Study on near shore typhoons wave height forecasting by the self organizing algorithm and data supplementing scheme investigation / 自組式網絡架構近岸海域颱風波高預測模式及數據補遺方式探討

碩士 / 國立成功大學 / 水利及海洋工程學系 / 103 / A forecasting model of near shore typhoon wave height developed as a alternative is proposed in this paper by the self-organizing GMDH (Group Method of Data Handling) algorithm with 4 parameters of wave height(H), wind speed(V), distance(L) and azimuth (θ) between the target location and typhoon center. Data of these 4 parameters/ variables observed at Hsin Chu and Long Dong data buoys as well as obtained from CWB could be used to calibrated, verified and construct the prior 1~ 6hrs near shore typhoon wave height forecasting model with 15 typhoon events data during 2006 to 2009 to provide the necessity of typhoon warning facility. A recursive GMDH model could be reorganized by using the update data to match the time variant properties in forecasting steps to improve the forecasting accuracy. Data supplementation model could be set up by the OTE(Ocean Tide Efficiency) concept and EMD(Empirical Mode Decomposition) technique and tested by the measured and acquired ADCP data of Taoyuan Yon An/Hsinchu Long Fung and Hsin Chu data buoy for lost its total 1/3,1/4 and 1/6 of monsoon wave height data. Typhoon wave height lost data supplementation is tested also with Hsin Chu and Long Dong data buoy’s data of 6 typhoon events for missed its total 1/3,1/2 and 2/3 of typhoon wave data by the same method. Prior 1~6hr typhoon wave height forecasting results show that the average RMSE, CC and error scale ratio are in between 20.0~47.8cm, 79.4~93.6% and 11.5~23.4%. Monsoon wave height missed data supplement results reveal that the average RMSE, CC and error scale ratio are in between 10.5~20.2cm, 78.1~ 87.7%, and 15.6~20.1% respectively, and typhoon wave height lost data supplement results reveal that the average RMSE, CC and error scale ratio are 53.4cm, 70.2%, and 23.6% by the EMD technique.

Identiferoai:union.ndltd.org:TW/103NCKU5083101
Date January 2015
CreatorsShih-HungChuang, 莊士宏
ContributorsTai-Wen Hsu, Pei-Hwa Yen, 許泰文, 顏沛華
Source SetsNational Digital Library of Theses and Dissertations in Taiwan
Languagezh-TW
Detected LanguageEnglish
Type學位論文 ; thesis
Format224

Page generated in 0.0016 seconds