Noble gas constraints on air-sea gas exchange and bubble fluxes

Air-sea gas exchange is an important part of the biogeochemical cycles of many climatically and biologically relevant gases including CO2, O-2, dimethyl sulfide and CH4. Here we use a three year observational time series of five noble gases (He, Ne, Ar, Kr, and Xe) at the Bermuda Atlantic Time series Study (BATS) site in tandem with a one-dimensional upper ocean model to develop an improved parameterization for air-sea gas exchange that explicitly includes separate components for diffusive gas exchange and bubble processes. Based on seasonal timescale noble gas data, this parameterization, which has a 1 sigma uncertainty of +/-14% for diffusive gas exchange and +/-29% for bubble fluxes, is more tightly constrained than previous parameterizations. Although the magnitude of diffusive gas exchange is within errors of that of Wanninkhof (1992), a commonly used parameterization, we find that bubble-mediated exchange, which is not explicitly included by Wanninkhof (1992) or many other formulations, is significant even for soluble gases. If one uses observed saturation anomalies of Ar (a gas with similar characteristics to O-2) and a parameterization of gas exchange to calculate gas exchange fluxes, then the calculated fluxes differ by similar to 240% if the parameterization presented here is used compared to using the Wanninkhof (1992) parameterization. If instead one includes the gas exchange parameterization in a model, then the calculated fluxes differ by similar to 35% between using this parameterization and that of Wanninkhof (1992). These differences suggest that the bubble component should be explicitly included in a range of marine biogeochemical calculations that incorporate air-sea gas fluxes.