Unmanned aerial vehicles for pre-harvest biomass estimation in willow (Salix spp.) coppice plantations

Bioenergy crops provide a form of renewable energy, contributing to reductions in greenhouse gas emissions. Short rotation coppice (SRC) bioenergy crops, including willow (Salix spp.) and poplar (Populus spp.) are planted commercially in the UK and more widely across Europe for use in heat and power generation [1] often on set-aside or arable agricultural land. However, available data on yield of such crops, in terms of available biomass pre-harvest and spatial variations in yield, are limited [1]. Identifying the best time to harvest crops and accurately estimating yield before harvest is vital as harvesting operations comprise a major component of the costs in SRC production [2]. A range of methods have been proposed to provide above-ground biomass or basal area estimates in willow or poplar SRC plantations, including measurements of light attenuation through the canopy [2], terrestrial laser scanning [3] and field inventory with use of allometric equations [e.g.1, 4]. However, such methods are time consuming and limited in their ability to detect and map spatial variations in yield within the crop or to detect changes in crop health. Small unmanned aerial vehicles (UAVs) are able to provide high spatial resolution imagery and have potential to be flexibly deployed as needed, providing ‘on-demand’ imagery over small areas. Such systems are increasingly being deployed in environmental monitoring and in agriculture, operating with a variety of sensor types. Photogrammetric methods, utilising data from UAVs, have been used in past studies to derive forest height [5, 6], whilst radiometric measurements obtained from modified consumer-grade cameras or multispectral sensors have been used to calculate vegetation indices and estimate arable crop development and health [e.g. 7, 8]. However, little research has been undertaken in the application of low-cost UAV-based sensor systems in monitoring and estimating biomass and growth of SRC crops. This paper tests the use of a small unmanned aerial vehicle, equipped with consumer-grade red, green, blue (RGB) and near-infrared cameras, for estimating the biomass of willow (Salix spp.) SRC based on canopy height information and normalised difference vegetation index (NDVI). The paper aims to determine the ability of such approaches to detect differences in pre-harvest biomass in different willow varieties and to compare the obtained estimates with those from terrestrial laser scanning and field inventory.