Overview
Global estimates of land use/land cover (LULC), derived from ESA Sentinel-2 imagery at 10m resolution. This dataset was generated by Impact Observatory, who used billions of human-labeled pixels (curated by the National Geographic Society) to train a deep learning model for land classification. The global maps were produced by applying this model to the Sentinel-2 scene collection on the Planetary Computer.
All of the annual LULC datasets are available under a Creative Commons BY-4.0 license.

10m Annual Land Use Land Cover (9-class) V2
Global land cover information with 9 classes for 2017-2023 at 10m resolution

Esri 10-Meter Land Cover (10-class)
Global land cover information for 2020 at 10m resolution

10m Annual Land Use Land Cover (9-class) V1
Global land cover information with 9 classes for 2017-2022 at 10m resolution