Elsevier

Journal of Marine Systems

Volume 159, July 2016, Pages 67-75
Journal of Marine Systems

Using eddy covariance to estimate air–sea gas transfer velocity for oxygen

https://doi.org/10.1016/j.jmarsys.2016.02.008Get rights and content

Highlights

  • We show the first atm. air–sea transfer velocities (k660) for O2 using the EC method.

  • The O2 fluxes are counter to the measured CO2-flux and within the same magnitude.

  • Our k660 shows a stronger wind dependence than used parametrizations of k660 for O2.

  • The strong wind dependence of k appears to coincide with the onset of breaking waves.

Abstract

Air–sea gas transfer velocity for O2 is calculated using directly measured fluxes with the eddy covariance technique. It is a direct method and is frequently used to determine fluxes of heat, humidity, and CO2, but has not previously been used to estimate transfer velocities for O2, using atmospheric eddy covariance data. The measured O2 fluxes are upward directed, in agreement with the measured air–sea gradient of the O2 concentration, and opposite to the direction of the simultaneously measured CO2 fluxes. The transfer velocities estimated from measurements are compared with prominent wind speed parameterizations of the transfer velocity for CO2 and O2, previously established from various measurement techniques. Our result indicates stronger wind speed dependence for the transfer velocity of O2 compared to CO2 starting at intermediate wind speeds. This stronger wind speed dependence appears to coincide with the onset of whitecap formation in the flux footprint and the strong curvature of a cubic wind-dependent function for the transfer velocity provides the best fit to the data. Additional data using the measured O2 flux and an indirect method (based on the Photosynthetic Quotient) to estimate oxygen concentration in water, support the stronger wind dependence for the transfer velocity of O2 compared to CO2.

Introduction

Air–sea exchange of oxygen is a research area of increasing scientific interest. Oxygen in the ocean is of crucial importance, as it is strongly linked to both the carbon and nitrogen cycles (Arrigo, 2005, Keeling et al., 2010) and is a key component in eutrophication (Karlsson et al., 2002). Gas transfer velocity describes the efficiency of the gas transfer process across the air–sea interface. Over the years numerous laboratory and field studies have been made to study gas transfer velocity. Several expressions have been proposed, relating the gas transfer velocity to different physical variables such as sea spray and bubbles (Monahan and Spillane, 1984, Wallace and Wirick, 1992, Woolf, 1997, Asher and Wanninkhof, 1998, McNeil and d’Asaro, 2007), surface films (Broecker et al., 1978, Frew, 1997, Salter et al., 2011), water-side convection (MacIntyre et al., 2002, Rutgersson and Smedman, 2010, Rutgersson et al., 2011) and rain (Ho et al., 1997, Ho et al., 2004, Zappa et al., 2009, Takagaki and Komori, 2007); still, wind speed holds as the most robust single parameter to describe the transfer velocity (Liss and Merlivat, 1986). The relative importance of these different processes for air–sea gas exchange is likely to vary for gases of different solubility and is still not fully understood (Garbe et al., 2014). It is well known that transfer velocities for water vapor and heat (corresponding to very high solubility) have a weaker wind speed dependence than CO2 (e.g. Rutgersson et al., 2001). Studies on the transfer velocities for acetone (Yang et al., 2014) and DMS (Huebert et al., 2010, Marandino et al., 2007) gases with relatively high solubility, has shown less wind speed dependence than the transfer velocity for CO2. By introducing eddy covariance measurements of oxygen (having a solubility 60 times lower than CO2), knowledge about the effect of a variety of processes on the transfer velocity could fundamentally increase.

The onset of breaking waves is expected to enhance the water-side turbulence (Kitaigorodskii, 1984, Woolf, 2005); in addition, it has been suggested that bubbles injection occurring in the near surface waters enhance the transfer process (Keeling, 1993, Woolf, 1993). Monahan and Spillane (1984) suggested a relation whereby the whitecap coverage scales to the cube of the wind speed. A cubic wind speed dependence for the transfer velocity, however, has only been verified for a few field studies (e.g. Wanninkhof and McGillis, 1999, Kihm and Körtzinger, 2010). Woolf (1997) proposed a relation of the transfer velocity to wind speed, using a theoretical model including injection of air bubbles in the bulk of the water. The importance of bubbles injection from wave breaking to the gas transfer velocity has been suggested to be inversely proportional to the gas solubility (Keeling, 1993, Asher et al., 1996, Woolf, 1997, Asher and Wanninkhof, 1998). However, the transfer velocities for different gases have also shown to be highly sensitive to void fraction and bubble distribution (Woolf, 1993, Woolf et al., 2007) within the water column.

Here we use the eddy covariance (EC) method to directly measure the air–sea gas flux of O2 and estimate a gas transfer velocity for oxygen. Only few previous studies exists using atmospheric EC data for oxygen (Andersson et al., 2014). Previous estimates of the gas transfer velocities of oxygen have instead been made using a budget method (Kihm and Körtzinger, 2010) or eddy covariance fluxes in water (McNiel and d’Asaro, 2007). Kihm and Körtzinger (2010) found a stronger wind speed dependence for the transfer velocity of O2 compared to prominent parametrizations for CO2. However, for most cases the gas transfer velocity of O2 is assumed to follow the same wind speed dependence as other nonreactive gases with low solubility. We will here use three different types of data to estimate the transfer velocity for oxygen. In all data sets the transfer velocities are determined from the measured EC O2 flux. For data set 1 (cf. 4.1) and 2 (cf. 4.2) measured air–sea gradient of oxygen are used in the computation of the transfer velocities, while for period 3 the concept of PQ (Photosynthetic Quotient cf. 3.3) was applied to determine the oxygen concentration in water within the flux footprint.

Section snippets

Theory

The air–sea flux, F, of a non-reactive gas such as O2 can be described byF=kCwαCawhere k is the transfer velocity, Cw denotes the gas concentration in the bulk water, α the Ostwald solubility constant, and Ca the gas concentration in the air near the water surface. This transfer efficiency is also commonly described in terms of the resistance R, which is inversely proportional to k. The two-layer film model (Liss and Slater, 1974) describes the total resistance as the sum of the air resistance

Östergarnsholm

The measurements used in this study are taken at the Östergarnsholm site in the Baltic Sea. This station is located at 57°27′N, 18°59′E (Fig. 1). The site has been running semi-continuously since 1995. The site includes a land-based 30-m tower with the base 1 ± 0.5 m above mean sea level, situated on the southern tip of a flat island in the Baltic Sea and mooring instruments mounted on a buoy 1 km SE of the tower.

The tower is equipped with both slow response instrument for profiles and high

Case2013

Data was collected from June 19 to June 23, 2013. During the field campaign winds between 3 and 12 m s 1 were measured, with two periods when conditions were stationary and with wind from the marine undisturbed sector, 80°–160° (Fig. 2a). The first period was found at the start of the measurements and covers 12 h of data showing mostly positive (upward directed) oxygen fluxes in the range 0.6–6.4 μmol m 2 s 1 (Fig. 2b), in agreement with the oxygen saturation in water varying between 105 % and 108 % (

Conclusions

Based on results from eddy covariance measurements in the Baltic Sea (Östergarnsholm), the first atmospheric air–sea fluxes and gas transfer velocities for O2 based on EC measurements are presented. These data are subject for corrections, where the density correction is the major one, introducing a mean uncertainty of 23% to the O2 flux. Within the periods that fulfil selection criteria's regarding wind direction and stationarity however, the O2 fluxes display an anti-correlation with the

Acknowledgments

We would like to give a special thank to Heidi Pettersson with staff at the Finnish Meteorological Institute for their help with deployment of the mooring instrument.

References (54)

  • A. Andersson et al.

    Using a high-frequency fluorescent oxygen probe in atmospheric eddy covariance applications

    J. Atmos. Ocean. Technol.

    (2014)
  • K.R. Arrigo

    Marine microorganisms and global nutrient cycles

    Nature

    (2005)
  • W.E. Asher et al.

    The effect of bubble-mediated gas transfer on purposeful dual-gaseous tracer experiments

    J. Geophys. Res.

    (1998)
  • W.E. Asher et al.

    The influence of bubble plumes on air–water gas transfer velocities

    J. Geophys. Res.

    (1996)
  • H.C. Broecker et al.

    The influence of wind on CO2 exchange in a wind wave tunnel, including the effects of monolayers

    J. Mar. Res.

    (1978)
  • X. Cai et al.

    Forward-in-time and backward-in-time dispersion in the convective boundary layer: the concentration footprint

    Bound.-Layer Meteorol.

    (2007)
  • D.J. Conley et al.

    Hypoxia-related processes in the Baltic Sea

    Environ. Sci. Technol.

    (2009)
  • D.J. Conley et al.

    Hypoxia is increasing in the coastal zone of the Baltic Sea

    Environ. Sci. Technol.

    (2011)
  • N.M. Frew

    The role of organic films in air–sea gas exchange

  • Garbe, C. S., A. Rutgersson, J. Boutin, B. Delille, C. W. Fairall, N. Gruber, J. Hare, D. Ho, M. Johnson, G. de Leeuw,...
  • D.T. Ho et al.

    The effect of rain on air–water gas exchange

    Tellus

    (1997)
  • D.T. Ho et al.

    Influence of rain on air–sea gas exchange: lessons from a model ocean

    J. Geophys. Res.

    (2004)
  • U. Högström et al.

    To what extent can we believe measurements on a land-based tower to represent upwind open sea conditions

    Boreal Environ. Res.

    (2008)
  • T.W. Horst et al.

    Footprint estimation for scalar flux measurements in the atmospheric surface layer

    Bound.-Layer Meteorol.

    (1992)
  • B.J. Huebert et al.

    Linearity of DMS transfer coefficient with both friction velocity and wind speed in the moderate wind speed range

    Geophys. Res. Lett.

    (2010)
  • K. Karlsson et al.

    Temporal and spatial large-scale effects of eutrophication and oxygen deficiency on benthic fauna in Scandinavian and Baltic waters: a review

    Oceanogr. Mar. Biol.

    (2002)
  • R.F. Keeling

    On the role of large bubbles in air–sea gas exchange and supersaturation in the ocean

    J. Mar. Res.

    (1993)
  • R.F. Keeling et al.

    Ocean deoxygenation in a warming world

    Mar. Sci.

    (2010)
  • C. Kihm et al.

    Air–sea gas transfer velocity for oxygen derived from float data

    J. Geophys. Res.

    (2010)
  • S. Kitaigorodskii

    On the fluid dynamical theory of turbulent gas transfer across an air–sea interface in the presence of breaking wind waves

    J. Phys. Oceanogr.

    (1984)
  • N. Kljun et al.

    A three-dimensional backward Lagrangian footprint model for a wide range of boundary-layer stratifications

    Bound.-Layer Meteorol.

    (2002)
  • E.A. Laws

    Photosynthetic quotients, new production and net community production in the open ocean

    Deep Sea Res. Part A

    (1991)
  • P.S. Liss et al.

    Air–sea gas exchange rates: introduction and synthesis

  • P.S. Liss et al.

    Fluxes of gases across the air–sea interface

    Nature

    (1974)
  • S. MacIntyre et al.

    The critical importance of buoyancy flux for gas flux across the air–water interface

  • C.A. Marandino et al.

    Eddy correlation measurements of the air/sea flux of dimethylsulfide over the North Pacific Ocean

    J. Geophys. Res. Atmos.

    (2007)
  • C.D. McNeil et al.

    Parameterization of air–sea gas fluxes at extreme wind speeds

    J. Mar. Syst.

    (2007)
  • Cited by (5)

    View full text