Application of time-frequency representation in the study of the cyclical behavior of wind speed in Algeria: wavelet transform

摘要

This study proposes a process of assessment of wind potential using descriptive statistics and time-frequency analysis. Via the time-frequency analysis, the variation of the spectral content of wind speed with respect to time is analyzed by the determination of the frequencies, the durations and the occurrence-time of significant fluctuations. The study is carried out for six sites in Algeria during a time interval of 1 year and over a frequency range from 1.1 × 10~(-2) to 2.4 cycles/day (i.e., periods from 10 h to about 90 days). The study has allowed to deduce that the windiest regions are located in the southern part of the country with annual mean from 5 to 6 m/s. The time-frequency analysis reveals the presence of three types of oscillations. In the warm season, diurnal oscillations have been distinguished especially in the costal and mountainous regions. However, in the cold season, synoptic oscillations of periods around 9-13 days have been observed. The Southern regions are characterized by long oscillations of periods around 33-50 days.