Sen2Cor - Sentinel-2 Level-2 Optical Processor Applied to Landsat-8 Data

Abstract

Sen2Cor is the official ESA Sentinel-2 ground segment processor for the generation of the Sentinel-2 Level 2A core products from the Level 1C top of atmosphere reflectance in fixed cartographic geometry. Sen2Cor can also be downloaded from the ESA website as a standalone tool for individual Level 2A processing by the users. It can be run either via command line or as a plugin of the Sentinel-2 Toolbox (SNAP-S2TBX). The Sentinel-2 A/B Level 2A products are bottom of atmosphere reflectance in cartographic geometry, which are widely distributed to the users since March 2018 over Copernicus Open Access Hub and Sentinel Hub. In this study, we test the capability of the Sen2cor algorithm for scene classification and atmospheric correction to be able to perform Landsat-8 Level 1 input data processing. Both instruments have eight overlapping spectral bands and the measurements are often used complimentarily for studies of vegetation and land parameters. However, there are also distinct differences between the two sensors, such as spectral bands response, calibration and viewing geometries, which are reflected in the differences between the L1 products. In the bands arrangements of Landsat-8, the water vapor band is missing. These have to be taken into account in the modification and upgrade of the Sen2Cor algorithm. The ability of Sen2Cor to process Landsat-8 scenes in the same manner as the Sentinel-2 ones is of interest for the Sen2Like framework, which is detailed by Telespazio France on a parallel presentation at this workshop. This will allow Sentinel-2 and Landsat-8 TOA reflectance to be converted to surface reflectance using the same atmospheric correction algorithm and a radiative transfer model adapted to the Landsat conditions. We address the necessary algorithmic modifications and the uncertainty due to the Level 1 to Level 2 processing methodology. A qualitative comparison of both Sen2Cor generated products and comparison to the Landsat-8 LaSrc products from USGS is also presented.