Description
This dataset classifies the land cover class present on the land surface at 10 m spatial resolution, with global coverage for custom time periods (minimum three months) from 2017 onwards. Pixels are assigned one of 15 land cover classes.
Usage notes
Data are delivered as composites of predictions made on Sentinel-2 observations over a date range. Custom date ranges are permitted, although annual composites are recommended to ensure sufficient Sentinel-2 observations for reliable predictions. Areas of interest (AOIs) must be > 2,500 ha and < 150,000 ha.
Licence
Commercial
Version
1.4
Release date
Made to order
CRS
EPSG:3857
Update schedule
Sub-weekly
Provider cost
$0.02/ha
Description
Class value provides the index of the predicted land cover class for a pixel. Predictions are made using a deep learning segmentation model trained to predict manually annotated land cover classifications from Sentinel-2 imagery.
Usage notes
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%).
Variable category
Land cover class
Variable type
Integer
Metadata & resources