, the YSI probe was out of order, which prevented us from fully correcting our pCO2 data. We are not attempting to use these data for quantitative analysis or flux calculation. As a general feature, pCO2 along these transects showed a very complex distribution pattern likely resulted from diverse hydrodynamic processes as revealed by Zhuang et al. (2003) based upon relevant data from the same cruise. Yet again, seawater was supersaturated with respect to the atmospheric CO2. It is thus evident to conclude that the NSCS shelf and slope surface waters were mostly over-saturated with respect to atmospheric CO2 throughout the three seasons covered by our cruises.
It is noteworthy that our spot-measured SST (26.7±0.4 °C, Table 1) in the outer shelf/slope areas during the 2002 survey was 0.6 °C higher than the long-term mean November SST, 26.1±0.7 °C around 20 °N 116 °E based on Wang et al. (2002) and NODC (Levitus) World Ocean Atlas 1998 (http://www.sciencedirect.com/science?_ob=RedirectURL&_method=externObjLink&_locator=url&_plusSign=%2B&_targetURL=http%253A%252F%252Fwww.cdc.noaa.gov%252Fcdc%252Fdata.nodc.woa98.html). Towards the end of our cruise, strong and cold wind came in, and the field-measured air temperature declined from 26.0–27.1 °C on 15 Nov 2002 to 17.9 °C at noon of 16 Nov 2002. Although the outer shelf/slope SST data thereafter are not available, it is reasonable to deduce that there was a significant SST drop in the area. Supportive information may also be identified by SST changes in the neighboring Western Pacific Ocean (WPO) before and after 16 Nov 2002. In the neighboring WPO waters, the weekly averaged 29 °C isothermal of SST moved from 18°N before 16 Nov 2002 to 9–14 °N in the following week (NOAA-CIRES Climate Diagnostics Center, http://www.sciencedirect.com/science?_ob=RedirectURL&_method=externObjLink&_locator=url&_plusSign=%2B&_targetURL=http%253A%252F%252Fwww.cdc.noaa.gov%252Fcdc%252Fdata.ncep.pac.ocean.html), which suggests a strong cooling event. This also suggests that our 2002 cruise data may not be typical of fall rather it is in a transitional period between fall and winter in the NSCS.
3.1.2. pCO2 variability along the coast
In the NSCS inner shelf/coastal areas, pCO2 depicted a much higher variability, both spatially and temporally as compared to the outer shelf (Fig. 2a). Such dynamic distribution had been observed in other continental shelves (Boehme et al., 1998, Frankignoulle and Borges, 2001 and DeGrandpre et al., 2002). Determinative processes that controlled pCO2 in these nearshore regions were complex and might be related to diverse hydrodynamic processes such as summer upwelling, river plumes and various fronts. However, what appeared to be clear from our data was that the Pearl River plume had a significant impact on surface water pCO2 distribution. A low salinity zone as shown in Fig. 2b manifested the extension of the Pearl River plume. This plume extended the farthest offshore in the spring of 2001. During the 2002 cruise, we did not observe significantly low salinity zones along the main transect. This is likely due to the fact that the transect A survey was not embarked until at a station located at 40 km offshore from the coast in Nov 2002 and partly that this was a low flow season. It appears that the influence of the Pearl River plume on the distribution of surface pCO2 occurred within 100 km during high flow times and 50 km during low flow times from the coastline (Fig. 2a). This influence was reflected by a very dynamic pCO2 distribution pattern observed in nearshore regions. A highly supersaturated CO2 (400 μatm to >600 μatm) was observed in the river plume in summer during the 2000 cruise (Fig. 1 and Fig. 2). As a comparison, the river plume area displayed pCO2 values as low as 180–250 μatm (Fig. 2a) in May 2001 in the same area (Fig. 1).
3.2. Influences of SST on surface pCO2 of the NSCS outer shelf and slope
During the surveys water temperature in the offshore region had a seasonal variation of 2 °C (Table 1) while the nearshore SST ranged in 25–29.5 °C (Fig. 2d). In order to filter the effect of temperature (Borges and Frankignoulle, 2002, Takahashi et al., 2002 and Copin-Montégut et al., 2004), we normalized aqueous pCO2 with the coefficient of 4.23% °C−1 (according to Takahashi et al., 1993 and Takahashi et al., 2002) to an annual average SST (26 °C) of the offshore water under survey. The results are presented in Fig. 2c, which shows that seasonal variations of normalized pCO2 on the shelf and the slope along transect A was substantially smaller than the in situ pCO2 as shown in Fig. 2a. Most significantly, the large difference in pCO2 between summer and spring disappears upon such normalization (Table 1). The normalized pCO2 in late fall remains high while the variation is in the range of standard deviation of all data (±17 μatm) (Table 1). This suggests that the seasonal variations of the sea surface pCO2 in the NSCS outer shelf and the slope were mainly influenced by the seasonal variations of SST.
In order to validate the major influence of SST on the outer shelf and slope pCO2, we plot relationships among pCO2, SST and surface salinity in the NSCS outer shelf and the slope (> 100 km away from the coast). As revealed by Fig. 3a, most field pCO2 data (in μatm) were dependent on SST and varied in a range of (370±20)*e0.0423 (SST-26). Exceptions were mainly associated with the outer shelf during May 24–25 of 2001 and on the slope area during Nov 13–14 of 2002. These two exceptions had the field-measured pCO2 data of 0–50 μatm higher than those predicted from the pCO2–SST relationship (Fig. 3a, also shown in Fig. 2c), while the surface salinity was both 33.8 (Fig. 3b). This might be a result of mixing from another water mass with high pCO2, which needs to be further investigated. In contrast to the positive relationship of pCO2–SST, the influence of salinity on pCO2 was generally random (Fig. 3b).
Display Full Size version of this image (34K)
Fig. 3. Relationships among pCO2, temperature and salinity in the NSCS outer shelf (>100 km away from the coast) and the slope. (a) pCO2–T. (b) pCO2–S. The two dashed lines in (a) represent functions of pCO2 (μatm)=390e0.0423 (SST-26) (the upper line) and of pCO2 (μatm)=350e0.0423 (SST-26) (the lower line), respectively.
The fact that SST played a major role in influencing the seasonal variability of pCO2 in the NSCS outer shelf and slope is consistent with much prior research at open oceans from a thermodynamic viewpoint (Millero, 1995 and Sabine et al., 2000) because temperature determines to a large extent the solubility of CO2 at a given salinity. It is also consistent with the generally low productivity in these sea areas (Cai et al., 2004). In a parallel study based on DIC and alkalinity depletion and plume travel time, Cai et al. (2004) estimated the summertime primary production in the inner shelf of transect A as 0.8 gC m−2 day−1, which was comparable to other large river plumes. However, in the open continental shelf and slope, the biological production decreased 80% to as low as 0.12–0.17 gC m−2 day−1 (Cai et al., 2004). Thus the combination of SST and primary productivity level may have caused the different variability of surface water pCO2 between the inner shelf/coast and the outer shelf/slope. Quantitative analysis of biological effect on pCO2 relative to the SST effect awaits however further investigations.
3.3. Sea–air CO2 flux
Based on CO2 sea–air flux calculations along transect A (Table 2) and then extrapolated over the entire NSCS region, we obtain an overview on the NSCS CO2 air–sea exchange in the seasons under survey. For the summer time, our estimation is 7 mmol CO2 m−2 day−1 for CO2 sea–air flux in the outer shelf and slope waters, which is much higher than previous estimations in the southern Taiwan Strait (0.1 mmol m−2 day−1, Zhang et al., 2000) and along the eastern boundary of the SCS (0–1.9 mmol m−2 day−1 in the basin and 0.3–5.5 mmol m−2 day−1 in the southern shelf, Rehder and Suess, 2001). As for the spring and fall seasons, CO2 sea–air fluxes in the outer shelf and slope waters were estimated as 1–3 mmol CO2 m−2 day−1.
Summary of pCO2 and sea–air CO2 fluxes in the Northern South China sea
a Mean aqueous pCO2 along transect A.
Observation time Aqueous pCO2a (μatm) Air pCO2b (μatm) Wind speedc (m s−1) Estimated fluxesd (mmol CO2 m−2 day−1) Survey average Seasonal average Outer shelf and slope regions of the northern SCS, 1×105 km2 24–25 May 2001 400 372.3 7 3.3 2.7 (spring) 29 May 2001 391 372.3 7 2.2 19–21 Jul 2000 411 349.1 7 7.5 7.5 (summer) 13–14 Nov 2002 393 362.8 4.4 1.4 1.4 (autumn) Inner shelf/coastal areas potentially influenced by the Pearl River plume, 3000 km2 24 May 2001 226 372.3 7 −18.1 29 May 2001 339 372.3 7 −4.0 19 Jul 2000 434 349.1 7 10.2 12 Nov 2002 405 362.8 4.4 2.0
b Mean atmospheric pCO2 along all shelf/slope transects from all cruises. During our 2000 cruise, CO2 mole fraction (xCO2) for dry air was measured to be 365.0–366.7 μmol mol−1, which was slightly lower than the nearby CMDL flask data (369.0 μmol mol−1) at Mariana Islands [GMI] (13°26′N 144°47′E). During the 2001 cruise, the measured xCO2 varied between 373.7 μmol mol−1 and 391.6 μmol mol−1, and 388 μmol mol−1 was used as a local average xCO2 although it was higher than the CMDL flask data (374.0 μmol mol−1). During the 2002 cruise, measured xCO2 was 370.0–376.4 μmol mol−1, which was consistent with the CMDL flask data (372.5 μmol mol−1). The CMDL flask data are from the Climate Monitoring and Diagnostics Laboratory (http://www.sciencedirect.com/science?_ob=RedirectURL&_method=externObjLink&_locator=url&_plusSign=%2B&_targetURL=http%253A%252F%252Fwww.cmdl.noaa.gov%252Fccgg%252Fiadv%252Findex.php).
c Mean wind speed along the cruise tracks. Recorded from an onboard weather station at 10 m height. The mean wind speeds were 7.3±1.5 m s−1 for the 19–21 Jul 2000 survey, 7.4±2.9 m s−1 for the 24–25 May 2001 survey, 6.7±2.4 m s−1 for the 29 May 2001 survey and 4.4±1.0 m s−1 for the 12–14 Nov 2002 survey.
d CO2 sea–air fluxes based on the Wanninkhof (1992) air–sea transfer coefficient. A positive value represents the net CO2 exchange from sea to the atmosphere and a negative value refers to the net CO2 exchange from the atmosphere to the sea.
While the NSCS offshore area is generally a source of CO2 to the atmosphere in the warm period of a year, the Pearl River plume may indeed become a significant seasonal sink during a spring bloom period (Fig. 2a). However, its influences may be spatio-temporally limited. For example, a spring bloom extending approximately 90 km was observed nearly throughout the duration during our 2001 cruise (Dai et al., unpublished). The affected area is only 3% of the shelf and slope areas in our study (Fig. 1c), although in its most intensive phase, the mean influx inside the plume can be as large as 5–6 times the efflux outside bloom regions that day. The duration of CO2 undersaturation induced by this bloom was also short (2 weeks, began on May 14 and ended after May 29). Therefore the bloom episodes may not change the overall source/sink pattern in the NSCS. In addition, the regional and episodic uptake of CO2 may be overlapped by excess CO2 sources discharged by the river, as observed in the nearshore area along transect A during the 2000 cruise.
During a recent cruise of ours to NSCS in Feb–Mar 2004 (late winter), we measured sea–air pCO2 differences in the region and estimated the flux as −2.2 mmol m−2 day−1 for the outer shelf/slope areas (Zhai et al., unpublished). Ignoring the interannual variability that needs to be further investigated, an annually integrated CO2 sea–air flux at the NSCS can be roughly estimated as 3.4 mmol m−2 day−1 or 1.2 mol m−2 year−1, suggesting that the northern South China Sea may indeed to be a weak/moderate source to the atmospheric CO2 on an annual base, with major uncertainty during the winter when enhanced CO2 uptake may occur in the region under lower temperature condition than our observations in Feb–Mar 2004.
At a global scale, most reported marginal air–sea CO2 flux measurements have been associated with middle-latitude regions, and most of these work reported that ocean margins acted as an atmospheric CO2 sink (e.g., Boehme et al., 1998, Tsunogai et al., 1999, Frankignoulle and Borges, 2001 and Thomas et al., 2004). The only reported low latitude margin that acted as a CO2 sink was in the Amazon River plume where riverine nutrients promoted CO2 uptake (Ternon et al., 2000). Cai et al. (2003), however, reported that the Southeastern U.S. continental shelf (low to middle latitude) was a significant source of CO2 to the atmosphere. They attributed this as a result of extensive organic carbon export from a highly productive coastal marsh and a relatively high annual sea surface temperature. Elsewhere in other tropical/subtropical regimes, there are examples that marginal seas can be a CO2 source, although with seasonal variations. The Arabian Sea acts as a moderate source for atmospheric CO2 in summer (July–September) and a weak source in other seasons (Sabine et al., 2000). The mean annual sea–air CO2 flux is 0.46 mol m−2 year−1 (Goyet et al., 1998). Close to the Omani coast, west of the Arabian Sea, the estimated CO2 outgassing flux varies from 0.27 to 1.5 mmol m−2 day−1 during most of the year and up to 10.5 mmol m−2 day−1 during the southwest monsoon season (July and August), driven by intense upwelling (Goyet et al., 1998). In the eastern Arabian Sea, the surface pCO2 levels were mostly higher than the atmospheric pCO2 throughout the year, although significant seasonal variation was found in coastal seawater pCO2 (266–685 μatm) due to influences of upwelling and river plumes (Sarma et al., 2000). Our study in the NSCS gives another case for the CO2 source/sink terms in a largest subtropical/tropical marginal sea. Based upon currently available data, we found that the area to be a source of atmospheric CO2 which is likely influenced by perennial high sea surface temperature and relatively low primary productivity. In addition, the phytoplanktonic community structure may also play a role in determining the source or sink of the SCS. South China Sea is a typical oligotrophic sea, where the phytoplanktonic community is characterized by small size cells (Jiao and Yang, 2002) that would have a fast turnover rate and lower export rate.
Similar to the latitudinal transition of relative importance of temperature
and biological effects on surface water pCO2 in different open
et al., 2002), there probably also exists latitudinal transition of the
marginal sea source/sink terms. We thus suggest that low latitude ocean margins
should not be overlooked while investigating the role of marginal seas in the
global ocean carbon cycle.
This research was supported by the Natural Science Foundation of China through grants #49825111, #40176025, #40228007 and #90211020 and by the Hi-Tech Research and Development Program of China (#2002AA635140). We thank Zhaozhang Chen and Ganning Zeng for assistance with CTD data collection, Zhaohui Wang for pCO2 data collection during our 2001 cruise and Kunming Xu for meteorological data collection during the 2002 cruise. Wuqi Ruan and Fan Zhang along with the crew of Yanping II provided much help during the sampling cruises. We appreciate constructive comments from George T.F. Wong during the preparation of the manuscript. Reviews and/or comments from Edward T. Peltzer, Rik Wanninkhof, Michel Frankignoulle and another anonymous reviewer greatly improved the quality of the paper.
Boehme et al., 1998 S.E. Boehme, C.L. Sabine and C.E. Reimer, CO2 fluxes from a coastal transect: a time-series approach, Mar. Chem. 63 (1998), pp. 49–67. SummaryPlus | Full Text + Links | PDF (911 K) | Abstract + References in Scopus | Cited By in Scopus
Borges and Frankignoulle, 2002 A.V. Borges and M. Frankignoulle, Distribution and air–water exchange of carbon dioxide in the Scheldt plume off the Belgian coast, Biogeochemistry 59 (2002), pp. 41–67. Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus
Cai et al., 2003 W.J. Cai, Z. Wang and Y. Wang, The role of marsh-dominated heterotrophic continental margins in transport of CO2 between the atmosphere, the land–sea interface and the ocean, Geophys. Res. Lett. 30 (2003) (16), p. 1849. Full Text via CrossRef
Cai et al., 2004 W.-J. Cai, M.H. Dai, Y.C. Wang, W.D. Zhai, T. Huang, S.T. Chen, F. Zhang, Z.Z. Chen and Z.H. Wang, The biogeochemistry of inorganic carbon and nutrients in the Pearl River estuary and the adjacent Northern South China Sea, Cont. Shelf Res. 24 (2004), pp. 1301–1319. SummaryPlus | Full Text + Links | PDF (750 K) | Abstract + References in Scopus | Cited By in Scopus
Chen, 2003 C.T.A. Chen, New vs. export production on the continental shelf, Deep-Sea Res., Part II 50 (2003), pp. 1327–1333. SummaryPlus | Full Text + Links | PDF (221 K) | Abstract + References in Scopus | Cited By in Scopus
Chen and Huang, 1995 C.T.A. Chen and M.H. Huang, Carbonate chemistry and the anthropogenic CO2 in the South China Sea, Acta Oceanol. Sin. 14 (1995) (1), pp. 47–57.
Chen and Wang, 1999 C.T.A. Chen and S.L. Wang, Carbon, alkalinity and nutrient budgets on the East China Sea continental shelf, J. Geophys. Res. 104 (1999) (C9), pp. 20675–20686. Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus
Chen et al., 2001 C.T.A. Chen, S.L. Wang, B.J. Wang and S.C. Pai, Nutrient budgets for the South China Sea basin, Mar. Chem. 75 (2001), pp. 281–300. SummaryPlus | Full Text + Links | PDF (559 K) | Abstract + References in Scopus | Cited By in Scopus
Copin-Montégut et al., 2004 C. Copin-Montégut, M. Bégovic and L. Merlivat, Variability of the partial pressure of CO2 on diel to annual time scales in the Northwestern Mediterranean Sea, Mar. Chem. 85 (2004), pp. 169–189.
DeGrandpre et al., 2002 M.D. DeGrandpre, G.J. Olbu, C.M. Beatty and T.R. Hammar, Air–sea CO2 fluxes on the US Middle Atlantic Bight, Deep-Sea Res., Part II 49 (2002), pp. 4355–4367. SummaryPlus | Full Text + Links | PDF (3076 K) | Abstract + References in Scopus | Cited By in Scopus
DOE, 1994 DOE In: A.G. Dickson and C. Goyet, Editors, Handbook of methods for the analysis of the various parameters of the carbon dioxide system in sea water, ver.2 (1994) ORNL/CDIAC-74.
Fasham et al., 2001 In: M.J.R. Fasham, B.M. Balino and M.C. Bowles, Editors, A new vision of ocean biogeochemistry after a decade of the Joint Global Ocean Flux Study (JGOFS). AMBIO, Special Rep. 10, Royal Swedish Academy of Sciences, Stockholm, Sweden (2001) 31 pp..
Frankignoulle and Borges, 2001 M. Frankignoulle and A.V. Borges, European continental shelf as a significant sink for atmospheric carbon dioxide, Glob. Biogeochem. Cycles 15 (2001), pp. 569–576. Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus
Goyet et al., 1998 C. Goyet, F.J. Millero, D.W. O'Sullivan, G. Eischeid, S.J. McCue and R.G.J. Bellerby, Temporal variations of pCO2 in surface seawater of the Arabian Sea in 1995, Deep-Sea Res., Part I 45 (1998), pp. 609–624.
Han, 1998 W.Y. Han, Marine chemistry in the South China Sea, Science Press, Beijing, China (1998) 289 pp. (in Chinese).
Hu et al., 2000 J.Y. Hu, H. Kawamura, H.S. Hong and Y.Q. Qi, A review on the currents in the South China Sea: seasonal circulation, South China Sea warm current and Kuroshio intrusion, J. Oceanogr. 56 (2000), pp. 607–624. Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus
Jiao and Yang, 2002 N.Z. Jiao and Y.H. Yang, Ecological studies on Prochlorococcus in China seas, Chin. Sci. Bull. 47 (2002), pp. 1243–1250. Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus
Liu et al., 2000 K.K. Liu, L. Atkinson, C.T.A. Chen, S. Gao, J. Hall, R.W. Macdonald, L.T. McManus and R. Quiñones, Exploring continental margin carbon fluxes on a global scale, EOS, Trans.-Am. Geophys. Union 81 (2000), pp. 641–644.
Liu et al., 2002 K.K. Liu, S.Y. Chao, P.T. Shaw, G.C. Gong, C.C. Chen and T.Y. Tang, Monsoon-forced chlorophyll distribution and primary production in the South China Sea: observations and a numerical study, Deep-Sea Res., Part I 49 (2002), pp. 1387–1412. SummaryPlus | Full Text + Links | PDF (1787 K) | Abstract + References in Scopus | Cited By in Scopus
McGillis et al., 2001 W.R. McGillis, J.B. Edson, J.D. Ware, J.W.H. Dacey, E.H. Jeffrey, C.W. Fairall and R. Wanninkhof, Carbon dioxide flux techniques performed during GasEx-98, Mar. Chem. 75 (2001), pp. 267–280. SummaryPlus | Full Text + Links | PDF (351 K) | Abstract + References in Scopus | Cited By in Scopus
Rabouille et al., 2001 C. Rabouille, F.T. Mackenzie and L.M. Ver, Influence of the human perturbation on carbon, nitrogen, and oxygen biogeochemical cycles in the global coastal ocean, Geochim. Cosmochim. Acta 65 (2001), pp. 3615–3641. Abstract | PDF (580 K) | Abstract + References in Scopus | Cited By in Scopus
Rehder and Suess, 2001 G. Rehder and E. Suess, Methane and pCO2 in the Kuroshio and the South China Sea during maximum summer surface temperatures, Mar. Chem. 75 (2001), pp. 89–109.
Sabine et al., 2000 C.L. Sabine, R. Wanninkhof, R.M. Key, C. Goyet and F.J. Millero, Seasonal CO2 fluxes in the tropical and subtropical Indian Ocean, Mar. Chem. 72 (2000), pp. 33–53. SummaryPlus | Full Text + Links | PDF (3594 K) | Abstract + References in Scopus | Cited By in Scopus
Sarma et al., 2000 V.V.S.S. Sarma, M.D. Kumar, M. Gauns and M. Madhupratap, Seasonal controls on surface pCO2 in the central and eastern Arabian Sea, Proc. Indian Acad. Sci. 109 (2000), pp. 471–479. Abstract + References in Scopus | Cited By in Scopus
Takahashi et al., 1993 T. Takahashi, J. Olafsson, J.G. Goddard, D.W. Chipman and S.C. Sutherland, Seasonal variation of CO2 and nutrients in the high-latitude surface ocean: a comparative study, Glob. Biogeochem. Cycles 7 (1993), pp. 843–878. Abstract + References in Scopus | Cited By in Scopus
Takahashi et al., 2002 T. Takahashi, S.C. Sutherland, C. Sweeney, A. Poisson, N. Metzl, B. Tilbrook, N. Bates, R. Wanninkhof, R.A. Feely, C. Sabine, J. Olafsson and Y. Nojiri, Global sea–air CO2 flux based on climatological surface ocean pCO2, and seasonal biological and temperature effects, Deep-Sea Res., Part II 49 (2002), pp. 1601–1622. SummaryPlus | Full Text + Links | PDF (2371 K) | Abstract + References in Scopus | Cited By in Scopus
Ternon et al., 2000 J.F. Ternon, C. Oudot, A. Dessier and D. Diverres, A seasonal tropical sink for atmospheric CO2 in the Atlantic Ocean: the role of the Amazon River discharge, Mar. Chem. 68 (2000), pp. 183–201. SummaryPlus | Full Text + Links | PDF (1396 K) | Abstract + References in Scopus | Cited By in Scopus
Thomas and Schneider, 1999 H. Thomas and B. Schneider, The seasonal cycle of carbon dioxide in Baltic Sea surface waters, J. Mar. Syst. 22 (1999), pp. 53–67. Abstract | PDF (1053 K) | Abstract + References in Scopus | Cited By in Scopus
Thomas et al., 2004 H. Thomas, Y. Bozec, K. Elkalay and H.J.W. de Baar, Enhanced open ocean storage of CO2 from shelf sea pumping, Science 304 (2004), pp. 1005–1008. Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus
Tsunogai et al., 1999 S. Tsunogai, S. Watanabe and T. Sato, Is there a “continental shelf pump” for the absorption of atmospheric CO2?, Tellus 51B (1999), pp. 701–712. Abstract + References in Scopus | Cited By in Scopus
Walsh et al., 1981 J.J. Walsh, G.T. Rowe, R.L. Iverson and C.P. McRoy, Biological export of shelf carbon is a sink of the global CO2 cycle, Nature 291 (1981), pp. 196–201. Full Text via CrossRef | Abstract + References in Scopus | Cited By in Scopus
Wang et al., 2000 S.L. Wang, C.T.A. Chen, G.H. Hong and C.S. Chung, Carbon dioxide and related parameters in the East China Sea, Cont. Shelf Res. 20 (2000), pp. 525–544. SummaryPlus | Full Text + Links | PDF (674 K) | Abstract + References in Scopus | Cited By in Scopus
Wang et al., 2002 D.X. Wang, Y. Du and P. Shi, Climatological atlas of physical oceanography in the upper layer of the South China Sea, Meteorology Press, Beijing, China (2002) 168 pp. (in Chinese).
Zhai et al., 2005 W.D. Zhai, M.H. Dai, W.J. Cai, Y.C. Wang and Z.H. Wang, High partial pressure of carbon dioxide and its maintenance mechanism in the Pearl River estuary, China, Mar. Chem. 93 (2005), pp. 21–32. SummaryPlus | Full Text + Links | PDF (1027 K) | Abstract + References in Scopus | Cited By in Scopus
Zhang et al., 2000 Y.H. Zhang, Z.Q. Huang, W.Q. Wang, X.B. Huang and Z.Y. Chen, A study of carbon dioxide in Taiwan Strait, J. Oceanogr. Taiwan Strait 19 (2000) (2), pp. 163–169 (in Chinese).
et al., 2003 W. Zhuang, J.Y. Hu, Z.G. He, G.N. Zeng and Z.Z. Chen, An
analysis on surface temperature and salinity from southern Taiwan Strait to
Zhujiang River estuary during July–August, 2000, J. Trop. Oceanogr.
22 (2003) (4), pp. 68–76 (in Chinese).
Volume 96, Issues 1-2, 11 August 2005, Pages 87-97