Multi-points cross calibration and validation of HJ-1B and AIRS thermal infrared channels

Abstract

On orbit radiometric calibration of thermal infrared sensors in China is mainly based on the single-point method with insitu measurements of large uniform targets and cross calibration with moderate resolution sensors which have high calibration accuracy. The single-point method is time-consuming and laborious, and the intercept can not be calculated. Accuracy of cross calibration with the moderate resolution sensors is restricted by the matching algorithm of entrance pupil radiances. Both of this two methods' insufficients have been affected the accuracy of on orbit radiometric calibration in China. According to the characteristics of AIRS hyperspectral data, multi-points cross calibration is performed between the AIRS and HJ-1B thermal infrared channles. The time and spatial varieties are eliminated through the similar imaging times and homogeneous regions. At the same time, radiative transfer calculation by MODTRAN is used for each pixel to normalize the observation geometry. Times series multi-points method with the single target, various times and homogeneous targets multi-points method with the single time, various targets are performed to obtain the calibration coefficients of HJ-1B thermal infrared channel. Onboard calibration coefficients calculated by the Look-up table method is used to validate the multi-points cross calibration results. After the elimination of abnormal points from the time series multi-points method and regulation of match factor with homogeneous targets multi-points method, the differences between the equivalent brightness temperature retrieved by the multi-points cross calibration and onboard calibration coefficients are 1.81 and 0.92 K, respectively. The results indicate that the homogeneous targets multi-points method for cross calibration of AIRS and HJ-1B thermal infrared channels has higher accurate.