A specific type of wave is identified as an infragravity wave (IG wave). These waves with very long wave periods, even up to minutes, are possibly dangerous for flood events, increased storm surge and/or failure of coastal defences. Current guidelines for dike safety are often determined for gravity wave dominated systems, so identifying IG dominated systems can be a tool to determine whether dike safety guidelines are sufficient or not. To analyse the IG waves, the non-hydrostatic simulation model SWASH is proposed. An extensive model validation was carried out, comparing the results of a lab experiment with the model output to assess the model’s accuracy to simulate the IG waves. Next, four parameters were chosen to be varied (offshore wave height and period, beach slope and vegetation type) to create several scenarios, mimicking real-life situations in the Wadden Sea and the Chesapeake Bay. To express the IG wave dominance, the Infragravity coefficient (IGC) is used, which is the low-frequency over the high-frequency wave height ratio. IG dominance is defined when the IGC>1.0, and a very IG dominant system is defined when IGC>2.0. For this study, the IGC is analysed at a dike toe. The effect of a slope on IG dominance shows that weak slopes (1:500) show very IG dominance overall, no matter the offshore conditions. As the slope gets steeper, the gravity waves start to dominate, starting at the lower wave heights. The wave heights show almost never IG dominance for low values (1m) and show more IG dominance as the offshore wave height increases. The offshore period has a minimal influence on IG dominance for short periods, but the IGC seems to decrease after an offshore period of 10s. This effect is mostly visible on steep slopes, and barely visible for weak slopes. When vegetation is added to the weak-sloped scenario (1:500), it is concluded that vegetation dissipates both the low -and high-frequency wave heights in case of low offshore wave heights (1 and 2m) and short periods (6 and 8s). For higher and longer waves, vegetation dissipates high-frequency gravity waves better than low-frequency IG waves. This shows an increase in IG dominance with the inclusion of vegetation. For very high and long waves (10m, 14s) the vegetation’s effect is minimal. These results from this study show that for a large number of scenarios, mainly for weak slopes, IG waves dominate over gravity waves.

The United States of America have a safety standard for flood protection of 1/100 year. However the flood protections in the Washington D.C. do not comply with this requirement. During this study the levees in Washington D.C. area were analysed and it was found, that in order to comply with the 1/100 safety standard the levees in the NationalMall needs to be heighten by 0.5mand those near the Anacostia river by 1.5m. The safety standard of 1/100 imposed in the United States of America is not based on exact calculation, therefore another design method was applied, which uses an optimal safety level based on the damage cost of floods and investment cost of flood protections. The calculation of the optimal return period was based on the ’Standaardmethode 2017’ and it was found that the optimal return period at the NationalMall levee is 1/263 years. At the southern bank of the Anacostia river this optimal return period is 1/373 year. With these return periods new levees were designed in order to ensure the safety of the area of Washington D.C. There were two new levees needed around the National mall. The first was situated north of the Lincolnmemory along the Potomac river and the second replaces the temporary flood defence by the 2nd street SW and had an L-shape facing the Potomac river. The Anacostia levee was stretched on the East side in order to comply with the new 1/373 year safety standard. The old and new National mall levees needed a height of 3.5 m and those by the Anacostia needed a height of 8.5 m. It is important to take cautionwith these new levees as some houses needed to be removed. An attempt was made at modelling the above mentioned floods in ADCIRC to improve the accuracy of the design water levels and to verify the final design. However, the model showed instabilities, preventing the results from being used. Explanations for these instabilities are the incoming water not being properly ramped up, the large vertical gradients in shallow areas and the upstream and downstream boundaries being too close to each other.