透過您的圖書館登入
IP:3.146.221.204
  • 期刊

半拉格朗日法與正定義數值方法之比較

The Comparision of Semi-Lagrangian Method to Positive Definite Schemes

摘要


本文以高斯波形、餘弦波及幣形圓柱三種不同形體的平流計算,來探討比較半拉格朗日法(semi-Lagrangian scheme)、及Smolarkiewicz法與Hsu-Arakawa法兩種正定義法。研究結果顯示,半拉格朗日法積分時距大小的選擇,不受限於穩定度,只受限於精確度及軌跡線上的風切大小。若半拉格朗日法使用的時距愈大,相位誤差愈大,而時距過小則因內插次數過多,精確度也會降低。因此定出最適當的時距以達最佳精確度,是使用半拉格朗日法重要的事情之一。經前述三種不同性質之形體的平流探討,半拉格朗日法振幅的維持,大致上比Smolarkiewicz法與Hsu-Arakawa法之正定義方法,及FD4都要好。不過對幣形圓柱之平流,半拉格朗日法有過量(overshoot)及負值(undershoot)的現象,大小可達10%以上,然而FD4完全不能掌握幣形圓柱的平流,而兩種正定義方法皆有合理表現。

並列摘要


Using Guassian cone, cosine cone and slotted cyclinder three different passive bodies, we compared semi-Largrangian method, the Smolarkiewicz method and the Hsu-Arakawa method. Research results shown the △t choice of semi-Largrangian method are limited by accuracy and the order of wind shear on trajectory, not limited by stability. The larger △t of semi-Largrangian method the more phase errors, but the smaller △t need the more interpolation times it reduces accuracy. So for the accuracy, use semi-Largrangian method it need choice suitably △t. The maintenance of amplitudes to semi-Largrangian method is better than the Smolarkiewicz method, the Hsu-Arakawa method and FD4. But to the advection of slotted cyclinder, the overshoot and undershoot to semi-Largrangian method are more than 10%.

被引用紀錄


陳政友(2016)。颱風邊界層似震結構對於颱風眼牆演變的影響探討〔碩士論文,國立臺灣大學〕。華藝線上圖書館。https://doi.org/10.6342/NTU201602609

延伸閱讀