Author(s): Felix Dacheneder and Jan Peter Balmes
Linked Author(s):
Keywords: Remote Sensing; SfM; Photogrammetry; River Restoration; Hybrid Numerical Model; CFD; LWD
Abstract:
New methods of remote sensing techniques such as unmanned aerial vehicles (UAV) and Structure from Motion (SfM) offer a solution for more comprehensive and spatially measurements to generate dense 3D point cloud data. This approach requires only a limited amount of ground control and is well suited to imagery acquired from low-cost, non-metric camera systems attached to UAV. We conducted a UAV campaign at an 850 m long restoration site at the river Boye, Germany, which is not yet hydraulically connected to the main channel. The aim is to generate different bathymetries for 3D hydraulic models with a high-resolution digital surface model (DSM) from aerial images. UAV photogrammetry at a flight altitude of 80 m and 20 m has been carried out with an additional ground survey using RTK-GPS. Ground checkpoints and cross sections surveyed at 100 m sections were utilized in assessing the vertical accuracy of the models. The results are point cloud data with average ground sampling distances between 0.5 and 2 cm. After developing the bathymetry, numerical simulations are performed using the software package FLOW-3D by Flow Science Inc. As for the most sections of the measure no three-dimensional investigations are required, the hybrid model contains one large 2Dshallow water mesh and smaller fully 3D meshes to get adequate calculation durations and accurate results in each section. The cell sizes are varying between dxyz = 0.0625-0.125 m from detail observations to sections without significant obstacles or vertical velocity ratios. The free surface is calculated via the Volume of Fluid (VOF) method and in the first step, turbulence modelling is included by the renormalization-group (RNG) turbulent model. This paper presents a comparison of RTK-GPS cross sections and UAV photogrammetry-based bathymetry. Further, the practical application of the hybrid model is shown through the first results of detailed 3D flow calculations in the region of implemented large woody debris (LWD).
Year: 2019