Data Driven Methods to Analyze Wave BUOY Observations

Recent data driven methods such as genetic programming (GP), model tree (MT), support vector regression (SVR), and locally weighted projection regression (LWPR) look attractive in carrying out a variety of tasks in coastal and ocean engineering. This paper provides an exposure to their applications made by authors to hindcast significant wave heights, to estimate wave surface spectral density and to inversely evaluate wind from waves. The field measurement used for this purpose mainly belonged to two selected locations in Arabian Sea along the west coast of India. The studies indicated that the various data driven methods can be effectively used to evaluate causal, temporal as well as inverse relationships between various wave and wind parameters and also for resolution of average wave parameters into a set of spectral densities. It is expected that the studies reported in this paper would pave the way for solutions to more complex and difficult tasks in coastal and ocean engineering such as estimation of uncertain design parameters and spatio-temporal evaluation of ocean environment.