Characterizing Long Island Sound outflows from HF radar using self-organizing maps

Jenq Chi Mau, Dong Ping Wang, David S. Ullman, Daniel L. Codiga

Research output: Contribution to journalArticlepeer-review

14 Scopus citations

Abstract

The surface outflows from the Long Island Sound are examined from one-year records of HF radar (CODAR) observations. Synoptic flow patterns are identified using manual classification, empirical orthogonal function (EOF) decomposition, and self-organizing maps (SOM). Four characteristic flow patterns for the spring/summer and fall/winter seasons each are obtained through a 2 × 2 SOM array. The SOM is confirmed by comparison with manual classification, and is shown to be a significant improvement over EOF classification. It is suggested that the degrees of freedom of the leading EOF modes can be used as a constraint on the otherwise arbitrary SOM dimension. The relationship between the flow patterns and the winds also can be conveniently examined in SOM. The outflows are shown to interact strongly with the ambient coastal currents, both of which are under the influence of the winds. This result challenges the conventional wisdom which often treats the outflows independent of the ambient currents. The advantage of using SOM in synthesizing and interpreting synoptic HF radar observations is clearly demonstrated.

Original languageEnglish
Pages (from-to)155-165
Number of pages11
JournalEstuarine, Coastal and Shelf Science
Volume74
Issue number1-2
DOIs
StatePublished - Aug 2007

Keywords

  • HF radar
  • Long Island Sound
  • SOM

Fingerprint

Dive into the research topics of 'Characterizing Long Island Sound outflows from HF radar using self-organizing maps'. Together they form a unique fingerprint.

Cite this