Spectral Wave Data Assimilation in SWAN Wave Model

• Main Authors: Yang-Ming Fan, 范揚洺
• Other Authors: Beng-Chun Lee
• Format: Others
• Language: en_US
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/86077357513156325939

博士 === 國立成功大學 === 水利及海洋工程學系碩博士班 === 96 === The purpose of this study is to enhance the accuracy of numerical wave forecast with data assimilation. Traditional forecast is consisting in: uses numerical wave model to do hindcast, and then adjusting the parameters by measured data. Use numerical wave model to forecast wave after finding the best parameters. A technique – spectral data assimilation combine model results with measured data. Using measured data to correct model results immediately and to do wave forecast. Review past research, we find that the extreme parameters were produced at grid point in the distance of measured station by using data of wave height to do data assimilation. In order to make this research look forward to do operating numerical wave model, we adopt sequential data assimilation to do data assimilation. In the process of data assimilation, the optimal interpolation of partition from Voorrips, etc. is adopted and improved. And the analytic data is from satellites and data buoy. Expect to set up the technique of spectral data assimilation. And then discuss the result which using spectral data assimilation to typhoon numerical wave model in order to verify the accuracy of spectral data assimilation model and analyze the wave character Taiwan water. The spectral data is near real time, on the contrary, the satellite even near real time, but there is no advantageous in time domain. Therefore, this study discuss the difference between spectral data assimilation and satellite data assimilation.