A global dataset of storm surges and waves for coastal hazard mapping from bias corrected unstructured coupled hindcast with cyclone events

The assessment of coastal vulnerability to hazards associated with extreme sea levels is strongly influenced by the combined effects of storm surges and waves. In addition, irregularities in available observations limit the reliable estimation of extreme return levels. To address this, we present a globally consistent dataset of wave heights and storm surges derived from a hindcast spanning 1950–2023, based on the high-resolution, unstructured and coupled global coastal model of Mentaschi et al. (2023), along with those from tropical cyclone events.

We apply a quantile mapping framework to debias the model with a focus on upper-tail values. We use along-track global L3 wave heights from satellite measurements, by the Copernicus Marine Service, to bias-correct the model-based significant wave heights. The corrected wave heights are independently validated using in-situ wave observations from ISPRA and Copernicus Marine Service buoy networks. For storm surges, in-situ coastal sea level observations from the GESLA3 database are employed. Historical tropical cyclone tracks and associated coastal water levels are simulated using the Deltares D-Flow Flexible Mesh (D-Flow FM) numerical model, whereas the wave heights during cyclonic events are obtained from satellite altimetry observations. The performance of the dataset is evaluated using state-of-the-art metrics tailored for the accuracy of extreme values. The results demonstrate substantial improvements in the representation of extremes. For example, extreme wave heights in the Italian Mediterranean region exhibit average (median) normalized biases below −30% in the original hindcast, which are reduced to within 0 and −10% after bias correction. Similarly, for storm surges, biases in the upper tail (above the 99.9th percentile) are reduced from −11.28% (−6.5%) to 0.38% (−0.55%) across selected global locations. In equatorial regions, where ERA5 wind forcing exhibits known deficiencies, extreme surge underestimation exceeding −40% is reduced to within −10%.

The dataset provides a robust foundation for determining the intensity of global coastal multi-hazards because its improved suitability for performing extreme value analysis and can be used to study the joint extremes arising from storm surges and waves.  

References

Mentaschi, L., Vousdoukas, M. I., García-Sánchez, G., Fernández-Montblanc, T., Roland, A., Voukouvalas, E., Federico, I., Abdolali, A., Zhang, Y. J., and Feyen, L.: A global unstructured, coupled, high-resolution hindcast of waves and storm surge, Front. Mar. Sci., 10, 1233679, 2023 https://doi.org/10.3389/fmars.2023.1233679     

Campos, R.M.; Gramcianinov, C.B.; de Camargo, R.; da Silva Dias, P.L. Assessment and Calibration of ERA5 SevereWinds in the Atlantic Ocean Using Satellite Data. Remote Sens. 2022, 14, 4918. https://doi.org/10.3390/rs14194918

Campos-Caba, R., Alessandri, J., Camus, P., Mazzino, A., Ferrari, F., Federico, I., Vousdoukas, M., Tondello, M., and Mentaschi, L.: Assessing storm surge model performance: what error indicators can measure the model's skill?, Ocean Sci., 20, 1513–1526, 2024,https://doi.org/10.5194/os-20-1513-2024.    

Tamizi, A., Young, I.R. A dataset of global tropical cyclone wind and surface wave measurements from buoy and satellite platforms. Sci Data 11, 106 (2024). https://doi.org/10.1038/s41597-024-02955-4 

Bahmanpour, M. H., Tilloy, A., Vousdoukas, M., Federico, I., Coppini, G., Feyen, L., and Mentaschi, L.: Transformed-Stationary EVA 2.0: A Generalized Framework for Non-Stationary Joint Extremes Analysis, EGUsphere [preprint], 2025.