Heathland above-ground biomass estimation using UAV and LiDAR

Zeren Yang , Jill Kowal, Elena Arrigoni, Tim Wilkinson, Elloise Budd, Justin Moat, Amanda Cooper

1. Heathland is an important carbon sink and habitat for biodiversity. It is heavily managed Heathland is an important carbon sink and habitat for biodiversity. It is heavily managed through grazing and prescribed fire and faces increasing threats from wildfire and forest encroachment. Because the proportion of overall heathland carbon stock is predominately below-ground, above-ground biomass (AGB) is often neglected. Quantifying both AGB and soil organic carbon (SOC) is therefore important for understanding total heathland carbon storage. 2. This study was conducted at Thursley National Nature Reserve, UK. We evaluated the effectiveness of hand-held Light Detection and Ranging (LiDAR) and unmanned aerial vehicle (UAV) Structure-from-Motion (SfM) for estimating above-ground biomass (AGB) and associated carbon stocks across dry (Calluna vulgaris) and wet (Erica tetralix) heathlands with different stand ages and fire histories. Vegetation was scanned before destructive biomass sampling, and the resulting three-dimensional (3D) point clouds were processed to derive structural metrics using canopy height models (CHMs) and voxelization. These metrics were then used to develop and evaluate regression models for AGB estimation, calibrated against the destructively sampled AGB. 3. In dry heathland, C. vulgaris stands aged 5, 19 and over 25 years had 95th-percentile canopy heights (H95) of 0.27 ± 0.01, 0.46 ± 0.04 and 0.76 ± 0.05 m, respectively, and AGB of 5.01 ± 0.55, 14.31 ± 2.19 and 31.92 ± 2.20 t ha⁻¹, respectively. Volumetric metrics provided better model fits than canopy height alone, with optimal resolutions of 20 mm for LiDAR voxel-derived and SfM CHM-derived volume, and 5 mm for LiDAR CHM-derived volume. LiDAR-derived models showed little evidence that biomass–structure relationships differed between C. vulgaris-dominated dry and E. tetralix-dominated wet heath, whereas the height-based model showed marginal evidence of heath-type-dependent relationships. 4. SOC were significantly greater in older dry (81.52±9.99 t ha⁻¹) than younger dry heath (56.08±6.27 t ha⁻¹) and were highest in wet heath (181.97±28.08 t ha⁻¹). Synthesis and applications: Three-dimensional structural metrics provide a non-destructive approach for monitoring AGB and associated carbon stocks in lowland heathlands. This method could enable repeated assessments of vegetation structure and biomass in response to management, while soil carbon measurements highlight the importance of considering below-ground carbon storage when evaluating heathland management strategies.