Influence of the precision of lidar data in surface water runoff estimation for road maintenance

Abstract

Roads affect the natural surface and subsurface drainage pattern of a hill or a watershed. Road drainage systems are designed with the objective of reducing the energy generated by the flowing water and the presence of excess water or moisture within the road. A poorly designed drainage may affect to road maintenance causing cut or fill failures, road surface erosion and degrading the engineering properties of the materials with which it was constructed. Surface drainage pattern can be evaluated from Digital Elevation Models typically calculated from point clouds acquired with aerial LiDAR platforms. However, these systems provide low resolution point clouds especially in cases where slopes with steep grades exist. In this work, Mobile LiDAR systems (aerial and terrestrial) are combined for surveying roads and their surroundings in order to provide complete point cloud. As the precision of the point clouds obtained from these mobile systems is influenced by GNSS outages, Gaussian noise with different standard deviation values is introduced in the point cloud in order to determine its influence in the evaluation of water runoff direction. Results depict an increase in the differences of flow direction with the decrease of cell size of the raster dataset and with the increase of Gaussian noise. The last relation fits to a second-order polynomial Differences in flow direction up to 42° are achieved for a cell size of 0.5 m with a standard deviation of 0.15 m.