| In this study the statistical characteristics of the annual NBS (Net Bassin Supply) which are located in different regions of North America have been studied. These characteristics series are considered as chronological series. We have used three techniques for the detection of the sudden shifting patterns in these NBS series. The first technique proposed by Hubert is based on segmentation procedure of hydrometeorological series, the second is based on the Bayesian method of Lee-Heghinian. Both of these two techniques show difficulties in interpreting and reproducing the analyzed behavior of these NBS series. We have proposed in this work, a technique which we have named Modified Haar wavelet approach (OHM), this technique gives a better interpretation and analysis for the detection of the sudden shifting patterns in the NBS series. Simulation results based on the use of the three methods are given and compared. |
