It is important to examine the parameterizations used in calculating air-sea exchange fluxes as they are essential in developing global carbon models and in carbon budget calculations. We quantify the potential biases involved in the parameterizations. Adopting a non-zero gas transfer velocity for low wind areas results in a significant increase in the CO2 flux in equatorial regions with a net increase of +0.2 Pg C yr-1 in the total sea-air global flux. The ocean "cool skin temperature" effect on CO2 flux estimation is found to be an order of magnitude smaller than early estimations. The previously unknown salty-skin effect has an opposite contribution that cancels the cool-skin effect. Comparing different wind speeds derived from satellite data and Global Circulation Models (GCM), the most significant divergence is found at the low wind equatorial regions regarding the CO2 flux estimation.