Sea ice cover influences the generation of surface ocean turbulence in ways that sometimes enhance, but mostly inhibit air-water gas exchange. Inhibition happens as ice cover reduces wind fetch, enhancement occurs when haline convection or sea ice drift creates additional surface turbulence. We used the bulk turbulence relationships within the Wave Age Gas Transfer model to estimate air-sea gas transfer velocity (kWAGT), based on sea ice cover and turbulence conditions in the ice-ocean boundary layer, throughout a year-long (2019–2020) ice drift campaign in the central Arctic Ocean. During the drift, sea ice cover averaged >97%, with a minimum of 58%, and boundary layer shear played a dominant role in the turbulence budget. Modeled turbulent kinetic energy dissipation was compared against 167 in-situ profiles of ocean dissipation to evaluate model performance and explore related processes. The modeled dissipation and observed dissipation profiles, averaged over 0–4 m depth, agreed within 1% of each other, with a mean dissipation of 5.8 × 10-7 W kg-1. Examining individual dissipation estimates by surface conditions, however, revealed poorest agreement in leads, especially leads covered by thin ice, which the model cannot detect. Dissipation from the model was used to produce a time series of kWAGT, revealing an average velocity of 0.034 m d-1 or 1% of the global average for the open ocean. Comparison with a widely used wind speed parameterization for gas exchange showed that wind speed scaling would overestimate k during 92% of the drift by 3.5 times on average, demonstrating how fetch limitation can suppress gas exchange, even as open water increases. These results suggest that photic zone processes, under-ice blooms, and attendant cycling of CO2 and O2 as well as CH4 can remain isolated from the atmosphere for an entire annual cycle in the central Arctic.