"The thermal infrared radiation has rich information content about surface properties, atmospheric gas composition and temperature profiling. Meteorological satellites and ground instruments are used to monitor upwelling and downwelling radiances, respectively. This study provides new analysis methodologies in different applications of hyperspectral infrared data in climate and weather monitoring. In the first study, an airborne spectrometer flying at tropopause level is proposed for stratospheric H2O sounding derived from downwelling radiances. A motivation is that variations of stratospheric water vapor impact significantly on the surface radiation budget. The advantage of using an airborne spectrometer is its ability to capture the small-scale H2O variability. A Line-By-Line radiative transfer model (LBLRTM) and Gauss-Newton iterative techniques are used to construct the retrieval algorithm. Our simulation experiments show that a spectrometer with 700-2000 cm-1 spectral coverage, 1 cm-1 spectral resolution, 10 s observation time, 12 cm aperture diameter, and realistic noise levels across different bands can reach a retrieval accuracy of 0.5 ppmv and 1 K for simultaneous retrieval of H2O and temperature in the lower to middle stratosphere. We also find that far infrared ( 700 cm-1) spectral measurements, because of their significantly higher noise level, provide little additional benefit for this application. In the second study, application of hyper-spectral radiance data for climate feedback analysis is discussed. Using General Circulation Model (GCM) profiles as input to the Moderate resolution atmospheric Transmission (MODTRAN) radiation code, spectral water vapor sensitivity Jacobians are developed to study the impact of water vapor on the Outgoing Longwave Radiation (OLR). It is shown that OLR changes linearly with water vapor change in the atmospheric window (800-1250 cm-1) and logarithmically in the water vapor absorption bands (the rotational band: 0-560 cm-1 and the vibrational-rotational band 1250-1850 cm-1). Using a Line-By-Line Radiative Transfer Model, LBLRTM, we develop second order H2O Jacobians and show that these terms provide a more accurate estimate of the sensitivity of the absorption to H2O perturbations in a given atmospheric layer. Based on these findings, to diagnose the water vapor radiative effect, we propose to rely on hybrid H2O scaling, i.e. linearly in window, and logarithmically in water vapor bands. In the third study, observed infrared radiance data are used to validate forecasts in range 0-24 h in radiance space. We adapt for cloudy sky computations a fast radiative transfer model commonly used at weather centers, RTTOV (Radiative Transfer for Television Operational Vertical Sounder), to generate synthetic radiance spectra from the output of a weather forecast model: the Global Environmental Multiscale (GEM) model. Synthetic radiance spectra are compared against Atmospheric Infrared Sounder (AIRS) observed radiance data. Novel in this study, spectral Jacobians of surface temperature, atmospheric temperature, and water vapor, obtained by prescribing unperturbed and perturbed profiles, are used for diagnosing the biases in GEM. Clear-sky diagnosis provides a test of radiation closure, in which we find radiance biases of GEM can be well explained by geophysical variable biases, such as surface temperature, and temperature and water vapor profiles. Cloud-related radiance biases are then further determined in all-sky conditions. Based on this spectral analysis, the following prominent GEM biases were identified: 1) too low (often 5 K) surface temperature over land, notably in daytime; 2) too high upper tropospheric humidity in the subtropical region (equivalent to a negative bias of 3 K in brightness temperature units); 3) a deficit of high clouds in western Pacific ocean Intertropical Convective Zone." --