The Entropy Analysis of Groundwater Level Time Series in Ardabil Plain

The hydrological and geomorphological changes in catchments is one of the most important challenges today. Analysis of hydrological time series such as groundwater level plays an important role in the behavior identification of them against various factors. In this study, the effect of wavelet based de-noising on the entropy of groundwater level time series in Ardabil plain has been investigated. Also, the effective sub-series of the groundwater level time series process were identified using three criteria: entropy, mutual information (MI) and linear correlation coefficient. The results showed that the entropy of the groundwater level time series increased using wavelet based de-noising method. The increase of entropy indicates an increase natural fluctuations in the groundwater level time series and thus indicates the occurrence of a favorable trend in it. Also, the results showed that MI and entropy criteria, due to nonlinear nature, can accurately demonstrate the dominant sub-series in the groundwater level process. A+D3 combination was considered as the dominant sub-series in most piezometers.