Quantifying the influence of surfactants on coastal air-sea gas exchange rates through measurements of dissolved organic carbon
Motivation: Accurate quantification of air-sea gas exchange rates is critical for understanding the carbon cycle, both locally and globally. Approximately half of the anthropogenic CO2 released to the atmosphere from 1800–1994 was absorbed by the ocean. Without this carbon sink, anthropogenic CO2 levels would be roughly 60 ppmv higher than they are today. A major source of uncertainty in our understanding of the carbon cycle is parameterization of the gas transfer coefficient, k. For example, the estimated annual air-sea flux of CO2 based on global measurements ranges from 1.1 to 1.9 Pg C yr-1, depending on which parameterization for the gas transfer coefficient is used.[2,3] Additionally, the gas transfer coefficient is a critical variable required for quantifying biological production and fluxes of photosynthetic O2, dimethyl sulfide, N2O, and other biogenic gases.
Coastal regions are major contributors to the ocean carbon cycle. The air-sea flux of CO2 from continental shelves is estimated at 0.2 to 0.4 Pg C yr-1 and the flux from estuaries and salt marshes is estimated at -0.1 to -0.4 Pg C yr-1 .[4,5] However, the most commonly-cited gas exchange parameterizations are based on open ocean measurements, and are likely inaccurate for coastal regions.[3,6] My aim is to develop a new parameterization for the gas transfer coefficient in coastal waters based on measurements of inert gases and a surfactant proxy. This parameterization could be used by scientists around the world to improve quantification of coastal carbon fluxes and biological production.
Background: Through funding from NSF, my advisor Rachel Stanley and I have developed a portable noble gas mass spectrometer. Measurements of multiple noble gases can be used to quantify air-sea gas exchange rates because the gases are biologically inert and respond in different ways to physical forcing such as temperature change and bubble-mediated gas injection. The portable instrument has a response time of ~10 min and is well suited for sample collection in heterogeneous environments such as coastal regions.
Most published gas exchange parameterizations calculate the gas transfer coefficient as a nonlinear function of the wind speed.[3,6] However, in coastal regions, using wind speed as the only environmental variable controlling gas exchange is insufficient to accurately quantify the gas transfer coefficient. Marine surfactants (surface active organic compounds) reduce the gas transfer rate by reducing turbulence at the air-sea interface. Surfactants are primarily derived from phytoplankton and are expected to be higher in abundance in productive coastal regions, compared to the open ocean.
Frew (1997) collected surface ocean water on a transect from Rhode Island to Bermuda and then transferred each water sample to a wind-wave tank, where he measured the gas transfer coefficient at a constant wind stress (Fig. 1). He observed a negative linear correlation between the gas transfer coefficient, and the surfactant or dissolved organic carbon concentration ([DOC]). This result suggests that the gas transfer coefficient has a first-order dependence on surfactant and DOC concentration.
I propose to measure [DOC] as a proxy for surfactants. Hunter and Liss (1982) demonstrated that surfactant concentration in estuarine waters scales linearly with [DOC] (Fig. 2). Developing a gas transfer coefficient parameterization that is a function of both wind speed and [DOC] will improve the accuracy of calculated gas exchange fluxes in coastal regions.
Proposed research and use of funding: This summer, I will be collaborating with members of the Physical Oceanography department at WHOI who have been funded to conduct a field and modeling campaign to better understand coastal circulation in the waters near Martha’s Vineyard, MA. As part of this research, they will be performing Lagrangian drifter release experiments in June and September 2014. During the drifter release experiments, over a period of ~1 week, I will be collecting noble gas measurements with my portable system while tracking individual drifters, and use the resulting data set to model gas exchange rates. Collecting data in a Lagrangian framework simplifies the mass balance model because advection can be neglected.I am requesting funding of $1294 to collect and analyze 140 samples for dissolved organic carbon concentration, as a proxy for surfactant concentration, during the drifter release experiments. I will analyze the samples in the lab of Dr. Vlahos at the University of Connecticut, who has generously agreed that I will not be required to pay any fees for my training on and use of the total organic carbon analyzer. Therefore, my budget only includes the consumables required for sample collection and analysis. This will greatly reduce the cost of this research. Additionally, I will be applying to the MIT Houghton Fund and/or the MIT Student Assistance Fund for funding to cover the costs associated with my travel to analyze the samples. This research cannot be funded through existing grants because the NSF grant that is currently funding my thesis research is intended to pay for the development of the noble gas mass spectrometer and it does not include any funding to measure additional parameters such as [DOC]. Additionally, that grant will expire before the drifter release experiments begin.
- Sabine, C.L., Feely, R.A., et al., 2004. Science 305, 367–371.
- Takahashi, T., Sutherland, S.C., et al., 2002. Deep Sea Res. II 49, 1601–1622.
- Ho, D.T., Law, C.S., et al. 2006. Geophys. Res. Lett. 33, 16611.
- Cai, W.J., 2011. Ann. Rev. Mar. Sci. 3, 123–145.
- Chen, C.T.A., Borges, A.B., 2009. Deep Sea Res. II 56, 578–590
- Wanninkhof, R., 1992. J. Geophys. Res. 97, 7373–7382.
- Stanley, R.H.R, Jenkins, W.J., Lott, D.E., Doney. S.C. 2009. J. Geophys. Res. 114, C11020.
- Frew, N. 1997. In: The Sea Surface and Global Change, edited by Liss, P.S. and Duce, R.A. 151–172.
- Hunter, K., Liss, P., 1982. Limnol. Oceanogr. 27, 322–335.