Detrended fluctuation analysis and entropy-complexity causality analysis of temperatures in an urbanized mountain stream

We present the results of applying detrended fluctuation analysis (DFA) to study correlations in mountain stream temperatures over a 9-year period for impacted and unimpacted conditions. Permutation entropy and complexity are also computed for stream temperature time series at various locations characterized by varying degrees of human impact, and a graphical representation of the “complexity-entropy causality plane” is used to examine the degree of disturbance caused by human activities. Urban development such as canopy removal and urban infrastructure can lead to a higher degree of anti-persistent correlations indicated by lower values of DFA exponents. This effect is more pronounced downstream due to cumulative urbanization effects with downstream distance. Seasonal variations also influence stream temperature dynamics: less anti-persistent behavior is observed in winter months, and stronger anti-persistent correlations during summer. The position in the complexity-entropy plane, reflecting distribution of short term temperature patterns, is also affected, where most sites experience a shift towards higher entropy values after restoration, and only for one site entropy is reduced. Our results indicate the important role urbanization plays on stream evolution greatly influenced by human activities. This information may provide valuable insights for watershed management and water resources management.

Duke Scholars

Author Chuanhui Gu DKU Faculty