The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
Download
Publications Copernicus
Download
Citation
Articles | Volume XLII-2/W16
https://doi.org/10.5194/isprs-archives-XLII-2-W16-201-2019
https://doi.org/10.5194/isprs-archives-XLII-2-W16-201-2019
17 Sep 2019
 | 17 Sep 2019

IMPROVEMENT OF EXISTING AND DEVELOPMENT OF FUTURE COPERNICUS LAND MONITORING PRODUCTS – THE ECOLASS PROJECT

E. Sevillano Marco, D. Herrmann, K. Schwab, K. Schweitzer, R. Almengor, F. Berndt, C. Sommer, and M. Probeck

Keywords: Copernicus Land Monitoring Service, ECoLaSS, Earth Observation, DIAS, HRL, LCLU, Sentinel, time series analysis

Abstract. The Horizon 2020 project ECoLaSS (Evolution of Copernicus Land Services based on Sentinel data) contributes to improving existing and developing next-generation Copernicus Land Monitoring Service (CLMS) products. The High Resolution Layers (HRLs) are currently produced in regular 3-year intervals at 10–20 meter spatial resolution for 39 European countries (EEA 39). Evolving scientific developments and user requirements are continuously analysed in a close stakeholder interaction process with the European Entrusted Entities (EEE), targeting a future pan-European roll-out of new/improved CLMS products and assessing transferability to global applications. Products and methods are being prototypically demonstrated. Representative sites (60,000–90,000 km2) were selected, covering boreal, Mediterranean, steppic, Atlantic, alpine and continental conditions. Improvements comprise yearly updates of enhanced dominant leaf types and tree cover change layers, better-quality permanent grassland classification and use categorisation. Novel products target agriculture products (i.e., crop mask, crop types). Temporal analysis, based on optical (Sentinel-2) and SAR (Sentinel-1) satellite data, makes use of temporal feature descriptors (multiple temporal statistical metrics) derived from spectral bands and indices (e.g., VV/VH ratio and NDVVVH from SAR data and NDWI, NDVI, Brightness and IRECI from optical data). Overall accuracies range from 77–98%. Rigorous benchmarking is applied to assess the prototypes’ operational readiness and technical maturity for integration into the CLMS architecture.