Premium
Crosstalk effect mitigation in black body warm‐up cool‐down calibration for Terra MODIS longwave infrared photovoltaic bands
Author(s) -
Sun Junqiang,
Madhavan Sriharsha,
Wang Menghua
Publication year - 2016
Publication title -
journal of geophysical research: atmospheres
Language(s) - English
Resource type - Journals
eISSN - 2169-8996
pISSN - 2169-897X
DOI - 10.1002/2016jd025170
Subject(s) - longwave , radiance , remote sensing , infrared , crosstalk , irradiance , physics , optics , shortwave , photovoltaic system , spectroradiometer , offset (computer science) , moderate resolution imaging spectroradiometer , calibration , environmental science , reflectivity , computer science , radiative transfer , geology , quantum mechanics , programming language , ecology , satellite , biology , astronomy
Abstract The crosstalk phenomena in Terra MODerate‐resolution Imaging Spectroradiometer (MODIS) midwave‐to‐longwave infrared (LWIR) photovoltaic (PV) bands (bands 27‐30) have recently been individually studied and characterized, and a correction algorithm has been developed. The routine calibration of the four LWIR PV bands uses an onboard black body (BB) based on a quadratic model for the relationship between the at‐aperture radiance and the background subtracted instrument response. While the crosstalk correction has been successfully applied in both the routine BB calibration (scan basis) to correct the crosstalk effect in the linear term and the Earth view (EV) radiance in our previous investigations on the bands 27–29, the most recent work on band 30 demonstrated a newfound necessity to include the impact of the crosstalk effect on the nonlinear term as well as the offset. In this paper, we analyze the calibration calculation under a variety of conditions consistent with the MODIS Collection 6 settings to examine the impact of the crosstalk effect in the two terms derived from the BB warm‐up‐cool‐down (WUCD) calibration. We show that with correct account of the crosstalk effect in the WUCD calibration, the sudden changes and other abnormal features that have been observed in the two terms for many years are effectively and remarkably removed. In addition, imagery for bands 27–29 using the calibration result fully corrected for the electronic crosstalk effect shows further improvement over previous results that account for only the corrected linear term, whereas for the band 30 different testing scenes validate the previous fully corrected findings.