Search

Analysis of daily streamflow variability in space and time is important for water resources planning, development, and management. The natural variability of streamflow is being complicated by anthropogenic influences and climate change, which may introduce additional complexity into streamflow records. To address the complexity in streamflow, daily discharge data recorded during the period 1989–2016 at twelve gauging stations on Brazos River in Texas (USA) were used to derive a set of novel quantitative tools: Kolmogorov complexity (KC) and its derivative-associated measures to assess complexity, and Lyapunov time (LT) to assess predictability. It was found that all daily discharge series exhibited long memory with an increasing down-flow tendency, while the randomness of the series at individual sites could not be definitively concluded. All Kolmogorov complexity measures had relatively small values with the exception of the USGS (United States Geological Survey) 08088610 station at Graford, Texas, which exhibited the highest values of the complexity measures. This finding may be attributed to the elevated effect of human activities at Graford, and proportionally lesser effect at other stations. In addition, complexity tended to decrease downflow, meaning that larger catchments were generally less influenced by anthropogenic activities. The correction on randomness of Lyapunov time (quantifying predictability) was found to be inversely proportional to the Kolmogorov complexity, which strengthened our conclusion regarding the effect of anthropogenic activities, considering that KC and LT were distinct measures, based on rather different techniques. • Daily streamflow complexity is analyzed using Kolmogorov complexity measures. • Highest complexity discrete spectrum indicates the highest streamflow randomness. • Lyapunov exponent and time of the streamflow series predictability are also studied. • The corrected Lyapunov time is inversely proportional to the Kolmogorov complexity. [ABSTRACT FROM AUTHOR]

The purpose of this paper was to choose an appropriate information dissimilarity measure for hierarchical clustering of daily streamflow discharge data, from twelve gauging stations on the Brazos River in Texas (USA), for the period 1989⁻2016. For that purpose, we selected and compared the average-linkage clustering hierarchical algorithm based on the compression-based dissimilarity measure (NCD), permutation distribution dissimilarity measure (PDDM), and Kolmogorov distance (KD). The algorithm was also compared with K-means clustering based on Kolmogorov complexity (KC), the highest value of Kolmogorov complexity spectrum (KCM), and the largest Lyapunov exponent (LLE). Using a dissimilarity matrix based on NCD, PDDM, and KD for daily streamflow, the agglomerative average-linkage hierarchical algorithm was applied. The key findings of this study are that: (i) The KD clustering algorithm is the most suitable among others; (ii) ANOVA analysis shows that there exist highly significant differences between mean values of four clusters, confirming that the choice of the number of clusters was suitably done; and (iii) from the clustering we found that the predictability of streamflow data of the Brazos River given by the Lyapunov time (LT), corrected for randomness by Kolmogorov time (KT) in days, lies in the interval from two to five days.

The purpose of this paper was to choose an appropriate information dissimilarity measure for hierarchical clustering of daily streamflow discharge data, from twelve gauging stations on the Brazos River in Texas (USA), for the period 1989-2016. For that purpose, we selected and compared the average-linkage clustering hierarchical algorithm based on the compression-based dissimilarity measure (NCD), permutation distribution dissimilarity measure (PDDM), and Kolmogorov distance (KD). The algorithm was also compared with K-means clustering based on Kolmogorov complexity (KC), the highest value of Kolmogorov complexity spectrum (KCM), and the largest Lyapunov exponent (LLE). Using a dissimilarity matrix based on NCD, PDDM, and KD for daily streamflow, the agglomerative average-linkage hierarchical algorithm was applied. The key findings of this study are that: (i) The KD clustering algorithm is the most suitable among others; (ii) ANOVA analysis shows that there exist highly significant differences between mean values of four clusters, confirming that the choice of the number of clusters was suitably done; and (iii) from the clustering we found that the predictability of streamflow data of the Brazos River given by the Lyapunov time (LT), corrected for randomness by Kolmogorov time (KT) in days, lies in the interval from two to five days.