Sen2Like: Paving the Way towards Harmonization and Fusion of Optical Data


Creative Commons License

Saunier S., Pflug B., Lobos I. M. , Franch B., Louis J., De Los Reyes R., ...More

REMOTE SENSING, vol.14, no.16, 2022 (SCI-Expanded) identifier

  • Publication Type: Article / Article
  • Volume: 14 Issue: 16
  • Publication Date: 2022
  • Doi Number: 10.3390/rs14163855
  • Journal Name: REMOTE SENSING
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, CAB Abstracts, Compendex, INSPEC, Veterinary Science Database, Directory of Open Access Journals
  • Keywords: Sentinel-2, Landsat, surface reflectance, Virtual Constellation, Analysis Ready Data, harmonization, data fusion, Copernicus, SURFACE REFLECTANCE, BIDIRECTIONAL REFLECTANCE, CROSS-CALIBRATION, GENERAL-METHOD, LANDSAT, BRDF, PRODUCTS, MISSION, MSI
  • Hacettepe University Affiliated: Yes

Abstract

Satellite Earth Observation (EO) sensors are becoming a vital source of information for land surface monitoring. The concept of the Virtual Constellation (VC) is gaining interest within the science community owing to the increasing number of satellites/sensors in operation with similar characteristics. The establishment of a VC out of individual missions offers new possibilities for many application domains, in particular in the fields of land surface monitoring and change detection. In this context, this paper describes the Copernicus Sen2Like algorithms and software, a solution for harmonizing and fusing Landsat 8/Landsat 9 data with Sentinel-2 data. Developed under the European Union Copernicus Program, the Sen2Like software processes a large collection of Level 1/Level 2A products and generates high quality Level 2 Analysis Ready Data (ARD) as part of harmonized (Level 2H) and/or fused (Level 2F) products providing high temporal resolutions. For this purpose, we have re-used and developed a broad spectrum of data processing and analysis methodologies, including geometric and spectral co-registration, atmospheric and Bi-Directional Reflectance Distribution Function (BRDF) corrections and upscaling to 10 m for relevant Landsat bands. The Sen2Like software and the algorithms have been developed within a VC establishment framework, and the tool can conveniently be used to compare processing algorithms in combinations. It also has the potential to integrate new missions from spaceborne and airborne platforms including unmanned aerial vehicles. The validation activities show that the proposed approach improves the temporal consistency of the multi temporal data stack, and output products are interoperable with the subsequent thematic analysis processes.