Measurements of Carbon Dioxide in the Seto Inland Sea of Japan

Transcription

1 Journal of Oceanography Vol. 49, pp. 559 to Measurements of Carbon Dioxide in the Seto Inland Sea of Japan EIJI YAMASHITA 1, FUKUICHI FUJIWARA 2, XIAOHU LIU 3 * and EIJI OHTAKI 4 1 Environmental Resources Research Center, Okayama University of Science, Okayama 700, Japan 2 Department of Public Health, City of Okayama, Okayama 700, Japan 3 Lanzhou Institute of Plateau Atmospheric Physics, Academia Sinica, Lanzhou, Gansu, P. R. China 4 College of Liberal Arts and Sciences, Okayama University, Okayama 700, Japan (Received 17 January 1993; in revised form 16 March 1993; accepted 5 April 1993) The carbon dioxide in seawater (pco 2 ) was measured in the Seto Inland Sea of Japan using newly developed equilibrator instrument designed to be free from the correction for addition or extraction of the carbon dioxide from carrier gas. The temperature dependence of pco 2 was about 4.5%pCO 2 / C for a single seawater sample which was processed as free from biological activity and change in total carbon dioxide content during an experiment. The decrease in pco 2 during daylight hours due to the photosynthetic fixation was about 30% of the daily mean of pco 2 during warm months and about 15% during cold months. The effect of carbon dioxide exchange between air and seawater on pco 2 was about 0.6 ppm in August and about 0.1 ppm in March. This is negligible small compared with the daily oscillation of carbon dioxide in seawater. 1. Introduction The carbon dioxide concentration in seawater plays an intrinsic role in the exchange of carbon dioxide between seawater and atmosphere. The equilibrator technique employing a nondispersive infrared gas analyzer (NDIR) has been used to measure the carbon dioxide concentration in seawater by many researchers (e.g., Takahashi, 1961; Gordon et al., 1971; Fushimi, 1987; Inoue et al., 1987; Oudot and Andrié, 1989). In traditional equilibrator techniques, seawater was continuously introduced into the equilibrator, and a constant volume of air was circulated in a closed circuit consisting of the equilibrator, air pump, chemical desiccant and NDIR. As noted by Takahashi (1961) in his pioneering work, if the air circuit is closed, the NDIR reading may be suffered by the pressure of the circulating air. Modifications were introduced in recent equilibrator techniques in order to avoid fault mentioned above (e.g., Wong and Chan, 1991). However, the equilibrator technique still needs a large amount of sample seawater to measure the carbon dioxide concentration. We developed a new type of equilibrator instrument for carbon dioxide measurement in seawater. Sample seawater of 100 ml is enough to determine the carbon dioxide by the present instrument. A measuring principle and measuring procedure of the instrument have been described in the previous paper (Ohtaki et al., 1993). The instrument has been applied to measure carbon dioxide in the Seto Inland Sea of Japan. In the present paper, the effects of temperature, biological activity and gas exchange across the air-sea interface on carbon dioxide in seawater are discussed. *Present address: College of Liberal Arts and Sciences, Okayama University, Okayama 700, Japan.

2 560 E. Yamashita et al. 2. Instrument A schematic diagram of the instrument used for the determination of carbon dioxide in seawater (pco 2 ) is shown in Fig. 1. The apparatus is consisting of a plastic chamber (equilibrator), a water bath, an infrared gas analyzer (NDIR) and recorder. The equilibrator has a cross section of 3 cm 2 and 15 cm tall. The equilibrator and sample bottle of seawater were kept in the water bath whose temperature was controlled by a regulator within an accuracy of 0.1 C to the seawater temperature under field conditions. A known concentration of carbon dioxide standard gas was continuously passed through a reference cell of the NDIR at a rate of 5 ml per min. Other carbon dioxide standard gases were passed through a measuring cell of the NDIR as a carrier gas. The carrier gas was flushed through a drying column with Mg(ClO 4 ) 2 at the final stage of the carrier gas circuit. The outlet of measuring cell and reference cell was opened to the atmosphere to maintain the carrier gas line at a barometric pressure. To check the precision of present instrument, consecutive measurements of pco 2 were carried out using sample seawater. pco 2, seawater temperature, and ph values measured are summarized in Table 1. Though the deviation of 3 ppm is obtained in the first group on August 20, results show that the present instrument is applicable to measure pco 2. It is also important to examine pressure broadning effects of two gas mixtures for seawater. The pco 2 was measured eight times using gas mixtures of CO 2 /N 2 (201.0, 299.6, 400.0, 501.0, and ppm) and CO 2 /air (201.0, 299.9, 396.0, 501.0, and ppm). Mean values of eight trials were ppm for CO 2 / Fig. 1. Schematic diagram of measuring instrument of pco 2.

3 Measurements of Carbon Dioxide in the Seto Inland Sea of Japan 561 Table 1. Precision of the instrument. pco 2 is measured using single seawater sample which is controlled at constant temperature and ph. Date Seawater temperature ( C) ph pco 2 (ppm) Aug Aug Aug N 2 mixtures and ppm for CO 2 /air mixtures. This result also denotes that the present instrument is adequate to measure pco 2 in seawater. The standard gas of CO 2 /N 2 mixtures was used as a carrier gas in the present study. 3. Time Variation of pco 2 in Seawater The pco 2 was measured as well as seawater temperature (Ts), ph, dissolved oxygen (DO) and salinity (S) at the Observatory for Environmental Research of Okayama University ( E, N) in the Seto Inland Sea of Japan. The water depth was about 7 m at the measuring site. The carbon dioxide concentration in the atmosphere (PCO 2 ) was also measured on the top of measuring tower, 10 m above the sea surface, at the Observatory. Figure 2 shows a time variation of pco 2 measured on May 22 to 23, It is apparent that the pco 2 shows well defined diurnal variation characterized by the low values in the daylight hours, and high values in the nighttime. This daily variation in pco 2 can be altered by the carbon dioxide exchange across air and sea interface, seawater temperature, and biological consumption of carbon dioxide in seawater through photosynthesis. 3.1 Effect of carbon dioxide exchange across air-sea interface on pco 2 The exchange rate of carbon dioxide can be calculated using the working equation proposed by Andrié et al. (1986): CO 2 flux = 0.24αK(pCO 2 PCO 2 ). Here, α is the solubility of carbon dioxide in seawater computed using the expression given by Weiss (1974). K is the transfer velocity of carbon dioxide. We assume that the µatm unit in pco 2 is equivalent to the ppm unit, because the total pressure is taken to be 1 atm in the present study. The results of CO 2 exchange rate on May 22 to 23 are given in Table 2. The CO 2 flux is a weak and its mean value is mol m 2 d 1. The negative sign means the downward transport of carbon dioxide from the atmosphere into sea water. Taking into account the mean

4 562 E. Yamashita et al. Fig. 2. Example of diurnal variation of pco 2 measured on May 22/23, Carbon dioxide in the atmosphere, PCO 2 seawater temperature, Ts, and dissolved oxygen, DO, are plotted for reference. Table 2. Carbon dioxide flux across the air and sea interface. Ts = seawater temperature, U = wind speed at 10 m height above sea surface, K = transfer velocity of carbon dioxide. Date Time Ts C U pco 2 ms 1 ppm PCO 2 ppm K cmh 1 CO 2 flux m-mol m 2 d 1 May May mean 0.23 depth of seawater (about 7 m) at the measuring site, we can estimate that the carbon dioxide of about mol kg 1 per day enters into seawater. This results in the increase of pco 2 of about 0.04 ppm per day in the present case. Here, the Revelle factor is assumed to be 8.5 (Oudot and Andrié, 1989) in the calculation. The pco 2 shows a seasonal variation with high values in summer months and low values in winter months (Ohtaki et al., 1993). It is also noted that the pco 2 shows lower values than PCO 2 from December to May, transporting carbon dioxide from the atmosphere to the sea surface, and higher values than PCO 2 from June to November, transporting carbon dioxide from the sea surface to the atmosphere. Sample calculations showed the downward flux of about mol m 2 d 1 in March and the upward flux of about mol m 2 d 1 in August. These results also show the decrease of pco 2 of about 0.6 ppm per day in August and the increase of pco 2 of about 0.1 ppm per day in March. Taking into account results mentioned above, we can conclude that the correction term from carbon dioxide exchange across air-sea interface is negligible small compared with the daily oscillation of pco 2.

5 Measurements of Carbon Dioxide in the Seto Inland Sea of Japan Temperature dependence of pco 2 In recent years, the temperature coefficient of pco 2 reported seems to have a consensus value of 4.3%pCO 2 / C (e.g., Peng et al., 1987). This is obtained under the assumption that the total inorganic carbon and the alkalinity are held constant. In order to examine the temperature coefficient of pco 2, seawater was sampled by passing through a filter with 0.3 µm mesh. The biological effect on pco 2 can be reduced by this filtering procedure. The sample bottle of seawater (cf. Fig. 1) was also sealed to reduce the carbon dioxide exchange across air-water interface during an experiment. The pco 2 was measured when the temperature of a single seawater sample was varied. Figure 3 gives the pco 2 as a function of seawater temperature, Ts. Though temperature range is limited within narrow ranges, the pco 2 increases with increasing temperature. The temperature coefficient of pco 2 is about 4.5%pCO 2 / C. This temperature coefficient is quite similar to the consensus value of 4.3%pCO 2 / C mentioned above. It is interesting to see how the temperature coefficient is modified for natural seawater in which the total inorganic carbon and the alkalinity alter. In the present study, seawater was sampled from two levels of 0.5 and 2.0 m under the sea surface. 47 data of pco 2 from 0.5 and 2 m levels were obtained from the present set of experiments. They show good agreement with each other and their regression line can be expressed by the equation: pco 2 (2 m) = 0.98pCO 2 (0.5 m) (r 2 = 0.97). Thus, the pco 2 data from 0.5 and 2 m are used to examine the temperature dependence of pco 2. Figure 4 shows changes in pco 2 as a function of seawater temperature, Ts. Though plotted points show scatter, the functional relationship between pco 2 and Ts is approximated as pco 2 = 141.1exp(0.0572Ts). It is noted that the temperature dependence of pco 2 is about 5.7%pCO 2 / C (about 18 ppm/ C at 300 ppm). This result implies that a general trend of pco 2 can be approximated by a simple seawater temperature, though the effects of salinity and ph have to be taken into account for the precise evaluation of pco 2. Fig. 3. Effect of temperature on pco 2 in a single water sample.

6 564 E. Yamashita et al. Fig. 4. Relationship between pco 2 and seawater temperature, Ts. Data are collected from following periods. : May 22/23, 1990, : July 23/24, 1990, : November 6/7, 1990, : March 2/3, 1991, : August 27/28, Fig. 5. Diurnal variation of pco 2 reduced to a constant temperature using temperature coefficient of 4.5%pCO 2 / C. Plotted values mean deviations from daily mean value of pco 2. Horizontal axis is scaled to express consecutive two days. : May 22/23, 1990, : March 2/3, 1991, : August 27/28, Biological effect on pco 2 In order to examine the biological effect, pco 2 data measured were processed to a constant temperature using the temperature coefficient of 4.5%pCO 2 / C mentioned in the previous section. Results are shown in Fig. 5 using data obtained from periods in low seawater temperature (March 2 to 3, 1991), in moderate seawater temperature (May 22 to 23, 1990) and in high seawater

7 Measurements of Carbon Dioxide in the Seto Inland Sea of Japan 565 temperature (August 27 to 28, 1991). In order to see clearly the biological effect, plotted data were produced by subtracting the daily mean values from the pco 2. Though the amplitude in pco 2 varies with periods measured, the pco 2 shows well defined diurnal oscillation with high values in the early morning around 6 h and low values in the afternoon around 16 h. pco 2, defined by the difference between maximum value and minimum one of pco 2, is 32 ppm for March, 103 ppm for May and 188 ppm for August. This corresponds to the ratio of pco 2 / pco 2 being about 0.15 for March, and 0.30 for May and August. Here, pco 2 is the daily mean of pco 2. These large decreases in pco 2 during daylight hours are associated with active photosynthesis by the phytoplankton in seawater, because the present experiment has been carried out at the most productive area in the Inland Sea. As the example of open oceans, Oudot and Andrié (1989) showed recently that the daytime decrease of about 4.5 ppm was occurred most of the time in eastern tropical Atlantic Ocean. This results in the value of pco 2 / pco 2 = The change in pco 2 results in the change in total dissolved carbon dioxide CO 2 (=CO 2 + HCO 3 + CO 2 3 ) in seawater. In the present study, the CO 2 was not measured. However, the CO 2, HCO 3 and CO 2 3 can be estimated from pco 2, ph, and first and second dissociation constants of carbonic acid (K 1 and K 2 ) in seawater as: CO 2 = α pco 2, HCO 3 = α pco 2 K 1 /H +, CO 3 2 = α pco 2 K 1 K 2 /(H + ) 2. Here, α is the solubility of carbon dioxide in seawater. H + is the activity of hydrogen ion defined by 10 ph. Values of K 1 and K 2 are computed using expression proposed by the CO 2 Subpanel of the JPOTS (Copin-Montegut, 1988). Again, we assume that the µatm unit in pco 2 is equivalent to the ppm unit. Values of CO 2, HCO 3, CO 3 2 and CO 2 calculated are tabulated in the Appendix. Values of CO 2 are plotted in Fig. 6 as a function of dissolved oxygen, DO. Though Fig. 6. Relationship between total carbon CO 2 and dissolved oxygen DO. Notations are the same as in Fig. 4.

8 566 E. Yamashita et al. there is a relatively large scatter in plotted points, the linear regression line for whole data shows the slope DO/ CO 2 = 1.3. This is quite similar to the slope expected from idealized marine photosynthesis (e.g., Baes et al., 1985). We have to accumulate similar data to deduce a significant relationship between CO 2 and DO. 4. Conclusions The pco 2 in the Seto Inland Sea of Japan was measured by a newly developed instrument which was designed to be free from the pressure effect of the carrier gas and the exchange of carbon dioxide between seawater and carrier gas. The primary results obtained in the present study are summarized as follows. 1) The effect of carbon dioxide exchange across air-sea interface on pco 2 is estimated to be about 0.04 ppm for data on May 22/23, about 0.6 ppm on August 27/28 and about 0.1 ppm on March 2/3. These correction terms are negligible small compared with the daily oscillation of pco 2. 2) The temperature coefficient of pco 2 is about 4.5%pCO 2 / C for a single seawater sample which is processed as free from biological activity and change in total carbon dioxide content of seawater during an experiment. 3) After correction of the temperature effect, the pco 2 shows well defined diurnal variation with high values in the early morning around 6 h and low values in the afternoon around 16 h. This diurnal oscillation is associated with the photosynthetic fixation by phytoplanktons. The decrease in pco 2 during daytime is about 30% of the daily mean of pco 2 during warm months and about 15% during cold months. It is noted that there is a possibility to have the ratio, DO/ CO 2 = 1.3, which is expected from the idealized marine photosynthesis. Similar work is required to deduce a significant relationship between DO and CO 2. Acknowledgements We wish to express our sincere thanks to Messrs. H. Shibata, T. Inoue, K. Watanabe, K. Masunaga and Ms. H. Omukai for their kind help in field observations. This work was partly supported by the Scientific Research on Priority Areas (No ) of The Ministry of Education, Science and Culture. Appendix (see pp ) References Andrié, C., C. Oudot, C. Genthon and L. Merlivat (1986): CO 2 fluxes in the tropical Atlantic during FOCAL cruises. J. Geophys. Res., 91, Baes, C. F., A. Björkström and P. J. Mulholland (1985): Uptake of carbon dioxide by the oceans. p In Atmospheric Carbon Dioxide and the Global Carbon Cycle, ed. by J. R. Trabalka, U.S. Department of Energy, Washington. Copin-Montegut, C. (1988): A new formula for the effects of temperature on the partial pressure of CO 2 in seawater. Mar. Chem., 25, Fushimi, K. (1987): Variation of carbon dioxide partial pressure in the western North Pacific surface water during the 1982/83 El Nino event. Tellus, 39B, Gordon, L. I., P. K. Park, S. W. Hager and T. R. Parsons (1971): Carbon dioxide partial pressures in north Pacific surface waters Time variations. J. Oceanogr. Soc. Japan, 27, Inoue, H., Y. Sugimura and K. Fushimi (1987): pco 2 and δ 13 C in the air and surface sea water in the western North Pacific. Tellus, 39B,

9 Measurements of Carbon Dioxide in the Seto Inland Sea of Japan 567 Ohtaki, E., E. Yamashita and F. Fujiwara (1993): Carbon dioxide in surface seawaters of the Seto Inland Sea, Japan. J. Oceanogr., 49, Oudot, C. and C. Andrié (1989): Short-term changes in the partial pressure of CO 2 in eastern tropical Atlantic surface seawater and in atmospheric CO 2 mole fraction. Tellus, 41B, Peng, T.-H., T. Takahashi, W. S. Broecker and J. Olafsson (1987): Seasonal variability of carbon dioxide, nutrients and oxygen in the northern North Atlantic surface water: observation and a model. Tellus, 39B, Takahashi, T. (1961): Carbon dioxide in the atmosphere and in Atlantic ocean water. J. Geophys. Res., 66, Weiss, R. F. (1974): Carbon dioxide in water and seawater: The solubility of a non-ideal gas. Mar. Chem., 2, Wong, C. S. and Y.-H. Chan (1991): Temporal variations in the partial pressure and flux of CO 2 at ocean station P in the subarctic northeast Pacific Ocean. Tellus, 43B,

10 568 E. Yamashita et al. Table A1. pco2 and carbonate species calculated from pco2, ph, and first and second dissociation constants of carbonic acid for seawater sampled at 0.5 m depth under sea surface. pco2 = carbon dioxide in seawater, Ts = seawater temperature, S = salinity, DO = dissolved oxygen, CO2 = CO2 + HCO3 + CO3 2. Date Time Ts C ph S pco 2 ppm DO CO 2 HCO 3 CO 3 µmolkg 1 µmolkg 1 2 CO 2 µmolkg May *** May July July *** *** Nov Nov

11 Measurements of Carbon Dioxide in the Seto Inland Sea of Japan 569 Table A1. (continued). Date Time Ts C ph S pco 2 ppm DO CO 2 HCO 3 CO 3 µmolkg 1 µmolkg 1 2 CO 2 µmolkg Mar Mar *** Aug Aug