以颱風為依據之風場最大風速探討

Abstract

由於全球氣候暖化，永續發展的意識推動，各國莫不減少石化能源的消耗，大力提倡綠能發電，而台灣也跟隨世界腳步大量建造離岸風力發電機。風力發電機的發電效率受到風速影響，但其結構能否抵抗強風則受最大風速影響，而台灣夏季時的最大風速則常受到颱風影響。針對颱風可能造成特定風場裡的最大風速，國內似乎缺乏較深入的研究，有待吾人探討。本研究收集一特定陸上氣象測站近14年來每年最大值風速和33年來颱風警報期內所量測到的逐時風速資料；針對前者，引進統計理論，求取該風場特定再現期可能最大風速強度，作為預測的比較值；針對後者，則引進Gumbel極值統計理論，依颱風路徑、颱風強度、Kmeans法與暴風半徑內等不同分類方式，預測該測站在未來1、5、10、25、50年可能遇到的最大風速，並以機率分佈呈現。本研究後半段延續前述研究結果，透過簡單的線性迴歸式，將陸上測站風速轉換為附近離岸測站風速，並依據離岸測站短期內量到的風速資料，預測離岸長期會遭遇到的最大風速。數值分析結果顯示，颱風造成的最大風速高於颱風以外整體考量的最大風速；在路徑分類中，橫越台灣的穿心颱帶來的最大風速最劇；在強度分類中，強烈颱風在所有分類中會有最大的長期風速，可作為考量最大風速的上限值；在暴風半徑內分類中，颱風中心區域的風速確實大於其餘周遭風速；在K-means分類中，成功辨識最大風速分群特徵，準確預測長期風速。因離岸風場缺乏長期觀測風速資料，本研究成功預測出離岸風場最大風速，可提供風場內結構與機械設備壽命期間內的風速設計參考。

Due to global warming and the promotion of sustainability awareness, countries around the world are reducing the consumption of fossil fuels and strongly advocating for green energy generation. Taiwan has followed suit by constructing a large number of offshore wind turbines. The power generation efficiency of wind turbines is influenced by wind speed, while their structural resilience against strong winds is affected by maximum wind speed, which, during Taiwan's summer, is often impacted by typhoons. However, there appears to be a lack of in-depth domestic research on the maximum wind speeds caused by typhoons in specific wind fields, warranting further investigation. This study collects data from a specific onshore meteorological station, including annual maximum wind speeds over the past 14 years and hourly wind speed measurements recorded during typhoon warning periods over the past 33 years. For the former, statistical theories are introduced to estimate the maximum wind speed intensity for specific return periods in the wind field as a reference for comparison. For the latter, the Gumbel extreme value statistical theory is employed to predict the maximum wind speed that the station may encounter in 1, 5, 10, 25, and 50 years, based on different classification methods such as typhoon paths, typhoon intensities, K-means clustering, and within storm radii, with results presented as probability distributions. In the second half of this study, building on the aforementioned findings, a simple linear regression model is applied to convert onshore station wind speeds into nearby offshore station wind speeds. Based on short-term wind speed data from offshore stations, long-term maximum wind speeds likely to be encountered offshore are predicted. Numerical analysis results reveal that maximum wind speeds caused by typhoons are higher than those considered under non-typhoon conditions. Among path classifications, typhoons crossing directly over Taiwan produce the strongest maximum wind speeds. For intensity classifications, severe typhoons yield the highest long-term wind speeds among all categories, providing an upper limit for maximum wind speed considerations. In storm radius classifications, wind speeds in the typhoon center region are indeed higher than those in surrounding areas. In K-means classifications, the largest cluster was successfully identified, enabling accurate predictions of long-term wind speeds. Due to the lack of long-term observational wind speed data for offshore wind fields, this study successfully predicts maximum wind speeds for offshore wind fields, providing a reference for wind speed design during the structural and mechanical equipment lifespans within the wind field.