The use of historical information for regional frequency analysis of extreme skew surge

Abstract

Abstract. The design of effective coastal protections requires an adequate estimation of the annual occurrence probability of rare events associated with a return period up to 10 3 years. Regional frequency analysis (RFA) has been proven to be an applicable way to estimate extreme events by sorting regional data into large and spatially distributed datasets. Nowadays, historical data are available to provide new insight on past event estimation. The utilisation of historical information would increase the precision and the reliability of regional extreme's quantile estimation. However, historical data are from significant extreme events that are not recorded by tide gauge. They usually look like isolated data and they are different from continuous data from systematic measurements of tide gauges. This makes the definition of the duration of our observations period complicated. However, the duration of the observation period is crucial for the frequency estimation of extreme occurrences. For this reason, we introduced here the concept of "credible duration". The proposed RFA method (hereinafter referenced as FAB, from the name of the authors) allows the use of historical data together with systematic data, which is a result of the use of the credible duration concept.

A Regional Frequency Analysis of Tide Gauges to Assess Pacific Coast Flood Risk

“…In both studies (Weiss and Bernardara, 2013;Wahl et al, 2017), nuances in model choice mostly affect the tail of the distribution (e.g., 100year LESLs). Our usage of the TG exceedance threshold directly as the local flood index is the equivalent to the location parameter, which Weiss and Bernardara (2013) find as an optimal index in some circumstances and follows other studies (Roth et al, 2012;Frau et al, 2018).…”

“…Positive shape parameters generally occur where extreme outliers (e.g., tropical storm surges) tend to occur. and coastal storm surge (Bardet et al, 2011;Bernardara et al, 2011;Weiss and Bernardara, 2013;Frau et al, 2018) studies.…”

“…With the TG regions established (Figure 2), the aggregated and normalized sets of TG threshold exceedances are fit with a GPD (Coles, 2001) using the penalized maximum likelihood method (Coles and Dixon, 1999;Frau et al, 2018) to estimate (median) regional ESL (RESL) probabilities and the 5th and 95th% levels (90% confidence interval) defined as:…”

A Stochastic Extreme Sea Level Model for the German Baltic Sea Coast

“…The parametric distribution fitted to z Peak is adjusted to maximize the likelihood function, which is simply the product of likelihood functions calculated for each complete set of observations, here that is the high‐resolution tide‐gauge data and the historical records. While this is not necessarily the most robust method of accounting for historical events (see Frau et al, ), it is useful as a demonstration. Using this adjusted distribution, three artificial ESLs produced by the model with peak water levels approximately equal to the 1872 event are shown in Figure .…”

La base de données TEMPETES : un support pour une expertise collégiale et interdisciplinaire des informations historiques de tempêtes et de submersions

“…L'ensemble de ces études montre que le dépouillement de données d'archives ainsi que la numérisation et l'interprétation de données d'archives et marégraphiques est un travail fastidieux qui prend beaucoup de temps. Toutefois, les quelques valeurs extrêmes historiques reconstruites peuvent être intégrées dans des études statistiques des niveaux marins et sont précieuses pour améliorer les estimations des niveaux extrêmes (BULTEAU et al, 2015, HAMDI et al, 2015, FRAU et al, 2018. Outre-Manche, un projet de base de données de tempêtes et surcotes historique existe déjà sous le nom de "surgewatch" (HAIGH et al, 2017 Thème 7 -Risques côtiers Les données intégrées dans la BD TEMPETES sont entièrement publiques, la BD TEMPETES est soumise à la licence open source (https://opendatacommons.org/licenses/odbl/1.0/).…”