Using global tide gauge data to validate and improve the representation of extreme sea levels in flood impact studies

• Published in: Global and planetary change Vol. 156; pp. 34 - 45
• Main Authors: Hunter, J.R., Woodworth, P.L., Wahl, T., Nicholls, R.J.
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.09.2017
• Subjects: DINAS-COAST data set; Extreme sea level parameters; GESLA-2 tide gauge data set; Major coastal cities flood risk; DINAS-COAST data set; Major coastal cities flood risk; GESLA-2 tide gauge data set; Extreme sea level parameters
• ISSN: 0921-8181; 1872-6364
• DOI: 10.1016/j.gloplacha.2017.06.007

The largest collection of tide gauge records assembled to date, called GESLA-2, has been used to provide reliable extreme sea level parameters at 655 locations around the world. This has enabled a rigorous assessment of the European Union-funded DINAS-COAST (D-C) data set of extreme sea level information for the global coastline that has been used in many published flood impact studies. We find the D-C extreme levels to be generally both too high, compared to those from GESLA-2, and too flat, when plotted as a function of return period. This leads to an over-estimation of the probability of extreme sea levels in the present day for most locations around the world, and also to an over-estimation of the probability of extreme sea levels in the future as sea level rises. A detailed impact study is conducted for the world's largest coastal cities following the approach of Hallegatte et al. (2013), resulting in similar conclusions for these particular locations. We suggest that most previous studies that have relied upon D-C information should be re-assessed in the light of these findings, using more recent modelling-based estimates of extreme sea level information. •The largest tide gauge data set (GESLA-2) was used to provide reliable extreme level parameters around the world.•These parameters have been compared to the DINAS-COAST (D-C) data set of global extreme sea level information.•D-C extreme levels are found to be generally both too high, and too flat, when plotted as a function of return period.•A detailed study of coastal cities confirms impacts to have been previously over-estimated using D-C.•Previous studies based on D-C should be re-assessed, using more recent modelling-based data sets of extremes levels.