3d vegetation as a predictor of erosivity

Rain splash, the initial stage of soil erosion by water, is influenced by both rainfall erosivity and soil surface erodibility. Vegetation alters rainfall erosivity—quantified as kinetic energy—by serving either as a protective, dispersive layer or as an amplifying, drip-aggregating layer. Previous research on vegetation's impact on throughfall kinetic energy (TKE) has primarily focused on point-based vegetation data and localized erosivity measurements. However, there is a growing need for spatially continuous, area-wide predictions of vegetation's effect on rainfall erosivity to enhance erosion modeling and conservation efforts. Recent studies emphasize the role of fine-scale tree structures in creating erosivity hotspots, known as drip points.

To address this gap, we employed lidar point clouds across multiple scales to investigate the relationship between 3D vegetation structure and TKE. UAV lidar data were used to derive vegetation cover and gap fractions within a voxel framework, identifying canopy layers that contribute leaf drips reaching the ground without re-interception. Furthermore, we linked field observations of active drip points to tree skeletons extracted from TLS point clouds to establish rules governing drip formation.

Our findings reveal that temperate forest vegetation's impact on erosivity surpasses values reported in prior studies focused on plantations. We observed a strong alignment between predicted and measured vegetation effects on TKE. We could demonstrate the potential of remote sensing for comprehensive, wall-to-wall predictions of vegetation's influence on rainfall erosivity. On the tree scale, lidar can improve our understanding of stemflow and re-interception dynamics on vegetation surfaces. These detailed findings can be scaled up to enhance landscape-level erosion predictions. Overall, lidar technology offers a promising solution to bridging data gaps in conventional erosion research.