Abstract Daily spectral radiance observations by NASA's Atmospheric Infrared Sounder contain detailed information about surface and atmospheric temperature and water vapor. We obtain climate geophysical trends from 20 years (2002/09–2022/08) of Atmospheric Infrared Sounder (AIRS) observations using a novel method operating mostly in radiance space. The observations are binned into 3 5 degree tiles using 16 day intervals, after which nominally clear scenes are selected for each tile to construct the spectral radiance time series. Deseasonalized spectral trends are then obtained, which are inverted using a physical retrieval to obtain geophysical trends. This approach is distinct from traditional use of radiances whereby trends are generated after operational retrievals or assimilation into Reanalysis models. Our approach rigorously ties the derived geophysical trends to the observed radiance trends, using far fewer computational resources and time. The retrieved trends are compared to trends derived from ERA5 and MERRA2 reanalysis geophysical fields, and NASA Level3 AIRS v7 and CLIMCAPS v2 data. Our retrieved surface temperature trends agree quite well with ERA5, CLIMCAPS, and the GISS surface climatology trends. Atmospheric temperature profile trends exhibit some variability among all these data sets, especially in the polar stratosphere. Water vapor profile trends are nominally similar among the data sets except for the AIRS v7 which exhibits drying trends in the mid troposphere. Spectral closure between observed trends and those computed by running the reanalysis and AIRS L3 monthly retrieval products through a radiative transfer code are discussed, with the major differences arising in the water vapor sounding region.
DeSouza‐Machado et al. (Mon,) studied this question.