Abstract
This research evaluates the complexity and chaotic behavior of streamflow data by introducing two novel complexity measures derived from Fourier spectral analysis and contrasting them with Kolmogorov complexity (KC) and its derivatives. The two proposed measures are the Participation of Non-seasonal Frequencies Measure (PNSF) and the Standard Triperiod Measure (STP). These measures are applied to monthly streamflow data from rivers in the Mendoza basin, Argentina, and the continental United States. The findings reveal that PNSF provides valuable insights into streamflow predictability by addressing yearly fluctuations and seasonalities, while STP effectively quantifies the periodic behavior and emphasizes the influence of seasonal cycles, making it a potentially universal parameter for hydrological classification of rivers. Additionally, a significant correlation exists between the proposed measures and established complexity metrics, indicating that these new measures enhance existing frameworks and deepen our understanding of streamflow predictability in complex systems. Future research will focus on further validating these measures and exploring their applicability across various hydrological contexts.

This publication has 23 references indexed in Scilit: