RAINER – Remote sensing based area wide prediction of forest vegetation influence on rainfall erosivity

• Applicant: Johannes Antenor Senn, Ph.D.
• Subject Area: Physical Geography
• Term: since 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 568805599

Soil loss is commonly associated with farmland; however, considerable erosion rates can occur even under closed forest vegetation. While the effect of vegetation on erosion processes is well-studied in uniform agricultural vegetation, such as vineyards, corn crops or rubber plantations, it remains underexplored in diverse forest ecosystems. Dispersion, accumulation and detachment of raindrops on plant surfaces alter the drop size distribution. This can, depending on the falling height, either amplify or reduce the kinetic energy of raindrops. Capturing the determinant vegetation traits or the resulting kinetic energy in point measurements under forest canopies is challenging due to the pronounced spatial variations in these processes. However, there is a need for methods to identify areas of increased erosion potential as a function of vegetation characteristics and a need for spatially continuous information on erosion potential as an input for spatially explicit erosion models. RAINER aims to advance the understanding of the effect of temperate forest vegetation on the kinetic energy of rainfall. The knowledge gained will be applied to validate and optimise a spatially continuous modelling approach for predicting the vegetation effect based on remotely sensed 3D structure. The research area is located in the südliche Hardt forest near Bruchhausen and comprises four field sites that are representative of the most widespread stand dominating broad-leaved and coniferous tree species in Central Europe. We will conduct a large-scale, two-year field campaign to gain reliable insights into the complex interplay of difficult-to-measure and highly variable parameters. We will collect three-dimensional structural data from a multi-temporal UAV-based lidar survey that we will apply in a modelling approach. The vegetation splash factor model identifies vegetation surfaces that contribute drips to uninterrupted throughfall to the open soil and vertically aggregates their protective or amplifying effect for each grid cell. The model results will be validated with observed differences between the kinetic energy of open rainfall and throughfall. This data will be collected using sand-filled splash-cups. In addition, we will take point-based field recordings of vegetation traits to ensure continuity with existing research practice. Replacing the simplified categorical crop factor values commonly used in erosion models with a detailed spatially continuous prediction of the effect of forest vegetation on erosivity can be highly beneficial in times of considerable shifts in precipitation.

• DFG Programme: Research Grants