FAQ

1. Why LEAF?

LEAF stands for Landscape Evolution and Forecasting. The toolbox was designed to meet the need to provide information on land surface condition as it changes in TIME and SPACE. At present it mainly does spatial mapping but we are working on temporal reanalysis and forecasting.

2. Is LEAF free and open?

Almost. LEAF is distributed under MIT license to control liability and attribution. LEAF uses Google Earth Engine. Currently Google Earth Engine is open for most use and free for most non-profit use. You must comply with the terms of use of Google Earth Engine.

3. Is LEAF the same as the ESA SNAP implementation of the Simplified Level 2 Prototype Processor (SL2P) for Sentinel 2 and Landsat data?

No.
LEAF is a correct implementation of SL2P ( Weiss and Baret, 2016).
The SNAP 'Biophysical Processor' had an incorrect implementation of the sampling used to define the calibration database (code available upon request). We can provide code used for both on request.

4. Have LEAF toolbox prediction algorithms been validated?

Yes. See references.
We would appreciate you flag any bugs or validation results by citing them as an Issue.

5. Can I have access to the prediction algorithms for my own software?

Yes. They use neural networks whose coefficients are coded as GEE assets.

6. How do I attribute use of leaf?

Fernandes, R. et al., 2021, "LEAF Toolbox", Canada Centre for Remote Sensing, https://github.com/rfernand387/LEAF-Toolbox/wiki, DOI: 10.5281/zenodo.4321298. (full list of names in footer!)

If you derived products using the SL2P algorithm (the default currently) please cite Weiss and Baret, 2016

Additionally cite the data providers for the input data: [Sentinel 2] https://sentinels.copernicus.eu/documents/247904/690755/Sentinel_Data_Legal_Notice), Landsat.