Assessment of spatiotemporal variability of precipitation using entropy indexes: a case study of Beijing, China

• Published in: Stochastic environmental research and risk assessment Vol. 36; no. 4; pp. 939 - 953
• Main Authors: Du, Longgang, Li, Xinxin, Yang, Moyuan, Sivakumar, Bellie, Zhu, Yanxin, Pan, Xingyao, Li, Zhijia, Sang, Yan-Fang
• Format: Journal Article
• Language: English
• Published: Berlin/Heidelberg Springer Berlin Heidelberg 01-04-2022; Springer Nature B.V
• Subjects: Aquatic Pollution; Chemistry and Earth Sciences; Climate; Climate change; Climate variability; Complexity; Computational Intelligence; Computer Science; Earth and Environmental Science; Earth Sciences; Entropy; Entropy (Information theory); Environment; Hydrologic data; Math. Appl. in Environmental Science; Mountain regions; Mountains; Original Paper; Permutations; Physics; Precipitation; Probability Theory and Stochastic Processes; Reliability analysis; Stability analysis; Statistics for Engineering; Suburban areas; Urban areas; Urbanization; Variability; Waste Water Technology; Water Management; Water Pollution Control; Weather stations; Beijing China; China; Precipitation; Entropy; Spatial pattern; Climate variability; Complexity
• ISSN: 1436-3240; 1436-3259
• DOI: 10.1007/s00477-021-02116-8

Entropy indexes are widely used to measure the complexity of precipitation and its spatiotemporal variability, due to their advantages in quantifying the complex and nonlinear interactions of climate variables. However, different entropy indexes may provide different outcomes on the complexity of precipitation, thus causing difficulties in the assessment of actual complexity. In this study, we identified and chose the optimal entropy index from six entropy indexes based on the reliability and stability of entropy values, to assess the complexity of precipitation. The six entropy indexes considered are: information entropy (IE), fuzzy entropy (FE), approximate entropy (AE), sample entropy (SE), wavelet entropy (WE), and permutation entropy (PE). We implemented the approach on the monthly precipitation data observed over the period 1968–2017 at 58 meteorological stations in Beijing, China. The results indicated that the SE index was the optimal index with the best reliability and stability and, thus, was used to examine the complexity of the 58 monthly precipitation series in Beijing. The results also suggested that the complexity of precipitation can be viewed in terms of four sub-regions in Beijing. The precipitation in the ‘Northern Mountain Area’, with high complexity, was more affected by the interactions between the terrain and climate variability, compared with that in the ‘Southern Mountain Area’ with low complexity. The precipitation in the ‘Central Urban Area’ was more significantly impacted by the extensive urbanization than that in the ‘Northeast Suburban Area’, causing higher complexity in the former. At temporal scale, the complexity of precipitation in the ‘Southern Mountain Area’ exhibited the largest increasing rate due to the highest impacts of climate variability, followed by the ‘Central Urban Area’ with a moderate increasing rate of complexity of precipitation due to the impacts of human activities. However, the complexity of precipitation in the ‘Northern Mountain Area’ and ‘Northeast Suburban Area’ exhibited small changes due to relatively weak sensitivity to climate variability and human activities.