Abstract
This work characterizes the main spatiotemporal variability present in flow series and relates these variabilities with climate patterns that could affect this variable. The Principal Component Analysis was applied over Brazilian streamflow time series, when each principal component is associated with one variability mode. The temporal variability of first six modes (that explain more than 80% of total variability) was compared with climate indexes performing a simple correlation and analyzing the power spectral density of modes and indexes. The results show similarities between variabilities patterns and climate indexes, the correlations are significant and both have spectral peaks in same frequencies. The main variability modes of river flow showed interannual fluctuations associated with El Niño Southern Oscillation phenomenon and significate correlations with sea surface temperature in the Tropical Atlantic, demonstrating the importance of remote influences associated which sea surface temperature anomalies in the tropical ocean. Climate patterns associated with multidecadal and decadal variability of the Atlantic and Pacific Oceans, respectively, have persistent relationships with the rivers flow modes.
Keywords:
streamflow variability; climate patterns; principal component analysis
