Climatology of the terms and variables of transformed Eulerian-mean (TEM) equations from multiple reanalyses: MERRA-2, JRA-55, ERA-Interim, and CFSR.

A 30-year (1980–2010) climatology of the major variables and terms of the transformed Eulerian-mean (TEM) momentum and thermodynamic equations is constructed by using four global atmospheric reanalyses, MERRA-2, JRA-55, ERA-Interim, and CFSR. Both the reanalysis ensemble mean (REM) and the differences of each reanalysis from the REM are investigated in the latitude-pressure domain for December-January-February and for June-July-August. For the REM investigation, two residual vertical velocities (the original one and one evaluated from residual meridional velocity) and two mass streamfunctions (from meridional and vertical velocities) are compared, and longwave (LW) and shortwave (SW) radiative heatings are also shown and discussed. For the TEM equations, the residual terms are also calculated and investigated for their potential usefulness, as the residual term for the momentum equation should include the effects of parameterised processes such gravity waves, while that for the thermodynamic equation should indicate the analysis increment. Inter-reanalysis differences are investigated for the mass streamfunction, LW and SW heatings, the two major terms of the TEM momentum equation (the Coriolis term and the Elliassen-Palm flux divergence term), and the two major terms of the TEM thermodynamic equation (the vertical temperature advection term and the total diabatic heating term). The spread among reanalysis TEM momentum balance terms is around 10 % in Northern-Hemisphere winter and up to 50 % in Southern-Hemisphere winter. The largest uncertainties in the thermodynamic equation (about 50 %) are found in the vertical advection, which does not show a structure consistent with the differences in heatings. The results shown in this paper provide basic information on the degree of agreement among recent reanalyses in the stratosphere and in the upper troposphere in the TEM framework. [ABSTRACT FROM AUTHOR]