Multi-source remote sensing data for forest monitoring

This dataset provides a multi-source collection of auxiliary variables designed to support forest monitoring and model development. It integrates remote sensing data from Sentinel-2, Landsat, and derived products, delivering consistent wall-to-wall coverage across seven European countries (Austria, Czechia, Finland, Germany, Italy, Norway, and Sweden). The dataset includes spectral and change-related predictors derived through advanced compositing and time-series processing methods. Rather than offering direct forest-change maps, it provides harmonized satellite-based predictors that can be readily used for model training, validation, and analysis. Applications include assessing forest dynamics, estimating biomass, monitoring canopy recovery, conserving biodiversity, tracking carbon, and assessing climate change impacts.

The Zenodo record contains a curated subset of the dataset, including wall-to-wall Landsat medoid composites for 2023, Landsat-derived change metrics, forest mask layers, forest type maps, selected Sentinel-2 and Sentinel-1 products, terrain variables, and the supplementary document "Complete_Dataset_Access_Links.pdf". 

Because the complete archive comprises multiple terabytes of data, the supplementary PDF provides direct public access links to the full country-level archives. These archives include all Sentinel-2 annual medoid composites (2017–2023), Sentinel-1 annual composites (2017–2024), Landsat annual medoid composites (1984–2023), Landsat-derived change metrics, forest mask layers, forest type maps, and terrain variables described in the associated publication.

Together, the files deposited in Zenodo and the publicly accessible archives referenced in the supplementary PDF provide access to the complete dataset.