Research Of Slope Micro-topography DEM Based On Terrestrial Laser Scanning

Technology of terrestrial laser scanning (TLS) has the ability to collect the target object’s3D point cloud with high efficiency and precision. On the other hand, micro-topography hasthe characters that the elevation of terrain surface is non-significant and the variation ofterrain surface caused by soil erosion is inconspicuous. Therefore, TLS has a huge advantagein monitoring erosion on micro-topography. The point cloud collected from TLS can reflectsubtle features of terrain surface. Meanwhile the method of non-contact survey ensured theintegrity of the micro-topography. The high resolution DEM interpolated from point cloudincreased the accuracy of micro-topography and can easily get the quantification of terrainfactor.Three slope runoff plots were set up in this study to simulate three kinds ofmicro-topography: PL1with plowing depths30cm; PL2with plowing depths20cm; PL3with bare slope. Topcon GLS-15003D laser scanning was used for surveying the slopemicro-topography.The blindness and randomness of the scanning and the large data volume may lead to hugedata calculation, high hardware requirements and data inaccuracy. The study started withprecision of micro-topography analysis. Horizontal resolution and point cloud density wereselected to analyze the effect on the DEM precision in the DEM generative process. Themethod to select the optimized horizontal resolution and the optimized point cloud densitywere also discussed.Three indexes including slope root mean square error, surface roughness and time wereselected to analyze the relationship with micro-topography DEM resolution. The resultsshowed that slope RMSE could quickly narrow the select range, simplify the data processingand make the aim of future research more clear. With the increase of resolution, the variationtrend of PL2and PL1was similar and soil roughness has a relationship of power functionwith resolution. Considering the soil roughness and time of DEM generating,4mm wasselected as the optimized horizontal resolution.Three indexes including mean error, standard error and root mean square error wereselected for the research of the optimized data density. Different point cloud density levelsand original point cloud density indexes were analyzed; the lowest point cloud density20% was selected as the best point cloud density.Micro-topography factors including slope and aspect were extracted and analyzed tostudy the time and space variation in different micro-topography factors.