Temporal Analysis of the Flows of the Rivers that form the Hydrographic Basin of Moquegua (Peru)

Abstract

The study of water sources such as rivers and lakes is increasingly important in the modern world, mainly due to their scarcity and the large number of cities flooded by heavy rains. In this paper, we aim to analyze the dynamics of the daily flow of water in the Torata, Tumilaca, and Huaracane rivers, which form the Moquegua watershed (Peru). To study these fluctuations, we apply the DFA method (estimate long-range autocorrelation in non-stationary time series) and the DCCA cross correlation coefficient (which strongly dominates the standard Pearson coefficient in time series analysis) to analyze the flow in the basin of Moquegua. Among the rivers that we analyzed descriptively, we identified greater variability in flows in the Huaracane River and less in the Torata, and in general a seasonal component of approximately one year. With the modeling via DFA, it was possible to identify persistent behavior and also annual seasonality. With the DCCA coefficient (cross correlation), as a function of the time scale, positive values of cross correlation are found between river flows at different levels (weak, medium, strong). With this study, considering the properties of the applied methods, it is possible to predict the oscillations of one river in relation to another, which can help the authorities in the management of freshwater over time.