The Timing, Magnitude, and Relative Composition of Extreme Total Water Levels Vary Seasonally Along the US Atlantic Coast

The annual maximum (AM) method, which subsamples time series to retain the maximum event per year, and the peak-over-threshold (POT) method, which extracts values exceeding a threshold to define extremes, have long histories in determining flood event frequency. In practice, extreme value distributions applied to AM and POT events often assume that the data comes from the same statistical population. Locations across the world, like the United States (U.S.) Atlantic coastline, however, experience high coastal water levels driven by various individual processes and storms with different driving mechanisms during different seasons. This research investigates when extreme total water levels (TWLs) occur during the year along the U.S. Atlantic coast and whether individual components, like waves, tides, and storm surge, contributing to TWLs vary across regions and during the year. From 1980 to 2020, extreme TWLs occurred during the extratropical and tropical seasons, with the relative proportion of extreme TWLs occurring during the extratropical season increasing northward. Still water levels drive spatial variability in extreme TWL magnitude, while wave climate drives differences in extreme TWL magnitude between extratropical and tropical seasons. Month-to-month variability in the composition of extreme TWLs varies more than spatial variability, highlighting the importance of understanding the components driving extremes at different times of the year. Variations across storm seasons in the processes contributing to extreme TWLs may have implications for how large-scale changes to the climate impact hazards along open sandy coastlines and influence the robustness of extrapolating rare events from models fit to a single population. To predict extreme sea level impacts, rare events (e.g., 100-year event) are often determined by evaluating the maximum event of the year or high values above a threshold. These techniques, however, often assume that the same storm type causes all extreme events. This might not be true in regions affected by extratropical (e.g., winter storms) and tropical (e.g., hurricanes) storms, potentially resulting in the incorrect prediction of extreme coastal sea levels. To address this, we analyze when extreme coastal sea levels happen during the year on open-coast sandy beaches along the U.S. Atlantic coast, and whether the proportion waves, tides, and storm surge contribute to driving extreme coastal sea levels varies by storm seasons. We find that extreme coastal sea levels occur year-round, and seasonal variations are more noticeable in the northeast because the influence of winter storms is stronger moving northwards. Tides influence regional variability in extreme coastal sea level magnitude, while waves control magnitude changes related to storm seasons. Moreover, extreme coastal sea level drivers vary more between storm seasons than across regions. Our research demonstrates that considering seasonal variations is essential to evaluate the impacts of weather and ocean hazards where winter storms and hurricanes occur frequently. The relative composition of extreme total water levels has more month-to-month variation than spatial variation Still water levels drive spatial variability of extreme total water levels, while wave climate drives variability across storm seasons Extreme total water levels occurring in different storm seasons can impact the robustness of rare event extrapolation methods