Regresar

Comprehending Dynamics of the Ecuadorian River Discharge Series Using Wavelet Analysis and Bandpass Filters

Abstract:

In this paper, two alternatives for a useful description of streamflow time patterns in Ecuadorian lowlands are proposed. In that region, streamflow rates are influenced by precipitation, which varies seasonally due to annual climate patterns, spatially due to small-scale weather patterns, and temporally over longer periods due to longer-period climate fluctuations. Using wavelet coefficients, different periodicities were identified throughout the series, with a dominant one-year frequency (annual cycle). Nevertheless, other frequencies are noteworthy. For instance, during the period 2005-2012, the frequency of 2 year suggested a larger difference between the highest peaks during the rainy season (usually from December to April), and the lowest valleys of the dry period (rest of the year), otherwise difficult to perceive using only the raw data. In addition, the filter Baxter-King has been successfully applied for seasonal separation, including extreme years (El Niño events of 1982-1983, and 1997-1998). Hence, the use of wavelets may open opportunities for further research, including gap analysis, and cross-wavelet analysis between river stations belonging to different subbasins, but sharing similar features.