R&D highlights edition 2019

PROJECTS Flood risk 14 A sound understanding of run- up is important to ensure the integrity of a dike or coastal defence structure because the run-up height determines the crest height of a dike or the stability of a coastal structure. Run-up is therefore measured frequently in the physical facilities at Deltares. Run-up measurements were also required as part of the KPP project “Study for the enhancement of flood risk management”. This study looked at the effect of water depth on wave overtopping and run-up. The hypothesis is that wave overtopping is lower in shallow water conditions because most of the large waves already break before reaching the dike. Until now, run-up has usually been measured visually or based on several wave gauges located on the slope. However, this approach is very time- consuming (in the case of visual inspection) or complex (when multiple wave gauges are used). A measurement technique using video recording was therefore adopted. Artificial intelligence was used to extract the necessary information from the video. To replace the traditional measurement techniques, a deep neural network was trained to identify the frame pixels which contain water. This approach makes it possible to measure run-up using a standard camera. The method has already been applied successfully in the WOODY project in the Delta Flume and it was also used on this project in the Scheldt flume. These results gave a unique insight into run-up behaviour. The video technique made it possible to investigate the effect of water depth on run-up distribution. This dataset makes it possible to relate water level exceedance curves to the run-up exceedance curves for different hydrodynamic conditions, which helps to understand the overtopping rates. The new technique was used not only to determine run-up height but also to look at the variability of the run-up over the flume width. These results suggest that there are wall effects in the wave flume. In other words, run-up near the walls is different than in themiddle of the flume. A second advantage of the video technique is that it allows for the computation of the front speed. Since some of the failure mechanisms are related to the front speed rather than the overtopping discharge, it is also important to measure these velocities accurately. Contact: Menno de Ridder, Menno.deRidder@deltares.nl, t +31 (0)6 5178 9665 Paul van Steeg, Paul.vanSteeg@deltares.nl , t +31 (0)88 335 8376 Alex Capel, Alex.Capel@deltares.nl , t +31 (0)88 335 8034 WAVE RUN-UP MEASUREMENTS USING ARTIFICIAL INTELLIGENCE Artificial intelligence was used to extract run-up observations from video recordings of an experiment in the Scheldt flume. The results from this newmeasuring technique gave a unique insight into run-up behaviour. The measured run-up signal (red) with the detection of the individual waves (orange peaks). The red line represents a moving average for the run-up signal. Run-up measurements in the Delta Flume Snapshot of the Scheldt flume results with the run-up measurements

RkJQdWJsaXNoZXIy Mjc4NjU=