General Comments This paper details temperature biases of two ECMWF datasets (ERA-interim and ERA-5) with respect to two independent observational datasets (ground-based lidar and satellite-based MLS). Biases are calculated from long-term comparisons of over 20 years for four NH lidars and ∼14 years for MLS, focusing on the "wintertime" period (October-March). In addition, lidar comparisons were made for Pre-2000, 2000-2007, and Post-2007 to assess impacts of inclusion of different satellite

more »

... rces. The results show a significant cold bias with respect to lidar in the upper stratosphere of ∼3-4 K in ERA-interim and a warm bias above 3hPa, both of which are reduced in ERA- C1 ACPD Interactive comment Printer-friendly version Discussion paper