Verification of ERA-Interim and ERA5 Reanalyses Data on Surface Air Temperature in the Arctic

The paper presents the results of the verification of ERA-Interim and ERA5 reanalyses data on surface air temperature obtained from drifting buoys, ground-based weather stations, and, for the first time, from measurements at the North Pole drifting stations. The North Pole station data were not assimilated in the reanalyses, which provides a rare opportunity for independent validation. The comparison with data of the North Pole drifting stations revealed that bias for the cold season in the Arctic basin is 2.25 degrees C for ERA-Interim and 3.92 degrees C for ERA5, respectively. The comparison with data of drifting buoys allows us to speculate about the cause of such large errors in the reanalyses. Some buoys were installed from the air-based platform. In this case, the temperature sensor of the buoy was potentially buried in the snow cover that shielded it from the cold atmosphere and contributed to the heating due to the heat flux from the underlying layer of sea water. The assimilation of such data could be one of the reasons for the overestimation of air temperature over drifting ice in the both reanalysis.