ON THE ROLE OF LARGE BUBBLES IN AIR-SEA GAS-EXCHANGE AND SUPERSATURATION IN THE OCEAN

A parameterization of bubble-induced gas exchange is presented in which the bubble contribution to gas exchange is expressed in terms of separate transfer velocities for ingassing (K(b)in) and outgassing (K(b)out). The difference between the ingassing and outgassing velocities (K(b)in - K(b)out) is further separated into two components, the first caused by the injection of small bubbles into the water, the second caused by gas exchange across the surface of hydrostatically compressed larger bubbles. It is argued that both K(b)out and the exchange contribution to the difference K(b)in - K(b)out should be largely independent of the dissolved concentrations of the major gases N2 and O2. A simple model is presented which allows K(b)out and the exchange contribution to the difference K(b)in - K(b)out to be estimated. The model incorporates data from laboratory simulation experiments on the bubble production spectrum. The results indicate that bubbles larger than 0.05 cm in radius, which have often been assumed to play a negligible role, contribute significantly to bubble-induced gas exchange and supersaturation in the ocean. The model is used to explore the sensitivity of bubble-induced gas exchange to the overall air entrainment rate, size and depth distributions of the bubbles, and to the gas exchange rates across the surface of individual bubbles. The model suggests that bubbles may make an important contribution to overall gas exchange at windspeeds above 10 m sec-1. In this regime gas transfer velocities should depend, not just on diffusivity, but also on the solubility of the gases. It is suggested that K(b)(out) should scale roughly as alpha-0.3D0.35 where a is the solubility and D is the diffusivity. The model results, in combination with measurements on inert gas supersaturations, suggest that the global-mean supersaturation of CO2 induced by bubbles is not larger than 0.3% and most probably is around 0.08%. A major uncertainty results from a lack of information on production rates and distributions of large bubbles. Several possible experiments are proposed for improving estimates of bubble-induced gas exchange and supersaturation.