Separating Storm surge from Tidal Level Based on Wavelet Packet Transform

摘要

W artykule przedstawiono algorytm realizacji transformaty Falkowej, analizuj?cy dane o astronomicznych parametrach ptywów oraz podniesieniu si? poziomu morza w czasie sztormu (nagon). Opracowana metoda pomaga w efektywnym odró?nieniu tych dwóch zjawisk, co ogranicza ryzyko fa?szywego alarmu. Na podstawie rzeczywistych danych, dokonano jej porównania z metodami liniowymi, co wykaza?o sprawniejsze dzia?anie.%A method of wavelet packet transformation (WPT) is presented in this paper, which regards the tidal level as a signal with containing different frequency of astronomical tidal components and storm surge components. The tidal level signal can be decomposed serials of frequency bands signals by using WPT to carry out time-frequency analysis for procuring the representative frequency bands of storm surge. Then the storm surge process gained by reconstructing the exclusive decomposed wavelet packet coefficients. Taken the typhoon NO.9608 measured tidal level of Xiamen tide station as an example, the storm surge calculates by using the presented method, which compared the value calculated by transformation linear method with the measured value. The result shows that the suggested method can preferably separate storm surge and provide a new way to separate storm surge from short-term tidal level material.