Verification of air/surface humidity differences from AIRS and ERA-Interim in support of turbulent flux estimation in the Arctic

Evaporation from the Arctic Ocean and its marginal seas is essential for air moisture, cloudiness, and precipitation, as well as for the associated feedbacks, which contribute to the Arctic amplification of climate warming. However, evaporation in the Arctic is still associated with large uncertainties. The Boisvert et al. (2013) moisture flux scheme (BMF13) is based on application of the Atmospheric Infrared Sounder (AIRS) data, which produces high-quality, global, daily atmospheric temperature and moisture profiles even in the presence of clouds. Comparing the results of BMF13 against the ERA-Interim reanalysis, we found differences up to 55Wm(-2) in the surface latent heat flux in the Beaufort-East Siberian Seas (BESS). We found out that the quality of the input data for the BMF13 and ERA-Interim flux schemes was the main cause for the differences. Differences in the input data sets cause moisture flux estimates to differ up to 1.6x10(-2)gm(-2)s(-1) (40Wm(-2) latent heat flux) in the BESS region, when both data sets were applied to the BMF13 scheme. Thus, the input data sets, AIRS version 6 and ERA-Interim reanalysis, were compared with a variety of in situ data. In skin temperature ERA-Interim had twice as large an error as AIRS version 6, but smaller errors in air specific humidity. The results suggested that AIRS data and the BMF13 scheme are a good option to estimate the moisture flux in the Arctic. However, the differences detected demonstrate a need for more in situ measurements of air temperature and humidity in the Arctic.