Sen2Cor version 2.10: Last evolutions and Focus on the update of Cloud Screening and Scene Classification algorithm

Kurzfassung

Sen2Cor is a Level-2A processor with the main purpose to correct single-date Sentinel-2 Level 1C products from the effects of the atmosphere in order to deliver a Level-2A surface reflectance product. Side products are Cloud Screening and Scene Classification (SCL), Aerosol Optical Thickness (AOT) and Water Vapour (WV) maps. The Sen2Cor version 2.10 has been developed with the aim to improve the quality of both the surface reflectance products and the Cloud Screening and Scene Classification (SCL) maps in order to facilitate their use in downstream applications like the Sentinel-2 Global Mosaic (S2GM) service. This version is planned to be used operationally within Sentinel-2 Ground Segment and for the Sentinel-2 Collection 1 reprocessing. The Cloud Screening and Scene Classification module is performed prior to the atmospheric correction and provides a Scene Classification map divided into 11 classes. This map does not constitute a land cover classification map in a strict sense. Its main purpose is to be used internally in Sen2Cor’s atmospheric correction module to distinguish between cloudy -, clear - and water pixels. Two quality indicators are also provided: a Cloud - and a Snow confidence map with values ranging from 0 to 100 (%). The presentation provides an overview of the last evolutions of Sen2Cor including the support of new L1C products with processing baseline >= 04.00 and the provision of additional L2A quality indicators. The different steps of the Cloud Screening and Scene Classification algorithm are recalled: cloud/snow -, cirrus -, cloud shadow detection, pixel recovery and post-processing with DEM information. It will also detail the latest updates of version 2.10 that makes use of the parallax properties of the Sentinel-2 MSI instrument to limit the false detection of clouds above urban and bright targets. Finally, SCL validation results with Sen2Cor 2.10 are included in the presentation. The recent improvements as well as the current limitations of the SCL-algorithm are presented. Some advices are given on the configuration choices and on the use of external auxiliary data files.