Pooling Seasonal Forecast Ensembles to Estimate Storm Tide Return Periods in Extra‐Tropical Regions

Abstract Low‐pressure systems and strong winds, when coinciding with high tides can generate severe storm tides, leading to coastal flooding and significant economic losses. Accurate estimates of storm tide frequency and intensity are crucial for flood hazard assessments and risk reduction. However, the limited observational records pose a challenge in estimating high return periods with low uncertainty. In this study, we evaluate the potential of pooling ensembles from the SEAS5 seasonal forecast archive to generate an extensive storm tide data set for robust return period estimates in extra‐tropical regions at large spatial scale. Using SEAS5 to force the hydrodynamic model GTSM, we generate 525 synthetic years of storm tides and apply extreme value analysis to estimate 40‐year and 500‐year return periods. Our findings demonstrate that SEAS5 produces unbiased and independent synthetic mean sea level pressure events across major extra‐tropical regions, including Europe, China, Russia, South America and Australia. In Europe, unbiased SEAS5‐derived storm tide extremes along the Atlantic coast are particularly well‐suited for return period analysis. The results show the benefits of using longer records to improve extreme return periods. SEAS5 not only reduces uncertainties in high return period estimates but also provides more extreme events, enhancing the reliability of extreme value distributions compared to short observational records.