The two-step framework for over-threshold modelling of environmental extremes proposed in Bernardara et al. (Nat Hazards Earth Syst Sci 14:635–647, 2014) for univariate analyses is generalized to an event-based framework applicable to multivariate analyses. The distinction between sequential values (temporal observations at a given time step) and the event-describing values (such as storm peaks in univariate Peaks-Over-Threshold extrapolations) is further detailed and justified. The classification of multivariate analyses introduced in Mazas and Hamm (Coast Eng 122:44–59, 2017) is refined and linked to the meaning of the concepts of event, sampling and return period that is thoroughly examined, their entanglement being highlighted. In particular, sampling is shown to be equivalent to event definition, identification and description. Event and return period definition are also discussed with respect to the source phenomena or to response (or structure) phenomena. The extreme event approach is thus proposed as a comprehensive framework for univariate and multivariate analyses for assessing natural hazards, seemingly applicable to any field of environmental studies.

Extreme events: a framework for assessing natural hazards

https://hal-univ-pau.archives-ouvertes.fr/hal-03200009/document

Coastal flooding event definition based on damages: Case study of Biarritz Grande Plage on the French Basque coast

The estimation of sea levels corresponding to high return periods is crucial for coastal planning and for the design of coastal defenses. This paper deals with the use of historical observations, that is, events that occurred before the beginning of the systematic tide gauge recordings, to improve the estimation of design sea levels. Most of the recent publications dealing with statistical analyses applied to sea levels suggest that astronomical high tide levels and skew surges should be analyzed and modeled separately. Historical samples generally consist of observed record sea levels. Some extreme historical skew surges can easily remain unnoticed if they occur at low or moderate astronomical high tides and do not generate extreme sea levels. The exhaustiveness of historical skew surge series, which is an essential criterion for an unbiased statistical inference, can therefore not be guaranteed. This study proposes a model combining, in a single Bayesian inference procedure, information of two different natures for the calibration of the statistical distribution of skew surges: measured skew surges for the systematic period and extreme sea levels for the historical period. A data‐based comparison of the proposed model with previously published approaches is presented based on a large number of Monte Carlo simulations. The proposed model is applied to four locations on the French Atlantic and Channel coasts. Results indicate that the proposed model is more reliable and accurate than previously proposed methods that aim at the integration of historical records in coastal sea level or surge statistical analyses.

Extreme Sea Level Estimation Combining Systematic Observed Skew Surges and Historical Record Sea Levels

Coastal planning implies the estimation of extreme sea levels. As the distribution of astronomical high tides can be predicted, most recent publications suggest focusing on the estimation of extreme skew surges. Historical information, record sea levels observed before the beginning of systematic tide gauge recordings, can improve estimations. The corresponding skew surges can be estimated but are not necessarily exhaustive. Indeed, some historical extreme skew surges can remain unnoticed if they are combined with low or moderate tides, or for a variety of reasons. To deal with this exhaustiveness issue, a previous publication proposed an unbiased method for combining systematic period skew surges with historical period extreme sea levels. This method appeared more reliable than previously proposed approaches. The present study aims at presenting a broader evaluation of this method, based on its application to nine sites located on the English Channel and North Sea coasts. The method is also improved to consider several historical periods and various types of historical information. Results confirm the method to be reliable, useful, and relevant. A number of recommendations is also formulated for the selection and use of historical information for sea level frequency analyses. 

https://onlinelibrary.wiley.com/doi/pdfdirect/10.1111/jfr3.12862

Extreme skew surge estimation combining systematic skew surges and historical record sea levels on the English Channel and North Sea coasts

Au cours de la derniere decennie, la tempete Xynthia et d'autres ont marque l'actualite de par leur intensite et les degâts causes lors de leur passage sur le territoire francais. Ces evenements peuvent donner l'impression qu'ils n'ont ete "jamais vu par le passe", mais le recensement des tempetes plus anciennes (1953, 1987, 1999…) montre que de tels phenomenes surviennent en fait regulierement sur le littoral metropolitain. Des lors, la collecte et l'analyse d'informations historiques apparaissent incontournables pour ameliorer la prevention des risques littoraux, en particulier liee a la submersion marine. Afin de mutualiser les informations disponibles en France, une base de donnees relationnelle et spatiale a ete developpee au sein de l'IRSN. A ce stade, plus de 650 evenements de tempete et/ou de submersion sur le littoral francais et des pays voisins ont pu etre recenses, sur une periode s'etendant de 1500 a nos jours. En parallele a l'elaboration de cette base de donnees, un groupe de travail interdisciplinaire s'est constitue. Il integre des ingenieurs, chercheurs, statisticiens et historiens appartenant a differents organismes. Un axe majeur de travail est l'integration de nouvelles donnees historiques issues de diverses archives dans la base de donnees, ainsi que leur interpretation et la quantification des niveaux marins associes.

/pdf/la-base-de-donnees-tempetes-un-support-pour-une-expertise-2bb8qrg9zq.pdf

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

Yann Ferret

Papers

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