Atmospheric Measurement Techniques | 2021
IMK/IAA MIPAS temperature retrieval version 8: nominal measurements
Abstract
Abstract. A new global set of atmospheric temperature profiles is retrieved from\nrecalibrated radiance spectra recorded with the Michelson Interferometer\nfor Passive Atmospheric Sounding (MIPAS). Changes with respect to previous\ndata versions include a new radiometric calibration considering the\ntime dependency of the detector nonlinearity and a more robust frequency\ncalibration scheme. Temperature is retrieved using a smoothing constraint,\nwhile tangent altitude pointing information is constrained using optimal\nestimation. ECMWF ERA-Interim is used as a priori temperature below 43\u2009km.\nAbove, a priori data are based on data from the Whole Atmosphere Community Climate\nModel Version 4 (WACCM4).\nBias-corrected fields from specified dynamics runs, sampled at the MIPAS\ntimes and locations, are used, blended with ERA-Interim between 43 and 53\u2009km.\nHorizontal variability of temperature is considered\nby scaling an a priori 3D temperature field in the orbit plane in a way that\nthe horizontal structure is provided by the a priori while the vertical\nstructure comes from the measurements.\nAdditional microwindows with better\nsensitivity at higher altitudes are used. The background continuum is\njointly fitted with the target parameters up to 58\u2009km altitude. The radiance\noffset correction is strongly regularized towards an empirically determined\nvertical offset profile. In order to avoid the propagation of uncertainties of\nO 3 and H 2 O a priori assumptions, the abundances of these species are\nretrieved jointly with temperature. The retrieval is based on HITRAN 2016\nspectroscopic data, with a few amendments.\nTemperature-adjusted climatologies of vibrational populations of CO 2 states\nemitting in the 15\u2009 µ m region are used in the radiative transfer\nmodeling in order to account for non-local thermodynamic equilibrium.\nNumerical integration in the radiative transfer model is now performed at\nhigher accuracy. The random component of the temperature uncertainty typically\nvaries between 0.4 and 1\u2009K, with occasional excursions up to 1.3\u2009K above 60\u2009km\naltitude.\nThe leading sources of the random component of the temperature error are\nmeasurement noise,\ngain calibration uncertainty, spectral shift, and uncertain CO 2 mixing\nratios. The systematic error is caused by uncertainties in spectroscopic data\nand line shape uncertainties. It ranges from 0.2\u2009K at 20\u2009km altitude for\nnorthern midlatitude summer conditions to 2.3\u2009K at 12\u2009km for tropical\nconditions. The estimated total uncertainty amounts to values between\n0.6\u2009K at 20\u2009km for midlatitude summer conditions to 2.5\u2009K at 12–15\u2009km for\ntropical conditions. The vertical resolution varies\naround 3\u2009km for altitudes below 50\u2009km. The long-term drift encountered in the\nprevious temperature product has been largely reduced.\nThe consistency between\nhigh spectral resolution results from 2002 to 2004 and the reduced spectral\nresolution results from 2005 to 2012 has been largely improved.\nAs expected, most\npronounced temperature differences between version 8 and previous data versions are\nfound in elevated stratopause situations.\nThe fact that the phase of temperature waves seen\nby MIPAS is not locked to the wave phase found in ECMWF analyses demonstrates\nthat our retrieval provides independent information and does not merely reproduce\nthe prior information.