Machine traffic during timber harvesting operations induces soil compaction, which is particularly evident in the formation of ruts. Visual inspection of rut formation is labor-intensive and limits the volume of data that can be collected. This study aims to contribute to the limited knowledge base regarding the extent of soil physical disturbance caused by machine traffic on steep slopes and to evaluate the utility of LiDAR and UAV photogrammetry techniques. The selected traffic trails included single-pass uphill, single-pass downhill, three-pass round trip, and five-pass round trip trails, with an average slope of 70.7%. Traditional methods were employed to measure rut depth using a pin board and to assess soil bulk density (BD) and soil porosity (SP) from soil samples. The results revealed that the average rut depth was 19.3 cm, while the deepest ruts were observed after a single pass (uphill: 20.0 cm; downhill: 22.7 cm), where BD and SP showed the most significant changes. This study provides a rare quantitative evaluation of the applicability of remote sensing methods in forestry by comparing surface height data collected via a pin board with that derived from a Mobile LiDAR System (MLS) and UAV photogrammetry using structure-from-motion (SfM). When compared to pin board measurements, the MLS data showed an R² value of 0.74 and an RMSE of 4.25 cm, whereas the SfM data had an R² value of 0.62 and an RMSE of 5.27 cm. For rut depth estimation, SfM (16.0 cm) significantly underestimated values compared to the pin board (19.3 cm) and MLS (19.9 cm). These findings not only highlight the potential and limitations of remote sensing methods for assessing soil disturbance in steep forest environments but also contribute to addressing the knowledge gaps surrounding the effects of soil compaction in steep terrain.

Mechanized timber harvesting on steep slopes causes soil disturbance; however, comprehensive post-harvest assessment remains challenging because terrain complexity and safety constraints render traditional field-based methods labor-intensive, spatially limited, and difficult to implement systematically. In this study, we developed and evaluated a convolutional neural network-based semantic segmentation model for detecting soil disturbances using high-resolution unmanned aerial vehicle (UAV) imagery in a steep-slope harvested area (2.50 ha, mean slope of 53.4%) in Republic of Korea. A U-Net semantic segmentation model was trained on manually annotated orthomosaic tiles incorporating RGB and digital elevation model (DEM) inputs. Ensemble predictions at an optimized threshold of 0.65 achieved Intersection over Union (IoU) of 0.55 and F1-score of 0.71. Although moderate, these values reflect the inherently challenging conditions of steep-slope forest terrain compared to similar studies conducted under gentler terrain. DEM-derived depth estimation enabled severity classification of the detected disturbances, with light disturbances predominating. Field validation using 38 pinboard measurements demonstrated reliable spatial detection (ρ = 0.567, RMSE = 6.45 cm). This approach provides an effective alternative to traditional monitoring practices in mountainous forests, where systematic trail planning is impractical, and may support evidence-based assessment of harvesting impacts for sustainable forest management.