Published January 27, 2021 | Version v1
Dataset Open

High resolution global industrial and smallholder oil palm map for 2019

  • 1. CREAF
  • 2. School of Biological and Environmental Sciences
  • 3. Borneo Futures
  • 4. Center for International Forestry Research
  • 5. Joint Research Centre

Description

The dataset contains 634 100x100 km tiles, covering areas where oil palm plantations were detected. The file 'grid.shp' contains the grid that covers the potential distribution of oil palm. The file 'grid_withOP.shp' shows the 100x100 grid squares with presence of oil palm plantations. The classified images (‘oil_palm_map’ folder, in geotiff format) are the output of the convolutional neural network based on Sentinel-1 and Sentinel-2 half-year composites. The images have a spatial resolution of 10 meters and contain three classes: [1] Industrial closed-canopy oil palm plantations, [2] Smallholder closed-canopy oil palm plantations, and [3] other land covers/uses that are not closed canopy oil palm. The file ‘Validation_points_GlobalOilPalmLayer_2019.shp’ includes the 13,495 points that were used to validate the product. Each point includes the attribute ‘Class’, which is the labelled class assigned by visual interpretation, and the attribute ‘predClass, which reflects the predicted class by the convolutional neural network. The ‘Class’ and ‘predClass’ values are the same as the raster files: [1] Industrial closed-canopy oil palm plantations, [2] Smallholder closed-canopy oil palm plantations, and [3] other land covers/uses that are not closed canopy oil palm.

See article for additional information:

Descals, Adrià, et al. "High-resolution global map of smallholder and industrial closed-canopy oil palm plantations." Earth System Science Data 13.3 (2021): 1211-1231.

 

Changelog v1:

- The analysis was extended to Sri Lanka, South India, and countries in Eastern Africa where oil palm can potentially grow.

- The validation dataset only includes the points drawn by simple random sampling and stratified random sampling in the grid cells where the IUCN industrial layer detected oil palm.

- The 'Class' and 'predClass' values in the validation dataset were reclassified with the same values as the raster images: [1] Industrial plantations, [2] Smallholder plantations, and [3] Other land covers/uses.

Files

High_resolution_global_industrial_and_smallholder_oil_palm_map_for_2019.zip

Files (101.8 MB)