SBAF for cross-calibration of Landsat-8 OLI and Sentinel-2 MSI over North African PICS

Abstract

Accurate radiometric cross calibration is critical for guaranteeing the consistency of measurements from different Earth observation sensors, and fully using the combined data in quantitative applications. It becomes even more indispensable with the rapid increase of remote sensing data availability from numerous sensors. The assessment of the Spectral Band Adjustment Factor (SBAF) is a key component of the cross-calibration method. The SBAF compensates for intrinsic differences in sensor response caused by Spectral Response Function (SRF) mismatches. Currently, Sentinel and Landsat data represent the most widely accessible medium spatial resolution multispectral satellite data. Hence, in this study, the SBAF of the Multi-Spectral Imager (MSI) on-board Sentinel-2 and the Operational Land Imager (OLI) on-board Landsat-8 was estimated over pseudo-invariant calibration sites (PICS) located in North Africa. The SBAF depends on the hyperspectral profile of the target and the sensor SRF. Here, the hyperspectral profile was derived from the Hyperion hyperspectral imager on-board the EO-1. Finally, it is important to highlight that an estimate of the SBAF is incomplete unless accompanied with its uncertainty. The uncertainty analysis of the SBAF was implemented using Monte Carlo simulation. The results obtained in this study can be utilized by any user who needs the SBAF of the OLI and MS1 over North Africa Desert sites.