The Advanced Baseline Imager (ABI) will be aboard the National Oceanic and Atmospheric Administration’s Geostationary Operational Environmental Satellite R-Series (GOES-R) to supply data needed for operational weather forecasts and long-term climate variability studies, which depend on high quality data. Unlike the heritage operational GOES systems that have two or four detectors per band, ABI has hundreds of detectors per channel requiring calibration coefficients for each one. This increase in number of detectors poses new challenges for next generation sensors as each detector has a unique spectral response function (SRF) even though only one averaged SRF per band is used operationally to calibrate each detector. This simplified processing increases computational efficiency. Using measured system-level SRF data from pre-launch testing, we have the opportunity to characterize the calibration impact using measured SRFs, both per detector and as an average of detector-level SRFs similar to the operational version. We calculated the spectral response impacts for the thermal emissive bands (TEB) theoretically, by simulating the ABI response viewing an ideal blackbody and practically, with the measured ABI response to an external reference blackbody from the pre-launch TEB calibration test.
The impacts from the practical case match the theoretical results using an ideal blackbody. The observed
brightness temperature trends show structure across the array with magnitudes as large as 0.1 K for and 12 (9.61 µm), and 0.25 K for band 14 (11.2 µm) for a 300 K blackbody. The trends in the raw ABI signal viewing the blackbody support the spectral response measurements results, since they show similar trends in bands 12 (9.61µm), and 14 (11.2 µm), meaning that the spectral effects dominate the response differences between detectors for these bands. We further validated these effects using the radiometric bias calculated between calibrations using the external blackbody and another blackbody, the ABI on-board calibrator. Using the detector-level SRFs reduces the structure across the arrays but leaves some residual bias. Further understanding of this bias could lead to refinements of the blackbody thermal model. This work shows the calibration impacts of using an average SRF across many detectors instead of accounting for each detector SRF independently in the TEB calibration. Note that these impacts neglect effects from the spectral sampling of Earth scene radiances that include atmospheric effects, which may further contribute to artifacts post-launch and cannot be mitigated by processing with detector-level SRFs. This study enhances the ability to diagnose anomalies on-orbit and reduce calibration uncertainty for improved system performance.