EFFECT OF WAVELET-BASED DE-NOISING METHOD ON CHANGE POINT DETECTION PROCESS IN HYDROLOGICAL TIME....

The influence of the de-noising method on the process of detecting change points in hydrological time series was investigated. The annual average of precipitation, stream flow, temperature, and moisture time series of Vanyar station in Iran were used in this article, while the annual average of precipitation, stream flow, temperature, and moisture time series of Tifton station in the United States of America were analysed. Data was divided into two types: first, the natural – noisy – form was analysed, then the data was edited and reformed – de-noised – form was studied once more. To create de-noised time-series data, a wavelet-based de-noising method was applied. Since both stations have had a known change in the last 30 years, there has been no substantial effect of this change in the phenomena time series, and the differences have been wiped out during the comparison using traditional change point detection methods. Before and after de-noising, the change point detection method shows a substantial difference. Using wavelet-based de-noising on hydrological time series, change point detection is greatly enhanced. The results show that the identified points of de-noised data are more dependable than the detected points of noisy data. De-noising is most effective on the Wilcoxon sum rank test, Pettit test, two-phase regression test, and standard normal homogeneity test, according to a comparison of applicable approaches.

Please see the link :- https://www.ikprress.org/index.php/JOGEE/article/view/900