We compare the daily, interannual, and decadal variability and trends in the thermal structure of the Arctic troposphere using eight observation-based, vertically resolved data sets, four of which have data prior to 1948. Comparisons on the daily scale between historical reanalysis data and historical upper-air observations were performed for Svalbard for the cold winters 1911/1912 and 1988/1989, the warm winters 1944/1945 and 2005/2006, and the International Geophysical Year 1957/1958. Excellent agreement is found at mid-tropospheric levels. Near the ground and at the tropopause level, however, systematic differences are identified. On the interannual time scale, the correlations between all data sets are high, but there are systematic biases in terms of absolute values as well as discrepancies in the magnitude of the variability. The causes of these differences are discussed. While none of the data sets individually may be suitable for trend analysis, consistent features can be identified from analyzing all data sets together. To illustrate this, we examine trends and 20-year averages for those regions and seasons that exhibit large sea-ice changes and have enough data for comparison. In the summertime Pacific Arctic and the autumn eastern Canadian Arctic, the lower tropospheric temperature anomalies for the recent two decades are higher than in any previous 20-year period. In contrast, mid-tropospheric temperatures of the European Arctic in the wintertime of the 1920s and 1930s may have reached values as high as those of the late 20th and early 21st centuries.
Bibliographical noteFunding Information:
This work was supported by the Swiss National Science Foundation (projects “Past climate variability from an upper-level perspective” and EVALUATE) and by the EU FP7 project ERA-CLIM. The Twentieth Century Reanalysis Project (20CR) used resources of the National Energy Research Scientific Computing Center and of the National Center for Computational Sciences at Oak Ridge National Laboratory, which are supported by the Office of Science of the US Department of Energy under Contract No. DE-AC02-05CH11231 and Contract No. DE-AC05-00OR22725, respectively. Support for the 20CR dataset is provided by the US Department of Energy, Office of Science Innovative and Novel Computational Impact on Theory and Experiment (DOE INCITE) program, and Office of Biological and Environmental Research (BER), and by the National Oceanic and Atmospheric Administration Climate Program Office. Constructive comments by two anonymous reviewers and also by P.D. Sardeshmukh of the University of Colorado CIRES/CDC and NOAA/ESRL/PSD on an earlier version of this manuscript are greatly appreciated.