Natural disturbances such as the 2018 windthrow Vaia and subsequent bark beetle outbreaks can strongly affect the protective function of mountain forests against rockfall. This study evaluated the relationship between the volume of horizontal deadwood, placed as natural barriers and their resulting terrain roughness relevant for rockfall mitigation, with the goal of improving the representation of biological legacies in rockfall modelling. We also assessed how much roughness changes in case study areas over time, thus analysing the effect of decay and settling on these structures. To achieve this, we combined UAV-based remote sensing with field surveys to quantify deadwood volume and monitor temporal changes. The RGB-based vegetation index (RGBVI) proved to be an effective and robust tool for improving the accuracy of roughness quantification derived from deadwood structures, even under variable lighting and sensor conditions. Our Models, particularly the logarithmic model for RG_10, which includes a saturation effect beyond approximately 400 m3/ha of deadwood, can predict the resulting roughness (RG_10 and RG_20) in stands with manually placed logs and fresh windthrown conditions. Six years after the Vaia event, surface roughness remained high in unmanaged areas, especially where horizontal deadwood volume exceeded 200 m3/ha.