Evaluation of Neuro Fuzzy and Numerical Wave Prediction Models in Lake Ontario
M. H. Kazeminezhad,A. Etemad-Shahidi,S. J. Mousavi
DOI: https://doi.org/10.2112/jcr-si50-062.1
2024-03-29
Journal of Coastal Research
Abstract:Kazeminezhad, M. H., Etemad-Shahidi, A. and MousavI, S. J., 2007. Evaluation of neuro fuzzy and numerical wave prediction models in Lake Ontario. Journal of Coastal Research, SI 50 (Proceedings of the 9th International Coastal Symposium), 317 321. Gold Coast, Australia, ISSN 0749.0208 For planning or designing maritime structures, wave observation with a long-term history is generally necessary. However, we may often encounter the problems that only short-term observed data were available or the records are incomplete in a station. In this study, the performances of a new generation spectral wind-wave model based on unstructured meshes and a neuro fuzzy model for predicting wave parameters are investigated. The data set used in this study comprises of wave data and over water wind data gathered from deep water location in Lake Ontario. The data set was divided into two groups. The first one that comprises of 26 days wind and wave measurement was used as training and checking data to develop neuro fuzzy models and also to calibrate the numerical model. The second one that comprises of 14 days wind and wave measurement was used to verify the neuro fuzzy and numerical models. Wind field was specified as constant in space and variable in time. Results show that whitecapping factor is the most important calibration factor in this case and satisfactory results can be obtained by tuning this factor. Results also indicate that the errors of proposed neuro fuzzy models in predicting wave parameters are more than those of on the numerical model.
geosciences, multidisciplinary,environmental sciences,geography, physical