Mapping Ghana’s forest and savanna ecosystems with drones

Last week I visited our rainforest-savanna transect in Ghana. We have been monitoring ecosystem carbon cycling and plant traits there for
the last few years, and currently are tracking the plots through this hot and dry El Niño year.

This time I was visiting with colleagues from Wageningen University in the Netherlands, in particular Harm Bartholomeus and Juha
Suomalainen. Our purpose was to map our plots with Unmanned Aerial Vehicles (UAVs or drones). The main instrument was a hyperspectral sensor aiming to map the forest canopy at wavelengths ranging from 450 nm to 950 nm. We hope to relate this to the
canopy leaf traits (leaf thickness, water content and chemistry, and photosynthesis rates) that we have been monitoring through
painstaking climbing of canopy trees and collection of branches. In our work in Peru we have been exploring the links with airborne
hyperspectral remote sensing – in Ghana we are exploring the potential of drones to collect similar data, albeit at a much smaller spatial
scale. In addition to this main drone, the team had a smaller, niftier Phantom drone, which was able to provide wonderful photographic
mapping and video footage.

In addition, another team from Wageningen have been scanning the plots with a 3D Terrestrial Laser Scanner (TLS), joined at the Ankasa rainforest site by Mat Disney’s team from UCL. Overall, we hope to have detailed understanding of the canopy and structure of almost every tree in our study plots.

The work is in collaboration with the Forest Research Institute of Ghana (Dr Stephen Adu-Bredu), and funded by my European Research Council Advanced Investigator Award, GEM-TRAIT, with the traits work supported by a Royal Society-Leverhulme Africa Award.