Underestimated risks of future storm surge extremes: A global study

Around the world, extreme sea levels are increasingly affecting low-lying and unprotected coasts. The rise in local mean sea level is widely assumed to be the dominant driver of changes in the frequency of coastal extremes, and most projections of future flood risk rely on the premise that extremes increase uniformly with mean sea level. Here, we show that this assumption does not always hold. Using non-stationary extreme-value models applied to long, high-quality tide-gauge records from the global GESLA-3 database, we find that roughly 15\% of sites exhibit statistically significant departures from this behaviour, with extreme sea levels rising faster or slower than local mean sea level. Regions such as the North Sea show surge amplification beyond what sea-level rise alone can explain, consistent with depth-sensitive wave dynamics in shallow basins. These findings highlight the need to incorporate non-stationary effects in coastal hazard assessments to support robust adaptation planning.