Deriving shrub biomass and carbon from affordable UAV observations

Measuring shrub cover and above ground biomass is important for habitat condition and carbon monitoring, particularly for early-successional woodland. Advances in unmanned aerial vehicle (UAV) remote sensing and artificial intelligence are creating opportunities to complement field-based surveying or provide effective alternatives.

While there is a wealth of biomass calculations and allometric equations available for trees, there is a contrasting lack of this information for shrubs. Here we show results combining a Maximum Entropy allometric model using Bayesian inference developed from destructive sampling, a U-NET deep learning model, and UAV imagery structure-from-motion, to identify individual hawthorn shrubs, extract shrub height and crown diameter and derived shrub biomass and carbon. Streamlining these steps into an accessible pipeline could result in an effective and affordable solution for shrub biomass mapping.