Shape Classification of AltiKa 40-Hz Waveforms using Linear Discriminant Analysis and Bayes Decision Rule in the Gujarat Coastal Region
Shape classification of the 40-Hz waveforms obtained by the recently launched AltiKa satellite has been attempted in the paper. Since retracking algorithms suitable for altimeter return echoes based on Brown model are not applicable for the echoes from coastal ocean, specific algorithms are to be devised for such echoes. In the coastal ocean, waveforms display a wide variety of shapes due to varying coastline geometry, and topography. Hence, a proper classification strategy is required for classifying the waveforms into various categories so that suitable retracker could be applied to each category for retrieving the oceanic parameters. The algorithm consists of three steps: feature selection, linear discriminant analysis, and Bayesian classifier. The classification algorithm has been applied to the waveforms in the close proximity of Gujarat coast. Independent validation has been done near the eastern coast of India. Confusion matrices obtained for both the coasts are quite encouraging. Individual examples of classification have been provided for the purpose of illustration.
No Reference information available - sign in for access.
No Citation information available - sign in for access.
No Supplementary Data.
No Article Media
Document Type: Research Article
Affiliations: Space Applications Centre, Ahmedabad, India
Publication date: September 10, 2015