Quantifying the effect of woody vegetation on the wave loads on a dike using remote sensing

Quantifying the effect of woody vegetation on the wave loads on a dike using remote sensing: Large scale physical model tests

Çete, Ceylan (TU Delft Civil Engineering and Geosciences)

Aarninkhof, S.G.J. (mentor)
Hofland, B. (graduation committee)
Oosterlo, P. (graduation committee)
van Steegh, Paul (graduation committee)

Delft University of Technology

Civil Engineering | Hydraulic Engineering

WOODY

2019-05-06

It is already acknowledged that Nature-based Solutions can be used to attenuate waves, however it is still uncertain to what extent the vegetation can contribute to decreasing the flood risk. So far mainly small-scale tests have been performed to quantify wave attenuating properties of vegetation. To quantify the effect of more extreme wave conditions (high-water levels and wave heights), full scale tests are required. In the Delta flume of Deltares large-scale physical model tests with a willow forest of 40 meters in front of a dike are conducted. Three different measuring methods: visual measurements, terrestrial laser scanning and video imaging, are used to quantify the wave run-up and wave-overtopping on the dike. This is the first time a ‘Machine Learning algorithm’ is used to obtain the wave run-up heights on a dike from video footages. It is also new that the wave overtopping volumes are determined by a laser scanner without using a wave overtopping tank, which was initially used to collect and measure the real overtopping volume.

Results by the laser scanner show an overestimation of the overtopping discharges at high water level for higher crest freeboards, making these results less reliable. However, more research and a thorough validation are required to confirm the accuracy of this method.
From results obtained by flume experiments can be concluded that remote sensing: laser scanner and video imaging, are accurate methods to measure the run-up on a dike. Thus, the camera in combination with Machine Learning is an accurate, simple and low-cost technique, to measure the wave run-up on a dike. Results from the camera not only give the run-up height, but also give new insights in the variations of the wave run-up over the dike.

It can also be concluded that a willow forest of 40 meters causes a significant reduction in the wave run-up and overtopping, for both willows in summer and winter state. Further research is needed, so these significant reductions can be implemented in the design and assessment of dikes.

The most commonly used models for designing a dike are the TAW (2002) and EurOtop. Comparing the obtained results, show that the determined wave run-up from the experiments are underestimated by the TAW (2002). The TAW is an empirical formula, based on a large data set of mainly small-scale experiments. Therefore, the difference between the test results and the TAW could likely be attributed to scale effects. However, more measurements at full scale are needed to confirm this.

Vegetation
Storm conditions
Dike height
Run-up
Overtopping
Remote sensing
Camera
Machine Learning
Laser scanner
Run-up reduction
Large Scale
Flume experiments
TAW (2002)
Delta Flume
Willows
Nature based flood defence
Woods versus Waves

http://resolver.tudelft.nl/uuid:88e2b629-efc9-48c8-b438-b250951200de

2020-04-29

Student theses

master thesis

© 2019 Ceylan Çete