OBSERVED BROADBAND RADIANT FLUX AND CLOUD RADIATIVE EFFECT (CRE) HAVE SERVED AS THE STANDARD DATASETS FOR EVALUATING GLOBAL CLIMATE MODEL FOR NEARLY THREE DECADES. A MAJOR CONSEQUENCE USING THIS APPROACH IS THAT COMPENSATING BIASES FROM DIFFERENT SPECTRAL BANDS COULD MAKE THE UNDERSTANDING OF THE BROADBAND DEFICIENCIES DIFFICULT. SPECTRAL RADIANCE IS THE INTEGRAND OF BROADBAND FLUX CONTAINING MUCH MORE INFORMATION ABOUT OUR ATMOSPHERE THAN ANY OTHER SINGLE OBSERVATION FROM SPACE. DUE TO WELL-KNOWN DIFFICULTIES IN RETRIEVING GEOPHYSICAL PARAMETERS FROM SUCH OBSERVATIONS RETRIEVAL PRODUCTS USUALLY HAVE LARGER UNCERTAINTY THAN THE CALIBRATED RADIANCES. THEREFORE DIRECT USE OF SPECTRAL OBSERVATIONS IN MODEL EVALUATIONS AND DIAGNOSTICS CAN REVEAL COMPENSATING BIASES IN THE MODEL THAT CANNOT BE REVEALED FROM BROADBAND FLUX DIAGNOSTICS ALONE. EXPOSING SUCH COMPENSATING BIASES USUALLY ALSO HELP US FIND CLUES TO CORRECT THEM. FROM THIS PERSPECTIVE HIGH-QUALITY HYPERSPECTRAL RADIANCE OBSERVATIONS FROM NASA SUCH AS AIRS AND CRIS ARE UNIQUELY VALUABLE FOR MODEL EVALUATIONS AND SIMULATION DIAGNOSTICS. FUNDED BY THE PREVIOUS TERRA/AQUA PROGRAMS THE PI HAS ESTABLISHED ALGORITHMS TO DERIVE SPECTRAL FLUXES OVER THE ENTIRE LONGWAVE (LW) SPECTRUM FROM COLLOCATED AIRS AND CERES OBSERVATIONS ON AQUA AND HAS PUBLISHED A SERIES OF ARTICLES TO USE SUCH SPECTRAL FLUXES TO EVALUATE GCMS. FUNDED BY ANOTHER NASA ROSES PROGRAM THE PI AND HIS COLLABORATORS HAVE ALSO DEVELOPED A LW SPECTRAL RADIATIVE KERNEL (SRK) TECHNIQUE. THE LW SRK CAN USE ROUTINE CMIP MONTHLY-MEAN MODEL OUTPUTS TO DIRECTLY COMPUTE SPECTRA FLUX ANOMALIES AND HENCE TO DERIVE SPECTRAL DETAILS OF LW RADIATIVE RESPONSES TO YEAR-TO-YEAR CLIMATE FLUCTUATIONS. AIRS ABOARD AQUA AND CRIS ABOARD S-NPP HAVE BEEN PROVIDING CALIBRATED AND STABLE SPECTRAL RADIANCE MEASUREMENTS SINCE SEPTEMBER 2002 AND MARCH 2012 RESPECTIVELY. HERE WE PROPOSE TO EXTEND THE SPECTRAL FLUX ALGORITHMS THAT WE HAVE DEVELOPED TO COLLOCATED CRIS AND CERES MEASUREMENTS AND TO USE THE MULTIPLE-YEAR SPECTRAL FLUX DATA MERRA-2 REANALYSIS CMIP MODEL OUTPUTS AND THE SRK TECHNIQUE TO FURTHER BROADEN THE USAGE OF NASA SPECTRAL OBSERVATIONS IN MODEL DIAGNOSTICS. SPECIFIC SUBTASKS ARE: 1. EXTENDING OUR ALGORITHM TO COLLOCATED CRIS AND CERES OBSERVATIONS ON S-NPP AND VALIDATING THE DERIVED SPECTRAL FLUX AS WE HAVE DONE FOR THE AIRS AND CERES OBSERVATIONS ON AQUA. 2. FOR THE PERIOD THAT BOTH AIRS AND CRIS HAVE OBSERVATIONS ASSESS THE CONSISTENCY OF OUR ALGORITHMS FOR THE AIRS&CERES AND FOR THE CRIS&CERES MEASUREMENTS. 3. USING THE LW SRK AND GEOPHYSICAL VARIABLES FROM THE MERRA-2 REANALYSIS WE WILL COMPUTE THE SHORT-TERM SPECTRAL RADIATIVE RESPONSES TO DETRENDED SURFACE TEMPERATURE VARIATIONS FROM 09/2002 TO 08/2017. THE RADIATIVE RESPONSES OF TEMPERATURE AND HUMIDITY CAN BE COMPUTED DIRECTLY. USING OBSERVED SPECTRAL FLUX FROM SUBTASKS 1&2 AND ADJUST METHOD THE SPECTRAL RADIATIVE RESPONSE DUE TO CLOUDS CAN BE ALSO DERIVED. 4. SIMILAR TO 3 USING THE LW SPECTRAL FLUX SIMULATOR AND THE LW SRK TO DERIVE THE SHORT-TERM SPECTRAL RADIATIVE RESPONSES OF THE CMIP5 GCMS. 5. UNDERSTAND THE DATA-MODEL DIFFERENCE FROM SUBTASKS 3&4. E.G. THEY WILL BE SORTED WITH RESPECT TO THE VERTICAL LEVEL WHERE THE SPECTRAL JACOBIAN OF EACH CHANNEL ATTAINS ITS MAXIMUM. BY DOING SO WE CAN AT LEAST TO THE FIRST ORDER ATTRIBUTE THE MODEL-OBSERVATION DISCREPANCIES TO CAUSES FROM DIFFERENT VERTICAL LEVELS. THE PROPOSED STUDIES WILL ESTABLISH A FRAMEWORK FOR THE CONTINUATION OF PRODUCING SPECTRAL FLUX DATA FROM MULTIPLE INSTRUMENTS AND FOR THE INTEGRATION OF SUCH SPECTRAL OBSERVATIONS INTO DATA-MODEL COMPARISON EFFORTS. IT WILL ALSO PROMOTE THE USE OF NASA SPECTRAL OBSERVATIONS TO THE ENTIRE CLIMATE MODELING COMMUNITY. THIS STUDY IS RELATED TO TWO BIG QUESTIONS FOR NASA'S EARTH SCIENCE RESEARCH: "HOW DOES THE EARTH SYSTEM RESPOND TO NATURE AND HUMAN-INDUCED CHANGES?" AND "HOW WILL THE EARTH SYSTEM CHANGE IN THE FUTURE?"