İstanbul Aydın Üniversitesi Dergisi (May 2013)
WAVELET TRANSFORMS IN TIME SERIES ANALYSIS
Abstract
The existing methods based on statistical techniques for long range forecasts of Indian summer monsoon rainfall have shown reasonably accurate performance, for past 10 years. Because of the limitation of such statistical techniques, new techniques may have to be tried to obtain better results. Time series analysis and forecasting using wavelet transforms has become a powerful tool to determine, within a time series, both the dominant modes of variability and how those modes vary in time. A preliminary investigation, using wavelet analysis, on temperature and rainfall time series is presented.
