A Dynamic Spatial-time Dark Level Correction Approach for Lunar Radiometric Calibration of LOIS

Lunar radiometric calibration is used to solve the problem of consistent radiometric calibration for multiple satellite platforms and remote sensors. However, the dark level will fluctuate when observing the Moon with the short-wave infrared spectrometer, which seriously affects the accuracy of the lunar radiation data. In this work, we proposed a dynamic spatial-time dark level correction approach to address the fluctuation of the dark level. This method employs the cold space signals in space and time dimensions to estimate the dark level for each frame individually and to reduce errors due to environmental variations. Experiments on lunar observations at multiple phase angles were conducted and the dark level correction results demonstrate that our proposed method is effective even in the short-wave infrared, and is also superior to currently existing techniques. For the single-band (1700 nm) image of the full moon, the mean background proportion of the proposed method is 1.00%, which is better than that of static dark correction method (2.25%) and linear dark correction method (5.93%).