Harmonizing Landsat-8 OLI and Sentinel-2 MSI: an assessment of surface reflectance and vegetation index consistency

Normalization of satellite images collected under various atmospheric conditions is critical for the comprehensive, long-term global surveillance of terrestrial surface alterations. This study utilized remote sensing data from the Sentinel-2A Multispectral Instrument (MSI) in polar orbit and the Landsat-8 Operational Land Imager (OLI) sensors, with multispectral global coverage of 10–30 m, to derive reflectance products using inversion algorithms. Validation and assessment were conducted using synchronous surface measurement spectra collected from four sites across three Chinese provinces in 2019. We corrected surface reflectance and derived vegetation indices across blue, green, red, near-infrared (NIR), and two short-wave infrared (SWIR) bands and normalized discrepancies. The phenological spatial distribution map for late rice in Jiangxi Province was constructed using normalized data outcomes. A robust linear correlation in reflectance across corresponding bands of the two satellite sensors was observed. The NIR and SWIR bands showed the most significant difference because of differences in their spectral response functions. A high degree of congruence was observed between Landsat-8 OLI and Sentinel-2 MSI sensor reflectance products, with root mean square error values consistently below 0.05. The derived conversion equations were highly accurate for harmonizing data from both sensor systems.