Skip to main content
Leaf uses Sentinel-2 L2A data (atmospherically corrected surface reflectance) from the Copernicus program. Sentinel-2 revisits each field every 3 to 5 days and is included at no additional cost.

Images produced per pass

Leaf generates 25 images for each satellite pass over a field:
NameResolutionTypeProjection
RGB.tif10 mTrue-colorEPSG:4326
RGB.pngTrue-colorEPSG:3857
NDVI.tif10 mRaw NDVI valuesEPSG:4326
NDVI_color.tif10 mColorized NDVIEPSG:4326
NDVI.pngColorized NDVIEPSG:3857
NDVI_relative.pngNDVI scaled to image min/maxEPSG:3857
NDVI_absolute.pngNDVI scaled to -1 to 1EPSG:3857
NDRE.tif10 mRaw NDRE valuesEPSG:4326
NDRE_color.tif10 mColorized NDREEPSG:4326
NDRE.pngColorized NDREEPSG:3857
NDRE_relative.pngNDRE scaled to image min/maxEPSG:3857
NDRE_absolute.pngNDRE scaled to -1 to 1EPSG:3857
multi_band.tif10 mAll bands stackedEPSG:4326
B01.tif60 mCoastal aerosolEPSG:4326
B02.tif10 mBlueEPSG:4326
B03.tif10 mGreenEPSG:4326
B04.tif10 mRedEPSG:4326
B05.tif20 mVegetation red edgeEPSG:4326
B06.tif20 mVegetation red edgeEPSG:4326
B07.tif20 mVegetation red edgeEPSG:4326
B08.tif10 mNIREPSG:4326
B8A.tif20 mVegetation red edgeEPSG:4326
B09.tif60 mWater vapourEPSG:4326
B11.tif20 mSWIREPSG:4326
B12.tif20 mSWIREPSG:4326
PNG files have no fixed resolution because they are upscaled 800% for display.

NDVI output formats

Leaf produces three NDVI PNG variants, each useful for different purposes:
  • NDVI_relative — Scales the color ramp between the minimum and maximum values of that specific image. Good for seeing variation within a single date.
  • NDVI_absolute — Scales from -1 to 1 with the full color ramp. Good for comparing across different dates.
  • NDVI (plain) — Same -1 to 1 range, but values below 0 are rendered as bright red. This is the recommended default for most use cases.
The same three variants exist for NDRE. If you want to apply your own color ramp, use the raw NDVI.tif or NDRE.tif files. These contain pre-calculated index values and can be imported into any GIS tool like QGIS.

Cloud masking

Leaf uses the cloud mask provided with Sentinel-2 L2A data to calculate the cloud coverage percentage for each image. You can filter images by maxClouds when querying processes.

What to do next

Last modified on March 24, 2026