Three things side by side: Google satellite imagery, the MapSPAM 2010 coffee grid, and the extracted points with the Google coffee probability. Lab index.
Basemap
Esri imagery needs no key.
MapSPAM 2010 coffee, 9 km cells
Opacity0.65
0 ha
Modelled statistics (IFPRI SPAM), not field truth: census totals spread over 5 arc-minute cells, about 8,600 ha each. Click a cell for arabica, robusta, banana and tea.
Extracted points, Google coffee probability
Min probability0.00
00.51
Probability from Google's Forest Data Partnership coffee model (2026a release, 2024 layer), sampled at the points when they were extracted. Grey points were never sampled from that layer: drawn polygons, NAEB tea plantations, CIAT annual crops. Click a point for its class and source.
Boundaries
Credits: imagery Google / Esri; MapSPAM 2010 v2.0 IFPRI, CC-BY 4.0; coffee probability "Produced by Google for the Forest Data Partnership", CC-BY 4.0; districts national layer.