Establishing coastal water level magnitude-duration-frequency curves for infrastructure design

• We develop the coastal water level-duration-frequency curves for extreme events • The WDF curves show significant sensitivity to the event sampling method • The WDF curves inform coastal communities about the effect of duration on intensity Extreme coastal floods are expected to be more frequent and/or intense in a warming climate. To protect both lives and assets, it is essential to continuously analyze the associated hazards and update engineering designs accordingly. While many studies have looked at how often and how high extreme sea levels get under various forcing conditions, much less attention has been paid to how long these extremes would last, which is just as important. The duration of elevated water levels can impact coastal areas in multiple ways, such as overwhelming drainage systems, bringing larger water volumes during prolonged events, and gradually weakening flood infrastructure over time. In this study, we address this overlooked aspect of extreme sea level events by introducing a framework for developing water level Magnitude-Duration-Frequency (MDF) curves using water level data from 217 coastal gauges distributed globally. Our analysis shows that the average water level difference between 1-hour and 24-hour durations is 0.8 meters for similar return periods, with a maximum difference of up to 3.63 meters for a 100-yr event. Additionally, we evaluate the impact of different event selection methods on the resulting MDF curves. Depending on the sampling methodology, differences in the developed MDF curves can reach up to half a meter across most studied regions. These curves can inform engineers and policymakers in developing more resilient flood mitigation strategies, refining risk assessments, and enhancing preparedness for future coastal hazards.