Solar signals in tropospheric re-analysis data: Comparing NCEP/NCAR and ERA40

Recent analyses of the NCEP/NCAR re-analysis data have demonstrated statistically significant variations in tropospheric temperatures, geopotential heights, water vapour distribution, and global circulation regimes in phase with the solar cycle over the last 44 years. The findings reveal a consistent pattern of atmospheric response to solar variability throughout the low- and mid-latitude troposphere. In order to more firmly establish the reality and nature of the detected Sun–climate relations, and to correctly identify the atmospheric processes involved, independent confirmation from other relevant data sets is warranted. In this paper we present the results of identical analyses of the NCEP/NCAR and ERA40 re-analysis data sets and show that the detected Sun–climate relations are substantially weaker in the ERA40 data than in the NCEP data. We argue that this largely is due to the presence of temporal inhomogeneities in ERA40 having a substantial impact on decadal signals, which strongly affects the observed Sun–climate relations as well as climate trends. The nature of the NCEP-ERA40 differences, their influence on the detection of Sun–climate relations, and their possible causes are discussed.