A Database for Global Storm Surges (1836 – present)

Abstract:

Storm surge is an abnormal rise in coastal water levels due to low atmospheric pressure and strong winds, which could be induced by tropical or extratropical cyclones. The greatest destruction from tropical cyclones is storm surge driven coastal flooding and half of the fatalities owed to Atlantic tropical cyclones are caused by storm surge. Hence, understanding the trends and variability of the storm surge climate in the past is crucial in order to adopt sound impact assessments and adaptation strategies. However, due to the absence of adequate observational data (especially in the tropics and southern hemisphere), it is challenging to carry out robust extreme value analysis, which is an important input for impact and adaptation assessments.

We use several atmospheric reanalysis datasets (ERA-20C, ERA-Interim, ERA5, MERRA-2, and 20CR) to model global storm surges with statistical and machine learning methods in order to extend the available storm surge records at locations where some data exists. The effects of temporal and spatial resolution of the atmospheric datasets on storm surge model accuracy are investigated by comparing reconstructed and observed storm surges. Centennial reanalysis products allow us to reconstruct storm surges back as early as 1836. Given the known uncertainties with atmospheric analyses in the 19^th and early 20^th century (due to inhomogeneous data assimilation over time), we conduct a thorough validation using long tide gauge records, many of which were recently extended by digitizing historic data, to assess how far back we can trust the reconstruction in different locations. Finally, a global storm surge database corresponding to each reanalysis dataset will be made available for open access.