Intermittent chaos in river flow dynamics

River flow dynamics are highly complicated, as a result of multiscale interactions between climatic inputs of tremendous variability and heterogeneities in landscape properties. While efforts made in the last two decades have led to many important understandings of river flow dynamics, such as non-Gaussian and heavy-tailed distributions of discharges, fractal and multifractal scalings of streamflow data, increase of dimensionality (or complexity) of the flow dynamics with the scale of aggregation, and and chaos-like behaviors, A definitive answer to the nature of river flow dynamics is still lacking. Among the outstanding issues are (1) What is the nature of the fractal scaling in river flow data? (2) Can the debate on whether river flow dynamics are chaotic or not be unambiguously resolved?

We show that river flow dynamics can be modeled as noisy intermittent chaos. The intermittency suggests an ON/OFF modeling of river flow dynamics, where ON and OFF periods are associated with high and low river discharges, respectively. The intermittent chaos is consistent with the long-range correlation and fractal scaling behaviors of streamflow data, sheds light on why there is an increase of dimensionality (or complexity) in the flow dynamics with the scale of aggregation, and explains why chaos cannot be found from river flow data with classic methods for detecting chaos.