Vol. 46 No. 6

SCIENCE IN CHINA (Series D)

June 2003

Investigations on distributions and fluxes of sea-air CO2 of the expedition areas in the Arctic Ocean WANG Weiqiang (ฆฟஜ)1,3, CHEN Liqi (чो୥)2,3, YANG Xulin (ཷ༢ॿ)1,3 & HUANG Xuanbao (ܻ༦ͯ)1 1. Third Institute of Oceanography, State Oceanic Administration, Xiamen 361005, China; 2. Chinese Arctic and Antarctic Administration, Beijing 100860, China; 3. Key Laboratory of Global Change and Marine-Atmospheric Chemistry, State Oceanic Administration, Xiamen 361005, China Correspondence should be addressed to Wang Weiqiang (email: [email protected]) Received January 9, 2003

Abstract The distributions and fluxes of sea-air carbon dioxide were investigated the first time based on the firsthand data collected during the First Chinese National Arctic Research Expedition. The results revealed that values of atmospheric CO2 partial pressure (Pa) measured in the summer during the expedition fell between 352 and 370 (u106CO2gAir1, same unit below) with an average value of 358. Particularly, Pa appeared high in the northern sea areas of Poitlay. However, the values of CO2 partial pressure at the surface layer of seawater (Pw) ranged from 98 to 580 with the difference between the low and high being 472. The average value of Pw was 242, which is 116 lower than that of the corresponding Pa. In addition, the distribution of Pw was roughly low in the west and north, but high in the east and south. These phenomena were closely related to plankton, ice, water temperature and circulation of the region. The estimation in carbon fluxes showed that the patterns in distribution were similar through different calculating methods with an exception in eastern sea areas of the region where a weak source of atmospheric CO2 was indicated. Most sea areas of the region were sinks or strong sinks of atmospheric CO2. However, the magnitudes in the fluxes were different. The average values varied from 6.57 (Liss method) to 26.32 mg˜CO2gm2gh1 (14C method) with a difference of about 4 times between the low and high, which is 2 to 10 times as high as the global average. Compared with the fluxes in the same region obtained using model of Takahashi, Feely et al., the values determined based on Wanninkhof coefficient calculation were 2.38 times as great as those obtained by them. Keywords: carbon flux, distribution, the Arctic Ocean, carbon dioxide.

“Greenhouse effect” causing global warming has been an important issue of studying climate change. In the latest 100 years, the earth surface temperature has been increased by about 0.4ćü 0.8ć[1,2]. And this has been becoming a hotspot of the world[3,4] due to its impact on many environmental problems such as sea level rising, more intensive and longer El Niño, frequent drought and flood disasters. Previous studies indicated that the enhancement of “greenhouse effect” is due mainly to the increase of surface temperature caused by the increase of carbon dioxide concentration in atmosphere at rate of 1ü2 u106CO2gAir1ga1[5,6], because of the large consumption of mineral fuels

570

SCIENCE IN CHINA (Series D)

Vol. 46

and undue exploitation of forest in huge areas since the last century. Occupying 71% of the total earth surface, the ocean plays the role of an adjuster for CO2 content in the air[7]. The quantity of CO2 in the ocean is 50 times as large as that in the atmosphere. Present models showed that 21% ü50% of carbon dioxide releasing from mankind activities enters into the ocean, thus reducing the increase rates of carbon dioxide in the atmosphere and lowering the air temperature [6]. Obviously knowledge of the ocean’s response to the increase of atmospheric carbon dioxide is a key to estimate the increase rate of the quantity and concentration of carbon dioxide in the air and the impacts of “greenhouse effect” on climate. Polar seas are among the focus areas in carbon monitoring programs chosen and coordinated by SCOR for their importance in global carbon cycle studies. There have been a number of expeditions and research programs in hydrology, meteorology, chemistry and biology on the Arctic Ocean in the world. After World War II, especially during the International Geophysical Years, countries around the Arctic Ocean have set up more than fifty land-based scientific research stations and carried out a number of studies in various scientific aspects[8]. In the Pt. Barrow, Alaska of the United States, air samples have been collected monthly since 1973 for analysis of partial pressure in the atmosphere. The measurements showed that the Pa was the highest in May and lowest in August every year with an average annual variation of about 15u106CO2gAir1[9,10], besides Pa was continuously increasing at rate of 1.4u106CO2gAir1ga1. Based on the concentrations of atmospheric CO2 obtained during many years’ observation, Keeling et al.[11] discussed the growth condition of vegetation in North Polar Region. Walsh once reviewed carbon cycle studies before 1989 and estimated the carbon sink in the Arctic Ocean in view of biogeochemistry[12]. Walsh et al.[13] had studied carbon and nitrogen cycle for Bering Sea and the Chukchi Sea on the basis of the relationship between organic substances and AOU (apparent oxygen utilization) in the seawater of the Arctic Ocean. Their results showed that the shallow continental region of the Arctic Ocean was of very high primary productivity in summer and an important sink for atmospheric CO2. During the first cruise of China National Arctic Research Expedition, a project named “Carbon cycle study on Arctic Ocean and the adjacent areas” has been fulfilled. A large number of valuable data of CO2 in air and seawater were produced around the navigation route, the Bering Sea, the Chukchi Sea and their adjacent sea areas in trawling mode. With the above-mentioned firsthand data this paper presents the distribution patterns of CO2 in air and seawater, and the exchange fluxes of air-sea carbon in the Chukchi Sea and its adjacent areas. 1

Data sources

From 2 July, 1999 to 9 September, 1999, the first Chinese National Arctic Research Expedition was fulfilled. The data of CO2 in air and surface seawater were obtained along the way of the Chukchi Sea and its adjacent areas at different stations on each hour sharp. The distribution of sampling stations is depicted in fig. 1.

No. 6

DISTRIBUTIONS & FLUXES OF SEA-AIR CO2 IN ARCTIC OCEAN

571

1.1

Major equipments Main equipments used in the expedition included Li6262 CO2/H2O infrared analyzer, DATAKER50 Series 2 digital recorder, FTS cold trap (made in USA), QGS-B80 infrared gas analyzer (made by China-Germany joint company), platinum resistor thermometer, Golden Great Wall PC, XMX-2042 self-balancing recorder Fig. 1. Distribution of the sampling stations in the Expedi(made in China), water-gas equilibrator (made in tion of Arctic Ocean. Canada), all of them worked well for more than two months during the expedition. 1.2

Standard materials Standard carbon dioxide gas with State Standard A (GBW) in concentrations at 285, 348, 401

(u106CO2gAir1) was provided by the State Research Center of Standard Materials of China. 1.3

Observation methods Infrared analysis was performed with a system consisting mainly of 2 infrared gas analyzers, 1 cold trap, 1 water-gas-equilibrator, 2 air pumps, 2 gas flowmeters, 3 bottles of standard CO2 gases and several control valves. The air sampling head was installed at the highest point on Xuelong, the expedition vessel. A plastic pipe was used to connect the head and the entrance of air pump in the laboratory to continuously pump the air samples into the cold trap. Water vapor in the air samples was removed before each analysis of air partial pressure with QGS-B80 infrared analyzer. The seawater samples were continuously pumped into the water-gas-equilibrator through the seawater pump on Xuelong. The gas in the water-gas-equilibrator was transferred into another cold trap with an air pump and analyzed with Li6262 infrared analyzer after the water vapor was removed. A triple valve in the analytical system was used to control two infrared analyzers to determine standard CO2 gas and samples under the same condition, for which the temperatures of the cold trap were kept between 40 and 45ć and gas intake velocity at 0.5 dm3gmin1. From 2 July to 5 August, the values of Pw of the surface seawater in flowing manner and those of Pa of the atmosphere in interrupted manner were determined with QGSB 80, and the absorption values were recorded with XWX-2042 auto-balancing recorder. The measurements from 5 August to 6 September were carried out with QGSB 80 for Pa of the air and Li6262 for Pw of the surface seawater, the absorption values given by these 2 analyzers were recorded simultaneously with DATAKER50-S series 2 digital recorder every minute. The partial pressures of CO2 in the air (Pa) and in the water-gas-equilibrator (Pw) were calculated from the absorption values of standard CO2, atmosphere and air in the water-gas equilibrator. The results were reported in terms of 106CO2gAir1. The measurement accuracy of the system was 1.0.

572

2

SCIENCE IN CHINA (Series D)

Vol. 46

Results and discussion

2.1 Distribution of partial pressure of atmospheric carbon dioxide within the expedition areas in the Arctic Ocean Fig. 2 presents the distribution of partial pressure of atmospheric CO2 (Pa) in the Arctic Ocean to 65°N during this expedition (from 19 o’clock on 12 July to 13 o’clock on 27 August, 1999, most data were obtained in August). The Pa values ranged from 352 to 370 with an average value of 358, which was very close to the Pa average of 358.6 at Pt. Barrow, Alaska of the United States, where samples were collected weekly in August of 1999. It was shown that the Pa level in the north sea areas of Poitlay was high (>360) but progressively decreased along the southwest and east and then reached a minimum (<353) in the north and northeast sea areas of Pt. Barrow. The above distribution pattern of air CO2 could be related to the weather condition during the expedition. The navigation logs of the vessel Xuelong read that there was much fog or thick fog on the sea surface during the tour back and forth to Tuktoyaktuk Harbor of Canada, and during the three days’ stay in the harbor. Since CO2 in the air was absorbed by fog, its level over the north and east sea areas of Barrow was obviously low.

Fig. 2.

Distribution of partial pressure of atmospheric CO2 -6

1

in the expedition region (u10 CO2gAir ).

Fig. 3.

Distribution of partial pressure CO2 of surface sea

water in the expedition region (u106CO2gAir1).

2.2 Distribution of partial pressure of carbon dioxide in surface seawater within the expedition areas in the Arctic Ocean The values of Pw obtained in the expedition areas showed great variation and complex distribution pattern. The Pw level ranged from 98 to 480 with the difference between the high and low being 372. However, its average was only 242, which is 116 lower than the corresponding Pa value, and accordingly, seawater in this area could be sink of atmospheric CO2. It is seen from fig. 3 that Pw was basically low in the west and north, but high in the east and south. The Pw values gradually decreased towards north in the Bering Strait and towards west in north part of Tuktoyaktuk with the minimum Pw below 150 at the north sea areas of Poitlay and the maximum Pw above 400 at the sea areas near Tuktoyaktuk. During the first three navigations of four sailings back and forth to Bering Strait of the expedition, algae blooms were often observed in the north region. It has been reported that biological

No. 6

DISTRIBUTIONS & FLUXES OF SEA-AIR CO2 IN ARCTIC OCEAN

573

productivity in this region was as high as 15 gCg m2gd1[13]. A great number of organisms in shape of small medusa were observed, attaching onto the plankton trawling net when sampling from stations C1 to C6, meanwhile the concentrations of DOC at station C14 and POC at station C21 were abnormally high, in addition, high concentrations of DOC, POC and chlorophyll a were observed for the sea areas northwestern to Poitlay, indicating that low Pw in the above sea areas could well result from biological activities. The sea ice condition in the expedition region could be another factor influencing the Pw distribution among others. For the Chukchi Sea, navigations or experiments were almost conducted in areas full of broken floating ice or at the edge of floating ice from 14 to 19 July for the oceanic stations from stations C1 to C14, from 4 to 11 August for the ice station No.1 and from 18 to 26 August for the ice station No.2[14], when the ice algae was often visual in the water under the floating ice which was pushed up by the vessel. Geological samples indicated that there was large amount of organic matter in the sediments from stations C1 to C8 and from stations C10 to C13[14], and all these factors proved that the marine productivity in areas with broken floating ice and at the edge of ice region of the Chukchi Sea was very high in summer, which was further confirmed by determination of plankton, DOC, POC and chlorophyll a. Besides, to some degree, floating ice blocked the material exchange between air and sea surface, and accordingly, it is easily understood that Pw is low in this region. As for the east part of the expedition region (from 12 to 17 August), the environmental conditions featured no ice but fog, even thick fog sometimes[14], light rain and very high seawater temperature, all of which possibly had relations with the high Pw too. 2.3

Distribution of partial pressure difference of carbon dioxide between the sea and air Fig. 4 illustrates that the distribution of difference between Pa and Pw, 'P (= PaPw) was

basically mirror symmetry with Pw. It roughly increased towards the west in the north part of Tuktoyaktuk and towards the north in the Bering Strait. Except Prudhoe Bay, the eastern sea areas of the Canada Basin and the right part of the Bering Strait, where seawater could be source of atmospheric carbon dioxide since 'P İ 0, most remaining regions could become sink or strong sink of atmospheric carbon dioxide because of their 'P > 0. With the most possibly complete data set of CO2 concentrations in air and water plus the related parameters before 1994 from 250000

Fig. 4. The distribution of 'P for the expedition region of Arctic Ocean (u106CO2gAir1).

stations in the world, Takahashi, Feely et al. modeled in 4°h5° square the distribution of global air-sea CO2 partial pressure difference and the exchange flux for 1990[15] when there was no El Niño (their difference was presented as 'P = PwPa). However their model used few CO2 data for air and seawater in the Arctic Ocean except for the polar seas of the North Atlantic. Moreover,

574

SCIENCE IN CHINA (Series D)

Vol. 46

there was almost no data for Bering Sea and the regions to the east of north polar (areas to the east from the line connecting Novosibirkiye Ostrova of Russia and Mouldboy of Canada), therefore, the observation and research of carbon cycle by the First Chinese National Arctic Research Expedition filled the gap in the data of air-sea CO2 partial pressure difference, which will make contributions to study on the global carbon cycle. By comparing fig. 4 with the results from Feel and others’ models[15], it is known that they were generally in agreement, i.e. there was a large CO2 partial pressure difference between air and sea for the Chukchi Sea and the adjacent regions, hence, seawater there could become a strong sink for air CO2. However, there existed a great difference in detail. The modeled result that the coastal and nearshore seawaters along Russia and Alaska could be a strong source of the atmospheric CO2 because of the CO2 partial pressure in seawater significantly larger than that in the atmosphere (>75) was obviously worked out due to lack of measurement data for this region. 2.4

Relationships between CO2 partial pressure and environmental factors Complexities of environmental factors affect the above distribution of carbon dioxide for the expedition region. In addition to the above biological activities and surface ice, other environmental elements have also played an important role. Discussed below are the relationships of Pw with seawater temperature and surface circulation. 2.4.1 Relationship of Pw and seawater temperature. Seawater temperature is one of the key factors in marine environment, which is more crucial in polar seas since seawater temperature is close to freezing point. It could be seen that the distribution of simultaneously determined surface seawater temperature was similar to that of Pw, i.e. low in the west and north and high in the east and south. The minimum temperature of surface seawater appeared in the north and northwest areas to Pt. Barrow, the maximum at the areas near Tuktoyaktuk. A quite significant relationship in form of quadratic curve at 0.01 significant level (N = 684, R = 0.5949) existed between water temperature and Pw for the expedition region, proving that the solubility of carbon dioxide in seawater was one of the key factors affecting its concentration distribution and role as source and sink. 2.4.2 Relationship of Pw and surface water circulation. Another important factor affecting the distribution of Pw is the surface water circulation system. There are three currents in the expedition region[8,16]. The Canadian coastal current comes from northeast and flows to west; the east Siberian coastal current comes from north Russian coast and flows to southeast. In summer these two coastal currents contain large amount of land fresh water and flow in nearly opposite direction. However, it was found in this expedition that their Pw content was very high in the former and very low in the latter. Both of them met and mixed in areas near the north end of Bering Strait with Bering current, which come from the Bering Sea and flow to the north, and are of high water temperature and Pw concentration. They then flow to the southwest off Ostrov Vrangelya forming a big mixing region at the triangle sea areas near Ostrov Vrangelya-Barrow-Bering Strait. This

No. 6

DISTRIBUTIONS & FLUXES OF SEA-AIR CO2 IN ARCTIC OCEAN

575

mixing zone is often a fishing ground with high biological productivity. In turn, low Pw, high DOC and POC would be observed. Previous study [13] showed that the seawater flowing through Bering Strait consisted of 60% seawater at Anodyrskiy Zaliv, 40% Alaska coastal water. Sea areas northwest to Saint Lawrence Island had very high productivity with carbon fixation as high as 10 gCgm2gd1. Our study found that seawater at Anodyrskiy Zaliv featured low Pw (<200), which conformed with the above high biological productivity. In contrast, Alaska coastal water had very high Pw (>360), obviously flowing across the Bering Strait without well mixing resulting in clear boundary. It is known by comparison of fig. 3 with the distribution and movement of surface water circulation that the movement, mixing and distribution of Pw for the expedition region were almost controlled by the surface circulation system forming a very similar distribution pattern to that of the circulation. 3

Estimation of air-sea CO2 fluxes in the expedition region

3.1

Estimation method of air-sea CO2 fluxes Provided that the transfer of CO2 from seawater to the atmosphere (or the opposite process) is a first order process and the transfer rate is proportional to their concentrations in water and air, the flux F of CO2 from gaseous phase to the liquid phase will be calculated as[7] F = KGhPaKLhCw,

where KG is the transfer rate of CO2 from gaseous phase to liquid phase, KL the transfer rate of CO2 from liquid phase to gaseous phase, Cw the concentration of CO2 in seawater, Pa the partial pressure of CO2 in the atmosphere. According to Henry’s law and the equilibrium conditions of gas on the gas-liquid interface, it is inferred that F = D × KL × Pa D × KL × Pw = D × KL ×'p, where D is the solubility of gas in liquid phase, Pw the partial pressure of CO2 in seawater. Therefore the exchange flux F can be calculated from measurement of 'p when the transfer rate KL of CO2 and solubility D are known. It has been shown by previous studies that the transfer rate KL is mainly related to wind velocity, D is the function of water temperature and salinity. Since KL cannot be directly determined, simulation experiments are usually used to develop the relationship between KL and wind velocity, yet different experiments give quite different KL values[7], and there has been no unified formula for KL calculation at present[3]. However, most simulated KL values fall in figures given by Liss et al.[17] and Tans et al.[18], accordingly, their results could be considered as the upper and lower limits of KL. The latest results of these parameters were given by Wanninkhof[19] which can be expressed as KL = K660 (660gSc1)0.5,

576

SCIENCE IN CHINA (Series D)

Vol. 46

2 where K660 = 0.31U N,10 and Sc = A  Bt + Ct2  Dt3.

The unit of KL is mgs1, UN,10 is the wind velocity at 10 m above the sea surface (mgs1), Sc is the ratio of dynamic viscosity to molecular diffusionˈa function of water temperature and equals 660 for seawater at 20ć (Sc20). For CO2 in seawater[19], A = 2073.1, B = 125.62, C = 3.6276, D = 0.043219. Wanninkhof indicated that the relationship between KL and wind velocity given by Liss et [17] al. can be shown as K660 = 0.17 U N2 ,10 . Jacobs et al.[20] provided the following formula based on their direct determination of KL from studies of marine aerosol and gas exchange of IGAC (International Global Atmospheric Chemistry): K660 = 0.54 U N2 ,10 . Thus for Wanninkhof method F = 0.31 × D × U N2 ,10 × (660gSc1)0.5×'p; for Liss method, F = 0.17× D × U N2 ,10 × (660gSc1)0.5×'p; for Jacobs method, F = 0.54 × D × U N2 ,10 × (660gSc1)0.5×'p. From

14

C method[7], the average residence time of CO2 molecule in atmosphere is about 7

years, and it was thus inferred that the air-sea exchange rate of CO2 was K =1.23 × 1010 molgcm2gPa1gmin1 (named as 14C method below). Though the 14C method did not consider short term environmental variation at the sampling stations, it took into account the average global status within seven years and thus excluded possible undue arguments appearing in experimental formulae when the surface wind velocity was zero or low, so this model has been used by many researchers in various global carbon cycle modeling studies. Our study adopted the 14C method[7] to estimate carbon flux for the expedition region. To facilitate comparing with results from other studies, estimations with methods of Liss et al.[17], Wanninkhof[19] and Jacobs et al.[20] (named as L method, W method and JA method, respectively) were also given. 3.2

The air-sea flux of CO2 observed for the Arctic Ocean Fig. 5 drawn with air-sea CO2 flux estimation for various stations shows the distribution of CO2 fluxes for the Arctic Ocean. Since only value of X is different in the formulae of air-sea flux calculation by L, W and JA methods (K660 = Xg U N2 ,10 ), these results suggested the same distribution trends, only contours were different, thereby, the result from W method is presented here. It can be seen from the figure that CO2 fluxes are high in northwest and low in the east of the sea, most areas are sinks of atmospheric CO2 and some sinks are strong, only Prudhoe Bay and the east areas to the Canada Basin are weak source.

No. 6

DISTRIBUTIONS & FLUXES OF SEA-AIR CO2 IN ARCTIC OCEAN

577

Fig. 5. Distribution of air-sea exchange flux of CO2 of the expedition region of Arctic Ocean (mgCO2gm2gh1).

In 1994, Takahashi and Feely modeled[15] with data available then the distribution of global average carbon fluxes, and they concluded that the flux for the Chukchi Sea was 1.0 mol CO2 · a-1 (same unit below), 2.38 times lower than the average flux of 2.38, which was calculated by experimental data in this study. Though the distribution trends from different calculation methods were similar, the numeric values were different. Table 1 summarizes their statistics. Table 1 Method Formula Unit Average Minimum Maximum Average/global average

Air-sea CO2 fluxes of the Arctic Ocean during the expedition Liss X = 0.17

6.57 12.15 58.66 2.55

Wanninkhof Jacobs X = 0.31 X = 0.54 F = X × KL × L × (P p) mg CO2 · m-2 · h-1 11.97 20.86 22.16 38.61 106.96 186.32 4.66 8.12

14 C KL = 1.23×10^10 mol/cm^2.pa.min F = KL × (P p)

26.32 74.68 88.41 10.24

It can be seen from table 1 that the results from different methods are basically of same magnitudes of 6.57ü26.32 mgCO2gm2gh1, among which the flux from Method 14C is the largest, flux from Liss Method the smallest being about 4 times greater in difference, while Jacobs Method and Method 14C gave similar number being twice greater difference. Method 14C produced an average flux of 26.32 mggCO2gm2 gh1, about twice as high as that for the Prydz Bay in the Antarctic[21, 22], 4.5 times that for Pacific Ocean, 10.2 times that of the globe[23]. 4

Conclusions

(1) Carbon cycle observation made during the First Chinese National Arctic Research Expedition filled the gap with the data of air-sea CO2 partial pressure for the Chukchi Sea and the adjacent sea areas in summer, which will make contributions to global carbon cycle studies. (2) The partial pressure of CO2 Pw at surface seawater in the expedition areas of the Arctic Ocean showed great variation and complicated spatial distributions in summer, basically low in

578

SCIENCE IN CHINA (Series D)

Vol. 46

west and north, high in east and south. Pw contents gradually decreased towards north of the Bering Strait and towards west from north part of Tuktoyaktuk with the minimum value of Pw <150 appearing at north sea areas of Point Lay, the maximum Pw > 400 near Tuktoyaktuk and an average of 242. The distribution pattern of Pw was mainly controlled by biological activities and the surface water circulation. (3) The distribution of difference between Pa and Pw, 'P (= Pa Pw), for the expedition region in summer was basically mirror symmetry with Pw, roughly increasing to the west in north part of Tuktoyaktuk, and to the north in the Bering Strait. Except for Prudhoe Bay and areas east to the Canada Basin and the right part of the Bering Strait, where 'P was below or equal to zero and could be source for atmospheric carbon dioxide, most regions could be sink or strong sink for atmospheric carbon dioxide because of their 'P˚0. (4) Estimation of the average carbon flux for the expedition region in summer ranged from 6.57 to 26.32 mgCO2gm2gh1, and generally this region was a strong sink for atmospheric CO2 in summer. The average carbon flux was about 2 to 10 times as high as global average flux or 2.38 times as many as the data given by model of Takahashi and Feely for this region when Wanninkhof coefficient was used in the calculation. References 1.

Hansen, J., Johnson, D., Lacis, A. et al., Climatic effects of atmospheric carbon dioxide, Science, 1983, 220:874—875.

2.

Gao, Z. Y., Chen, L. Q., Wang, W. Q., Study of carbon cycle in South and North Polar and the global change, References Frontiers (in Chinese), 2002ˈ9(1ü2): 263ü270.

3.

Zhang, Y. H., Wang, W. Q., Chen, L. Q., The progress of marine carbon dioxide study, Advance in Earth Science (in Chi-

4.

Chen, B. Q., Guo, Y. F., The study and progress of global climate change, Environmental Science (in Chinese), 1993, (4):

5.

Manabe, S., Tang, W., Fu, L L., Century-scale effects of increased atmospheric CO2 on the ocean-atmosphere system,

6.

Zhan, B. Q., Zhao, Y. P., CO2, oceans and climate, Oceanography and Limnology (in Chinese), 1989, 20(1): 92—99.

7.

Riley, J. P., Skelow, G., Chemical Oceanography, Vol. 2 (in Chinese), (eds. Cui, Q. C., Qian, Z. G., Tang, S. Q.), Beijing:

8.

Mcwhinnie, M. A., Polar Research: Its Present and Future (in Chinese), (translated by Gao, Y. X., Guo, J. L., An, N. et al.),

nese), 2000, 15(5): 559ü564. 14—23. Nature, 1993, 364: 215ü218.

Ocean Press, 1982, 1—211. Beijing: Ocean Press, 1981, 114—126. 9.

Keeling,C. D., Whorf,.T. P., Atmospheric Carbon Dioxide Record from Barrow in 1974ü2000, La Jolla: Scripps Institu-

10.

Keeling, C. D., Whorf, T.P., Wahlen, M. et al., Interannual extremes in the rate of rise of atmospheric carbon dioxide since

11.

Keeling, C.D., Chin, J.F.S., Whorf, T.P., Increased activity of northern vegetation inferred from atmospheric CO2 meas-

12.

John, J. J., Arctic carbon sink: Present and future global biogeochemical cycles, Prog. Oceanogr., 1989, 3(4): 393—411.

13.

Walsh, J. J., Mcroy, C. P., Coachman, L. K. et al., Carbon and nitrogen cycling within the Bering/Chukchi Seas: Source

tion of Oceanography, University of California, 92093—92144. 1980, Nature, 1995, 375: 666—670. urements, Nature, 1996, 382: 146—149.

regions for organic matter effecting AOU demands of the Arctic Ocean, Prog. Oceanogr., 1989, 22˖ 279—361.

No. 6

DISTRIBUTIONS & FLUXES OF SEA-AIR CO2 IN ARCTIC OCEAN

579

14.

First Chinese National Arctic Research Expedition Team, The Expedition Report of First Chinese National Arctic Re-

15.

Takahashi, T., Feely, R.A., Weiss, R.F. et al., Global air-sea flux of CO2:An estimate based on measurements of sea-air

16.

Chinese Arctic and Antarctic Administration, and

17.

Liss, P. S., Merlivat, Air-sea gas exchange ratesüIntroduction and synthesis, in the Role of Air-Sea Exchange in Geo-

18.

Tans, P. P., Conway, T. J., Nakazawa, T., Latitudinal distribution of the source and sinks of atmospheric carbon dioxide de-

search Expedition (in Chinese), Beijing: Ocean Press, 2000, 1—39, 93—105, 119—137, 155—191. pCO2 difference, Proc. Natl. Acad. Sci. USA, 1997, 94: 8292ü8299. Colloquium Paper National Center of Mapping and Survey of the Antarctic, Map of The

Arctic Region (in Chinese), Nuhan: Wuhan Technology University of Surveying and Mapping Press, 1998. chemical Cycling, Dordrecht: Reidel Pub., 1986, 113—127. rived from surface observation and an atmospheric transport model, J. Geophys. Res., 1989, 94: 5151—5173. 19.

Wanninkhof, R., Relationship between wind speed and gas exchange over the ocean, J. Geophys. Res., 1992ˈ97(C5):

20.

Jacobs, M. J., Kohoiek, W., Oost, W. A., Air-sea fluxes and transfer velocity of CO2 over the North Sea: Result from AS-

7373—7382. GAMAGE, Tellus, 1999, 51B: 629ü641. 21.

Wang, W. Q., Huang, X. B., Zhang, Y. H., Difference of air sea carbon dioxide partial pressure and the fluxes in South In-

22.

Zhang, Y. H., Wang, W. Q., Huang, X. B, Distribution of carbon dioxide in Pulitze Bay of South Polar and adjacent sea

dia Ocean, Taiwan Strait (in Chinese), 1998, 17(3): 262ü268. areas and its air-sea fluxes, Polar Study (in Chinese), 1997, 9(2): 158—162. 23.

Keeling, C. D., Carbon dioxide in surface ocean water, J. Geophys. Res., 1968, 73: 4543ü4553.