We have estimated the impact of extra-tropical hurricanes in the NW Atlantic on air-sea CO2 exchanges, using two approaches. On one hand the conventional representation of the air-sea CO2 exchange rate is given in terms of the gas transfer velocity formula, expressed as a function of wind speed by Wanninkhof. In a second approach, we used a new formula for gas transfer velocity as a function of a breaking-wave parameter, as proposed by Zhao and Toba. This is based on whitecap coverage and depends on only friction velocity and wave age.

Using climatological fields as background fields, we found that there is absorbance in the northern part of our domain of implementation, the Labrador Sea and Grand Banks, and efflux in the southern part, the Gulf Stream area, consistent with recent previous studies, for tropical and extratropical regions. Limitations in these calculations are that the parameterizations for gas transfer velocity may not be valid for high wind speeds, i.e. in excess of 30 m/s. This is a concern for either the Zhao-Toba wave-breaking formulation or the Wanninkhof wind algorithm. We show that the variability in air-sea parameters such as winds and waves, for example between coupled and uncoupled atmosphere-ocean runs, surpasses the variability implicit between Zhao-Toba wave-breaking or Wanninkhof-wind formulations for the gas transfer velocity. Moreover, coupled model Zhao-Toba simulations seem more realistic in the sense that maximum gas transfer velocities occur along the storm tracks.