臺灣主要港口附近海域混合波浪統計特性及推算模式建置研究(1/4)

Abstract

港灣工程規劃設計所需之設計波浪須由長期實測波浪資料來決定。近年來為解決現有記錄的時間不足的狀況，交通部運輸研究所港灣技術研究中心已發展類神經颱風波浪推算模式，但此技術的建立還是基於往昔的實測資料，在環境有所改變的情況下這些波浪推算模式的可靠性仍需再評估。為提升國內波浪推算技術的正確性本研究除建立蘇澳港類神經網路颱風波浪推算模式，並針對花蓮港與蘇澳港以M5模式樹發展新的波浪推算模式，再對類神經與M5模式的適用性進行評估。 本研究以MIKE 21 SW二維風浪數值模式模擬各場颱風對各港口造成的波場以供設計波浪的推算。另為響應國際航海學會(PIANC)對氣候變遷造成海事工程之影響議題，本研究針對交通部運輸研究所港灣技術研究中心在花蓮港與蘇澳港長期監測的波浪資料利用Seasonal Kendall趨勢檢測法進行分析，此法能同時考慮季節性的變化與資料缺漏的狀況，估算出合理的資料趨勢以評估颱風及波浪受氣候變遷影響的顯著性，並針對設計波高的決定標準提出相關對策與建議。

Design waves for marine structures are commonly determined by extreme analysis on long-term annual maximum wave data. However insufficient data are available for extreme analysis. The problem of insufficient data samples was solved by supplementation of simulated waves which were calculated by ANN wave simulation model developed by the Harbor and Marine Technology Center of Taiwan (IHMT). The accuracy of these wave data samples for extreme analysis should be promoted considering alternative approach and the effect on wave climate by recent global climate change. This project aims to not only establish the Ann wave simulation model for the Suao harbor, but also to develop a new M5 wave model. Comparison on the simulated result by ANN and M5 models with observed data will show the prediction accuracy and applicability of both wave models. A numerical wave model of MIKE 21 SW will be used to calculate waves at both Hualien and Suao harbors for chosen typhoons. The best predictions on wave among ANN, M5 and numerical wave models will be determined by comparing the simulated waves with observed data. Design waves for both Hualien and Suao harbors will be determined from most accurate wave samples. Responding to the issue of the impact of climate change on marine works proposed by World Association for Waterborne Transport Infrastructure (PIANC), Seasonal Kendall method of detecting a trend is used to determine the trend of long-term observed typhoon’s waves of IHMT. Applicable countermeasures for design wave will be suggested by this research.