Land Cover 9-Class

Impact Observatory — version 1.2 (2025)

Description

This dataset classifies land cover at 10 m spatial resolution, with global spatial coverage annually from 2017 onwards. Pixels are assigned one of 9 land cover classes.

Dataset ID
a4bb9aea
Category
Land use & land cover
Type
Raster
CRS
EPSG:3857
Spatial coverage
Global
Spatial resolution
10 m
Temporal coverage
2017+
Temporal resolution
Annual

Usage notes

New time points are delivered in January for the previous year. Data are delivered as annual composites of predictions made on multiple Sentinel-2 observations.

Variables

land_cover_class

Typeuint8

Description

Index of the predicted land cover class. Predictions are made using a deep learning segmentation model trained to predict > 5 billion manually annotated land cover classifications from Sentinel-2 imagery.

Usage notes

Grass and Shrub classes are merged into a single Rangeland class (index 11 ). Overall accuracy is 76% compared to a majority consensus. Lowest accuracy is for Flooded Vegetation (57%) and Rangeland (58%). Highest accuracy is for Crops (90%) and Water (87%).

IndexNameDescription
1WaterWater with little to no accompanying vegetation, rock, outcrops, or infrastructure. May not include sporadic or ephemeral water bodies.
2TreesA clustering of tall dense vegetation with closed or dense canopy.
4Flooded VegetationAny vegetation mixed with water throughout most of the year, including seasonally flooded areas.
5CropsCultivated cereals, grasses, or other non-tree crops.
7Built AreaHuman-made structures, including rail/road networks, buildings, and large impervious surfaces (e.g. parking lots).
8Bare GroundRock, soil, or sand with little to no vegetation for the entire year.
9Snow/IceHomogenous areas of permanent snow or ice. Typically only found in mountainous and high latitude regions.
10CloudsNo land cover information due to persistent cloud cover.
11RangelandLow vegetation with little to no taller vegetation. Includes patchy landscapes of grass, shrubs, scrub, and bare ground.

Pricing

This dataset has no data acquisition cost.

Resources

Provider

Impact Observatory

www.impactobservatory.com

Impact Observatory is a nature data analytics company based in Washington DC. Impact Observatory uses AI-based approaches to build land cover products that inform environmental analysis for governments, industries, and markets.