Evaluation of the uncertainty in the spectral band adjustment factor (SBAF) for cross-calibration using Monte Carlo simulation

Abstract

Cross-calibration is one of the various methods applied for Earth Observation Satellites sensor calibration. In the cross-calibration procedure, one sensor is calibrated against another sensor, in which the radiometric calibration is better known, via near-simultaneous imaging of a common ground target. One of the most important steps during the cross-calibration is the Spectral Band Adjustment Factor (SBAF) assessment. The SBAF is used to compensate the differences in the spectral responses of the sensors, avoiding large uncertainties in cross-calibration results. The investigation described in this work focussed on the evaluation of the SBAF's inherent uncertainties using Monte Carlo Simulation method. Basically, the Monte Carlo approach is based on calculating multiple integral by random sampling. The SBAFs were developed for analogous Landsat 8 Operational Land Imager and CBERS 4 Multispectral Camera spectral bands. The Hyperion hyperspectral sensor on-board Earth Observing-1 was utilized to understand the spectral profile of the target and to derive the SBAF. This study was performed over two pseudo invariant calibration sites: Algodones Dunes and Libya-4. The spectral uncertainty of the SBAFs using Monte Carlo was found to be within 0.01-1.79%. The results suggested that the uncertainty of the SBAFs is dependent on the correlation between the input variables: the higher the correlation, the lowest is the SBAF uncertainty.