Clustering Time Series of Sea Levels: Extreme Value Approach Articles uri icon

publication date

  • July 2010

start page

  • 215

end page

  • 225

issue

  • 4

volume

  • 136

International Standard Serial Number (ISSN)

  • 0733-950X

Electronic International Standard Serial Number (EISSN)

  • 1943-5460

abstract

  • In this paper, long (>40 years) hourly tide gauge records from the North Atlantic are analyzed. A new time series clustering approach which combines Bayesian methodology, extreme
    value theory, and classification techniques is adopted for the analysis
    of the regional variability of sea-level extremes. The tide gauge
    records are clustered on the basis of their corresponding predictive
    distributions for 25-, 50-, and 100-year return values. The results of
    the cluster analysis show a clear distinction between the higher
    latitude stations for which the return values are largest and the
    remaining locations. This distinction reflects in the U.S. east coast
    the transition between the Scottian shelf and Gulf of Maine area and the
    mid-Atlantic Bight area. For the stations at lower latitudes the
    results show a grouping based on return levels that is not a function of
    geographical proximity but reflects local effects in extreme sea levels
    associated with the specific location of each tide gauge.