The Physical and Practical Implications of a CO2Gas Transfer Coefficient that Varies as the Cube of the Wind Speed

Abstract

Wanninkhof and McGillis [1999] recently presented further observational evidence supporting the contention that there is a cubic relationship between wind speed and the air-sea gas transfer (piston) velocity for CO 2 . They acknowledged that Monahan and Spillane [1984] were the first to propose that the gas transfer coefficient is proportional to whitecap coverage and that whitecap coverage scales approximately as u 3 . While Wanninkhof and McGillis' findings give further support to the contention that at moderate and high wind speeds it is the fraction of the sea surface covered by Stage-A whitecaps (spilling wave crests) that determines the magnitude of the air-sea gas transfer coefficient, it should be noted that a cubic expression is only an approximation of the true dependence of whitecap coverage on wind speed [Monahan and O'Muircheartaigh, 1980], and that other factors such as wind duration need to be taken into account [Monahan and O'Muircheartaigh, 1986]. Since global maps of fractional oceanic whitecap coverage can now be generated using satellite-borne microwave radiometers, it follows that the potential exists for routinely producing similar global maps of CO 2 gas transfer coefficient (piston velocity).