ForestCO2 project: Canopy Height data

This data was produced as part of the ForestCO2 project (EO and in situ based information framework to support generating Carbon Credits in forestry), which was funded by the Science Fund of the Republic of Serbia, grant no. 6686.

The Canopy Height Map (CHM) for Serbia was produced using the model from https://www.nature.com/articles/s41559-023-02206-6, applied to the 5 images with the least cloud cover from the period May-September for each year in the 2020-2024 period. Final predictions for each of the 5 years were obtained by averaging the predictions from the individual models for all images of that year. The output was then masked to exclude pixels with CH<5m and to exclude pixels with no vegetation (for this, the SCL maps from the Sentinel-2 L2A products were used).

Forest Above Ground Biomass (AGB) is modelled by adapting the power equation from the paper https://essd.copernicus.org/articles/15/4927/2023 where the authors estimated yearly AGB for France using the tree canopy height. We have slightly adapted the coefficient from the paper to produce the Serbia AGB map, with the final equation for relating the height and biomass of a single tree being Biomass = 0.6 x H^1.9. 

Forest disturbance is derived from the 2020 and 2024 CHM layers by subtracting the latter from the former to produce the total canopy height change for the time period. The change layer is filtered to exclude the pixels with change smaller than 10m, to exclude the changes caused by agricultural activities and the Canopy Height model (in)accuracy.