PB280450.JPG

Trees outside forests in global drylands (TOFDRY)

ERC starting grant




TOFDRY aims to quantify the worlds non-forest trees by using PlanetScope satellite imagery and a deep learning technique which is able to identify objects within imagery at unprecedented accuracy.

Contact: Florian Reiner




 
Screenshot-20210626090533-1574x539.png

PlanetScope




We form global mosaics at 3 m resolution from millions of PlanetScope scenes having a size of several hundred terabyte.




Screenshot-20210503204242-1758x728.png

Deep Learning




We develop deep learning models being able to identify trees and groups of trees - at continental scale and with a precision at tree level.




DSC03462.JPG

Tree cover maps at unprecedented accuracy




Tree cover maps at 1-m resolution will be linked with various variables, like soil fertility, agroforestry, carbon stocks, climate, etc.




 
erc.png