Analyzing Long-term Persistence of Precipitation and Runoff of Rhiss and Nekkor basins: A Case Study from Morocco

Characterizing the behaviour of precipitation and river flow time series in the temporal domain was always a way to understand the complexity of the dynamical processes in hydrological systems. In order to characterize the intrinsic characteristic of the hydrological processes of the Rhiss and Nekkor basins, a power spectrum analysis was deduced from instrumental time series of precipitation and flow. On the background of flow and precipitation fluctuations, the power spectrum shows discrete components in the form of 1/f-noise. A power-law with 1/f β noise factor was found in precipitation and flow indicating a stochastic behavior with long-range persistence. The spectral exponent β for precipitation shows values ranging between 0.16 and 0.43. On the other, hand river flow shows a power-law dependence on frequency S(f) with an exponent of ½. The existence of long-range persistence was also evaluated by the analysis of Hurst exponent, where we have found a strong long-term persistence for precipitation (H>0.79) and flow (H>0.73) with slow decay. We noted that the river flow of a power-law with β=0.5, yields a power-law rescaled range with Hurst exponent close to H=0.73, indicating a universal power-law dependence on frequency. The existence of pink-noise energy signature in precipitation and river flow indicates that the climatologic and hydrologic systems are characterized by more low frequency components and operating in large scales and maybe related to large-scale climatic phenomena. Our results suggested that the two hydrologic systems play an important role in the generation of new characteristics of precipitation signal.