Wavelet Transform Based Method for River Stream Flow Time Series Frequency Analysis and Assessment in Tropical Environment
| Autor: | Kai Lun Chong, Ahmed El-Shafie, Sai Hin Lai |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Discrete wavelet transform
business.industry Wavelet transform Pattern recognition Time–frequency analysis symbols.namesake Fourier transform Wavelet Fourier analysis symbols Time domain Artificial intelligence business Continuous wavelet transform Water Science and Technology Civil and Structural Engineering Mathematics |
| Zdroj: | Water Resources Management. 33:2015-2032 |
| ISSN: | 1573-1650 0920-4741 |
| Popis: | The main aim of this study is to perform a time series frequency analysis and assessment for stream flow over the Johor River using a comparative method between wavelet transform (WT) and Fourier transform (FT). One of the wavelet analyses used was the discrete wavelet transform (DWT), which revealed the periodic wavelet components responsible for the trend detection. Using the FT, the periodicities that governed the trend can be obtained; however, in terms of the time domain analysis, FT seems to be lacking compared to the WT. The conditions for using DWT are discussed, and the selection decisions for such discretization are considered. Besides, using the global wavelet spectrum (GWS) and the continuous wavelet transform (CWT), the dominant periodicity components can be further well described in time frequency characteristic. In addition, the integration of the WT and Mann–Kendall (MK) test allows the determination of possible trends present in the stream flow dataset series. It is shown that the wavelet analysis is more suitable than the Fourier analysis as it exhibits good extraction of the time and frequency characteristics, especially for a nonstationary data series. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|
Full text from SpringerLink