Reproduce “Multi-site modelling and reconstruction of past extreme skew surges along the French Atlantic coast” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Nathan Huet, Philippe Naveau, Anne Sabourin
Researchers developed statistical methods for reconstructing historical extreme skew surge time series at data-sparse French Atlantic coast stations, utilizing a multivariate peaks-over-threshold framework with two extreme value theory approaches: a nonparametric machine learning method (ROXANE) and a parametric multivariate generalized Pareto model (MGPRED). The study also introduced a novel, objective method for selecting extreme thresholds, contributing to improved long-range return period estimations for coastal risk management.
Key insights
- Explicitly modeling spatial dependence among extreme events allows for robust reconstruction of historical extreme time series at data-sparse locations by leveraging data from nearby long-record stations.
- Nonparametric machine learning approaches (like ROXANE) and parametric models (like MGPRED) offer complementary strengths in multivariate extreme value analysis, with ROXANE often performing better for the most extreme values and MGPRED providing confidence intervals.
- Defining the extreme threshold as the largest zero of the second derivative of the fitted EGP density provides an objective and robust method for threshold selection in POT analysis.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.