Storm surge time series de-clustering using correlation analysis

• DOI: 10.1016/j.wace.2024.100701
• Publication Date: 2024-06-01T18:35:14Z
• Authors: Ariadna Martín; Thomas Wahl; Alejandra R. Enri...; Robert Jane
• Keywords: Storm Surge

The extraction of individual events from continuous time series is a common challenge in many extreme value studies. In the field environmental science, various methods and algorithms for event identification (de-clustering) have been applied past. distinctive features events, such as their temporal evolutions, durations, inter-arrival times, vary significantly one location to another making it difficult identify independent series. this study, we propose new automated approach detect series, by identifying standard duration across locations using correlations. To account inherent variability at given site, incorporate deviation through soft-margin approach. We apply method 1,485 tide gauge records global coast gain insights into typical durations storm surges along different coastline stretches. results highlight effects both local characteristics seasonality on derived durations. Additionally, compare obtained with other commonly used de-clustering techniques showing that these are more sensitive chosen threshold.