PARTITIONING SEQUENCES AND ARRAYS OF DIRECTIONAL OCEAN WAVE SPECTRA INTO COMPONENT WAVE SYSTEMS

A method for characterizing wave systems from temporal sequences and spatial arrays of directional ocean surface gravity-wave spectral estimates is presented. The method is structured in two major stages. In the first data-intensive stage, individual spectra are parameterized in terms of the directions, periods, and significant wave heights of wave systems represented in that spectrum. In the second stage, wave systems from spectra are associated with similar wave systems in neighboring spectra, and ''large-scale wave systems'' representing activity occurring continuously over groups of contiguous spectra are determined. This high-level description of the ocean surface wave field represents a considerable data reduction that preserves directional information in a more meaningful way than either average wave directions computed over the entire spectrum or wind sea-swell decompositions. Each large-scale wave system is normally caused by a distinct meteorological event. Data from the Labrador Sea Extreme Waves Experiment (LEWEX) are used to illustrate applications to numerical wave model and buoy spectra, and these methods should be useful in other spectral data comparison experiments. Other potential applications include the interpretation and archival of global wave-model analyses and the compression of wave spectra in a manner sufficient to preserve information necessary for numerical wave model data-assimilation schemes.