Uncertainties in gas exchange parameterization during the SAGE dual-tracer experiment

A dual tracer experiment was carried out during the SAGE experiment using the inert tracers SF6 and 3He, in order to determine the gas transfer velocity, k, at high wind speeds in the Southern Ocean. Wind speed/gas exchange parameterization is characterised by significant variability and we examine the major measurement uncertainties that contribute to that scatter. Correction for the airflow distortion over the research vessel, as determined by computational fluid dynamics (CFD) modelling, had the effect of increasing the calculated value of k by 30%. On the short time scales of such experiments, the spatial variability of the wind field resulted in differences between ship and satellite QuikSCAT winds, which produced significant differences in transfer velocity. With such variability between wind estimates, the comparison between gas exchange parameterizations from diverse experiments should clearly be made on the basis of the same wind product. Uncertainty in mixed layer depth of ∼10% arose from mixed layer deepening at high wind speed and limited resolution of vertical sampling. However the assumption of equal mixing of the two tracers is borne out by the experiment. Two dual tracer releases were carried out during SAGE, and showed no significant difference in transfer velocities using QuikSCAT winds, despite the differences in wind history. In the SAGE experiment, duration limitation on the development of waves was shown to be an important factor for Southern Ocean waves, despite the presence of long fetches.