Some aspects of temperature sounding of the atmosphere

Based on measurements with the interferometer spectrometer SI-1 on board of Meteor satellites the following problems are investigated: the systematic effect of the spectral response function in deriving effective radiation temperatures; the selection of ‘optimal’ spectral channels within the 15 μm region, and effects of additional molecular absorbers on the determination of the temperature profile. It can be shown that using broad-band window channels (half width ?20cm^?1) the derived surface temperature is significantly different from the surface temperature derived from narrow spectral band channels (e.g. for the HIRS window channel 8 this difference is about 1° C). Simulations of some combinations of spectral channels including one proposed as an optimal approach show no significant difference in the derived temperature profile. Neglecting the ozone absorption of the 14.1 μm band leads to a change in radiance of about 1.5 mW/m²sr cm^?1 and to a systematic error of the derived temperature of about ?2° C in the troposphere.