LiDAR-Derived DEM and Raw Height Comparisons along Profile Corridor Gradients within a Forest

Abstract

We compared field based and airborne LiDAR-derived profile corridor measurements across forest canopy types and terrain ranging from 37% to 49% slope. Both LiDAR-derived DEM and raw LiDAR point elevations were compared to field data. Primary objectives included examining whether canopy type or terrain slope influenced LiDAR-derived pro- file measurements. A secondary objective included comparing cable logging payloads based on field measured profile elevations to payloads based on LiDAR-derived elevations. Average RMSE elevation errors were slightly lower for profile point to LiDAR DEM values (0.43 m) than profile point to nearest LiDAR elevation point (0.49 m) with differ- ences being larger when sites within forest clearings were removed from analysis. No statistically significant relation- ship existed between field measured ground slopes and associated profile point and LiDAR DEM elevation differences but a mild correlation existed when LiDAR raw point elevation differences were compared. Our payload analysis de- termined the limiting payload distance and had consistent results across study sites. The DEM-based profile outper- formed the nearest point profile by 5% on average. Results suggest that forest analysts should consider using the nearest LiDAR DEM value rather than the nearest LiDAR point elevation for terrain heights at discrete locations, particularly when forest canopy occludes locations of interest.