Changes in the seasonal sea level cycle (SSLC) can modulate the flooding risk along coastlines. Here, we use harmonic analysis to quantify changes in the amplitude and phase of the annual component of the sea level cycle at 663 tide gauge locations along the global coastline where long records are available. We identify coastal hotspots by applying clustering methods revealing coherent regions with similar patterns of variability in the annual sea level cycle (ASLC). Results show that for most tide gauges the annual amplitude reached its maximum after 1970 and its peak typically occurs during the fall season of the respective hemisphere. Many tide gauges exhibit non-stationarity in the annual cycle in terms of amplitude and/or phase. For example, at 125 tide gauges we find significant trends in the amplitude (either increasing or decreasing) while several sites (36 in total), mostly in the Mediterranean and around Pacific islands, experienced phase changes leading to shifts in the timing of the peak of the annual cycle by more than a month. Our results highlight the importance of accounting for potential non-stationarity in seasonal mean sea level (MSL) cycles along coastlines.

Storm surge events are a key driver of widespread flooding, particularly when combined with astronomical tides superimposed on mean sea level (MSL). Coastal storms exhibit seasonal variability which translates into a seasonal cycle in storm surge activity. It has been demonstrated before that the seasonal cycle shows significant interannual variations including possible long-term trends. Understanding these changes is critical as both changes in the amplitude and the phase of their seasonal cycle may alter the compound flood potential. Changes in the seasonal storm surge cycle and the compounding with the MSL cycle remain largely unknown. A comprehensive analysis of the storm surge seasonal cycle and its links to the MSL seasonal cycle is performed using tide gauge observations from a quasi-global dataset. Harmonic analysis is used to assess the mean and changing storm surge seasonal cycles over time. Extreme value analysis is applied to explore the effect of seasonal changes on storm surge return levels. We also quantify the influence of large-scale climate modes, and we compare how the seasonalities of storm surge and MSL have changed relative to each other. The peak of the storm surge cycle typically occurs during winter for tide gauges outside of tropical cyclone regions, where there is greater variability in the phase of the storm surge cycle. The timing of the peak varied by more than a month at 21% of the tide gauges analyzed. The MSL and storm surge cycles peaked at least once within 30 days at 74% of tide gauges.