This dataset is part of a Publisher Catalog, and not managed by Google Earth Engine. Contact Land & Carbon Lab for bugs or view more datasets from the Land & Carbon Lab Catalog. Learn more about Publisher datasets.
- Catalog Owner
- Land & Carbon Lab
- Dataset Availability
- 2001-01-01T00:00:00Z–2024-01-01T00:00:00Z
- Dataset Provider
- World Resources Institute Google DeepMind
- Contact
- Land & Carbon Lab
- Tags
Description
This dataset maps the dominant driver of tree cover loss from 2001-2023 globally at 1 km resolution. Produced by the World Resources Institute (WRI) and Google DeepMind, the data were developed using a global neural network model (ResNet) trained on a set of samples collected through visual interpretation of very high-resolution satellite imagery. The model used satellite imagery (Landsat 7 & 8, Sentinel-2) and ancillary data to classify seven driver categories: permanent agriculture, hard commodities, shifting cultivation, logging, wildfires, settlements and infrastructure, and other natural disturbances. An independent stratified random sample collected through interpretation of very high resolution satellite imagery was used to estimate the accuracy of the map.
A driver is defined as the direct cause of tree cover loss, and can include both temporary disturbances (natural or anthropogenic) or permanent loss of tree cover due to a change to a non-forest land use (e.g., deforestation). The dominant driver is defined as the direct driver that caused the majority of tree cover loss within each 1 km cell over the time period. Classes are defined as follows:
- Permanent agriculture: Long-term, permanent tree cover loss for small- to large-scale agriculture. This includes perennial tree crops, as well as pasture and seasonal crops and cropping systems, which may include a fallow period. Agricultural activities are considered "permanent" if there is visible evidence that they persist following the tree cover loss event and are not a part of a temporary cultivation cycle.
- Hard commodities: Loss due to the establishment or expansion of mining or energy infrastructure.
- Shifting cultivation: Tree cover loss due to small- to medium-scale clearing for temporary cultivation that is later abandoned and followed by subsequent regrowth of secondary forest or vegetation.
- Logging: Forest management and logging activities occurring within managed, natural or semi-natural forests and plantations, often with evidence of forest regrowth or planting in subsequent years. Includes clear-cut and selective logging, establishment of logging roads, forest thinning, and salvage or sanitation logging.
- Wildfire: Tree cover loss due to fire with no visible human conversion or agricultural activity afterward. Fires may be started by natural causes (e.g. lightning) or may be related to human activities (accidental or deliberate).
- Settlements and infrastructure: Tree cover loss due to expansion and intensification of roads, settlements, urban areas, or built infrastructure (not associated with other classes).
- Other natural disturbances: Tree cover loss due to other non-fire natural disturbances (e.g., landslides, insect outbreaks, river meandering). If loss due to natural causes is followed by salvage or sanitation logging, it is classified as forest management.
Limitations: This product does not distinguish between the loss of natural forest and planted trees (e.g., plantations, tree crops, or agroforestry systems). While tree cover loss associated with the permanent agriculture, hard commodities, and settlements and infrastructure classes represent a close approximation of deforestation (permanent conversion of forest to another land use), these classes may sometimes include the clearing of planted trees. For example, clearing and replanting an orchard would be included in the permanent agriculture class, but is not deforestation of a natural forest. Similarly, replacement of natural forest with wood fiber plantations is not distinguished from routine harvesting within existing plantations established before 2000, as these are both included in the logging class.
This product shows the dominant driver in each 1km cell over the entire period. It does not show multiple drivers if they occur in the same cell at smaller scales, nor does it detail the sequence of drivers if multiple occurred at different times within the period. Additionally, these data are limited in scope to attributing drivers to tree cover loss as mapped by the Global Forest Change v1.11 tree cover loss product, and therefore the detection of loss is subject to the accuracy of that product.
For a full description of the methods, technical specifications, definitions, accuracy, and limitations, please see the publication: https://doi.org/10.1088/1748-9326/add606. The data is also available for download on Zenodo and the WRI Data Explorer.
Bands
Pixel Size
1111.95 meters
Bands
| Name | Min | Max | Scale | Description |
|---|---|---|---|---|
classification |
1 | 7 | Most likely class based on raw probabilities. |
|
probability_1 |
0 | 250 | 0.004 | Probability of "Permanent agriculture" class (scaled to [0-250]). |
probability_2 |
0 | 250 | 0.004 | Probability of "Hard commodities" class (scaled to [0-250]). |
probability_3 |
0 | 250 | 0.004 | Probability of "Shifting cultivation" class (scaled to [0-250]). |
probability_4 |
0 | 250 | 0.004 | Probability of "Logging" class (scaled to [0-250]). |
probability_5 |
0 | 250 | 0.004 | Probability of "Wildfire" class (scaled to [0-250]). |
probability_6 |
0 | 250 | 0.004 | Probability of "Settlements and infrastructure" class (scaled to [0-250]). |
probability_7 |
0 | 250 | 0.004 | Probability of "Other natural disturbances" class (scaled to [0-250]). |
classification Class Table
| Value | Color | Description |
|---|---|---|
| 1 | #E39D29 | Permanent agriculture |
| 2 | #E58074 | Hard commodities |
| 3 | #E9D700 | Shifting cultivation |
| 4 | #51A44E | Logging |
| 5 | #895128 | Wildfire |
| 6 | #A354A0 | Settlements and infrastructure |
| 7 | #3A209A | Other natural disturbances |
Terms of Use
Terms of Use
Citations
Sims, M., Stanimirova, R., Raichuk, A., Neumann, M., Richter, J., Follett, F., MacCarthy, J., Lister, K., Randle, C., Sloat, L., Esipova, E., Jupiter, J., Stanton, C., Morris, D., Slay, C. M., Purves, D., and Harris, N. (2025). Global drivers of forest loss at 1 km resolution. Environmental Research Letters. doi:10.1088/1748-9326/add606
Explore with Earth Engine
Code Editor (JavaScript)
Map.setCenter(-9.22,20.65,3) var drivers = ee.Image('projects/landandcarbon/assets/wri_gdm_drivers_forest_loss_1km/v1_1_2001_2023'); var drivers_class = drivers.select(['classification']); var vis = { "min":1, "max": 7, "palette": ['E39D29','E58074','e9d700','51a44e','895128','a354a0','3a209a'] }; Map.addLayer(drivers_class, vis, 'Drivers of Forest Loss, 2001-2023'); var permAg_prob = drivers.select(['probability_1']); //Select a probability band var probVis = { min: 0, max: 250, palette: ['#440154','#481567','#482677','#453781','#3b528b','#2c728e','#21908d','#27ad81','#5ec962','#aadc32','#fde725'] }; Map.addLayer(permAg_prob, probVis, 'Probability band for permanent agriculture', false);