SEARCH

SEARCH BY CITATION

Keywords:

  • CO2;
  • Southern Ocean;
  • tracers;
  • air-sea gas exchange

Abstract

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[1] Gas transfer velocities were determined in the Southern Ocean during the Southern Ocean Iron Fertilization experiment (SOFex) using the dual deliberate tracer technique. The decrease of the purposefully injected tracers, sulfur hexafluoride and helium-3, could be well described by gas exchange parameterizations with wind speed that satisfy global constraints based on bomb-14C uptake. The concentration decrease of tracers could be predicted slightly better with established relationships if gas transfer was modeled as a function of the cube rather than the square of the wind speed, particularly over a time interval with high and variable winds. However, both fits can model the concentration decrease within the uncertainty of the observations. This suggests that it will be singularly difficult to definitively determine if a quadratic or cubic dependence of gas exchange with wind is more appropriate based on deliberate tracer measurements. However, these results show that gas exchange rates in the Southern Ocean are not anomalous when compared with the rest of the ocean. Thus this cannot account for discrepancy between observational and model-based estimates of uptake of CO2 in the Southern Ocean. Using a high-quality wind speed field obtained from the QuikSCAT satellite Seawinds scatterometer and an established surface water pCO2 climatology, the CO2 uptake in the Southern Ocean (>34°S) is reassessed. The total uptake rates are similar to previous observation-based estimates, but the analysis shows that the uptake rate is sensitive to wind speed product used and the wind speed distribution.

1. Introduction

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[2] Accurate quantification of air-sea gas transfer rates is necessary to constrain fluxes of climate relevant gases on short timescales. To a large extent, our knowledge of gas fluxes is limited by a dearth of field observations. Recent advances in measuring gas fluxes by micrometeorological approaches promise to greatly increase the number of observations and improve our quantitative understanding of the mechanisms controlling air-sea fluxes [Fairall et al., 2000; McGillis et al., 2001]. However, gas exchange measurements based on inventory changes of deliberate tracers remain a powerful technique, as they are a robust integrative method to estimate gas transfer over a period of days. Several gas transfer studies using the dual deliberate tracer technique with helium-3 (3He) and sulfur hexafluoride (SF6) have been performed in various ocean environments [Nightingale et al., 2000a; Wanninkhof et al., 1993, 1997; Watson et al., 1991b]. A synthesis of the dual tracer gas exchange measurements was prepared by Nightingale et al. [2000b], who showed a strong relationship of the derived gas transfer velocities with wind speed. The coefficients of the second-order polynomial relationship developed by Nightingale et al. [2000b] were determined by a least squares fit through the data. The relationship is applicable for the time period over which the gas transfer velocities are determined (1–4 days). Because of the nonlinear nature of the relationship of gas exchange with wind speed, variability of winds will affect the absolute gas transfer value, in addition to the magnitude of the wind. Since the variability around a mean wind speed differs, there inherently will be scatter in gas transfer velocities plotted against time-averaged wind.

[3] No gas transfer velocity measurements have been published in the Southern Ocean to date. This region encompasses roughly 30% of the world's ocean but is undersampled compared with other ocean basins. From an air-sea gas exchange perspective, the particular issue that warrants further investigation in this region is the significant mismatch between air-sea CO2 fluxes determined from gas transfer velocities and the air-water partial pressure difference of CO2 (ΔpCO2), and those inferred from atmospheric observations and models [Gurney et al., 2002; Keeling et al., 1996] and ocean inverse models [Gloor et al., 2003]. The observations suggest an uptake at latitudes greater than 34°S of 1.4–2.5 Pg C per year using the pCO2 climatology of Takahashi et al. [2002] and the quadratic or cubic wind speed formulations of Wanninkhof [1992] and Wanninkhof and McGillis [1999]. Estimates based on atmospheric and ocean models yield uptakes of <1.2 Pg C per year. The difference between gas fluxes inferred from models and estimates based on pCO2 observations in the Southern Ocean are significantly greater than for other basins. Lack of ΔpCO2 data in the Southern Ocean has been implicated as a possible reason for the large differences between ocean and atmospheric derived estimates. However, biases in parameterization of gas exchange and wind speed are another possible cause.

[4] Parameterization of gas exchange with wind speed is problematic since wind is an important factor but not the only forcing function for gas transfer. This is, in part, the reason that different parameterizations yield roughly a factor of 2 difference in gas transfer values over the oceans. Parameterizations with other parameters have been proposed, including those with whitecap coverage [Monahan, 2002], wave slope, and friction velocity [Jähne et al., 1987]. Efforts have also been directed at relating gas transfer to wave slopes inferred from satellite altimeters [Glover et al., 2002] with a focus on more fully utilizing remote sensing products. An important objective of these parameterizations is to use the daily global fields that can be obtained from scatterometers and radiometers in orbit to determine gas transfer velocities on a global scale. In particular, synoptic global coverage of QuikSCAT greatly increases the fidelity of global ocean flux estimates as shown below. Many investigations continue to rely on wind as a parameter to infer gas transfer velocities; therefore, continued improvement of such relationships is warranted.

[5] There are several possible causes for biases in air-sea gas fluxes in the Southern Ocean. The Southern Ocean environment is characterized by high winds and very long fetches in the downwind direction of prevailing westerly winds. This causes an environment with sustained significant wave heights and large swells. One hypothesis is that the swells lead to sheltering effects and decreased gas transfer on the lee side of the waves. This could lead to weaker increases of gas exchange with wind when compared with other environments. Another possibility is that current parameterizations of gas exchange with wind speed do not adequately account for bubble mediated gas transfer. In general, because of the lack of gas exchange measurements in this region, observational constraints will improve our confidence in the calculated fluxes.

[6] The dual deliberate tracer study was performed as part of the Southern Ocean fertilization experiment (SOFex) [Coale et al., 2004]. The major objective in SOFex was to determine the biological and chemical response to a deliberate infusion of iron to the surface ocean. In order to unequivocally determine the water in which the iron was injected, an inert chemical tracer, SF6, was added [Watson et al., 1991a]. An opportunistic project in this effort was to simultaneously inject a small amount of the light isotope of helium, 3He, along with the SF6 to quantify the air-sea gas exchange. Here we present the methodology for injection and analysis followed by the description of the gas exchange calculations. The results are elaborated upon in terms of gas exchange-wind speed relationships using QuikSCAT winds that are applied to the recent climatological pCO2 data for the Southern Ocean to estimate the Southern Ocean CO2 sink. On the basis of these results, the Southern Ocean sink is at the lower end of the range described by Takahashi et al. [2002].

2. Methods

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[7] During the SOFex study, a mixture of iron and SF6 was infused at a location north of the Polar Front and a location south of the Polar Front from the research vessel Revelle. In the southern injection, 3He was added to the injection mixture to perform the dual deliberate tracer method of determining air-sea gas exchange. The injection procedures were similar to those described by Upstill-Goddard et al. [1991]. It involved saturating two 4800-L containers of seawater each topped with a chamber enclosing 1 L of headspace. The headspace was recirculated to an aerator at the bottom of the tank. Pure SF6 was bled into the recirculation loop at a rate of 200 mL min−1 over a period of about 30 hours, and the excess gas was vented from the 1-L chamber through a ball flowmeter. The approach to saturation was monitored by analyzing aliquots of water from the tank by equilibration of a 40-mL water sample with a 10-mL nitrogen headspace in a syringe followed by analysis with a thermal conductivity detector (TCD). Once the SF6 had reached saturation in the tanks, 6 L of 85% 3He, with balance 4He, was added over a period of 1 hour to one of the tanks, in the same fashion as the SF6, until the 3He levels in water reached constant values, as indicated by TCD analyses of the water aliquots measured in the same fashion as the SF6.

[8] The outflow of the spiked tanks was joined and metered at a rate of 3.4 L min−1 into the ocean at about 3-m depth at the edge of the ship's wake about 20 m behind the ship. The iron sulfate solution was fed into the ocean on the opposite side of the wake. A regular “lawnmower pattern” with nominal 0.5 km spacing was followed during the injection, covering an area of about 15 km by 15 km (Figure 1) over 42 hours. On the basis of the volume of spiked water released and the TCD analyses of SF6 and 3He concentrations in the tank, we estimate that approximately 44 L of SF6 and 4.6 L 3He (STP) were injected into the mixed layer.

image

Figure 1. Contour plot of SF6 concentrations measured during the first 2.5 days after injection. The solid circles show the track occupied during the SF6, 3He, and iron sulfate injection. The plus symbols show the locations of surface underway SF6 measurements. The SF6 patch and cruise tracks are shown in Cartesian coordinates.

Download figure to PowerPoint

[9] Surface water and bottle samples from hydrocasts were analyzed for SF6 with different instruments. Discrete analyses from casts were run on an instrument patterned after Law et al. [1998]. In short, samples were taken in 550-mL borosilicate glass bottles with glass stoppers and about 264 mL of sample was sucked into an evacuated chamber through a showerhead. The total volume dispensed was reproducible to within 1 mL using a liquid level sensor near the top of the chamber that triggered the intake valve to shut. The SF6 that entered the headspace during filling and the remaining SF6 dissolved in the water was stripped with ultra high purity nitrogen onto a Carboxen 1000 trap held at −55°C. The trap was isolated and heated to 150°C, after which the gases were swept onto a 1.5 m × 0.3 cm OD molecular sieve 5A column where SF6 was separated from oxygen and other gases and measured with an electron capture detector (ECD). Total extraction and analysis time was about 10 min. The detector was calibrated with five standards with concentrations of 5.7, 55.1, 112, 167, and 345 pptv. Reproducibility of the measurements based on 30 sets of duplicate samples in the mixed layer taken from the same Niskin bottle in the tracer patch was 0.8 ± 0.6%.

[10] The underway SF6 system was used as the primary means to map the iron fertilized patch. Water from the ship's intake system, at nominally 5 m depth at the bow, was pumped through a 40-micron filter and subsequently through a LiquiCel permeable membrane extractor. The LiquiCel consists of a bundle of gas permeable microtubes where the SF6 in the water stream is extracted through the membranes into a nitrogen stream that is diverted to a 1.5 mL sample loop. The loop is injected at 3-min intervals into a molecular sieve 5A column and analyzed with an ECD that was calibrated with the standard gases listed above. The efficiency of extraction varied over the course of the study due to changes in operating conditions, such as variations in air and water flow, and because of fouling of the gas permeable tubes. The efficiency was quantified by periodically taking aliquots of water before and after the LiquiCel and analyzing their SF6 content. Results were corrected for efficiency of extraction. The efficiency ranged from 25–50% during the surveys with slowly changing trends that were well quantified by measuring the gas and water flows through the LiquiCel and analyses of discrete aliquots. The underway sampling system experienced some lag and hysteresis effects caused largely by significant residence times of water in the seawater intake lines. On the basis of measurements of temperature at the bow inlet and at the LiquiCel, water residence time in the intake lines was estimated at 3 to 5 min. Combined with the 3-min analysis time, underway analysis results were delayed by about 6 min from the time the water entered the bow intake. For typical survey speeds of 10 km/hour, this corresponds to spatial offsets of 1 km between water sampled and analysis. Hysteresis effects were more difficult to quantify but led to more slowly decreasing apparent concentrations when steaming out of the patch, particularly along sharp boundaries. The offsets and hysteresis effects were not quantitatively accounted for because dispersion in the seawater line and retention in the LiquiCel were not properly characterized during the study.

[11] The samples for helium analysis were drawn from the Niskin bottles by flushing 15 cm3 copper tubes with sample water. The tubes were sealed on each end with a pinch clamp to prevent exchange with the atmosphere. At the University of Miami noble gas laboratory, the dissolved gases were extracted and purified over a number of cold and charcoal traps. The helium fraction was then admitted into a mass spectrometer for isotopic analysis [Jenkins and Clarke, 1976]. The precision for both helium isotopes was ±0.5% based on the reproducibility of the air standards.

[12] The first Fe injection, along with the addition of SF6 and 3He, began at 0900 GMT on January 24, 2002 (year day, JD, 24.31) and continued for nearly 2 days until January 26, 2002 (JD 26.06). Immediately following the injection, the patch was surveyed for 2.5 days. A second Fe injection then took place but without the addition of SF6 or 3He. This sequence of injection and survey was repeated two more times for a total of four Fe infusions, four survey periods, but only a single addition of SF6 and 3He. The underway SF6 system operated from the first survey onward throughout the subsequent injection and survey periods. A total of 8 depth profiles were taken during the survey period with about 12 Niskin bottles sampled for SF6 and 10 for He per cast down to 100 m. Mixed layer depths were about 33 m during the first half of the investigation and increased to 42 m after a storm on JDs 30 and 31.

[13] Contour plots of SF6 survey results with ship track and locations of sampling stations overlain are shown in Figure 2. The contours were mapped in an advective (Lagrangian) frame by plotting the ship's position relative to a drogued drifter that was deployed in the patch. The drifter did not follow the patch exactly and moved out of the patch over a period of days. It was repositioned twice near what was believed to be the middle of the tracer patch during the study. All station locations used for the gas exchange analysis were well within the patch.

image

Figure 2. Three manifestations of the SF6 patch: (a) A contour plot from observations between JD 26.1 and 28.5; (b) from JD 30.3 to 32.4; and (c) from JD 33.5 to 36.1 with spurious contours removed from the survey tracked blanked out. The sample points are depicted with a plus symbol. The plots are in a Lagrangian reference frame relative to a drogued drifter. The stations at which depth profiles were obtained are marked as numbered circles. P. Strutton, SUNY Stony Brook, supplied the locations relative to the drifter. Note that Figures 1 and 2a are the same data depicted in Cartesian and Lagrangian reference frame, respectively.

Download figure to PowerPoint

image

Figure 2. (continued)

Download figure to PowerPoint

image

Figure 2. (continued)

Download figure to PowerPoint

[14] The cruise objectives and SF6 survey procedures were not optimized to quantify the total mass of SF6 in the water column through time. Nevertheless, the cruise track suggests that mapping was reasonably comprehensive. A mass balance estimate based on integrating contour plots (Figure 2) shows that a reasonable mass balance of the tracer could be established for the surveys. An estimated 44 L of SF6 and 4.6 L of 3He (STP) was added to the mixed layer. A volume integral of the first survey initiated immediately after injection (Figure 2a) yielded 39 L of SF6, and the following two surveys yielded inventories of 31 L SF6 and 19 L SF6, respectively. The decrease in SF6 between the injection and subsequent surveys is in accord with the estimated gas loss through the interface. The mass balance suggests that the tracer patch remained intact and that much of it was sampled during the study. A further check is that the ratio of 3He and SF6 in the injection mixture is 1:9.6 (volume: volume). The initial ratio for the first stations sampled was 1:11, suggesting reasonable consistency between the total amount and ratios of gases injected with those subsequently observed in the water column.

3. Discussion

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

3.1. Determination of Gas Transfer Velocities

[15] Air-sea gas transfer rates were determined from the change in ratio of 3He/SF6 over time. This method takes advantage of the large difference in diffusion coefficients between SF6 and 3He. The concentration of the 3He and SF6 will decrease over time due to dispersion and air-sea gas transfer. Dispersion will decrease the concentration of both gases at the same rate, while the 3He will be lost by gas exchange approximately 3 times faster than SF6 because of the diffusivity dependence of air-water gas transfer. The relative rate of gas transfer over the air-water interface is

  • equation image

where image and image are the gas transfer velocities of 3He and SF6, respectively. Sc is the Schmidt number defined as the kinematic viscosity of seawater divided aqueous diffusion coefficient of the gas and n is the Schmidt number exponent which is −2/3 for smooth (calm) surfaces and −1/2 for wavy surfaces [Jähne et al., 1987]. This Sc dependency breaks down in situations with bubble entrainment. Studies comparing SF6 and He exchange in wave breaking simulation tanks suggest that the dependency becomes less negative under breaking wave conditions. Initial studies [Wanninkhof et al., 1995] suggested a rapid increase (less negative) in the exponent in the presence of breaking waves, but subsequent laboratory studies and theoretical arguments suggest that for He and SF6 the Schmidt number exponent increases by <20% under extreme breaking wave conditions [Asher and Wanninkhof, 1998; Woolf, 1997].

[16] The known relative dependency of 3He and SF6 transfer makes it possible to determine the absolute gas transfer velocity as shown by Wanninkhof et al. [1993],

  • equation image

where Δt is the time interval between t1 and t2, and Xt1, Xt2 are the ratio of 3He and SF6 concentrations in the water sample corrected for background concentrations at t1 and t2. In this formulation it is implicitly assumed that the concentration decreases due to dispersion and mixing are first-order processes, and that the surface layer is well mixed on the timescale of concentration decrease due to gas exchange (≈days). Dilution processes by mixing and dispersion can be assumed to be first-order processes if the samples are taken well within the tracer patch and if the dilution due to dispersion is of similar or smaller magnitude as loss due to gas transfer [Gulliver et al., 2002]. For the study in the southern patch of SOFex these conditions are fulfilled, except for the first few days when the streaks injected at 3-m depth were spreading vertically and laterally to form a single patch. The concentrations of SF6 in the mixed layer with depth are reasonably homogeneous based on vertical profiles. The loss due to gas exchange can account for all of the concentration decreases in the latter part of the study, suggesting limited lateral spreading.

[17] Four to six colocated 3He and SF6 samples were obtained in the mixed layer at each of the eight stations that were sampled. The concentration profiles for the five stations used in the analysis are shown in Figure 3. The gas transfer values were determined by averaging the ln(3He/SF6) values for the mixed layer samples for each station. The standard deviations of the 4 to 6 ln(3He/SF6) values varied from station to station but without clear correlation with mixed layer concentration or time (Figure 4). During the first 2 days of sampling immediately after the 3He and SF6 injection, concentration ratios did not decrease systematically with time, likely because of heterogeneity caused by mixing of the injection streaks and low wind speeds. In the first days after injection the streaks coalesced, and the dilution of SF6 was probably not a first-order process with respect to concentration decrease. The data from the stations sampled on the first 3 days were excluded from the analysis. During the following 8 days, the concentration ratio decreased with time, with larger drops in the ln(3He/SF6) occurring in stormy periods (Figure 4).

image

Figure 3. Depth profiles of SF6 (open circles in 10−15 mol/L) and 3He (solid circles in 10−15 cc/gr) for the five sampling stations used to determine the gas transfer velocity. The year day (YD) of measurements are listed on the figure header.

Download figure to PowerPoint

image

Figure 4. Average 3He/SF6 ratio for each station over the time period. The error bars are the standard deviations for the points in the depth profile and the line is a least squares fit through all points. The first three data points are not used in the analysis because the 3He/SF6 ratios are likely influenced by factors other than gas transfer alone.

Download figure to PowerPoint

[18] The calculated gas exchange velocities of 3He for each time interval after JD 28 were normalized to a Schmidt number of 660 (equivalent to that of CO2 in seawater at 20°C) and related to wind speeds obtained from the bow mast at 22 m above the water surface (u22) normalized to a height of 10 m (u10) using a constant drag coefficient of 1.2 × 10−3, yielding u10 = 0.93 u22. The results shown in Figure 5 as solid squares show a general trend of gas exchange with wind. Two frequently used relationships expressing gas transfer as a function of the square of the wind speed [Wanninkhof, 1992] and the cube of the wind speed [Wanninkhof and McGillis, 1999] are included for comparison. Two sets of curves are shown, those applicable to steady (or short-term) winds and those for climatological winds. Although the time interval (0.5 to 2 days) would suggest that comparison with the short-term relationships would be more applicable, the wind speed variability for each of the intervals was significant. The trend of the data favoring a strong (≈u3) dependence of wind might be, in part, related to the k at 9.8 m s−1 falling below the trend. Since each k value was determined from ln(3He/SF6) measurements at adjacent times, any anomalously low k value would be followed by an anomalously high k value. Thus if ln(3He/SF6) values for JD 30.23 (station 25) are anomalously high, this biases the k value for 9.8 m s−1 low and the value for 11.3 m s−1 high. However, no artifacts in sampling or analysis of samples from station 25 are apparent. This station profile did show a mid-depth maximum in SF6 (Figure 3), but the ln(3He/SF6) values were constant in the mixed layer.

image

Figure 5. Gas transfer velocities determined for the study plotted against wind speed. For comparison, the long and short-term relationships from Wanninkhof [1992] (shaded dashed and solid lines, respectively) and Wanninkhof and McGillis [1999] (black dashed and solid lines, respectively) are shown. The solid squares are the observed values, kobs, while the plus symbols image and open squares image are the inferred values for steady winds assuming a quadratic and cubic relationship, respectively. Average winds are determined from ship-based observations normalized to 10-m height. Gas transfer velocities are normalized to a Schmidt number of 660. The error bars are the standard deviations of the observed values based on the uncertainty in the 3He and SF6 ratios shown in Figure 4.

Download figure to PowerPoint

[19] The unique wind speed distributions for each time interval can account for some of the scatter of k versus u10 in Figure 5. Nonlinear dependencies yield higher gas transfer values for periods with greater wind speed variability. For each time period, this effect was assessed by determining k for instantaneous or steady winds, kinst, assuming a particular relationship between gas exchange and wind,

  • equation image

R is the enhancement caused by the variability in the winds and is expressed as R = equation imageu10avx where Σ(u10x) is the average of the sum of the 10-min wind speeds squared (x = 2) or cubed (x = 3). The value for ax is 0.34 when x = 2 and 0.0277 when x = 3. The values for R for each time period are shown in Table 1. They range from 1.12 to 1.24 for x = 2 and from 1.35 to 1.74 for x = 3. This implies that the variability in the wind “enhances” the gas exchange by 12–24% over each time period over which k is determined, compared with a steady wind situation if a quadratic dependency is assumed. A cubic dependency of k with wind leads to enhancements of 35–74%. The resulting kinst values that account for variability are depicted in Figure 5 as crosses and open squares for quadratic and cubic dependencies, respectively. In each case, the kinst values are closer to the steady wind speed relationships than the unnormalized values.

Table 1. Summary of Wind and Gas Transfer Values for the SOFex Studya
Year Dayu10, m s−1equation imageuav−2equation imageuav−3k660, cm hr−1St. Dev., cm hr−1image cm hr−1image cm hr−1
  • a

    The values listed are for the preceding time interval. All gas exchange values, k, are normalized to a Schmidt number of 660. Here k660 is the average gas transfer velocity over the time period, while image and image are the inferred steady or instantaneous gas transfer velocities based on the wind speed distributions over the time interval using a quadratic or cubic dependence, respectively. The values listed on the last line are the averages for the entire study. The standard deviation for the intervals is determined from propagating the uncertainty in the ln(3He/SF6) for each profile. The standard deviation for the entire period is estimated from the error in the slope of the linear fit through the four points.

26.27       
27.063.9      
27.197.1      
28.106.3      
30.249.81.121.3527.717.724.720.5
31.4211.31.241.6666.125.753.339.7
35.307.81.241.7923.912.319.413.4
36.159.31.121.3544.1113.339.432.7
28.10–36.158.81.241.7631.63.225.518.0

[20] To quantitatively determine the optimal coefficient for a quadratic or cubic dependency, a least squares difference was determined from the observed ln(3He/SF6) decrease over each time interval versus the decrease expected using a quadratic or cubic dependency with 10-min averaged winds obtained from the ship's anemometer,

  • equation image
  • equation image

[21] The results shown in Figure 5 are weighted linearly with concentration decrease according to [ln(3He/SF6)obs–ln(3He/SF6)modeled]2 Δln(3He/SF6)obs to assign a greater weight for time intervals with greater concentration decrease. The minimum in the weighted squares of the difference is 0.34 and 0.0277 for a2 and a3, respectively. The coefficient of 0.34 is higher than the coefficient of 0.31 for the quadratic dependency proposed by Wanninkhof [1992], while the coefficient of 0.0277 is similar to the coefficient of 0.0283 suggested for the cubic dependency by Wanninkhof and McGillis [1999]. The minima in the weighted least squares difference is similar for the cubic and the quadratic dependencies, suggesting that either fit is applicable.

[22] To further investigate if a quadratic or cubic dependence yields a better fit to the data, we focus on the time period with high winds when the difference would be most pronounced. Very strong winds were experienced during the middle part of the study from JD 30.2 to JD 31.4 with maximum 10-min winds (U10) reaching over 20 m s−1. For this time interval there were also very low winds, further contrasting the dependencies. The optimal fit through the two points is accomplished for k = 0.403 u2 (660/367)0.5 and k = 0.02663 u3 (660/367)0.5. The coefficient for the cubic dependence is close to the coefficient for the best fit of the entire experiment, while the coefficient for the quadratic dependence is 10% higher. This suggests that a cubic dependence, which implies a very weak dependency at low wind speeds and a strong dependency at high wind speeds, is an appropriate way to model the concentration decreases for this study.

[23] The difficulty in discerning whether a quadratic or cubic functional dependence is more appropriate for the study can be well illustrated from the modeled decrease in ln(3He/SF6) compared with the measured mixed layer ln(3He/SF6) over the course of the experiment following the approach of Kuss et al. [2004]. The concentration trends are shown in Figure 6 using 10-min averaged wind speeds normalized to 10-m height using cubic and quadratic relationships with the appropriate proportionalities. The initial modeled ln(3He/SF6) ratio at the start of the experiment was adjusted such that the modeled values would equal the value at JD 28. Since the winds at the start of the study were low, little change in modeled ln(3He/SF6) was noted for the first 2 days with any of the parameterizations. The modeled decreases in ln(3He/SF6) are very similar, illustrating the challenge of providing unique parameterizations with the limited data sets available from deliberate tracer studies. The modeled trends clearly show the effect of varying winds with very sharp decreases during periods of high wind and invariant ratios during quiescent periods. The modeled trends show that by the end of the study the use of different coefficients and dependencies yields differences in 3He/SF6 of less than 10%, which is significantly smaller than the uncertainty in the measurements.

image

Figure 6. Modeled changes in ln(3He/SF6) over time using a quadratic and cubic dependence of gas exchange with wind speed. The squares are the average ln(3He/SF6) observed at each station. The relationships depicted with thin lines are the quadratic and cubic relationships of Wanninkhof [1992] and Wanninkhof and McGillis [1999], while the thick lines are the best fits for this study as outlined in the text. The relationships shown are k = 0.31 u102 (thin solid line); k = 0.34 u102 (thick solid line); k = 0.0283 u103 (thin dashed line); and k = 0.0277 u103 (thick dashed line). All model curves have been corrected to the gas exchange of helium by multiplying the coefficients by the inverse square root of the ratio of Schmidt numbers for He at in situ temperature and CO2 at 20°C: (357/660)−0.5.

Download figure to PowerPoint

[24] The SF6 losses due to air-sea gas exchange between injection and survey 1 (determined at the midpoint, JD 26.1), survey 2 (JD 30.3), and survey 3 (JD 33.5), based on the decrease in 3He/SF6 ratio, are 5.7, 24.2, and 29.5 L of SF6 (STP). When these values are added to the estimated SF6 remaining in the patch, based on a volume integral of SF6 measured on the surveys, this accounts for 102, 125, and 110% of the total SF6 injected. Within the overall uncertainty of the gas exchange and water column inventories this suggests that all the SF6 is accounted for during the study.

3.2. Air-Sea CO2 Flux Estimates for the Southern Ocean

[25] The magnitude of the Southern Ocean CO2 sink is much debated, with significant discrepancies between estimates based on atmospheric and oceanic inverse models, and oceanic estimates based on ΔpCO2 [Gloor et al., 2003; Gurney et al., 2002]. Using the gas transfer parameterization described here, along with QuikSCAT wind data and the pCO2 climatology of Takahashi et al. [2002], an estimate of Southern Ocean fluxes is made for the region >34°S from August 2001 through July 2002. This estimate differs from Takahashi et al. [2002] in that it is for a different year and uses a different wind speed product, which includes an estimate of the influence of wind speed variability. The similarity is that the same ΔpCO2 field is used. Although the optimal coefficients for the gas exchange relationships derived here are slightly different from those proposed by Wanninkhof [1992] and Wanninkhof and McGillis [1999], we chose to use the published relationships since they fall well within the uncertainty range of our results. The satellite QuikSCAT wind product from August 2001 through July 2002 is used and compared with the climatological NCEP winds for 1995 or the 41-year averaged NCEP winds used by Takahashi et al. [2002].

[26] The variability in the wind was assessed by determining the second and third moments of the wind product. The gas transfer velocities were determined according to

  • equation image

where u10av is the average wind speed, a = 0.31 when x = 2, and a = 0.0283 when x = 3. The R is the enhancement caused by the variability in the winds and is expressed as R = equation imageu10avx where equation image is the average of the sum of the wind speeds squared or cubed. Note that Σ(u10x)/n is the second or third moment if x = 2 or 3, respectively. For this analysis, monthly averaged winds over the 4° × 5° grid were determined along with the second and third moments of the wind using about 16 × 103 observations per pixel per month for the QuikSCAT product. The resulting gas transfer velocities were applied to the 960 K pCO2 climatology of Takahashi et al. [2002], which is an update based on more data compared with the original study presented by Takahashi et al. [1997].

[27] To assess the agreement of wind products, the average and moments of the QuikSCAT winds are compared with the ship-based winds for the 2-week study period. For the entire year, a blended product of QuikSCAT and special sensor microwave/imager (SMM/I) data are compared with the QuikSCAT record. The average, second, and third moments of the wind are determined from each satellite overpass for each 4° × 5° grid. Any particular point on the surface is sampled about twice a day with the QuikSCAT sensor. The large number of observations per month (≈1.6 104) for the QuikSCAT product is attributable to the 4° × 5° region in which the data are binned.

[28] The average ship winds for the period from January 24 to February 6 was 8.8 m s−1 for 30-s averages with a standard deviation of 4.31 m s−1 (n = 1872). The equation imageu10av−2 is 1.24, while equation imageu10av−3 is 1.76. The corresponding QuikSCAT wind statistics for the area covering the southern patch of SOFex (66°S–66.7°S and 171.4°W–172.4°W) are: average = 10.0 m s−1 and equation imageu10av−2 is 1.23, while equation imageu10av−3 is 1.73 for an average of 35 observations per day. Figure 7 shows the daily average winds and moments for the 15-day observational period. The good correspondence of the second and third moments suggests that the spatial variability obtained from QuikSCAT near the study site can statistically capture the temporal variability, as indicated by the ship's observations in accordance with the Taylor hypothesis. Slightly higher scatterometer winds compared with ship observations at higher winds have also been observed before and are believed to be caused by biases in local winds under stormy conditions [Wentz, 2002]. The comparison suggests a reasonable correspondence between the shipboard winds and QuikSCAT winds for the study period, including the moments expressing the effect of variability on the gas transfer.

image

Figure 7. Comparison of wind speeds and wind speed distributions for in situ winds from the R/V Revelle (dark shading) and QuikSCAT winds (white): (a) Daily average winds, (b) equation imageu10av−2, and (c) equation imageu10av−3 for January 24 to February 14, 2002.

Download figure to PowerPoint

[29] The average, second, and third moments for QuikSCAT over the entire year are compared with the blended QuikSCAT and SMM/I data set. The SSM/I has a higher temporal coverage of up to 8 times a day and yields, on average, 5.7 104 values per month. From this comparison, the effect of number of observations on the moments was determined. From August 2001 through July 2002, the QuikSCAT values for average wind, equation imageu10av−2, and equation imageu10av−3 were 9.76 m s−1, 1.15, and 1.46, respectively. For the blended product, the values were 10.6 m s−1, 1.25, and 1.91. This suggests that the QuikSCAT observations miss an appreciable fraction of the variability over the annual cycle that is captured with the blended product. The combination of higher winds and greater variability of the blended product leads to 28% greater gas transfer velocities for quadratic dependencies and 67% greater gas transfer velocities for cubic dependencies, illustrating the importance of accurate and high fidelity wind speed measurements to constrain fluxes in this region. A more detailed study of wind speed variability and the effect on global air-sea CO2 fluxes is forthcoming.

[30] Table 2 presents the fluxes for CO2 for both a cubic and quadratic dependence and compares the 41-year averaged NCEP wind product, 1995 NCEP wind product, and QuikSCAT wind product for the period August 2001 through July 2002. The uptake for the quadratic and cubic relationship for QuikSCAT winds, including the second and third moments, ranges from 1.4 to 1.6 Pg C yr−1. These values can be compared with estimates ranging from 1.4 to 2.6 Pg C yr−1 using the “long-term” wind equations of Wanninkhof [1992] and Wanninkhof and McGillis [1999], k = 0.39 u10av2, and k = 1.09 u10av − 0.333 u10av2 + 0.078 u10av3, respectively. The very high uptake of 2.6 Pg C yr−1 is, in part, due to the incorrect NCEP wind product used by Takahashi et al. [2002]. The good agreement of the quadratic relationship between this analysis and that of Takahashi et al. [2002] using the 1995 NCEP wind product is caused by the compensating effect of the 1995 NCEP winds being lower than the QuikSCAT winds, but the long-term relationship yielding a higher k.

Table 2. Summary of Air-Sea Fluxes in the Southern Ocean (>34°S)a
FormulationWind ProductReferenceu10av, m s−1equation imageuav−2equation imageuav−3Uptake, Pg C yr−1
  • a

    W-Long: long term relationship of Wanninkhof [1992], k = 0.39 u2 (Sc/660)−0.5; W&McG-Long: long term relationship of Wanninkhof and McGillis [1999], k = (1.09u − 0.33u2 + 0.078u3) (Sc/660)−0.5. NCEP-41: The 41-year averaged wind product from NCEP used by Takahashi et al. [2002]; NCEP-95: NCEP wind product for 1995; and QuikSCAT: scatterometer winds from August 2001 through July 2002. T-2002: Takahashi et al. [2002]. To determine the fluxes, the monthly averaged ΔpCO2 of Takahashi et al. [2002] were multiplied by the appropriate formulation using the monthly averaged wind speed for each 4° × 5° grid box. For the cases that the wind speed distribution function R = (Σuxuavx) is used, the R is for monthly values for each grid box.

W-LongNCEP-41T-200210.25  −1.72
W&McG-LongNCEP-41T-200210.25  −2.55
W-LongNCEP-95T-20029.26  −1.42
W&McG-LongNCEP-95T-20029.26  −1.92
0.31 R u2NCEP-95this work9.261.15 −1.29
0.0283 R u3NCEP-95this work9.26 1.54−1.40
0.31 R u2QuikSCATthis work9.761.15 −1.40
0.0283 R u3QuikSCATthis work9.76 1.46−1.59

[31] Boutin et al. [2002] performed a global analysis of the influence of gas transfer parameterizations using ERS-1 scatterometer winds. For the region >40°S they obtained values of 1.35 Pg C yr−1 and 1.51 Pg C yr−1 for the quadratic and cubic relationships, respectively, with the appropriate statistical estimates of variability, again suggesting that the overall effect of the difference between these two dependencies is small for the Southern Ocean. Our analysis shows an uptake of ≈0.3 Pg C yr−1 between 34°S and 40°S yielding an uptake of ≈1.2 Pg C yr−1 for >40°S, suggesting that the analysis of Boutin et al. [2002] with ERS-1 winds yielded a slightly greater uptake.

[32] The relationship with long-term averaged winds assumes a coefficient for the difference in steady versus variable winds of equation imageu10av−2 of 1.26 based on a global wind speed pattern following a Weibull distribution, as suggested by Wentz et al. [1984]. The observed coefficient using the QuikSCAT winds for the time period was 1.15. For the stronger nonlinear dependence of the cubic relationship, the less variable winds observed with QuikSCAT gives equation imageu10av−3 of 1.46 compared with a Weibull distribution yielding equation imageu10av−3 of 1.91. This leads to a significant difference in estimated uptake using the long-term averaged winds compared with QuikSCAT. A Weibull distribution assumption with a cubic dependency gives an uptake of −1.9 Pg C yr−1 compared with −1.6 Pg C yr−1 in our analysis. The high uptake rate using the long-term cubic dependence suggests that it is inappropriate to assume that global wind speed distributions are valid on a regional basis [Wanninkhof et al., 2002]. The uptake rates between the quadratic and cubic relationships are more concordant when enhancement factors are based on observed wind speed variability rather than using the long-term relationships. This is in large part because the global equation imageu10av−3 estimate of 1.91 is about 30% higher than the values determined in the Southern Ocean. When using observed wind speed distributions, the greater exchange at high winds using the cubic relationship is offset by lower gas transfer velocities at low winds compared with a quadratic dependence (see Figure 5).

[33] The results from the in situ gas exchange experiment do not suggest any significant difference in the gas exchange-wind speed relationship in the Southern Ocean when compared with the established empirical relationships. The difference in resulting CO2 fluxes between the quadratic and cubic wind speed dependence is significantly less than previous analyses. The discrepancy of model derived CO2 fluxes and uptakes estimated from observations of ΔpCO2 must be sought elsewhere. It has been suggested that the lack of observations of ΔpCO2 for the entire Southern Ocean for the winter season negatively biases the ΔpCO2 values. Metzl et al. [2001] show that the ΔpCO2 in the Indian Ocean in the winter is less than that suggested by the Takahashi climatology. However, changes in Southern Ocean uptake values in the updated climatology of Takahashi et al. [2002], which substantially increased the amount of Southern Ocean observations, show only a modest decrease in the Southern Ocean sink from −1.65 to −1.42 Pg C yr−1 using the long-term quadratic relationship of Wanninkhof [1992].

[34] A key question arising from this analysis is how to capture the influence of variability in wind speed on the CO2 fluxes. Although the QuikSCAT winds provide a time series record of high accuracy, the two observations per day, on average, appear to be insufficient to characterize the full spectrum of variability over a full year. From a mechanistic perspective, this requires capturing the variability in wind at the frequency of the response of gas transfer to surface forcing. Wind-wave tank studies suggest that this is less than 10 min [Jähne et al., 1989], but the exact magnitude is not constrained in the natural environment.

[35] The first gas exchange determination in the Southern Ocean using the dual deliberate tracer approach suggests a clear dependence of gas exchange on wind speed that is similar to previous parameterizations by Wanninkhof [1992] and Wanninkhof and McGillis [1999]. The SOFex results show that during a period of high and variable winds the cubic dependence with the coefficient proposed by Wanninkhof and McGillis [1999] provides a slightly better fit than the quadratic dependence of Wanninkhof [1992]. Applying our derived gas exchange wind speed parameterization to the climatological ΔpCO2 field does not lead to significant differences in the Southern Ocean CO2 uptake estimate when compared with that of Takahashi et al. [2002]. However, the difference in flux between the quadratic and a cubic dependence of gas exchange on wind speed is significantly less because the variability in winds is better accounted for by using remotely sensed wind products. This analysis points to the importance of properly accounting for the effect of variable winds on gas transfer.

Acknowledgments

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

[36] The ambitious SOFex study was organized by Kenneth Coale, Moss Landing, Kenneth Johnson of MBARI, and Ken Buesseler of WHOI. The success of the study can be directly attributed to their diligence and perseverance in all aspects of the work. Craig Neill of CNN Consulting, together with K. F. S., performed the deliberate tracer work on the Revelle, and the high-quality data is testament to their analytical skills. Joaquin Trinanes of AOML provided the QuikSCAT and SMM/I wind speed data through the support of NOAA/NESDIS. Editorial assistance by Gail Derr of AOML is greatly appreciated. Reviews and comments by David Ho and Phil Nightingale helped improve this work. The field work and data analysis was sponsored by the National Science Foundation under grant OCE 0000365. R. W. acknowledges NOAA/OAR for salary support.

References

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information
  • Asher, W. E., and R. Wanninkhof (1998), The effect of bubble-mediated gas transfer on purposeful dual gaseous-tracer experiments, J. Geophys. Res., 103, 10,55510,560.
  • Boutin, A. J., A. J. Etcheto, A. L. Merlivat, and A. Y. Rangama (2002), Influence of gas exchange coefficient parameterisation on seasonal and regional variability of CO2 air-sea fluxes, Geophys. Res. Lett., 29(8), 1182, doi:10.1029/2001GL013872.
  • Coale, K. H., et al. (2004), Southern Ocean iron enrichment experiment: Carbon cycling in high- and low-Si waters, Science, 304, 408414.
  • Fairall, C. W., J. E. Hare, J. B. Edson, and W. R. McGillis (2000), Parameterization and micrometeorological measurement of air-sea gas transfer, Boundary Layer Meteorol., 96, 63105.
  • Gloor, M., N. Gruber, J. Sarmiento, C. L. Sabine, R. A. Feely, and C. Rödenbeck (2003), A first estimate of present and pre-industrial air-sea CO2 flux patterns based on ocean interior carbon measurements and models, Geophys. Res. Lett., 30(1), 1010, doi:10.1029/2002GL015594.
  • Glover, D. M., N. M. Frew, S. J. McCue, and E. J. Bock (2002), A multi-year time series of global gas transfer velocity from the TOPEX/Poseidon dual frequency normalized radar backscatter algorithm, in Gas Transfer at Water Surfaces, Geophys. Monogr. Ser., vol. 127, edited by M. Donelan et al., pp. 325333, AGU, Washington, D. C.
  • Gulliver, J. S., B. Erickson, A. J. Zaske, and K. S. Shimon (2002), Measurement uncertainty in gas exchange coefficients, in Gas Transfer at Water Surfaces, Geophys. Monogr. Ser., vol. 127, edited by M. Donelan et al., pp. 239245, AGU, Washington, D. C.
  • Gurney, K. R., et al. (2002), Toward robust regional estimates of CO2 sources and sinks using atmospheric transport models, Nature, 415, 626629.
  • Jähne, B., K. O. Münnich, R. Bösinger, A. Dutzi, W. Huber, and P. Libner (1987), On parameters influencing air-water gas exchange, J. Geophys. Res., 92, 19371949.
  • Jähne, B., P. Libner, R. Fischer, T. Billen, and E. J. Plate (1989), Investigating the transfer process across the free aqueous viscous boundary layer by the controlled flux method, Tellus, Ser. B, 41, 177195.
  • Jenkins, W. J., and W. B. Clarke (1976), The distribution of 3He in the western Atlantic Ocean, Deep Sea Res., 23, 481494.
  • Keeling, R. F., S. C. Piper, and M. Heinmann (1996), Global and hemispheric CO2 sinks deduced from changes in atmospheric O2 concentration, Nature, 381, 218.
  • Kuss, J., K. Nagel, and B. Schneider (2004), Evidence from the Baltic Sea for enhanced CO2 air-sea transfer velocity, Tellus, Ser. B, 56, 175182.
  • Law, C. S., A. J. Watson, M. J. Liddicoat, and T. Stanton (1998), Sulphur hexafluoride as a tracer of biogeochemical and physical processes in an open-ocean iron fertilisation experiment, Deep Sea Res., Part II, 45, 977994.
  • McGillis, W. R., J. B. Edson, J. E. Hare, and C. W. Fairall (2001), Direct covariance air-sea CO2 fluxes, J. Geophys. Res., 106, 16,72916,745.
  • Metzl, N., C. Brunet, A. Jsabuad-Jan, A. Poisson, and B. Scauer (2001), Summer and winter air-sea fluxes in the Southern Ocean, in Extended Abstracts, 6th International Carbon Dioxide Conference, edited by T. Nakazawa, pp. 685689, World Meteorol. Org., Sendai, Japan.
  • Monahan, E. C. (2002), The physical and practical implications of a CO2 gas transfer coefficient that varies as the cube of the wind speed, in Gas Transfer at Water Surfaces, Geophys. Monogr. Ser., vol. 127, edited by M. Donelan et al., pp. 193198, AGU, Washington, D. C.
  • Nightingale, P. D., P. S. Liss, and P. Schlosser (2000a), Measurements of air-sea gas transfer during an open ocean algal bloom, Geophys. Res. Lett., 27, 21172120.
  • Nightingale, P. D., G. Malin, C. S. Law, A. J. Watson, P. S. Liss, M. I. Liddicoat, J. Boutin, and R. C. Upstill-Goddard (2000b), In situ evaluation of air-sea gas exchange parameterizations using novel conservative and volatile tracers, Global Biogeochem. Cycles, 14, 373387.
  • Takahashi, T., R. A. Feely, R. Weiss, R. Wanninkhof, D. W. Chipman, S. C. Sutherland, and T. T. Takahashi (1997), Global air-sea flux of CO2: An estimate based on measurements of sea-air pCO2 difference, Proc. Natl. Acad. Sci. USA, 94, 82928299.
  • Takahashi, T., et al. (2002), Global sea-air CO2 flux based on climatological surface ocean pCO2 and seasonal biological and temperature effects, Deep Sea Res., Part II, 49, 16011622.
  • Upstill-Goddard, R. C., A. J. Watson, J. Wood, and M. I. Liddicoat (1991), Sulphur hexafluoride and helium-3 as seawater tracers: Deployment techniques and continuous underway analysis for sulphur hexafluoride, Anal. Chim. Acta, 249, 555562.
  • Wanninkhof, R. (1992), Relationship between gas exchange and wind speed over the ocean, J. Geophys. Res., 97, 73737381.
  • Wanninkhof, R., and W. M. McGillis (1999), A cubic relationship between gas transfer and wind speed, Geophys. Res. Lett., 26, 18891893.
  • Wanninkhof, R., W. Asher, R. Weppernig, H. Chen, P. Schlosser, C. Langdon, and R. Sambrotto (1993), Gas transfer experiment on Georges Bank using two volatile deliberate tracers, J. Geophys. Res., 98, 20,23720,248.
  • Wanninkhof, R., W. E. Asher, and E. C. Monahan (1995), The influence of bubbles on air-water gas exchange: Results from the gas transfer experiments during WABEX-93, in Third International Symposium on Air-Water Gas Transfer, edited by B. Jähne, and E. C. Monahan, pp. 239254, Aeon Verlag and Studio, Hanau, Germany.
  • Wanninkhof, R., et al. (1997), Gas exchange, dispersion, and biological productivity on the west Florida Shelf: Results from a Lagrangian tracer study, Geophys. Res. Lett., 24, 17671770.
  • Wanninkhof, R., S. C. Doney, T. Takahashi, and W. R. McGillis (2002), The effect of using time-averaged winds on regional air-sea CO2 fluxes, in Gas Transfer at Water Surfaces, Geophys. Monogr. Ser., vol. 127, edited by M. Donelan et al., pp. 351357, AGU, Washington, D. C.
  • Watson, A. J., P. S. Liss, and R. A. Duce (1991a), Design of a small-scale iron fertilization experiment, Limnol. Oceanogr., 36, 19601965.
  • Watson, A. J., R. C. Upstill-Goddard, and P. S. Liss (1991b), Air-sea exchange in rough and stormy seas measured by a dual tracer technique, Nature, 349, 145147.
  • Wentz, F. J. (2002), The status of measuring ocean winds from space, Eos Trans. AGU, 83(47), Fall Meet. Suppl., Abstract OS62D-01.
  • Wentz, F. J., S. Peteherych, and L. A. Thomas (1984), A model function for ocean radar cross sections at 14.6 GHz, J. Geophys. Res., 89, 36893704.
  • Woolf, D. K. (1997), Bubbles and their role in gas exchange, in The Sea Surface and Global Change, edited by P. S. Liss, and R. A. Duce, pp. 173205, Cambridge Univ. Press, New York.

Supporting Information

  1. Top of page
  2. Abstract
  3. 1. Introduction
  4. 2. Methods
  5. 3. Discussion
  6. Acknowledgments
  7. References
  8. Supporting Information

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.