Regime-related regularities in river flow revealed by Aksentijevic-Gibson complexity

Abstract A streamflow time series contains a large quantity of information and finding the instruments that are capable of accessing this information remains an important task. Using the Kolmogorov complexity (KC) and its derivatives, it has previously been shown that the degree of randomness in streamflow partly depends on its orographic characteristics. This paper applied a change-based complexity measure (Aksentijevic-Gibson complexity – AG) to investigate the spatiotemporal regularities of monthly streamflow of seven rivers from Bosnia and Herzegovina for a period of 1965-1986, and evaluate their hydrological effects. A unique advantage of this complexity measure is its ability to quantify large-scale complexity and to zoom in on low-level structure and detect patterns that are normally inaccessible. Ten time series created from the seven rivers were examined and temporal and structural complexity profiles of different streamflows and annual periodicities associated with bimodal/mixed flow regimes were obtained. The AG complexity can be employed as a standard instrument for the analysis of hydrological data and it holds the promise of uncovering patterns and rhythms in the data that cannot be captured by KC and other complexity measures.