Airborne LiDAR-derived elevation data in terrain trafficability mapping

Heavy off-road traffic causes soil compaction and rutting, which can significantly reduce the yield of forest stands. Reliable information on terrain trafficability, that is, the ability of terrain to support the passage of vehicles, would enable significant enhancement of wood procurement planning and reduction of soil damage. The objective here was to determine the feasibility of airborne scanning light detection and ranging (LiDAR)-derived digital terrain models (DTM) in terrain trafficability mapping. Soil damage was inventoried from a total of 13 km of forwarding trails, and a logistic regression model was fitted for predicting the event of soil damage. DTM-derived soil wetness indices performed well as predictor variables, and DTM-derived local binary patterns also proved useful in terrain trafficability mapping. A prediction accuracy of 83.6% (Cohen's kappa of 0.38) was observed for soil damage probability modelling, using only DTM-derived predictors, and a corresponding accuracy of 85.0% (kappa of 0.45) was achieved when an existing soil map was used as well. In addition to the topography-related features, soil stoniness proved to be a critical factor affecting soil resistance to rutting. Our results indicate that the utilisation of LiDAR-derived elevation data for terrain trafficability mapping is a feasible method in sustainable forest management.