Radiometric nonlinearity and the correction strategies for infrared hyperspectral benchmark sounder

Abstract

The infrared (IR) benchmark sounder is designed to detect the tiny change of long-term global climate by measuring the spectrally resolved IR radiance emitted from Earth to space with high accuracy. Besides, the IR sounder also serve as a space-borne radiometric reference to convert the international fleet of weather sounders into a climate benchmarking system with excellent global coverage and similar measurement accuracy. In order to achieve high accuracy, the benchmark sounder must be tuned to be a linear response system and be well radiometrically calibrated. So the nonlinearity response in an IR detector signal chain needs to be corrected prior to the linear radiometric calibration. There are some algorithmic approaches being commonly used to correct the nonlinear measurements. These methods use the measured nonlinear interferograms to polynomially fit the corrected linear interferograms, without considering the physical root of non-linearity. However, they work well only when the detector nonlinearity is small. Regarding the large nonlinearity, a correction method is proposed in this paper. It follows the nonlinearity response mechanism of the IR detector, and uses the to-be-solved linear interferogram to polynomially fit the measured non-linear interferogram signal formally, and then derive the correction coefficients from the established equations. According to the correction evaluation and methods comparison using the simulated data as proxy measurements, the proposed method is appropriate for both small and large degree of quadratic nonlinearity detectors.