A comparison of satellite‐derived carbon dioxide transfer velocities from a physically based model with GasEx cruise observations

Abstract

A method has been developed to use satellite observations with the physically based National Oceanic and Atmospheric Administration–Coupled Ocean Atmosphere Response Experiment (NOAA‐COARE) gas transfer model (COAREG) to determine global estimates of carbon dioxide (CO2) transfer velocities. Ship observations from three Gas Exchange (GasEx) experiments conducted in 1998 (GasEx‐98), 2001 (GasEx‐01), and 2008 (SO GasEx) were used to validate the satellite estimates. Satellite retrievals of near‐surface temperature, humidity, and wind speed along with observations of sea surface temperature, longwave and shortwave radiative fluxes at the surface were used to execute the gas transfer model. These satellite products exhibited good agreement with collocated in situ measurements even in higher‐latitude environments where the retrieval techniques had not yet been extensively validated. Transfer velocities computed with COAREG using both the satellite‐ and ship‐derived inputs further showed good agreement with an overall bias of −1.01 cm/h and a root mean square (RMS) difference of 7.43 cm/h. These results imply that COAREG can be applied with satellite inputs with accuracy comparable to that achieved with ship‐based observations. COAREG transfer velocities were found to have a bias of 6.32 cm/h and RMS difference of 14.02 cm/h when compared with covariance‐derived transfer velocities. Comparison of satellite‐derived COAREG transfer velocities with empirical‐derived wind‐only models shows generally higher transfer velocities for wind speeds exceeding 7 m/s. However, these higher transfer velocities show good agreement with covariance‐derived observations from both GasEx‐98 and SO GasEx.