Mapping rill soil erosion in agricultural fields with UAV‐borne remote sensing data

Soil erosion by water is a main form of land degradation worldwide. The problem has been addressed, among others, in the United Nations Sustainability Goals. However, for mitigation consequences and adequate management affected areas, reliable information on magnitude spatial patterns needed. Although such need often addressed modelling, precise monitoring necessary calibration validation models to study landscapes. Conventional methods quantification rill are based labour-intensive field measurements. In contrast, remote sensing techniques promise fast, non-invasive, systematic larger-scale surveying. Thus, objective this was develop evaluate automated transferable methodologies mapping extent rills from single acquisition data. Data collected an uncrewed aerial vehicle (UAV) used deliver highly detailed digital elevation model (DEM) analysed area. Rills were classified two with different settings. One approach series decision rules applied DEM-derived geomorphological terrain attributes. second utilized random forest machine learning algorithm. tested three agricultural fields representing vegetation covers. Our showed that proposed can assure recognition accuracies between 80-90% depending characteristics. some cases, however, sensitive very small incisions similar geometry other features. Additionally, their performance influenced structure cover. Besides these challenges, introduced capable fully automatically at scale can, therefore, support fast flexible assessment magnitudes.