Non-destructive monitoring of grassland canopy height using a UAV

Grasslands cover one third of the agricultural area in Europe and provide a substantial part of the ruminant feed requirements in the dairy and meat industry. Furthermore, these ecosystems are very effective in preventing soil erosion, building soil fertility, require minimal pesticide use, and contribute multiple positive effects on the environment. Permanent grasslands, and the species they contain, have evolved spontaneously to adapt to the pedo-climatic conditions in which they occur. Hence, these different ecotypes are crucial in breeding under climate change scenarios. One of the main component species in North-West European grasslands is Lolium perenne (perennial ryegrass). Due to its rapid growth and establishment, high yield potential and feed quality, perennial ryegrass is also the most frequently sown grassland species in Europe.
We are currently developing a methodology, based on the use of unmanned aerial vehicles (UAVs) combined with different (spectral range) sensors, to monitor perennial ryegrass growth in a non-destructive way. The field trial used for this is located in Merelbeke (Belgium), and is part of GrassLandscape, a FACCE-JPI ERA-NET+ project. In this trial, 500 accessions sampling the natural diversity of the species across Europe and North Africa are grown in replicated micro-swards. As this set of accessions displays a very broad range of genetic and phenotypic diversity, including growth characteristics, it is an ideal target for the development of the methodology. In addition, the results obtained are directly useful to describe the growth characteristics of accessions adapted to different climatological conditions and their value for breeding purposes.
Frequent flights at low altitudes are performed to collect RGB imagery at high spatial and temporal resolution throughout the growing season to estimate canopy height and its temporal evolution. These data are combined with manual on-field measurements of canopy height, performed with an Herbometer, for validation. Correlations between the UAV data and on-field measurements will be calculated to determine the robustness and applicability of the methodology. Finally, canopy height data derived from UAV flights in combination with vegetation indices will be used to estimate the biomass volume and yield.