SBTN Natural Lands

WRI — version 1.1 (February 2025)

Description

This dataset classifies the land surface into natural and non-natural land cover classes for the year 2020, with global coverage at 30 m spatial resolution. Pixels are assigned one of 21 classes, alongside a binary index indicating whether it is natural or non-natural land.

Dataset ID
bf249342
Category
Land use & land cover
Type
Raster
CRS
EPSG:4326
Spatial coverage
Global
Spatial resolution
30 m
Temporal coverage
2020
Temporal resolution
Single time point

Usage notes

This dataset supports baseline assessments of natural land conversion for organisations setting science-based targets for nature (e.g. SBTN Land Target #1: no conversion of natural ecosystems). The dataset intentionally overestimates natural lands to protect from conversion. Variable spatial resolution of input data sources may create spatial artefacts at 30 m resolution. Some input data sources are also from earlier than 2020. A list of input datasets is found here (Appendix D).

Variables

binary_class

Typeuint8

Description

Indicates whether a pixel is natural or non-natural land. Classifications are tied to land cover class assignments.

Usage notes

Binary class is suitable for baseline assessments against SBTN Land Target #1 (no conversion of natural ecosystems). Overall reported accuracy is 92.1%, with ~ 6% of natural pixels being misclassified as non-natural land and ~ 15% of non-natural pixels being misclassified as natural land.

IndexName
0Non-natural
0Natural

class_value

Typeuint8

Description

The index of the land cover class assigned to a pixel. The data is a composite of > 30 global and regional datasets combined through a decision tree analysis to delineate natural versus non-natural land cover classes.

Usage notes

Overall reported accuracy is 92.1%. Forest land cover classes follow Accountability Framework initiative definitions. Other classes follow definitions of respective input data sources, so may be inconsistent between regions or ecosystem types. Variable spatial resolution of input data sources may yield spatial artefacts at 30 m resolution. A minority of input data sources are also from earlier than 2020. Natural forests may be overestimated in temperate and boreal regions. Small-scale mixed/rotational agriculture may be misclassified as natural land cover classes.

Binary classLand cover classIndexDescription
NaturalForest2Areas of tree cover (> 5 m height, > 0.5 ha). Excludes planted forests and perennial tree crops.
NaturalShort vegetation3Areas of short stature vegetation (< 5 m height) dominated by grass or shrubs. Excludes cultivated grasslands, cropland, and short tree crops.
NaturalWater4Areas where surface water is present at least 20% of the year.
NaturalMangroves5Areas dominated by mangrove forests.
NaturalBare6Exposed rock, soil, or sand with less than < 10% vegetation cover. Excludes livestock, cropland, or tree crops.
NaturalSnow/Ice7Areas of persistent glaciers or snow.
NaturalWetland forest8Natural forests with water-saturated soil, inundated either seasonally or permanently.
NaturalPeat forest9Natural wetland forests over peat.
NaturalWetland short vegetation10Natural short vegetation with water-saturated soil, inundated either seasonally or permanently.
NaturalPeat short vegetation11Natural short vegetation over peat.
Non-naturalCropland12Annual or perennial crop areas, including herbaceous crops, plantations, tree crops, pasture, and other agricultural land.
Non-naturalBuilt-up13Man-made surfaces for infrastructure, commercial, residential, and mining use.
Non-naturalTree cover14Perennial tree crops and planted forests. Includes short and long rotation timber forest plantations.
Non-naturalShort vegetation15Cultivated short vegetation (< 5 m height), (e.g. pasture, plantation crops, cultivated grasslands).
Non-naturalWater16Artificial surface water (e.g. aquaculture areas, or artificial dams).
Non-naturalWetland tree cover17Non-natural tree cover with water-saturated saturated soil, inundated either seasonally or permanently.
Non-naturalPeat tree cover18Non-natural wetland tree cover over peat.
Non-naturalWetland short vegetation19Non-natural short vegetation with water-saturated soil, inundated either seasonally or permanently.
Non-naturalPeat short vegetation20Non-natural short vegetation over peat.
Non-naturalBare21Exposed rock, soil, or sand areas with < 10% vegetation cover and a high density of ruminant livestock.

Pricing

This dataset has no data acquisition cost.

Resources

Provider

WRI

www.wri.org

The World Resources Institute (WRI) is a non-profit research organisation based in Washington DC. The WRI does research that addresses policy across food systems, land and water, energy, and cities to protect nature and mitigate climate change globally.

Land & Carbon Lab

landcarbonlab.org

The Land & Carbon Lab is a joint venture by the WRI and Bezos Earth Fund. Land & Carbon Lab provides geospatial data and insights for ecosystems and land to improve landscape restoration and sustainable use of resources.

WRI - SBTN Natural Lands